Lotti Marco MD - Cancer of the Oesophago-Gastric Junction

CarcinomaoftheGastricCardia:Transhiatal Approach

MarcoLottiMDDivisionofAdvancedSurgicalOncology

DepartmentofGeneralSurgery1PapaGiovanniXXIIIHospital– Bergamo- Italy

17

XXVII CONGRESSO NAZIONALE SPIGC

SOCIETÀ POLISPECIALISTICA ITALIANA DEI GIOVANI CHIRURGHI

Brescia, 11-13 giugno 2015

PROGRAMMA PRELIMINARE

PROGRAMMA SCIENTIFICO

11 - 13 giugno 2015

©2015MarcoLotti– [email protected]

RecurrencewithsurgeryalonePercentage of Siewert I Siewert II Siewert IIIPerioperative Deaths 3- 5 2.5 -5 5– 6.5Recurrenceat2and5years 45- 70 55 - 75 65- 72• haematogenous 55 55 30 - 45• local/anastomotic 35 30 25• peritoneal 7 15 35 -55• lymphnode 20 30 25

- coeliacaxis 40 25 /- porta 30 33 /- retrocrural/aortocaval 10 25 80- cervical/mediastinal 30 25 20


TheRandomizedTrials


• TRANSHIATALESOPHAGECTOMY:thetumoranditsadjacentlymphnodesweredissectedenbloc.A3-cm-widegastrictubewasconstructed.Theleftgastricarterywastransectedatitsorigin,withresectionoflocallymphnodes.Celiaclymphnodesweredissectedonlywhentherewasclinicalsuspicionofinvolvement.Anastomosisintheneck.

• TRANSTHORACICESOPHAGECTOMY:thethoracicduct,azygos vein,ipsilateral pleura,andallperiesophageal tissueinthemediastinumweredissectedenbloc.Theparacardial,lesser-curvature,left-gastric-artery(alongwithlesser-curvature),celiactrunc,common-hepatic-artery,andsplenic-arterynodesweredissected,andagastrictubewasconstructed.

• Amean(±SD)of16±9nodeswereidentifiedintheresectionspecimenaftertranshiatalresection,and31± 14aftertransthoracicresection(p<0.001).

1662

·

N Engl J Med, Vol. 347, No. 21

·

November 21, 2002

·

www.nejm.org

The New England Journal of Medicine

EXTENDED TRANSTHORACIC RESECTION COMPARED WITH LIMITED TRANSHIATAL RESECTION FOR ADENOCARCINOMA OF THE ESOPHAGUS

J

AN

B.F. H

ULSCHER

, M.D., J

OHANNA

W.

VAN

S

ANDICK

, M.D., A

NGELA

G.E.M.

DE

B

OER

, P

H

.D.,B

AS

P.L. W

IJNHOVEN

, M.D., J

AN

G.P. T

IJSSEN

, P

H

.D., P

AUL

F

OCKENS

, M.D., P

EEP

F.M. S

TALMEIER

, P

H

.D., F

IEBO

J.W.

TEN

K

ATE

, M.D., H

ERMAN

VAN

D

EKKEN

, M.D., H

UUG

O

BERTOP

, M.D., H

UGO

W. T

ILANUS

, M.D.,

AND

J. J

AN

B.

VAN

L

ANSCHOT

, M.D.

A

BSTRACT

Background

Controversy exists about the best sur-gical treatment for esophageal carcinoma.

Methods

We randomly assigned 220 patients withadenocarcinoma of the mid-to-distal esophagus or ad-enocarcinoma of the gastric cardia involving the dis-tal esophagus either to transhiatal esophagectomy orto transthoracic esophagectomy with extended en bloclymphadenectomy. Principal end points were overallsurvival and disease-free survival. Early morbidity andmortality, the number of quality-adjusted life-yearsgained, and cost effectiveness were also determined.

Results

A total of 106 patients were assigned to un-dergo transhiatal esophagectomy, and 114 to undergotransthoracic esophagectomy. Demographic charac-teristics and characteristics of the tumor were similarin the two groups. Perioperative morbidity was higherafter transthoracic esophagectomy, but there was nosignificant difference in in-hospital mortality (P=0.45).After a median follow-up of 4.7 years, 142 patientshad died — 74 (70 percent) after transhiatal resectionand 68 (60 percent) after transthoracic resection (P=0.12). Although the difference in survival was not sta-tistically significant, there was a trend toward a surviv-al benefit with the extended approach at five years:disease-free survival was 27 percent in the transhiatal-esophagectomy group, as compared with 39 percentin the transthoracic-esophagectomy group (95 per-cent confidence interval for the difference, ¡1 to 24percent [the negative value indicates better survivalwith transhiatal resection]), whereas overall survivalwas 29 percent as compared with 39 percent (95 per-cent confidence interval for the difference, ¡3 to 23percent).

Conclusions

Transhiatal esophagectomy was asso-ciated with lower morbidity than transthoracic esoph-agectomy with extended en bloc lymphadenectomy.Although median overall, disease-free, and quality-adjusted survival did not differ statistically betweenthe groups, there was a trend toward improved long-term survival at five years with the extended transtho-racic approach. (N Engl J Med 2002;347:1662-9.)

Copyright © 2002 Massachusetts Medical Society.

From the Departments of Surgery (J.B.F.H., J.W.S., H.O., J.J.B.L.), MedicalPsychology (A.G.E.M.B., P.F.M.S.), Cardiology (J.G.P.T.), Gastroenterology(P.F.), and Pathology (F.J.W.K.), Academic Medical Center, University ofAmsterdam, Amsterdam; the Departments of Surgery (B.P.L.W., H.W.T.) andPathology (H.D.), Erasmus University Hospital Rotterdam, Rotterdam; andRADIAN and Medical Technology Assessment (P.F.M.S.), Nijmegen — allin the Netherlands. Address reprint requests to Dr. van Lanschot at the Aca-demic Medical Center at the University of Amsterdam, Department of Sur-gery, Suite G4-112, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands,or at [email protected].

ONG-TERM survival after surgery with cur-ative intent for adenocarcinoma of the distalesophagus and gastric cardia is only 20 per-cent.

1,2

Surgery is generally considered to of-fer the best chance for cure, but opinions differ on howto improve survival by surgery. One strategy aims atdecreasing early postoperative risk by the use of lim-ited cervicoabdominal (transhiatal) esophagectomywithout formal lymphadenectomy. Another is in-tended to improve long-term survival by performinga combined cervicothoracoabdominal resection, withwide excision of the tumor and peritumoral tissues andextended lymph-node dissection in the posterior medi-astinum and the upper abdomen (transthoracic esoph-agectomy with extended en bloc lymphadenectomy).

1-5

We studied whether transthoracic esophagectomywith extended en bloc lymphadenectomy sufficientlyimproves overall, disease-free, and quality-adjusted sur-vival over the rates with transhiatal esophagectomy tocompensate for the possibly higher perioperative mor-bidity and mortality and the increased costs of thetreatment.

METHODS

Study Design

The study was performed in two academic medical centers, eachperforming more than 50 esophagectomy procedures per year. Theeligible patients had histologically confirmed adenocarcinoma of themid-to-distal esophagus or adenocarcinoma of the gastric cardia in-volving the distal esophagus, had no evidence of distant metastases(including the absence of histologically confirmed tumor-positivecervical lymph nodes and unresectable celiac lymph nodes), anddid not have unresectable local disease. These patients were random-ly assigned to undergo transhiatal esophagectomy or transthoracicesophagectomy with extended en bloc lymphadenectomy betweenApril 1994 and February 2000.

Patients had to be older than 18 years of age and in adequatephysical condition to undergo surgery (as indicated by their assign-ment to American Society of Anesthesiologists class I or II

6

). Ex-clusion criteria were previous or coexisting cancer, previous gastricor esophageal surgery, receipt of neoadjuvant chemotherapy or ra-

L

The New England Journal of Medicine Downloaded from nejm.org on May 25, 2015. For personal use only. No other uses without permission.

Copyright © 2002 Massachusetts Medical Society. All rights reserved.

2 Clinical Classification Systems of Adenocarcinoma of the Esophagogastric Junction 25

2.2.4 Surgical Consequences

type I II III

Fig. 2.5

Differentlymphadenectomyinthemediastinum

andevenintheabdomen©2015MarcoLotti– [email protected]

whereas the pathologists decided in favor of type II relativelyoften. To our knowledge, this issue has never been system-atically studied, but this problem is well known from clinical

practice. Despite the substantial discrepancy between theendoscopic and pathologic classification, the randomizationand stratification process has resulted in a well-balanced

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

391624334043475473110TTE28101721323539526995THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

391624334043475473110TTE28101721323539526995THE

Numbers at risk

THE

TTE

FIGURE 2. Overall survival of all patients aftertranshiatal (drawn line) or transthoracic (dottedline) esophagectomy (P ! 0.71) based on perprotocol analysis and after exclusion of patientswho did not undergo surgical resection.

Test for interaction:Lokalisation:

# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06

Outcome is overall survival

Outcome is locoregional recurrence

Type I (n=90)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



Type II (n=115)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



0 (n=55)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



1-8 (n=104)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



9+ (n=46)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



Overall (n=205)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



Type I (n=90)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



Type II (n=115)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



0 (n=55)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



1-8 (n=104)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



9+ (n=46)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



Overall (n=205)


# positive nodes:

Lokalisation:

# positive nodes:

p=0.39

p=0.07

p=0.55

p=0.06



HR TTE better THE better

0.2 0.4 0.6 0.8 1 1.25 1.75 2.5 5

Location

Location

>8 (n=46)

>8 (n=46)

FIGURE 3. Subgroup analysis with test for inter-action for overall survival and locoregional re-currence.

Omloo et al Annals of Surgery • Volume 246, Number 6, December 2007

© 2007 Lippincott Williams & Wilkins996

ORIGINAL ARTICLES

Extended Transthoracic Resection Compared With LimitedTranshiatal Resection for Adenocarcinoma of the Mid/Distal

EsophagusFive-Year Survival of a Randomized Clinical Trial

Jikke M. T. Omloo, MD,* Sjoerd M. Lagarde, MD,* Jan B. F. Hulscher, MD,*Johannes B. Reitsma, MD, PhD,† Paul Fockens, MD, PhD,‡ Herman van Dekken, MD, PhD,§

Fiebo J. W. ten Kate, MD,¶ Huug Obertop, MD,! Hugo W. Tilanus, MD, PhD,!and J. Jan B. van Lanschot, MD!

Objective: To determine whether extended transthoracic esopha-gectomy for adenocarcinoma of the mid/distal esophagus improveslong-term survival.Background: A randomized trial was performed to compare surgi-cal techniques. Complete 5-year survival data are now available.Methods: A total of 220 patients with adenocarcinoma of the distalesophagus (type I) or gastric cardia involving the distal esophagus(type II) were randomly assigned to limited transhiatal esophagec-tomy or to extended transthoracic esophagectomy with en bloclymphadenectomy. Patients with peroperatively irresectable/incur-able cancer were excluded from this analysis (n ! 15). A total of 95patients underwent transhiatal esophagectomy and 110 patientsunderwent transthoracic esophagectomy.Results: After transhiatal and transthoracic resection, 5-year sur-vival was 34% and 36%, respectively (P ! 0.71, per protocolanalysis). In a subgroup analysis, based on the location of theprimary tumor according to the resection specimen, no overallsurvival benefit for either surgical approach was seen in 115 patientswith a type II tumor (P ! 0.81). In 90 patients with a type I tumor,a survival benefit of 14% was seen with the transthoracic approach(51% vs. 37%, P ! 0.33). There was evidence that the treatment

effect differed depending on the number of positive lymph nodes inthe resection specimen (test for interaction P ! 0.06). In patients(n ! 55) without positive nodes locoregional disease-free survivalafter transhiatal esophagectomy was comparable to that after trans-thoracic esophagectomy (86% and 89%, respectively). The samewas true for patients (n ! 46) with more than 8 positive nodes (0%in both groups). Patients (n ! 104) with 1 to 8 positive lymph nodesin the resection specimen showed a 5-year locoregional disease-freesurvival advantage if operated via the transthoracic route (23% vs.64%, P ! 0.02).Conclusion: There is no significant overall survival benefit foreither approach. However, compared with limited transhiatal resec-tion extended transthoracic esophagectomy for type I esophagealadenocarcinoma shows an ongoing trend towards better 5-yearsurvival. Moreover, patients with a limited number of positivelymph nodes in the resection specimen seem to benefit from anextended transthoracic esophagectomy.

(Ann Surg 2007;246: 992–1001)

The incidence of adenocarcinoma of the esophagus and gas-troesophageal junction is rapidly rising. It is an aggressive

disease with early lymphatic and hematogenous dissemination.Long-term survival rates barely exceed 25%, even after surgerywith curative intent.1,2 Surgery is still considered the best cura-tive treatment option. However, much controversy concerningthe optimal surgical approach exists.

Two main operation techniques are currently advo-cated. Limited transhiatal esophagectomy (THE) (withoutformal lymphadenectomy) aims at decreasing early postop-erative morbidity and mortality. Whereas extended transtho-racic esophagectomy (TTE) with en bloc lymphadenectomyis intended to improve long-term survival by performing acombined (cervico) thoracoabdominal resection, with wideexcision of the tumor and peritumoral tissues and extendedlymph node dissection in the posterior mediastinum andupper abdomen.

From the *Departments of Surgery, †Clinical Epidemiology, Biostatistics,and Bioinformatics, and ‡Gastroenterology, Academic Medical Center,University of Amsterdam, Amsterdam, The Netherlands; §Department ofPathology, Erasmus Medical Center, Rotterdam, The Netherlands; ¶De-partment of Pathology, Academic Medical Center, University of Amster-dam, Amsterdam, The Netherlands; and !Department of Surgery, Eras-mus Medical Center, Rotterdam, The Netherlands.

Supported by ZonMw Health Care Efficiency Research (945-04-510; toJ. M. T. O.).

Supported by the Maag Lever Darm Stichting (Dutch Digestive Foundation,04-77; to S. M. L.).

Supported by the Dutch Health Care Insurance Funds Council (1996-041; toJ. B. F. H.).

Reprints: Jikke M. T. Omloo, Department of Surgery, Academic MedicalCenter, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam,The Netherlands. E-mail: [email protected].

Copyright © 2007 by Lippincott Williams & WilkinsISSN: 0003-4932/07/24606-0992DOI: 10.1097/SLA.0b013e31815c4037

Annals of Surgery • Volume 246, Number 6, December 2007992

1662

·


·

November 21, 2002

·

www.nejm.org



J

AN

B.F. H

ULSCHER

, M.D., J

OHANNA

W.

VAN

S

ANDICK

, M.D., A

NGELA

G.E.M.

DE

B

OER

, P

H

.D.,B

AS

P.L. W

IJNHOVEN

, M.D., J

AN

G.P. T

IJSSEN

, P

H

.D., P

AUL

F

OCKENS

, M.D., P

EEP

F.M. S

TALMEIER

, P

H

.D., F

IEBO

J.W.

TEN

K

ATE

, M.D., H

ERMAN

VAN

D

EKKEN

, M.D., H

UUG

O

BERTOP

, M.D., H

UGO

W. T

ILANUS

, M.D.,

AND

J. J

AN

B.

VAN

L

ANSCHOT

, M.D.

A

BSTRACT

Background


Methods


Results


Conclusions





1,2


1-5


METHODS

Study Design



6


L



1666 · N Engl J Med, Vol. 347, No. 21 · November 21, 2002 · www.nejm.org


percent) and 57 after transthoracic resection (50 per-cent). Local–regional recurrence occurred in 14 per-cent and 12 percent of patients, respectively; distant re-currence in 25 percent and 18 percent; and both in 18percent and 19 percent (P=0.60). For the transhiataland transthoracic procedures, the median disease-freeinterval was 1.4 years (95 percent confidence inter-val, 0.8 to 2.0) and 1.7 years (95 percent confidenceinterval, 0.7 to 2.7), respectively (P=0.15) (Fig. 1).The estimated rate of disease-free survival at five yearswas 27 percent (95 percent confidence interval, 19 to38 percent) after transhiatal resection, as comparedwith 39 percent (95 percent confidence interval, 30 to48 percent) after transthoracic resection. The 95 per-cent confidence interval for the difference in the rateswas ¡1 percent to 24 percent (the negative value indi-cates that survival was better with transhiatal resection).

At the end of follow-up, 142 patients had died —

74 in the transhiatal-esophagectomy group (70 per-cent) and 68 in the transthoracic group (60 percent;P=0.12). Thirteen patients died of causes unrelatedto cancer. The median overall survival was 1.8 years (95percent confidence interval, 1.2 to 2.4) after transhi-atal resection and 2.0 years (95 percent confidence in-terval, 1.1 to 2.8) after transthoracic resection with ex-tended en bloc lymphadenectomy (P=0.38) (Fig. 2).The estimated rate of overall survival at five years was29 percent (95 percent confidence interval, 20 to 38percent) after transhiatal resection, as compared with39 percent (95 percent confidence interval, 30 to 48percent) after transthoracic resection. The 95 percentconfidence interval for the difference was ¡3 percentto 23 percent. The median number of quality-adjust-ed life-years after transhiatal resection was 1.5 (95 per-cent confidence interval, 0.8 to 2.1), as compared with1.8 (95 percent confidence interval, 1.1 to 2.4) after

Figure 1. Kaplan–Meier Curves Showing Disease-free Survival among Patients Randomly Assigned to TranshiatalEsophagectomy or Transthoracic Esophagectomy with Extended en Bloc Lymphadenectomy.

100

80

60

40

20

0876543210

Years

Transhiatal esophagectomy

Cum

ulat

ive

Dis

ease

-fre

e S

urvi

val (

%)

NO. AT RISK

TranshiatalmesophagectomyTransthoracicmesophagectomy

106

114

68

69

47

53

32

39

20

31

15

20

11

13

4

7

Transthoracic esophagectomy



Fiveyearslaterthesurvivalbenefitislessthanexpected

“Althoughsurvivaldidnotdifferstatisticallybetweenthegroups,therewas

atrendtowardimproved5ysurvivalwiththeextendedtransthoracicapproach.”

“Thereisnosignificantsurvivalbenefitforeitherapproach.However,transthoracic

esoph.fortypeIesophagealadk showsanongoingtrendtowardsbetter5ysurvival.”©2015MarcoLotti– [email protected]

ORIGINAL ARTICLES







(Ann Surg 2007;246: 992–1001)











1662

·


·

November 21, 2002

·

www.nejm.org



J

AN

B.F. H

ULSCHER

, M.D., J

OHANNA

W.

VAN

S

ANDICK

, M.D., A

NGELA

G.E.M.

DE

B

OER

, P

H

.D.,B

AS

P.L. W

IJNHOVEN

, M.D., J

AN

G.P. T

IJSSEN

, P

H

.D., P

AUL

F

OCKENS

, M.D., P

EEP

F.M. S

TALMEIER

, P

H

.D., F

IEBO

J.W.

TEN

K

ATE

, M.D., H

ERMAN

VAN

D

EKKEN

, M.D., H

UUG

O

BERTOP

, M.D., H

UGO

W. T

ILANUS

, M.D.,

AND

J. J

AN

B.

