Glutathione S-Transferase P1??Ile105Val Polymorphism is Associated??with Haematological Toxicity in...

transcript

Approval for publication Signed Date Number of amended pages returned

Drugs Aging 2008; 25 (6): 1ORIGINAL RESEARCH ARTICLE 1170-229X/08/0006-0001/$48.00/0

Glutathione S-Transferase P1Ile105Val Polymorphism isAssociated with HaematologicalToxicity in Elderly Rectal CancerPatients ReceivingPreoperative ChemoradiotherapyMarco Agostini,1 Lara Maria Pasetto,2 Salvatore Pucciarelli,1 Salvatore Terrazzino,3,4

Alessandro Ambrosi,1,5 Chiara Bedin,1 Francesca Galdi,1 Maria Luisa Friso,6Claudia Mescoli,7 Emanuele Urso,1 Alberta Leon,3 Mario Lise1,8 and Donato Nitti1

1 Clinica Chirurgica II, Department of Oncological and Surgical Sciences, University of Padova,Padova, Italy

2 Oncologia Medica 2, Istituto Oncologico Veneto IRCCS, Padova, Italy3 Research & Innovation Laboratories, Padova, Italy4 Studio Nutrizione & Benessere, Como, Italy5 Istituto Oncologico Veneto IRCCS, Padova, Italy6 Radioterapia, Istituto Oncologico Veneto IRCCS, Padova, Italy7 Anatomia Patologica, Istituto Oncologico Veneto IRCCS, Padova, Italy8 Chirurgia Oncologica I, Centro di Riferimento Oncologico, IRCCS, Aviano, Italy

Background: Increasing evidence suggests that common gene polymorphismsAbstractmay influence the toxicity of various cytotoxic agents used in the treatment ofcancer.Objective: To evaluate the predictive value of acute toxicity of methylenetetrahy-drofolate reductase 677T polymorphism, glutathione S-transferase P1 (GSTP1)substitution of isoleucine with valine at codon 105 (Ile105Val) polymorphism andthe tandem repeat polymorphism in the thymidylate synthase gene promoter inelderly patients with rectal cancer receiving preoperative chemoradiotherapy(CRT).Method: From 1994 to 2002, 166 Caucasian patients underwent surgery follow-ing CRT for mid-low rectal cancer at a single institution, 42 (male-to-female ratio,25 : 17) of whom were aged ≥65 years (median age 70 years, range 65–79). Thepre-treatment clinical stage was tumour (T) stage 3–4 in 38 patients and node (N)-positive in 29 patients. Patients received external-beam radiotherapy with conven-tional fractionation and fluorouracil-based chemotherapy. Blood samples wereused to extract and amplify DNA. Gene polymorphisms were determined bypolymerase chain reaction and restriction enzyme digestion. Acute toxicity topreoperative therapy was reported according to the National Cancer Institute

2 Agostini et al.

Common Toxicity Criteria, version 2. Univariate and multivariate analyses wereperformed using one-way analysis of variance and linear regression, respectively.Results: Haematological toxicity (grade 1–2) was observed in 15 of 40 patientsfor whom toxicity data were available and gastrointestinal toxicity (grade 1–4) in24 of these same 40 patients. At univariate analysis, female sex (p = 0.036) andGSTP1 Ile105Val (p = 0.0376) were associated with haematological toxicity. Atmultivariate analysis, GSTP1 Ile105Val polymorphism (p = 0.041) was the onlyfactor found to be associated with haematological toxicity. Patients carrying theVal/Val genotype in the GSTP1 gene had a lower risk of haematological toxicity(odds ratio = 0.322, 95% CI 0.101, 0.957) than patients with the Ile/Ile genotype.Conclusion: GSTP1 Ile105Val polymorphism is a promising marker of potentialhaematological toxicity in elderly patients with rectal cancer receiving preopera-tive CRT.

