Isolated Lung Perfusion with

Gemcitabine for the Treatment of

Pulmonary Metastases

Experimental Study in a Rat Model

Bart P. van Putte

Promotores: prof. dr. A. Brutel de la Rivière1

prof. dr. P.E.Y. van Schil2

Co-promotor: dr. J.M.H. Hendriks2

1 Department of Cardiothoracic Surgery, University Medical CentreUtrecht, Utrecht, The Netherlands2 Department of Thoracic and Vascular Surgery, University HospitalAntwerp, Antwerp, Belgium

Van Putte, Bart Pieter

Isolated Lung Perfusion with gemcitabine for the treatment ofpulmonary metastases. Experimental study in a rat model.ISBN: 90-393-3479-X

Cover: Josephine Walta, Bart van PutteFigure on the back: pulmonary micrometastatic disease.

The research described in this thesis was performed at the Laboratoryof Experimental Surgery of the University of Antwerp under thesupervision of prof. dr. PEY Van Schil.

Printed by Eli Lilly and Company

© by the author, Culemborg, 2003All rights reserved. No part of this book may be reproduced, stored ina retrieval system, or transmitted, in any form or by any means,without prior permission of the holder of the copyright.

Isolated lung perfusion with gemcitabine for the

treatment of pulmonary metastases

Experimental study in a rat model

Geïsoleerde longperfusie met gemcitabine voor de

behandeling van longmetastasen

Experimentele studie in een rattenmodel

(with summary in Dutch)

Proefschrift ter verkrijging van de graad van doctor aan de

Universiteit Utrecht op gezag van de Rector Magnificus, prof. dr. WH

Gispen, ingevolge het besluit van het College voor Promoties in het

openbaar te verdedigen op donderdag 18 september 2003 des ochtends

om 10.30 uur

door

Bart P. van Putte

Geboren op 23 januari 1977 in Maasland

Promotiecommissie:

prof. dr. I.H.M. Borel Rinkesprof. dr. J-W.J. Lammersprof. dr. A.F.A.M. Schobbenprof. dr. E.E. Voest

Financial support by the Prof. R.L.J. van Ruyven Foundation for thepublication of this thesis is gratefully acknowledged. Further supportis gratefully acknowledged by Eli Lilly Nederland, Sulzer Medica,Medtronic, Sorin Biomedica and Jostra.

Aan Nienke Aan mijn ouders

Contents

CONTENTS

General introduction and outline of this PhD thesis 1

Chapter 1:

Pneumonectomy for lung metastases: report of ten cases. 17

Thorac Cardiovasc Surg 2003;51:38-41.

Chapter 2:

Isolated lung perfusion for the treatment of pulmonary metastases: 27

current review of work in progress.

Surg Oncol 2003, in press.

Chapter 3:

Isolated lung perfusion with gemcitabine in a rat: pharmacokinetics 45

and survival.

J Surg Res 2003;109:118-22.

Chapter 4:

Pharmacokinetics after pulmonary artery perfusion with 61

gemcitabine.

Ann Thorac Surg 2003, in press.

Contents

Chapter 5:

Single-Pass isolated lung perfusion versus recirculating isolated 77

lung perfusion with melphalan in a rat model.

Ann Thorac Surg 2002;74:893-898.

Chapter 6:

Gemcitabine prolongs survival in a rat model of metastatic 97

pulmonary adenocarcinoma.

Submitted.

Chapter 7:

Combination chemotherapy with gemcitabine in a rat model of 117

isolated lung perfusion for the treatment of pulmonary metastases.

Submitted.

General discussion 133

Summary (English and Dutch) 149

Acknowledgements 157

Curriculum Vitae 163

General introduction and outline of

this PhD thesis

General introduction

2

General introduction and outline of this PhD thesis

With exception of lymph nodes, the lungs are the most common site of

metastatic involvement for all invasive cancer types. The most obvious

reason is the filtering of circulating tumour cells by the pulmonary capillairy

bed. The incidence of pulmonary metastases is unknown, and has been

estimated to occur in approximately 50% of patients with non-pulmonary

malignancy.

The most common primary malignancies are breast, colon, kidney, uterus,

prostate and nasopharyngeal carcinomas. Other primary tumour types that

especially metastasise to the lungs are melanoma, germ cell tumour, soft

tissue sarcoma and osteogenic sarcoma. The high incidence of pulmonary

metastases in autopsy studies may overestimate both the true incidence of

initial lung involvement and its clinical significance.

Current treatment of pulmonary metastases consists of surgical resection and

adjuvant chemotherapy. Results from the international database, reported by

Pastorino et al, show a 5-year survival after metastasectomy of almost 70%

for patients with germ cell tumours, 37% for epithelial tumours, 30% for

sarcomas and 20% for melanomas. The mean 5-year survival of pulmonary

metastases is 37% after surgical resection and adjuvant chemotherapy and

0% if not treated.

Isolated lung perfusion (ILuP) is an experimental surgical technique for the

treatment of pulmonary metastases in order to improve the current 5-year

survival of approximately 40% after resection of lung metastases. The main

advantage is that higher lung levels can be obtained without systemic

toxicity which reduces overall toxicity. ILuP as an adjuvant treatment during

surgical resection aims to destroy micrometastases but its use as induction

therapy should be explored as well.

General introduction

3

In this thesis the results are presented of isolated lung perfusion with

gemcitabine for the treatment of pulmonary metastases. This study was set

up in 1999 at the Laboratory of Experimental Surgery at the University of

Antwerp. The primary aim was to find out the efficay of gemcitabine for the

treatment of pulmonary micrometastases from colorectal adenocarcinoma.

Its use in the clinical setting of isolated lung perfusion was studied and

compared to intravenous therapy.

Gemcitabine and melphalan are the main drugs studied in this thesis.

Therefore, their mechanisms of action and their clinical activity are

described at the end of this general introduction.

As an introduction, the role of pneumonectomy for the current treatment of

lung metastases is discussed in chapter 1.

A short review about the experimental work is presented in chapter 2.

Several drugs like melphalan, doxorubicin, cisplatin and TNF-α have been

investigated in this model. Melphalan was the most successful anticancer

drug used in isolated lung perfusion which resulted in a currently ongoing

clinical trial.

In chapter 3, in vitro and in vivo (acute) toxicity and pharmacokinetics of

gemcitabine in isolated lung perfusion are compared to systemic intravenous

treatment using the unique rat model of isolated lung perfusion which was

firstly described by Weksler and modified in our laboratory by Hendriks in

1999.

The disadvantage of ILuP is its invasive nature which limits repetitive

treatment. In preparation of less invasive techniques, chapter 4 contains a

description of a pharmacokinetic study to the first-pass effect of gemcitabine

in a rat model of selective perfusion of the pulmonary artery during blood

flow occlusion without clamping the pulmonary veins.

General introduction

4

In chapter 5, two specific techniques of isolated lung perfusion namely

single-pass perfusion versus reperfusion, were compared using melphalan in

the perfusion circuit. Furthermore, the influence of red blood cells as a

possible carrier of melphalan was evaluated using rabbit red blood cells

which were processed according to a specific protocol.

In chapter 6, long-term toxicity and efficacy of isolated lung perfusion with

gemcitabine are presented. For this study a model of unilaterally induced

pulmonary metastases was used which prevented death of the rats due to

bulky disease on the untreated (right) side. Furthermore, lung fibrosis was

studied in a dose escalating toxicity study whereby isolated lung perfusion

with gemcitabine was compared with intravenous infusion.

Because of their different mechanisms of cytotoxicity, chapter 7 describes

the in vivo efficacy of combinations of gemcitabine, cisplatin and melphalan

for the treatment of pulmonary metastatic colorectal adenocarcinoma in a rat

model of isolated lung perfusion.

Discussion (English) and summary (English and Dutch)

General introduction

5

Gemcitabine

Cellular actions

2’,2’-Difluorodeoxycytidine (dFdC, gemcitabine) is a nucleoside analogue

of deoxycytidine in which two fluorine atoms are incorporated in the

deoxyribofuranosyl ring. dFdC enters the cell by several transport

mechanisms either facilitated or energy-dependent processes. These

mechanisms are currently not fully understood. dFdC is phosphorylated by

deoxycytidine kinase which results in gemcitabine di- (dFdCDP) and

triphosphate (dFdCTP). Studies using radioactive dFdC showed that

dFdCTP was the major cellular metabolite compared to dFdCMP and

dFdCDP [1]. This observation implies that phosphorylation to dFdCMP is

the rate-limiting step in the formation of dFdCTP. After incorporation of

phosphorylated gemcitabine into the DNA, one nucleotide has to be added

until DNA polymerase is not capable any more to proceed (“masked chain

termination”). Because gemcitabine is locked in the DNA, proofreading

exonucleasen are unable to remove gemcitabine nucleotides from this

penultimate position. Polymerization experiments showed that this

phenomenon could be explained by an alteration of the primer configuration

at the 3’-terminus due to the incorporated analogue and therefore a change of

its substrate properties [2]. This effect on polymerase progression is more

severe than just a pause, suggesting that if polymerases can extend beyond

single dFdC residues, incorporation of multiple residues has a seriously

inhibitory effect either on DNA replication or (re)synthesis in response to

damage [3]. Incorporation of gemcitabine is strongly correlated with DNA

synthesis inhibition and loss of viability. Mainly S-phase cells proved to be

sensitive to gemcitabine finally resulting in apoptosis characterized by

generation of internucleosomal DNA fragmentation and apoptotic

General introduction

6

morphology [1]. Two types of DNA fragmentation can be found: (1)

nucleosomal-sized DNA fragmentation which is calcium-dependent and (2)

large-sized double-stranded DNA fragmentation which is not calcium-

dependent [4]. Beside incorporation in the DNA mainly responsible for loss

of viability, gemcitabine is also incorporated into the RNA which is

concentration-dependent as well as the incorporation into the DNA. RNA

synthesis in the CCRF-CEM cell line (human leukemic cell line) is inhibited

after incubation during 24 hours with gemcitabine at 1µM [5,6].

dFdC Elimination is dependent on the cellular concentrations of dFdCTP [1].

Monophasic kinetics was observed at dFdCTP concentrations less than 100

µmol/L while biphasic kinetics with a prolonged terminal elimination phase

was seen at concentrations greater than 100 µmol/L [4]. Metabolic clearance

by deamination activity converts dFdC to dFdU which is predominated in

cells containing relatively low dFdCTP levels. In contrast, dFdC is excreted

from cells having high levels of dFdCTP. So it can be concluded that

deamination is the major elimination route in cells with less than 100 µmol/L

while excretion and dephosphorylation is more important in cells with

dFdCTP levels greater than 100 µmol/L [1]. These results were confirmed

by studies blocking deaminase activity that changed the elimination route

from mono- with a terminal half-life of 3.6 hours to biphasic with a terminal

half-life of 19 hours [4]. Subsequent studies showed direct inhibition of

deaminase activity by dFdCTP and indirectly by depletion of the dCTP pools

that are required for deaminase activity [1].

Compared with ara-C, gemcitabine serves as a better transport substrate, is

phosphorylated more efficiently, eliminated more slowly and contains three

mechanisms of self-potentiating action.

General introduction

7

Firstly, gemcitabine diphosphate (dFdCDP) reduces activity of

ribonucleotide reductase which normally produces deoxyribonucleotides and

is time- and concentration-dependent. This results in a reduction of mainly

nuclear dCTP resulting in an increased phosphorylation and therefore

incorporation of gemcitabine nucleotides into the DNA which are

competitive with nuclear dCTP [7]. Furthermore, a reduction of cellular

deoxynucleotide pools necessary for DNA synthesis and repair results in

decreased DNA repair activity.

Secondly as mentioned before, a reduction of nuclear dCTP also results in a

reduced clearance of gemcitabine by deoxycytidine monophosphate

deaminase contributing to the concentration-dependent elimination kinetics.

The combination of slow elimination and efficient phosphorylation are

responsible for accumulation of high concentrations of gemcitabine [8,9].

A third self-potentiating mechanism is a concentration-dependent formation

of active gemcitabine di- and triphosphate which results in an increased ratio

of gemcitabine triphosphate and deoxycytidine triphosphate [9,10].

In conclusion, these metabolic properties and mechanisms of self-

potentiating action enhance effective accumulation and prolonged

intracellular retention that are paralleled the incorporation of gemcitabine

into the DNA and loss of viability. This shows evidence for a mechanistic

relationship between the metabolism of gemcitabine and its biologic actions

[4].

Anti-tumour activity

Clinical activity of intravenous single-agent treatment with gemcitabine was

shown in ovarian, breast, colorectal, breast, pancreas and NSCLC [11,12].

Gemcitabine is extremely well tolerated even in heavily pre-treated patients

General introduction

8

[13]. The combination of its mechanistic and metabolic properties suggests a

possibility of synergistic effects if combined with some other

chemotherapeutics [14]. Response rates in treatment of advanced breast

cancer with gemcitabine vary from 25 to 46% [13]. Combinations of

gemcitabine with several products like doxorubicin, epirubicin and

paclitaxel, cisplatin, paclitaxel, docetaxel, 5-FU and vinorelbine have been

studied [15-21]. The best results have been achieved with gemcitabine

combined with epirubicin and paclitaxel [16].

Furthermore, gemcitabine seems to be a promising radiation sensitizer [22].

Radiosensitization occurs under conditions in which cells demonstrate

concurrent redistribution into S phase and deoxyadenosine triphosphate pool

depletion. Prolonged clinical benefit response and disease stabilization was

achieved in patients with localized, unresectable pancreatic cancer [23].

Melphalan

Mechanism of action

Melphalan is a bifunctional alkylating agent that is actively transported into

cells by the high-affinity L-amino acid transport system. Since this system

also transports the amino acids glutamine and leucine, high concentrations of

either of these aminoacids will reduce the uptake in tumour cells in vitro

(24). A second less effective transport system may contribute to melphalan

uptake next to alanine, serine and cysteine. In addition to aminoacids,

melphalan uptake is also inhibited by tamoxifen, doxorubicin,

aminophylline, chlorpromazine, and indomethacin (25).

Once transported into the cell, cytotoxicity is accomplished by forming

either DNA-interstrand, DNA-intrastrand, or DNA-protein cross-links (26).

Probably, melphalan quickly forms monoadducts, which then only slowly

General introduction

9

convert to cytotoxic DNA-DNA or DNA-protein bonds (27). Similar to other

alkylating agents, melphalan is non-cell cycle phase specific which makes it

active against both resting and rapidly dividing tumour cells (28).

Drugs and compounds that increase the cytotoxicity of melphalan include 1-

deamino-8-D-arginine vasopressin (increases intracellular concentrations),

nicotinamide (inhibition of poly [adenosine diphosphate-ribosel polymerase,

which increases the half-life ), tissue oxygenation, and intracellular

glutathione (GSH) (29-31).

Antitumour effect in clinical trials

The activity of intravenously infused melphalan at doses less than 65 mg/m2

has been reviewed by Sarosy (33). At this dose, melphalan has documented

activity in ovarian cancer, rhabdomyosarcoma, pancreatic carcinoma,

osteogenic sarcoma, and glioma. At doses less than 65 mg/m2, melphalan has

little activity in breast, lung, or colorectal cancer. However, with high dose

melphalan given intravenously an extraordinarily high response rate is seen

in "resistant" solid tumours such as melanoma (34,35) and colorectal

carcinoma (36,37). The observed response rate of 45% after high-dose

melphalan (HDM) and bone marrow transplantation (BMT) in metastatic

colon cancer as reported by Leff is impressively high (36). Unfortunately,

this response is of brief duration. Leff et al noted nine objective responses

among 20 evaluable patients, with a median duration of survival of 198 days.

Spitzer treated 14 patients, previously treated with, and refractory to,

conventional chemotherapy (37). They received melphalan 120 to 180

mg/m2 (total dose) divided as 40-60 mg/m2 iv daily for 3 days with the

modulator misonidazole. The response rate was six of 14 patients (43%) and

all responses were PRs. The median duration of response was 4 months.

General introduction

10

Isolated Limb Perfusion

Melphalan was identified as a potential agent for use in isolated limb

perfusion (ILP) because in vitro studies demonstrated that it was taken up

and retained by melanocytes and it was one of the few alkylating agents

available in the late 1950s when this operation was conceived (40).

The amount of melphalan administered during ILP has varied over time.

Initial series calculated melphalan dose on the basis of body weight with a

maximum tolerated dose of 1.75 mg/kg and a 2.0 mg/kg dose-limiting

regional toxicity (41). However, this calculation does not take into account

the wide variation of body habitus and the distribution of weight among

individuals. Therefore, melphalan dose is now calculated on the basis of

limb volume as proposed by Wieberdink (42, 43). The most widely used

dose based on limb volume is 10mg/L limb volume for the lower extremity

and 13mg/L limb volume for the upper extremity perfusion (44). A phase I

trial of escalating melphalan dose defined 14 mg/L as the MTD, with 16

mg/L or a total dose greater than 150 mg resulting in dose-limiting regional

toxicity (45).

The majority of melphalan leaves the perfusate during the course of a 60-

min treatment, but most of the dose is absorbed by the tissue of the perfused

extremity within the first 30 min. The initial loss with a half-life (T1/2) of

approximately 5-10 min is interpreted as rapid uptake of melphalan by the

tissue. The terminal portion of the curve with a half-life of approximately

35-50 min is interpreted as due to the hydrolysis of melphalan, with a lesser

component of loss due to absorption to the filters and tubing of the perfusion

apparatus. Based on the AUC, the authors suggest that perfusion with

melphalan for longer than 30 to 40 min is unnecessary but further studies are

required (46).

General introduction

11

Two important prospective trials that have evaluated the use of ILP with

melphalan in an adjuvant setting for localized, high-risk melanoma showed a

significant decrease in recurrent disease but no difference in overall survival

rate between treatment groups (47, 48).

A few prospective randomized trials of therapeutic ILP have been reported,

both normothermic (48-49) and hyperthermic (50-52). No trials have

compared ILP with systemic treatment since the response rates of the latter

are too low. Generally, complete responses for therapeutic ILP with

melphalan alone are 40-60%, with an overall response of approximately

80%.

Interest in ILP as a surgical therapy has been renewed since adding highdose

Tumour Necrosis Factor to a melphalan ILP.

References1. Heinemann V, Xu YZ, Chubb S, Sen A, Hertel LW, Grindey GB, Plunkett W. Cellular

elimination of 2’-2’-difluorodeoxycytidine 5’phosphate: a mechanism of self-

potentiating. Cancer Res 1992;52:533-39.

2. Huang P, Chubb S, Hertel LW, et al. Action of 2’,2’-difluorodeoxycytidine on DNA

synthesis. Cancer Res 1991;51:6110-7.

3. Yang L-Y, Li L, Liu L, et al. Gemcitabine suppresses the repair of cisplatin adducts in

plasmid DNA by extracts of cisplatin resistant human colon carcinoma cells. Proc Am

Assoc Cancer Res 1995;36:357 (abstract).

4. Plunkett W, Huang P, Searcy CE, Gandhi V. Gemcitabine: preclinical pharmacology and

mechanisms of action. Sem Oncol 1996;23:3-15).

5. Van Haperen R, Veerman G, Vermorken JB, Peters GJ. Biochem Pharmacol

1993;46:762-6.

6. Heinemann V, Schulz L, Issels RD, Plunkett W. Gemcitabine: a modulator of

intracellular nucleotide and deoxynucleotide metabolism. Semin Oncol 1995;22:11-8.

7. Plunkett W, Huang P, Gandhi V. Preclinical characteristics of gemcitabine. Anticancer

drugs 1995;6:7-13.

General introduction

12

8. Huang P, Plunkett W. Induction of apoptosis by gemcitabine. Semin Oncol 1995;22:19-

25.

9. Storniolo AM, Allerheiligen SR, Pearce HK. Preclinical pharmacologic and phase I

studies of gemcitabine. Sem Oncol 1997;2:S7-2-S7-7.

10. Huang P, Plunkett W. Fludarabine- and gemcitabine-induced apoptosis: incorporation of

analogues into DNA is a critical event. Cancer Chemother Pharmacol 1995;36:181-8.

11. Storniolo AM, Enas NH, Brown CA, Voi M, Rothenberg ML, Schilsky R. An

investigational new drug treatment program for patients with gemcitabine: results for

over 3000 patients with pancreatic carcinoma. Cancer (Phila.)1999;85:1261-8.

12. Sandler A, Ettinger DS. Gemcitabine: single-agent and combination therapy in non-small

cell lung cancer. Oncologist 1999;4:241-51.

13. Carmichael J, Walling. Advanced breast cancer: investigational role of gemcitabine. J.

Eur J Cancer 1997;33:S27-30.

14. Van Moorsel CJ, Veerman G, Bergman AM, Guechev A, Vermorken JB, Postmus PE,

Peters. Combination chemotherapy studies with gemcitabine. GJ. Semin Oncol

1997;24:S7-17-S7-23.

15. Rivera E, Valero V, Syrewicz L et al. Phase I study of stealth liposomal doxorubicin in

combination with gemcitabine in the treatment of patients with metastatic breast cancer.

J Clin Oncol 2001;19:1716-22.

16. Gemcitabine, epirubicine, and paclitaxel combinations in advanced breast cancer. Semin

Oncol 2001;28:15-7.

17. Nagourney RA. Gemcitabine plus cisplatin in breast cancer. Oncology 2001;15:28-33.

18. Murad AM, Guimaraes RC, Aragao BC, Scalabrini’Neto AO, Rodrigues VH, Garcia R.

Gemcitabine and paclitaxel as salvage therapy in metastatic breast cancer. Oncology

2001;15:25-7.

19. Georgoulas VA. Docetaxel/gemcitabine: salvage chemotherapy in anthracycline-

pretreated patients with advanced breast cancer. Oncology 2001;15:18-24.

20. Hortobagyi GN. Treatment of advanced breast cancer with gemcitabine and vinorelbine.

Oncology 2001;15:15-7.

21. Awada A, Biganzoli L, Cufer T et al. An EORTIC-IDBBC phase I study of gemcitabine

and continuous infusion of 5-fluorouracil in patients with metastatic breast cancer

resistant to anthracyclines or pre-treated with both anthracyclines and taxanes. Eur J

Cancer 2002;37:773-8.

General introduction

13

22. Lawrence TS, Eisbruch A, Shewach DS. Gemcitabine-mediated radiosensitization.

Semin Oncol 1997;24:S7-24-S7-28.

23. Epelbaum R, Rosenblatt E, Nasrallah S, Faraggi D, Gaitini D, Mizrahi S, Kuten A. Phase

II study of gemcitabine combined with radiation therapy in patients with localized,

unresectable pancreatic cancer. J Surg Oncol 2002;81-138-43.

24. Vistica DT, Toal NJ, Rabinowitz M. Amino-acid conferred protection against

melphalan: interference with leucine protection of melphalan cytotoxicity by the basic

aminoacids in cultured murine L1210 leukemic cell. Mol Phan-nacol 1978; 14: 1136-42.

25. Goldenberg CJ, Froese EK. Antagonism of the cytocidal activity and uptake of

melphalan by tamoxifen in human breast cancer cells in vitro. Biochem Pharmacol 1985;

34: 763-70.

26. Kohn KW. Molecular mechanisms of crosslinking of alkylating agents and platinum

complexes. In: Sartorelli AC, Lazo JS, Bertino JR (eds.): Molecular actions and targets

for cancer chemotherapeutic agents. San Diego, Academic, 198 1, pp 3 -16.

27. Ross WE, Ewig RAG, Kohn KW. Differences between melphalan and nitrogen mustard

in the formulation and removal of DNA crosslinks. Cancer Res 1978; 38: 1502-6.

28. Calebresi P, Parks CE Jr. Antiproliferative agents and drugs used for

immunosuppression. In: Gieman AG, Goodman LS, Gilman A (eds.): The

pharmacologic basis of therapeutics (ed.6). New York, Macmillan, 1980, pp 1169-1313.

29. Brown DM, Horsman MR, Hirst DG et al. Enhancement of melphalan cytotoxicity in

vivo and in vitro by inhibitors of poly (ADP-ribose) polymerase. Int J Radiat Oncol Biol

Phys 1984; 10: 1665-8.

30. Muler L, Kobrinsky NL, Goldenberg CJ. Modulation of membrane transport of

alkylating agents and aminoacids by an analogue of vasopressin in murine L5178Y

lymphoblasts in vitro. Biochem Pharmacol 1987; 36: 169-76.

31. Teicher BA, Crawford JM, Holden SA et al. Effects of various oxygenation conditions

on the enhancement of fluosol-DA of melphalan antitumour activity. Cancer Res 1987;

47: 5036-41.

32. Suzukake K, Petro BJ, Vistica DT. Preliminary communications: reduction of

glutathione content of L-PAM resistant L1210 cells confers drug sensitivity. Biochem

Pharmacol 1982; 31: 121-124.

33. Sarosy G, Leyland-Jones B, Soochan P et al. The systemic administration of intravenous

melphalan. J Clin Oncol 1988; 6: 1768-82.

General introduction

14

34. Kessinger A. High-dose chemotherapy with autologous bone marrow transplantation or

malignant melanoma. J Am Acad Dermatol 1985; 12: 337-43.

35. Knight WA Ill, Page CP, Kuhn JG et al. High-dose L-PAM and autologous marrow

infusion for refractory solid tumours. Proc Am Soc Clin Oncol 1984; 3: 150.

36. Leff RS, Thompson JM, Johnson DB et al. Phase II trial of high-dose melphalan and

autologous bone marrow transplantation for metastatic colon carcinoma. J Clin Oncol

1986; 4: 1586-91.

37. Spitzer TR, Lazarus HM, Creger RJ et al. High-dose melphalan, misonidazole, and

autologous bone marrow transplantation for the treatment of metastatic colorectal

carcinoma. Am J Clin Oncol 1989;12: 145 -5 1.

38. Creech Jr. Krementz ET, Ryan RF, Winblad JN. Chemotherapy of cancer: regional

perfusion utilizing extracorporeal circuit. Ann Surg 1958;148: 616-32.

39. Scott RN, Blaclde R, Kerr DJ et al. Melphalan in isolated limb perfusion for malignant

melanoma, bolus or divided dose, tissue levels, the pH effect. In: Jakesz R, Rainer H,

editors. Progress in regional cancer therapy. Berlin: Springer-Verlag 1990. p 195-200.

40. Wieberdink J, Benckhuysen C, Braat RP, van Slooten EA, Olthuis GAA. Dosimetry in

isolation perfusion of the limbs by assessment of perfused tissue volume and grading of

toxic tissue reactions: Eur J Cancer Clin Oncol 1982; 18: 905-10.

41. Benckhuysen C, Kroon BBR, van Geel AN, Wieberdink J. Regional perfusion treatment

with melphalan for melanoma in a limb: an evaluation of drug kinetics. Eur J Surg Oncol

1988; 14: 157-63.

42. Lejeune FJ, Ghanem GE. A simple and accurate new method for cytostatic dosimetry in

isolation perfusion of the limbs based on exchangeable blood volume determination.

Cancer Res 1987; 47: 639-43.

43. Boddie AW, Briele H, Krementz E et al. A phase I study of melphalan in 40'C isolated

limb perfusion using packed red blood cells and lactated ringers perfusate (abstract).

Proc ASCO 1992; 11:351.

44. Briele H, Djuric M, Jung DT et al. Pharmacokinetics of melphalan in clinical isolation

perfusion of the extremities. Cancer Res 1985; 45: 1885-9.

45. Koops HS, Vaglini M, Suciu S et al. Prophylactic isolated limb perfusion for localized,

high risk limb melanoma: results of a multicenter randomized phase II trial. European

Organization for Research and Treatment of Cancer Malignant Melanoma Cooperative

Group Protocol 18832, the World Health Organization Melanoma Program Trial 15, and

General introduction

15

the North American Perfusion Group Southwest Oncology Group 8593. J Clin Oncol

1998; 16: 2906-12.

46. Hafstrom L, Rudenstam CM, Blomquist E et al. Regional hyperthermic perfusion with

melphalan after surgery for recurrent malignant melanoma of the extremities. J Clin

Oncol 199 1; 9: 2091-4.

47. Klaase JM, Kroon BBR, van Geel AN, Eggermont AMM, Franklin HR, Hart AMM.

Prognostic factors for tumour response and limb recurrence-free survival interval in

patients with advanced melanoma of the limbs treated with regional isolated perfusion

with melphalan. Surgery 1994; 115:39-45.

48. Bulman A, Jamicson CVV. Isolated limb perfusion with melphalan in the treatment of

malignant melanoma. Br J Surg 1980; 67: 660-2.

49. Kroon BBR, Klaase JM, van Geel AN, Eggermont ANN, Franklin HR, van Dongen JA.

Results of a double perfusion schedule with melphalan in patients with melanoma of the

lower limb. Eur J Cancer 1993; 29A: 325-8.

50. Skene Al, Bulman AS, Williams TF, Thomas JM, Westbury G. Hyperthermic isolated

limb perfusion with melphalan in the treatment of advanced malignant melanoma of the

lower limb. Br J Surg 1990; 77: 765-7.

51. Lejeune FI, Deloof T, Ewalenko P et al. Objective regression of unexcised melanoma in-

transit metastases after hyperthermic isolation perfusion of the limbs with melphalan. In:

Recent results in cancer research. Vol 86. Berlin: Springer-Verlag 1983. P 268-76.

52. Storm FK, Morton DL. Value of therapeutic hyperthermic limb perfusion in advanced

recurrent melanoma of the lower extremity. Am J Surg 1985; 150: 32-5.

Chapter 1

Chapter 1

Pneumonectomy for lung metastases: report of ten cases

JMH Hendriks1, BP Van Putte1, S Romijn1, J Van Den Brande 2,

JB Vermorken2 and PEY Van Schil1

1 Department of thoracic and vascular surgery,2 Department of medical oncology,

University Hospital Antwerp, Antwerp, Belgium

Published in: Thorac Cardiovasc Surg 2003;51:38-41.