VAN

L

ANSCHOT

, M.D.

A

BSTRACT

Background


Methods


Results


Conclusions





1,2


1-5


METHODS

Study Design



6


L



distribution of the 2 tumor types between the 2 treatmentarms. Therefore, the impact of this potential confounder isprobably limited.

In the present study, none of the patients received(neo)adjuvant chemo- and/or radiotherapy. However, manyWestern institutes have now incorporated neoadjuvant ther-apy in their standard treatment protocols. It is unclearwhether the outcome of the present study would be influencedby the introduction of multimodality therapy. It is conceiv-able that the potential benefit of an extended lymphadenec-tomy is dependent on the number of positive lymph nodes.Therefore, we subdivided our patients into 3 groups: patients

without positive lymph nodes in the resection specimen,patients with 1 to 8 positive nodes, and patients with morethan 8 positive nodes. Theoretically, an artificially low num-ber of positive nodes could be removed by the surgeon andidentified by the pathologist after performing a limited trans-hiatal resection. These patients would be erroneously consid-ered to have a limited number of positive nodes (N ! 1–8),and thus contaminate this subgroup. Even though the totalnumber of resected lymph nodes was higher after an extendedlymphadenectomy,5 the number of positive nodes found inthe resection specimen did not differ significantly betweenboth techniques (Table 1). Therefore, the phenomenon of

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

24101520242426273347TTE145912161920263343THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

24101520242426273347TTE145912161920263343THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

156913161821274063TTE15589161619263652THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

156913161821274063TTE15589161619263652THE

Numbers at risk

THE

TTE

A

B

FIGURE 4. A, Overall survival of patients with typeI adenocarcinoma of the esophagus after transhia-tal (drawn line) or transthoracic (dotted line)esophagectomy (P ! 0.33). B, Overall survival ofpatients with type II adenocarcinoma of the gas-tric cardia after transhiatal (drawn line) or aftertransthoracic (dotted line) esophagectomy (P !0.81).

Annals of Surgery • Volume 246, Number 6, December 2007 Extended Versus Limited Esophagectomy

© 2007 Lippincott Williams & Wilkins 997

distribution of the 2 tumor types between the 2 treatmentarms. Therefore, the impact of this potential confounder isprobably limited.

In the present study, none of the patients received(neo)adjuvant chemo- and/or radiotherapy. However, manyWestern institutes have now incorporated neoadjuvant ther-apy in their standard treatment protocols. It is unclearwhether the outcome of the present study would be influencedby the introduction of multimodality therapy. It is conceiv-able that the potential benefit of an extended lymphadenec-tomy is dependent on the number of positive lymph nodes.Therefore, we subdivided our patients into 3 groups: patients

without positive lymph nodes in the resection specimen,patients with 1 to 8 positive nodes, and patients with morethan 8 positive nodes. Theoretically, an artificially low num-ber of positive nodes could be removed by the surgeon andidentified by the pathologist after performing a limited trans-hiatal resection. These patients would be erroneously consid-ered to have a limited number of positive nodes (N ! 1–8),and thus contaminate this subgroup. Even though the totalnumber of resected lymph nodes was higher after an extendedlymphadenectomy,5 the number of positive nodes found inthe resection specimen did not differ significantly betweenboth techniques (Table 1). Therefore, the phenomenon of

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

24101520242426273347TTE145912161920263343THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

24101520242426273347TTE145912161920263343THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

156913161821274063TTE15589161619263652THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7 8 9 10 11

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

surv

ival

156913161821274063TTE15589161619263652THE

Numbers at risk

THE

TTE

A

B

FIGURE 4. A, Overall survival of patients with typeI adenocarcinoma of the esophagus after transhia-tal (drawn line) or transthoracic (dotted line)esophagectomy (P ! 0.33). B, Overall survival ofpatients with type II adenocarcinoma of the gas-tric cardia after transhiatal (drawn line) or aftertransthoracic (dotted line) esophagectomy (P !0.81).

Annals of Surgery • Volume 246, Number 6, December 2007 Extended Versus Limited Esophagectomy

© 2007 Lippincott Williams & Wilkins 997

“Patientswerestratifiedtoatype1andtype2,accordingtotheendoscopyreport.ThegastroenterologistsandsurgeonswereperhapsabittoopronetocallatumortypeIesophageal.Forthatreasonseveraltumorswerecalledesophagealwhentheywereactuallyinthegastrocardial region.Wethoughtitwasbettertolookattheactuallocalizationsiteinthepathologyreportandthatisthereasonforthedifferenceinnumbers.”

“Nodifferenceisnotnoresults”isdifficulttoaccept


stage migration does not seem to be of major importance inthis comparison. We previously reported in 74 patients whounderwent extended TTE that stage migration was mainlybecause of positive nodes in the upper abdomen, whereaspositive subcarinal and paratracheal nodes only changedstaging in a minority of patients.16 The present analysisshows, even though these conclusions are based on smallsubgroups, that an extended lymphadenectomy does not offerbetter locoregional control of the disease in patients withoutpositive lymph nodes. Neither it is useful in patients whohave many (more than 8) positive lymph nodes. It makes

sense that removal of negative lymph nodes does not have animpact on disease control. Limited transhiatal resectionseems to be adequate to cure patients without positive lymphnodes; a microscopically radical resection of the tumor wasachieved in 72% of the patients, which is identical to theresults of the extended technique (also 72%). On the otherhand, patients with many (more than 8) positive lymph nodesare beyond surgical cure and probably need (additional)systemic therapy. Because the aim of a more extended resec-tion is to gain more locoregional disease control and patientswith 1 to 8 positive lymph nodes significantly benefited from

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

01931TTE12615THE

Numbers at risk

THE

TTE

00

0 1 2 3 4 5

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

regio

nal

dis

ease

free

surv

ival

THE

TTE

A B

C

FIGURE 5. A, Locoregional disease-free survival of all patients without positive lymph nodes in the resection specimen aftertranshiatal (drawn line) or transthoracic (dotted line) esophagectomy (P ! 0.64). B, Locoregional disease-free survival of allpatients with 1 to 8 positive lymph nodes in the resection specimen after transhiatal (drawn line) or transthoracic (dotted line)esophagectomy (P ! 0.02). C, Locoregional disease-free survival of all patients with more than 8 positive lymph nodes in theresection specimen after transhiatal (drawn line) or transthoracic (dotted line) esophagectomy (P ! 0.24).



ORIGINAL ARTICLES







(Ann Surg 2007;246: 992–1001)











“Basedonthisbestavailableevidence,wefavoranextendedtransthoracicprocedurefortypeIesophagealcarcinoma,especiallyifthereisalimitednumberofsuspiciousnodes,anda(limited)transhiatalprocedurefortypeIIcarcinomaofthegastriccardia.“




0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

01931TTE12615THE

Numbers at risk

THE

TTE

00

0 1 2 3 4 5

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

regio

nal

dis

ease

free

surv

ival

THE

TTE

A B

C




ORIGINAL ARTICLES







(Ann Surg 2007;246: 992–1001)













5yrOS:TTE39%THE19%

Meaning:TTEisusefulto9%ofpts whoundergoTTE

Look:thisis“locoregional diseasefree”

survival(!!!)



0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al

361012202627TTE37913202528THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

0 1 2 3 4 5 6 7

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

reg

ion

ald

isea

sefr

eesu

rviv

al27816243352TTE1456173252THE

Numbers at risk

THE

TTE

01931TTE12615THE

Numbers at risk

THE

TTE

00

0 1 2 3 4 5

follow-up (years)

0,0

0,2

0,4

0,6

0,8

1,0

loco

regio

nal

dis

ease

free

surv

ival

THE

TTE

A B

C




ORIGINAL ARTICLES







(Ann Surg 2007;246: 992–1001)















type I II III

Fig. 2.5

TRANSHIATALAPPROACH:totalgastrectomywithD2lymphadenectomy(includingsplenectomy).Additionaldissectionofthelymphnodesalongtheleftinferiorphrenicvesselsandthepara-aorticnodeslateraltotheaortaandabovetheleftrenalveinwasdoneincurablepatients.Mediastinalresectionincludedtheloweroesophagus andonlytheperioesophageal lymphnodes.

TRANSTHORACICAPPROACH:thesameprocedureasthatforTHwasdoneintheabdominalcavity,includinglymphadenectomy.Throughanobliqueincisionovertheleftthoraxathoroughmediastinal nodaldissectionbelowtheleftinferiorpulmonaryveinwasundertakenwithoesophagectomy ofsufficientlength.

Amedianof68nodeswereidentifiedintheresectionspecimenaftertranshiatal resection,and60aftertransthoracicresection.

644 http://oncology.thelancet.com Vol 7 August 2006

Articles

Left thoracoabdominal approach versus abdominal-transhiatal approach for gastric cancer of the cardia or subcardia: a randomised controlled trialMitsuru Sasako, Takeshi Sano, Seiichiro Yamamoto, Motonori Sairenji, Kuniyoshi Arai, Taira Kinoshita, Atsushi Nashimoto, Masahiro Hiratsuka, for the Japan Clinical Oncology Group (JCOG9502)

SummaryBackground Because of the inaccessibility of mediastinal nodal metastases, the left thoracoabdominal approach (LTA) has often been used to treat gastric cancer of the cardia or subcardia. In a randomised phase III study, we aimed to compare LTA with the abdominal-transhiatal approach (TH) in the treatment of these tumours.

Methods Between July, 1995, and December, 2003, 167 patients were enrolled from 27 Japanese hospitals and randomly assigned to TH (n=82) or LTA (n=85). The primary endpoint was overall survival, and secondary endpoints were disease-free survival, postoperative morbidity and hospital mortality, and postoperative symptoms and change of respiratory function. The projected sample size was 302. After the fi rst interim analysis, the predicted probability of LTA having a signifi cantly better overall survival than TH at the fi nal analysis was only 3·65%, and the trial was closed immediately. Analysis was by intention to treat. This study is registered with ClinicalTrials.gov, number NCT00149266.

Findings 5-year overall survival was 52·3% (95% CI 40·4–64·1) in the TH group and 37·9% (26·1–49·6) in the LTA group. The hazard ratio of death for LTA compared with TH was 1·36 (0·89–2·08, p=0·92). Three patients died in hospital after LTA but none after TH. Morbidity was worse after LTA than after TH.

Interpretation Because LTA does not improve survival after TH and leads to increased morbidity in patients with cancer of the cardia or subcardia, LTA cannot be justifi ed to treat these tumours.

IntroductionBy contrast with the notable decrease in the incidence of distal gastric cancer, frequency of adenocarcinoma in the oesophagogastric junction has increased, especially in developed countries.1–3 The Siewert classifi cation for these tumours is now widely accepted.4 Studies of adjuvant treatment for gastric cancer with chemotherapy or chemoradiotherapy have included tumours in the oesophagogastric junction.5,6 However, no evidence suggests that oesophagogastric-junction tumours can be treated in the same way as gastric cancers; if thoracotomy is mandatory for oesophagogastric-junction tumours, they should not be included in studies on the treatment of gastric cancers. So far, only one prospective randomised controlled trial7 has been undertaken to compare the eff ects of surgical treatments in Siewert type 1 and 2 tumours in the oesophagogastric junction. Although the trial was slightly underpowered, it suggested that extended transthoracic resection resulted in better survival than a restricted transhiatal resection. However, a systematic review8 comparing surgical treatments for lower oesophageal carcinoma showed a higher morbidity for transthoracic resection than for transhiatal resection, but with similar survival.

In eastern Asian countries, including Japan, most tumours in the oesophagogastric junction are of Siewert type 2 and 3.9 The occurrence of lower mediastinal lymph-node metastasis from type 2 and 3 tumours is

reported to be 10–40%.10–16 Some researchers10,11 claim that a thoracotomy is needed to thoroughly dissect the mediastinal nodes and to obtain a safe surgical margin, although mediastinal lymph-node metastasis is an indicator of poor prognosis. Other studies12,13 recommend the use of a transhiatal resection, because patients with mediastinal-lymph-node metastasis have poor prognosis even if a more extensive procedure was done. Advances in circular stapling devices have enabled surgeons to make safe intrathoracic or mediastinal anastomosis without thoracotomy.

In 1995, the Gastric Cancer Surgical Study Group of the Japan Clinical Oncology Group (GCSSG/JCOG) initiated a multicentre, randomised controlled trial with the aim to compare the eff ects of the left thoracoabdominal approach (LTA) with the abdominal-transhiatal approach (TH) on patients with cancer of the cardia or subcardia (JCOG 9502).

Methods PatientsOur study was designed as a multicentre, prospective, randomised phase III trial. The study protocol was approved by the clinical trial review committee of JCOG and the institutional review boards of all 27 participating Japanese hospitals before the initiation of the study, and all patients provided written informed consent. Eligibility criteria included: histologically proven adenocarcinoma

Lancet Oncol 2006; 7: 644–51

Published Online July 12, 2006DOI:10.1016/S1470-2045(06)

70766-5

See Refl ection and Reaction page 613

National Cancer Centre, Tokyo, Japan (Prof M Sasako MD,

T Sano MD, S Yamamoto PhD); Kanagawa Cancer Centre,

Kanagawa, Japan (M Sairenji MD); Metropolitan

Komagome Hospital, Tokyo, Japan (K Arai MD); National

Cancer Centre East Hospital, Kashiwa, Japan

(T Kinoshita MD); Niigata Cancer Centre Hospital,

Niigata, Japan (A Nashimoto MD); and Osaka Medical Centre for Cancer and

Cardiovascular Diseases, Osaka, Japan (M Hiratsuka MD)

Correspondence to: Prof Mitsuru Sasako, National Cancer Centre Hospital, Tokyo

104-0045, [email protected]



type I II III

Fig. 2.5

Differentlymphadenectomyinthelowermediastinum


Articles

http://oncology.thelancet.com Vol 7 August 2006 647

the predictive probability that LTA was signifi cantly better than TH was lower than 5%.

We retrospectively calculated two-sided p values for overall survival only in the updated analysis. Analysis was by intention to treat.

Overall survival curves and disease-free survival curves were estimated by the Kaplan-Meier method and compared by the log-rank test, stratifi ed by factors used in the randomisation. Disease-free survival was measured as the date of randomisation to the date of the fi rst observation of disease recurrence or death from any cause. Overall survival was also assessed by the Cox regression analysis stratifi ed by tumour stage (T2 vs T3 and 4) and Borrman type (type 0–2 vs 3 and 5), the adjustment variables in randomisation; and adjusted with para-aortic nodal metastasis (negative vs positive), mediastinal nodal metastasis (negative vs positive), residual tumour (R0 vs R1 and 2), washing cytology (negative vs positive), and number of positive nodes (0 vs 1–6 vs 7–15 vs ≥16), as potential confounding factors. Operative morbidity and mortality were based on the proportion of the number of events divided by all eligible patients. Statistical analysis was done with SAS version 8.02. This study is registered with ClinicalTrials.gov, number NCT00149266.

Role of the funding sourceThe sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had fi nal responsibility for the decision to submit for publication.

Results Between July, 1995, and December, 2003, 82 patients were randomly assigned to TH and 85 to LTA (fi gure 1). One patient was ineligible because of a second lesion in the distal stomach. Ten cases of protocol violation were reported. Two patients underwent adjuvant chemotherapy after surgery because of positive peritoneal lavage cytology. Nine patients did not complete the lymph-node dissection required in the protocol, one of whom underwent adjuvant chemotherapy.

The fi rst interim analysis was done as scheduled in the amended protocol for the 165 patients enrolled before October, 2003, and survival analysis was undertaken for 146 patients for whom data were available. At the time, the observed number of events was 28% of the expected number at the fi nal analysis, and the alpha used for this analysis was 0·00173. The median overall survival time

TH (n=82) LTA (n=85)*

Type of gastrectomy

Total 79 80

Proximal 3 3

Not resected 0 2

Reconstruction method

Roux-en-Y 75 76

Interposition 5 3

Other 2 4

Length of resected oesophagus (cm)

Median (range) 4·2 (2·0–9·5) 4·5 (2·0–8·5)

Splenectomy

Yes 78 81

No 4 4

Pancreatic-tail resection

Yes 22 13

No 60 72

Thoracotomy

Intercostal 3 79

Transabdominal 10 3

None 69 3

Dissected lymph nodes (median [range])

Total 68 (14–147) 60 (16–160)

Mediastinal 2 (0–13) 8 (0–24)

Para-aortic 7 (0–63) 6 (0–60)

Operation time (min)

Median (range) 305 (100–620) 338 (73–635)

Blood loss (mL)

Median (range) 673 (55–3500) 655 (55–2174)

Allogeneic blood transfusion

Yes 25 39

No 57 46

Data are number of patients unless stated otherwise. *Two patients undergoing simple laparotomy without gastrectomy not included in data for reconstruction method, length of resected oesophagus, and dissected lymph nodes; one of these patients was not included in data for operation time and blood loss.

Table 2: Details of surgical procedures

(Continued from previous page)

Oesophageal invasion (cm)

Median (range) 1·6 (0–4·5) 1·2 (0–7·0)

Washing cytology

Negative 69 73

Positive 11 9

Not done 2 3

Residual tumour

R0 76 75

R1/2 6 10

Para-aortic nodal metastasis

Positive 13 9

Negative 59 64

Not dissected 10 12

Mediastinal nodal metastasis

Positive 3 9†

Negative 79 74

Not dissected 0 2

Data are number of patients unless stated otherwise. *Data not available for two patients in LTA group who did not undergo resection because of peritoneal seeding. †Includes fi ve patients with Siewert type 2 tumours and four with other types.

Table 1: Baseline characteristics

Articles

http://oncology.thelancet.com Vol 7 August 2006 647

the predictive probability that LTA was signifi cantly better than TH was lower than 5%.

We retrospectively calculated two-sided p values for overall survival only in the updated analysis. Analysis was by intention to treat.

Overall survival curves and disease-free survival curves were estimated by the Kaplan-Meier method and compared by the log-rank test, stratifi ed by factors used in the randomisation. Disease-free survival was measured as the date of randomisation to the date of the fi rst observation of disease recurrence or death from any cause. Overall survival was also assessed by the Cox regression analysis stratifi ed by tumour stage (T2 vs T3 and 4) and Borrman type (type 0–2 vs 3 and 5), the adjustment variables in randomisation; and adjusted with para-aortic nodal metastasis (negative vs positive), mediastinal nodal metastasis (negative vs positive), residual tumour (R0 vs R1 and 2), washing cytology (negative vs positive), and number of positive nodes (0 vs 1–6 vs 7–15 vs ≥16), as potential confounding factors. Operative morbidity and mortality were based on the proportion of the number of events divided by all eligible patients. Statistical analysis was done with SAS version 8.02. This study is registered with ClinicalTrials.gov, number NCT00149266.

Role of the funding sourceThe sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had fi nal responsibility for the decision to submit for publication.

Results Between July, 1995, and December, 2003, 82 patients were randomly assigned to TH and 85 to LTA (fi gure 1). One patient was ineligible because of a second lesion in the distal stomach. Ten cases of protocol violation were reported. Two patients underwent adjuvant chemotherapy after surgery because of positive peritoneal lavage cytology. Nine patients did not complete the lymph-node dissection required in the protocol, one of whom underwent adjuvant chemotherapy.