Introduction fluorouracil, the new agent oxaliplatin, a DNAcross-linking agent[10] active in metastatic disease,appears to be an excellent candidate for neoadjuvantThe incidence of rectal cancer is approximately‘downsizing’ of rectal cancer because of its rapid217 000 cases per year in Europe and 35 000 percytoreductive capacity without increased acute tox-year in Italy.[1] For patients aged ≥85 years, colorec-icity.tal cancer represents one-third of all neoplasms, and

Current anti-cancer drug treatments are effective70% of patients with colorectal cancer are aged ≥65in only a minority of patients and it is not yetyears.[2] Since patients with locally advanced dis-possible to reliably predict which patient is likely toease (transmural and/or positive nodes) have a highbenefit from a specific chemotherapeutic drug treat-risk of local relapse, the current standard approachment or which patient will experience life-threaten-to treatment is a combination of chemoradiotherapying drug toxicity. Thus, it is clinically relevant to(CRT) followed by radical surgery. Since 1994, indefine predictors of toxicity, especially in elderlyour institution, fit patients with locally advancedpatients, in order to identify individuals predisposedrectal cancer have been treated with preoperativeto a high risk of toxicity from standard doses of anti-CRT (external-beam radiotherapy [usually 45 Gy incancer drugs. Less than 20% of elderly patients are25 fractions, 1.8 Gy/day]; and fluorouracil-basedincluded in clinical trials.[11] Although old age ischemotherapy).[3] Using this approach, no local re-often associated with increased haematological tox-lapse was found and 16% of pathological completeicity, age alone is not a sufficient reason to excluderesponses were achieved. More recently, higherelderly patients from neoadjuvant treatment or todoses of radiotherapy (50.4 Gy) and preoperativereduce the dose of a cytotoxic drug.[12,13]fluorouracil administration by continuous infusion

have been used.[4,5] This treatment is associated with Several studies suggest that gene polymorphismsimilar global or acute toxicities independently of may influence the toxicity of different cytotoxicage (2% vs 5% incidence of grade 3–4 nausea and/or agents.[14-16] Among these, the tandem repeat (TR)vomiting with bolus vs continuous infusion, respec- polymorphism[17] in the thymidylate synthasetively; 15% vs 1% incidence of grade 3–4 stomatitis, (TYMS) gene, methylenetetrahydrofolate reductaserespectively; 15% vs 5% incidence of grade 3–4 (MTHFR) C677T polymorphism[18] and glutathionediarrhoea, respectively; and 8% vs 4% incidence of S-transferase P1 (GSTP1) substitution of isoleucinegrade 3–4 neutropenia, respectively);[6,7] only hand- with valine at codon 105 (Ile105Val) polymor-foot syndrome was more common in patients receiv- phism[19] seem to affect the therapeutic efficacy and/ing continuous infusion.[8,9] In combination with or toxicity of the major drugs currently used. This is

Polymorphism of Haematotoxicity of Chemotherapy for Rectal Cancer 3

particularly the case for enzymes for which identifi- and storage of blood samples in the departmentalcation of polymorphic genetic variations is playing a biological tissue bank.significant role in the tailoring of chemotherapy. TR

Treatmentpolymorphism (i.e. two [2R] or three [3R] TRs of a28 base-pair sequence) in the 5′-flanking untranslat-

Radiotherapyed region of the TYMS gene promoter and the C-to-TRadiotherapy was delivered with a linear acceler-substitution at nucleotide 677 (C677T) in the

ator using 6-mV photons and a three- or four-fieldMTHFR gene influence the efficacy of fluorouracil-box technique with the patient in the prone position.based chemotherapies.[17] Moreover, glutathioneThe deliverable total dose was 45 Gy/25 fractions intransferase enzymes are involved in the detoxifica-5 weeks to the posterior pelvis until 2003, then 50.4tion of the products of UV radiation-induced oxida-Gy/28 fractions (45 Gy/25 fractions in 5 weeks totive stress.[20]

the posterior pelvis followed by 5.4 Gy/3 fractionsThe aim of this study was to assess the role of TRboost to the tumour bed), as specified by the Interna-polymorphism in the TYMS gene, MTHFR C677Ttional Commission on Radiation Units and Mea-polymorphism and GSTP1 Ile105Val polymorphismsurements 50 report, with daily fractions of 1.8 Gyas predictive markers of acute toxicity in elderlyon 5 consecutive days per week.patients affected by locally advanced rectal cancer

and receiving preoperative CRT. ChemotherapyFluorouracil was administered concomitantly