Chapter 1

18

Abstract

Today, pulmonary resection for lung metastases is a widely accepted

treatment if complete resection can be achieved. However, 5-year survival is

only 40%. Many patients develop recurrences, but some reports have

demonstrated that salvage operations can result in a long-term survival. A

resection of a complete lung or a resection of more than a lung is still

controversial since procedure-related morbidity or mortality does not

outweigh the survival benefit. We report on a series of 10 consecutive

patients who underwent a primary pneumonectomy or an operation on the

residual lung after pneumonectomy with curative intent for pulmonary

metastases. 5-Year survival rates for the 10 patients after pneumonectomy

alone or with additional resection was 45 %, which was not significantly

different from those who underwent a more minor resection with a 5-year

survival of 39% (p = 0.40). Since there is currently no alternative proven

therapy for patients with isolated pulmonary metastases, a primary or

completion pneumonectomy may be offered to selected patients as long as

sufficient pulmonary reserve is present, and a complete resection can be

achieved.

Introduction

At present, pulmonary resection for lung metastases is a widely accepted

treatment if complete resection can be achieved. However, 5-year survival

rate is only approximately 40% [1]. Many patients develop recurrences that

are often inoperable due to insufficient pulmonary reserve or location, or the

number of metastases. Intrathoracic recurrences occur in more than 50% of

epithelioma and sarcoma metastasis patients [2,3]. Some reports have

demonstrated that salvage operations can result in long-term survival

Chapter 1

19

provided that a complete resection of all metastatic disease is accomplished

[4,5]. However, a resection of a complete lung alone or with additional

resection remains controversial since procedure-related morbidity or

mortality does not outweigh survival benefit. Two patients who underwent a

resection beyond pneumonectomy in our institution were reported in 1999

[6]. In this communication, we report on a series of 10 consecutive patients

who underwent a primary pneumonectomy or an operation on the residual

lung after pneumonectomy with curative intent for pulmonary metastases.

Materials and Methods

Over a l0-year period from January 1990 to December 2000, 10 patients

were operated on for recurrent pulmonary metastatic disease resulting in a

pneumonectomy alone or with additional resection in order to achieve a

complete resection. All patients had a complete preoperative work-up to

exclude extrathoracic metastatic disease. These investigations included a

computed tomographic (CT) scan of the thorax, a bone scan, a CT scan of

the abdomen and a PET scan in the last two patients. When the thoracic CT

scan showed enlarged mediastinal lymph nodes, a cervical mediastinoscopy

was performed. There were 7 male and 3 female patients with a mean age of

42 years (range, 17 to 61 years) at the time of the initial operation. These

patients are described in Table 1, and represent 14% (10/73) of all

pulmonary metastasis operations at our Department of Thoracic Surgery

over the same period [3].

Chapter 1

20

Results

In 3 patients, a primary pneumonectomy was performed, meaning that

removal of an entire lung was necessary at the first metastasectomy. The first

patient developed lung metastases 180 months after resection of a

leiomyosarcoma in the back. After the initial right pneumonectomy, he

underwent an additional wedge resection on the left side due to tumour

recurrence. Three nodules were removed, and he survived for 50 months

after the initial pulmonary resection. He finally died of recurrent disease

outside the lung (Table 1, patient no. 2).

The second patient developed metastatic disease 38 months after a sigmoid

resection due to a colon adenocarcinoma. Five nodules were removed in a

left pneumonectomy operation, and the patient died because of a cerebral

primary tumour 66 months after the operation (Table 1, patient no. 7).

The third patient had a right pneumonectomy for lung metastases 12 months

after resection of a leiomyosarcoma in the pubic region. A total of 2 nodules

were resected; the patient survived for only 8 months, dying of recurrent

disease in this left lung (Table 1, patient no. 8).

All other patients (n = 7) had repeated salvage surgery for lung metastases

resulting in a completion pneumonectomy alone or with additional resection.

Indications were one teratocarcinoma in two patients and one liposarcoma,

leiomyosarcoma, adenocystic carcinoma, osteosarcoma and chondrosarcoma

in one each of the others. The number of procedures ranged from 2 to 3

resections for a total of 5 to 11 lesions. The mean diameter of the pulmonary

nodule was 66 mm (range, 25 to 140 mm). The mean time interval between

the initial operation of the primary tumour and resection of the lung

metastases was 37 months (range, 0 to 55 months).

Chapter 1

21

No. Age Gender Histology Location ofPrimary

Number ofprocedures

Type ofPneumonectomySide Alive

Follow-up inmonths

after 1" Ms.

1 57 male Liposarcoma Femur R 3 Completion L yes 302 22 male Leiomyosarcoma Back 2 Primary R no 503 23 male Teratocarcinoma Testis L 2 Completion R yes 1934 61 male Leiomyosarcoma Pubic region 3 Completion R no 485 53 female Adenocystic ca Tongue 2 Completion R no 416 43 male Teratocarcinoma Testis R 2 Completion R no 477 57 male Colon adenoca Sigmoid 1 Primary L no 668 72 female Leiomyosarcoma Back 1 Primary R no 89 17 male Osteosarcoma Knee L 3 Completion L yes 107

10 19 female Chondrosarcoma Femur L 3 Completion R yes 101

Table 1. Patient characteristics

Four patients are still alive 30, 101, 107 and 193 months after the first

pulmonary metastasectomy, respectively (Table 1).There was lymph node

involvement in three cases. These included stations 7 (subcarinal) and 9

(inferior pulmonary ligament), stations 11 (interlobar) and 7, and station 5

(aortopulmonary window respectively, according to the Naruke lymph node

map [7]. These 3 patients with lymph node involvement survived for 23, 41

and 78 months.

The postoperative course was uneventful in 9 further patients. One patient

(Table 1, no. 4) died on the 11th postoperative day of aspiration pneumonia

in the left lung after a right completion pneumonectomy for metastases of a

leiomyosarcoma in the pubic region resected 7 years before! He had

undergone three metastasectomies for a total of 11 metastatic nodules during

these 7 years. The survival analysis is depicted in Fig. 1. The 5-year survival

rate for the 10 patients who underwent pneumonectomy alone or with

additional resection was 45 %, not significantly different from those patients

who underwent a lesser resection with a 5-year survival of 39 % (p = 0.40).

Chapter 1

22

Fig 1.

Survival of patientswith lungmetastasesundergoing apneumonectomy ora lesser resection

In addition, the survival curve starting from the date of the pneumonectomy

is depicted in Fig. 2 (5-year survival of 45 %).

Fig 2.

Survival of patients withlung metastases startingfrom the data of thepneumonectomy

Chapter 1

23

Author No. of patients

with CR

MST in months Operative mortality in % Survival at 5 years in

%

McGovern 1988 [8]

Putnam 1993 [9]

AI-Kattan 1995 [10]

Spaggiari 1998 [11]

Koong 1999 [12]

Bernard 2001 [13]

Present Series

19

19

5

42

112 PP and 31 CP

32

10

20

27

NS

6.5 (1-144)

17 for PP and 29

for CP

NS

50

0

10.5

0

4.8

4 for PP and 3 for CP

9.4

10 (single case)

40.8

21

NS (3 alive> 5 years) 17

17

20 for PP and 30 for CP

NS

45

Table 2. Literature review

Discussion

Surgical resection is a widely accepted treatment for pulmonary metastases

in certain solid tumours. Many patients develop recurrent disease after

metastasectomy, which is reflected by the low 10-year survival rate of 26%

[2]. The optimal therapy is still unknown, but systemic chemotherapy does

not result in a major survival advantage

Regarding the operation, the sole criterion is that only complete resection

can result in long-term survival [2]. Both the primary role and extent of

surgery for recurrent lung metastases remain controversial. The reported

operative mortality rates from various centres ranges from 0% to 10.5% [8-

13] (Table 2). Several reports favour salvage surgery, having demonstrated

long-term survival after repeated resections for recurrent pulmonary

metastases from soft-tissue sarcoma [4,5]. This is probably due to the fact

that the majority of metastatic sarcoma patients have "lung-only metastases"

[14], while more than 50 % of all relapses from sarcoma are intrathoracic

[2,3]. However, this long-term survival was also shown in a heterogeneous

group of other tumours [8,15-17]. Five-year survival rates of around 20% are

seen (range, 9 to 41 %), and are lower for primary pneumonectomy than for

Chapter 1

24

completion pneumonectomy [8-13] (Table 2). This long-term survival may

be lower than the 5-year survival rate of 36% for all pulmonary

metastasectomy cases [2], but it still remains superior to no resection at all.

In conclusion, since no alternative proven therapy is available for patients

with isolated pulmonary metastases, a primary or completion

pneumonectomy may be offered to select patients as long as sufficient

pulmonary reserve is present and complete resection can be achieved. Since

recurrent disease within the lung is probably due to micrometastatic disease

present at the time of initial operation, isolated lung perfusion with high-

dose chemotherapy as an adjunct to surgical resection may be a promising

therapy. Excellent results have been obtained in animal models, but the

results of clinical trials to determine whether there is a real survival benefit

are still pending in patients with pulmonary metastatic disease [18-20].

References1. McCormack P, Bains M, Beattie E Jr, Martini N. Pulmonary resection in metastatic

carcinoma. Chest 1978; 73: 163 – 166.

2. Pastorino U, Buyse M, Friedel G, et al. Long-term results of lung metastasectomy:

prognostic analyses based on 5206 cases. The International Registry for Lung

Metastases. J Thorac Cardiovasc Surg 1997; 113: 37 – 49.

3. J Hendriks JM, Romijn S, van Putte B, et al. Long-term results of surgical resection of

lung metastases. Acta Chir Belg 2001; 101: 267 – 272.

4. Pogrebniak H, Roth J, Steinberg S, Rosenberg S, Pass H. Reoperative pulmonary

resection in patients with soft tissue sarcoma. Ann Thorac Surg 1991; 52: 197 – 203.

5. Casson AG, Putnam JB, Natarajan G, et al. Efficacy of pulmonary metastasectomy for

recurrent soft tissue sarcoma. J Surg Oncol1991; 47: 1-4.

6. Hendriks JM, van Schil PE, Schrijvers D. Bilateral pulmonary resection for lung

metastases: report of two cases. Eur J Surg Oncol 1999; 25: 552 – 553.

7. Naruke T, Suemasu K, Ishikawa S. Lymph node mapping and curability at various levels

of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 1978; 76: 832-839.

Chapter 1

25

8. Me Govern E, Trastek V, Pairolero P, Payne W. Completion pneumonectomy:

indications, complications and results. Ann Thorac Surg 1988; 46: 141-146.

9. Putnam J. Douglas M, Natarajan G. Roth J. Extended resection of pulmonary metastases:

is the risk justified? Ann Thorac Surg 1993; 55: 1440-1446.

10. AI-Kattan K, Goldstraw P. Completion pneumonectomy: indications and outcome. J

Thorac Cardiovasc Surg 1995; 110: 1125 –1129.

11. Spaggiari L, Grunenwald H, Gerard Ph, Solli P, Le Chevalier Th. Pneumonectomy for

lung metastases: indications, risks, and outcome. Ann Thorac Surg 1998; 66: 1930-1933.

12. Koong H, Pastorino U, Ginsberg R, for the International Registry of Lung Metastases. Is

there a role for pneumonectomy in pulmonary metastases: Ann Thorac Surg 1999; 68:

2039 –2043.

13. Bernard A, Deschamps C, Allen M, et al. Pneumonectomy for malignant disease: factors

affecting early morbidity and mortality. J Thorac Cardiovasc Surg 2001; 121: 1076 –

1082.

14. Van Geel A, Pastorino U, Jauch K, et al. Surgical treatment of lung metastases: The

European Organization for Research and Treatment of Cancer-Soft Tissue and Bone

Sarcoma Group study of 255 patients. Cancer 1996; 77: 675-682.

15. Kandioler D, Kromer E, Tuchler H, et al. Long-term results after repeated surgical

removal of pulmonary metastases. Ann Thorac Surg 1998; 65: 909-912.

16. Jaklitsch M, Mery C. Lukanich J, et al. Sequential thoracic metastasectomy prolongs

survival by re-establishing local control within the chest. J Thorac Cardiovasc Surg

2001; 121: 657 – 667.

17. Grunenwald 0, Spaggiari L, Girard Ph. Baldeyrou P, Filaire M, Dennewald G.

Completion pneumonectomy for lung metastases: is it justified? Eur J Cardiothoracic

Surg 1997; 12: 694-697.

18. Weksler B, Burt M. Isolated lung perfusion with antineoplastic agents for pulmonary

metastases. In: Faber LP, McCormack PM eds. Chest Surgery Clinics of North America.

Metastatic disease to the lung. Pennsylvania: W.B. Saunders 1998; 8: 157 –182.

19. Hendriks JM. Pulmonary metastases. Therapeutic Application of Isolated Lung

Perfusion. PhD Thesis, Antwerp University, 2001.

20. Burt M, Liu 0, Abolhoda A, et al. Isolated lung perfusion for patients with unresectable

metastases from sarcoma: a phase I trial. Ann Thorac Surg 2000; 69: 1542 –1549.

Chapter 2

Chapter 2

Isolated Lung Perfusion for the treatment of pulmonary

metastases: current review of work in progress

BP Van Putte, JMH Hendriks, S Romijn and PEY Van Schil

Department of thoracic and vascular surgery,

University Hospital Antwerp, Antwerp, Belgium

Will be published in: Surg Oncol 2003 (in press)

Chapter 2

28

Abstract

Surgical resection of lung metastases is a widely accepted procedure but

long-term results are disappointing with a 5-year survival rate of

approximately 40%. Pulmonary metastasectomy is only indicated when

complete resection can be achieved. A better survival is reported in patients

with a single metastasis or a disease-free survival of more than 3 years.

Intravenous chemotherapy has no major impact on survival because high-

dose therapy is limited by systemic side effects. Isolated lung perfusion has

the advantage of both selectively delivering an agent into the lung while

diverting the venous effluent. This allows the drug to be given in a

significantly higher dose compared to intravenous therapy, while drug levels

in critical organs are kept low enough to avoid significant morbidity.

Isolated lung perfusion has proven to be effective for the treatment of lung

metastases in animal models while the procedure is technically safe in

humans. However, the real clinical value and survival benefit remain to be

determined in ongoing clinical trials.

The aim of this paper was to update the literature on isolated lung perfusion

for the treatment of lung metastases. Furthermore, some proposals are made

in order to improve the ultimate prognosis of these patients.

Introduction

Cancer remains the second cause of death with almost 550 000 deaths

expected for the USA in 2002 [1]. Currently, 40-50% of all cases of

malignancy are cured [2]. The non-cured group mostly suffers from

metastases. At the moment of diagnosis of the primary tumour, metastases,

which are too small to be detected, so called micrometastases, may already

be present. Furthermore, widespread metastatic disease may be symptomless

Chapter 2

29

at an early stage.

With exception of the lymph nodes, the lungs are the most common site of

metastatic involvement for all invasive cancer types. The most obvious

reason is the filtering of circulating tumour cells by the pulmonary capillary

bed. The precise incidence of pulmonary metastases is unknown, but has

been estimated to occur in up to 50% of patients with non-pulmonary

malignancy [2].

The cumulative incidence and prevalence of pulmonary metastases vary

widely according to tumour type and observed time interval between the

occurrence of the primary cancer and its metastases. Some tumour types

preferentially metastasise to the lungs, such as sarcomas, germ cell tumours

and some paediatric malignancies. Gastrointestinal tumours will first

metastasise to the liver and from there to the lungs. The high incidence of

pulmonary metastases in autopsy studies may overestimate both the true

incidence of initial lung involvement and its clinical significance.

Current treatment of pulmonary metastases consists of surgical resection and

adjuvant chemotherapy. Results from the international database, reported by

Pastorino et al., showed a 5-year survival after metastasectomy of almost

70% for patients with germ cell tumours, 37% for epithelial tumours, 30%

for sarcomas and 20% for melanomas [3]. Furthermore, this study

investigated prognostic factors for lung metastasectomy. Radiological

staging is unreliable with an accuracy in patients undergoing unilateral

thoracotomy of 75%, while accuracy in patients undergoing median

sternotomy or bilateral thoracotomy was only 37%. Thorough, intraoperative

staging is therefore required to identify and resect all metastases. Pastorino

et al. demonstrated the prognostic value of the number of metastases with a

Chapter 2

30

Group Prognostic factors Median survival

Group 1Resectable, no risk factor (DFI > 36 months and

single metastasis)61 months

Group 2Resectable, 1 risk factor (DFI < 36 months or

multiple metastases)34 months

Group 3Resectable, 1 risk factor (DFI < 36 months and

multiple metastasis)24 months

Group 4 Unresectable 14 months

Table 1. Prognostic groups in patients with lung metastases based on threeprognostic parameters: respectability, DFI and number of metastases [3].

5-year survival rate of 43% in case of a single metastasis dropping to 27% in

case of 4 or more metastases. They also showed that the proportion of

patients that underwent a second metastasectomy was in accordance with the

probability of intrathoracic relapse. The 5-year survival of patients who had

a second metastasectomy was remarkably good (44% versus 34% for

patients having a single operation). Based on three parameters of prognostic

significance (resectability, disease-free interval and number of metastases)

four prognostic groups could be identified (table 1) taking all parameters into

account [3]. In 1999 we published a mini-review on isolated lung perfusion

[4]. In this manuscript we updated the results with isolated lung perfusion,

both experimentally and clinically.

Isolated Lung Perfusion

Experimental research in animal models

In selected cases, pulmonary metastasectomy is an effective treatment for

patients with isolated lung metastases. However, 5-year survival rate for all

tumour types is only 30-40% [3]. Intravenous therapy is dose-limited due to

Chapter 2

31

systemic side effects. So novel techniques should be developed in order to

improve prognosis. It was Creech in 1959 who first reported a model of

isolated lung perfusion to achieve high local drug levels without apparent

systemic toxicity [5]. He perfused both lungs simultaneously with divided

circuits for systemic and pulmonary circulations. In 1960, Pierpont and

Blades and in 1961, Jacobs described a method of isolated in vivo perfusion

of a single lung. As shown by radioisotope studies no leakage of the drug

outside the pulmonary circulation was present [6,7]. Johnston put the base

for further clinical and experimental studies by demonstrating isolated

single-lung perfusion to be a safe and reproducible technique [8-10].

Doxorubicin lung levels were determined in a dog model without control of

the bronchial arterial blood flow and with apparently no systemic toxicity.

This can be explained by a recent study which shows that a significant

portion of both pulmonary and metastatic tumour vasculature is fed by the

pulmonary arterial circulation [11]. However, bronchial artery infusion is the

desired route in primary bronchogenic carcinoma, which is mainly fed by the

bronchial circulation [12].

Later on, Weksler described a rat model of unilateral isolated lung perfusion,

which was later modified by Wang and by Hendriks [13-15]. Briefly, in this

model a left thoracotomy is performed after inducing anaesthesia.

Subsequently, the rib retractor and lung are placed anteriorly and the hilum

is dissected free from the posterior side. After clamping the pulmonary

artery and vein with curved microclips, an introductory needle is placed

through the chest wall. A perfusion catheter is introduced into the chest

through the needle and secured by a 4/0 silk tie after insertion into the

pulmonary artery. Perfusate is delivered through this catheter. In addition, a

pulmonary venotomy is performed and two venous catheters are placed into

Chapter 2

32

the superior and inferior pulmonary vein to collect the venous effluent

(figure 1). Nawata et al. recently described a rat model of bilateral isolated

lung perfusion performing a left and right thoracotomy sequential with a

time interval of one week minimally [16]. Several drugs have already been

tested in the unilateral model and some of them proved to be very effective.

These are summarized in table 2.

Initially doxorubicin was studied and Baciewicz et al. did not observe any

toxicity using this drug [17]. Weksler reported a rat study in which

micrometastases were induced in the left lung at day 0. After seven days, rats

were treated with left isolated lung perfusion using doxorubicin. Animals

were sacrificed after three weeks and eradication of metastases was observed

in the left lungs while the right lungs had massive tumour replacement [18].

Significantly longer survival and even complete remission after six weeks

was observed by Abolhoda after isolated lung perfusion with doxorubicin in

a rodent model of metastatic pulmonary sarcoma [19].

Based on the antitumour effect of TNF-α, Weksler reported that a single

treatment of isolated lung perfusion with TNF-α can be done safely and is

effective in decreasing the number of sarcoma lung metastases using a rat

model [20,21]. Although combining melphalan with TNF-α resulted in

excellent synergistic effects in isolated limb perfusion for the treatment of

melanoma, synergistic effects using a single treatment with TNF-α in

combination with melphalan could not been confirmed by Hendriks in a

model of adenocarcinoma metastases [22].

However, a significant reduction in number of lung nodules was observed

after single treatment with melphalan in a rat model of metastatic sarcoma as

well as a reduction in number of nodules and a significantly longer survival

Chapter 2

33

Figure 1. Experimental setting of a rat model of isolated lung perfusion.“A” shows a reperfusion circuit in which the perfusate is circulatingbetween the perfusate reservoir and the lung.“B” shows a single-pass perfusion system in which the perfusate is suckedtowards a reservoir after passing the lung once.

in a rat model of metastatic adenocarcinoma [22-24]. Pogrebniak et al.

applied TNF-α in a swine model of isolated lung perfusion. They concluded

that the use of TNF-α is feasible when intraoperative systemic serum levels

are monitored to ensure adequate pulmonary isolation [25].

Ng reported significantly less pulmonary nodules after isolated lung

perfusion with a fluorinated pyrimidine (FUDR) using a rat model of

pulmonary metastatic colorectal adenocarcinoma [26]. In vitro assays

described by Schrump showed enhanced paclitaxel-mediated cytotoxicity of

5 to 100 times in cultured cancer cell lines under hyperthermic conditions

Chapter 2

34

compared to normothermic conditions. However, no paclitaxel toxicity was

observed in cultured normal human bronchial epithelial cells after exposure

under normothermic or hyperthermic conditions [27].

Cisplatin proved to be very effective in treatment of pulmonary metastatic

sarcoma with enhanced antitumour effect in combination with digitonin [28].

Isolated lung perfusion can be performed antegradely by way of the

pulmonary artery or retrogradely by way of the pulmonary veins, which

results in drug distribution to the areas perfused by the pulmonary and

bronchial circulation as well. The feasibility and toxicity of retrograde

perfusion with doxorubicin was studied in a rat model and compared to

antegrade perfusion [29]. However, no significant difference was measured

between retrograde and antegrade isolated left lung perfusion regarding the

uptake of doxorubicin in lung tissue [29]. No significant differences in

weight of the rats were observed during the follow-up period. Furthermore,

feasibility of hyperthermic retrograde isolated lung perfusion was

demonstrated with paclitaxel in a sheep model [28].

Rickaby et al. investigated the influence of hyperthermia in a dog model of

isolated lung perfusion. Temperatures up to 44.4º Celsius did not have any

detectable influence on the measured variables [30].

Recent studies in our laboratory showed cell kill of adenocarcinoma cells

due to gemcitabine to be time and concentration-dependent [31]. This

finding is the basic argument for treatment of lung metastases with isolated

lung perfusion to deliver high local drug levels without severe systemic

toxicity. Isolated lung perfusion resulted in significantly higher gemcitabine

lung levels compared to systemic serum levels after isolated lung perfusion

using the maximum tolerated dose of 320 mg/kg gemcitabine in a rat model

of metastatic pulmonary adenocarcinoma [31]. Gemcitabine also resulted in

Chapter 2

35

significantly longer survival in this rat model [32]. In contrast to melphalan,

which is dose-dependent, gemcitabine proved not to be dose-dependent

within the studied flow range [33-35]. Further research is necessary to find

out gemcitabine to be concentration- and/or flow-dependent. If a drug is

concentration-dependent recalculation of the amount of the drug to be

administered to other species is less complicated when a single bolus

perfusion is applied in which the concentration of the drug delivered is

stable.

Author Product Model Species Dose Time

Efficacy

Decrease number

nodules (p) Survival

(p)

Abolhoda [19] Doxorubicin sarcoma Rodent 6.4 mg/kg 15 min <0.0001***

Weksler [18] Doxorubicin sarcoma Rat 6.4 mg/kg 10 min <0.001

Weksler [20] TNF-α sarcoma Rat 1.7 mg/kg 10 min 0.01**

Hendriks [22] TNF-α + melphalan carcinoma Rat 2 mg/kg + 1.2 mg/kg 25 min <0.0001*

Hendriks [22] TNF-α carcinoma Rat 300 µg 25 min NS**

Nawata [24] Melphalan sarcoma Rat 8 mg/kg 20 min 0.01*

Hendriks [23] Melphalan carcinoma Rat 2 mg/kg 25 min <0.0001** 0.0002***

Ng [26] FUDR carcinoma Rat 1400 mg/kg 20 min <0.05*

Van Putte [32] Gemcitabine carcinoma Rat 320 mg/kg 25 min 0.02***

Tanaka [28] Cisplatin sarcoma Rat 2 mg/kg 10 min NS***

Tanaka [28] Cisplatin + digitonin sarcoma Rat 2 mg/kg + 1.6 µmol/kg 10 min <0.0001***

Table 2. Animal studies investigating efficacy of isolated lung perfusion forthe treatment of pulmonary metastases* p-value compared to iv treatment** p-value compared to the untreated right lung*** p-value compared to controls

Chapter 2

36

Less invasive models for the treatment of pulmonary metastases

In preparation of less invasive methods of selective perfusion of the lung, we

investigated the first-pass effect of gemcitabine during low flow pulmonary

artery perfusion (0.2 ml/min) using blood flow occlusion of the pulmonary

artery without control of the pulmonary veins. Low flow pulmonary artery

perfusion with 16 mg (64 mg/kg) of gemcitabine resulted in significantly

lower systemic gemcitabine serum levels and significantly higher lung levels

compared to intravenous infusion of 40 mg (160 mg/kg) of gemcitabine [34].

This emphasizes the feasibility of chemotherapy infusion of gemcitabine by

sequential selective catheterisation of the pulmonary artery using blood flow

occlusion without control of the pulmonary veins.

Schneider reported a rat model of chemoembolization of solitary metastasis

in the lung [36]. Efficacy was comparable to isolated lung perfusion but

chemoembolization seems advantageous because it only requires injection of

degradable starch microspheres (DSM) together with a cytotoxic drug into

the pulmonary artery by an endovascular catheter.

Ex vivo model of isolated lung perfusion

Beside this extensive research in the rat model, Linder described an ex-vivo

model in which human lungs can be perfused by isolated lung perfusion in

order to study physiology, pharmacokinetics and different surgical

procedures [37]. The lung specimens used are obtained after resection in

cases of bronchial carcinoma. Reproducibility of the experiments was

evaluated by the net weight gain, wet-to-dry ratio, angiography of the

pulmonary artery, pulmonary vascular resistance, fluorescence of the lung

surface and alveolar gas diffusion into the perfusate.

The uptake of doxorubicin was investigated in this model and compared with

Chapter 2

37

the uptake of a non-toxic glucuronide prodrug (HMR 1826) from which

doxorubicin is released by the action of beta-glucuronidase present in high

levels in many tumours. Seven-fold higher lung levels of doxorubicin were

measured in the lung cancer cells after perfusion with the non-toxic prodrug

compared to perfusion with doxorubicin itself [38]. Another study showed

ten-times higher levels of doxorubicin in the lung tissue than in the tumour

tissue while perfusion with the prodrug of doxorubicin resulted in equal

levels in the tumour and lung tissue as well [39]. Furthermore, ex vivo study

investigating the uptake of gemcitabine and melphalan into the human has

started as a result of the successful application of both drugs in the rat model

[22,32,35].

Human studies of isolated lung perfusion (table 3)

Although no animal models or ex vivo models can exactly define human

tolerance to any drug, four clinical phase I trials have been performed using

tumour necrosis factor, platinum and doxorubicin respectively [40-43].

Except for one study by Ratto et al., all patients had inoperable pulmonary

metastatic disease and perfusions were performed in the antegrade way.

However, in contrast to all rat studies using the single-pass perfusion

technique, a reperfusion circuit was applied in these four studies. Because of

the continuous uptake of the drug into the lung, the concentration in the

circuit is not constant while the concentration delivered is stable in a single-

pass perfusion system. However, previous animal studies in our laboratory

did not show significant differences in final lung and tissue melphalan levels

between single-pass perfusion and reperfusion [35].

All human studies concluded that the human model of isolated lung

perfusion is feasible and results in higher lung and tumour levels while

Chapter 2

38

minimizing systemic toxicity. Although efficacy was not the primary aim of

these studies and no study was conducted as a phase II trial, responses were

disappointingly low. Based on results from animal models and on the

maximally tolerated dose in isolated limb perfusion, specific doses were

determined. However, large variations in final lung and effluent levels were

measured. Therefore, effective drug levels might probably not be reached in

most of the patients with these agents. As noted before, melphalan is one of

the most successful drugs tested in sarcoma and adenocarcinoma animal

models. Currently, a phase I trial is ongoing in our hospital in cooperation

with the St. Antonius hospital (Nieuwegein, the Netherlands). Increasing

doses of melphalan are administered using isolated lung perfusion in patients

undergoing resection of lung metastases, which are melphalan sensitive.

Until February 2003, 11 patients have been treated in this trial without major

morbidity or mortality

Table 3. Phase I trials on isolated lung perfusion for the treatment ofpulmonary metastases (NA: not available)

Author Product Population n Dose Time Flow

Temper

ature of

the lung

Lung

levels

Toxicity

Burt (41) doxorubicin Unresectable 8 40-80

mg/m2

20 min 300-500

mL/min

37°C 1.3-57.3

µg/g

Substantial lung injury

above 40 mg/m 2

Pass (40) TNF-α Unresectable 16 0.3-6 mg 90 min Pressure-

controlled

38-

39.5°C

NA Only one patient had

evidence of systemic

toxicity

Ratto (42) platinum Resectable 6 200

mg/m2

60 min 200-280

mL/min

37-

37.5°C

NA No severe systemic and

pulmonary toxicity

Putnam (43) doxorubicin Unresectable 16 60-75

mg/m2

30 min NA 37°C 74-2750

µg/g

Operative mortality 18.8 %

Early morbidity 23 %

Ongoing

(Antwerp +

Nieuwegein)

melphalan Resectable >20 15-60

mg

30 min 1000

mL/min

37 and

42°C

NA No systemic and pulmonary

toxicity up to 60mg

Chapter 2

39

References1. Jemel A, Thomas A, Murray T, Thun M. Cancer statistics. Ca Cancer J Clin,

2002;52:23-47.