The fi rst interim analysis was done as scheduled in the amended protocol for the 165 patients enrolled before October, 2003, and survival analysis was undertaken for 146 patients for whom data were available. At the time, the observed number of events was 28% of the expected number at the fi nal analysis, and the alpha used for this analysis was 0·00173. The median overall survival time

TH (n=82) LTA (n=85)*

Type of gastrectomy

Total 79 80

Proximal 3 3

Not resected 0 2

Reconstruction method

Roux-en-Y 75 76

Interposition 5 3

Other 2 4

Length of resected oesophagus (cm)

Median (range) 4·2 (2·0–9·5) 4·5 (2·0–8·5)

Splenectomy

Yes 78 81

No 4 4

Pancreatic-tail resection

Yes 22 13

No 60 72

Thoracotomy

Intercostal 3 79

Transabdominal 10 3

None 69 3

Dissected lymph nodes (median [range])

Total 68 (14–147) 60 (16–160)

Mediastinal 2 (0–13) 8 (0–24)

Para-aortic 7 (0–63) 6 (0–60)

Operation time (min)

Median (range) 305 (100–620) 338 (73–635)

Blood loss (mL)

Median (range) 673 (55–3500) 655 (55–2174)

Allogeneic blood transfusion

Yes 25 39

No 57 46

Data are number of patients unless stated otherwise. *Two patients undergoing simple laparotomy without gastrectomy not included in data for reconstruction method, length of resected oesophagus, and dissected lymph nodes; one of these patients was not included in data for operation time and blood loss.

Table 2: Details of surgical procedures

(Continued from previous page)

Oesophageal invasion (cm)

Median (range) 1·6 (0–4·5) 1·2 (0–7·0)

Washing cytology

Negative 69 73

Positive 11 9

Not done 2 3

Residual tumour

R0 76 75

R1/2 6 10

Para-aortic nodal metastasis

Positive 13 9

Negative 59 64

Not dissected 10 12

Mediastinal nodal metastasis

Positive 3 9†

Negative 79 74

Not dissected 0 2

Data are number of patients unless stated otherwise. *Data not available for two patients in LTA group who did not undergo resection because of peritoneal seeding. †Includes fi ve patients with Siewert type 2 tumours and four with other types.

Table 1: Baseline characteristics 644 http://oncology.thelancet.com Vol 7 August 2006

Articles

Left thoracoabdominal approach versus abdominal-transhiatal approach for gastric cancer of the cardia or subcardia: a randomised controlled trialMitsuru Sasako, Takeshi Sano, Seiichiro Yamamoto, Motonori Sairenji, Kuniyoshi Arai, Taira Kinoshita, Atsushi Nashimoto, Masahiro Hiratsuka, for the Japan Clinical Oncology Group (JCOG9502)

SummaryBackground Because of the inaccessibility of mediastinal nodal metastases, the left thoracoabdominal approach (LTA) has often been used to treat gastric cancer of the cardia or subcardia. In a randomised phase III study, we aimed to compare LTA with the abdominal-transhiatal approach (TH) in the treatment of these tumours.

Methods Between July, 1995, and December, 2003, 167 patients were enrolled from 27 Japanese hospitals and randomly assigned to TH (n=82) or LTA (n=85). The primary endpoint was overall survival, and secondary endpoints were disease-free survival, postoperative morbidity and hospital mortality, and postoperative symptoms and change of respiratory function. The projected sample size was 302. After the fi rst interim analysis, the predicted probability of LTA having a signifi cantly better overall survival than TH at the fi nal analysis was only 3·65%, and the trial was closed immediately. Analysis was by intention to treat. This study is registered with ClinicalTrials.gov, number NCT00149266.

Findings 5-year overall survival was 52·3% (95% CI 40·4–64·1) in the TH group and 37·9% (26·1–49·6) in the LTA group. The hazard ratio of death for LTA compared with TH was 1·36 (0·89–2·08, p=0·92). Three patients died in hospital after LTA but none after TH. Morbidity was worse after LTA than after TH.

Interpretation Because LTA does not improve survival after TH and leads to increased morbidity in patients with cancer of the cardia or subcardia, LTA cannot be justifi ed to treat these tumours.

IntroductionBy contrast with the notable decrease in the incidence of distal gastric cancer, frequency of adenocarcinoma in the oesophagogastric junction has increased, especially in developed countries.1–3 The Siewert classifi cation for these tumours is now widely accepted.4 Studies of adjuvant treatment for gastric cancer with chemotherapy or chemoradiotherapy have included tumours in the oesophagogastric junction.5,6 However, no evidence suggests that oesophagogastric-junction tumours can be treated in the same way as gastric cancers; if thoracotomy is mandatory for oesophagogastric-junction tumours, they should not be included in studies on the treatment of gastric cancers. So far, only one prospective randomised controlled trial7 has been undertaken to compare the eff ects of surgical treatments in Siewert type 1 and 2 tumours in the oesophagogastric junction. Although the trial was slightly underpowered, it suggested that extended transthoracic resection resulted in better survival than a restricted transhiatal resection. However, a systematic review8 comparing surgical treatments for lower oesophageal carcinoma showed a higher morbidity for transthoracic resection than for transhiatal resection, but with similar survival.

In eastern Asian countries, including Japan, most tumours in the oesophagogastric junction are of Siewert type 2 and 3.9 The occurrence of lower mediastinal lymph-node metastasis from type 2 and 3 tumours is

reported to be 10–40%.10–16 Some researchers10,11 claim that a thoracotomy is needed to thoroughly dissect the mediastinal nodes and to obtain a safe surgical margin, although mediastinal lymph-node metastasis is an indicator of poor prognosis. Other studies12,13 recommend the use of a transhiatal resection, because patients with mediastinal-lymph-node metastasis have poor prognosis even if a more extensive procedure was done. Advances in circular stapling devices have enabled surgeons to make safe intrathoracic or mediastinal anastomosis without thoracotomy.

In 1995, the Gastric Cancer Surgical Study Group of the Japan Clinical Oncology Group (GCSSG/JCOG) initiated a multicentre, randomised controlled trial with the aim to compare the eff ects of the left thoracoabdominal approach (LTA) with the abdominal-transhiatal approach (TH) on patients with cancer of the cardia or subcardia (JCOG 9502).

Methods PatientsOur study was designed as a multicentre, prospective, randomised phase III trial. The study protocol was approved by the clinical trial review committee of JCOG and the institutional review boards of all 27 participating Japanese hospitals before the initiation of the study, and all patients provided written informed consent. Eligibility criteria included: histologically proven adenocarcinoma

Lancet Oncol 2006; 7: 644–51

Published Online July 12, 2006DOI:10.1016/S1470-2045(06)

70766-5

See Refl ection and Reaction page 613

National Cancer Centre, Tokyo, Japan (Prof M Sasako MD,

T Sano MD, S Yamamoto PhD); Kanagawa Cancer Centre,

Kanagawa, Japan (M Sairenji MD); Metropolitan

Komagome Hospital, Tokyo, Japan (K Arai MD); National

Cancer Centre East Hospital, Kashiwa, Japan

(T Kinoshita MD); Niigata Cancer Centre Hospital,

Niigata, Japan (A Nashimoto MD); and Osaka Medical Centre for Cancer and

Cardiovascular Diseases, Osaka, Japan (M Hiratsuka MD)

Correspondence to: Prof Mitsuru Sasako, National Cancer Centre Hospital, Tokyo

104-0045, [email protected]

THLTA


Left thoracoabdominal and transhiatal approaches to total gastrectomy 343

Table 1 Patient characteristics

TH group(n= 82)

LTA group(n= 85)

Age (years)* 60 (36–75) 63 (38–75)Sex ratio (M : F) 71 : 11 63 : 22Borrmann type

0–2 36 373 or 5 46 48

Siewert classification†Type II 52 43Type III 27 36Non-OGJ tumour 3 4

Tumour size (cm)*† 6⋅2 (2⋅5–19) 7⋅0 (2⋅0–18)Histological type†

Differentiated 42 43Undifferentiated 40 40

Clinical tumour category‡§cT2 20 20cT3/4 62 65

Pathological tumour category†‡§pT1b 2 1pT2a 10 6pT2b 24 35pT3 39 37pT4 7 4

Pathological node category†‡pN0 14 15pN1 24 27pN2 30 25pN3/4 14 16

Pathological node category†§pN0 14 15pN1 35 28pN2 16 26pN3 17 14

No. of positive nodes*† 5 (0–53) 5 (0–52)Histological oesophageal invasion (cm)*† 1⋅6 (0–4⋅5) 1⋅2 (0–7⋅0)Residual tumour

R0 76 75R1/2 6 10

*Values are median (range). †Data not available for two patients in the leftthoracoabdominal (LTA) group who did not undergo surgical resectionowing to M1 disease. ‡Japanese Classification of Gastric Carcinoma, 12thedition16; §International Union Against Cancer (UICC) TNMclassification, 6th edition17. TH, transhiatal; OGJ, oesophagogastricjunction.

disease), no further treatment was allowed unless recur-rence was diagnosed.

Statistical analysis

The primary endpoint was overall survival (OS). Sec-ondary endpoints were disease-free survival (DFS),morbidity and mortality, postoperative symptoms andpostoperative respiratory function. All in-hospital deathsand deaths within 1 month of surgery were defined ashospital mortality. Operative procedures and pathology

0

0·1

No. at riskTHLTA

8285

a Overall survival

7263

6152

5144

4738

4231

4028

3724

3123

2418

1913

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

0

0·1

No. at riskTHLTA

7675

b Disease-free survival

5649

4637

4133

3727

3625

3525

3120

2620

2115

1711

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Dis

ease

-fre

e su

rviv

al

6 7 8 9 10

THLTA

Fig. 2 Kaplan–Meier curves of a overall and b disease-freesurvival in all randomized patients by treatment group. TH,transhiatal approach; LTA, left thoracoabdominal approach.a Hazard ratio (HR) 1⋅42 (95 per cent c.i. 0⋅98 to 2⋅05; P = 0⋅970and P = 0⋅060, 1- and 2-sided log rank test respectively); b HR1⋅28 (0⋅87 to 1⋅89; P = 0⋅892 and P = 0⋅215, 1- and 2-sided logrank test respectively)

results were recorded according to the 12th edition of theJapanese Classification of Gastric Carcinoma (JCGC)16.Tumour stage is reported here using the sixth editionof the TNM classification17. All tumours were clas-sified as Siewert type II, type III, or non-OGJ if thetumour epicentre was located more than 5 cm distal to theOGJ based on pathological examination of the resectedspecimen.

© 2015 The Authors. BJS published by John Wiley & Sons Ltd www.bjs.co.uk BJS 2015; 102: 341–348on behalf of BJS Society Ltd.



TH group(n= 82)

LTA group(n= 85)


0–2 36 373 or 5 46 48









R0 76 75R1/2 6 10





0

0·1

No. at riskTHLTA

8285

a Overall survival

7263

6152

5144

4738

4231

4028

3724

3123

2418

1913

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

0

0·1

No. at riskTHLTA

7675


5649

4637

4133

3727

3625

3525

3120

2620

2115

1711

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Dis

ease

-fre

e su

rviv

al

6 7 8 9 10

THLTA






TH group(n= 82)

LTA group(n= 85)


0–2 36 373 or 5 46 48









R0 76 75R1/2 6 10





0

0·1

No. at riskTHLTA

8285

a Overall survival

7263

6152

5144

4738

4231

4028

3724

3123

2418

1913

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

0

0·1

No. at riskTHLTA

7675


5649

4637

4133

3727

3625

3525

3120

2620

2115

1711

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Dis

ease

-fre

e su

rviv

al

6 7 8 9 10

THLTA




344 Y. Kurokawa, M. Sasako, T. Sano, T. Yoshikawa, Y. Iwasaki, A. Nashimoto et al.

Table 2 Sites of first recurrence

TH group (n= 82) LTA group (n= 85) P*

Lymph nodes 12 (15) 19 (22) 0⋅235Peritoneum 9 (11) 10 (12) 1⋅000Liver 8 (10) 9 (11) 1⋅000Lung 5 (6) 5 (6) 1⋅000Pleura 3 (4) 1 (1) 0⋅362Other 5 (6) 2 (2) 0⋅271

Values in parentheses are percentages. TH, transhiatal; LTA, leftthoracoabdominal. *Fisher’s exact test, two-sided.

The original intention was to recruit 302 patients toachieve a one-sided α of 0⋅05 and statistical power of 80per cent to detect a difference between the two groups,assuming a 5-year survival rate of 15⋅5 per cent for the THapproach versus 26⋅0 per cent for the LTA procedure. Theprojected accrual period was 4 years. After 8 years of slowaccrual, the JCOG Data and Safety Monitoring Committeeapproved an amendment to the sample size and analysisplan. The amended sample size was 250, with a one-sidedα of 0⋅1 and power of 80 per cent, with an accrual periodof 12 years in total and 8 years of follow-up. Three interimanalyses were planned.

OS was measured from the date of randomization tothe date of death from any cause. Among patients whounderwent R0 resection, DFS was measured from the dateof randomization to the date of the first observation ofdisease recurrence or death from any cause. OS and DFScurves were estimated using the Kaplan–Meier methodand compared with the log rank test. Subgroup analysis wasperformed by means of Cox regression to assess statisticalinteractions between treatment approach and 12 patientcharacteristics. Postoperative factors were also included inthe Cox regression analysis to estimate their influence onsurvival. Two-sided P values were calculated for all tests.As the study was planned for one-sided testing, one-sidedP values are presented for the primary endpoint. P < 0⋅050was judged to be statistically significant. All analyses werebased on an intention-to-treat basis. Statistical analyseswere performed with SAS® version 9⋅2 (SAS Institute,Cary, North Carolina, USA).

Results

Between July 1995 and December 2003, 167 patients wereenrolled, of whom 82 were randomly assigned to the THand 85 to the LTA approach (Fig. 1). Baseline characteristicsof the two groups were similar, except for Siewert classifica-tion (Table 1). There were 95 Siewert type II and 63 type IIItumours. Seven patients had large gastric tumours invad-ing the oesophagus that could not be classified by Siewert

type. At operation, 141 patients (62 TH, 79 LTA) under-went mediastinal node dissection and 145 (72 TH, 73 LTA)had para-aortic node dissection. The rate of metastasis inmediastinal nodes was 5 per cent (3 of 62) in the TH groupand 11 per cent (9 of 79) in the LTA group. The metastasisrate in para-aortic nodes was 18 per cent (13 of 72) and 12per cent (9 of 73) respectively.

Operative details, including morbidity and mortality,postoperative symptoms and postoperative respiratoryfunction, have been reported previously13,18. Median dura-tion of surgery was 33 min longer for the LTA procedurethan for the TH approach (P = 0⋅127). Median blood losswas similar in the two groups (655 versus 673 ml for LTAand TH group respectively; P = 0⋅949), but allogeneicblood transfusion was used more frequently in the LTAgroup (39 of 85, 46 per cent) than in the TH group (25of 82, 30 per cent) (P = 0⋅056). Patients in the LTA grouphad a higher morbidity rate: 42 (49 per cent) versus 28 (34per cent) (P = 0⋅060). For six selected major complications(pancreatic fistula, abdominal abscess, pneumonia, anas-tomotic leak, empyema thoracis and mediastinitis), theincidence was significantly higher following the LTA thanthe TH procedure: 35 (41 per cent) versus 18 (22 per cent)(P = 0⋅008). There were two treatment-related deaths,both in the LTA group.

Median follow-up for all censored patients was 10⋅6(range 5⋅1–17⋅1) years at the last follow-up in December2012. There had been 52 and 63 deaths in the TH and LTAgroup respectively, with 42 and 50 patients respectivelydying from cancer. The 5- and 10-year OS rates for allrandomized patients were 51 (95 per cent c.i. 40 to 61) and37 (26 to 47) per cent respectively for the TH approach,and 37 (26 to 47) and 24 (15 to 34) per cent for the LTAapproach (Fig. 2a). The log rank test showed marginaldifferences between the groups (2-sided P = 0⋅060, 1-sidedP = 0⋅970), and the hazard ratio (HR) for the LTA versus theTH approach was 1⋅42 (95 per cent c.i. 0⋅98 to 2⋅05). Afteradjustment for cT category and Borrmann type, stratifiedP values were 0⋅102 in two-sided and 0⋅949 in one-sidedlog rank tests, and the HR was 1⋅36 (0⋅94 to 1⋅98). Inmultivariable Cox regression analysis with seven baselinevariables (age, sex, Borrmann type, Siewert classification,tumour size, histological type, cT category (JCGC 12thedition)), the HR was essentially unchanged: HR 1⋅33(0⋅91 to 1⋅95). Per-protocol analysis (145 patients) alsoshowed a HR of 1⋅44 (0⋅96 to 2⋅15) (2-sided P = 0⋅076,1-sided P = 0⋅962) (Fig. S1, supporting information).

Among patients who underwent R0 or R1 resection,initial sites of recurrence were lymph nodes (31 patients),peritoneum (19), liver (17), lung (10), pleura (4) and othersites (7). The pattern of recurrence was similar in the two


167patients

346 Y. Kurokawa, M. Sasako, T. Sano, T. Yoshikawa, Y. Iwasaki, A. Nashimoto et al.

0

0·1

No. at riskTHLTA

5243

a Siewert type II

4739

4134

3429

3123

2618

2616

2414

2014

1610

138

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

THLTA

0

0·1

No. at riskTHLTA

2736

b Siewert type III

2421

1917

1715

1615

1613

1412

1310

119

88

65

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

Fig. 4 Kaplan–Meier curves of overall survival in patients with aSiewert type II and b Siewert type III tumours by treatmentgroup. TH, transhiatal approach; LTA, left thoracoabdominalapproach. a Hazard ratio (HR) 1⋅19 (95 per cent c.i. 0⋅72 to 1⋅95;P = 0⋅496, 2-sided log rank test); b HR 1⋅67 (0⋅90 to 3⋅11;P = 0⋅102, 2-sided log rank test)

groups (Table 2). The 5- and 10-year DFS rates were 47(95 per cent c.i. 36 to 58) and 36 (25 to 47) per cent forthe TH approach, and 37 (26 to 48) and 26 (16 to 37) percent for the LTA approach (P = 0⋅215) (Fig. 2b). The HRfor the LTA group compared with the TH group was 1⋅28(0⋅87 to 1⋅89).

There were no significant interactions between treat-ment effects and the patient characteristics examined(Fig. 3). For the 95 patients with Siewert type II tumours,5- and 10-year OS rates were 50 (95 per cent c.i. 36 to

63) and 35 (22 to 48) per cent respectively for the THapproach, and 42 (27 to 56) and 29 (16 to 43) per cent forthe LTA approach (HR 1⋅19, 95 per cent c.i. 0⋅72 to 1⋅95;P = 0⋅496) (Fig. 4a). Among 63 patients with Siewert typeIII tumours, 5- and 10-year OS rates were 59 (39 to 75) and44 (26 to 62) per cent respectively for the TH approach,and 36 (21 to 51) and 22 (10 to 38) per cent for the LTAapproach (HR 1⋅67, 0⋅90 to 3⋅11; P = 0⋅102) (Fig. 4b). Forthe subgroup of patients with type III tumours, excludingthose with major complications from the survival analysis,the difference in OS between the groups was significant(HR 2⋅00, 0⋅99 to 4⋅05; P = 0⋅050).