Methods with radiotherapy by continuous infusion at a fixeddose of 225 mg/m2 daily ± oxaliplatin 60 mg/m2

weekly or by bolus 450 mg/m2 weekly, for approxi-Eligibility Criteria

mately 5.5 weeks, from the first to the last day ofradiotherapy. In some cases, capecitabine was sub-From January 1994 to June 2002, 166 Caucasianstituted for fluorouracil ± oxaliplatin 50 mg/m2patients with a primary adenocarcinoma of the mid-weekly and administered at a dosage of 825 mg/m2low rectum underwent preoperative CRT followedtwice daily on Monday to Friday of weeks 1–5 (totalby surgery at the Clinica Chirurgica II, University ofof 25 days dosing).Padova, Padova, Italy. For the purpose of this study

the following inclusion criteria were used: Toxicity

• age ≥65 years; Haematological and non-haematological toxicitywere monitored weekly during treatment and 7–10• histologically proven adenocarcinoma, tumourdays after completion of CRT, according to thelocation up to 11 cm from the anal verge asNational Cancer Institute Common Toxicity Crite-assessed by rigid proctoscopy;ria, version 2.[21] Patients were divided in two• clinical stage II–III (tumour [T] stage 3–4 and/orgroups according to toxicity grade (0 or ≥1).lymph node [N] involvement) as assessed by

transrectal ultrasonography and/or pelvic CTGenotypingscan and an abdominal and thoracic CT scan;

• no previous neoadjuvant or adjuvant chemother- Blood samples were obtained for DNA isolationapy or radiotherapy and Eastern Cooperative On- and subsequent genotype determination.TYMS,[22,23]

cology Group performance status ≤1. MTHFR[24] and GSTP1[25] polymorphisms were de-Of 166 patients considered, 42 were aged ≥65 termined by isolation of constitutional DNA from

years. peripheral blood samples of each patient, withAll procedures were reviewed and approved by a QIAamp® DNA Mini Kit (Qiagen GmbH, Hilden,

local ethics committee (protocol number 740 P), and Germany) and with subsequent polymerase chainall patients provided informed consent for the use reaction (PCR) amplification of target regions, ac-

4 Agostini et al.

cording to standardized protocols and literature data. In the multivariate analysis a logistic regressionTarget regions were then analysed with electrophor- model was considered and the most significant sub-esis. set of variables selected using a stepwise procedure.

A p-value of ≤0.05 was considered statistically sig-Homozygotes for the triple repeat variant in thenificant. Statistical analysis was performed usingTYMS promoter (TR 3/3) had a PCR product of 248SPSS software (SPSS Inc., Chicago, IL, USA).base pairs (bp), homozygotes for the double repeat

variant (TR 2/2) had a product of 220 bp, whileResultsheterozygotes (TR 2/3) had both 220- and 248-bp

products.The GSTP1 PCR product (176 bp in size) was Patients, Tumour and

digested overnight at 55°C with 2.5 U of BsmAI, Treatment Characteristicsand the fragments were separated on 3% agarose gel

From 166 Caucasian patients who underwent sur-and visualized after staining with ethidium bromide.gery following CRT for mid-low rectal cancer in ourHomozygotes for the Ile variant (Ile/Ile) had aninstitution, 42 fulfilled the inclusion criteria. Theseintact fragment of 176 bp, homozygotes for the Valpatients had a median age of 70 years (range 65–79),variant (Val/Val) underwent complete digestion of25 of whom were men and 17 were women.the PCR product resulting in fragments measuringTumours were located in the middle rectum in 1591 and 85 bp, while heterozygotes (Ile/Val) hadpatients and in the lower rectum in 27 patients.fragments measuring 176, 91 and 85 bp.Tumours were clinically staged as T3–4 in 38 (91%)The MTHFR PCR product (measuring 198 bp)patients, and N-positive in 29 (70%) patients (tablewas digested overnight at 37°C with 2.5 U of HinfI,I). Fluorouracil was delivered by continuous infu-and the fragments were separated on 3% agarose gelsion in 28 patients, eight of whom also receivedand visualized after staining with ethidium bromide.oxaliplatin, or by bolus in 13 patients (31%), four ofWild-type MTHFR (677CC) was characterized bywhom also received carboplatin. All patients under-an intact 198-bp fragment, heterozygotes (677CT)went surgery and in 34 cases (81%) the procedureby 198-, 175- and 23-bp fragments and homozy-was radical (R0). Details of treatment modalitiesgotes for the T variant (677TT) by 175 and 23 bp-and surgical procedures performed are summarizedsized fragments.in table II.For the statistical analysis, a recessive model was

assumed in which an individual must have twoToxicitycopies of the variant allele to have a different meta-

bolic pathway. Toxicity data were available for only 40 patients.Of these, 11 (27.5%) patients reported grade 3–4toxicities. Grade 3 gastrointestinal or skin toxicityDefinitions and Endpointswas observed in seven (17.5%) and two (5%) pa-tients, respectively. Grade 4 gastrointestinal toxicity

The primary objective of the study was to assesswas observed in two (5%) patients. No grade 3–4

the impact of the selected gene polymorphisms onhaematological, neurological or urological toxicity

haematological and non-haematological toxicity inwas found. Toxicity data are summarized in table

patients aged ≥65 years.III.