2. Koehne C-H, Cunningham D, Di Costanzo, Glimelius B, Blijham G, Aranda E,

Scheithauer W, Rougier P, Palmer M, Wils J, Baron B, Pignatti F, Schoeffski P, Micheel

S, Hecker H. Clinical determinants of survival in patients with 5-fluorouracil-based

treatment for metastatic colorectal cancer: results of a multivariate analysis of 3825

patients. Ann Oncol 2002;10:308-17.

3. Pastorino U, Buyse M, Friedel G, Ginsberg RJ, Girard P, Goldstraw P, Johnston M,

McCormack P, Pass H, Putnam JB. Long-term results of lung metastasectomy:

prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37-49.

4. Hendriks J, Van Schil P. Isolated lung perfusion for the treatment of pulmonary

metastases . Surg Oncol 1999;7:59-63.

5. Creech O, Krementz ET, Ryan RF, Winblad JN. Chemotherapy of cancer: regional

perfusion utilizing an extracorporeal circuit. Ann Surg 1959;148:616-32.

6. Pierpont H, Blades B. Lung perfusion with chemotherapeutic agents. J Thorac

Cardiovasc Surg 1960;39:159-65.

7. Jacobs JK, Flexner JM, Scott HW Jr. Selective isolated lung perfusion of the right lung. J

Thorac Cardiovasc Surg 1961;42:546-52.

8. Johnston MR, Minchin R, Shull JH, et al. Isolated lung perfusion with adriamycin. A

preclinical study. Cancer 1983;52:404-9.

9. Johnston MR, Christensen CW, Minchin RF, et al. Isolated total lung perfusion as a

means to deliver organ-specific chemotherapy: Long-term studies in animals. Surgery

1985:35-44.

10. Johnston MR, Minchin R, Dawson CA. Lung perfusion with chemotherapy in patients

with unresectable metastatic sarcoma to the lung or diffuse bronchioloalveolar

carcinoma. J Thorac Cardiovasc Surg 1995;110:368-73.

11. Mochizuki A, Kurihara Y, Yokote K, Nakajima Y, Osada H. Discrimination of solitary

pulmonary nodules based on vascular supply patterns with first-pass dynamic CT. Lung

cancer 2000;29(suppl 1):242.

12. Miller BJ, Rosenbaum AS. The vascular supply to metastatic tumours of the lung.

Surgery, Gynaecology and Obstetrics 1967;1009-12.

13. Weksler B, Schneider A, Ng B, Burt M. Isolated lung perfusion in the rat: an

Chapter 2

40

experimental model. J Appl Physiol 1993;74:2739-9.

14. Wang HY, Port JL, Hochwald SN, Burt ME. Revised technique of isolated lung

perfusion in the rat. Ann Thorac Surg 1995;60:211-2.

15. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

16. Nawata S, Abolhoda A, Ross HM, Brooks A, Burt ME. Sequential bilateral isolated lung

perfusion in the rat: an experimental model. Ann Thorac Surg 1997;63:796-9.

17. Baciewicz FA, Arredondo M, Chaudhuri B, Crist KA, Basilius D, Bandyopadhyah S,

Thomford NR, Chaudhuri PK. Pharmacokinetics and toxicity of isolated lung perfusion

with doxorubicin. J Surg Res 1991;50:124-8.

18. Weksler B, Lenert J, Ng B, Burt M. Isolated single lung perfusion with doxorubicin is

effective in eradicating soft tissue sarcoma lung metastases in a rat model. J Thorac

Cardiovasc Surg 1994;107:50-4.

19. Abolhoda A, Brooks A, Nawata S, Kaneda Y, Cheng H, Burt ME. Isolated lung

perfusion with doxorubicin prolongs survival in a rodent model of pulmonary

metastases. Ann Thorac Surg 1997;64:181-4.

20. Weksler B, Blumberg D, Lenert JT, Ng B, Fong Y, Burt ME. Isolated single-lung

perfusion with TNF-α in a rat sarcoma lung metastases model. Ann Thorac Surg

1994;58:328-32.

21. Vassalli P. The pathophysiology of tumour necrosis factor. Annu Rev Immunol

1992;10:411-52.

22. Hendriks JMH, Van Schil PEY, De Boeck G, Van Oosterom A, Van Marck E, Eyskens

EJM. Isolated lung perfusion with melphalan and tumour necrosis factor for pulmonary

adenocarcinoma. Ann Thorac Surg 1998;66:1719-25.

23. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-71.

24. Nawata S, Abecasis N, Ross HM, Abolhoda A, Cheng H, Sachar KS, Burt ME. Isolated

lung perfusion with melphalan for the treatment of metastatic pulmonary sarcoma. J

Thorac Cardiovasc Surg 1996;112:1542-8.

25. Pogrebniak HW, Witt C, Terrill R, Kranda K, Travis WD, Rosenberg SA, Pass HI.

Isolated lung perfusion with tumour necrosis factor: a swine model in preparation of

human trials. Ann Thorac Surg 1994;57:1477-83.

26. Ng B, Lenert JT, Weksler B, Port JL, Ellis JL, Burt ME. Isolated lung perfusion with

Chapter 2

41

FUDR is an effective treatment for colorectal adenocarcinoma lung metastases in rats.

Ann Thorac Surg 1995;59:205-8.

27. Schrump DS, Zhai S, Nguyen DM, Weiser TS, Fisher BA, Terrill RE, Flynn BM, Duray

PH, Figg WD. Pharmacokinetics of paclitaxel administered by hyperthermic retrograde

isolated lung perfusion techniques. J Thorac Cardiovasc Surg 2002;123:686-694.

28. Tanaka T, Kaneda Y, Li T, Matsuoka T, Zempo N, Esato K. Digitonin enhances the

antitumour effect of cisplatin during isolated lung perfusion. Ann Thorac Surg

2001;72:1173-8.

29. Krueger T, Kümmerle A, Vallet C, Pan Y, Decosterd L, Ris HB. Antegrade and

retrograde isolated lung perfusion (ILP) with doxorubicin in the rat: comparison of

pharmacokinetics and toxicity. Eur Resp J 2002;20(suppl38):179s.

30. Rickaby DA, Fehring JF, Johnston MR, Dawson CA. Tolerance of the isolated lung to

hyperthermia. J Thorac Cardiovasc Surg 1991;101:732-9.

31. Van Putte BP, Hendriks JMH, Romijn S, Pauwels B, Guetens G, De Bruijn E, Van Schil

PEY. Isolated lung perfusion with gemcitabine: pharmacokinetics and survival. J Surg

Res 2003;109;118-22.

32. Van Putte BP, Hendriks JMH, Vermeulen PB, Romijn S, Van Marck E, Van Schil PEY.

Isolated lung perfusion with gemcitabine prolongs survival in a rat model of metastatic

pulmonary adenocarcinoma. Submitted.

33. Romijn S, Hendriks JMH, Van Putte BP, Guetens G, Van Schil PEY. Variations in flow,

duration and concentration do not change the final lung concentrations of melphalan

after isolated lung perfusion in rats. Eur Surg Res 2003;35:50-3.

34. Van Putte BP, Hendriks JMH, Pauwels B, Romijn S, Guetens G, De Bruijn E and Van

Schil PEY. Pharmacokinetics of gemcitabine in a rat model of metastatic pulmonary

adenocarcinoma. Ann Thorac Surg 2003, in press.

35. Van Putte BP, Hendriks JMH, Romijn S, Guetens G, De Boeck, De Bruijn E, Van Schil

PEY. Single-pass isolated lung perfusion versus recirculating isolated lung perfusion

with melphalan in a rat model. Ann Thorac Surg 2002;74:893-8.

36. Schneider P, Kampfer S, Loddenkemper C, Foitzik T, Buhr HJ. Chemoembolization of

the lung improves tumour control in a rat model. Clin Cancer Res 2002;8:2463-2468.

37. Linder A, Friedel G, Fritz P, Kivisto KT, McClellan M, Toomes H. The ex-vivo isolated,

perfused human lung model: description and potential applications. Thorac Cardiovasc

Surg 1996;44:140-6.

38. Murdter TE, Sperker B, Kivisto KT, McClellan M, Fritz P, Friedel G, Linder A, Bosslet

Chapter 2

42

K, Toomes H, Dierkesmann R, Kroemer HK. Enhanced uptake of doxorubicin into

bronchial carcinoma: beta-glucoronidase mediates release of doxorubicin from a

glucuronide prodrug (HMR 1826) at the tumour site. Cancer Res 1997;57:2440-5.

39. Murdter TE, Linder A, Friedel G, McClellan M, Bohnenstengel F, Sperker B, Kroemer

HK, Toomes H, Freitag L, Fritz P. Pharmacokinetics of cyclophosphamide, adriamycine

and adriamycine prodrug (HMR 1928) using an ex vivo isolated perfused human lung

model (IHLP). Pneumologie 2000;54:494-8.

40. Pass HI, Mew DJY, Kranda KC, Tmeck BT, Donington JS, Rosenberg SA. Isolated lung

perfusion with tumour necrosis factor for pulmonary metastases. Ann Thorac Surg

1996;61:1609-17.

41. Burt ME, Liu D, Abolhoda A, Ross HM, Kaneda Y, Jara E, Casper ES, Ginsberg RJ,

Brennan MF. Isolated lung perfusion for patients with unresectable metastases from

sarcoma: a phase I trial. Ann Thorac Surg 2000;69:1542-9.

42. Ratto GB, Toma S, Civalleri D, Passerone GC, Esposito M, Zaccheo D, Canepa M,

Romano P, Palumbo R, De Cian F, Scarano F, Vannozzi M, Spessa E, Fantino G.

Isolated lung perfusion with platinum in the treatment of pulmonary metastases from soft

tissue sarcomas. J Thorac Cardiovasc Surg 1996;112:614-22.

43. Putnam JB. New and evolving treatment methods for pulmonary metastases. Sem Thorac

Cardiovasc Surg 2002;14:49-56.

Chapter 3

Chapter 3

Isolated lung perfusion with gemcitabine in a rat:

pharmacokinetics and survival

BP Van Putte1, JMH Hendriks1, S Romijn1, B Pauwels2,G Friedel4,

G Guetens 3, EA De Bruijn3and PEY Van Schil1

Department of Thoracic and Vascular Surgery1 and Department of

Medical Oncology2, University Hospital Antwerp, Edegem, Belgium;

Department of Oncology and Radiotherapy, Catholic University of

Leuven, Leuven, Belgium3 ; Department of Thoracic Surgery, Klinik

Schillerhoehe, Gerlingen, Germany4

This chapter has been published in J Surg Res 2003;109:118-22.

Presented at the Research Seminar on isolated lung perfusion of theEuropean Respiratory Society, April 26-28, 2002, Antwerp (Belgium);presented at the Annual Congress of the European Society for SurgicalResearch, May 22-25, 2002, Szeged,(Hungary); presented at the FirstInternational CC531S Meeting, June 14-15, 2002, Maastricht (TheNetherlands); presented at the Annual Congress of the RespiratorySociety, September 14-18, 2002, Stockholm (Sweden).

Chapter 3

46

Abstract

Background: Toxicity and pharmacokinetics of gemcitabine (GCB) were

evaluated in a rat model of ILuP and compared to intravenous (iv) infusion.

Materials and methods: CC531S adenocarcinoma cells were incubated in

vitro for 24 hours with GCB. Cell survival was determined 4 days after GCB

treatment with the sulphorhodamine B test. In a first in vivo experiment,

Wag/Rij rats underwent left ILuP with 20mg/kg (n=3), 40mg/kg (n=6),

80mg/kg (n=6), 160mg/kg (n=6), 320mg/kg (n=6) of GCB and a control

group (n=6) with buffered starch. After 3 weeks, right pneumonectomy was

performed. Furthermore, survival was determined for rats treated with iv

infusion of 40mg/kg (n=10), 80mg/kg (n=10), 160mg/kg (n=10), 320mg/kg

(n=6) of GCB and a control group (n=6) with saline (0.9% NaCl). In a second

experiment lung and serum GCB levels were determined for rats treated with

iv infusion (160mg/kg, n=6) and rats which had ILuP (160mg/kg, n=6; 320

mg/kg, n=6).

Results: Incubation of the CC531S adenocarcinoma cells with GCB led to a

50% decrease (p<0.05) in the number of cells compared to controls at a dose

of 23.6 nanoM. After 90 days, the mortality for rats treated with 320mg/kg iv

GCB was 100% compared to 17% after ILuP for the same dose. ILuP with

160mg/kg and 320mg/kg resulted in significantly higher lung levels of GCB

compared to iv therapy without any systemic leakage.

Conclusions: GCB ILuP is well tolerated to a maximum dose of 320 mg/kg

and results in significantly higher GCB lung levels with undetectable serum

levels compared to iv treatment.

Chapter 3

47

Introduction

Colorectal adenocarcinoma remains a serious health problem with 55 000

deaths in 1996 in the United States, and it accounts for a significant number

of isolated lung metastases [1].

Current treatment after isolated pulmonary metastases consists of surgical

resection and systemic chemotherapy. Five-year survival of resection is only

30% to 40% [2,3] and many patients become inoperable due to local

recurrences [4]. Systemic chemotherapy is dose- limited because of high

systemic toxicity and severe side effects. In order to prevent side effects and

to obtain higher local concentrations, Wang et al. developed the technique of

isolated lung perfusion (ILuP) as a new approach for treatment of pulmonary

metastases [5]. Several chemotherapeutics against colorectal adenocarcinoma

have already been tested. Ng et al. reported a significant reduction of the

number of pulmonary nodules in animals treated with ILuP high-dose FUDR

compared to intravenously (iv) treated animals [6]. Hendriks et al. observed a

significant reduction of pulmonary nodules with melphalan and a partial

response to TNF [7]. They also found a significantly longer survival of rats

treated with melphalan and two rats even had tumour-free lungs at sacrifice

[8]. Some recent studies show gemcitabine (GCB) to be active in colorectal

cancer [9,10]. This study was designed to determine sensitivity of CC531S

adenocarcinoma to GCB and to evaluate toxicity and pharmacokinetics of

GCB in a rat model of ILuP compared to iv infusion.

Material and methods

Cell line

In this study CC531S adenocarcinoma cell line was used to determine the

sensitivity for the cytotoxic effect of GCB.

Chapter 3

48

CC531S was cultured in RPMI-1640 medium, supplemented with 10% fetal

calf serum (Life Technologies, Merelbeke, Belgium). Cultures were

maintained in exponential growth in a humidified atmosphere at 37°C under

5% CO2/95% air.

Cell survival after treatment with gemcitabine

Cells were harvested from exponential phase cultures by trypsinisation,

counted and plated in 48-wells plates. In order to assure exponential growth

during the experiments seeding density was 1000 cells per well. Following

plating and a 24-hour recovery period, cells were treated with GCB (0-200

nM) dissolved in phosphate buffered saline (PBS) during 24 hours. PBS was

added to control cells. Each concentration was tested in sextuple within the

same experiment. After 24-hour incubation with GCB, cells were washed

with drug free medium. After four days, survival was determined by the

sulphorhodamine B (SRB) assay.

The SRB assay was performed according to the method of Skehan and

Papazisis with minor modifications. Culture medium was aspirated prior to

fixation of the cells by addition of 200 µL 10% cold trichloroactic acid. After

1 hr incubation at 4°C, cells were washed 5 times with deionized water. Then

the cells were stained with 200 µL 0.1% SRB (ICN, Asse, Belgium)

dissolved in 1% acetic acid for at least 15 minutes and subsequently washed

4 times with 1% acetic acid to remove unbound stain. The plates were left to

dry at room temperature and bound protein stain was solubilized with 200 µL

10 mM unbuffered TRIS base (tris(hydroxymethyl)aminomethane) and

transferred to 96 wells plates for reading the optical density at 540 nm

(Biorad 550 microplate reader, Nazareth, Belgium).

Chapter 3

49

The survival rates (%) were calculated by: mean OD (optical density) of

treated cells / mean OD of control cells x 100. The dose response curves were

fitted to the sigmoid inhibition model: E(survival)=1-(Cγ / Cγ+IC50γ)

whereby C is the concentration of the drug and γ is a constant, using

Winnonlin (Pharsight, USA) to calculate IC50 values, i.e. the concentration

GCB causing 50% growth inhibition.

Animals

Male inbred Wag-Rij strain rats (weight, approximately 250g), obtained from

Harlan-CPB (Zeist, The Netherlands), were used for all experiments.

Animals were treated in accordance with the Animal Welfare Act and the

“Guide for the Care and Use of Laboratory Animals” (NIH Publication 86-

23, revised 1985). The rats were transported in sterile conditions, housed in

suspended mesh wired cages under standard laboratory conditions and ad

libitum fed a standard pellet diet (standard rat chow, Hope Farms, Woerden,

The Netherlands). The experimental protocols were approved by the

Institutional Animal Care Use Committee, University of Antwerp.

Left Isolated Lung Perfusion

Isolated left lung single-pass perfusion was performed according to the

technique described by Wang et al. [5]. This was modified in our laboratory

and full details have been reported previously [13]. Briefly, rats were

anaesthetized with isoflurane in a mixture of nitrous oxide (NO2) and oxygen

(O2). Isoflurane was administered in a concentration of 4% and the NO2:O2

ratio was 3:1. After 5 minutes, rats were intubated with a 16-gauge Insyte-W

catheter by translaryngeal illumination and afterwards ventilated with a

volume-controlled ventilator (Harvard Rodent Ventilator, South Natick, MA)

Chapter 3

50

37°C

ventilator

Roller pump

O2 +N2O

Isoflurane

Warmth padlung

Perfusate recipient

suction

single-pass perfusion circuit

Water reservoirperfusate

[13]. Once connected to the ventilator, the NO2:O2 ratio was 1:1 (0.5 L/min)

and the rate of ventilation was 75 strokes/min with a tidal volume of 10

mL/kg. Subsequently, the left chest was shaved and prepared with a 70%

alcohol solution, and a left thoracotomy between the third and fourth

intercostal space was performed. After placing a rib retractor, the left lung

was luxated anteriorly and the hilum was dissected free. The pulmonary

artery and vein were clamped with curved microclips. A PE-10 perfusion

catheter (Becton Dickinson, Bornem, Belgium) was introduced into the chest

through a 16-G Angiocath, which was placed through the chest wall. The PE-

10 catheter was inserted into the pulmonary artery and fixed with a 4/0 silk

tie. Perfusate was delivered through this catheter. The effluent was collected

at the venous side by a catheter in proximity of the venotomy (figure 1).

Figure 1.

Experimental setting

Chapter 3

51

For the pharmacokinetic studies, two catheters (PE-90) were placed in the

superior and inferior pulmonary vein in order to collect the effluent [14].

In all experiments, rats were perfused during 25 minutes followed by a 5-

minute washout with buffered starch at a rate of 0.5 mL/min. In addition,

blood samples were taken out of the left ventricle and subsequently

centrifuged for 20 minutes and frozen in liquid N2 for later analysis.

GCB (Gemcitabine, Eli Lilly Benelux) solutions were prepared by

reconstituting lyophilized powder in the supplied diluent and performing

appropriate dilutions with buffered hydroxy-ethyl starch (Haes Steril, BHE).

GCB in lung and tumour tissue was measured by gas chromatography-mass

spectrometry according to the method described by De Boeck et al. [15].

Statistical analysis

MN levels were analyzed by Student´s t-test. Significance was defined as

p<0.05.

Experiment 1: In vitro toxicity of gemcitabine in CC531S

adenocarcinoma

CC531S cells were treated with different concentrations of GCB during 24

hours. PBS was added to control cells. Each concentration was tested in

sextuple within the same experiment. After 24-hour incubation with GCB,

cells were washed with drug free medium. After four days, survival was

measured by the sulphorhodamine B (SRB) assay. This experiment was

performed three times.

Chapter 3

52

Experiment 2: In vivo toxicity of gemcitabine in rats

Seventy-five rats were randomized into eleven groups. Groups 1 to 5 had left

ILuP with 20 mg/kg (n=3), 40 mg/kg (n=6), 80 mg/kg (n=6), 160 mg/kg

(n=6) and 320 mg/kg (n=6) of GCB respectively. Group 6 (n=6) had left

ILuP with buffered starch. Rats were perfused during 25 minutes followed by

a 5-minute washout. After three weeks a right pneumonectomy was

performed. Groups 7 to 10 were infused iv with 40 mg/kg (n=10), 80 mg/kg

(n=10), 160 mg/kg (n=10) and 320 mg/kg (n=6) of GCB respectively. Group

11 (n=6) was infused intravenously with saline (0.9 % NaCl). All animals

were sacrificed after 90 days. Weight of the animals was followed during the

whole study period.

Experiment 3: Pharmacokinetics of gemcitabine

Eighteen rats were randomized into three groups (n=6, each). Group 1and 2

had left ILuP with 160 mg/kg and 320 mg/kg of GCB respectively. Rats were

perfused during 25 minutes followed by a 5-minute washout. Effluent was

collected every three minutes during perfusion and every two minutes during

washout and subsequently frozen in liquid N2 for later analysis. After

washout, a blood sample was taken which was frozen after centrifugation

during 20 minutes. Subsequently, the left lung was removed and frozen in

liquid N2 for later analysis. Group 3 was infused iv with 160 mg/kg of GCB.

After 30 minutes the left lung was removed and frozen in liquid N2. Also a

blood sample was taken and after 20-minute centrifugation frozen in liquid

N2.

Chapter 3

53

CC531s

0,0

50,0

100,0

150,0

0,1 1 10 100 1000conc GMC (nM)

% s

urv

ival

Results

Mortality

Procedure related survival was 100% for all protocols.

Experiment 1: In vitro toxicity of gemcitabine in CC531S adenocarcinoma

The dose response curve of CC531S cells after treatment with GCB (0-200

nM) during 24 hours is shown in figure 2. The IC50 is 23.6 ± 1.1 nM.

Figure 2.

Dose response curve ofCC531S cells after treatmentwith gemcitabine

Experiment 2: In vivo toxicity of gemcitabine in rats

Survival rate before right contralateral pneumonectomy was 100% in group 1

to 4 and in the control group and 83% in group 5. One rat in group 1 (ILuP

20 mg/kg) died eight days after right contralateral pneumonectomy and one

rat in group 4 (ILuP 160 mg/kg) died immediately after right

pneumonectomy (table 1). There was no difference in mean weight of the

different ILuP groups before pneumonectomy. However, a significant

difference (p=0.03) was found in mean weight between ILuP 20 mg/kg and

ILuP 80 mg/kg after pneumonectomy (figure 3). A significant difference

(p<0.0001) was found in mean weight between iv 40 mg/kg and iv 320

mg/kg (figure 4). Survival rate was 100% in group 7 (iv 40 mg/kg) and the

Chapter 3

54

mean weight iv groups

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

day0

day2

day4

day6

day8

day10

day12

day14

day16

day18

day20

day22

day24

day26

day28

week6

week8

week1

0

week1

2

time

mea

n r

elat

ive

wei

gh

t

mean iv40

mean iv320

control group (saline), 70% in group 8 (iv 80 mg/kg), 80% in group 8 (iv 160

mg/kg) and 0% in group 10 (iv 320 mg/kg) (figure 5).

Table 1. Survival (%) ILuP groups

Figure 3.

Weight of ILuPgroups

Figure 4.

Weight ivgroups

group postoperatively postpneumonectomy after 90 days

ILuP control

ILuP 20 mg/kg

ILuP 40 mg/kg

ILuP 80 mg/kg

ILuP 160 mg/kg

ILuP 320 mg/kg

100%

100%

100%

100%

100%

83%

100%

67%

100%

100%

83%

83%

100%

67%

100%

100%

83%

83%

mean weight ILuP groups

0

0,5

1

1,5

2

day0

day3

day6

day9

day1

2

day1

5

day1

8

day2

1

day2

4

day2

7

day3

0

day3

3

wee

k6

wee

k9

wee

k12

t ime

mea

n re

lativ

e w

eigh

t

mean ilup20

mean ilup80

mean ilup320

ILuPpneumonectomy

Chapter 3

55

survival iv groups

0

20

40

60

80

100

120

day0

day4

day8

day12

day16

day20

day24

day28

week8

week1

2

time

rela

tive

surv

ival

(%

)

iv40iv80iv160iv320

Figure 5.

Survival iv groups

Experiment 3: Pharmacokinetics of gemcitabine

Left ILuP with 160 and 320 mg/kg of GCB resulted in significantly higher

left lung GCB levels compared to rats treated with iv 160 mg/kg of GCB. No

GCB was detected in systemic circulation in the ILuP groups (table 2).

Figure 6 and 7 show GCB effluent levels in function of perfusion time. A

plateau was reached after seven minutes.

lung levels

levels (µg/g)

p serum

(µg/mL)

ILuP 160 mg/kg

ILuP 320 mg/kg

iv 160 mg/kg

1.5 ± 1.6

2.5 ± 1.8

0.2 ± 0.1

p=0.045¹

p=0.14²

p=0.0143

0

0

92.2±63.6

Table 2. Gemcitabine lung and serum levels¹ ILuP 160 mg/kg compared to iv 160 mg/kg; ²ILuP 320 mg/kg compared toILuP 160 mg/kg; 3 ILuP 320 mg/kg compared to iv 160 mg/kg; ILuP: isolatedlung perfusion; iv: intravenous.

Chapter 3

56

gemcitabine effluent (ILuP 160 mg/kg)

01000

20003000

40005000

0 3 6 7 12 15 18 21 24 27 30 32 34

time (minutes)

gem

cita

bin

ef

flu

ens

(mic

rog/

mL)

gemcitabine effluent (ILuP 320 mg/kg)

-20000

2000400060008000

100001200014000

0 3 6 9 12 15 18 21 24 27 30 32 34 36

time (minutes)

gem

ctab

in e

fflu

ens

(mic

rog

/mL

)Figure 6.

Effluent ILuP with 160 mg/kg

Figure 7.

Effluent ILuP with 320mg/kg

Discussion

Colorectal adenocarcinoma gives rise to a large number of pulmonary

metastases and therefore, it remains a serious health problem [1]. In case of

lung metastases no effective therapy is available resulting in long-term

survival. Many patients develop local recurrences after pulmonary

metastasectomy probably due to micrometastases present at the time of initial

therapy [1,16]. Systemic chemotherapy is dose-limited due to its severe side

effects. Isolated lung perfusion as a novel experimental technique is

promising as it can deliver a high local concentration without systemic side

effects. A rat model for pharmacokinetic and efficacy studies was developed

by Weksler [17] and modified in our laboratory in order to perform

experiments of ILuP in the Wag/Rij rat [13]. Several drugs like FUDR and

melphalan/TNF have already been tested with good results [6,7]. Recently, a

cytotoxic effect of GCB against HT29 colon carcinoma cells was observed

Chapter 3

57

by Tonkinson making it a promising agent for further studies of ILuP [9].

Our in vitro results show a higher sensitivity of CC531S adenocarcinoma

(IC50=23.6±1.1 nM) compared to HT29 colon carcinoma (IC50= 32.0±9.0

nM) [9]. Furthermore, toxicity of GCB was determined after treatment with

ILuP and iv infusion. Both the control groups of left ILuP and iv infusion had

a survival percentage of 100%. Intravenous infusion of 80 mg/kg and 160

mg/kg was sublethal with 70% and 80% survival respectively. All animals

died after iv infusion of 320mg/kg. However, all animals survived after ILuP

with doses of GCB up to 320 mg/kg which was the highest dose we could

administer because of maximal solubility of GCB. Only a significant

reduction in mean weight between group ILuP 20 mg/kg and ILuP 80 mg/kg

after pneumonectomy at three weeks is demonstrated. In this way, it has been

proven that treatment with GCB by ILuP is much less toxic compared to iv

infusion. Pharmacokinetic studies show significantly higher GCB lung levels

reached by ILuP with 160 and 320 mg/kg compared to iv infusion with 160

mg/kg. The pharmacokinetic benefit of ILuP has been confirmed and is in

agreement with results obtained before [7, 18, 19].

Further studies will be performed to study efficacy of gemcitabine in an in

vivo model of lung metastases. More specifically, survival will be evaluated

in a rat model of unilateral lung metastases in which gemcitabine is

administered by isolated lung perfusion. In this specific model melphalan and

TNF have already been studied in our laboratory [7].

In conclusion, GCB is an effective drug against colon adenocarcinoma in

vitro. When given by isolated lung perfusion, higher lung levels are reached

and less toxicity is observed. In this way, GCB is a promising agent for use

by ILuP in the treatment of pulmonary metastases from colon

Chapter 3

58

adenocarcinoma. Further experimental and clinical studies are needed to

determine any survival benefit.

Acknowledgements

We wish to thank Marleen Nysten and August Van Laer for their assistance

during all experiments. We thank Eli Lilly Benelux for providing

gemcitabine.

References1. Parker SL, Tong T, Bolden S, Wingo PA. Cancer statistics, 1996. CA Cancer J Clin

1996;46:5-27.

2. Pastorino U, Buyse M, Friedel G, Ginsberg RJ, Girard P, Goldstraw P, Johnston M,

McCormack P, Pass H, Putnam JB, Toomes H. Long-term results of lung

metastasectomy: prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg

1997;113:37-49.

3. Lehnert T, Knaebel HP, Duck M, Bulzebruck H, Herfarth C. Sequential hepatic and

pulmonary resections for metastatic colorectal cancer. Br J Surg 1999;86:241-3.

4. Yano T, Fukuyama Y, Yokoyoma H, Tanaka Y, Miyagi J, Kuninaka S, Asoh H, Ichinose

Y. Failure in resection of multiple pulmonary metastases from colorectal cancer. J Am

Coll Surg 1997;185:120-2.