The rate of metastasis in mediastinal lymph nodes was10 per cent (8 of 81) for type II tumours compared with5 per cent (3 of 55) for type III. The metastasis rate inpara-aortic lymph nodes was 9 per cent (8 of 86) for type IIcompared with 22 per cent (12 of 54) for type III tumours.The 5-year OS rate for 22 patients with pathologicallyconfirmed metastasis in the para-aortic nodes was 18 (95per cent c.i. 6 to 36) per cent.

Discussion

This final analysis, based on 10-year follow-up data, hasconfirmed the conclusion reached in the earlier publicationof the interim analysis13. Compared with the TH approach,the LTA approach offered no improvement in OS or DFS.The LTA technique did not reduce rates of cancer recur-rence in lymph nodes, but was associated with greater mor-bidity and mortality. Although the LTA approach involvedno increase in blood loss, more patients assigned to thisapproach received allogeneic blood transfusion, mainly tocorrect haemodynamic instability in the early postoperativeperiod13. LTA cannot, therefore, be justified for the treat-ment of adenocarcinoma of the OGJ or gastric cardia if thelength of oesophageal invasion is 3 cm or less.

Omloo and colleagues19 reported the final results ofa Dutch trial comparing a right thoracic with the THapproach for Siewert type I and II tumours. In their trial,subgroup analysis revealed a 14 per cent 5-year OS advan-tage with the right thoracic approach for patients withtype I tumours, but no difference for patients with typeII tumours (27 per cent versus 31 per cent in the THgroup). The present study confirms the results of the Dutchstudy for patients with type II tumours based on surgi-cal approach. For patients with type III tumours, the LTAapproach was associated with a 22 per cent lower 10-yearOS rate than the TH technique, and this difference inOS between the two groups occurred within 1 year aftersurgery. Although the reason for worse survival of patientswho underwent LTA in this subgroup is unclear, the greater


35%vs 29%p=0.496

44%vs 22%p=0.05

Randomized clinical trial

Ten-year follow-up results of a randomized clinical trialcomparing left thoracoabdominal and abdominal transhiatalapproaches to total gastrectomy for adenocarcinoma of theoesophagogastric junction or gastric cardia

Y. Kurokawa1, M. Sasako2, T. Sano3, T. Yoshikawa6, Y. Iwasaki4, A. Nashimoto7, S. Ito8, A. Kurita9,J. Mizusawa5 and K. Nakamura5 for the Japan Clinical Oncology Group (JCOG9502)1Department of Gastroenterological Surgery, Osaka University Graduate School of Medicine, Osaka, 2Department of Surgery, Hyogo College ofMedicine, Nishinomiya, 3Department of Surgery, Cancer Institute Hospital, 4Department of Surgery, Tokyo Metropolitan Cancer and InfectiousDisease Centre, Komagome Hospital and 5Japan Clinical Oncology Group Data Centre, National Cancer Centre, Tokyo 6Department of Surgery,Kanagawa Cancer Centre, Yokohama 7Department of Surgery, Niigata Cancer Centre Hospital, Niigata 8Department of Gastroenterological Surgery,Aichi Cancer Centre Hospital, Nagoya and 9Department of Surgery, National Hospital Organization Shikoku Cancer Centre, Matsuyama, JapanCorrespondence to: Professor M. Sasako, Department of Surgery, Hyogo College of Medicine, 1–1, Mukogawa-cho, Nishinomiya, Hyogo 663–8501, Japan(e-mail: [email protected])

Background: The optimal surgical approach for treatment of oesophagogastric junction (OGJ) canceris controversial. A randomized clinical trial (JCOG9502) comparing transhiatal (TH) and left thoraco-abdominal (LTA) approaches was stopped after the first interim analysis owing to limited efficacy forLTA resections. Complete 10-year follow-up data are now available.Methods: Patients with histologically proven adenocarcinoma of the OGJ or gastric cardia withoesophageal invasion of 3 cm or less were randomized to a TH or LTA approach. Both groups underwenttotal gastrectomy and splenectomy with D2 nodal dissection plus para-aortic lymphadenectomy abovethe left renal vein. For LTA, a thorough mediastinal lymphadenectomy below the left inferior pulmonaryvein was also mandatory. The primary endpoint was overall survival.Results: A total of 167 patients (82 TH, 85 LTA) were enrolled. The 10-year overall survival rate was37 (95 per cent c.i. 26 to 47) per cent for the TH approach and 24 (15 to 34) per cent for the LTAtechnique (P =0⋅060). The hazard ratio for death was 1⋅42 (0⋅98 to 2⋅05) for the LTA technique. Subgroupanalysis based on the Siewert classification indicated non-significant survival advantages in favour of theTH approach.Conclusion: LTA resections should be avoided in the treatment of adenocarcinoma of the OGJ or gastriccardia. Registration number: NCT00149266 (https://www.clinicaltrials.gov).

Paper accepted 4 December 2014Published online 21 January 2015 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.9764

Introduction

The incidence of adenocarcinoma of the oesophago-gastric junction (OGJ) has increased in developed countriesover the past 20 years1,2. Although surgery is consideredessential as part of a curative treatment strategy for mostpatients, survival remains poor even in those who undergoR0 resection, with or without additional therapy3. Toimprove the R0 resection rate and long-term outcomes,extended surgery with en bloc lymphadenectomy has beenattempted for many years. When considering tumoursarising from the cardia (Siewert type III4), or those at theOGJ (Siewert type II) with minimal oesophageal extension

where total gastrectomy seems appropriate, left thoraco-abdominal (LTA) and transhiatal (TH) approaches havebeen advocated for curative resection. There is no clearinformation to indicate whether the operative approachinfluences long-term outcome.

In East Asian countries, including Japan, the majority ofOGJ tumours are Siewert types II and III5. The incidenceof lower mediastinal lymph node metastasis from typeII and III tumours is reported to range from 10 to 40per cent6–12. Some institutions prefer the LTA to theTH approach in order to perform lymph node dissectionin the lower mediastinal field and obtain a safe surgicalmargin6,7, whereas others prefer the TH technique owing

© 2015 The Authors. BJS published by John Wiley & Sons Ltd on behalf of BJS Society Ltd.This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in anymedium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.



TH group(n= 82)

LTA group(n= 85)


0–2 36 373 or 5 46 48









R0 76 75R1/2 6 10





0

0·1

No. at riskTHLTA

8285

a Overall survival

7263

6152

5144

4738

4231

4028

3724

3123

2418

1913

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Ove

rall

surv

ival

6 7 8 9 10

0

0·1

No. at riskTHLTA

7675


5649

4637

4133

3727

3625

3525

3120

2620

2115

1711

0·2

0·3

0·4

0·5

0·6

0·7

0·8

0·9

1·0

1 2 3 4 5

Dis

ease

-fre

e su

rviv

al

6 7 8 9 10

THLTA




“Leftthoraco-abdominalresectionsshouldbeavoidedinthetreatmentofadenocarcinomaoftheesophago-gastricjunctionorgastriccardia.“©2015MarcoLotti– [email protected]

TheMeta-Analyses


Sensitivity AnalysisWhen high-quality studies, studies with direct statistic method,

studies matched for stage or the final survival results of JCOG9502considered were re-performed for meta-analyses in sensitivityanalyses, the results were shown in Table 7. No changes ofoutcomes were observed in terms of the 5-y overall survival rate,postoperative morbidity and mortality, compared to the primaryresults. When RCT and non-RCT studies were combined, therewas a significant increase in postoperative morbidity with

thransthoracic resection (OR = 0.55, 95% CI: 0.31- 0.99,P = 0.04). This contrasts with the lack of evidence of differencein 5-year overall survival and 30 day mortality (Table 7).

Discussion

The incidence of GEJ cancers has been increased [2–4].Surgery is still considered as the potential curative treatment.Proximal margin length and lymph node involvement are

Figure 2. Forest plot of 5-year overall survival rates for RCTs and non-RCTs. a: RCTs; b: non-RCTs. The 95% confidence interval (CI) for thehazard ratio for each study is represented by a horizontal line and the point estimate is represented by a square. The size of the square correspondsto the weight of the study in the meta-analysis. The 95% CI for pooled estimates is represented by a diamond. Data for a fixed-effects model areshown as there was no statistical heterogeneity. df = degrees of freedom; I2 = percentage of the total variation across studies due to heterogeneity;IV = Inverse Variance; SE = standard error; Z = test of overall treatment effect.doi:10.1371/journal.pone.0037698.g002

Transthoracic Resection for GEJ Cancers

PLoS ONE | www.plosone.org 6 June 2007 | Volume 7 | Issue 6 | e37698

Transthoracic Resection versus Non-TransthoracicResection for Gastroesophageal Junction Cancer: AMeta-AnalysisKun Yang1., Hai-Ning Chen2., Xin-Zu Chen1, Qing-Chun Lu2, Lin Pan2, Jie Liu1, Bin Dai1, Bo Zhang1*, Zhi-

Xin Chen1, Jia-Ping Chen1, Jian-Kun Hu1

1 Department of Gastrointestinal Surgery, West China Hospital, Sichuan University, Chengdu, Sichuan Province, People’s Republic of China, 2 West China School of

Medicine, Sichuan University, Chengdu, Sichuan Province, People’s Republic of China

Abstract

Background: The aim of this meta-analysis is to evaluate the impact of transthoracic resection on long-term survival ofpatients with GEJ cancer and to compare the postoperative morbidity and mortality of patients undergoing transthoracicresection with those of patients who were not undergoing transthoracic resection.

Method: Searches of electronic databases identifying studies from Medline, Cochrane Library trials register, and WHO TrialRegistration etc were performed. Outcome measures were survival, postoperative morbidity and mortality, and operationrelated events.

Results: Twelve studies (including 5 RCTs and 7 non-RCTs) comprising 1105 patients were included in this meta-analysis,with 591 patients assigned treatment with transthoracic resection. Transthoracic resection did not increase the 5-y overallsurvival rate for RCTs and non-RCTs (HR = 1.01, 95% CI 0.80- 1.29 and HR = 0.89, 95% CI 0.70- 1.14, respectively). Stratified bythe Siewert classification, our result showed no obvious differences were observed between the group with transthoracicresection and group without transthoracic resection (P.0.05). The postoperative morbidity (RR = 0.69, 95% CI 0.48- 1.00 andOR = 0.55, 95% CI 0.25- 1.22) and mortality (RD = 20.03, 95% CI 20.06- 0.00 and RD = 0.00, 95% CI 20.05- 0.05) of RCTs andnon-RCTs did not suggest any significant differences between the two groups. Hospital stay was long with thransthoracicresection (WMD = 25.80, 95% CI 210.38- 21.23) but did not seem to differ in number of harvested lymph nodes, operationtime, blood loss, numbers of patients needing transfusion, and reoperation rate. The results of sensitivity analyses weresimilar to the primary analyses.

Conclusions: There were no significant differences of survival rate and postoperative morbidity and mortality betweentransthoracic resection group and non-transthoracic resection group. Both surgical approaches are acceptable, and that oneoffers no clear advantage over the other. However, the results should be interpreted cautiously since the qualities ofincluded studies were suboptimal.

Citation: Yang K, Chen H-N, Chen X-Z, Lu Q-C, Pan L, et al. (2012) Transthoracic Resection versus Non-Transthoracic Resection for Gastroesophageal JunctionCancer: A Meta-Analysis. PLoS ONE 7(6): e37698. doi:10.1371/journal.pone.0037698

Editor: Robert S. Phillips, University of York, United Kingdom

Received July 14, 2011; Accepted April 25, 2012; Published June 4, 2012

Copyright: ! 2012 Yang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was funded by Projects (2008SZ0168) in the Science & Technology Pillar Program, Scientific and Technological Department of SichuanProvince, People’s Republic of China( http://www.scst.gov.cn/info/) and a grant from National Natural Science Foundation of China (No.81071777). (http://www.nsfc.org.cn/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: [email protected]

. These authors contributed equally to this work.

Introduction

Gastroesophageal junction (GEJ) cancer has been graduallyconsidered as an entity separate from both esophageal cancer andgastric cancer [1]. Although a decline in incidence of gastriccarcinoma, there has been a tendency of proximal migration ofcarcinoma in Western countries [2–4]. A kind of classificationproposed by Siewert & Stein, which includes three types, waswidely accepted for GEJ cancer [5]. According to the classifica-tion, type 1 is defined as tumors whose centers are located 1 to5 cm above the gastroesophageal junction (distal esophagealadenocarcinoma); type 2, adenocarcinoma with its epicenter

located between 1 cm proximal and 2 cm distal of the GEJ, isdefined as a true cardia carcinoma; and the center of the type 3tumor lies 2 to 5 cm distal to the GEJ (subcardial gastriccarcinoma) [5].

Surgery is the mainstay treatment although the prognosis ispoor. Controversies, especially on operation route, still exist. Thedebate on the question whether transthoracic (TT) resection ornon- transthoracic resection is better for GEJ cancer remainscontinuing. Transthoracic resection was advocated with intent toprolong the survival, because mediastinal lymph nodes could beobserved and dissected under the direct vision and a safe surgical

PLoS ONE | www.plosone.org 1 June 2007 | Volume 7 | Issue 6 | e37698

Therewerenosignificantdifferencesofsurvivalrateandpostoperativemorbidityandmortalitybetweentransthoracicresectiongroupandnon-transthoracicresectiongroup.

Bothsurgicalapproachesareacceptable,andnooneoffersclearadvantageovertheother.

However,theresultsshouldbeinterpretedcautiouslysincethequalitiesofincludedstudiesweresuboptimal.


10189 August 7, 2014|Volume 20|Issue 29|WJG|www.wjgnet.com

resections. Transthoracic resection has given rise to sev-eral other approaches, including the left thoracic, right thoracic and even thoracoabdominal approaches, and different medical centers favor an individual transthoracic approach. Despite the potential for a wider resection of tumor margins and thoracic lymphadenectomy, pulmo-nary complications have been continuously reported after transthoracic resections. Transhiatal resection was first described by Grey Turner in 1933[18], and this approach often achieves lower short-term morbidity and mortality without formal thoracic lymphadenectomy. Until now, the proper approach for surgical treatment of cancers of the esophagogastric junction was still up for debate[19]. To clarify some of these issues, we undertook the present meta-analysis, concentrating on the objective analysis of the two approaches that are used to treat cancer of the esophagogastric junction.

Considering postoperative complications, pulmonary complications, such as pneumonia, atelectasis, acute respiratory distress syndrome and pulmonary embo-

lism, are the most prominent and often account for the majority of postoperative deaths[20]. Our results showed a significantly higher incidence of pulmonary complica-tions in the TT group, although Chou et al[21] reported no difference between the two groups in 2005. One potential rationale for explaining this is that thoracic sur-gery or combined upper abdominal resection carries a higher risk of postoperative pulmonary complications[22]. Nevertheless, early extubation and aggressive pulmo-nary rehabilitation[23] may reduce the rate of pulmonary complications. Anastomotic leakage, the main cause of mortality[24], was not significantly different between the groups. Previous studies have reported various outcomes in terms of anastomotic leakage[5,21,25]. This wide varia-tion can be attributed to different manual anastomoses, surgical anastomosis sites and levels of expertise[24,26]. Cardiovascular complications were also pooled with no significant difference between the groups, consistent with

that reported in a recent study comparing the two resec-tion approaches[27]. In terms of 30-d hospital mortality,

Study or subgroup log(hazard ratio) SE Weight Hazard ratio Hazard ratio

IV, random, 95%CI IV, random, 95%CI

Nakamura 2008 0.49 0.24 23.4% 1.63 (1.02, 2.61) Omloo 2007 -0.03 0.19 28.7% 0.97 (0.67, 1.41) Sasako 2006 0.3 0.24 23.4% 1.35 (0.84, 2.16) Zheng 2010 -0.28 0.23 24.4% 0.76 (0.48, 1.19)

Total (95%CI) 100.0% 1.11 (0.81, 1.54) Heterogeneity: Tau2 = 0.06; χ 2 = 6.53, df = 3 (P = 0.09); I 2 = 54% Test for overall effect: Z = 0.65 (P = 0.51)

0.01 0.1 1 1.0 100Favours transthoracic Favours transhiatal


IV, fixed, 95%CI IV, fixed, 95%CI

Omloo 2007 -0.05 0.32 100.0% 0.95 (0.51, 1.78)

Total (95%CI) 100.0% 0.95 (0.51, 1.78) Heterogeneity: Not applicable Test for overall effect: Z = 0.16 (P = 0.88)




Nakamura 2008 0.62 0.44 16.0% 1.86 (0.78, 4.40) Omloo 2007 -0.06 0.24 53.8% 0.94 (0.59, 1.51) Sasako 2006 0.17 0.32 30.2% 1.19 (0.63, 2.22)

Total (95%CI) 100.0% 1.13 (0.80, 1.59) Heterogeneity: χ 2 = 1.88, df = 2 (P = 0.39); I 2 = 0% Test for overall effect: Z = 0.67 (P = 0.50)




Nakamura 2008 0.54 0.59 28.2% 1.72 (0.54, 5.45) Sasako 2006 0.5 0.37 71.8% 1.65 (0.80, 3.40)

Total (95%CI) 100.0% 1.67 (0.90, 3.08) Heterogeneity: χ 2 = 0.00, df = 1 (P = 0.95); I 2 = 0% Test for overall effect: Z = 1.63 (P = 0.10)


Figure 5 Forest plot of overall survival in the transthoracic group vs transhiatal group of cancers of the esophagogastric junction. A: All Siewert types; B: Siewert Ⅰ; C: Siewert Ⅱ; D: Siewert Ⅲ. IV: Inverse variance.

Wei MT et al . TT vs TH for esophagogastric cancer

A

B

C

D

Transthoracic vs transhiatal surgery for cancer of the esophagogastric junction: A meta-analysis

Ming-Tian Wei, Yuan-Chuan Zhang, Xiang-Bing Deng, Ting-Han Yang, Ya-Zhou He, Zi-Qiang Wang

Ming-Tian Wei, Yuan-Chuan Zhang, Xiang-Bing Deng, Ting-Han Yang, Ya-Zhou He, Zi-Qiang Wang, Department of Gas-trointestinal Surgery, West China Hospital, Sichuan University, Chengdu 610041, Sichuan Province, ChinaAuthor contributions: Wei MT and Wang ZQ designed the research; Wang ZQ provided supervision; Wei MT and He YZ performed the research; Deng XB and Yang TH performed a literature search and collected the data; Wei MT and Zhang YC analyzed the data and wrote the paper.Supported by National Natural Science Foundation of China, No. 81172373Correspondence to: Zi-Qiang Wang, Professor, Department of Gastrointestinal Surgery, West China Hospital, Sichuan Uni-versity, No. 37 Guo Xue Alley, Chengdu 610041, Sichuan Prov-ince, China. [email protected]: +86-28-85422480 Fax: +86-28-81654035Received: November 20, 2013 Revised: February 13, 2014Accepted: March 5, 2014Published online: August 7, 2014

AbstractAIM: To compare the efficacy and safety of the trans-thoracic and transhiatal approaches for cancer of the esophagogastric junction.