The dependencies between the considered vari-ables (age, sex, polymorphisms, type of treatment) Distribution of Genotypesand haematological and non-haematological toxicitywere evaluated by means of Fisher’s exact test and TR variants in the TYMS gene promoter togetherp-values adjusted for multiplicity with Holm’s with MTHFR (C677T) and GSTP1 (Ile105val)method. polymorphisms were determined in all 42 patients

40%) and Val/Val (n = 7, 17%). The observed allelefrequencies were similar to those previously report-ed in patients with rectal cancer.[19,21-27]

Association between Gene Polymorphismsand Acute Toxicity

None of the factors evaluated was associatedwith non-haematological toxicity.

Haematological toxicity was found to be asso-ciated with sex (p = 0.036; females more prone totoxicity) and GSTP1 Ile105Val polymorphism (p =0.037; p-values adjusted for multiplicity withHolm’s method = 0.1128). However, in multivariateanalysis, only GSTP1 Ile105Val polymorphism (p =0.041) [table V] was found to be associated withhaematological toxicity. The risk of havinghaematological toxicity was lower in patients carry-ing the Val/Val genotype in the GSTP1 gene than inthose with the Ile allele variant (odds ratio = 0.32,95% CI 0.101, 0.957).

Table I. Patient and tumour characteristics

Characteristics n %

Age (y)

≥65–70 24 57

>70–75 9 21.5

>75–80 9 21.5

Male 25 60

Female 17 40

Tumour location (from the anal verge) [cm]

≤7 27 64

>7 15 36

Clinical tumour (T) stage

T2 2 4.5

T3 28 67

T4 10 24

Not available 2 4.5

Clinical node (N) stage

N0 12 29

N1 29 69

Not available 1 2

In order to eliminate the potential confounding(table IV). The distribution of the genotypes was as effects of heterogeneous treatments, an exploratoryfollows: (i) TYMS polymorphism: TR 2/2 (n = 9, analysis was also performed in patients treated with22%), TR 2/3 (n = 18, 43%), TR 3/3 (n = 14, 33%), fluorouracil alone. However, this reduced the studynot determined 1 (2%); (ii) MTHFR polymorphism: sample size and consequently statistical power, such677CC (n = 14, 33%), 677CT (n = 19, 45%) and that none of the factors evaluated was associated677TT (n = 9, 22%); and (iii) GSTP1 Ile105Val with haematological toxicity (TYMS, p = 0.212;polymorphism: Ile/Ile (n = 18, 43%), Ile/Val (n = 17, GSTP1, p = 0.376; MTHFR, p = 0711).

Table II. Treatment modality

Modality Variable n %

Total radiotherapy dose delivered (median [range] Gy) 49.1 [45–58]

Fluorouracil mode of administration Bolus 13 31

Continuous infusion 28 67

Not available 1 2

Drugs Fluorouracil alone 30 71

Fluorouracil + oxaliplatin 8 19

Fluorouracil + carboplatin 4 10

Surgical procedure Abdominoperineal resection 11 26

Local excision 2 5

Hartmann’s 3 7

Low anterior resection or colo-anal anastomosis 26 62

Radical (R0) surgery Yes 34 81

No 6 14

Undetermineda 2 5

a Patients who underwent local excision.

6 Agostini et al.

that was difficult to evaluate statistically and none ofthe factors considered was found to be associatedwith haematological toxicity. Thus, keeping in mindthe preliminary approach evaluated in this study, weplan to confirm our results by increasing the numberof cases homogeneous for a chemotherapy regimen.It should also be noted, however, that since allpatients underwent radiation treatment and fluoro-

Table III. Acute toxicitya

Toxicity Grade

0 1 2 3 4

Haematological 25 10 5 0 0

Gastrointestinal 16 12 3 7 2

Skin 25 7 6 2 0

Urological 36 4 0 0 0

Neurological 37 3 0 0 0

a In two patients toxicity data were not available.uracil administration, it is not possible to discrimi-nate as to whether the toxicity reported in this studyDiscussionwas associated with radiation or chemotherapy.