5. Wang HY, Port JL, Hochwald SN, Burt ME. Revised technique of isolated lung perfusion

in the rat. Ann Thorac Surg 1995;60:211-2.

6. Ng B, Lenert JT, Weksler B, Port JL, Ellis JL, Burt ME. Isolated lung perfusion with

FUDR is an effective treatment for colorectal adenocarcinoma lung metastases in rats.

Ann Thorac Surg1995;59:205-8.

7. Hendriks JMH, Van Schil PEY, De Boeck G, Lauwers PRM, Van Oosterom AAT, Van

Marck EAE, Eyskens EJM. Isolated lung perfusion with melphalan and tumour necrosis

factor for metastatic pulmonary adenocarcinoma. Ann Thorac Surg 1998;66:1719-25.

8. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-271.

Chapter 3

59

9. Tonkinson JL, Worzalla JF, Teng CH, Mendelsohn LG. Cell cycle modulation by a

multitargeted antifolate, LY231514, increases the cytotoxicity and antitumour activity of

gemcitabine in HT29 colon carcinoma. Cancer Res 1999;59:3671-6.

10. Adjei AA, Erlichman C, Sloan JA, Reid JM, Pitot HC, Goldberg RM, Peethambaram P,

Atherton P, Hanson LJ, Alberts SR, Jett J. Phase I and pharmacologic study of sequences

of gemcitabine and the multitargeted antifolate agent in patients with advanced sold

tumours. J Clin Oncol 2000;18:1748-57.

11. Skehan P, Storeng R, Scudiero D, Monks A, McMahon, Vistica D, Warren JT, Bokesch

H, Kenney S, Boyd MR. New colorimetric cytotoxicity assay for anticancer-drug

screening. J Natl Cancer Inst 1990;82:1107-12.

12. Papazisis KT, Geromichalos GD, Dimitriadis KA, Kortsaris AH. Optimization of the

sulphorhodamine B colorimetric assay. J Immunol Methods 1997;208:151-8.

13. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

14. Van Putte BP, Hendriks JMH, Romijn S, Guetens G, De Boeck G, De Bruijn EA, Van

Schil PEY. Single-pass isolated lung perfusion versus recirculating isolated lung

perfusion with melphalan in a rat model. Ann Thorac Surg, submitted for publication.

15. De Boeck G, Van Cauwenberghe K, Eggermont AM, Van Oosterom AT, De Bruijn EA.

Determination of melphalan and hydrolysis products in body fluids by GC-MS. High Res

Chromatog 1997;12:697-700.

16. Hendriks J, Romijn S, Van Putte B, Vermorken J, Van Marck E, Eyskens E, Van Schil P.

Long-term results of surgical resection of lung metastases. Acta Chir Belg 2001:267-72.

17. Weksler B, Schneider A, Ng B, Burt M. Isolated single-lung perfusion in the rat: an

experimental model. J Appl Physiol 1993;74:2736-9 18. Nawata S, Abecasis N, Ross

HM, Abolhoda A, Cheng H, Sachar KS, Burt ME. Isolated Lung perfusion with

melphalan for the treatment of metastatic pulmonary sarcoma. J Thorac Cardiovasc Surg

1996;12:1542-8.

18. Weksler B, Ng B, Lenert JT, Burt ME. Isolated single-lung perfusion with doxorubicin is

pharmacokinetically superior to intravenous injection. Ann Thorac Surg 1993;56:209-14.

Chapter 4

Chapter 4

Pharmacokinetics after Pulmonary Artery Perfusion

with Gemcitabine.

BP Van Putte1, JMH Hendriks1, S Romijn1; B Pauwels2; G De

Boeck3, G Guetens 3, E De Bruijn3 and PEY Van Schil1

1Department of Thoracic and Vascular Surgery and 2Department of

Medical Oncology, University Hospital Antwerp, Edegem, Belgium,3Department of Oncology and Radiotherapy, Catholic University of

Leuven, Belgium

Will be published in: Ann Thorac Surg 2003 (in press)

Presented at the annual congress of the Society of Thoracic Surgeons2003, San Diego, CA, USA, January 30th 2003 and at the springmeeting of the Dutch Society of Cardiothoracic Surgery, UMCUtrecht, Utrecht, April 12th 2003.

Chapter 4

62

Abstract

Background: Isolated lung perfusion (ILuP) proved to be superior for the

treatment of lung metastases compared to intravenous injection (IV).

However, its invasive character inhibits repetitive treatment. Blood flow

occlusion (BFO), as a regional therapy, with gemcitabine (GCB) was

evaluated in a rat model. Lung levels of GCB were examined with different

exposure times and flow rates and compared with ILuP and IV. Cell kill was

studied in vitro.

Methods: In vitro: Survival of CC531 adenocarcinoma cells was determined

after 10, 20 and 40 min of exposure to GCB. In vivo: 48 Wag/Rij rats

underwent BFO with GCB at a rate of 0.2 mL/min and 0.5 mL/min during

10, 20, 30 and 40 min. Statistical analysis was performed using the Student’s

t-test.

Results: In vitro: The dose of GCB resulting in 50% growth inhibition was

9.1 µg/mL, 7.2 µg/mL and 2.2 µg/mL after 10, 20 and 40 min exposure

respectively. In vivo: No significant difference in lung levels of GCB was

observed between a flow rate of 0.2 mL/min compared to 0.5 mL/min at any

exposure time point (p<0.05). Lung tissue was saturated after 20 min. BFO

resulted in a lower plasma levels and higher lung levels of GCB compared to

IV of the MTD of 40mg.

Conclusions: Growth inhibition of CC531 cells in vitro increased with

exposure time, while lung tissue was saturated after 20 minutes of BFO. No

difference in GCB lung levels were seen after BFO compared to ILuP.

Systemic exposure after IV was higher compared to BFO but did not result

in higher lung levels.

Chapter 4

63

Introduction

Isolated lung perfusion is an experimental surgical procedure for the

treatment of pulmonary metastatic disease in order to improve current 5-year

survival of around 40% [1]. The aim of this technique is to achieve high

local lung concentrations without systemic exposure. It can be a treatment

option to make bulky metastatic disease operable or to prevent intrathoracic

recurrences after removal of macroscopic metastatic disease of the lung.

Although the procedure itself is tolerated well [2], it is a highly invasive

procedure unlikely to be performed once again. Therefore, regional infusion

techniques like bronchial artery infusion or pulmonary artery perfusion

without control of the venous circulation are preferred since these

procedures can be repeated in time [3-4]. Only drugs with a good first-pass

effect can be selected since systemic exposure must be limited to a

minimum. The aim of the present study is to evaluate the in-vitro toxicity of

gemcitabine in an adenocarcinoma cell line, and to investigate lung and

serum levels of gemcitabine (GCB) during perfusion of the pulmonary

artery.

Material and methods

Animals

Male inbred WAG-Rij strain rats (weight, approximately 225g), obtained

from Iffa Credo (Brussels, Belgium), were used for all experiments. Animals

were treated in accordance with the Animal Welfare Act and the “Guide for

the Care and Use of Laboratory Animals” (NIH Publication 86-23, revised

1985). The rats were transported in sterile conditions, housed in suspended

mesh wired cages and fed a standard pellet diet ad libitum (standard rat

chow, Hope Farms, Woerden, The Netherlands). The experimental protocols

Chapter 4

64

were approved by the Institutional Animal Care and Use Committee,

University Hospital of Antwerp.

Cell line

The CC531s tumour cell line was used in this study to determine the

sensitivity for the cytotoxic effect of gemcitabine. This cell line was derived

from a chemically-induced adenocarcinoma of the colon of a WAG rat [5].

CC531s was cultured in RPMI-1640 medium, supplemented with 10% fetal

calf serum (Invitrogen, Merelbeke, Belgium). Cultures were maintained in

exponential growth in a humidified atmosphere at 37°C under 5% CO2 and

95% air.

Pulmonary artery perfusion

Pulmonary artery perfusion is performed as an isolated left lung perfusion

procedure without venous control [6]. Briefly, anaesthesia was induced with

isoflurane in a mixture of nitrous oxide (N2O) and oxygen (O2). Isoflurane

was administered in a concentration of 4%; N2O:O2 ratio was 3:1. Next, rats

were intubated by translaryngeal illumination and connected to a volume-

controlled ventilator at a rate of 75 strokes/min and a tidal volume of 10

mL/kg. After a left thoracotomy, the pulmonary artery is clamped with a

curved microclip. A PE-10 perfusion catheter was introduced into the

pulmonary artery. Perfusate was delivered through this catheter at a chosen

rate. At the completion of lung perfusion, the rat was killed by a venous cut

down of the caval vein.

The perfusate was temperature-controlled at 37ºC throughout the whole

perfusion. Rats were placed on a heating pad immediately after induction

and body temperature was kept constantly between 34°C and 37°C.

Chapter 4

65

Gemcitabine and gemcitabine processing and measurement

Gemcitabine (difluorodeoxycytidine, Ely Lilly Benelux, Brussels, Belgium)

perfusate solutions were prepared by reconstituting non-lyophilized powder

in the supplied diluent and performing appropriate dilutions with BHE prior

to the experiments.

Gas chromatography-mass spectrometry method was used for measuring

gemcitabine levels in lung tissue and serum. Deoxycytidine was used as an

internal standard. Samples were extracted over a Chrompack Spherisorb

ODS-2 reversed phase column [7,8].

Experiment 1: Cell survival after treatment with gemcitabine

Cells were harvested from exponential phase cultures by trypsinisation,

counted and plated in 48-wells plates. In order to assure exponential growth

during the experiments seeding density was 1000 cells per well. Following

plating and a 24 hr recovery period, cells were treated with gemcitabine (0,

10, 20 and 40 µM) dissolved in phosphate buffered saline (PBS) during 10,

20 or 40 minutes. PBS was added to control cells. Each concentration was

tested in sextuple within the same experiment. After incubation with

gemcitabine, cells were washed with drug free medium. Four days after

treatment the survival was determined by the sulforhodamine B (SRB) assay.

The SRB assay was performed according to the method of Skehan [9] and

Papazisis [10], with minor modifications. Culture medium was aspirated

prior to fixation of the cells by addition of 200 µl 10% cold trichloroactic

acid. After 1 hr incubation at 4°C, cells were washed 5 times with deionized

water. Then the cells were stained with 200 µl 0.1% SRB (ICN, Asse,

Belgium) dissolved in 1% acetic acid for at least 15 minutes and

subsequently washed 4 times with 1% acetic acid to remove unbound stain.

Chapter 4

66

The plates were left to dry at room temperature and bound protein stain was

solubilized with 200 µl 10 mM unbuffered TRIS base

(tris(hydroxymethyl)aminomethane) and transferred to 96 wells plates for

reading the optical density at 540 nm (Biorad 550 microplate reader,

Nazareth, Belgium).

The survival rates were calculated by: mean OD (optical density) of treated

cells / mean OD of control cells x 100%. The dose response curves were

fitted to the sigmoid inhibition model: E(survival)=1-(Cγ / Cγ+IC50γ), using

Winnonlin (Pharsight, USA) to calculate IC50 values, the concentration

gemcitabine causing 50% growth inhibition.

Experiments were performed at least three times. All data are presented as

the mean ± standard deviation.

Experiment 2

Forty-eight Wag/Rij rats underwent pulmonary artery perfusion. These rats

were randomized into eight groups of 6 rats each. Gemcitabine (GCB) was

given at a concentration of 2.7 mg/mL during 10, 20, 30 and 40 minutes

respectively. Half of the groups were perfused at a rate of 0.5 mL/min while

the other groups were perfused at 0.2 mL/min. At the end of the procedure,

the lung was removed for determination of the wet-to-dry ratio and analysis

of GCB levels.

In 6 rats in the group with a flow rate of 0.2 mL/min perfused during 30

minutes serum samples were collected at 20 and 30 minutes for later analysis

of serum GCB levels.

Ten rats received the maximum tolerated intravenous dose of GCB, 160

mg/kg (40 mg per rat). Serum samples were collected at 6, 12, 15 and 21

minutes and stored at -70ºC for measurement of GCB lung and serum levels.

Chapter 4

67

dose response curves of gemcitabine

0,0

20,0

40,0

60,0

80,0

100,0

0 10 20 30 40conc gemc (µM)

% s

urv

ival

10min20min40min

Statistical Analysis

All data are presented as mean ± standard deviation (SD). Data were

analyzed using Student’s t-test. Significance was defined as p<0.05.

Results

Experiment 1

The dose-response curve is shown in figure 1. The IC50 value of CC531 for

gemcitabine at 10 minutes is 30.1±2.9 µM, at 20 minutes 21.7±3.3 µM and

at 40 minutes 6.5±3.1 µM.

Figure 1. Mean surviving fraction of CC531 colonies measured as afunction of gemcitabine concentration. Exposure times are 10, 20 and40 minutes.

Experiment 2

Lung levels of GCB are stabilized after 10 minutes for an infusion rate of 0.5

mL/min, and after 20 minutes for a rate of 0.2 mL/min (Table 1 and Figure

Chapter 4

68

2). No difference in lung levels of GCB is seen at different times between

the two flow rates (Table 2).

No significant difference in wet-to-dry ratio was measured between both

flow groups or between different perfusion times (Figure 3).

No significant difference in serum levels of GCB was observed after

intravenous infusion in time (Figure 4, upper curve). Serum levels during left

lung perfusion at 0.2 mL/min (Figure 4, lower curve) were significantly

lower (p=0.02) while lung levels were significantly higher (p=0.003)

compared to intravenous infusion (0.2 microgram/g; reference 11).

Figure 2. Gemcitabinelung levels in functionof exposure time

Figure 3. Wet-to-dryratio as a function ofperfusion time andflow rate

Lung levels of gemcitabine

0

600

1200

1800

10 min 20 min 30 min 40 min

perfusion time (minutes)

con

cen

trat

ion

gem

cita

bin

e (n

g/g

)

0,2mL/min0,5mL/min

wet-to-dry ratio

1

1,04

1,08

1,12

1,16

10 20 30 40perfusion time (minutes)

wet

-to

-dry

rat

io

0,5 mL/min0,2 mL/min

Chapter 4

69

gemcitabine serum levels

0

1

2

3

4

6 12 15 21 30time after ILuP and iv infusion (minutes)

gem

cita

bine

leve

ls

(mg/

mL)

BFO 0,2mL/miniv

Figure 4. Gemcitabine serum levels after IV infusion and BFO (blood flowocclusion). The IV treated rats received a bolus of 40 mg on t=0. Rats thathad BFO at 0.2 mL/min received 16 mg during 30 minutes (SD<0.025mg/mL in the BFO group)

Flow rate Dose delivered Concentration delivered

0.5 mL/min

10 min: 13.3 mg

20 min: 26.7 mg

30 min: 40 mg

40 min: 53.3 mg

2.7 mg/mL

0.2 mL/min

10 min: 5.3 mg

20 min: 10.7 mg

30 min: 16 mg

40 min: 20.4 mg

2.7 mg/mL

Table 1. Dose and concentration of gemcitabine delivered by BFO (bloodflow occlusion)

Chapter 4

70

0.20mL/min 10 minutes 20 minutes 30 minutes 40 minutes

Mean±SD (µg/g) 0.63±0.13 0.94±0.21 0.97±0.41 1.35±0.6

p-value 0.0071 0.902 0.113

0.50mL/min 10 minutes 20 minutes 30 minutes 40 minutes

Mean±SD (µg/g) 0.62±0.37 0.90±0.53 0.76±0.38 1.19±0.77

p-value 0.171 0.632 0.133

p-value 0.96a 0.90 a 0.39 a 0.69 a

Table 2. Gemcitabine lung levels after BFO (blood flow occlusion) .1 20 minutes vs 10 minutes; 2 30 minutes vs 20 minutes;3 40 minutes vs 30minutes;a 0.50mL/min vs 0.20mL/min

Discussion

Complete surgical resection of pulmonary metastatic disease results in a 5-

year survival of around 40% [1]. Different novel target techniques to achieve

a higher local drug concentration in the lung are explored in an attempt to

improve this survival [12]. Isolated lung perfusion (ILuP) as a surgical

technique was already described in 1959 by Creech to obtain high lung

levels without systemic exposure [13]. In addition, it avoids metabolisation

by the liver or kidney. The technique was re-invented by Johnston during the

eighties [2] and tested in large animals by several investigators [2,14-16].

The lung proved to be an ideal organ since a complete vascular isolation was

possible without systemic exposure. The lung could also tolerate

temperatures as high as 44°C. Next with the development of rat models of

isolated lung perfusion, several drugs were tested for their tumour efficacy.

All these experiments showed isolated lung perfusion superior to intravenous

infusion, both pharmacokinetically and therapeutically as well [17-21]. This

resulted in some phase I trials of isolated lung perfusion with different

Chapter 4

71

agents, some of them already finished [22-25]. The main disadvantage of the

technique of isolated lung perfusion is its invasive character. It necessitates a

thoracotomy while its effect has to be exerted during a single procedure.

Therefore, it is doubtful that ILuP becomes useful to treat bulky metastatic

disease or to make inoperable patients curable. Regional infusion techniques

like bronchial artery infusion or pulmonary artery perfusion have the

advantage they can be repeated, but drugs need to have a good first-pass

effect in order to minimize systemic leakage through an open venous

circulation. Since lung metastases are mainly supplied by the pulmonary

artery, pulmonary artery perfusion by endovascular way is the preferred

technique. Of these techniques, blood flow occlusion (BFO) as described by

Wang [4] proved to be superior, and has been tested with success for

doxorubicin [4, 26]. The aim of this study was to assess the use of

gemcitabine by BFO, and compare it with ILuP which was tested in a

previous experiment [11]. As a first step, the IC50 of gemcitabine was

explored for an adenocarcinoma cell line of which a rat model of pulmonary

metastatic disease is present at our laboratory [20]. These in-vitro

experiments showed that 50% of the cells were killed between a

concentration of at least 2.2 mg/mL for 40 minutes and 7.2 mg/mL for 20

minutes. In a next experiment, plasma and lung concentrations were

compared between BFO, IV injection and ILuP. The concentrations after

ILuP were extrapolated from a previous experiment. ILuP with 40 mg (160

mg/kg) resulted in lung levels of 1.5 µg/g of GCB [11]. The flow rate of 0.5

mL/min used for the ILuP experiments was chosen for the BFO as well. In

addition, a lower perfusion rate (0.2 mL/min) was selected in order to reduce

the total amount of drug given for the same perfusion time (Table 2). The

concentration of GCB within the perfusate was the same for all BFO groups.

Chapter 4

72

For the intravenous treated animals, the maximum tolerated dose of 160

mg/kg was given. In this way, maximum systemic exposure achieving the

highest lung levels after intravenous injection were compared with regional

techniques like ILuP and BFO. Although the same concentration was

delivered by BFO in all groups, the rate of 0.2 mL/min resulted in the same

final lung levels compared to 0.5 mL/min for all groups, while the total

amount of GCB given was only half. Although the dose of GCB after

intravenous injection (40 mg; MTD) was significantly higher compared to

BFO at a rate of 0,2 ml/min which resulted in higher serum levels, lung

levels after intravenous injection were significantly lower (0.2 µg/g)

compared to BFO. Gemcitabine lung levels after ILuP for 30 minutes with a

dose of 40 mg (1.5 µg/g; reference 11) were not different from BFO at the

lowest rate for 20, 30 and 40 minutes. However, the total amount of GCB

given was less compared to ILuP.

In conclusion, our experiments demonstrate that the same gemcitabine levels

in lung tissue can be obtained using ILuP or BFO perfusion. Compared with

IV drug administration, the plasma levels are lower while the lung tissue

concentrations are higher with BFO perfusion. Compared to ILuP, a higher

and non-toxic systemic exposure is present after BFO while no different lung

levels are seen for a lower total amount of GCB given. In addition, BFO is

technically less demanding and can be used clinically by catheterization of

the pulmonary artery.

Acknowledgements

The authors thank A. Van Laer for technical assistance during all

experiments.

Chapter 4

73

References1. Pastorino U, Buyse M, Friedel G et al. Long-term results of lung metastasectomy:

prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997; 113: 37-

49.

2. Minchin RF, Johnston MR, Aiken MA, Boyd MR. Pharmacokinetics of

Doxorubicin in Isolated Lung of Dogs and Humans Perfused in Vivo. J Pharm And

Exp Therap 1984; 229: 193-8.

3. Osaki T, Oyama T, Takenoyama M et al. Feasibility of induction chemotherapy

using bronchial arterial infusion for locally advanced non-small cell lung cancer: a

pilot study. Surg Today 2002; 32: 772-8.

4. Wang H-Y, Ng B, Blumberg D, Port JL, Hochwald SN, Burt ME. Pulmonary

Artery Perfusion of Doxorubicin With Blood Flow Occlusion: Pharmacokinetics

and Treatment in a Metastatic Sarcoma Model. Ann Thorac Surg 1995; 60: 1390-4.

5. Marquet RL, Westbroek DL, Jeekel J. Interferon treatment of a transplantable rat

colon adenocarcinoma: importance of tumour site. Int J Cancer 1984; 33: 689-92.

6. Hendriks JMH, Van Schil PEY, De Boeck G, Van Oosterom A, Van Marck E,

Eyskens EJM. Isolated lung perfusion with melphalan and Tumour Necrosis Factor

for pulmonary adenocarcinoma. Ann Thorac Surg 1998; 66: 1719-25.

7. Pestieau SR, Stuart OA, Chang D, Jacquet P, Sugarbaker PH. Pharmacokinetics of

intraperitoneal gemcitabine in a rat model. Tumori 1998; 84: 706-11.

8. Van Moorsel CJ, Kroep JR, Pinedo HM et al. Pharmacokinetic schedule finding

study of the combination of gemcitabine and cisplatin in patients with solid

tumours. Ann Oncol 1999; 10: 441-8.

9. Skehan P, Storeng R, Scudiero D et al. New colorimetric cytotoxicity assay for

anticancer-drug screening. J Natl Cancer Inst 1990; 82: 1107-12.

10. Papazisis KT, Geromichalos GD, Dimitriadis KA, Kortsaris AH. Optimization of

the sulforhodamine B colorimetric assay. J Immunol Methods 1997; 208: 151-8.

11. Van Putte BP, Hendriks JM, Romijn S et al. Isolated Lung Perfusion with

Gemcitabine in a Rat: Pharmacokinetics and Survival. J Surg Res 2002; in press.

12. Putnam JB. New and evolving treatment methods for pulmonary metastases. Sem

Thoracic Cardiovasc Surg 2002; 14: 49-56.

13. Creech O, Krementz ET, Ryan RF, Winblad JN. Chemotherapy of cancer: regional

perfusion utilizing an extracorporeal circuit. Ann Surg 1958; 148: 616-32.

Chapter 4

74

14. Furrer M, Lardinois D, Thormann W et al. Isolated Lung Perfusion: Single-Pass

System Versus Recirculating Blood Perfusion in Pigs. Ann Thorac Surg 1998; 65:

1420-5.

15. Ratto GB, Esposito M, Leprini A et al. In Situ Lung Perfusion with Cisplatin.

Cancer 1993; 71: 2962-70.

16. Pogrebniak HW, Witt CJ, Terrill R et al. Isolated lung perfusion with tumour

necrosis factor: a swine model in preparation of human trials. Ann Thorac Surg

1994; 57: 1477-83.

17. Abolhoda A, Brooks A, Nawata S, Kaneda Y, Cheng H, Burt ME. Isolated Lung

Perfusion With Doxorubicin Prolongs Survival in a Rodent Model of Pulmonary

Metastases. Ann Thorac Surg 1997; 64: 181-4.

18. Port JL, Ng B, Ellis JL, Nawata S, Lenert JT Burt ME. Isolated Lung Perfusion

With FUDR in the Rat: Pharmacokinetics and Survival. Ann Thorac Surg 1996; 62:

848-52.

19. Tao-Sheng LI, Kazuro S, Kazuhiro U, Kouichi N, Sumihiko N, Kensuke E. Isolated

Lung Perfusion with Cisplatin in a Rat Lung Solitary Tumour Nodule Model.

Anticancer Research 1998; 18: 4171-6.

20. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model

of metastatic pulmonary adenocarcinoma. Eur Surg Res 1999; 31: 267-71.

21. Weksler B, Blumberg D, Lenert J, Ng B, Fong Y, Burt ME. Isolated single-lung

perfusion with TNF-α in a rat sarcoma lung metastases model. Ann Thorac Surg

1994; 58: 328-32.

22. Johnston MR, Minchen RF, Dawson CA. Lung perfusion with chemotherapy in

patients with unresectable metastatic sarcoma to the lung or diffuse

bronchioloalveolar carcinoma. J Thoracic Cardiovasc Surg 1995; 110: 368-73.

23. Burt ME, Liu D, Abolhoda A et al. Isolated Lung Perfusion for Patients With

Unresectable Metastases From Sarcoma: A Phase I Trial. Ann Thorac Surg 2000;

69: 1542-9.

24. Pass HI, Mew DJY, Kranda KC, Temeck BK, Donington JS, Rosenberg SA.

Isolated Lung Perfusion With Tumour Necrosis Factor for Pulmonary Metastases.

Ann Thorac Surg 1996; 61: 1609-17.

Chapter 4

75

25. Ratto G, Toma S, Civalleri D et al. Isolated lung perfusion with platinum in the

treatment of pulmonary metastases from soft tissue sarcomas. J Thorac Cardiovasc

Surg 1996; 112: 614-22.

26. Furrer M, Lardinois D, Thormann W et al. Cytostatic Lung Perfusion by Use of an

Endovascular Blood Flow Occlusion Technique. Ann Thorac Surg 1998; 65: 1523-

8.

Chapter 5

Chapter 5

Single-Pass Isolated Lung Perfusion versus

Recirculating Isolated Lung Perfusion with Melphalan

in a Rat Model

BP Van Putte1, JMH Hendriks1, S Romijn1,G Guetens 2, G De

Boeck2, EA De Bruijn2 and PEY Van Schil1

1Department of Thoracic and Vascular Surgery, University of

Antwerp, Antwerp, Belgium2Department of Oncology and Radiotherapy, Catholic University of

Leuven, Leuven, Belgium

Published in: Ann Thorac Surg 2002;74:893-8

Presented at the Annual Congress of the Society of ThoracicSurgeons, Fort Launderdale, January 2003 and at the Spring Meetingof the Dutch Society of Cardiothoracic surgery, Scheveningen, April2002.

Chapter 5

78

Abstract

Background: Isolated lung perfusion (ILuP) with melphalan (MN) is

superior to intravenous infusion for the treatment of pulmonary carcinoma

and sarcoma metastases. However, it is unknown whether a bolus injection

of MN into the perfusion circuit or ILuP with a fixed concentration of MN

will result in the highest lung levels.

Methods: ILuP with 0.5mg MN was performed in Wag/Rij rats for 30

minutes either by a single-pass system (SP) (fixed concentration) (n=10) or

by reperfusion (RP) (bolus injection) (n=10). In a separate experiment, rats

were perfused with blood as the perfusate. In a third experiment, tumour

levels were compared between SP, RP or intravenous therapy with a dose of

0.5mg. For induction of pulmonary metastases, 0.5x10E6 single

adenocarcinoma cells were injected intravenously and therapy was given on

day 30. For comparison of drug concentrations, unpaired Student’s t test was

applied. Statistical significance was accepted at p less than 0.05.

Results: Lung perfusion studies were successfully performed without

systemic leakage. Temperature of perfusate and rats was 34-37°C. A

significantly higher hematocrit (mean: 27.9) compared with buffered starch

(mean: 2.5) did not result in higher MN lung levels or lower wet-to-dry ratio.

Tumour levels were significantly higher after ILuP compared with

intravenous therapy. However, no difference in tumour and lung levels was

seen between single-pass and reperfusion.

Conclusion: Both ILuP techniques resulted in significant higher MN lung

levels than after intravenous therapy. Since no difference was seen between

single-pass and recirculating perfusion, MN can be injected as a bolus into

the closed perfusion circuit.

Chapter 5

79

Introduction

Although surgical resection remains the primary treatment for the treatment

of pulmonary metastases, the 5-year survival is only 30% to 40% [1].

Systemic chemotherapy is of limited value, and high-dose regimens in order

to increase response rates are limited by severe systemic side effects. To

achieve higher local concentrations in the lung with minimal systemic

exposure, isolated lung perfusion (ILuP) can be a promising technique. First

described by Creech and associates in 1958 [2], numerous agents like

doxorubicin, tumour necrosis factor, cisplatin, 5FU and melphalan have been

tested since, both in large animals and small animals as well [3-7]. For

melphalan, both acute and long-term experiments showed reversible short-

term toxicity, minimal long-term toxicity and excellent pharmacokinetic

profiles. Compared with intravenous treatment, ILuP resulted in a

significantly higher eradication of tumour nodules and a significantly longer

survival time. These findings have been confirmed for both carcinoma and

sarcoma lung metastases models in the rat [7-9]. In addition, isolated limb

perfusion with melphalan showed synergistic effects with tumour necrosis

factor for melanoma and sarcoma [10]. Because isolated lung perfusion in

patients with tumour necrosis factor was tolerated up to a dose of 6 mg,

melphalan seems to be a promising agent for further clinical studies, as a

single agent or in combination with tumour necrosis factor. So far, all

isolated lung perfusion experiments with melphalan were single-pass

perfusions, which is an excellent technique for pharmacokinetic studies.

Clinical ILuP, however, is performed with a closed extracorporeal circuit

with a priming volume of 1.0 L or less, mainly to reduce the amount of

expensive perfusate fluids needed compared with single-pass perfusions.

However, it is unknown whether the dose or the concentration of melphalan

Chapter 5

80

will result in the highest lung levels. If it is the concentration that will

determine this level, it will be very difficult to maintain the concentration

constantly in a closed circuit with reperfusion. A dilutional effect will

develop during perfusion, because a lower concentration at the venous side

will be mixed with the higher concentration at the arterial side. In

preparation for clinical phase I studies, these pharmacokinetic parameters

need to be defined, and therefore, single-pass left lung ILuP (SP) was

compared with recirculating isolated left lung perfusion (RP) in a rat model

to determine the maximal final melphalan levels in lung and tumour tissue.