METHODS: An electronic and manual search of the literature was conducted in PubMed, EmBase and the Cochrane Library for articles published between March 1998 and January 2013. The pooled data included the following parameters: duration of surgical time, blood loss, dissected lymph nodes, hospital stay time, anasto-motic leakage, pulmonary complications, cardiovascular complications, 30-d hospital mortality, and long-term survival. Sensitivity analysis was performed by exclud-ing single studies.

RESULTS: Eight studies including 1155 patients with cancer of the esophagogastric junction, with 639 pa-tients in the transthoracic group and 516 in the tran-shiatal group, were pooled for this study. There were

no significant differences between two groups concern-ing surgical time, blood loss, anastomotic leakage, or cardiovascular complications. Dissected lymph nodes also showed no significant differences between two groups in randomized controlled trials (RCTs) and non-RCTs. However, we did observe a shorter hospital stay (WMD = 1.92, 95%CI: 1.63-2.22, P < 0.00001), lower 30-d hospital mortality (OR = 3.21, 95%CI: 1.13-9.12, P = 0.03), and decreased pulmonary complications (OR = 2.95, 95%CI: 1.95-4.45, P < 0.00001) in the tran-shiatal group. For overall survival, a potential survival benefit was achieved for type Ⅲ tumors with the tran-shiatal approach.

CONCLUSION: The transhiatal approach for cancers of the esophagogastric junction, especially types Ⅲ, should be recommended, and its long-term outcome benefits should be further evaluated.

© 2014 Baishideng Publishing Group Inc. All rights reserved.

Key words: Transthoracic surgery; Transhiatal surgery; Cancer of the esophagogastric junction; Meta-analysis

Core tip: Surgical resection is the optimum therapy for cancer of the esophagogastric junction, and the transthoracic and transhiatal approaches are the two major surgical approaches used worldwide. However, considerable debate exists on the superior benefits of

the two approaches regarding their efficacy and safety.

We conducted this meta-analysis to address the issue. The results indicated a shorter hospital stay, lower 30-d hospital mortality and decreased pulmonary complica-tions with the transhiatal approach compared with the transthoracic approach. Moreover, a potential survival benefit was achieved for type Ⅲ tumors using the tran-shiatal approach.

Wei MT, Zhang YC, Deng XB, Yang TH, He YZ, Wang ZQ. Transthoracic vs transhiatal surgery for cancer of the esophago-

META-ANALYSIS

Submit a Manuscript: http://www.wjgnet.com/esps/Help Desk: http://www.wjgnet.com/esps/helpdesk.aspxDOI: 10.3748/wjg.v20.i29.10183


World J Gastroenterol 2014 August 7; 20(29): 10183-10192 ISSN 1007-9327 (print) ISSN 2219-2840 (online)


Transthoracic vs transhiatal surgery for cancer of the esophagogastric junction: A meta-analysis

Ming-Tian Wei, Yuan-Chuan Zhang, Xiang-Bing Deng, Ting-Han Yang, Ya-Zhou He, Zi-Qiang Wang

Ming-Tian Wei, Yuan-Chuan Zhang, Xiang-Bing Deng, Ting-Han Yang, Ya-Zhou He, Zi-Qiang Wang, Department of Gas-trointestinal Surgery, West China Hospital, Sichuan University, Chengdu 610041, Sichuan Province, ChinaAuthor contributions: Wei MT and Wang ZQ designed the research; Wang ZQ provided supervision; Wei MT and He YZ performed the research; Deng XB and Yang TH performed a literature search and collected the data; Wei MT and Zhang YC analyzed the data and wrote the paper.Supported by National Natural Science Foundation of China, No. 81172373Correspondence to: Zi-Qiang Wang, Professor, Department of Gastrointestinal Surgery, West China Hospital, Sichuan Uni-versity, No. 37 Guo Xue Alley, Chengdu 610041, Sichuan Prov-ince, China. [email protected]: +86-28-85422480 Fax: +86-28-81654035Received: November 20, 2013 Revised: February 13, 2014Accepted: March 5, 2014Published online: August 7, 2014

AbstractAIM: To compare the efficacy and safety of the trans-thoracic and transhiatal approaches for cancer of the esophagogastric junction.

METHODS: An electronic and manual search of the literature was conducted in PubMed, EmBase and the Cochrane Library for articles published between March 1998 and January 2013. The pooled data included the following parameters: duration of surgical time, blood loss, dissected lymph nodes, hospital stay time, anasto-motic leakage, pulmonary complications, cardiovascular complications, 30-d hospital mortality, and long-term survival. Sensitivity analysis was performed by exclud-ing single studies.

RESULTS: Eight studies including 1155 patients with cancer of the esophagogastric junction, with 639 pa-tients in the transthoracic group and 516 in the tran-shiatal group, were pooled for this study. There were

no significant differences between two groups concern-ing surgical time, blood loss, anastomotic leakage, or cardiovascular complications. Dissected lymph nodes also showed no significant differences between two groups in randomized controlled trials (RCTs) and non-RCTs. However, we did observe a shorter hospital stay (WMD = 1.92, 95%CI: 1.63-2.22, P < 0.00001), lower 30-d hospital mortality (OR = 3.21, 95%CI: 1.13-9.12, P = 0.03), and decreased pulmonary complications (OR = 2.95, 95%CI: 1.95-4.45, P < 0.00001) in the tran-shiatal group. For overall survival, a potential survival benefit was achieved for type Ⅲ tumors with the tran-shiatal approach.

CONCLUSION: The transhiatal approach for cancers of the esophagogastric junction, especially types Ⅲ, should be recommended, and its long-term outcome benefits should be further evaluated.


Key words: Transthoracic surgery; Transhiatal surgery; Cancer of the esophagogastric junction; Meta-analysis

Core tip: Surgical resection is the optimum therapy for cancer of the esophagogastric junction, and the transthoracic and transhiatal approaches are the two major surgical approaches used worldwide. However, considerable debate exists on the superior benefits of

the two approaches regarding their efficacy and safety.

We conducted this meta-analysis to address the issue. The results indicated a shorter hospital stay, lower 30-d hospital mortality and decreased pulmonary complica-tions with the transhiatal approach compared with the transthoracic approach. Moreover, a potential survival benefit was achieved for type Ⅲ tumors using the tran-shiatal approach.

Wei MT, Zhang YC, Deng XB, Yang TH, He YZ, Wang ZQ. Transthoracic vs transhiatal surgery for cancer of the esophago-

META-ANALYSIS

Submit a Manuscript: http://www.wjgnet.com/esps/Help Desk: http://www.wjgnet.com/esps/helpdesk.aspxDOI: 10.3748/wjg.v20.i29.10183


World J Gastroenterol 2014 August 7; 20(29): 10183-10192 ISSN 1007-9327 (print) ISSN 2219-2840 (online)


Foroverallsurvival,nosignificancewasfoundineitherallSiewert’stypesorsingleSiewert’s type.

A potentialsurvivalbenefitwasachievedfortypeIIItumorsusingthetranshiatalapproach comparedwiththetransthoracicapproach.

Weconcludethat,forcancersoftheesophagogastricjunction (especiallyforSiewert’s typeIIItumors)thetranshiatalapproachshouldberecommendedastheoptimalchoice.


Lymphaticpathways

whereas a median of 6 (range, 4–8) lymph node stations perpatient was identified when sulfur colloid was used (P !0.002; Mann–Whitney U test). The counting rate in thenodes with accumulation of nanocolloid ranged from 40 to450 cps (median, 226 cps), versus 30 to 2400 cps (median,230 cps) in the lymph nodes with accumulation of sulfurcolloid.In the 1 patient with paraesophageal lymph node drainage

on preoperative scintigraphy (Fig. 1; patient 9), a blueparaesophageal lymph node was identified but no radioac-tive accumulation could be detected by "-probe in theresection specimen ex vivo. In this same patient, another 6active lymph nodes were detected by perioperative probemeasurements and by postoperative probe measurements ofthe specimen.Of the 19 radioactive nodes measured by "-probe in the

specimens, only 7 were blue. In the lymph node stations thatremained in situ, the presence of patent blue could not beassessed accurately and therefore has not been evaluated.

DISCUSSION

This study showed that lymposcintigraphy of the gastriccardia is feasible and adds to our knowledge of the lym-phatic pathways of the gastric cardia. In the future, thistechnique might be helpful to determine the optimal surgicalprocedure for patients with cancer of the gastric cardia.The study used techniques recently developed for the

visualization of lymphatic drainage in other tumor types(especially breast cancer and melanoma). In these types ofcancer, the objective is to determine the sentinel nodes of

the tumor in a particular lymph node area: for example, theaxilla in breast cancer or the groin in melanoma of the lowerlimb. This study, however, did not investigate the visual-ization of sentinel nodes in stricto sensu. The objective wasto assess the feasibility of lymphoscintigraphy of the gastriccardia and to determine the normal lymphatic drainage,including first- and second-echelon lymph nodes of thenormal, nonmalignant gastric cardia. Baciewicz et al.showed the feasibility of lymphoscintigraphy of the esoph-agus in a canine model. After submucosal injection ofradiolabeled 99mTc-antimony sulfide colloid in 6 dogs,lymph nodes were identified on nuclear scans (10). Theexpected position of lymph nodes based on the scans cor-related with the location of the radiolabeled nodes at ana-tomic dissection.In the first part of this study, medium-sized nanocolloid

was administered as is used in, for example, breast cancer,but the radiolabeled colloid was subsequently changed intothe small-sized sulfur colloid to enhance transport of theradiotracer. Indeed, sulfur colloid appeared to be superior tonanocolloid for this purpose, because with sulfur colloidmore lymph nodes could be visualized. Moreover, the en-hanced transport from the injection site to these lymphnodes resulted in less hindrance from overprojection by theinjection site. This study indicated that 1 late scintigram(3–4 h after injection) is sufficient to obtain optimal infor-mation on lymphatic drainage of the upper digestive tract.Endoscopic injection of the radiolabeled colloid at the gas-troesophageal junction was relatively easy. Because theinjection site was marked by a clip, patent blue could be

TABLE 1Number of Active Lymph Node Stations Identified In Vivo During Surgery and in Resection Specimen After Surgery,

in Relation to Preoperative Lymphoscintigraphy

Lymph node station

Nanocolloid Sulfur colloid Total

Patient1

Patient2

Patient3

Patient4

Patient5

Total Patient6

Patient7

Patient8

Patient9

Patient10

Total RA B

RA B RA B

Intraoperative probemeasurement

Greater curvature, leftand right ## ## ## ## ## 0 0 ## ## ## ## ## 0 0 0 0

Supra- and infrapyloric ## ## ## ## ## 0 0 ## ## ## ## ## 0 0 0 0Celiac trunk ## $# ## $$ $$ 3 $$ $$ $$ $$ $$ 5 8Splenic hilum ## ## ## ## ## 0 ## ## $$ ## $# 2 2Splenic artery ## $$ ## ## ## 1 $$ $$ $$ $$ $$ 5 6Hepatoduodenal ligament ## ## ## ## ## 0 ## ## $# ## ## 1 1Mesenteric root ## ## ## ## ## 0 $$ ## $# ## ## 2 2Paraaortal ## ## ## ## ## 0 $$ ## $$ $$ ## 3 3

Probe measurement ofspecimen

Paracardial, right $$* ## ## $$ ## 2 1 ## $$ ## $$ $$ 3 0 5 1Paracardial, left $$ $$* ## ## ## 2 1 $$ ## $$ $$ ## 3 0 5 1Lesser curvature ## $$ $$* $$* $$* 4 3 $$ $$* $$ $$ $$* 5 2 9 5Paraesophageal # # # # # 0 0 # # # #$* # 1 1 1 1

Total 2 4 1 3 2 12 5 6 4 8 7 5 30 3 42 8

*Blue nodes (evaluated only in resection specimens).RA % radioactive; B % blue; ## % lymph node station negative for radioactivity by probe and by lymphoscintigraphy; $$ % lymph node station

positive for radioactivity by probe and by lymphoscintigraphy; $# % lymph node station positive for radioactivity by probe and negative bylymphoscintigraphy (false negative); #$ % lymph node station negative for radioactivity by probe and positive by lymphoscintigraphy.

250 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 45 • No. 2 • February 2004

according to stations (data is not presented) [3,20,21,24e26,29].Node metastasis in the lower mediastinal stations ranged from 7.5to 23.8%, whereas patients with upper mediastinal node involve-ment were below 4%. Patients with lymph node involvementextending to the proximal field of the chest (no. 106, 107) tendedto have a higher ratio of involved to total lymph nodes [34]. Onestudy found positive nodes in the chest in 24% of the patients [35].

Five-year survival related to each lymph node station

In Fig. 3 the 5-Y-S is shown in percent with median and rangeaccording to each lymph node station [3,19,20,25,30,36].

Stations in D1 has the best 5-Y-S rates overall, ranging from 0 to50%. Involvement of no. 4sb reduced the survival rate, although theincidence of metastasis was correspondingly low.

In D2 stations the 5-Y-S ranges from 0 to 53%. The best survivalrate is seen in station no. 7, the stationwith the highest lymph nodeinvolvement in this group. The lowest survival rate is seen in stationno. 9, with survival rates below 14.7%.

Metastasis to the lymph node stations in D3 shows overall verypoor survival. Data is not presented due to insufficiency of reporteddata. No patients with involvement of station no. 5, 12 and 111survived after five years. Only one study found patients alive after

five years with involvement of station 6 [19]. Few studies reportedsurvival in station 16 and 110, ranging from 0 to 23.8%[3,20,25,30,36].

Additional findings

The presence of lymph nodemetastasis seems to be significantlycorrelated to the depth of tumor invasion [28,36e38]. Ielpo et al.found an increasing rate of distal nodal involvement as the T-stageincreases [35]. Furthermore, lymph node metastases were signifi-cantly correlated to the size of the tumor, where large tumors(>4 cm) had significantly greater risk of lymph node involvement[37].

Some studies have investigated the influence of lymph nodemetastasis on prognosis. The majority concluded that the presenceof lymph node metastasis (Nþ) was an independent prognosticfactor, andwas associatedwith poor survival [3,25,27,37,39,40]. Onlyone study found no significantly difference between Nþ/N" [29].

A limited number of studies investigated survival rateswhen thefrequency of lymph node involvement was divided into N-stage(Table 2). There was an observable correlation between higher N-stage and poorer survival, with 5-Y-S rates from 2 to 17.4% in N3tumors compared to 5-Y-S in N0 up to 82.7% [5,19,30].

Figure 2. Lymph node metastasis according to JGCA stations in % with median and range.

Figure 3. 5-Y-S according to JGCA lymph node stations in % with median and range.

C. Okholm et al. / Surgical Oncology 23 (2014) 140e146 143

Cardia cancer was defined as adenocarcinoma of the GEJ. Thesearch was specified to Siewert's type II, since this was consideredthe true cardia carcinoma [13]. Siewert type I is mainly treated likeesophagus cancer, since most tumors arise from areas of intestinalmetaplasia in Barrett's epithelium as a consequence of reflux. Incontrast, type III tumors represent proximal gastric cancer andshould be approached in accordance with gastric cancer guidelines[14]. The characterization and optimal treatment for Siewert's typeII cardia cancer remains debatable, and is therefore the primaryfocus of this review.

Classifications of lymph node stations

The Japanese Gastric Cancer AssociationClassification guidelines published by JGCA for gastric cancer

including definitions of the lymph node stations, aimed atproviding a common language for the clinical and pathologicaldescription of gastric cancer. In 1998, the second English editionfrom JGCA classified lymph nodes into three groups depending onthe location of the primary tumor [15]. These three groups weresubdivided into 33 stations based on the results of studies of

lymphatic flow and observed survival associated with metastasis ineach station (Fig. 1).

The third English edition from JGCA (2011) [16] defined lymphnode stations 1e12 and 19, 20, 110 and 111 for tumors withesophageal invasion as regional lymph node metastasis; otherswere classified as distant metastasis. Moreover, an edition withspecial focus on treatment guidelines was published, in which theextent of lymphadenectomy was defined according to the type ofgastrectomy [17].

In principle, a D1 or a D1þ lymphadenectomy was indicated fornode negative tumors, and D2 was indicated for node positive orT2eT4 tumors. Since the pre- and perioperative diagnosis of lymphnodemetastases remains unreliable, it was recommended that a D2lymphadenectomy should be performed whenever nodal involve-ment was suspected. Although it is not prerequisite, the examina-tion of 16 or more regional lymph nodes was recommended for N-status determination [17].

Since most of the included studies describe lymph node stationsand dissections according to the JGCA 1998 edition [15], the resultsin this review will be presented according to this edition.

Results

Based on the search methods described, a total of 17 articleswere identified describing the pattern of lymph node spread andsix articles were identified describing the associated five-yearsurvival (5-Y-S). The six articles comprised of retrospective co-horts with a number of included patients ranging from 62 to 225(Table 1).

Pattern of nodal spread in cardia cancer

In Fig. 2, the incidence of lymph node metastases is presentedaccording to the JGCA lymph node stations divided into D1eD3dissections [3,18e33].

The highest incidence of lymph node metastasis is seen inregional lymph nodes, especially station no. 1e3, with a range from13.7 to 72.7%. The incidence in station no.4 is lower, ranging from0 to 31.8%. In the D2 stations the highest incidence of metastasis isobserved in no. 7, 9 and 11, ranging from 0 to 45.5%. The incidencesin D3 stations are remarkably lower with fewer than 20% of nodesinvolved.

Incidences and survival rates associated with mediastinallymph node involvement is poorly reported and few are classified

Table 1Included studies.

n Siewert n of LN stations 5-Y-S

Yamashita et al. [30] 225 II 19 xHoskokawa et al. [20] 179 I,II,III 17 xMine et al. [25] 150 II 15 xFujitani et al. [19] 86 II 13 xSaito et al. [3] 110 II 16 xWang et al. [28] 42 II 17Fang et al. [18] 51 II 16Matsumoto et al. [23] 58 I,II,III 16Nakamura et al. [26] 30 II 11Ichikura et al. [21] 31 II 10Feith et al. [14] 485 II 6Mattioli et al. [24] 116 II 8Yamamoto et al. [29] 67 II 4Kakeji et al. [22] 129 I,II,III 15Pedrazzani et al. [36] 62 I,II,III 14 xGertler et al. [32] 224 I,II,III 7Meier et al. [31] 167 II,III 9Yuasa et al. [33] 40 II 8

Figure 1. Station numbers of regional lymph nodes. Reprinted with permission fromthe Japan Esophageal Society, from the article “Japanese Classification of EsophagealCancer, tenth edition: printed in part I”, printed in Esophagus (2009) 6: 1e25.