The detoxifying action of glutathione transferaseThe elderly are the largest group of oncological

usually protects cellular macromolecules againstpatients but are often under-treated.[11] In reality, age

damage from carcinogenic and cytotoxic agents.[28]

alone is not a sufficient reason to withhold adjuvant,GSTP1 polymorphism is widely expressed in human

neoadjuvant or palliative treatment or to reduce theepithelial tissues and over-expressed in some cancer

dose of a cytotoxic drug, although old age can betypes, including colon tumours.[29] Increased levels

more often associated with increased haematologi-of GSTP1 in tumours may contribute to chemother-

cal toxicity.[12,13] To better determine the predictiveapy resistance, since cyclophosphamide, vincristine,

value of acute toxicity induced by gene polymorph-cisplatin and etoposide are its substrates.[30-32] Pre-

isms in elderly patients with rectal cancer receivingvious studies revealed that substitution of Ile105Val

preoperative CRT, a number of polymorphismsreduces enzyme activity.[33,34] Although no other

were investigated in the current study. To date,data are available on the association between GSTP1

despite current evidence suggesting that drug-relat-Ile105Val polymorphism and toxicity, a prognostic

ed toxicity may be influenced by common genevalue has been reported in cancer patients. Individu-

polymorphisms,[14-16] the role of polymorphic vari-als with two Val alleles have shown a significant

ants as a predictive factor of toxicity in elderlysurvival benefit after combined oxaliplatin/fluoro-

patients receiving CRT has not been extensivelyuracil treatment with a median survival of 24.9

investigated.months compared with only 7.9 months for patients

In the current study, we retrospectively evaluated with metastatic colorectal cancer with two wild-typethe association between toxic adverse reactions andgermline polymorphic variants in the GSTP1,MTHFR and TYMS genes in Caucasian patients aged≥65 years. In the multivariate analysis, only GSTP1Ile105Val polymorphism was found to be associatedwith haematological toxicity. Patients carrying theGSTP1 Ile allele variant reported a significantlyhigher risk of grade 1–2 haematological toxicitythan those with the Val/Val genotype.

Although all patients analysed were treated withfluorouracil, some (n = 12) also received oxaliplatinand carboplatin. Therefore, to eliminate the poten-tially confounding effects of such heterogeneoustreatments, we also performed an exploratory ana-lysis in patients treated with fluorouracil alone.However, this reduced the sample number to a size

Table IV. Thymidylate synthase (TYMS), methylenetetrahydrofo-late reductase (MHTFR) and glutathione S-transferase P1 (GSTP1)polymorphisms determined in the study

Polymorphism Variant n %

TYMS TR genotype, n = 42 2/2 9 22

2/3 18 43

3/3 14 33

Not determineda 1 2

GSTP1 genotype, n = 42 105Ile/105Ile 18 43105Ile/105Val 17 40105Val/105Val 7 17

MTHFR genotype, n = 42 C677C 14 33

C677T 19 45

T677T 9 22

a DNA not amplifiable.

Ile = isoleucine; TR = tandem repeat; Val = valine.

Table V. Univariate and multivariate analysis of variables predictive of haematological toxicity

Polymorphisms Univariate analysis Multivariate analysis

haematological toxicity OR 95% CI p-value

no yes p-value

TYMS TR genotype (n = 39)a,b 0.216 NS

2/2, 2/3 17 8 1

3/3 7 7 2.082 0.451, 9.948

GSTP1 genotypea 0.0376 0.041105Ile/105Ile 8 9 1105Ile/105Val 10 6 0.543 0.106, 2.612105Val/105Val 7 0 0 0.00, 0.890

MTHFR genotypea 0.711 NS

C677C 9 4 1

C677T 11 8 1.611 0.300, 9.866

T677T 5 3 1.330 0.136, 12.050

a In two patients toxicity data were not available.

b In one patient TYMS TR genotype was not available.

GSTP1 = glutathione S-transferase P1; Ile = isoleucine; MTHFR = methylenetetrahydrofolate reductase; NS = not significant; OR = oddsratio; TR = tandem repeat; TYMS = thymidylate synthase; Val = valine.

alleles.[19] A similar result was reported in a retro- be increased when the GSTP1 enzyme reduces itsactivity.spective study of 240 patients with breast cancer

treated with radiotherapy or various alkylating In an evaluation of TYMS promoter polymor-agents.[35] Individuals with the 105Val allele had phism in 50 patients with metastatic colorectal can-superior survival (hazard ratios for death were 0.8 cer undergoing fluorouracil-based chemotherapy,and 0.3 for Ile/Val and Val/Val genotypes, respec- Pullarkat et al.[26] found that the toxicity rate wastively) compared with individuals with the 105Ile significantly lower in patients with the TR 3/3 geno-allele. Evidently, the 105Val allele is an independent type than in TR 2/2 and 2/3 genotype carriers.