With the single-pass technique, a constant concentration is presented to the

lung, whereas in the reperfusion circuit, a bolus (a dose) is injected into the

circuit, but the concentration within the perfusate is changing continuously.

The total amount delivered is the same but the volume of the perfusate will

be different. In addition, blood as a perfusate was used in order to study the

effect of the presence of red blood cells on the maximal melphalan tissue

concentration because red blood cells can act as a carrier for MN, and to

investigate the development of pulmonary oedema [11].

Material and Methods

Animals

Male inbred WAG-Rij strain rats (weight, approximately 200g), obtained

from Harlan-CPB (Zeist, The Netherlands), were used for all experiments.

Animals were treated in accordance with the Animal Welfare Act and the

“Guide for the Care and Use of Laboratory Animals” (NIH Publication 86-

23, revised 1985). The rats were transported in sterile conditions, housed in

suspended mesh wired cages and ad libitum fed a standard pellet diet

(standard rat chow, Hope Farms, Woerden, The Netherlands). The

Chapter 5

81

experimental protocols were approved by the Institutional Animal Care Use

Committee, University Hospital Antwerp.

Tumour

The CC531S tumour cell line was used for assessment of MN levels in lung

metastases. This cell line was derived from a chemically induced

adenocarcinoma of the colon of a WAG rat. CC531S was cultured in

complete medium and maintained by serial passage. Complete medium

consisted of RPMI-1640 medium, supplemented with 10% heat activated

fetal bovine serum (FBS, both from Life Technologies, Merelbeke,

Belgium), 2 mmol/L glutamine, 50 mg/mL streptomycin and 50 U/mL

penicillin [3].

Induction of lung metastases

CC531S cells were prepared for injection by dispersal with a solution of

0.25% ethylinediamine tetraacetic acid (EDTA) and 0.05% trypsin in Hank

’s balanced salt solution (HBSS). Cell viability was determined with a

tryptan blue exclusion method using a hemocytometer [8]. Because the

original technique for induction of lung metastases resulted in small

pulmonary metastases not suitable for pharmacokinetic analysis, a new

model was developed. Ten rats were injected with only 500.000 CC531S

adenocarcinoma cells in 1mL of RPMI 1640 into the left femoral vein. Half

of the rats were killed on day 15 and half on day 30. On day 15, only a few

tumour nodules were visible still too small for selective excision. However,

on day 30 large metastases were visible macroscopically as separate nodules,

easily to excise for further analysis.

Chapter 5

82

Blood

In order to perfuse the lung with buffered hetastarch (BHE) enriched with

red blood cells, many rats need to be sacrificed for one experiment since

only 6 mL/100 g whole blood can be retrieved from a single rat. Because red

blood cells (RBC) seem to be of surprising uniformity in all mammals [10],

we were not restricted to the use of rodent RBCs, and washed rabbit

erythrocytes were chosen to enrich the perfusate. Rabbit blood was drawn

directly from the left ventricle of living rabbits after systemic heparinization.

The red cells were processed within a few hours, using sterile techniques to

remove plasma, white cells and platelets. The blood was spun at 3,500g for

10 minutes and the supernatant plasma was removed. The cells were diluted

with 0.9% saline and spun in the same way as in the first process. The red

blood cells were then diluted with BHE solution to a hematocrit of around

30%. Thereafter, leucocytes were removed through a commercially available

leukocyte filter.

Technique of Single-Pass Isolated Left Lung Perfusion (SP) and Isolated

Left Lung Reperfusion (RP)

Single-pass isolated left lung perfusion was performed according to the

technique as described by Hendriks et al [12]. Briefly, anaesthesia was

induced with isoflurane. Rats were intubated by translaryngeal illumination

and connected to the ventilator. Isoflurane was titrated between 0.5% and

1.5% according to muscle relaxation, heart rate and pupil size. Ventilation

was accomplished with a volume-controlled ventilator at a rate of 75

strokes/min and a tidal volume of 10 mL/kg. After a left thoracotomy, the

lung and rib retractor were placed anteriorly and the hilum was dissected

free. The bronchus was deprived from its surrounding tissue in order to

Chapter 5

83

exclude the bronchial arterial flow to the lung. After clamping the

pulmonary artery and vein with curved microclips, a 16-G Angiocath was

placed through the chest wall. A PE-10 perfusion catheter was introduced

into the chest through the Angiocath and secured by a 4/0 silk tie after

insertion into the pulmonary artery. Perfusate was delivered through this

catheter. In addition, a pulmonary venotomy was performed and two venous

catheters (PE-90) were placed into the superior and inferior pulmonary vein

to collect the venous effluent. In the case of single-pass perfusion, the

effluent is discarded at the venous side. With isolated left lung reperfusion,

the venous effluent is collected into the warm water bath, and from this bath

reinfused into the pulmonary artery (fig. 1).

For both SP and RP, the perfusate is temperature-controlled throughout the

whole perfusion. Rats were placed on a heating pad immediately after

induction, and body temperature was kept constantly between 34°C and

37°C.

In all experiments, animals were perfused on the left side for 25 minutes at a

rate of 0.5 mL/min, followed by a 5-minute washout with buffered

hetastarch (BHE) at 0.5 mL/min.

Chapter 5

84

Figure 1. The rat was placed on a heating pad after orotracheal intubationand connected to a ventilator. Through a left thoracotomy, the pulmonaryartery and veins were cannulated. The perfusate was drawn from a warmwater bath by a roller pump and infused into the pulmonary artery. Forsingle-pass perfusion, the effluent coming out of the pulmonary veins wasdiscarded. In case of reperfusion, the effluent was sent back by the rollerpump to the warm water bath and recirculated into the pulmonary artery. A= reperfusion circuit; B = single-pass perfusion.

Melphalan (Alkeran, 50 mg/vial, Wellcome, Waterloo, Belgium) perfusate

solutions were prepared by reconstituting lyophilized powder in the supplied

diluent and performing appropriate dilutions with BHE prior to the

experiments.

Chapter 5

85

Experiment 1: Single-pass Isolated Lung Perfusion versus Isolated

Lung Reperfusion with buffered starch and red blood cells as

perfusates

Twenty-eight rats were randomized into four groups. Group 1 (n=10) and

group 3 (n=4) had SP with 0.5 mg of MN, whereas group 2 (n=10) and

group 4 (n=4) had RP with 0.5 mg of MN. While group 1 and 2 had BHE as

the perfusate, this was BHE enriched with red blood cells (RBHE) for

groups 3 and 4. During perfusion, pulmonary effluent samples were

collected at 5-minute intervals throughout the perfusion for groups 1 and 2.

At the completion of a 25-minute MN perfusion and 5-minute BHE washout,

the left lung was removed and frozen at -70°C for later analysis.

Experiment 2: Determination of melphalan levels within the lung

metastases

Twelve rats were randomized into three groups (n=4 each). On day 0, all rats

received 500,000 CC531S adenocarcinoma cells in the left femoral vein. On

day 30, group 1 had SP with 0.5 mg of MN and group 2 had RP with 0.5 mg

of MN. Group 3 received 0.5 mg of intravenous MN. The left lungs of group

1 and 2 were perfused for 25 minutes with BHE as the perfusate followed by

a 5-minute BHE washout. Thirty minutes after the onset of therapy,

pulmonary metastases in all groups were separated from normal lung tissue

and frozen at -70°C for later analysis.

Melphalan Processing and Measurement

Gas chromatography-mass spectrometry described by De Boeck et al. was

used for measuring melphalan levels in lung tissue and serum [13]. P-[Bis(2-

Chapter 5

86

chloroethyl)amino]-phenylacetic acid methyl ester was used as an internal

standard. Samples were extracted over trifunctional C18 silica columns.

Tabel 1. Hematocrit and temperaturea group 1 compared to group 2b group 2 compared to group 4c group 3 compared to group 1d group 3 compared to group 4

SP: single-pass perfusionRP: reperfusion

Statistical Analysis

All data are presented as mean ± standard deviation (SD). Comparison

between groups was done by an unpaired Student’s t-test. Significance was

defined as p less than 0.05.

Results

Mortality

All rats survived injection of tumour cells and the procedure related survival

was 100% for all experiments.

group procedure n Temp

reservoir (ºC)

Temp rat(ºC)

hematocrit W/D

1 SP, BHE 4 35.8±0.5

(NSa)

35.8±1.1

(NSa)

1.4±0.6 6.4±0.5

(NSa)

2 RP, BHE 4 36.0±0.8

(NSb)

35.1±0.8

(NSb)

3.6±1.1

(p<0.001b)

6.5±0.5

(NSb)

3 SP, RBC 4 35.5±1.0

(NSc)

34.6±0.5

(NSc)

27.8±3.2

(p<0.001c)

6.5±0.7

(NSc)

4 RP, RBC 4 36.6±0.5

(NSd)

34.8±1.0

(NSd)

28.1±1.6 6.6±0.7

(NSd)

Chapter 5

87

Experiment 1

The temperature of the reservoir and the rats was not significantly different

between groups (Table 1).

Hematocrit levels were significantly lower for group 1 and 2 (BHE

perfusate) compared with groups 3 and 4 (red blood cell-BHE perfusate). No

significant difference was seen in wet-to-dry ratios between the groups.

(table 1).

Group Procedure n Left lung (µg/g) p

1 left SP, BHE 10 29.3±11.8 NSa

2 left RP, BHE 10 27.1±11.4 NSb

3 left SP, RBC 4 18.7±12.9 0.09c

4 left RP, RBC 4 17.3±10.2 NSd

Table 2. Left lung melphalan levels after perfusion with 0.5 mg MNa group 1 compared to group 2b group 2 compared to group 4c group 3 compared to group 1d group 3 compared to group 4

SP: single-pass perfusionRP: reperfusionMN: melphalan

No significant difference was seen in the left lung MN levels of rats treated

with single-pass compared to reperfusion (Table 2). Also no statistically

significant difference was observed in left lung MN levels of rats treated

with reperfusion with RBC-BHE as the perfusate compared with BHE, or

between single-pass and reperfusion with RBC as the perfusate (Table 2). A

significant difference was seen in rats treated with single-pass perfusion with

RBC-BHE compared with BHE. Figure 2 shows MN effluent levels over

time.

Chapter 5

88

MN effluent levels

05

101520

0 5 10 15 20 25 30

time (minutes)

MN

eff

luen

s (m

icro

g/m

l)

RP BHE

RP RBCSP BHESP RBC

Experiment 2

Tumour melphalan levels of intravenously treated rats were significantly less

compared with single-pass perfusion or reperfusion (Table 3). However, no

statistically significant differences in final tumour melphalan levels were

seen between a single-pass perfusion compared with a recirculating

procedure (p=0.133) (Table 3).

Table 3.

Lung metastasesmelphalan levels(acompared withgroup 3)

Figure 2.

Melphalaneffluent levels

Comment

While the lungs are the most common site of metastases [14], it is known

that for sarcoma and carcinoma metastases, up to 50 % of the patients have

recurrent disease exclusively in the lung after resection of the primary

disease, probably due to the presence of micrometastases at the time of the

initial operation [1,15]. Recurrent disease can be inoperable due to size,

number or location of these lung metastases. Therefore, it is important to

have local control of microscopic disease after surgical resection of visible

Group procedure n Tumour (µg/g) p

1 SP 0.5 mg MN 4 8.3±5.9 0.031a

2 RP 0.5 mg MN 4 16.6±7.5 0.014a

3 iv 0.5 mg MN 4 1.5±0.06 NS

Chapter 5

89

metastatic disease in order to achieve a long-term survival. Systemic

chemotherapy has been ineffective because it is limited by severe systemic

side effects. Isolated lung perfusion can deliver very high doses of

chemotherapy with minimal systemic exposure, while avoiding

metabolisation of the drug in liver or kidney. Therefore, isolated lung

perfusion seems to be a promising technique to eradicate micrometastatic

disease at the time of surgical resection of macrometastatic disease in order

to prevent the development of recurrent disease.

Since the early 1980s, the technique of isolated lung perfusion has been

tested with success in large animals like dogs and pigs [4,16,17]. Later, rat

models were developed for tumour efficacy studies with different

chemotherapeutic agents. Numerous agents like tumour necrosis factor,

FUDR, cisplatin, and doxorubicin have been tested with success in these rat

models, and single-pass left lung perfusion was the only technique used [18].

A single-pass system was selected as a procedure for technical reasons and

because this method facilitates drug kinetic studies by ensuring a constant

drug concentration in the perfusate. Technically seen, it is very demanding to

cannulate the pulmonary veins of the rat and even more difficult to restore

the venotomy after perfusion in case of survival experiments. Therefore, no

cannulation is performed in the single-pass technique, but the effluent is

discarded by a suction catheter and closure of the small venotomy is

accomplished by pressure after the procedure [19].

Weksler and associates showed in their rat model of ILuP that the perfusate

doxorubicin concentration and duration of perfusion were the only factors

determining the final lung concentration of doxorubicin. The dose of

doxorubicin, and the total amount and rate of perfusion were not important

in determining the final lung doxorubicin concentration [20]. They did not

Chapter 5

90

investigate these values in a model of pulmonary metastases. Their studies

were finalized by a phase I dose-escalating trial of ILuP with doxorubicin in

sarcoma metastases [21]. For this clinical study, a recirculating perfusion

system was preferred because it minimizes the amount of perfusate and drug

needed, vascular isolation becomes more complete and greater flexibility is

obtained in regulation of the perfusion environment [22]. The drug was

added to the circuit as a bolus and not as a constant concentration. The fact

that drug concentration was crucial in the rat experiments for the final lung

concentration might explain the large variations in perfusate and lung levels

of doxorubicin between patients receiving the same amount of drug [21]. It

also shows the difficulty of calculating the amount of drug to give each

patient. In contrast with isolated limb perfusion, where doses are calculated

by the amount of water displaced by the limb, no such value is available is

available for lung perfusion studies [10].

In our laboratory, a rat model of isolated single-lung perfusion was

developed, and experiments with melphalan were extensively performed in

preparation of human clinical trials since 1995. These animal experiments

showed low operative mortality and negligible long-term injury to the lungs

[7,8]. We have shown that ILuP with melphalan results in significantly

higher final lung melphalan levels compared with a high intravenous

injection of melphalan [7]. Studies of experimental pulmonary metastases

showed statistically significant eradication of experimental sarcoma and

carcinoma metastases in rats undergoing isolated single-lung perfusion with

melphalan compared to intravenous therapy [7-9]. In preparation of a clinical

phase I dose-escalating trial of isolated lung perfusion with melphalan, a

number of questions were unanswered. First, it is unknown how the dose

should be calculated because lung mass can be very different between two

Chapter 5

91

patients. No value such as limb volume for isolated limb perfusion is

available for the lung. Second, in all clinical studies performed so far, a

closed circuit was used in order to lessen the amount of perfusate needed and

to control values like flow, air leakage and temperature. In these studies, the

drug was given as a bolus into the circuit proximal to the pulmonary artery

(a dose), but not offered to the lung as a constant concentration. However,

Weksler and associates demonstrated that drug concentration was crucial for

the final lung concentration of doxorubicin [20]. For melphalan these values

were not known. In this study, both lung and tumour MN levels were

determined, as no correlation has been made between normal lung and

tumour tissue. Tumour melphalan levels were significantly higher in the

single-pass group compared with MN levels in the intravenous group.

However, there was no significant difference between single-pass and

reperfusion group, and a steady concentration was reached after 15 minutes.

So, not the concentration but the total amount of drug was critical for the

final lung concentration, which is in contrast with rat studies of doxorubicin

[20]. In this way, it seems that a bolus of MN into the circuit will result in

the same final lung and tumour levels compared with single-pass perfusion

with a fixed concentration. A higher concentration within the lung was

achieved compared with the tumour. In the reperfusion group, a higher

concentration was reached in the tumour compared with the single-pass

group, but the difference was not significant.

So far, both experimental and clinical studies are only available for

doxorubicin. Although doxorubicin has proven to be a superior agent for

isolated lung perfusion studies compared with intravenous therapy, the

clinical dose-escalating study was disappointing, and for the highest dose,

important side effects were seen [21]. Putnam and associates showed smaller

Chapter 5

92

dose increases can result in lower morbidity [23]. The highest dose used by

Burt and associates, 80 mg/kg, was twice the lower dose, whereas Putnam

and associates included a dose of 60 mg/kg without any morbidity. These

two studies show that pharmacokinetics will behave totally differently from

the three-compartment model described for intravenous administration of

chemotherapy. These dose-escalating studies should be performed for every

agent that was promising in animal studies. The human isolated lung model

described by Linder and associates extrapolates animal studies into the

human situation, and provides useful pharmacokinetic information before

starting phase I dose-escalating trials [24].

No significant differences in wet-to-dry ratios were seen between the groups

in our experimental study (mean of all groups, 6.5 ± 0.5). Because both the

pulmonary veins were cannulated, the wet-to-dry ratios were higher

compared with results obtained by Weksler and associates [20]. In our

previous work, we used BHE as the perfusate fluid of choice based on the

experiments by Weksler and associates [20]. These experiments did not

investigate the addition of RBCs within the perfusate. However, it is possible

that these RBCs will act as a better carrier for melphalan [25] and will

diminish the amount of pulmonary edema during perfusion. In addition,

RBCs will decrease anoxia-induced pulmonary damage. Despite this

hypotheses, no significantly higher MN levels were reached with RBCs

within the perfusate compared with BHE, and no difference was seen in the

amount of pulmonary edema. Therefore, BHE will remain the perfusate fluid

of choice for both laboratory and clinical studies.

In conclusion, no difference in final lung concentration was seen between

single-pass left lung perfusion and isolated left lung reperfusion. The single-

pass left lung perfusion acted as a model to deliver a fixed concentration of

Chapter 5

93

drug to the lung. This was compared with recirculating isolated lung

perfusion as a model of a closed circuit, in which a bolus was injected at a

certain time point. The total amount of drug delivered in the two models was

identical. In addition, the administration of RBCs to the perfusate did not

increase the final lung concentration, and the amount of pulmonary oedema

was not statistically significant. These findings will simplify isolated lung

perfusion with melphalan because buffered starch will be the perfusate of

choice and melphalan can be injected as a bolus into the circuit, after

stabilization of the main values, such as temperature, flow, and leakage.

Dose-escalating phase I studies will determine whether melphalan is a useful

agent in the setting of isolated lung perfusion.

Acknowledgements

We wish to thank Marleen Nysten and August Van Laer for assistance in all

experiments. This work was covered by a research grant from the University

of Antwerp.

References

1. Pastorino U, Buyse M, Friedel G et al. Long-term results of lung metastasectomy:

prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37-49.

2. Creech O, Krementz ET, Ryan RF, Winblad JN. Chemotherapy of cancer: regional

perfusion utilizing an extracorporeal circuit. Ann Surg 1958;148:616-32.

3. Weksler B, Lenert J, Ng B, Burt M. Isolated single lung perfusion with doxorubicin is

effective in eradicating soft tissue sarcoma lung metastases in a rat model. J Thorac

Cardiovasc Surg 1994;107:50-4.

4. Pass HI, Mew DJ, Kranda KC, Temeck BK, Donington JS, Rosenberg SA. Isolated lung

perfusion with tumour necrosis factor for pulmonary metastases. Ann Thorac Surg

1996;61:1609-17.

Chapter 5

94

5. Ratto GB, Toma S, Civalleri D, et al. Isolated lung perfusion with platinum in the

treatment of pulmonary metastases from soft tissue sarcomas. J Thorac Cardiovasc Surg

1996;112:614-22.

6. Ng B, Lenert JT, Weksler B, Port JL, Ellis JL, Burt ME. Isolated lung perfusion with

FUDR is an effective treatment for colorectal adenocarcinoma lung metastases in rats.

Ann Thorac Surg 1995;59:205-8.

7. Hendriks JMH, Van Schil PEY, De Boeck G, et al. Isolated lung perfusion with

melphalan and tumour necrosis factor for metastatic pulmonary adenocarcinoma. Ann

Thorac Surg 1998;66:1719-25.

8. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-271.

9. Nawata S, Abecasis N, Ross HM et al. Isolated lung perfusion with melphalan for the

treatment of metastatic pulmonary sarcoma. J Thorac Cardiovasc Surg 1996;12:1542-8.

10. Lejeune F, Kroon B, Difilippo F, et al. Isolated limb perfusion: the European experience.

Surg Oncol Clin N Am 2001;10:821-32.

11. Barni S, Gerzeli G, Gervaso MV. Cytofluorometric study of the erythrocyte

glycoconjugates content in different species of vertebrates. Acta Histochem

1983;73:251-7.

12. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

13. De Boeck G, Van Cauwenberghe K, Eggermont AM, Van Oosterom AT, De Bruijn EA.

Determination of melphalan and hydrolysis products in body fluids by GC-MS. High

Res Chromatog 1997;12:697-700.

14. Crow J, Slavin G, Kreel L. Pulmonary metastases: a pathologic and radiologic study.

Cancer 1981; 47:2595-602.

15. J. Hendriks, S. Romijn, B. Van Putte, J. Vermorken, E. Van Marck, E. Eyskens, P. Van

Schil. Surgery for lung metastases: a retrospective analysis. Acta Chirurgica Belgica

2002. In press.

16. Johnston MR, Christensen CW, Minchin RF et al. Isolated total lung perfusion as a

means to deliver organ-specific chemotherapy: long-term studies in animals. Surgery

1985;98:35-44.

Chapter 5

95

17. Pogrebniak HW, Witt CJ, Terrill R et al. Isolated lung perfusion with tumour necrosis

factor: a swine model in preparation of human trials. Ann Thorac Surg 1994;57:1477-83.

18. Weksler B, Ng B, Lenert JT, Burt ME. Isolated single-lung perfusion: a study of optimal

perfusate and other pharmacokinetic factors. Ann Thorac Surg 1995;60:624-9.

19. Wang H, Port J, Hochwald S, Burt M. Revised technique of isolated lung perfusion in

the rat. Ann Thorac Surg 1995;60:211-2.

20. Weksler B, Ng B, Lenert JT, Burt ME. Isolated single-lung perfusion: a study of optimal

perfusate and other pharmacokinetic factors. Ann Thorac Surg 1995;60:624-9.

21. Burt ME, Liu D, Abolhoda A et al. Isolated lung perfusion for patients with unresectable

metastases from sarcoma: a phase I trial. Ann Thorac Surg 2000;69:1542-9.

22. Johnston MR. Lung perfusion and other methods of targeting therapy to lung tumours.

Chest surgery clinics of North America 1995;1:139-156.

23. Putnam JB, Benjamin R, Rha S, et al. Early results of isolated single lung perfusion for

treatment of unresectable sarcomatous metastases. Abstract presented at the 80th Annual

AATS meeting, April 30 to May 3, 2000, Toronto, Ontario.

24. Linder A, Friedel G, Fritz P, Kivisto KT, McClellan M, Toomes H. The ex-vivo isolated,

perfused human lung model: description and potential applications. Thorac Cardiovasc

Surg 1996;44:140-6.

25. Ahmed AE, Hsu TF, el-Azhary RA, Moawad H, Costanzi J. Macromolecular

interactions of [14-C-ring] melphalan in blood. Biochem Pharmacol 1982;31(8):1615-9.

Chapter 6

Gemcitabine prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma

BP Van Putte1, JMH Hendriks1, S Romijn1, PB Vermeulen2,

E Van Marck2 and PEY Van Schil1

Departments of Thoracic and Vascular Surgery1and Pathology2,

University Hospital Antwerp, Edegem, Belgium

Submitted for publication

Presented at the Annual Congress of the European Society ofCardiothoracic Surgery, October 28-30, 2002, Istanbul (Turkey);presented at the autumn meeting of the Dutch Society of Surgery,November 1, 2002, Utrecht (the Netherlands)

Chapter 6

98

Abstract

Objective: Chronic toxicity and efficacy after isolated left lung perfusion

(ILuP) with gemcitabine (GCB) were studied in a rat model of metastatic

pulmonary adenocarcinoma.

Methods: Toxicity: Seventy-five rats were randomized between intravenous

injection (IV) and ILuP on day 0. Doses of 20, 40, 80, 160 and 320mg/kg

GCB and buffered hespan were given in the ILuP arm. Doses of 40, 80, 160

and 320mg/kg (n=6) and buffered hespan were given in the IV arm. At day

90, lungs were harvested and processed for histologic analysis. Efficacy: On

day 0, thirty-three rats were randomized into five groups: Groups 1 (n=5)

and 2 (n=10) received tumour cells IV for induction of bilateral lung

metastases, whereas group 2 received IV treatment with GCB 160mg/kg at

day 7 (n=10). Groups 3, 4 and 5 underwent a 10-min occlusion of the right

pulmonary artery during tumour cell injection for induction of unilateral left

lung metastases. On day 7, group 3 received no treatment (n=4). Group 4

and 5 underwent left ILuP with GCB 320mg/kg and 480 mg/kg (n=7 each).

Results: Toxicity: After IV treatment, all rats receiving GCB 320mg/kg died

within one week while GCB 160mg/kg had a survival rate of 60%. After

ILuP with GCB 160mg/kg and GCB 320mg/kg survival rates were 83% in

both groups. A slight increase in collagen deposits was seen for ILuP

320mg/kg after 3 months compared to controls. Efficacy: Median survival of

group 4 (38±4 days) was significantly longer compared to controls (group 3;

28±2 days; p=0.02).

Conclusions: ILuP with GCB prolongs survival in experimental metastatic

adenocarcinoma while no major acute or chronic toxicity is observed.

Chapter 6

99

Introduction

Colorectal adenocarcinoma is a major cause of pulmonary metastases [1].

Complete resection for pulmonary metastases has extensively been studied

but results in a 5-year survival of approximately 40 % for pulmonary

metastases from colorectal adenocarcinoma [2]. Systemic chemotherapy

does not have a significant effect on long-term survival and results in severe

side effects like bone marrow depression and hair loss [3]. In order to

prevent systemic side effects and to obtain higher local drug concentrations,

Weksler et al. developed a rat model of Isolated Lung Perfusion (ILuP) as a

new strategy for the treatment of pulmonary metastases. This technique was

subsequently modified by Wang [4,5]. ILuP is especially focused on

destroying micrometastases and can be combined with metastasectomy to

remove manually palpable metastases. In preparation of human studies

several drugs have now been tested in this rat model with different success

rates. Melphalan proved to be the most active drug tested in the rat model of

metastatic pulmonary sarcoma and colorectal adenocarcinoma [6,7]. In the

latter study from our laboratory unilateral pulmonary metastases were

induced by intravenous (IV) injection of adenocarcinoma cells while

clamping the right pulmonary artery during ten minutes in order to induce

metastases on the left side. Rats had left ILuP with melphalan at day seven

and survival was compared to rats which were IV treated. Mean survival was

81±12 days compared to 28±3 days in the IV treated group while some

animals of the ILuP group were even disease-free histologically after

sacrification after 90 days [7]. Significantly more alveolar cell hyperplasia

was noted in the acute phase after ILuP with melphalan [7]. Ueda et al.

confirmed the mild acute toxicity with melphalan but did not observe long-

term adverse effects [8]. Due to these hopeful results a phase I trial has been

Chapter 6

100

started in our hospital in cooperation with the Sint Antonius Hospital

(Nieuwegein, the Netherlands).

Gemcitabine (GCB) is a deoxycytidine analogue with single agent activity in

non-small cell lung carcinoma, colorectal adenocarcinoma and carcinoma of

the pancreas, breast and ovary [9-12]. Although pulmonary toxicity of

gemcitabine has only seldom been reported and is usually mild, respiratory

distress syndrome and interstitial pneumonitis after IV GCB infusion have

been described [13,14]. In a recent study we determined the survival of

CC531S adenocarcinoma cells exposed to different concentrations of GCB.

The IC50 value which is the concentration of GCB resulting in a growth

inhibition of 50% was 23.6 nM which means that the CC531 S cells are

highly sensitive to GCB [15]. Furthermore, this study showed the maximally

tolerated dose after ILuP with GCB to be higher compared to intravenous

therapy resulting in significantly higher lung levels after ILuP [15].

No data are available on efficacy and chronic pulmonary toxicity of ILuP

with doses of GCB that are much higher than the doses of the currently used

IV therapy. Therefore this study aimed to investigate the efficacy and the

histological side effects of GCB on lung tissue after ILuP compared to

systemic administration.

Material and methods

Animals

Male inbred Wag-Rij strain rats (weight, approximately 250g), obtained

from Iffa Credo (Brussels, Belgium), were used for all experiments. The

animals were treated in accordance with the Animal Welfare Act and the

“Guide for the Care and Use of Laboratory Animals” (NIH Publication 86-

23, revised 1985). The rats were transported in sterile conditions, housed in

Chapter 6

101

suspended mesh wired cages under standard laboratory conditions and ad

libitum fed a standard pellet diet (standard rat chow, Hope Farms, Woerden,

The Netherlands). The experimental protocols were approved by the

Institutional Animal Care Use Committee, University of Antwerp.

Technique of Left Isolated Lung Perfusion

Isolated left lung single-pass perfusion was performed according to the

technique described by Wang [5]. This technique was modified in our

laboratory and full details have been reported previously [15-17]. Briefly,

rats were anaesthetised with isoflurane in a mixture of nitrous oxide (NO2)

and oxygen (O2). Isoflurane was administered in a concentration of 4% and

the NO2:O2 ratio was set to 3:1. After 5 minutes, rats were intubated with a

16-gauge Insyte-W catheter by translaryngeal illumination and afterwards

ventilated with a volume-controlled ventilator (Harvard Rodent Ventilator,

South Natick, MA) [15]. Once connected to the ventilator, the NO2:O2 ratio

was 1:1 (0.5 L/min) and the rate of ventilation was 75 strokes/min with a

tidal volume of 10 mL/kg. Subsequently, the left chest was shaved and a left

thoracotomy between the third and fourth intercostal space was performed.