C. Okholm et al. / Surgical Oncology 23 (2014) 140e146142

according to stations (data is not presented) [3,20,21,24e26,29].Node metastasis in the lower mediastinal stations ranged from 7.5to 23.8%, whereas patients with upper mediastinal node involve-ment were below 4%. Patients with lymph node involvementextending to the proximal field of the chest (no. 106, 107) tendedto have a higher ratio of involved to total lymph nodes [34]. Onestudy found positive nodes in the chest in 24% of the patients [35].

Five-year survival related to each lymph node station

In Fig. 3 the 5-Y-S is shown in percent with median and rangeaccording to each lymph node station [3,19,20,25,30,36].

Stations in D1 has the best 5-Y-S rates overall, ranging from 0 to50%. Involvement of no. 4sb reduced the survival rate, although theincidence of metastasis was correspondingly low.

In D2 stations the 5-Y-S ranges from 0 to 53%. The best survivalrate is seen in station no. 7, the stationwith the highest lymph nodeinvolvement in this group. The lowest survival rate is seen in stationno. 9, with survival rates below 14.7%.

Metastasis to the lymph node stations in D3 shows overall verypoor survival. Data is not presented due to insufficiency of reporteddata. No patients with involvement of station no. 5, 12 and 111survived after five years. Only one study found patients alive after

five years with involvement of station 6 [19]. Few studies reportedsurvival in station 16 and 110, ranging from 0 to 23.8%[3,20,25,30,36].

Additional findings

The presence of lymph nodemetastasis seems to be significantlycorrelated to the depth of tumor invasion [28,36e38]. Ielpo et al.found an increasing rate of distal nodal involvement as the T-stageincreases [35]. Furthermore, lymph node metastases were signifi-cantly correlated to the size of the tumor, where large tumors(>4 cm) had significantly greater risk of lymph node involvement[37].

Some studies have investigated the influence of lymph nodemetastasis on prognosis. The majority concluded that the presenceof lymph node metastasis (Nþ) was an independent prognosticfactor, andwas associatedwith poor survival [3,25,27,37,39,40]. Onlyone study found no significantly difference between Nþ/N" [29].

A limited number of studies investigated survival rateswhen thefrequency of lymph node involvement was divided into N-stage(Table 2). There was an observable correlation between higher N-stage and poorer survival, with 5-Y-S rates from 2 to 17.4% in N3tumors compared to 5-Y-S in N0 up to 82.7% [5,19,30].

Figure 2. Lymph node metastasis according to JGCA stations in % with median and range.

Figure 3. 5-Y-S according to JGCA lymph node stations in % with median and range.

C. Okholm et al. / Surgical Oncology 23 (2014) 140e146 143

scopically (volume, 1.0 mL) into the submucosa of the gastriccardia. After the first 5 patients, nanocolloid was exchanged forsulfur colloid to investigate whether the use of a smaller particlesize would show more intense adjacent lymph nodes and moredistant lymph nodes on the lymphoscintigraphy. The endoscopicprocedures were performed with a standard upper gastrointestinalendoscope (GIF-100 or GIF-140; Olympus Optical Co.). Duringthis preoperative procedure, the endoscope was introduced withthe patient in the left lateral position and was advanced into theesophagus. The injection site was chosen at the posterior wall ofthe lesser curvature, about 3 cm distal from the tubular esophagus.The radioactive tracer was injected into the submucosa using astandard 21-gauge needle (Boston Scientific Microvasive). Thisneedle had a dead volume of 1.0 mL, for which the total volumeand activity to be injected was corrected. After injection, theinjection site was marked with an endoscopically placed hemoclip(Olympus Optical Co.) for future reference.

LymphoscintigraphyLymphoscintigraphy of the lower chest and upper abdomen was

performed early (15–30 min) and late (4 and 20 h) after injectionof the radioactive tracers with a dual-head !-camera (Millennium;GE Medical Systems). Planar anterior and posterior acquisitions(600 s; 128" 128 matrix) were obtained at all time points. SPECT(128 " 128 matrix) was performed at 4 and 20 h after injection. Inthe last 5 patients, tracer accumulation was localized by CT(Hawkeye; GE Medical Systems) (Fig. 1).

Before surgery, paraesophageal lymph node drainage was pre-dicted scintigraphically by consensus reading (nuclear medicinephysician and surgeon). The specific lymph node stations wereretrospectively evaluated using the visualization on the variousscintigraphic acquisitions in combination with the knowledge ofthe peri- and postoperative probe measurements (reference stan-dard; vide infra).

Patent Blue Administration, Surgical Resection, andProbe MeasurementSurgery was started shortly after the last scintigraphic acquisi-

tion. Immediately after induction of anesthesia, a second endo-scopic procedure was performed. The endoscopically placed he-moclip was identified in the cardia, and at the same site, patentblue (sterile patent V 2.5% in aquadest and mannitol 5%; rawmaterial from Brunschwig Chemie) was injected into the submu-cosa using a similar standard 21-gauge needle (Fig. 2). Subse-quently, all patients underwent subtotal esophagectomy plus gas-tric cardia resection with limited lymph node dissection underdirect vision by a transhiatal approach via a widened hiatus. Aftercompletion of the resection, nondissected regional lymph nodestations were systematically measured in vivo for radioactivityusing a !-probe (Europrobe; PI Medical). Background measure-ments were performed on the jejunum and liver. Thereafter, theresection specimen itself was measured ex vivo for radioactivity,by !-probe, in the Nuclear Medicine Department. For verification,a scintigram of the resection specimen was obtained (Fig. 3).

FIGURE 1. Radioactivity in paraesopha-geal lymph nodes (PE) as identified preop-eratively by lymphoscintigraphy in combi-nation with CT. Shown are CT scan (left),scintigram (middle), and a combination ofCT scan and scintigram (right). This sagittalprojection also shows the injection site (IS)at the gastric cardia and celiac trunk nodes(CTN). See also patient 9 in Table 1.

248 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 45 • No. 2 • February 2004

Status and prognosis of lymph node metastasis in patients with cardiacancer: a systematic review of English publications dealing withadenocarcinoma of the cardia was conducted to elucidate patterns ofnodal spread and prognostic implications.


INDEXOFTHEBENEFITOFLYMPHNODEDISSECTIONofeachstationmultiplicationofthefrequencyofmetastasisatthestationandthe

5-yearsurvivalrateofpatientswithmetastasisatthatstationDissectionofperigastricnodes 4sb,4d,5,and6,offeredonlymarginaltherapeuticbenefit.Involvementofthelymphnodesinthesestationsappearedtorepresentdistantratherthanlocoregional metastasis.Theindexofestimatedbenefitfrommediastinallymphnodedissectionwasmarginal,exceptfortheperiesophageal nodes(110).Thisresultmightexplainwhytheoncologicoutcomesbetweenextendedesophagectomy andtotalgastrectomy arecomparable.

Yamashita et al Annals of Surgery ! Volume 254, Number 2, August 2011

TABLE 3. Incidence, 5-year Survival Rate, and Calculated Index of Estimated Benefit from LymphNode Dissection

Nodal station

Number ofpatients with

metastatic nodes

Number of patientsin whom the station

was dissectedIncidence of lymphnode metastasis (%)

5-year overall survivalrate of patients withmetastatic nodes (%) Index

No. 1 86 225 38.2 36.0 13.8No. 2 52 225 23.1 30.5 7.0No. 3 79 225 35.1 38.9 13.7No. 4sa 9 225 4.0 25.4 1.0No. 4sb 3 225 1.3 0.0 0.0No. 4d 0 169 0.0 NA NANo. 5 1 169 0.6 0.0 0.0No. 6 2 169 1.2 0.0 0.0No. 7 47 225 20.9 18.3 3.8No. 8a 14 225 6.2 35.7 2.2No. 9 23 225 10.2 14.7 1.5No. 10 6 147 4.1 16.7 0.7No. 11p 25 225 11.1 23.8 2.6No. 11d 12 173 6.9 32.4 2.2No. 12a 0 102 0.0 NA NANo. 110 7 95 7.4 23.8 1.8No. 111 3 95 3.2 0.0 0.0No. 112 1 79 1.3 0.0 0.0No. 16a2 8 73 11.0 12.5 1.4No. 16b1 7 38 18.4 0.0 0.0

Lymph node stations are defined as: 1 = right cardial; 2 = left cardial; 3 = lesser curvature; 4sa = along the short gastric artery; 4sb = alongthe left gastroepiploic artery; 4d = along the right gastroepiploic artery; 5 = suprapyloric; 6 = infrapyloric; 7 = along the left gastric artery; 8a =along the common hepatic artery (anterosuperior side); 9 = along the celiac artery; 10 = splenic hilum; 11p = along the proximal splenic artery;11d = along the distal splenic artery; 12a = along the proper hepatic artery; 110 = paraesophageal in the lower thorax; 111 = supradiaphragmatic;112 = posterior mediastinal; 16a2 = around the abdominal aorta (from the upper margin of the celiac trunk to the lower margin of the left renalvein); 16b1 = around the abdominal aorta (from the lower margin of the left renal vein to the upper margin of the inferior mesenteric artery)

NA indicates not analyzed.

TABLE 4. Survival Analysis According to the Number ofMetastatic Nodes in the Paracardial and Lesser CurveRegions

Overall Survival (median [95% CI], years)0 13.6 (10.0–13.8)1–2 9.7 (4.5–20.7)3– 2.1 (1.6–3.1)

5-year survival (% [95% CI])0 76.6% (64.6–85.4)1–2 62.3% (49.2–73.9)3– 22.4% (13.2–35.4)

Disease-free survival at 5 years (% [95% CI])0 69.1% (57.3–78.9)1–2 58.3% (45.0–70.4)3– 17.4% (9.5–29.7)

site of nodal recurrence, whereas nodal recurrence in the mediastinalnodes was less frequent (n = 4).

DISCUSSIONThe preponderance of male patients in our study was rea-

sonably consistent with the reports of other studies from Germany(5.4:1)6 and Taiwan (9.2:1)13 investigating Siewert type II EGJ tu-mors. Although the exact reason for the male predominance of thistype of cancer remains unknown, it seems to be a definite feature ofthis type of tumor, irrespective of the origin of the population.

The survival rate in our series was comparable with that re-ported from other studies: 5-year OS rate of 59.6% in a series of 51

patients13 and of about 40% in a series of 205 patients.6 Therefore,total gastrectomy for advanced carcinomas and proximal gastrectomyfor early carcinomas as a routine procedure for Siewert type II tumorsappears to be justified. To the best of our knowledge, there is no reporton the survival data by the tumor stage for patients with this tumorentity. Interestingly, the survival rates of pT2 and pT3 patients weresimilar, and the pT classification was not identified as an independentdeterminant of the survival. Meanwhile, the pN classification wasfound to be a powerful indicator of the clinical outcome, as shownin Figure 1. These findings suggested a possible biological differencebetween true cancer of the cardia and gastric cancer, with a cleardifference in the survival rate between pT2 and pT3 tumors beingreported for the latter.14

The most important prognostic factor has been reported tobe the lymph node status in EGJ cancers.6 Accordingly, the newnodal staging in the 7th TNM classification is based on the numberof metastatic nodes. It had been considered previously that all theregional nodes of the stomach were potentially involved by metastasisin patients with adenocarcinoma of the gastric cardia.15 Even aftera precise anatomic-topographic differentiation of this tumor entity,Siewert et al found a small number of patients with parapyloric nodemetastasis in their cohort with type II adenocarcinoma. Consistentwith their finding, in our patient series also, we found 1 patient withsuprapyloric node metastasis and 2 with infrapyloric node metastasis.However, these patients had involvement of multiple node stations andeventually died of the disease.

This study was undertaken in an attempt to define the optimalextent of lymph node dissection for true cancer of the cardia. Ourresults clearly indicated that dissection of the paracardial and lessercurve lymph nodes offered significant therapeutic benefit, suggesting

Copyright © 2011 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

278 | www.annalsofsurgery.com C⃝ 2011 Lippincott Williams & Wilkins



Nodal station


metastatic nodes









5-year survival (% [95% CI])0 76.6% (64.6–85.4)1–2 62.3% (49.2–73.9)3– 22.4% (13.2–35.4)













Nodal station


metastatic nodes









5-year survival (% [95% CI])0 76.6% (64.6–85.4)1–2 62.3% (49.2–73.9)3– 22.4% (13.2–35.4)











ORIGINAL ARTICLE

Optimal Extent of Lymph Node Dissection for Siewert Type IIEsophagogastric Junction Carcinoma

Hiroharu Yamashita, MD,∗ Hitoshi Katai, MD,∗ Shinji Morita, MD,∗ Makoto Saka, MD,∗ Hirokazu Taniguchi, MD,†

and Takeo Fukagawa, MD∗

Objective: To determine the optimal extent of lymph node dissection forcarcinomas of the true cardia, otherwise called Siewert type II esophagogastricjunction (EGJ) carcinomas.Background: In patients with cancer of the EGJ, comparable outcomes havebeen obtained with extended esophagectomy and total gastrectomy. The issueof the optimal surgical approach for EGJ tumors has been under debate.Nodal involvement is a strong predictor of survival, however, the optimalextent of prophylactic lymphadenectomy for Siewert type II tumors remainsto be elucidated.Methods: We retrospectively evaluated the distributions of the metastaticnodes, the recurrence pattern, and the oncological outcomes in a single-centerlarge cohort of 225 patients with Siewert type II tumors. To assess the thera-peutic outcomes of respective node dissection, we applied an index calculatedby multiplication of the incidence of metastasis by the 5-year survival rate ofpatients with metastasis in the respective node stations.Results: The incidence of nodal metastasis was high in the right paracardial(38.2%), lesser curve (35.1%) and left paracardial (23.1%) nodes, and also thenodes along the left gastric artery (20.9%). Involvement of the suprapancreaticnodes along the celiac artery, splenic artery and common hepatic artery wasfound in 23, 25, and 14 patients, respectively. According to the index ofestimated benefit from lymph node dissection, dissection of the paracardialand lesser curve nodes yielded the highest therapeutic benefit. The numberof metastatic nodes in these areas was as predictive of the disease-free andoverall survivals as the TNM pN category. The 5-year overall survival rates inpatients with no or 1–2 metastatic nodes were 76.6% and 62.3%, respectively,whereas the 5-year survival rate in those with 3 or more positive nodes wasonly 22.4%, comparable with the rate of 17.4% in patients with TNM pN3tumors.Conclusions: Clear anatomic distinction of EGJ tumors is likely to provideinsight into the appropriate extent of lymphadenectomy. Dissection of theparacardial and lesser curve nodes is essential for staging as well as for ob-taining therapeutic benefit in surgery for in EGJ carcinomas (Siewert type II).

(Ann Surg 2011;254:274–280)

E sophagogastric junction (EGJ) carcinoma, representing carci-noma involving the anatomical border between the esophagus and

the stomach, has attracted considerable attention recently, because ofthe marked increase in its incidence.1 In fact, the percentage of EGJcarcinomas among all gastric carcinomas has been estimated to haveincreased by 4-fold at our hospital during the past 4 decades.2 Thistype of carcinoma has been treated mainly as esophageal carcinoma

From the *Gastric Surgery Division and †Pathology Division, National CancerCenter Hospital, Tokyo, Japan.

Drs. H. Yamashita, H. Katai, S. Morita, M. Saka, H. Taniguchi, and T. Fukagawahave no conflicts of interest or financial ties to disclose.

This manuscript has not received any funding from commercial sponsors.Reprints: Hitoshi Katai, MD, Gastric Surgery Division, National Cancer Center

Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan, E-mail: [email protected].

Copyright C⃝ 2011 by Lippincott Williams & WilkinsISSN: 0003-4932/11/25402-0274DOI: 10.1097/SLA.0b013e3182263911

in the United States and as gastric carcinoma in Japan. The previousTNM staging systems for esophageal and gastric carcinomas weredifferent, and the resultant stage grouping for tumors straddling theEGJ was rather confusing. A new TNM classification (7th edition)has attempted harmonization between these tumors and provided aclear description of EGJ carcinomas.3

EGJ tumors have been classified according to the Siewert sys-tem based on their epicenters; type I, 5 to 1 cm above the EGJ; typeII, 1 cm above to 2 cm below the EGJ; and type III, 2–5 cm belowthe EGJ. Siewert type II and III EGJ carcinomas are the predominantsubtypes in Japan, in marked contrast to the extremely rare type Itumors.2 Notably, Siewert type II cancer, representing true carcinomaof the cardia, has been estimated to account for more than half of thetotal cases of EGJ cancer diagnosed during this last decade.2

The surgical management of patients with adenocarcinoma ofthe EGJ had been a matter of debate in both Japan and western coun-tries. Because type II tumors are located just at the esophagogastricborder, marked discrepancies in the selection of the surgical approachare noted. Extended esophagectomy is preferentially selected at someinstitutions,4,5 whereas extended total gastrectomy is the mainly em-ployed procedure in others.6,7 The former is based on the principlethat the proximal resection margin has a great impact on the survival,and the latter is based mainly on experimental and clinical obser-vations that the main lymphatic pathways from the cardia advancepreferentially downward to the para-celiac, and para-aortic nodes8;the main goal of either procedure, however, remains complete tumorremoval. As a matter of fact, the survival rates are comparable be-tween the 2 procedures, therefore, both procedures are considered tobe justified.

The beneficial effect of systematic lymph node dissection forgastric carcinoma has been well documented and acknowledged. Theoptimal extent of lymphadenectomy (D2) for this cancer has been de-fined in the Japanese Classification of Gastric Carcinoma,9 based onthe retrospective historical data of the involved nodes in patients withgastric carcinoma. On the other hand, the optimal extent of lymphnode dissection for Siewert type II EGJ carcinoma is poorly definedin this classification. Siewert et al. showed the distribution of themetastatic nodes in patients with type II adenocarcinoma. In their co-hort of 186 patients, they found that mainly the paracardial and lessercurve nodes, followed in frequency by the nodes in the lower me-diastinum, suprapancreatic nodes and nodes along the greater curvewere involved in patients with Siewert type II EGJ cancers. Further-more, they found positive parapyloric nodes in 3 of their patients,which might lend support to and justify their recommended strategyof extended total gastrectomy for type II EGJ carcinoma. In addi-tion to the lack of a precise description of the greater curve nodes,however, the association between the incidence of the nodal involve-ment in the respective node stations and the outcomes remains to beclarified.

In this study, we retrospectively examined the medical recordsof patients with Siewert type II EGJ cancer, which represents truecarcinoma of the cardia arising from the epithelium of the EGJ, tometiculously determine the anatomic sites of the lymph nodes in-volved and thereby clarify the optimal extent of prophylactic lymph


274 | www.annalsofsurgery.com Annals of Surgery ! Volume 254, Number 2, August 2011

single-centerlargecohortof225patientswithSiewert typeIItumors


The frequency of metastasis of each regional lymph

node, the 5-year survival rate of the patients with nodal

involvement, and the IEBLD for each station are shown inTables 3 and 4. The IEBLDs of stations located near the

EGJ were generally high. The index was higher than 5 in

stations 1 (right paracardia), 3 (lesser curvature), and 7

(along the left gastric artery) in patients with Siewert type

II AEG compared to stations 1, 2 (left paracardia), 3, 7, and

9 (along the celiac artery) in those with Siewert type IIIAEG.