Differences between this study and our study in theprognostic factor in breast cancer because the survi-number, stage, site of tumour (colorectal vs rectal)val benefit was observed irrespective of the treat-and acute toxicity events could help explain thesement used.[35] Although direct involvement ofvariations. However, two other functional polymor-GSTP1 Ile105Val polymorphism in the detoxifica-phisms have been recently described in the TYMStion of cisplatin occurs by formation of cisplatin-gene: a glycine-to-cysteine substitution in theglutathione adducts,[30] no biochemical informationsecond repeat of the TR 3 allele, which also affectsis yet available on the possible impact of the detoxi-the level of TYMS expression in patients by abolish-fying effect of GSTP1 on oxaliplatin and fluoro-ing an upstream stimulatory factor-1 (USF1)-bind-

uracil pathways. On the other hand, GSTP1 is in-ing site,[36] and a 6-bp deletion located in the 3′un-

volved in the detoxification of reactive oxygen spe- translated region, 447 bp downstream from the stopcies,[20] which may act as intermediaries in the codon, which has been associated with decreasedcytotoxicity of many chemotherapeutic agents, in- response to fluorouracil chemotherapy.[37] There-cluding fluorouracil. Although larger clinical stud- fore, in addition to TR polymorphism, these twoies of the influence of GSTP1 Ile105Val polymor- other polymorphisms of the TYMS gene should alsophism on the toxic profile of fluorouracil-based be considered, both individually and as a haplotype,chemotherapy are still warranted, our results support by assessing the clinical value of TYMS polymorph-the hypothesis that toxic adverse reactions of fluoro- isms as a predictive marker of toxicity in response to

TYMS-targeting anticancer drugs.uracil, alone or in association with oxaliplatin, may

8 Agostini et al.

In contrast to other investigators,[18-26,28,30-38] but References1. EUCAN. Cancer incidence, mortality and prevalence in thein agreement with Cohen et al.,[27] we found no

European Union: 1998 estimates [online]. Available fromassociation between MTHFR C677T polymorphism URL: http://www-dep.iarc.fr/eucan/eucan.htm [Accessed

2008 Apr 1]and acute toxicity. In their series, Toffoli et al.[38]

2. Grobovsky L, Kaplon M, Krozser-Hamati A, et al. Features ofand Chiusolo et al.[18] evaluated toxicity in 43 pa-cancer in frail elderly pts (pts) (85 years of age) [abstract no.

tients with advanced ovarian carcinoma and in 61 2469]. Proc Am Soc Clin Oncol 2000; 19: 1393. Pucciarelli S, Friso ML, Toppan P, et al. Preoperative combinedpatients with acute leukaemia, respectively, and

radiotherapy and chemotherapy for middle and lower rectalfound a significant association between MTHFRcancer: preliminary results. Ann Surg Oncol 2000; 7 (1): 38-44

C677T polymorphism and acute treatment toxicity. 4. The Advanced Colorectal Meta-Analysis Project. Modulation offluorouracil with leucovorin in patients with advanced colorec-However, differences in the sample size and drugstal cancer: evidence in terms of response rate. J Clin Oncolused (in both studies, patients were treated with 1992; 10: 896-903

methotrexate therapy) makes comparison difficult. 5. O’Connell MJ, Martenson JA, Wieand HS, et al. Improvingadjuvant therapy for rectal cancer by combining protracted-Although a trend has been observed toward a lowerinfusion fluorouracil with radiation therapy after curative sur-toxicity rate in patients with the MTHFR C677C gery. New Engl J Med 1994; 331: 502-7

genotype than in those with the T allele, this associa- 6. Sobrero A, Guglielmi A, Cirillo M, et al. 5-fluorouracil modu-lated by leucovorin, methotrexate and mitomycin: highly ef-tion has not been definitively proven.fective, low-cost chemotherapy for advanced colorectal can-