After placing a rib retractor, the left lung was luxated anteriorly and the

pulmonary artery and vein were dissected free before they were clamped

with curved microclips. A PE-10 perfusion catheter (Becton Dickinson,

Bornem, Belgium) was introduced into the chest through a 16-G Angiocath,

which was placed through the chest wall. The PE-10 catheter was inserted

into the pulmonary artery and fixed with a 4/0 silk tie. The perfusate was

delivered through this catheter. The effluent was sucked by a catheter at the

venous side in proximity of the venotomy.

Chapter 6

102

In all experiments, rats were perfused during 25 minutes followed by a 5-

minute washout with buffered starch at a rate of 0.5 mL/min. GCB

(gemcitabine, Eli Lilly Benelux) solutions were prepared by reconstituting

lyophilized powder in the supplied diluent and performing appropriate

dilutions with buffered hydroxy-ethyl starch (Haes Steril, BHE). After

perfusion the catheter was removed and the arteriotomy was closed with a

transverse suture (10/0 Ethilon, Ethicon, Belgium). The microclamps were

removed and the lung was returned into its anatomic position. Pressure was

applied using moist gauze over the lung in order to tamponade the bleeding

from the venotomy. A 16-gauge catheter connected to a 50-mL syringe was

introduced into the left chest cavity through a separate puncture wound to

facilitate lung reexpansion. The thoracotomy incision was closed in three

layers with 4/0 vicryl (Ethicon, Belgium). The animal was supplied with

room air until it began to breathe spontaneously. Subsequently, the pleural

drain and the endotracheal tube were removed.

Tumour

The CC531S tumour cell line was used to induce lung metastases. This cell

line was derived from a chemically induced adenocarcinoma of the colon of

a WAG rat [18]. CC531S was cultured in complete medium and maintained

by serial passage. Complete medium consisted of RPMI-1640 medium,

supplemented with 10% heat activated fetal bovine serum (Life

Technologies, Merelbeke, Belgium), 2 mmol/L glutamine, 50 mg/mL

streptomycin and 50 U/mL penicillin.

Chapter 6

103

Induction of pulmonary metastases

Unilateral and bilateral lung metastases were induced according the method

which was extensively described before [19-21].

Briefly, after anaesthesia and intubation, a right thoracotomy was performed

through the third intercostal space for induction of unilateral lung

metastases. Under microscopic view a longitudinal incision was made along

the posterior side of the superior caval vein in order to visualize the right

pulmonary artery which was subsequently clamped. After repositioning the

rat, a tumour cell suspension (2.0 × 106 viable CC531 S adenocarcinoma

cells in 1 mL of RPMI) was injected IV into the left femoral vein which was

prepared in advance. The groin incision was closed with a running suture of

4-0 vicryl (Ethicon, Belgium). After a 10-minute occlusion of the right

pulmonary artery, the clamp was removed and the thoracotomy was closed

as described above.

For induction of bilateral metastases a tumour cell suspension (2.0 × 106

viable CC531 S adenocarcinoma cells in 1 mL of RPMI) was injected into

the left femoral vein. The groin incision was closed as described above.

Histopathological examination

All lungs were prefixed in 4% formaldehyde in a mixture of 0.1M

natriumcacodilaat (pH=7.4) and 1% CaCl2. Samples taken for light

microscopy investigations were dehydrated with ethanol, cleared with toluol

and embedded in paraffin wax. Sections of 4 µm thick were processed by

Masson Trichrome staining in order to assess collagen fibrosis.

The lung sections were evaluated by a pathologist (P.B.V) without

knowledge of the respective treatment schedule. Collagen fibrosis in the

alveolar septa was graded by own definitions, ranging from absent over mild

Chapter 6

104

to moderate fibrosis. Normal lung tissue is shown in figure 1a. The collagen

fibres surrounding larger bronchi were taken as an internal control of the

staining intensity. Mild fibrosis was defined as slight broadening of some

alveolar septa due to collagen fibres (figure 1b). Moderate fibrosis was

defined as easily recognizable broadening of a considerable amount of

alveolar septa due to collagen deposition (figure 1c). Pleural fibrosis was

defined as broadening of the pleura. Reproducibility was tested by re-

evaluation of the slides with an interval of one week by the same pathologist.

In less than 10%, a shift of grade 3 to 2 or vice versa was observed.

Statistical analysis

Survival rates were assessed by Kaplan-Meier analysis and differences in

survival by a log rank test. Toxicity data were statistically evaluated using a

χ2-test. Significance was accepted at p < 0.05.

Experiment 1: Chronic pulmonary toxicity of gemcitabine in rats

Seventy-five rats were randomized into eleven groups. Groups 1 to 5 had left

ILuP with 20 mg/kg (n=3), 40 mg/kg (n=6), 80 mg/kg (n=6), 160 mg/kg

(n=6) and 320 mg/kg (n=6) of GCB respectively. Group 6 (n=6) had ILuP

with buffered starch. Rats were perfused during 25 minutes followed by a 5-

minute washout. After three weeks a right pneumonectomy was performed

to assess left lung function. Groups 7 to 10 were infused IV with 40 mg/kg

(n=10), 80 mg/kg (n=10), 160 mg/kg (n=10) and 320 mg/kg (n=6) of GCB

respectively. Group 11 (n=6) was infused IV with saline (0.9% NaCl). All

animals were sacrificed after 90 days and lungs were removed and fixed in

formaldehyde and subsequently processed and stained as described before.

Chapter 6

105

Experiment 2: Efficacy of gemcitabine

Thirty-three rats were randomized into 5 groups.

At day 0, two groups received 2.0 × 106 tumour cells IV for induction of

bilateral lung metastases. Group 1 (n=5) had no treatment and served as the

control group. Group 2 (n=10) received IV treatment with GCB 160mg/kg at

day 7.

At day 0, group 3, 4 and 5 underwent a 10-minute occlusion of the right

pulmonary artery during IV injection of 2.0 × 106 tumour cells for induction

of unilateral left pulmonary metastases as described above. Group 3 (n=4)

had no further treatment and served as the control group. On day 7, group 4

(n=7) and 5 (n=7) underwent left ILuP with GCB 320 mg/kg and 480 mg/kg

respectively.

All animals were followed up until death. In order to confirm unilateral or

bilateral micrometastases, a lung biopsy was taken at day 7. Two hemoclips

were placed at the edge of the lung in order to perform a wedge resection.

Biopsies for confirmation of bilateral micrometastases were taken by a

staged thoracotomy with a time interval between the right and left

thoracotomy of two hours.

Results

Experiment 1: Chronic pulmonary toxicity of gemcitabine in rats

Survival rates of experiment 1 were described in a former paper of our

laboratory [15]. Briefly, survival rates for animals treated with left ILuP 20,

40, 80 mg/kg GCB and buffered starch were 100% and 83% for animals

which had left ILuP 160 and 320 mg/kg GCB. Animals treated with IV 40,

80 and 160 mg/kg GCB survived for 100%, 50% and 60% while animals

treated with IV 320 mg/kg GCB died within one week after infusion.

Chapter 6

106

Histopathological examination showed a significant slight increase of

fibrosis in left lung of animals treated with left ILuP 160 and 320 mg/kg

GCB compared to their untreated right lung (p=0.002). This increase of

fibrosis compared to their right lung was not significant for animals treated

with ILuP 20, 40, 80 mg/kg GCB (p=0.40) and buffered starch (p=0.11). A

non significant tendency in increase of fibrosis in left lung was observed in

rats which had left ILuP 160 and 320 mg/kg GCB compared to ILuP with

20, 40, 80 mg/kg GCB (p=0.06) and buffered starch (p=0.078). Left ILuP

160 and 320 mg/kg GCB resulted in significantly more fibrosis compared to

IV 160 and 320 mg/kg GCB (p=0.002). IV infusion did not result in a

significant increase in fibrosis compared to the control group (p=0.20 and

p=0.24). No significant difference in fibrosis was seen between left and right

lung of rats which underwent iv infusion (table 1-2, fig. 1). Furthermore,

some diffuse alveolar and interstitial infiltrates were observed.

Chapter 6

107

Figure 1a.

Normal lungtissue (Masson,magnificationx400)

Figure 1b.

Mild fibrosis(Masson,magnificationx400)

Figure 1c.

Moderate fibrosiswith pleuralfibrosis (Masson,magnificationx400)

Chapter 6

108

EFT Normal Lung

Tissue (Figure 1a)

Mild Fibrosis

(Figure 1b)

Moderate

Fibrosis

(Figure 1c)

N N % N % N %

ILuP control 6 4 66.7 2 33.3 0 0

ILuP 40 mg/kg 6 4 66.7 1 16.7 1 16.7

ILuP 80 mg/kg 6 3 50 2 33.3 1 16.7

ILuP 160 mg/kg 6 1 16.7 1 16.7 4 66.7

ILuP 320 mg/kg 6 1 16.7 1 16.7 4 66.7

IV control 6 5 83.3 1 16.7 0 0

IV 40 mg/kg 10 9 90 1 10 0 0

IV 80 mg/kg 10 9 90 1 10 0 0

IV 160 mg/kg 10 10 100 0 0 0 0

IV 320 mg/kg 6 6 100 0 0 0 0

Table 1. Pulmonary fibrosis in the left lungILuP: isolated left lung perfusion; IV: intravenous; N: number of rats.

RIGHT Normal Lung

Tissue (Figure 1a)

Mild Fibrosis

(Figure 1b)

Moderate

Fibrosis

(Figure 1c)

N N % N % N %

ILuP control 6 6 100 0 0 0 0

ILuP 40 mg/kg 6 5 83.3 1 16.7 0 0

ILuP 80 mg/kg 6 3 50 3 50 0 0

ILuP 160 mg/kg 6 6 100 0 0 0 0

ILuP 320 mg/kg 6 6 100 0 0 0 0

IV control 6 6 100 0 0 0 0

IV 40 mg/kg 10 9 90 1 10 0 0

IV 80 mg/kg 10 9 90 1 10 0 0

IV 160 mg/kg 10 9 90 1 10 0 0

IV 320 mg/kg 6 6 100 0 0 0 0

Table 2. Pulmonary fibrosis in the right lungILuP: isolated left lung perfusion; IV: intravenous; N: number of rats

Chapter 6

109

Experiment 2: Efficacy of gemcitabine

As described before, rats treated with IV 320 mg/kg died within 1 week

while 83% survived ILuP with the same dose. Therefore 160 mg/kg was

used to determine IV efficacy and 320 and 480 mg/kg were used to evaluate

efficacy in ILuP. The occurrence of pulmonary micrometastases was

confirmed histologically for the unilateral groups as well as for the bilateral

groups. No significant difference (p=0.87) in median survival was seen

between group 1 (28±2 days, bilateral metastases without iv infusion) and 2

(27±2 days, bilateral metastases with 160 mg/kg). Median survival was

significantly longer (p=0.02) in group 4 (38±4 days, unilateral metastases

with ILuP 320 mg/kg) compared to group 3 (28±2 days, unilateral

metastases without ILuP) (table 3). Five animals in group 5 (unilateral

metastases with ILuP 480 mg/kg) died during ILuP due to acute toxicity

while two animals survived for 34 and 36 days in the ILuP 480 mg/kg group.

Group Treatment Median survival

Group 1 (n=5) bilateral controls 28 ± 2 days

Group 2 (n=10) IV GCB 160 mg/kg 27 ± 2 days

Group 3 (n=4) unilateral controls 28 ± 2 days

Group 4 (n=7) ILuP GCB 320 mg/kg 38 ± 4 days (ip=0.01; up=0.02)

Table 3. Survival after IV injection of CC531S adenocarcinoma cells without(group 1-2) and with concomitant right pulmonary artery occlusion (group3-4); ILuP: isolated left lung perfusion; IV: intravenous; GCB:gemcitabine ; i: Group 4 compared to Group 1

u: Group 4 compared to Groups 2 and 3

Chapter 6

110

Discussion

Pulmonary metastases from colorectal adenocarcinoma are responsible for

56,600 deaths in the United States each year [1]. Despite combination

therapy 5-year survival rate is only 30-40% [2]. Surgical resection only

eradicates manually palpable nodules while microscopic disease remains

unaffected. On the other hand, systemic chemotherapy results in high

systemic toxicity with only low local lung levels. Therefore, a model of ILuP

was developed by Weksler in order to obtain higher local drug

concentrations without systemic toxicity [4]. Many drugs like melphalan,

FUDR, TNF, doxorubicin and cisplatin have been evaluated successfully in

this model in preparation of human studies [7,17,20,21,22-25].

GCB (difluorodeoxycytidine, dFdC) is a relatively new anticancer

nucleoside which is very promising in treatment of non-small cell lung

carcinoma, ovarian carcinoma, soft tissue carcinoma, pancreatic, breast and

colon carcinoma [9-12]. GCB is a pro-drug and must be phosphorylated

intracellularly by deoxycytidine kinase to an active form. Incorporation of

GCB di- and tri-phosphate into the elongating DNA strand makes it unable

for the DNA polymerases to proceed with DNA synthesis [26].

This study investigated survival and pulmonary toxicity after left ILuP with

GCB. Pulmonary toxicity in systemic administration is usually mild and self-

limiting [27]. In literature, some cases of acute respiratory distress syndrome

(ARDS), interstitial pneumonitis and systemic capillary leak syndrome

(SCLS) have been described [14,26,27]. Our study showed some diffuse

interstitial infiltrates in accordance with the case report described by Rosado

[26]. Pulmonary fibrosis in systemic administration has not been reported till

now.

Chapter 6

111

In this study, all animals treated IV with GCB 320mg/kg died within one

week after infusion while 83% survived after ILuP with 320 mg/kg GCB.

However, histological examination showed a significant slight increase in

fibrosis after ILuP with 160 and 320 mg/kg compared to the untreated right

lung (p=0.002) and compared to IV infusion with the same doses (p=0.002).

A non-significant tendency in increase of fibrosis was seen after ILuP 160

and 320 mg/kg compared to ILuP with the lower doses (p=0.06) and the

ILuP control group (p=0.08). There seems to be a dose dependency in level

of fibrosis. Apparently, IV infusion did not result in increase of fibrosis

compared to the IV control group. Fibrotic changes, septal thickening and

interstitial infiltrates were described earlier for melphalan and cisplatin

[7,20,25]. Further research is necessary to determine the clinical

consequences of pulmonary fibrosis.

Furthermore, efficacy of GCB was evaluated in this study. Systemic

administration of 160 mg/kg GCB did not prolong survival while ILuP 320

mg/kg GCB resulted in significantly longer survival compared to the control

group (p=0.02). Another group was treated with ILuP 480 mg/kg. Five of

seven rats died during ILuP due to acute pulmonary toxicity. In literature,

only two studies reported significantly longer survival after ILuP using 2

mg/kg melphalan (p=0.0002) and 6.4 mg/kg doxorubicin (p<0.0001) using

the unilateral model of pulmonary metastases [15,22].

Our results show that GCB is a promising product for the treatment of

pulmonary metastatic colorectal adenocarcinoma. The unique combination

of metabolic properties and mechanistic characteristics suggests that GCB is

likely to be synergistic if combined with other drugs that damage DNA [28].

Therefore, combination of GCB with melphalan should be tested in order to

lower the doses of GCB and melphalan and to decrease pulmonary toxicity.

Chapter 6

112

Van Moorsel et al. described synergistic activity after combination therapy

in vitro using GCB and mitomycin C which is an alkylating agent like

melphalan [29]. This effect is possibly based on the property of GCB to

enhance DNA alkylation by mitomycin C.

In conclusion, GCB has proved to be an effective drug against colorectal

CC531S adenocarcinoma in this pulmonary metastatic rat model of ILuP.

Fibrotic septal thickening after 90 days is slight and seems to be dose

dependent. However, pulmonary toxicity does not seem to be more serious

compared to other effective drugs tested in this model.

References1. Jemel A, Thomas A, Murray T, Thun M. Cancer statistics, CA Cancer J Clin

2002;52(1):23-47.

2. Pastorino U, Buyse M, Friedel G, Ginsberg RJ, Girard P, Goldstraw P, Johnston M,

McCormack P, Pass H, Putnam JB, Toomes H. Long-term results of lung

metastasectomy: prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg

1997;113:37-49.

3. Lanza LA, Putnam JB, Benjamin RS, Roth JA. Response to chemotherapy does not

predict survival after resection of sarcomatous pulmonary metastases. Ann Thorac Surg

1991;51:219-24.

4. Weksler B, Schneider A, Ng B, Burt M. Isolated single-lung perfusion in the rat: an

experimental model. J Appl Physiol 1993;74:2736-9.

5. Wang HY, Port JL, Hochwald SN, Burt ME. Revised technique of isolated lung

perfusion in the rat. Ann Thorac Surg 1995;60:211-2.

6. Nawata S, Abecasis N, Ross HM, Abolhoda A, Cheng H, Sachar KS, Burt ME. Isolated

lung perfusion with melphalan for the treatment of metastatic sarcoma. J Thorac

Cardiovasc Surg 1996;112:1542-8.

7. Hendriks JMH, Van Schil PEY, De Boeck G, Lauwers PRM, Van Oosterom AAT, Van

Marck EAE, Eyskens EJM. Isolated lung perfusion with melphalan and tumour necrosis

factor for metastatic pulmonary adenocarcinoma. Ann Thorac Surg 1998;66:1719-25.

Chapter 6

113

8. Ueda K, Sugi K, Li TS, Saeki K, Nawata S, Esato K. The long-term evaluation of

pulmonary toxicity following isolated lung perfusion with melphalan in the rat.

Anticancer Res 1999;19:141-148.

9. Anderson H, Lund B, Bach F, Thatcher N, Walling J, Hansen HH. Single agent activity

of weekly gemcitabine in advanced non-small cell lung cancer: a phase II study. J Clin

Oncol 1994;12:1821-6.

10. Kornmann M, Butzer U, Blatter J, Beger HG, Link KH. Pre-clinical evaluation of the

activity of gemcitabine as a basis for regional chemotherapy of pancreatic and colorectal

cancer. Eur J Surg Oncol 2000;26:583-7.

11. Carmichael J, Possinger K, Philip P, Beykirch M, Kerr H, Walling J, et al. Advanced

breast cancer: a phase II trial with gemcitabine. J Clin Oncol 1995;13:2731-6.

12. Millward MJ, Rischin D, Toner GC, Bishop JF, Walcher V, Hutton-Potts J, et al.

Activity of gemcitabine in ovarian cancer patients resistant to paclitaxel. Proc Am Soc

Clin Oncol 1995;14:776.

13. Gupta N, Ahmed I, Steinberg H, Patel D, Nissel-Horowitz S, Mehrota B. Gemcitabine-

induced pulmonary toxicity: case report and review of the literature. Am J Clin Oncol

2002;25:96-100.

14. Pavlakis N, Bell DR, Millward MJ, Levi JA. Fatal pulmonary toxicity resulting from

treatment with gemcitabine. Cancer 1997;80:286-91.

15. Van Putte BP, Hendriks JMH, Romijn S, Guetens G, De Bruijn E, Van Schil PEY.

Isolated lung perfusion with gemcitabine in a rat: pharmacokinetics and survival. J Surg

Res 2003;109:118-22.

16. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

17. Van Putte BP, Hendriks JMH, Romijn S, Guetens G, De Boeck G, De Bruijn E, Van

Schil PEY. Single-pass isolated lung perfusion versus recirculating isolated lung

perfusion with melphalan in a rat model. Ann Thorac Surg 2002;74:893-8.

18. Marquet RL, Westbroek DL, Jekel J. Interferon treatment of a transplantable rat colon

adenocarcinoma: importance of tumour site. Int J Cancer 1984;33:689-92.

19. Weksler B, Blumberg D, Lenert JT, Ng B, Fong Y, Burt ME. Isolated Lung Perfusion

with TNF-α in a rat sarcoma lung metastases model. Ann Thorac Surg 1994;58:328-32.

20. Tanaka T, Kaneda Y, Li T, Matsuoka T, Zempo N, Esato K. Digitonin enhances the

antitumour effect of cisplatinum during isolated lung perfusion. Ann Thorac Surg

2001;72:1173-8.

Chapter 6

114

21. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-71.

22. Abolhoda A, Brooks A, Nawata S, Kneda Y, Cheng H, Burt ME. Isolated lung perfusion

with doxorubicin prolongs survival in a rodent model of pulmonary metastases. Ann

Thorac Surg 1997;64:181-4.

23. Braakhuis BJ, Ruiz Van Haperen VW, Boven E, Veerman G, Peters GJ. Schedule-

dependent antitumour effect of gemcitabine in vivo model system. Semin Oncol

1995;22:42-6.

24. Port JL, Ng B, Ellis JL, Nawata S, Lenert JT, Burt ME. Isolated lung perfusion with

FUDR in the rat: pharmacokinetics and survival. Ann Thorac Surg 1996;62:848-52.

25. Ratto GB, Toma S, Civalleri D, Passerone GC, Esposito M, Zaccheo D, Canepa M,

Romano P, Palumbo R, De Cian F, Scarano F, Vannozi M, Spessa E, Fantino G. Isolated

lung perfusion with platinum in the treatment of pulmonary metastases from soft tissue

sarcomas. J Thorac Cardiovasc Surg 1996;112:614-622.

26. Rosado MF, Kett DH, Schein RM, Baraona FJ, Sridhar KS. Severe pulmonary toxicity in

a patient treated with gemcitabine. Am J Clin Oncol 2002;25:31-3.

27. De Pas T, Curigliano G, Franceschielli L, Catania C, Spaggiari L, De Braud F.

Gemcitabine-induced systemic capillary leak syndrome. Ann Oncol 2001;12:1651-2.

28. Plunkett W, Huang P, Gandhi V. Preclinical characteristics of gemcitabine. Anticancer

drugs 1995;6(suppl 6) :7-13.

29. Van Moorsel CJA, Veerman G, Bergman AM, Guechev A, Vermorken JB, Postmus PE,

Peters GJ. Combination chemotherapy studies with gemcitabine. Semin Oncol

1997;24(suppl.2):S7-17-S7-23.

Chapter 7

Combination chemotherapy with gemcitabine in a rat

model of Isolated Lung Perfusion for the treatment of

pulmonary metastases

BP Van Putte1, JMH Hendriks1, S Romijn1, B Pauwels2

and PEY Van Schil1

Departments of Thoracic and Vascular Surgery1and Medical

Oncology2,

University Hospital Antwerp, Edegem, Belgium

Submitted for publication

Chapter 7

118

Abstract

Introduction: Isolated lung perfusion (ILuP) is an experimental surgical

technique for the treatment of lung metastases. Currently, ILuP with a single

agent does not result in complete remission. Because of their different

mechanisms of cytotoxicity we studied the in vivo efficacy of combinations

of gemcitabine (GCB), cisplatin (CDDP) and melphalan (MN) administered

by ILuP.

Material and methods: In order to induce left pulmonary metastases, sixty-

three rats had intravenous injection of 2.0×10E6 viable CC531S colorectal

adenocarcinoma cells with clamping of the right pulmonary artery during ten

minutes. At day 7, except for the control group (n=21), all rats underwent

left ILuP during 25 minutes at a flow rate of 0.5 mL/min using GCB (n=7),

CDDP (n=9), MN (n=7), GCB/CDDP (n=6), GCB/MN (n=6) and

CDDP/MN (n=7). Death due to metastatic disease or sacrifice at day 90 was

the endpoint of the study. Survival analysis was performed by the Kaplan

Meier method and differences in survival assessed by the log rank test.

Statistical significance was defined as p<0.05.

Results: All treated rats lived significantly longer compared to control rats

(p<0.00001) while ILuP GCB/MN was superior compared with all other

groups (p≤0.011). ILuP monotherapy with MN resulted in longer survival

compared to GCB (p=0.017) and CDDP (p=0.01). ILuP MN/CDDP

(p=0.029 vs CDDP; p=0.0038 vs MN) and ILuP GCB/MN (p=0.0007 vs

MN; p=0.0004 vs GCB) resulted in important synergistic actions. No

synergistic activity was observed after ILuP GCB/CDDP.

Conclusion: ILuP monotherapy or combination therapy using GCB, CDDP

or MN resulted in significantly longer survival compared to controls. ILuP

monotherapy with MN and combination therapy with MN/GCB gave the

best survival results.

Chapter 7

119

Introduction

Isolated lung perfusion (ILuP) with chemotherapy is an experimental

surgical technique for the treatment of pulmonary metastases in order to

improve the current mean 5-year survival of approximately 40%. Based on

the finding that cell kill of tumour cells due to chemotherapeutics is

concentration-dependent, ILuP aims to obtain higher local drug levels

without systemic toxicity in order to destroy micrometastases before

resection of manually palpable metastases [1]. Therefore, animal models of

isolated lung perfusion using sarcoma and carcinoma cell lines were set up

to test toxicity, pharmacokinetics and antitumour activity of several drugs

like doxorubicin, TNF-α, melphalan, cisplatin, FUDR and gemcitabine [2-

10]. Former studies in our laboratory showed significantly longer survival in

rats with lung metastases which had ILuP monotherapy using gemcitabine

and melphalan compared to control rats [6,10]. Although significantly longer

survival was achieved using ILuP monotherapy, none of these drugs resulted

in complete remission.

Gemcitabine (GCB) is a deoxycytidine analogue with proven clinical

activity in ovarian, breast, colorectal, pancreas, bladder cancer, small cell

lung cancer (SCLC) and non-small cell lung cancer (NSCLC) [11,12]. After

entering the cell, GCB is phosphorylated and incorporated into the DNA.

Subsequently, one nucleotide has to be added until DNA polymerase is not

capable any more to proceed (“masked chain termination’’). Mainly S-phase

cells proved to be sensitive to GCB. Beside actions on DNA, GCB has some

self-potentiating mechanisms within the cell. Because of its low toxicity

profile and its mechanistic and metabolic properties GCB seems to be a very

interesting drug for combination therapy.

Chapter 7

120

Cisplatin (CDDP) is a platinum compound and has proven clinical activity

against many of earlier mentioned types of cancer. CDDP does not only

affect the DNA but the mitochondrial RNA, phospholipid membranes and

the cytoskeleton as well. Melphalan (MN) is an alkylating agent and it is not

cell phase specific resulting in DNA-interstrand, or DNA-protein cross-links

causing inhibition of DNA synthesis.

Because of their different mechanisms of cytotoxicity we studied the in vivo

efficacy of combinations of GCB, CDDP and MN for the treatment of

pulmonary metastatic colorectal adenocarcinoma in a rat model of ILuP.

Material and methods

Animals

Male inbred Wag-Rij strain rats (weight, approximately 250g), obtained

from Iffa Credo (Charluff River, Belgium), were used for all experiments.

Animals were treated in accordance with the Animal Welfare Act and the

“Guide for the Care and Use of Laboratory Animals” (NIH Publication 86-

23, revised 1985). The rats were transported in sterile conditions, housed in

suspended mesh wired cages under standard laboratory conditions and ad

libitum fed a standard pellet diet (standard rat chow, Hope Farms, Woerden,

The Netherlands). The experimental protocols were approved by the Ethical

Committee of the University of Antwerp.

Cell line

In this study CC531S adenocarcinoma cell line was used to determine the

sensitivity for the cytotoxic effect of GCB.

CC531S was cultured in RPMI-1640 medium, supplemented with 10% fetal

calf serum (Life Technologies, Merelbeke, Belgium). Cultures were

Chapter 7

121

maintained in exponential growth in a humidified atmosphere at 37°C under

5% CO2/95% air.

Left Isolated Lung Perfusion

Left ILuP was performed according to the technique described by Wang et

al. [5]. This was modified in our laboratory and full details have been

reported previously by Hendriks et al. [13-15]. Briefly, rats were

anaesthetized with isoflurane in a mixture of nitrous oxide (NO2) and oxygen

(O2). Isoflurane was administered in a concentration of 4% and the NO2:O2

ratio was 3:1. After 5 minutes, rats were intubated with a 16-gauge Insyte-W

catheter by translaryngeal illumination under direct vision and afterwards

ventilated with a volume-controlled ventilator (Harvard Rodent Ventilator,

South Natick, MA) [13]. Once connected to the ventilator, the NO2:O2 ratio

was set to 1:1 (0.5 L/min) and the rate of ventilation was 75 strokes/min with

a tidal volume of 10 mL/kg. Subsequently, the left chest was shaved and

prepared with a 70% alcohol solution followed by a left thoracotomy

between the third and fourth rib. After placing a rib retractor, the left lung

was luxated anteriorly and the hilum was dissected free under microscopic

view (x16 magnification; Carl Zeiss, Belgium). The pulmonary artery and

vein were clamped with curved microvascular clips (Kleinert-Kurz

WK65145). A PE-10 perfusion catheter (Becton Dickinson, Bornem,

Belgium) was introduced into the chest through a 16-G Angiocath, which

was placed through the chest wall. The PE-10 catheter was inserted into the

pulmonary artery and fixed with a 4/0 silk tie. Perfusate was delivered

through this catheter. The effluent was collected at the venous side by a

catheter in proximity of the venotomy. The left thorax was flushed with 2

mL saline every five minutes.

Chapter 7

122

In all experiments, rats were perfused during 25 minutes followed by a 5-

minute washout with buffered starch using a roller-pump at a rate of 0.5

mL/min. At completion of the perfusion, the arteriotomy was repaired with

10-0 nylon suture (Ethilon, Ethicon, Dilbeek, Belgium). No venotomy repair

was performed but pressure was applied with gauze placed over the lung to

control the bleeding from the venotomy. The thoracotomy incision was

closed in layers after introducing a 16-gauge catheter connected to a 50-mL

syringe into the left chest cavity. When animals recovered, the chest tube

and endotracheal tubes were removed.