On the contrary, the IEBLDs of stations located far from

the EGJ were low. It was zero in stations 4d, 5, 6, and 12a

Table 3 Frequency of lymphnode metastasis and 5-yearsurvival in patients with Siewerttype II adenocarcinoma of theesophagogastric junction basedon lymph node station

IEBLD Index of estimatedbenefit from lymph nodedissection

Lymphnodestation

Number ofpatients withmetastatic nodes

Number of patients inwhom the station wasdissected

Incidence oflymph nodemetastasis (%)

5-year survival rate ofpatients withmetastatic nodes (%)

IEBLD

1 36 92 39.1 36.6 14.3

2 12 92 13.0 15.9 2.1

3 34 92 37.0 45.4 16.8

4sa 2 92 2.2 0 0

4sb 2 92 2.2 50.0 1.1

4d 0 73 0 0 0

5 2 73 2.7 0 0

6 0 73 0 0 0

7 20 92 21.7 40.6 8.8

8a 3 85 3.5 50.0 1.8

9 15 87 17.2 22.9 3.9

10 3 52 5.8 0 0

11p 10 76 13.2 19.0 2.5

11d 2 65 3.1 0 0

12a 0 14 0 0 0

19 4 23 17.4 0 0

20 2 21 9.5 50.0 4.8

Table 4 Frequency of lymphnode metastasis and 5-yearsurvival in patients with Siewerttype III adenocarcinoma of theesophagogastric junction basedon lymph node station

IEBLD Index of estimatedbenefit from lymph nodedissection

Lymphnodestation

Number ofpatients withmetastatic nodes

Number of patients inwhom the station wasdissected

Incidence oflymph nodemetastasis (%)

5-year survival rate ofpatients withmetastatic nodes (%)

IEBLD

1 21 40 52.5 52.2 27.4

2 8 40 20.0 50.0 10.0

3 20 40 50.0 43.8 21.9

4sa 1 40 2.5 0 0

4sb 3 40 7.5 50.0 3.8

4d 4 38 10.5 25.0 2.6

5 2 38 5.3 50.0 2.6

6 0 38 0 0 0

7 8 40 20.0 40.0 8.0

8a 2 38 5.3 0 0

9 6 39 15.4 33.3 5.1

10 3 31 9.7 0 0

11p 4 36 11.1 0 0

11d 2 30 6.7 0 0

12a 0 12 0 0 0

19 1 5 20 0 0

20 0 2 0 0 0

378 H. Goto et al.

123

ORIGINAL ARTICLE

The optimal extent of lymph node dissection for adenocarcinomaof the esophagogastric junction differs between Siewert type IIand Siewert type III patients

Hironobu Goto • Masanori Tokunaga • Yuichiro Miki • Rie Makuuchi •

Norihiko Sugisawa • Yutaka Tanizawa • Etsuro Bando • Taiichi Kawamura •

Masahiro Niihara • Yasuhiro Tsubosa • Masanori Terashima

Received: 9 October 2013 / Accepted: 28 February 2014 / Published online: 22 March 2014! The Author(s) 2014. This article is published with open access at Springerlink.com

AbstractBackground The incidence of adenocarcinoma of the

esophagogastric junction (AEG) has been increasing

worldwide. We investigated the clinicopathological char-acteristics of patients with Siewert type II and III AEGs

and clarified the optimal intra-abdominal lymph node dis-

section in these patients.Methods This study included 132 patients with AEG who

underwent curative resection at Shizuoka Cancer Center

from September 2002 to December 2012. We used theindex of estimated benefit from lymph node dissection

(IEBLD) to assess the efficacy of lymph node dissection of

each station. The clinicopathological characteristics andIEBLDs of each station were compared between patients

with Siewert type II and III AEGs.

Results We analyzed 92 patients with Siewert type IIAEG and 40 patients with Siewert type III AEG. The

incidence of lymph node metastasis was high in both

groups (64.1 % in type II AEG and 75.0 % in type IIIAEG). The 5-year survival rates were similar for the

patients with Siewert type II and III AEGs, at 54.0 and53.4 %, respectively. The IEBLDs of stations located near

the esophagogastric junction were generally high in both

groups, while the IEBLDs of lower perigastric lymph

nodes were higher in Siewert type III than in Siewert typeII AEG cases.

Conclusions The IEBLDs were similar between Siewert

type II and III AEGs at all stations except for lower peri-gastric lymph nodes. Total gastrectomy should be selected

as a standard treatment for Siewert type III AEG, whereas

in Siewert type II AEG, preservation of the distal part ofthe stomach may be an acceptable procedure.

Keywords Gastric cancer ! Adenocarcinoma of the

esophagogastric junction ! Siewert type II ! Siewert type III

Introduction

The incidence of adenocarcinoma of the esophagogastric

junction (AEG) has been increasing recently in bothEastern and Western countries [1]. In Eastern countries,

westernized lifestyle habits and the increased incidence of

gastroesophageal reflux disease are thought to be possiblereasons, with the incidence of AEG likely to increase

further [2].Siewert et al. [3] classified AEG into three subgroups

according to the location of the tumor’s epicenter. Siewert

type I AEG is the most prevalent type in Western countriesand is generally treated as an esophageal cancer [4]. The

standard surgical procedure for Siewert type I AEG is a

subtotal esophagectomy with proximal gastrectomythrough thoracotomy [5]. Siewert type II and type III AEGs

are more common than Siewert type I AEG in Eastern

countries and are mostly treated as a gastric cancer with atrans-hiatal approach [6].

In contrast, the seventh edition of TNM classification

categorized AEG as an esophageal cancer irrespective ofthe Siewert type, and indeed, a current concern of surgeons,

H. Goto ! M. Tokunaga (&) ! Y. Miki ! R. Makuuchi !N. Sugisawa ! Y. Tanizawa ! E. Bando ! T. Kawamura !M. TerashimaDivision of Gastric Surgery, Shizuoka Cancer Center,1007 Shimonagakubo, Nagaizumi-cho, Sunto-gun,Shizuoka 411-8777, Japane-mail: [email protected]

M. Niihara ! Y. TsubosaDivision of Esophageal Surgery, Shizuoka Cancer Center,Shizuoka, Japan

123

Gastric Cancer (2015) 18:375–381

DOI 10.1007/s10120-014-0364-0

Accordingtotheresultsofthepresentstudy,totalgastrectomy shouldbeselectedasastandardtreatmentforSiewert typeIIIAEG.

Incontrast,preservationofthedistalpartofthestomachmaybeanacceptableprocedureinpatientswithSiewert typeIIAEG,becausethepresentstudydidnotshowsurvivalbenefitforlowerperigastriclymphnodedissection.

TheIEBLDofthesplenichilar lymphnodeswaszeroinpatientswithSiewerttypeIIandIIIAEG.

single-centerlargecohortof132patientswithSiewert typeIItumors


Original article

Lymphadenectomy around the left renal vein in Siewert type IIadenocarcinoma of the oesophagogastric junction

S. Mine, T. Sano, N. Hiki, K. Yamada, S. Nunobe and T. YamaguchiDepartment of Gastroenterological Surgery, Cancer Institute Hospital, 3-8-31, Ariake, Koto-ku, Tokyo, 135-8550, JapanCorrespondence to: Dr S. Mine (e-mail: [email protected])

Background: The extent of lymphadenectomy in patients with Siewert type II adenocarcinoma ofthe oesophagogastric junction is controversial. The aim of this study was to investigate lymph nodeinvolvement around the left renal vein.Methods: Lymph node involvement and prognosis in patients with Siewert type II cancers treated byR0–1 surgical resection were investigated, with regard to lymphadenectomy around the left renal vein.Based on the incidence of involvement at each node, the node stations were divided into three tiers (firsttier, more than 20 per cent involvement; second tier, 10–20 per cent involvement; third tier, less than10 per cent involvement).Results: Of 150 patients with type II oesophagogastric adenocarcinoma, 94 had left renal veinlymphadenectomy. The first lymph node tier included nodes along the lesser curvature, right cardia,left cardia and left gastric artery, with involvement of 28·0–46·0 per cent and a 5-year survival rate of42–53 per cent in patients with positive nodes. The nodes around the lower mediastinum, left renal vein,splenic artery and coeliac axis constituted the second tier, with involvement of 12·7–18 per cent and a5-year survival rate of 11–35 per cent. With regard to the left renal vein, the incidence of involvementwas 17 per cent and the 5-year rate survival rate was 19 per cent. Multivariable analysis showed that leftrenal vein lymphadenectomy was an independent prognostic factor in patients with pathological tumourcategory pathological T3–4 disease (hazard ratio 0·51, 95 per cent confidence interval 0·26 to 0·99;P = 0·048).Conclusion: Left renal vein nodal involvement is similar to that seen along the splenic artery, in thelower mediastinum and coeliac axis, with similar impact on patient survival.

Paper accepted 6 September 2012Published online in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.8967

Introduction

The incidence of adenocarcinoma at the oesophagogastricjunction has increased dramatically in the past fewdecades in European countries and the USA1–3. InEastern countries, including Japan, the incidence has alsobeen rising4,5, although not so rapidly as in the West.For most patients, surgery alone has been associatedwith poor survival, making the development of effectivemultidisciplinary treatments imperative6. Nevertheless,complete surgical removal is necessary even in the contextof multidisciplinary treatment.

The Siewert classification for junctional cancers7 isused widely, dividing tumours into three types. Siewerttype I represents lower oesophageal adenocarcinomagenerally considered to arise on the background of Barrett’s

oesophagus; type II tumours are those that seem to arise atthe oesophagogastric junction (cardia cancers) and typeIII are considered gastric adenocarcinomas arising inthe subcardial region. With regard to Siewert type II,no operative procedure has been recognized as standardregarding either approach, extent of resection or degreeof lymphadenectomy. Stein and colleagues8,9 maintainedthat Siewert type II adenocarcinoma should be treated bytotal gastrectomy via a transhiatal approach. On the otherhand, in a study in which more than half of the patients hadSiewert type II lesions, patients treated by oesophagectomyhad a better prognosis than those having gastrectomy10.

The lymph nodes around the left renal vein areconsidered as extraregional in the tumour node metastasis(TNM) classification11 and lymphadenectomy here wouldnot be standard in most institutions12–14. Previous studies

© 2012 British Journal of Surgery Society Ltd British Journal of Surgery 2013; 100: 261–266Published by John Wiley & Sons Ltd

264 S. Mine, T. Sano, N. Hiki, K. Yamada, S. Nunobe and T. Yamaguchi

Table 2 Pathological features

No. of patients (n = 150)

Tumour categorypT1 20 (13·3)pT2 23 (15·3)pT3 70 (46·7)pT4 37 (24·7)

Node categorypN0 45 (30·0)pN1 33 (22·0)pN2 41 (27·3)pN3 31 (20·7)

Metastasis categorypM0 129 (86·0)pM1 21 (14·0)

Stage (TNM, 7th edn)I 30 (20·0)II 18 (12·0)III 81 (54·0)IV 21 (14·0)

ResectionR0 144 (96·0)R1 6 (4·0)

No. of dissected lymph nodes* 47 (7–107)No. of metastatic lymph nodes* 2 (0–35)Histological type

Differentiated 99 (66·0)Undifferentiated 46 (30·7)Other 5 (3·3)

Values in parentheses are percentages unless indicated otherwise; *valuesare median (range). (p)TNM, (pathological) tumour node metastasis.

Table 3 Incidence and 5-year survival rate in relation to lymphnode station

Lymph node stationIncidence ofinvolvement*

5-year survivalrate of patients

with positivenodes (%)

First tierLesser curvature (no. 3) 69 of 150 (46·0) 45Right cardia (no. 1) 67 of 150 (44·7) 42Left cardia (no. 2) 52 of 150 (34·7) 44Left gastric artery (no. 7) 42 of 150 (28·0) 53

Second tierLower mediastinum 18 of 99 (18) 35Left renal vein area (no. 16A2lat) 16 of 94 (17) 19Splenic artery (no. 11) 23 of 139 (16·5) 23Coeliac axis (no. 9) 19 of 150 (12·7) 11

Third tierProper hepatic artery (no. 12) 3 of 39 (8) 0Splenic hilum (no. 10) 5 of 107 (4·7) 40Upper or middle mediastinum 2 of 46 (4) 50Greater curvature (no. 4) 5 of 146 (3·4) 0Common hepatic artery (no. 8) 3 of 130 (2·3) 0Suprapyloric (no. 5) 1 of 102 (1·0) 0Infrapyloric (no. 6) 0 of 102 (0) —

*Number of patients with involved nodes as a proportion of number whounderwent lymphadenectomy.

Table 4 Multivariable analysis using Cox proportional hazardsmodel of prognostic factors

Hazard ratio P

Neoadjuvant chemotherapy (yesversus no)

2·07 (0·98, 4·38) 0·056

Node involvement (pN2–3 versuspN0–1)

2·51 (1·29, 4·90) 0·007

Metastasis (pM1 versus pM0) 1·26 (0·66, 2·40) 0·479Lymphadenectomy of no. 16A2lat

(yes versus no)0·51 (0·26, 0·99) 0·048

Values in parentheses are 95 per cent confidence intervals. pN,pathological node category; pM, pathological metastasis category.

Table 5 Characteristics of patients with involvement ofno. 16A2lat (left renal vein)

No. 16A2lat

Positive(n = 16)

Negative(n = 78) P

Tumour category 0·520†pT1 0 4pT2 1 7pT3 12 44pT4 3 23

Node category < 0·001†pN0 0 15pN1 0 20pN2 4 29pN3 12 14

Tumour size (mm)* 49 (25–90) 55 (12–135) 0·194‡Length of oesophageal

invasion (mm)*18 (2–55) 22 (5–70) 0·426‡

*Values are median (range). pT, pathological tumour category; pN,pathological node category. †χ2 test; ‡Mann–Whitney U test.

ratio (HR) 2·51, 95 per cent confidence interval 1·29 to4·90; P = 0·007) and lymphadenectomy of no. 16A2lat(HR 0·51, 0·26 to 0·99; P = 0·048) were independentlyprognostic (Table 4).

Higher pN status was significantly related to no. 16A2latinvolvement (P < 0·001), but there was no statisticallysignificant difference between patients with and withoutno. 16A2lat involvement in terms of pT (P = 0·520),tumour size (P = 0·194) or length of oesophageal invasion(P = 0·426) (Table 5).

Discussion

In line with previous findings9,12,19, the present studyconfirmed that nodal station numbers 3 (lesser curvature),1 (right cardia), 2 (left cardia) and 7 (left gastric artery)were most frequently involved in type II junctional

© 2012 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery 2013; 100: 261–266Published by John Wiley & Sons Ltd

Nodestationsweredividedintothreetiers:firsttier,morethan20percentinvolvement;secondtier,10–20percentinvolvement;

thirdtier,lessthan10percentinvolvement.©2015MarcoLotti– [email protected]

Shouldwelookbetterattheholeweareabouttogothrough?


Themammaliandiaphragmisprimarilyarespiratorymuscle.However,itshouldbeconsideredastwoseparatemusclesconsistingofthecrural andthecostaldiaphragms.

• Duringhumandevelopmentmyoblastsoriginatinginthebodywallandderivedfromthecervicalsegmentsinvadetwopleuroperitonealmembranesandformthecostaldiaphragm.

• Ontheotherhand,thetwocrura developinthemesenteryoftheesophagus.


Thecrural diaphragmhastwofunctions:respiratoryandgastrointestinal

• Theupperpartisfullymuscularandmeasures2.5cminlength.

• Thelowerpartformsagutterthatisopenanteriorly.

• Thecentralfibershavearelativelycirculararrangement.

• Theperipheralfibersareorientedinacraniocaudal direction.

• Theuniquearrangementofitsmusclefibersresultsintwodifferenttypesofactionsontheesophagus whenitcontracts:averticalorcraniocaudal motion,andacircumferentialsqueeze.


• Allofthemaneuversthatincreasegastroesophagealpressuregradientsareaccompaniedbycontractionofdiaphragmaticsphincter.

• Thustherapidchangesinpressuregradientsbetweenesophagusandstomach,causedbyskeletalmusclecontractionofthechestandabdomen,areantagonizedbyrapidlycontractingskeletalsphinctermusclesofthediaphragmaticsphincter.


Whatdoesthehiatusdowhenitdoesn’tknowitis

observed


Lymphaticflow fromtheesophagus


Lymphaticflowwithintheesophagus


FEATURE

Adenocarcinoma of the Gastroesophageal JunctionInfluence of Esophageal Resection Margin and Operative Approach

on Outcome

Andrew P. Barbour, MD, PhD,* Nabil P. Rizk, MD,* Mithat Gonen, PhD,† Laura Tang, MD, PhD,‡Manjit S. Bains, MD,* Valerie W. Rusch, MD,* Daniel G. Coit, MD,* and Murray F. Brennan, MD*

Objective: To determine whether the length of esophageal resectionor the operative approach influences outcome for patients withadenocarcinoma of the gastroesophageal junction (GEJ).Summary Background Data: While R0 resection remains themainstay of curative treatment of patients with GEJ cancer, theoptimal length of esophageal resection remains controversial.Methods: Patients with Siewert I, II, or III adenocarcinoma whounderwent complete gross resection without neoadjuvant therapywere identified from a prospectively maintained database. Proximalmargin lengths were recorded ex vivo as the distance from the grosstumor edge to the esophageal transection line. Operative approacheswere grouped into gastrectomy (limited esophagectomy) or esoph-agectomy (extended esophagectomy).Results: From 1985 through 2003, 505 patients underwent R0/R1gastrectomy (n ! 153) or esophagectomy (n ! 352) withoutneoadjuvant treatment. There were no differences in R1 resectionrate, number of nodes examined or operative mortality betweengastrectomy and esophagectomy. Univariate analysis found "3.8cm to be the ex vivo proximal margin length (approximately 5 cm insitu) most predictive of improved survival. Multivariable analysis inpatients who underwent R0 resection with !15 lymph nodes exam-ined (n ! 275) found the number of positive lymph nodes, T stage,tumor grade, and ex vivo proximal margin length "3.8 cm to beindependent prognostic factors. Subset analysis found that the ben-efit associated with "3.8 cm margin was limited to patients with T2or greater tumors and "6 positive lymph nodes.Conclusions: In patients not receiving neoadjuvant therapy, the goalfor patients with adenocarcinoma of the GEJ should be R0 resectionincluding at least 15 lymph nodes, preferably with 5 cm of grosslynormal in situ proximal esophagus for those with "6 positive lymphnodes. The operative approach may be individualized to achievethese goals.