Our study evaluated a relatively small number of cer. Br J Cancer 2001; 84 (8): 1023-87. Sargent DJ, Goldberg RM, Jacobson SD, et al. A pooled ana-elderly patients with cancer who were not homoge-

lysis of adjuvant chemotherapy for resected colon cancer inneous with respect to their chemotherapy regimens.elderly patients. N Engl J Med 2001; 345 (15): 1091-7

Our results may represent a preliminary approach 8. The Advanced Colorectal Cancer Meta-Analysis Project. Meta-analysis of randomized trials testing the biochemical modula-for toxicity studies of neoadjuvant treatments fortion of fluorouracil by methotrexate in metastatic colorectalrectal cancer, but it is absolutely necessary to con- cancer. J Clin Oncol 1994; 12: 960-9

firm these data in a larger, prospective case study 9. The Meta-Analysis Group in Cancer. Efficacy of intravenouscontinuous infusion of fluorouracil compared with bolus ad-that is homogeneous for chemotherapy regimens.ministration in advanced colorectal cancer. J Clin Oncol 1998;16: 301-8

10. Raymond E, Chaney SG, Taamma A, et al. Oxaliplatin: a reviewConclusion of preclinical and clinical studies. Ann Oncol 1998; 9: 1053-7111. Hessman O, Bergkvist L, Strom S. Colorectal cancer in patients

over 75 years of age: determinants of outcome. Eur J SurgGSTP1 Ile105val polymorphism in genomic Oncol 1997; 23 (1): 13-9DNA seems to be associated with haematological 12. Koperna T, Kisser M, Schulz F. Hepatic resection in the elderly.

World J Surg 1998; 22 (4): 406-12toxicity in elderly patients with rectal cancer receiv-13. Balducci L, Yates G. General guidelines for the management ofing preoperative fluorouracil-based CRT.[39] Pro- older patients with cancer. Oncology 2000; 14: 221-7

spective larger-scale trials confirming this observa- 14. Relling MV, Dervieux T. Pharmacogenetics and cancer therapy.Nat Rev Cancer 2001; 1: 99-108tion may allow for more individualized therapy,

15. Ulrich CM, Robien K, McLeod HL. Cancer pharmacogenetics:particularly in the neoadjuvant treatment of elderly polymorphisms, pathways and beyond. Nat Rev Cancer 2003;patients with cancer. 3: 912-20

16. Nagasubramanian R, Innocenti F, Ratain MJ, et al. Pharmaco-genetics in cancer treatment. Ann Rev Med 2003; 54: 437-52

17. Jakobsen A, Nielsen JN, Gyldenkerne N, et al. ThymidylateAcknowledgementssynthase and methylenetetrahydrofolate reductase gene poly-morphism in normal tissue as predictors of fluorouracil sensi-tivity. J Clin Oncol 2005; 23: 1365-9This study, conducted in Padova (Italy), was partially

18. Chiusolo P, Reddiconto G, Casorelli I, et al. Preponderance ofsupported by grants from the Italian Ministry of Health (Pro-methylenetetrahydrofolate reductase C677T homozygositygramma speciale ex. art.12, 2004), AIRC (Associazione Ital-among leukemia patients intolerant to methotrexate. Anniana Ricerca Cancro). The authors have no conflicts of inter-Oncol 2002; 13 (12): 1915-8

est that are directly relevant to the content of this study. The19. Stoehlmacher J, Park DJ, Zhang W, et al. Association between

authors are grateful to Miss Marta Briarava for collecting and glutathione S-transferase P1, T1, and M1 genetic polymor-managing clinical and research data, and to Antonette Leon phism and survival of patients with metastatic colorectal can-for her assistance with the English language. cer. J Natl Cancer Inst 2002; 94: 936-42

20. Edvardsen H, Kristensen VN, Grenaker Alnaes GI, et al. 31. O’Brien M, Kruh GD, Tew KD. The influence of coordinateGermline glutathione S-transferase variants in breast cancer: overexpression of glutathione phase II detoxification generelation to diagnosis and cutaneous long-term adverse effects products on drug resistance. J Pharmacol Exp Ther 2000; 294:after two fractionation patterns of radiotherapy. Int J Radiat 480-7Oncol Biol Phys 2007; 67 (4): 1163-71

32. Tsuchida S, Sato K. Glutathione transferases and cancer. Crit21. Trotti A, Byhardt R, Stetz J, et al. Common toxicity criteria:

Rev Biochem Mol Biol 1992; 27: 337-84version 2.0 – an improved reference for grading the acute33. Watson MA, Stewart RK, Smith GB, et al. Human glutathioneeffects of cancer treatment: impact on RT. Int J Radiat Oncol