Gemcitabine (Difluorodeoxycytidine, Eli Lilly, Brussels, Belgium),

melphalan (Alkeran, 50 mg/vial, Wellcome, Waterloo, Belgium) and

cisplatin (cis-dichlorordiammineplatinum, CDDP, Bristol-Myers, Waterloo,

Belgium) solutions were prepared by reconstituting lyophilized powder in

the supplied diluent and performing appropriate dilutions with buffered

hydroxy-ethyl starch (Haes Steril, BHE) prior to the experiments. The

optimal dose of gemcitabine, cisplatin and melpalan to be administered by

ILuP was determined in former experiments [1,5,16]

Induction of unilateral pulmonary metastases

At day 0, sixty-three rats were infused with 2.0x10E6 viable CC531S

adenocarcinoma cells through the left femoral vein while clamping the right

pulmonary artery during ten minutes for induction of left pulmonary

metastases as described in previous papers [5,10]. Briefly, anaesthesia was

induced as described before. After a right thoracotomy, the rib retractor was

placed in the third intercostal space. After luxating the right upper and

middle lobe posteriorly, the right pulmonary artery was dissected free under

microscopic view (x16 magnification; Carl Zeiss, Belgium). Subsequently, a

longitudinal incision was made along the posterior border of the superior

Chapter 7

123

caval vein and the right main pulmonary artery was identified. Finally, an

occluding curved microvascular clamp (Kleinert-Kurz WK65145) was

placed over the main pulmonary artery. The rat was repositioned in order to

get access to the left femoral vein for infusion of the tumour cells. The clamp

on the right pulmonary artery was removed after ten minutes. The groin

incision was subsequently closed with a running suture. After full lung

reexpansion, the chest was closed in three layers with 4-0 vicryl and the

animal was allowed to recover from anaesthesia.

Statistical analysis

Survival curves were constructed by the Kaplan Meier method and

differences in survival assessed by the Log Rank test using SPSS version

11.0 for Windows (Chicago, Ill). Significance was determined as p<0.05.

Experiment: survival after isolated lung perfusion using

melphalan, gemcitabine and cisplatin (fig. 1)

At day 0, sixty-three rats were infused with 2.0x10E6 viable CC531S

adenocarcinoma cells through the left femoral vein while clamping the right

pulmonary artery during ten minutes for induction of left pulmonary

metastases as described before. At day 7, all rats were randomized into 7

groups. Groups 1 to 3 underwent ILuP with 320 mg/kg GCB (n=7), 3 mg/kg

CDDP (n=9) and 2 mg/kg MN (n=7) respectively. Groups 4 to 6 had ILuP

using 320 mg/kg GCB combined with 3 mg/kg CDDP (n=6), 320 mg/kg

GCB combined with 2 mg/kg MN (n=6) and 3 mg/kg CDDP combined with

2 mg/kg MN (n=7) respectively. Group 7 was the untreated control group

(n=21) (fig.1). The end point of the study was death due to metastatic

disease.

Chapter 7

124

N Day 0 Day 7 Endpoint

N=7 Left ILuP GCB

N=7 Left ILuP MN

N=9 Left ILuP CDDP

N=6 Left ILuP GCB+MN

N=6 Left ILuP GCB+CDDP

N=7 Left ILuP CDDP+MN

N=21

IV infusion of 2.0x10E6

viable CC531S

adenocarcinoma cells

while clamping the right

pulmonary artery

No treatment

Death due to

metastatic

disease or

sacrifice at

day 90

Figure 1. Experimental setting.ILuP: isolated lung perfusion; IV : intravenous; GCB: gemcitabine (320mg/kg); MN: melphalan (2 mg/kg); CDDP: cisplatin (3 mg/kg)

Results

All rats that had ILuP monotherapy or combination therapy lived

significantly longer compared to untreated control animals (p<0.00001) .

Rats that underwent ILuP with MN had significantly longer survival

compared to ILuP with GCB (p=0.0017) and to ILuP with CDDP (p=0.01).

No significant difference was observed in survival between ILuP using GCB

or CDDP (p=0.096). Combination ILuP using GCB/MN resulted in

significantly longer survival compared to all other groups (p≤0.011). 67% Of

the rats in the ILuP GCB/MN group survived the follow-up period of 90

days while no rats did in other groups. Combination ILuP using MN/CDDP

resulted in significantly longer survival compared with CDDP (p=0.029) and

with MN (p=0.0038).

Chapter 7

125

survival after ILuP

0 20 40 60 80

GCB

CDDP

MN

MN/CDDP

GCB/CDDP

GCB/MN

controls

trea

tmen

t

days

Figure 2. Mean survival after Isolated Lung Perfusion using single-agenttherapy and combinations of gemcitabine, melphalan and cisplatin.

ILuP GCB CDDP MN GCB/MN GCB/CDDP CDDP/MN No ILuP

N 7 9 7 6 6 7 21

Mean ± SE (days) 38±1 42±3 44±2 79±7 47±2 56±4 28±1

% survival at day 90 0 0 0 67 0 0 0

P-value vs GCB/MN 0.0004 0.0015 0.0007 0.0016 0.011 <0.00001

Table 1. Survival after ILuP mono or combination therapy (SE: standarderror).

Significantly longer survival was seen after combination ILuP using

GCB/CDDP compared with GCB (p=0.0004) and with MN/CDDP

(p=0.028) while no longer survival was observed compared with CDDP

(p=0.36) (figure 2 and table 1).

Chapter 7

126

Discussion

In order to improve efficacy of isolated lung perfusion (ILuP) for the

treatment of pulmonary metastases, this study investigated synergistic

actions in tumour cytotoxicity using combinations of gemcitabine (GCB),

cisplatin (CDDP) and melphalan (MN). Survival was determined in a rat

model of pulmonary metastatic colorectal adenocarcinoma using CC531S

adenocarcinoma cells. One week after induction of micrometastases in the

left lung, rats were treated with left ILuP and survival until death due to

metastatic disease was compared with untreated control animals.

ILuP is based on the concentration dependency of cytotoxicity of

chemotherapeutic drugs [1]. Therefore, it aims to deliver higher local drug

levels without systemic leakage compared to intravenous treatment in order

to destroy micrometastases and therefore to prevent recurrences. Animal

studies investigating efficacy of ILuP with several drugs like GCB, MN,

FUDR, TNF, doxorubicin and CDDP were promising but did not result in

complete remission with exception of a single rat [2-10]. However, MN was

the most successful drug tested which resulted in a currently ongoing phase I

trial at the University Hospital Antwerp (Antwerp, Belgium) and the

Antonius Hospital (Nieuwegein, the Netherlands) including 16 patients until

June 2003. Furthermore, four phase I trials have been described discussing

doxorubicin, CDDP and TNF [17-20]. Although it was not the aim of these

studies to evaluate efficacy, results were disappointing. An explanation for

these disappointing results could be the invasive nature of this technique

which limits repetitive treatment. Furthermore, all phase I trials evaluated

single-agent therapy while the currently used intravenous treatment for

cancer often applies combination chemotherapy [21].

Because of their different mechanisms of cytotoxicity, synergistic actions

were investigated in the present study using combinations of GCB, MN and

Chapter 7

127

CDDP. ILuP with each single drug resulted in significantly longer survival

compared to control animals. Cells were stored under standardized

circumstances. Survival of rats treated with GCB were comparable with

former results in our laboratory while CDDP was only tested in sarcoma

cells before [9,10].

Theoretically, GCB and CDDP interact at different levels from cellular

uptake until incorporation into the DNA. GCB might interact with either the

uptake of CDDP or the binding of CDDP to DNA. Otherwise, CDDP could

interact with the cellular uptake, the phosphorylation or incorporation into

DNA [22]. Van Moorsel et al. concluded from in vitro studies that the effect

of GCB on CDDP accumulation or DNA platination might be important

factors in this synergistic action. The best synergistic effect was observed

when GCB was administered four hours before CDDP [23]. This finding

explains that only a non-significant synergistic tendency was found in the

present study in which both drugs were delivered simultaneously (p=0.0004

compared with GCB; p=0.36 compared with CDDP). However, repetitive

administration might be possible in less invasive models as blood flow

occlusion without clamping of the pulmonary veins or by intravenous pre-

treatment with GCB (15).

The mechanism of cytotoxicity of GCB is totally different from MN. MN

can be activated to an alkylating agent resulting in DNA crosslinks while

GCB is incorporated into the DNA after phosphorylation to GCB

triphosphate [24]. Van Moorsel et al. suggested that GCB might enhance the

DNA alkylation by mitomycin C which is also an alkalyting agent while

mitomycin C did not have any effect on GCB triphosphate DNA

incorporation [24]. In contrast to these small synergistic actions between

GCB and alkalyting agents, the present study shows an important synergistic

action using GCB combined to MN (p=0.0007 compared with MN ;

Chapter 7

128

p=0.0004 compared with GCB). However, one rat was in complete

histologic remission at sacrifice.

Finally, ILuP with MN and CDDP was evaluated resulting in significant

synergistic action (p=0.029 compared with CDDP; p= 0.0038 compared with

MN).

In conclusion, ILuP monotherapy with GCB, CDDP or MN enhances

survival of rats with pulmonary metastases while MN is the strongest agent

evaluated. Synergistic actions in survival were observed after ILuP with

GCB/MN and MN/CDDP while GCB/MN is the strongest combination

tested. Therefore, we advocate a phase I trial with ILuP using GCB

combined with MN for the treatment of lung metastases.

Acknowledgements

The authors want to thank A. Van Laer for excellent technical assistance

during all in vivo experiments. Furthermore we appreciate the kind supply of

gemcitabine by Eli Lilly Benelux (Brussels, Belgium).

References1. Van Putte BP, Hendriks JMH, Romijn S, Pauwels B, Friedel G, Guentens G, De Bruijn

E, Van Schil PEY. Isolated lung perfusion with gemcitabine in a rat: pharmacokinetics

and survival. J Surg Res 2003;109:118-22.

2. Weksler B, Lenert J, Ng B, Burt M. Isolated single lung perfusion with doxorubicin is

effective in eradicating soft tissue sarcoma lung metastases in a rat model. J Thorac

Cardiovasc Surg 1994;107:50-4.

3. Abolhoda A, Brooks A, Nawata S, Kaneda Y, Cheng H, Burt ME. Isolated lung

perfusion with doxorubicin prolongs survival in a rodent model of pulmonary

metastases. Ann Thorac Surg 1997;64:181-4.

4. Weksler B, Blumberg D, Lenert JT, Ng B, Fong Y, Burt ME. Isolated single-lung

perfusion with TNF-α in a rat sarcoma lung metastases model. Ann Thorac Surg

1994;58:328-32.

Chapter 7

129

5. Hendriks JMH, Van Schil PEY, De Boeck G, Van Oosterom A, Van Marck E, Eyskens

EJM. Isolated lung perfusion with melphalan and tumour necrosis factor for pulmonary

adenocarcinoma. Ann Thorac Surg 1998;66:1719-25.

6. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-71.

7. Nawata S, Abecasis N, Ross HM, Abolhoda A, Cheng H, Sachar KS, Burt ME. Isolated

lung perfusion with melphalan for the treatment of metastatic pulmonary sarcoma. J

Thorac Cardiovasc Surg 1996;112:1542-8.

8. Ng B, Lenert JT, Weksler B, Port JL, Ellis JL, Burt ME. Isolated lung perfusion with

FUDR is an effective treatment for colorectal adenocarcinoma lung metastases in rats.

Ann Thorac Surg 1995;59:205-8.

9. Tanaka T, Kaneda Y, Li T, Matsuoka T, Zempo N, Esato K. Digitonin enhances the

antitumour effect of cisplatin during isolated lung perfusion. Ann Thorac Surg

2001;72:1173-8.

10. Van Putte BP, Hendriks JMH, Vermeulen PB, Romijn S, Van Marck E, Van Schil PEY.

Isolated lung perfusion with gemcitabine prolongs survival in a rat model of metastatic

pulmonary adenocarcinoma. Submitted.

11. Storniolo AM, Enas NH, Brown CA, Voi M, Rothenberg ML, Schilsky R. An

investigational new drug treatment program for patients with gemcitabine: results for

over 3000 patients with pancreatic carcinoma. Cancer (Phila.)1999;85:1261-8.

12. Sandler A, Ettinger DS. Gemcitabine: single-agent and combination therapy in non-small

cell lung cancer. Oncologist 1999;4:241-51.

13. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

14. Van Putte BP, Hendriks JMH, Romijn S, Guetens G, De Boeck G, De Bruijn EA, Van

Schil PEY. Single-pass isolated lung perfusion versus recirculating isolated lung

perfusion with melphalan in a rat model. Ann Thorac Surg 2002;74:893-8.

15. Van Putte BP, Hendriks JMH,Romijn S, Pauwels B; De Boeck B, Guetens G; De Bruijn

E, Van Schil PEY. Pharmacokinetics after Pulmonary Artery Perfusion with

Gemcitabine. Ann Thorac Surg 2003. In press.

16. Li TS, Kaneda Y, Saeki K, Ueda K, Zempo N, Esato K. Pharmacokinetic difference

between rat tumour and lung tissues following isolated lung perfusion with cisplatin. Eur

J Cancer 1999;35:1846-50.

Chapter 7

130

17. Pass HI, Mew DJY, Kranda KC, Tmeck BT, Donington JS, Rosenberg SA. Isolated lung

perfusion with tumour necrosis factor for pulmonary metastases. Ann Thorac Surg

1996;61:1609-17.

18. Burt ME, Liu D, Abolhoda A, Ross HM, Kaneda Y, Jara E, Casper ES, Ginsberg RJ,

Brennan MF. Isolated lung perfusion for patients with unresectable metastases from

sarcoma: a phase I trial. Ann Thorac Surg 2000;69:1542-9.

19. Ratto GB, Toma S, Civalleri D, Passerone GC, Esposito M, Zaccheo D, Canepa M,

Romano P, Palumbo R, De Cian F, Scarano F, Vannozzi M, Spessa E, Fantino G.

Isolated lung perfusion with platinum in the treatment of pulmonary metastases from soft

tissue sarcomas. J Thorac Cardiovasc Surg 1996;112:614-22.

20. Putnam JB. New and evolving treatment methods for pulmonary metastases. Sem Thorac

Cardiovasc Surg 2002;14:49-56.

21. Hobday TJ, Goldberg RM. Perspectives on the role of sequential or combination

chemotherapy for first-line and salvage therapy in advanced colorectal cancer.

22. Peters GJ, Bergman AM, Ruiz Van Haperen VWT, et al. Interaction between cisplatin

and gemcitabine in vitro and in vivo. Semin Oncol 1995;22(suppl11):72-79.

23. Van Moorsel CJA, Veerman G, Bergman AM, Guechev A, Vermorken JB, Postmus PE,

Peters GJ. Combination chemotherapy studies with gemcitabine. Semin Oncol 1997;24

(suppl7):S1-17-S7-23.

24. Dorr RT, Bowden GT, Alberts TS, et al. Interactions of mitomycin C with mammalian

DNA detected by alkaline alution. Cancer Res 1985;45:3510-16.

General discussion

General discussion

134

General discussion

Cancer remains the second cause of death with almost 550 000 deaths

expected for the USA in 2002 [1]. Currently, 40-50% of all cases of

malignancy is cured [2]. The non-cured group mostly suffers from

metastases. At the moment of diagnosis of the primary tumour, metastases

which are too small to be detected, so called micrometastases, may already

be present. Furthermore, widespread metastatic disease may be symptomless

at an early stage.

With exception of the lymph nodes, the lungs are the most common site of

metastatic involvement for all invasive cancer types. The most obvious

reason is the filtering of circulating tumour cells by the pulmonary capillary

bed. The precise incidence of pulmonary metastases is unknown, but has

been estimated to occur in up to 50% of patients with non-pulmonary

malignancy [2].

The cumulative incidence and prevalence of pulmonary metastases vary

widely according to tumour type and disease-free time interval. Some

tumour types preferentially metastasise to the lungs, such as sarcomas, germ

cell tumours and some paediatric malignancies. Gastrointestinal tumours will

first metastasise to the liver and from there to the lungs. The high incidence

of pulmonary metastases in autopsy studies may overestimate both the true

incidence of initial lung involvement and its clinical significance.

Current treatment of pulmonary metastases consists of surgical resection and

adjuvant chemotherapy resulting in a 5-year survival of 30-40% [3]. Surgical

treatment only resects manually palpable metastases while micrometastases

remain untouched and finally result in recurrences. Systemic chemotherapy

fails because of genetic drug resistance and the inability to achieve sufficient

drug levels in lung and tumour tissue. So, novel techniques should be

General discussion

135

developed in order to improve prognosis. It was Creech in 1959 who first

reported a model of isolated lung perfusion to achieve high local drug levels

without apparent systemic toxicity [4]. Later on, this model was transposed

to the rat by Weksler and modified by Wang and subsequently by Hendriks

[5-7]. In preparation of clinical trials several drugs were tested like

melphalan, FUDR, TNF, doxorubicin and cisplatin [8-15].

Part of this PhD thesis aimed to investigate the application of gemcitabine

(GCB) for the treatment of pulmonary metastatic colorectal adenocarcinoma

in a rat model of isolated lung perfusion (ILuP). Clinical activity of single-

agent treatment with gemcitabine is currently described for ovarian, breast,

colorectal, breast, pancreas and non-small cell lung cancer [16-17].

Concluding from our studies, GCB is an effective drug against colon

adenocarcinoma in vitro showing time- and concentration-dependent cell kill

[18,19]. Basically, concentration-dependent cell kill is the main argument for

application of isolated lung perfusion resulting in significantly higher drug

levels compared to the currently used intravenous (IV) chemotherapeutic

treatment. However, studying an equilibrium between achieving high local

drug levels and avoiding major pulmonary toxicity on the other hand, is

essential for optimal efficacy. Furthermore, isolated lung perfusion with

GCB resulted in significantly higher lung levels and less acute toxicity when

compared to IV treatment. The maximally tolerated dose after ILuP was 320

mg/kg and 160 mg/kg after IV infusion [18]. Therefore, the pharmacokinetic

benefit of ILuP has been confirmed and is in agreement with results obtained

before in rat studies using melphalan and doxorubicin [8,11,13].

Pulmonary toxicity in systemic administration of GCB is usually mild and

self-limiting [20]. In literature, some cases of acute respiratory distress

syndrome (ARDS), interstitial pneumonitis and systemic capillary leak

General discussion

136

syndrome (SCLS) have been described [20-22]. Our chronic

histopathological studies of rat lungs which underwent ILuP with GCB

showed mild and dose-dependent pulmonary fibrosis of the interstitial and

pleural tissue while obviously no fibrosis was observed in the IV treated and

control rats [23]. Pulmonary fibrosis in ILuP or after systemic administration

has not been reported till now. Furthermore, some diffuse interstitial

infiltrates were observed which are in accordance with the case report

described by Rosado [22]. Further investigations studying functional

pulmonary capacity are planned to determine possible significant side-

effects.

Efficacy of GCB was determined using the unilateral model of pulmonary

(micro)metastases. After induction of unilateral micrometastases at day 0,

rats were treated after one week with ILuP using the MTD of 320 mg/kg of

GCB. This resulted in significantly longer survival compared to untreated

control animals which were not treated. In literature, only two studies

reported significantly longer survival after ILuP with 2 mg/kg melphalan

(p=0.0002) and 6.4 mg/kg doxorubicin (p<0.0001) using the unilateral

model of pulmonary metastases [9,12].

However, even better results were described in rats treated with melphalan

(84 days survival after inducing micrometastases and ILuP with melphalan

versus 39 days after ILuP with GCB) [12,23].

Although no animal models or ex vivo models can exactly define human

tolerance to any drug, four clinical phase I trials have been performed using

tumour necrosis factor, platinum and doxorubicin respectively [24-27].

Except for one study by Ratto et al., all patients had inoperable pulmonary

metastatic disease and perfusions were performed in the antegrade way. A

recent study supported the antegrade way of perfusion by showing that a

General discussion

137

significant portion of both pulmonary and metastatic tumour vasculature is

fed by the pulmonary arterial circulation [28]. All rat studies selected

antegrade single-pass perfusion for technical reasons and because this

method facilitates drug kinetic studies by ensuring a constant drug

concentration in the perfusate. Technically seen, it is very demanding to

cannulate the pulmonary veins of the rat and even more difficult to restore

the venotomies after perfusion in case of survival experiments. However a

recirculating perfusion system was preferred for clinical trials because it

minimizes the amount of perfusate and drug needed, vascular isolation

becomes more complete and greater flexibility is obtained in regulation of

the perfusion environment. In contrast to the single-pass perfusion technique,

the drug was added as a bolus into the perfusion circuit but due to the

continuous uptake of the drug into the lung, the concentration in the

perfusate decreases in function of time. The fact that drug concentration was

crucial in the rat experiments for the final lung concentration might explain

the large variations in perfusate and lung levels of doxorubicin between

patients receiving the same amount of drug [25]. It also shows the difficulty

of calculating the amount of drug to give each patient. In contrast with

isolated limb perfusion, where doses are calculated by the amount of water

displaced by the limb, no such value is available for lung perfusion studies

[29]. However, as mentioned before, previous animal studies in our

laboratory did not show significant differences in final lung and tissue

melphalan levels between single-pass perfusion and reperfusion [30]. These

findings will simplify isolated lung perfusion with melphalan because

melphalan can be injected as a bolus into the circuit, after stabilization of the

main values, such as temperature, flow, and leakage.

General discussion

138

Based on promising experimental results from our laboratory, a phase I trial

has been started at the University Hospital Antwerp (Antwerp, Belgium) in

cooperation with the St. Antonius Hospital (Nieuwegein, the Netherlands).

Increasing doses of melphalan are administered using ILuP in patients

undergoing resection of lung metastases which are melphalan sensitive.

Until June 2003, 16 patients have been treated in this trial without major

morbidity or mortality. Comparison of results of these clinical trials with the

currently used intravenous treatment is difficult as lung and tumour levels

during intravenous chemotherapeutic treatment are not available.

Furthermore, the impact of flow, retrograde perfusion, ventilation and

sequential perfusion remain unclear.

Although efficacy was not the primary aim of the four described clinical

phase I trials which studied doxorubicin, cisplatin and tumour necrosis factor

in ILuP, survival results were disappointing [24-27]. This can be explained

by the invasive nature of this technique which limits repetitive treatment and

the fact that in most studies ILuP was applied in patients with irresectable

lung metastases. Moreover, all clinical trials evaluated single-agent therapy

while the currently used intravenous treatment for cancer often applies

combination chemotherapy [31].

Before the lungs are involved, colorectal cancer preferentially metastasises

to the liver. Beside surgery and systemic chemotherapy for the treatment of

liver metastases, new experimental techniques are under investigation [32].

In order to improve the current 5-year survival of liver metastases, Savier et

al. recently described a feasibility study of percutaneous isolated hepatic

perfusion (IHP) with melphalan in humans; three courses were given with

time intervals of three and six weeks [33]. During the therapy period, an

implantable device connected to a silicone catheter remained in situ. The

General discussion

139

device was implanted subcutaneously while the catheter was inserted into

the gastroduodenal artery. For each course a double-balloon catheter was put

into the caval vein and a single-balloon catheter into the portal vein for

isolation of the liver circulation. Further studies are necessary to investigate

the clinical impact of repetitive percutaneous treatment on survival of liver

metastases. However, double-balloon techniques are very hard to apply in

the pulmonary circulation because of the complicated anatomic access of the

pulmonary veins. We hypothesized that repetitive, less invasive treatment of

pulmonary metastases can only be achieved by selective pulmonary artery

perfusion during blood flow occlusion without control of the pulmonary

veins.

Therefore, we investigated the first-pass effect of gemcitabine during low

flow pulmonary artery perfusion (0.2 ml/min) using blood flow occlusion of

the pulmonary artery without control of the pulmonary veins in preparation

of less invasive methods of selective perfusion of the lung. Low flow

pulmonary artery perfusion with GCB resulted in significantly lower

systemic GCB serum levels and significantly higher lung levels compared to

intravenous infusion [19]. This emphasizes the feasibility of chemotherapy

infusion of gemcitabine by sequential selective catheterisation of the

pulmonary artery using blood flow occlusion without control of the

pulmonary veins.

Because of their different mechanisms of cytotoxicity, synergistic actions

were investigated in the present study using combinations of GCB, MN and

CDDP. ILuP with each single drug resulted in significantly longer survival

compared to control animals. Survival of rats treated with GCB were

comparable with former results in our laboratory while CDDP was only

tested in sarcoma cells before [15,23].

General discussion

140

Theoretically, GCB and CDDP interact at different levels from cellular

uptake until incorporation into the DNA. GCB might interact with either the

uptake of CDDP or the binding of CDDP to DNA. Otherwise, CDDP could

interact with the cellular uptake, the phosphorylation or incorporation into

DNA [35]. Van Moorsel et al. concluded from in vitro studies that the effect

of GCB on CDDP accumulation or DNA platination might be important

factors in this synergistic action. The best synergistic effect was observed

when GCB was administered four hours before CDDP [36]. This finding

explains that only a non significant synergistic tendency was found in the

present study in which both drugs were delivered simultaneously (p=0.0004

compared with GCB; p=0.36 compared with CDDP). However, repetitive

administration might be possible in less invasive models as blood flow

occlusion without clamping of the pulmonary veins or by intravenous

pretreatment with GCB (19).

The mechanism of cytotoxicity of GCB is totally different from MN. MN

can be activated to an alkylating agent resulting in DNA crosslinks while

GCB is incorporated into the DNA after phosphorylation to GCB

triphosphate [37]. Van Moorsel et al. suggested that GCB might enhance the

DNA alkylation by mitomycin C which is also an alkalyting agent while

mitomycin C did not have any effect on GCB triphosphate DNA

incorporation [37]. In contrast to these small synergistic actions between

GCB and alkalyting agents, the present study shows a important synergistic

action using GCB combined to melphalan (p=0.0007 compared with MN,

p=0.0004 compared with GCB). However, one rat was in complete

histologic remission at sacrifice.

General discussion

141

Finally, ILuP with MN and CDDP was evaluated resulting in significant

synergistic action (p=0.029 compared with CDDP; p= 0.0038 compared with

MN).

Pulmonary ischemia/reperfusion (IR) injury as a result of ILuP, lung emboli,

cardiac surgery, shock or transplantation was recently investigated in our

laboratory by differentiating cellular infiltrates and quantifying lung oedema,

apoptosis and necrosis after one hour of warm ischemia followed by 30

minutes, 1, 2, 3 and 4 hours of reperfusion [38]. ILuP was applied as a flush

procedure in this model. Currently, the role of the neutrophil and

macrophage in occurrence of pulmonary IR injury is quite controversial

while the role of the T-cell is even unknown [39-43]. An important increase

in number of neutrophils was observed after 0.5-4 hours of reperfusion and

after reperfusion and flushing as well which might be an indication for

adhesion and activation of these cells. Macrophages doubled in lung tissue

after IR while a biphasic response was observed for neutrophils. This is the

first study that showed a possible involvement (a four times increase) of T-

cells in lung tissue after 1 hour of warm ischemia and 30 minutes of

reperfusion. Furthermore, apoptosis in total absence of necrosis was

observed together with important alveolar oedema [38]. Further studies are

planned to investigate if this inflammatory response is reversible and antigen

specific .

In conclusion, isolated lung perfusion with gemcitabine for the treatment of

pulmonary metastases results in significantly higher lung levels without

systemic toxicity and significantly longer survival compared to systemic

treatment. Synergistic actions have been observed with GCB in combination

with cisplatin and melphalan even resulting in complete remission.

Furthermore, an important first-pass effect has been shown during selective

General discussion

142

pulmonary artery perfusion without clamping the pulmonary veins. Finally,

the reperfusion technique will simplify isolated lung perfusion with

melphalan because melphalan can be injected as a bolus into the circuit, after

stabilization of the main values, such as temperature, flow, and leakage.

Future research

Feasibility of less invasive models of isolated lung perfusion for the

treatment of pulmonary microscopic and macroscopic metastatic disease has

to be investigated. It should be emphasized that less invasive models will

offer the possibility of repetitive application.

After validation of the current results obtained from animal studies, new

drugs and combination of drugs have to be evaluated.

Alternatives to isolated lung perfusion or chemoembolization should be

explored further. Pharmacokinetics of drugs or non-toxic prodrugs that have

proved to be effective in animal models should be investigated in ex-vivo

models.

Based on experimental animal results, clinical phase I, II and eventually

phase III trials have to be designed. Specific attention should be paid to the

optimal dose or concentration administered to the patient in order to obtain a

constant drug level in the lung and tumour tissue. Methods to increase drug

uptake in metastatic cells should be improved.

Finally, clinical parameters should be determined that predict specific

patterns of metastatic development. By applying significant results of these

trials, isolated lung perfusion will find its way to definite clinical application.

In this way, ultimate prognosis in patients with lung metastases will

hopefully be improved.

General discussion

143

References1. Jemel A, Thomas A, Murray T, Thun M. Cancer statistics. Ca Cancer J Clin,

2002;52:23-47.

2. Koehne C-H, Cunningham D, Di Costanzo, Glimelius B, Blijham G, Aranda E,

Scheithauer W, Rougier P, Palmer M, Wils J, Baron B, Pignatti F, Schoeffski P, Micheel

S, Hecker H. Clinical determinants of survival in patients with 5-fluorouracil-based

treatment for metastatic colorectal cancer: results of a multivariate analysis of 3825

patients. Ann Oncol 2002;10:308-17.

3. Pastorino U, Buyse M, Friedel G, Ginsberg RJ, Girard P, Goldstraw P, Johnston M,

McCormack P, Pass H, Putnam JB. Long-term results of lung metastasectomy:

prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37-49.

4. Creech O, Krementz ET, Ryan RF, Winblad JN. Chemotherapy of cancer: regional

perfusion utilizing an extracorporeal circuit. Ann Surg 1959;148:616-32.

5. Weksler B, Schneider A, Ng B, Burt M. Isolated lung perfusion in the rat: an

experimental model. J Appl Physiol 1993;74:2739-9.

6. Wang HY, Port JL, Hochwald SN, Burt ME. Revised technique of isolated lung

perfusion in the rat. Ann Thorac Surg 1995;60:211-2.

7. Hendriks JMH, Van Schil PEY, Eyskens EJM. Modified technique of isolated left lung

perfusion in the rat. Eur Surg Res 1999;31:93-96.