(Ann Surg 2007;246: 1–8)

Adenocarcinoma of the gastroesophageal junction (GEJ)remains a significant clinical problem that is increasing in

incidence1 and is associated with a poor prognosis.2–4 Themajority of patients present with advanced disease and lessthan 50% undergo curative treatment.5 Siewert and Steinhave proposed a clinical classification system, types I to III,for GEJ cancer to aid clinicians in developing treatmentstrategies for this heterogeneous clinical entity.6

While surgical extirpation of the primary tumor re-mains the mainstay of curative treatment of patients with GEJcancer, the extent of esophageal resection remains controver-sial. Tumor infiltration of the proximal or distal resectionmargin has been associated with diminished survival in mostseries.7–10 To minimize the risk of positive proximal margins,resection of up to 10 cm of grossly normal esophagus prox-imal to the most cephalad extent of the primary tumor hasbeen advocated by several groups.2,7,9,11,12 However, proxi-mal resection margin infiltration occurs more frequently inpalliative resections13 and does not universally translate topoor survival in curative resections.14,15 Shorter proximalmargins have also been correlated with local recurrence inboth proximal gastric cancer16 and squamous cell carcinomaof the esophagus,9,17 but the relationship between extendedesophageal resection and survival has not been well studied.

The operative approaches to patients with cancer of theGEJ vary widely and include esophagectomy via transtho-racic (TTE) or transhiatal (THE) approaches, total gastrec-tomy (TG) or proximal gastrectomy via laparotomy or leftthoracoabdominal incisions.18,19 Extended resections,20,21 in-cluding D2 lymphadenectomy,22 have been proposed in anattempt to reduce the incidence of positive margins, improveloco-regional control and survival. Lymph nodes in the lowermediastinum are involved in up to 40% of cases.3,23–26

However, the majority of positive nodes associated withSiewert type II and III cancers involve the paracardial, leftgastric, and lesser curve nodes, which are included in bothabdominal and thoracic approaches and the incidence oflymph node positivity is not influenced by the surgical ap-proach.27 The majority of patients experience hematogenousrecurrence and lymph node recurrences that are predomi-nantly found in para-aortic or porta hepatic nodes, stationsthat are not included in either operation.28

From the Departments of *Surgery, †Epidemiology-Biostatistics, and ‡Pa-thology, Memorial Sloan-Kettering Cancer Center, New York, NY.

Reprints: Murray F. Brennan, MD, Memorial Sloan-Kettering Cancer Cen-ter, Department of Surgery, 1275 York Avenue, New York, NY, 10021.E-mail: [email protected].

Copyright © 2007 by Lippincott Williams & WilkinsISSN: 0003-4932/07/24601-0001DOI: 10.1097/01.sla.0000255563.65157.d2

Annals of Surgery • Volume 246, Number 1, July 2007 1

FEATURE

Adenocarcinoma of the Gastroesophageal JunctionInfluence of Esophageal Resection Margin and Operative Approach

on Outcome

Andrew P. Barbour, MD, PhD,* Nabil P. Rizk, MD,* Mithat Gonen, PhD,† Laura Tang, MD, PhD,‡Manjit S. Bains, MD,* Valerie W. Rusch, MD,* Daniel G. Coit, MD,* and Murray F. Brennan, MD*

Objective: To determine whether the length of esophageal resectionor the operative approach influences outcome for patients withadenocarcinoma of the gastroesophageal junction (GEJ).Summary Background Data: While R0 resection remains themainstay of curative treatment of patients with GEJ cancer, theoptimal length of esophageal resection remains controversial.Methods: Patients with Siewert I, II, or III adenocarcinoma whounderwent complete gross resection without neoadjuvant therapywere identified from a prospectively maintained database. Proximalmargin lengths were recorded ex vivo as the distance from the grosstumor edge to the esophageal transection line. Operative approacheswere grouped into gastrectomy (limited esophagectomy) or esoph-agectomy (extended esophagectomy).Results: From 1985 through 2003, 505 patients underwent R0/R1gastrectomy (n ! 153) or esophagectomy (n ! 352) withoutneoadjuvant treatment. There were no differences in R1 resectionrate, number of nodes examined or operative mortality betweengastrectomy and esophagectomy. Univariate analysis found "3.8cm to be the ex vivo proximal margin length (approximately 5 cm insitu) most predictive of improved survival. Multivariable analysis inpatients who underwent R0 resection with !15 lymph nodes exam-ined (n ! 275) found the number of positive lymph nodes, T stage,tumor grade, and ex vivo proximal margin length "3.8 cm to beindependent prognostic factors. Subset analysis found that the ben-efit associated with "3.8 cm margin was limited to patients with T2or greater tumors and "6 positive lymph nodes.Conclusions: In patients not receiving neoadjuvant therapy, the goalfor patients with adenocarcinoma of the GEJ should be R0 resectionincluding at least 15 lymph nodes, preferably with 5 cm of grosslynormal in situ proximal esophagus for those with "6 positive lymphnodes. The operative approach may be individualized to achievethese goals.

(Ann Surg 2007;246: 1–8)

Adenocarcinoma of the gastroesophageal junction (GEJ)remains a significant clinical problem that is increasing in

incidence1 and is associated with a poor prognosis.2–4 Themajority of patients present with advanced disease and lessthan 50% undergo curative treatment.5 Siewert and Steinhave proposed a clinical classification system, types I to III,for GEJ cancer to aid clinicians in developing treatmentstrategies for this heterogeneous clinical entity.6

While surgical extirpation of the primary tumor re-mains the mainstay of curative treatment of patients with GEJcancer, the extent of esophageal resection remains controver-sial. Tumor infiltration of the proximal or distal resectionmargin has been associated with diminished survival in mostseries.7–10 To minimize the risk of positive proximal margins,resection of up to 10 cm of grossly normal esophagus prox-imal to the most cephalad extent of the primary tumor hasbeen advocated by several groups.2,7,9,11,12 However, proxi-mal resection margin infiltration occurs more frequently inpalliative resections13 and does not universally translate topoor survival in curative resections.14,15 Shorter proximalmargins have also been correlated with local recurrence inboth proximal gastric cancer16 and squamous cell carcinomaof the esophagus,9,17 but the relationship between extendedesophageal resection and survival has not been well studied.

The operative approaches to patients with cancer of theGEJ vary widely and include esophagectomy via transtho-racic (TTE) or transhiatal (THE) approaches, total gastrec-tomy (TG) or proximal gastrectomy via laparotomy or leftthoracoabdominal incisions.18,19 Extended resections,20,21 in-cluding D2 lymphadenectomy,22 have been proposed in anattempt to reduce the incidence of positive margins, improveloco-regional control and survival. Lymph nodes in the lowermediastinum are involved in up to 40% of cases.3,23–26

However, the majority of positive nodes associated withSiewert type II and III cancers involve the paracardial, leftgastric, and lesser curve nodes, which are included in bothabdominal and thoracic approaches and the incidence oflymph node positivity is not influenced by the surgical ap-proach.27 The majority of patients experience hematogenousrecurrence and lymph node recurrences that are predomi-nantly found in para-aortic or porta hepatic nodes, stationsthat are not included in either operation.28

From the Departments of *Surgery, †Epidemiology-Biostatistics, and ‡Pa-thology, Memorial Sloan-Kettering Cancer Center, New York, NY.

Reprints: Murray F. Brennan, MD, Memorial Sloan-Kettering Cancer Cen-ter, Department of Surgery, 1275 York Avenue, New York, NY, 10021.E-mail: [email protected].

Copyright © 2007 by Lippincott Williams & WilkinsISSN: 0003-4932/07/24601-0001DOI: 10.1097/01.sla.0000255563.65157.d2

Annals of Surgery • Volume 246, Number 1, July 2007 1

tionship between proximal margin and operative approach.Gastrectomy with limited esophagectomy resulted in signif-icantly shorter proximal margins for Siewert I (median, 1.75cm; range 0.3–8.0 cm), Siewert II (median, 2.0 cm; range,0.1–6.5 cm) and Siewert III tumors (median, 1.5 cm; range,0.2–7.0 cm) compared with extended esophagectomy (me-dian, 4.0; range, 0.0–13.0 cm for type I, P ! 0.009; median,5.5 cm; range, 0.3–16.0 cm for type II, P " 0.0001, Wil-coxon; and median, 5.5 cm; range, 0.0–13.5 cm for type III,P " 0.0001, Wilcoxon). The majority of patients with Siew-ert type I (90%) and type II (80%) underwent extendedesophagectomy (Table 1, P " 0.0001, Fisher exact test). Incontrast, 52% of patients with Siewert type III tumors weretreated by gastrectomy with limited esophagectomy.

There were no significant differences between the gas-trectomy with limited esophagectomy and extended esopha-gectomy groups with respect to gender, procedure-relatedmortality (Table 3), or the number of lymph nodes removed(Table 4). Similarly, there was no significant difference in theproportion of gastrectomy and esophagectomy patients thathad !15 nodes examined (63% and 68%, respectively; Table3). However, the gastrectomy patients were statistically sig-nificantly older (Table 3, P ! 0.004, Wilcoxon), higherpathologic stage (Table 3, P ! 0.0001, "2), and had signifi-cantly more positive lymph nodes (Table 4, P ! 0.003,Wilcoxon) compared with the extended esophagectomygroup. There was no difference in the rate of R1 resectionsbetween the 2 surgical approaches. Furthermore, when theinvolved margins were analyzed by site, the incidences ofmicroscopically positive proximal, distal or deep marginswere similar for the 2 surgical approaches (Table 3).

Analysis of survival was performed on the 275 patientsthat underwent curative (R0) resection with !15 lymphnodes examined. Figure 2 reveals that the gastrectomy withlimited esophagectomy group had statistically significantly

poorer OS (median, 22 months; 5-year OS, 27%) comparedwith the extended esophagectomy group (median, 37 months;5-year OS, 37%, P ! 0.02, log rank) when not corrected forstage.

Multivariable Survival AnalysesActuarial survival analyses by the Kaplan–Meier

method did not find Siewert type (Table 5, P ! 0.38, logrank) to be associated with OS for R0 resected patients with!15 lymph nodes examined. Multivariable analysis was thenperformed for the 275 patients that underwent R0 resectionwith !15 lymph nodes examined and no missing variables(33 of 308 patients with !15 lymph nodes examined did nothave complete data). A proportional hazards regressionmodel was developed using prognostic variables that havebeen validated in the development of a nomogram for gastricand GEJ cancer,36,37 including age at diagnosis, gender, pTstage (AJCC), tumor size (cm), number of positive lymphnodes, number of negative lymph nodes, and tumor grade inaddition to Siewert type, operative approach and gross prox-imal margin #3.8 cm (Table 5). This analysis found numberof positive nodes (HR, 1.06 per positive node; 95% CI,1.04–1.09, P " 0.01), AJCC T stage (Table 6, P " 0.01),proximal margin #3.8 cm (HR, 0.69; 95% CI, 0.50–0.97, P !0.03) and poor differentiation (HR, 1.44; 95% CI, 1.01–2.04,

FIGURE 1. Esophageal transection greater than 3.8 cmabove the most cephalad extent of the tumor was highlypredictive of improved OS for 275 patients who underwentR0 resection with !15 lymph nodes removed (P ! 0.0004,log rank).

TABLE 3. Clinicopathologic Variables for Patients Treatedby Gastrectomy Compared With Those Treated byEsophagectomy

Clinicopathologic Factor Gastrectomy Esophagectomy P

Male 119 (78) 292 (83) 0.117Female 34 (22) 60 (17)Age (yr) (median) 68 65 0.004Postop. death 7 (4.6) 16 (4.5) 0.988Siewert type I 12 (8) 100 (28) "0.001Siewert type II 77 (50) 199 (57)Siewert type III 64 (42) 53 (15)pT stage 1 23 (15) 87 (24) 0.004pT stage 2 29 (19) 80 (23)pT stage 3 96 (63) 183 (52)pT stage 4 5 (3) 2 (1)pN stage 0 49 (32) 147 (42) 0.008pN stage 1 58 (38) 144 (41)pN stage 2 34 (22) 49 (14)pN stage 3 12 (8) 12 (3)AJCC stage I 33 (22) 123 (35) 0.001AJCC stage II 43 (21) 66 (19)AJCC stage III 70 (46) 150 (43)AJCC stage IV 17 (11) 13 (4)R1 14 (9) 35 (10) 0.782Positive proximal margin 7 (5) 7 (2)* 0.136Positive distal margin 1 (1) 4 (1)* 0.813Positive deep margin 6 (4) 26 (7) 0.167Proximal margin #3.8 cm 15 (10) 201 (57) "0.001!15 nodes removed 96 (63) 240 (68) 0.218Total 153 352

*Two patients had both proximal and distal margins involved.

Barbour et al Annals of Surgery • Volume 246, Number 1, July 2007


tionship between proximal margin and operative approach.Gastrectomy with limited esophagectomy resulted in signif-icantly shorter proximal margins for Siewert I (median, 1.75cm; range 0.3–8.0 cm), Siewert II (median, 2.0 cm; range,0.1–6.5 cm) and Siewert III tumors (median, 1.5 cm; range,0.2–7.0 cm) compared with extended esophagectomy (me-dian, 4.0; range, 0.0–13.0 cm for type I, P ! 0.009; median,5.5 cm; range, 0.3–16.0 cm for type II, P " 0.0001, Wil-coxon; and median, 5.5 cm; range, 0.0–13.5 cm for type III,P " 0.0001, Wilcoxon). The majority of patients with Siew-ert type I (90%) and type II (80%) underwent extendedesophagectomy (Table 1, P " 0.0001, Fisher exact test). Incontrast, 52% of patients with Siewert type III tumors weretreated by gastrectomy with limited esophagectomy.

There were no significant differences between the gas-trectomy with limited esophagectomy and extended esopha-gectomy groups with respect to gender, procedure-relatedmortality (Table 3), or the number of lymph nodes removed(Table 4). Similarly, there was no significant difference in theproportion of gastrectomy and esophagectomy patients thathad !15 nodes examined (63% and 68%, respectively; Table3). However, the gastrectomy patients were statistically sig-nificantly older (Table 3, P ! 0.004, Wilcoxon), higherpathologic stage (Table 3, P ! 0.0001, "2), and had signifi-cantly more positive lymph nodes (Table 4, P ! 0.003,Wilcoxon) compared with the extended esophagectomygroup. There was no difference in the rate of R1 resectionsbetween the 2 surgical approaches. Furthermore, when theinvolved margins were analyzed by site, the incidences ofmicroscopically positive proximal, distal or deep marginswere similar for the 2 surgical approaches (Table 3).

Analysis of survival was performed on the 275 patientsthat underwent curative (R0) resection with !15 lymphnodes examined. Figure 2 reveals that the gastrectomy withlimited esophagectomy group had statistically significantly

poorer OS (median, 22 months; 5-year OS, 27%) comparedwith the extended esophagectomy group (median, 37 months;5-year OS, 37%, P ! 0.02, log rank) when not corrected forstage.

Multivariable Survival AnalysesActuarial survival analyses by the Kaplan–Meier

method did not find Siewert type (Table 5, P ! 0.38, logrank) to be associated with OS for R0 resected patients with!15 lymph nodes examined. Multivariable analysis was thenperformed for the 275 patients that underwent R0 resectionwith !15 lymph nodes examined and no missing variables(33 of 308 patients with !15 lymph nodes examined did nothave complete data). A proportional hazards regressionmodel was developed using prognostic variables that havebeen validated in the development of a nomogram for gastricand GEJ cancer,36,37 including age at diagnosis, gender, pTstage (AJCC), tumor size (cm), number of positive lymphnodes, number of negative lymph nodes, and tumor grade inaddition to Siewert type, operative approach and gross prox-imal margin #3.8 cm (Table 5). This analysis found numberof positive nodes (HR, 1.06 per positive node; 95% CI,1.04–1.09, P " 0.01), AJCC T stage (Table 6, P " 0.01),proximal margin #3.8 cm (HR, 0.69; 95% CI, 0.50–0.97, P !0.03) and poor differentiation (HR, 1.44; 95% CI, 1.01–2.04,

FIGURE 1. Esophageal transection greater than 3.8 cmabove the most cephalad extent of the tumor was highlypredictive of improved OS for 275 patients who underwentR0 resection with !15 lymph nodes removed (P ! 0.0004,log rank).

TABLE 3. Clinicopathologic Variables for Patients Treatedby Gastrectomy Compared With Those Treated byEsophagectomy

Clinicopathologic Factor Gastrectomy Esophagectomy P

Male 119 (78) 292 (83) 0.117Female 34 (22) 60 (17)Age (yr) (median) 68 65 0.004Postop. death 7 (4.6) 16 (4.5) 0.988Siewert type I 12 (8) 100 (28) "0.001Siewert type II 77 (50) 199 (57)Siewert type III 64 (42) 53 (15)pT stage 1 23 (15) 87 (24) 0.004pT stage 2 29 (19) 80 (23)pT stage 3 96 (63) 183 (52)pT stage 4 5 (3) 2 (1)pN stage 0 49 (32) 147 (42) 0.008pN stage 1 58 (38) 144 (41)pN stage 2 34 (22) 49 (14)pN stage 3 12 (8) 12 (3)AJCC stage I 33 (22) 123 (35) 0.001AJCC stage II 43 (21) 66 (19)AJCC stage III 70 (46) 150 (43)AJCC stage IV 17 (11) 13 (4)R1 14 (9) 35 (10) 0.782Positive proximal margin 7 (5) 7 (2)* 0.136Positive distal margin 1 (1) 4 (1)* 0.813Positive deep margin 6 (4) 26 (7) 0.167Proximal margin #3.8 cm 15 (10) 201 (57) "0.001!15 nodes removed 96 (63) 240 (68) 0.218Total 153 352

*Two patients had both proximal and distal margins involved.

Barbour et al Annals of Surgery • Volume 246, Number 1, July 2007


Withformalinfixationthefreshesophagushasbeenshowntocontractbyupto27%.Hence,thetrueinsitumarginlengthtreatedbysurgeryalonewouldbeatleast5cm


Conclusion


• TheEGJisaborderlineregion, locatedbetweenthethoraxandtheabdomen,inwhichphysicalforcesarecontinuouslyinactiontoregulateflowbyvaryingpressureandresistance.

• Itisaphysiologicconceptmorethanananatomicstructure:infact,itisanapparatus.

• Youcan’tfiguretheEGJwithoutathoroughassessmentofthehiatus=>it’sworkforYOU,notforthepathologist.

• Thehiatusisone(ifnotthemostimportant)engineofthelymphaticflowattheEGJ:thedirectionoftheflowfollowstheactionofthehiatus.


• ThetherapeuticvalueofextendedmediastinallymphadenectomyseemsmarginalforSiewert IIandIIIEGJC.Mediastinal nodespositivitymeansadvanceddisease.

• Thegreatdealoflymphadenectomyisintheabdomen,andthemostimportantphaseisintheareaundertheactionofthehiatus.

• Theneedtogointothethoraxwillbefortheadequatemargin,notforlymphadenectomy.

• Ifitisbelowthehiatusyoucansafelygothroughitsofar.• Ifitlooksthroughthehiatuswatchoutandmindtheneedforan

adequatemarginandtheoptionforagastrictubereconstruction.• Ifthehiatusisembraced,itisesophagus:youwon’tgetthe

adequatemargintranshiatally.©2015MarcoLotti– [email protected]

ThankYou


Date post:	15-Feb-2017
Category:	Health & Medicine
Upload:	marco-lotti
View:	59 times
Download:	0 times

Lotti Marco MD - Cancer of the Oesophago-Gastric Junction

Health & Medicine