Biol Phys 2000; 47: 13-47 S-transferase P1 polymorphisms: relationship to lung tissueenzyme activity and population frequency distribution. Carci-22. Van Triest B, Peters GJ. Thymidylate synthase: a target fornogenesis 1998; 19: 275-80combination therapy and determinant of chemotherapeutic re-

sponse in colorectal cancer. Oncology 1999; 57: 179-94 34. Srivastava SK, Singhal SS, Hu X, et al. Differential catalytic23. Agostini M, Pucciarelli S, Calandra P, et al. Genetic hetero- efficiency of allelic variants of human glutathione S-transfer-

geneity of variable number tandem repeats in thymidylate ase P1 in catalyzing the glutathione conjugation of thiotepa.synthase gene in colorectal cancer patients. Int J Biol Markers Arch Biochem Biophys 1999; 366: 89-942004; 19 (4): 332-6

35. Sweeney C, McClure GY, Fares MY, et al. Association between24. Terrazzino S, Agostini M, Pucciarelli S, et al. A haplotype of the survival after treatment for breast cancer and glutathione S-

methylenetetrahydrofolate reductase gene predicts poor tumortransferase P1 Ile105Val polymorphism. Cancer Res 2000; 60:response in rectal cancer patients receiving preoperative5621-4chemoradiation. Pharmacogenet Genomics 2006; 16 (11):

817-24 36. Mandola MV, Stoehlmacher J, Muller-Weeks S, et al. A novelsingle nucleotide polymorphism within the 50 tandem repeat25. Harries LW, Stubbins MJ, Forman D, et al. Identification ofpolymorphism of the thymidylate synthase gene abolishesgenetic polymorphisms at the glutathione S-transferase P1

locus and association with susceptibility to bladder, testicular USF-1 binding and alters transcriptional activity. Cancer Resand prostate cancer. Carcinogenesis 1997; 18: 641-4 2003; 63 (11): 2898-904

26. Pullarkat ST, Stoehlmacher J, Ghaderi V, et al. Thymidylate 37. Dotor E, Cuatrecases M, Martinez-Iniesta M. Tumor thymidy-synthase gene polymorphism determines response and toxicity late synthase 1494del6 genotype as a prognostic factor inof 5FU chemotherapy. Pharmacogenomics J 2001; 1 (1): 65-70 colorectal cancer patients receiving fluorouracil-based adju-

27. Cohen V, Panet-Raymond V, Sabbaghian N, et al. Methylenete- vant treatment. J Clin Oncol 2006; 24 (10): 1603-11trahydrofolate reductase polymorphism in advanced colorectal

38. Toffoli G, Russo A, Innocenti F, et al. Effect of methylenete-cancer: a novel genomic predictor of clinical response totrahydrofolate reductase 677C-->T polymorphism on toxicityfluoropyrimidine-based chemotherapy. Clin Cancer Res 2003;and homocysteine plasma level after chronic methotrexate9: 1611-5treatment of ovarian cancer patients. Int J Cancer 2003; 103:

28. Mannervik B, Alin P, Guthenberg C, et al. Identification of three294-9classes of cytosolic glutathione transferase common to several

mammalian species: correlation between structural data and 39. Pasetto LM, Friso ML, Pucciarelli S, et al. Rectal cancerenzymatic properties. Proc Natl Acad Sci U S A 1985; 82: neoadjuvant treatment in elderly patients. Anticancer Res7202-6 2006; 26: 3913-24

29. Moscow JA, Fairchild CR, Madden MJ, et al. Expression ofanionic glutathione-S-transferase and P-glycoprotein genes inhuman tissues and tumors. Cancer Res 1989; 49: 1422-8 Correspondence: Dr Salvatore Pucciarelli, Clinica Chirurgica

II, Department of Oncological and Surgical Sciences,30. Ban N, Takahashi Y, Takayama T, et al. Transfection of gluta-thione S-transferase (GST)-P1 antisense complementary DNA University of Padova, Via Giustiniani 2, 35128 Padova,increases the sensitivity of a colon cancer cell line to ad-

Italy.riamycin, cisplatin, melphalan, and etoposide. Cancer ResE-mail: puc@unipd.it1996; 56: 3577-82

Glutathione S-Transferase P1??Ile105Val Polymorphism is Associated??with Haematological Toxicity in...

Documents