8. Weksler B, Lenert J, Ng B, Burt M. Isolated single lung perfusion with doxorubicin is

effective in eradicating soft tissue sarcoma lung metastases in a rat model. J Thorac

Cardiovasc Surg 1994;107:50-4.

9. Abolhoda A, Brooks A, Nawata S, Kaneda Y, Cheng H, Burt ME. Isolated lung

perfusion with doxorubicin prolongs survival in a rodent model of pulmonary

metastases. Ann Thorac Surg 1997;64:181-4.

10. Weksler B, Blumberg D, Lenert JT, Ng B, Fong Y, Burt ME. Isolated single-lung

perfusion with TNF-α in a rat sarcoma lung metastases model. Ann Thorac Surg

1994;58:328-32.

11. Hendriks JMH, Van Schil PEY, De Boeck G, Van Oosterom A, Van Marck E, Eyskens

EJM. Isolated lung perfusion with melphalan and tumor necrosis factor for pulmonary

adenocarcinoma. Ann Thorac Surg 1998;66:1719-25.

General discussion

144

12. Hendriks JMH, Van Schil PEY, Van Oosterom AAT, Kuppen PJK, Van Marck E,

Eyskens E. Isolated lung perfusion with melphalan prolongs survival in a rat model of

metastatic pulmonary adenocarcinoma. Eur Surg Res 1999;31:267-71.

13. Nawata S, Abecasis N, Ross HM, Abolhoda A, Cheng H, Sachar KS, Burt ME. Isolated

lung perfusion with melphalan for the treatment of metastatic pulmonary sarcoma. J

Thorac Cardiovasc Surg 1996;112:1542-8.

14. Ng B, Lenert JT, Weksler B, Port JL, Ellis JL, Burt ME. Isolated lung perfusion with

FUDR is an effective treatment for colorectal adenocarcinoma lung metastases in rats.

Ann Thorac Surg 1995;59:205-8.

15. Tanaka T, Kaneda Y, Li T, Matsuoka T, Zempo N, Esato K. Digitonin enhances the

antitumor effect of cisplatin during isolated lung perfusion. Ann Thorac Surg

2001;72:1173-8.

16. Storniolo AM, Enas NH, Brown CA, Voi M, Rothenberg ML, Schilsky R. An

investigational new drug treatment program for patients with gemcitabine: results for

over 3000 patients with pancreatic carcinoma. Cancer (Phila.)1999;85:1261-8.

17. Sandler A, Ettinger DS. Gemcitabine: single-agent and combination therapy in non-small

cell lung cancer. Oncologist 1999;4:241-51.

18. Van Putte BP, Hendriks JMH, Romijn S, Pauwels B, Friedel G, , De Bruijn E, Van Schil

PEY. Isolated lung perfusion with gemcitabine in a rat: pharmacokinetics and survival. J

Surg Res 2003;109:118-22.

19. Van Putte B, Hendriks JMH, Romijn S, Guentens G, De Bruijn E, Van Schil PEY.

Pharmacokinetics after pulmonary artery perfusion with gemcitabine. Ann Thorac Surg

2003, in press.

20. De Pas T, Curigliano G, Franceschielli L, Catania C, Spaggiari L, De Braud F.

Gemcitabine-induced systemic capillary leak syndrome. Ann Oncol 2001;12:1651-2.

21. Pavlakis N, Bell DR, Millward MJ, Levi JA. Fatal pulmonary toxicity resulting from

treatment with gemcitabine. Cancer 1997;80:286-91.

22. Rosado MF, Kett DH, Schein RM, Baraona FJ, Sridhar KS. Severe pulmonary toxicity in

a patient treated with gemcitabine. Am J Clin Oncol 2002;25:31-3.

23. Van Putte BP, Hendriks JMH, Pauwels B, Vermeulen P, Van Marck E, Van Schil PEY.

Isolated Lung Perfusion with gemcitabine prolongs survival in a rat model of metastatic

pulmonary adenocarcinoma with slight long-term fibrosis. Submitted.

General discussion

145

24. Pass HI, Mew DJY, Kranda KC, Tmeck BT, Donington JS, Rosenberg SA. Isolated lung

perfusion with tumour necrosis factor for pulmonary metastases. Ann Thorac Surg

1996;61:1609-17.

25. Burt ME, Liu D, Abolhoda A, Ross HM, Kaneda Y, Jara E, Casper ES, Ginsberg RJ,

Brennan MF. Isolated lung perfusion for patients with unresectable metastases from

sarcoma: a phase I trial. Ann Thorac Surg 2000;69:1542-9.

26. Ratto GB, Toma S, Civalleri D, Passerone GC, Esposito M, Zaccheo D, Canepa M,

Romano P, Palumbo R, De Cian F, Scarano F, Vannozzi M, Spessa E, Fantino G.

Isolated lung perfusion with platinum in the treatment of pulmonary metastases from soft

tissue sarcomas. J Thorac Cardiovasc Surg 1996;112:614-22.

27. Putnam JB. New and evolving treatment methods for pulmonary metastases. Sem Thorac

Cardiovasc Surg 2002;14:49-56.

28. Mochizuki A, Kurihara Y, Yokote K, Nakajima Y, Osada H. Discrimination of solitary

pulmonary nodules based on vascular supply patterns with first-pass dynamic CT. Lung

cancer 2000;29(suppl 1):242.

29. Lejeune F, Kroon B, Difilippo F, et al. Isolated limb perfusion: the European experience.

Surg Oncol Clin N Am 2001;10:821-32.

30. Van Putte BP, Hendriks JMH, Romijn S, Guentens G, De Bruijn E, Van Schil PEY.

Single-Pass Isolated Lung Perfusion versus Recirculating Isolated Lung Perfusion with

Melphalan in a Rat Model. Ann Thorac Surg 2002;74:893-8.

31. Hobday TJ, Goldberg RM. Perspectives on the role of sequential or combination

chemotherapy for first-line and salvage therapy in advanced colorectal cancer. Clin

Colorectal Cancer. 2002;3:170-2.

32. Ruers T, Bleichrodt RP. Treatment of liver metastases, an update on the possibilities and

results. Eur J Cancer 2002;38:1023-33.

33. Savier E, Azoulay D, Huguet E, Lokiec F, Gil-Delgado M, Bismuth H. Percutaneous

isolated hepatic perfusion for chemotherapy. Arch Surg 2003;138:325-32.

34. Van Putte BP, Hendriks JMH, Romijn S, Pauwels B, Van Schil PEY. Isolated lung

perfusion using combinations of gemcitabine, melphalan and cisplatin. Submitted.

35. Peters GJ, Bergman AM, Ruiz Van HaperenVWT, et al. Interaction between cisplatin

and gemcitabine in vitro and in vivo. Semin Oncol 1995;22(suppl11):72-79.

36. Van Moorsel CJA, Veerman G, Bergman AM, Guechev A, Vermorken JB, Postmus PE,

Peters GJ. Combination chemotherapy studies with gemcitabine. Semin Oncol

1997;24(suppl7):S1-17-S7-23.

General discussion

146

37. Dorr RT, Bowden GT, Alberts TS, et al. Interactions of mitomycin C with mammalian

DNA detected by alkaline alution. Cancer Res 1985;45:3510-16.

38. Van Putte BP, Persy V, Hendriks JMH, Van Marck E, Van Schil PEY, De Broe ME.

Cellular infiltrates and injury evaluation in a rat model of warm ischemia-reperfusion.

Submitted.

39. Kuhnle GEH, Reichenspurner H, Lange T, Wagner F, Groh J, Messmer K, et al.

Microhemodynamics and leukocyte sequestration after pulmonary ischemia and

reperfusion in rabbits. J Thorac Cardiovasc Surg 1998;115:937-44.

40. Takeyoshi I, Iwanami K, Kamoshita N, Takahashi T, Kobayashi J, Tomizawa N,

Kawashima Y, Matsumoto K, Morishita Y. Effect of lazaroid U-74389 on pulmonary

ischemia-reperfusion injury in dogs. J Invest Surg 2001;14:83-92.

41. Thomas DD, Sharar SR, Winn RK, Chi EY, Verrier ED, Allen MD, Bishop MJ. CD18-

Independent mechanism of neutrophil emigration in the rabbit lung after ischemia-

reperfusion. Ann Thorac Surg 1995;60:1360-6.

42. Eppinger MJ, Deen GM, Bolling SF, et al. Mediators of ischemia reperfusion injury of

rat lung. Am J Pathol 1997;150:1773-84.

43. Fiser SM, Tribble CG, Long SM, Kaza AK, Kern JA, Kron IL. Pulmonary macrophages

are involved in reperfusion injury after lung transplantation. Ann Thorac Surg

2001;71:1134-9.

Summary

150

SummaryIsolated lung perfusion (ILuP) is an experimental surgical technique for the

treatment of lung metastases in order to improve the 5-year survival of

approximately 40% after surgical resection. Based on in vitro studies that

show cell kill to be concentration dependent, ILuP as an adjuvant treatment

before surgical resection of manually palpable metastases aims to achieve

significantly higher lung and tumour drug levels without apparent systemic

toxicity in order to destroy micrometastases for prevention of recurrences.

However, repetitive treatment is quite unattractive due to its invasive nature.

This PhD thesis aims to review the literature of this research area (chapter 1

and 2) and to investigate in vitro and in vivo toxicity (chapter 3),

pharmacokinetics (chapter 4) and efficacy (chapter 6) of gemcitabine (GCB)

using a rat model of left ILuP. Furthermore, synergistic actions of GCB were

studied using combinations with melphalan and cisplatin (chapter 7).

Finally, two different techniques of isolated lung perfusion (i.e. single-pass

perfusion versus reperfusion) were compared (chapter 5).

In vitro studies using CC531S adenocarcinoma cells exposed to GCB during

different periods of time showed cytotoxic activity to be concentration- and

time-dependent as well. Concerning (tumour) cell kill as the major target of

chemotherapeutic treatment, these in vitro results imply the importance of

achieving high local drug levels and maintenance of a certain drug level as

well (chapter 3 and 4).

Using a rat model of isolated left lung perfusion, toxicity of GCB in vivo

was investigated in a dose escalating study that showed a maximally

tolerated dose (MTD) of 320 mg/kg after left ILuP and 160 mg/kg after

intravenous infusion. All treated lungs from this rat study were

microscopically investigated after sacrification at 3 months. Only a slight

increase of collagen deposits was observed which was significant in rats

treated with left ILuP 160 and 320 mg/kg while no fibrosis was seen in the

intravenously treated rats (chapter 3 and 6).

The unilateral pulmonary metastatic model was used in order to study in

vivo anti-tumour efficacy of left ILuP with GCB. After induction of

micrometastases in the left lung, left ILuP with GCB resulted in a

Summary

151

significantly longer survival compared to control rats which were not treated

(chapter 6).

GCB seems to be a very interesting drug for combination therapy because of

its low toxicity profile and mechanistic and metabolic properties. Because of

their different mechanisms of cytotoxicity, we studied the in vivo efficacy of

combinations of gemcitabine, cisplatin and melphalan for the treatment of

pulmonary metastatic colorectal adenocarcinoma in our rat model of isolated

lung perfusion. All single drugs and combinations resulted in a significant

elongation of survival compared to controls. However, ILuP with melphalan

and melphalan combined with gemcitabine gave the best survival results

(chapter 7).

In preparation of less invasive techniques of ILuP, a rat model was

developed to obtain selective pulmonary artery perfusion during blood flow

occlusion without clamping the pulmonary veins. An important first-pass

effect of GCB in the lung was observed resulting in significantly lower

systemic GCB serum levels compared to GCB serum levels after intravenous

infusion of a lower dose (chapter 4).

Finally, in preparation of a clinical trial, two different techniques of ILuP

were compared using melphalan. The single-pass (SP) perfusion technique is

used in all rat studies and implies delivery of a constant concentration of the

drug to the lung while the reperfusion (RP) technique is currently used in an

ongoing clinical phase I trial in the University Hospital Antwerp (Antwerp,

Belgium) and the St. Antonius Hospital (Nieuwegein, the Netherlands)

which implies a delivery of a decreasing concentration of the drug used due

to the uptake into the lung. No significant differences in final lung and

tumour levels were observed between SP and RP if saturation occurs. These

findings will simplify isolated lung perfusion with melphalan because

melphalan can be injected as a bolus into the circuit, after stabilization of the

main values, such as temperature, flow, and leakage (chapter 5).

152

SamenvattingGeïsoleerde Longperfusie (ILuP) is een experimentele chirurgische

behandeling voor de behandeling van longmetastasen om de huidige 5-

jaarsoverleving van ongeveer 40% na chirurgische resectie te verbeteren. In

vitro studies toonden aan dat celdood bij blootstelling aan een

chemotherapeuticum concentratieafhankelijk is. Als een adjuvante

behandeling voorafgaand aan chirurgische resectie van manueel palpabele

metastasen, heeft ILuP als doelstelling om significant hogere long- en

tumorconcentraties te bereiken zonder dat systemische toxiciteit optreedt.

Het is de bedoeling om micrometastasen te vernietigen ter voorkoming van

recidieven. Herhaalde behandeling is technisch bijzonder moeilijk omwille

van het invasieve aspect van de behandeling.

Dit proefschrift heeft als doelstelling om een overzicht te geven van de

literatuur op het gebied van ILuP (hoofdstuk 1 en 2) als behandeling van

longmetastasen. Verder worden in vitro en in vivo toxiciteit (hoofdstuk 3),

farmacokinetiek (hoofdstuk 4) en efficaciteit (hoofdstuk 6) van gemcitabine

(GCB) onderzocht waarbij gebruik wordt gemaakt van een rattenmodel van

ILuP. Bovendien wordt synergistische activiteit onderzocht van GCB in

combinatie met cisplatinum en melphalan (hoofdstuk 7). Uiteindelijk worden

twee verschillende technieken (namelijk de single-pass perfusietechniek

versus de reperfusietechniek) vergeleken (hoofdstuk 5).

In vitro studies hebben laten zien dat cytotoxiciteit zowel concentratie- als

tijdsafhankelijk is. Hierbij werd gebruik gemaakt van de CC531S

adenocarcinomacellijn die werd blootgesteld aan GCB in functie van de tijd.

Het bereiken van celdood is de belangrijkste doelstelling van een

chemotherapeutische behandeling. Daarom impliceren deze in vitro

resultaten het belang van het behalen van een hoge lokale concentratie van

Summary

153

een chemotherapeuticum maar vooral ook van het handhaven van deze

concentratie (hoofdstuk 3 en 4).

In vivo toxiciteit van GCB werd onderzocht met behulp van een rattenmodel

van ILuP. Hierbij werd een opbouwend dosisschema gebruikt dat een

maximaal tolereerbare dosis (MTD) liet zien van 320 mg/kg voor

behandeling met ILuP en 160 mg/kg na intraveneuze inspuiting. Na drie

maanden werden alle ratten opgeofferd en werden de longen microscopisch

onderzocht. Slechts een lichte toename van collageendeposities werd gezien

die significant was voor de groepen behandeld met ILuP 160 en 320 mg/kg.

In de intraveneus behandelde groepen werd echter geen fibrose aangetroffen

(hoofdstuk 3 en 6).

Een unilateraal model van pulmonaal gemetastaseerd adenocarcinoma werd

gebruikt voor het bepalen van in vivo anti-tumor activiteit van ILuP van de

linker long met GCB. In het dit model worden micrometastasen geïnduceerd

in de linker long. Een significante verlenging van overleving werd gezien na

ILuP met GCB in vergelijking met controledieren die niet werden behandeld

(hoofdstuk 6).

GCB is een erg interessant product voor combinatietherapie omwille van het

lage toxiciteitsprofiel en de mechanistische en metabole eigenschappen. De

in vivo efficaciteit van GCB werd bestudeerd in combinatie met melphalan

en cisplatin omwille van hun verschillende mechanismen van cytotoxiciteit.

Alle producten die als monotherapie werden gebruikt evenals de

verschillende combinaties resulteerden in een significante verlenging van

overleving in vergelijking met controle dieren. Melphalan monotherapie en

melphalan/gemcitabine combinatietherapie gaven echter de beste

overlevingsresultaten (hoofdstuk 7).

ln voorbereiding op minder invasieve technieken van ILuP werd een

rattenmodel ontwikkeld waarbij de pulmonaalarterie selectief werd

geperfundeerd met gelijktijdige bloedflowocclusie aldaar, terwijl de

pulmonaalvenen niet geoccludeerd werden. Een belangrijk first-pass effect

van GCB ter hoogte van de long werd waargenomen dat resulteerde in

significant hogere longconcentraties en significant lagere serumconcentraties

in vergelijking met intraveneuze inspuiting van een lagere dosis GCB

(hoofdstuk 4).

154

Ten slotte werden twee verschillende technieken van ILuP met melphalan

vergeleken in voorbereiding van een fase I studie. De single-pass

perfusietechniek (SP) wordt gebruikt in de rattenmodellen en impliceert het

afgeven van een constante concentratie van het product terwijl de

reperfusietechniek momenteel gebruikt wordt in een fase I studie in het

Universitair Ziekenhuis Antwerpen (Antwerpen, België) en het Sint

Antonius Ziekenhuis (Nieuwegein, Nederland). Reperfusie leidt tot het

afgeven van een dalende concentratie van het product als gevolg van de

opname door het longweefsel. Geen significante verschillen werden gemeten

tussen SP en RP in long- en tumorweefsel als verzadiging optreedt. Deze

bevindingen vereenvoudigen ILuP met melphalan, omdat melphalan als een

bolus kan worden toegevoegd aan het circuit na stabilisatie van de

belangrijkste variabelen zoals temperatuur, flow en systemische lekkage

(hoofdstuk 5).

Acknowledgements

Acknowledgements

158

First of all, I want to thank prof. dr. Paul Van Schil and dr. Jeroen Hendriks.

They offered me the opportunity to cooporate in their laboratory work while

I was still a medical student. Throughout these years they gave me a lot of

intellectual freedom, but at the same time they continuously helped me with

their comments and suggestions, interpretation of the data, structuring the

presentations and writing the different articles. Furthermore, I want to thank

prof. dr. Aart Brutel de la Rivière. He initially accepted me as a PhD student

and later on he gave me the opportunity to start with my training in

cardiothoracic surgery by the first of October 2003.

During the summer of 1999, I started working with dr. Jeroen Hendriks

while he was finishing his final experiments for his PhD thesis. I was very

lucky to start working with him because he set up and modified the rat

model of isolated lung perfusion some years before and has learnt me a lot of

different anaesthetic and surgical techniques in the rat.

From the beginning I have worked with my dear friend and colleague drs.

Sander Romijn. I want to thank him for the cooperation in the lab, for

joining me during the trips to several congresses and for the understanding

and support that were necessary to combine study, research and friendship.

Most of the experiments have been done by evening or night. This implied a

lot of understanding, flexibility, support and timing of the laboratory

assistant August Van Laer and the logistic head of the laboratory prof. dr.

Dirk Ysebaert. Thank you very much for this.

For almost all the experiments many lung and serum specimens had to be

processed and analyzed for measuring the gemcitabine levels and for further

microscopic investigations. Therefore, I want to thank prof. dr. Ernst De

Bruijn, dr. Gunther Guetens and dr. Gert De Boeck from the oncology

Acknowledgements

159

laboratory of the Katholic University Leuven and dr. Peter Vermeulen,

Frank Rylant and prof. dr. Erik Van Marck from the pathology laboratory of

the University of Antwerp.

During the different experiments I tried to link in vitro and in vivo results to

strongen the final conclusions. All the in vitro work was done by dr. Bea

Pauwels in cooperation with dr. Annelies Korst, prof. Dr. Filip Lardon and

prof. dr. Jan Vermorken. Thank you all.

In 2001, a clinical phase I trial with melphalan started at the University

Hospital Antwerp in cooperation with the Antonius Hospital (Nieuwegein,

the Netherlands). An important issue in this research area still remains how

to extrapolate data from the rat model to a clinical human model. Therefore

we visited dr. Godehard Friedel from the Schillerhoehe Klinik (Gerlingen,

Germany) who had developed an ex vivo model using human lungs and

lobes after pneumonectomy or lobectomy respectively in order to study

pharmacokinetics during isolated lung perfusion. I thank dr. Godehard

Friedel and Renate Vlaeminck for the still continuouing cooperation.

In the winter of 2002, I have got the opportunity to study pulmonary

ischemia-reperfusion injury. During the months following, I learnt some

immunohistochemistric and microscopic techniques with help of prof. dr.

Paul Van Schil, prof. dr. Marc De Broe, dr. Veerle Percy and Bertie Zoete.

Thank you for this.

I am furthermore indebted to dr. Patricia Bakker, dr. Erik Jansen, dr. Jaap

Lahpor and dr. Steve Woolley for accepting me as a trainee in cardiothoracic

surgery at the University Medical Centre in Utrecht. I have appreciated the

layout advices of drs. Tjander Maikoe and Josephine Walta very much.

Acknowledgements

160

I am specially indebted to my parents and sister Marileen for their

continuous support, advices and understanding. Thank you for this.

Last but not least, Nienke, thank you for your patience and understanding. I

know last year was a hard period for both of us starting a new job in new

surroundings while I was continuously working in the hospital, at home

behind my computer or during the weekends in Antwerp. It would have been

impossible without you.

Culemborg, August 2003

Curriculum Vitae

• Born at January 23th 1977, in Maasland

• Gymnasium, Sint Stanislas College, Delft (1989-1995)

• Candidate MD, University of Antwerp, Belgium, magna cum laude

(1995-1998)

• MD, University of Antwerp, Belgium, magna cum laude (1998-2002)

• Assistant cardiothoracic surgery, UMC Utrecht (2002-now)

Publications

1. J. Hendriks, S. Romijn, B. Van Putte , J. Vermorken, E. Van Marck, P.

Van Schil. Surgery for lung metastases: a retrospective analysis. Acta

Chir Belg 2001;101:267-72.

2. B. P. Van Putte, J. Hendriks, S. Romijn, G. Guentens, E. De Bruijn, P.

Van Schil. Single-Pass Isolated Lung Perfusion versus Recirculating

Isolated Lung Perfusion with Melphalan in a Rat Model. Ann Thorac

Surg 2002;74:893-8.

3. J. Hendriks, B. Van Putte , S. Romijn, J. Van Den Branden, J.B.

Vermorken, P. Van Schil. Pneumonectomy for of lung metastases: report

of ten cases. Thorac Cardiovasc Surg 2003;51(1):38-41.

4. S. Romijn, J. Hendriks, B. Van Putte, E. De Bruijn, P. Van Schil.

Variations in flow, duration and concentration do not change the final

lung concentration of melphalan after isolated lung perfusion in rats. Eur

Surg Res 2003;35(1)50-3.

5. B. P. Van Putte, J. M. H. Hendriks, S. Romijn, B. Pauwels, G. Friedel,

G. Guentens, E. De Bruijn, P. E. Y. Van Schil. Isolated lung perfusion

with gemcitabine in a rat: pharmacokinetics and survival. J Surg Res

2003;109:118-22.

Curriculum Vitae

164

6. B. Van Putte, S. Romijn, J. Hendriks, E. De Bruijn, P. Van Schil.

Pharmacokinetics after pulmonary artery perfusion with gemcitabine. Ann

Thorac Surg 2003, in press.

7. B. Van Putte , J. Hendriks, S Romijn, P. Van Schil. Isolated lung

perfusion for the treatment of pulmonary metastases: current mini review

of work in progress. Surg Oncol 2003, in press.

8. C. J. Vrints, C. Moerenhout, B. P. Van Putte, J. Bosmans, M. Claes.

Long (≥25 mm) slotted tube stents for long coronairy artery leasions. A

save conduct or a bridge too far for the interventional cardiologist??

Submitted.

9. B. Van Putte , J. Hendriks, B. Pauwels, P. Vermeulen, E. Van Marck, P.

Van Schil. Toxicity and efficay of Isolated Lung Perfusion with

gemcitabine in a rat model of metastatic pulmonary adenocarcinoma.

Submitted.

10. B. P. Van Putte, V. Persy, J Hendriks, E Van Marck, P. Van Schil, M.

De Broe. Cellular infiltrates and injury evaluation in a rat model of warm

ischemia-reperfusion. Submitted.

11. B.P. Van Putte and P. Bakker. Subtotal stenosis of the anonymous vein

after unsuccessful pacemaker implantation for resynchronization therapy.

Submitted.

12. B. Van Putte, J. Hendriks, S. Romijn, B. Pauwels, JB Vermorken, P.

Van Schil. Isolated lung perfusion using combinations of gemcitabine,

melphalan and cisplatin. Submitted.

Curriculum vitae

165

Abstracts of Oral Presentations

1. B. Van Putte , S. Romijn, J. Hendriks, P. Van Schil. Isolated lung

perfusion: single pass versus recirculating perfusion using two different

perfusates. MMSC, Maastricht, March 2001.

2. J. Hendriks, B. Van Putte , S. Romijn, P. Van Schil. A retrospective

analysis of treatment of pulmonary metastases. Nederlandse

Thoraxvereniging (NVT), Nieuwegein, June 2001. Netherlands Heart J

2001;9:209.

3. J. Hendriks, B. Van Putte , S. Romijn, P. Van Schil. Pulmonary

metastasectomy: a 10-year experience. Belgium Association of Thoracic

Surgeons (BACTS), Brussels, Decembre 2001.

4. J. Hendriks, B. Van Putte , S. Romijn, G. Guentens, E. De Bruijn, P.

Van Schil. Isolated lung perfusion: single-pass system versus

recirculating perfusion with melphalan in rats. Society of Thoracic

Surgery (STS), Fort Launderdale, USA, January 2002.

5. B. Van Putte , J. Hendriks, S. Romijn, B. Pauwels, G. Friedel, G.

Guentens, P. Van Schil. Isolated Lung Perfusion with gemcitabine in a

rat: pharmacokinetics and survival. 37th Congress of the European

Society for Surgical Research (ESSR), Szeged, Hungary, May 2002. Eur

Surg Res 2002;34(1):55.

6. S. Romijn, J. Hendriks, B. Van Putte , G. Guentens, E. De Bruijn, P.

Van Schil. Variations in flow, duration and concentration do not change

the final lung concentrations of melphalan after isolated lung perfusion

in rats. 37th Congress of the European Society for Surgical Research

(ESSR), Szeged, Hungary, May 2002. Eur Surg Res 2002;34(1):55.

7. B. P. Van Putte, J. Hendriks, S. Romijn, B. Pauwels, P. Van Schil.

Survival and pharmacokinetics of gemcitabine in a rat model of isolated

Curriculum Vitae

166

lung perfusion. European Respiratory Society (ERS), Annual Congress

2002. Eur Resp J 2002;20(suppl 38):179s.

8. S. Romijn, J. Hendriks, B. Van Putte , P. Van Schil. European

Respiratory Society (ERS), Annual Congress 2002. Eur Resp J

2002;20(suppl 38):179s.

9. B. P. Van Putte, J. M. Hendriks, S. Romijn, B. Pauwels, P.E.Van Schil.

In vitro and in vivo toxicity of gemcitabine in a rat model of isolated

lung perfusion using CC531S cells, 1st International CC531S

symposium, Maastricht (NL), June 2002.

10. S. Romijn, J. Hendriks, B. Van Putte , P. Van Schil. Pharmacokinetics

of melphalan. 1st International CC531S symposium, Maastricht (NL),

June 2002.

11. J. Hendriks, S. Romijn, B. Van Putte , P. Van Schil. Phase I trial with

melphalan in treatment of pulmonary metastases.

Grens(z)landsymposium (Leuven), September 13th 2002.

12. B. P. Van Putte, V. Persy, J Hendriks, P. Van Schil, M. De Broe. New

insights in warm ischemia-reperfusion lung injury. Evolving basic

concepts on ischemic injury, 20-21 September 2002, Antwerp

(Belgium).

13. J Hendriks, B. P. Van Putte, S. Romijn, B. Pauwels, P. Van Schil.

Gemcitabine prolongs survival in a rat model of metastatic pulmonary

adenocarcinoma. Eur Soc Cardiothoracic Surg, 29-30 October 2002,

Istanbul (Turkey).

14. .B. P. Van Putte , J. M. Hendriks, S. Romijn, P.E.Van Schil.

Gemcitabine prolongs survival in a rat model of metastatic pulmonary

adenocarcinoma. Ned Vereniging Voor Heelkunde, 1 November 2002,

Utrecht (The Netherlands).

Curriculum vitae

167

15. B. P. Van Putte, S. Romijn, J Hendriks, P. Van Schil. Gemcitabine

prolongs survival in a rat model of metastatic pulmonary

adenocarcinoma SEOHS, November 15th, 2002, Rotterdam (the

Netherlands).

16. S. Romijn, B. P. Van Putte, J Hendriks, P. Van Schil. SEOHS,

November 15th, 2002, Rotterdam (the Netherlands).

17. B. P. Van Putte, J. M. Hendriks, B. Pauwels, G. Guetens, P.E.Van

Schil. Pharmacokinetics of gemcitabine in a rat model of metastatic

pulmonary adenocarcinoma. STS 2003, San Diego, CA, USA, 30th Jan

2003.

18. B. P. Van Putte, J Hendriks, S. Romijn, A. Brutel de la Rivière, P. Van

Schil. The first-pass effect of gemcitabine during selective pulmonary

artery perfusion using blood Flow Occlusion. Nederlandse

Thoraxvereniging (NVT), Utrecht, April 12th 2003.

19. B. P. Van Putte, J Hendriks, S. Romijn, P. Van Schil. The first-pass

effect of gemcitabine during selective pulmonary artery perfusion using

blood Flow Occlusion. 38th ESSR, Ghent, Belgium, May 2003. Eur Surg

Res 2003;35:190.

Presentaties (invited speaker)

1. B. Van Putte. Isolated lung perfusion using gemcitabine: toxicity,

pharmacokinetics and efficacy. European Respiratory Society (ERS),

research seminar of isolated lung perfusion. Antwerp, April 28th 2002.

2. B. Van Putte. Warm ischemia-reperfusion injury in lung tissue.

European Respiratory Society (ERS), research seminar of isolated lung

perfusion. Antwerp, April 28th 2002.