Cancer and Leukemia Group B Gastrointestinal Cancer Committee Richard M

Cancer and Leukemia Group B Gastrointestinal Cancer Committee Richard M. Goldberg,1Donna Niedzwiecki,2 Monica Bertagnolli,3 A.William Blackstock,4 Joel E.Tepper,1and Robert J. Mayer3

Abstract The Cancer and Leukemia Group B Gastrointestinal Cancer Committee was organized in the late 1970s under the leadership of Michael Perry and Philip Schein and began full-scale operations in the mid-1980s.Today, it is a multidisciplinary team of surgeons, radiation and medical oncologists, statisticians, quality-of-life experts, pharmacogeneticists, basic scientists, and community oncologists who design studies that are conducted across the United States.The Committee has done trials in patients with esophageal, gastric, pancreatic, colon, rectal, and anal cancers. New initia- tives are under way in hepatocellular cancer, cholangiocarcinoma, and neuroendocrine tumors originating in the gastrointestinal tract. The Committee has increasingly concentrated on translational studies using tumor blocks, germ-line DNA, and plasma to evaluate biological correlates of tumor response and clinical outcomes. A broad program of pharmacogenomics has been incor- porated for virtually all studies, including trials that prospectively use polymorphisms in drug- metabolizing genes to assign treatments. Future efforts aim to evolve new standards of care, evaluate new therapies, and answer relevant biological questions in gastrointestinal cancer.

The Committee’s History advances in gastrointestinalcancer through randomized trials conducted by cooperative oncology groups (4). Shortly When the Cancer and Leukemia Group B (CALGB) was thereafter, preliminary data were presented, suggesting the founded in 1956, treatment for gastrointestinalcancer was benefit of adding leucovorin and/or levamisole to 5-fluoroura- rudimentary and not a component of the Group’s research cil(5-FU) in the advanced disease and adjuvant settings for agenda. During the early 1970s, the National Cancer Institute patients with colorectal cancer (5, 6). In 1986, the CALGB funded the GastrointestinalTumor Study Group to develop Board of Directors discussed the potentialfor resuming clinicaltrialsin gastrointestinalcancers. Severalfounding investigative efforts in gastrointestinalcancer because of the institutionalmembers of GastrointestinalTumor Study Group success of other groups in accruing to large trials in (i.e., Georgetown University, McGill University, Mt. Sinai gastrointestinalcancer, the high frequency of gastrointestinal Hospital, and Roswell Park Cancer Institute) were also active cancers, and the interest expressed by CALGB surgeons and participants in the CALGB. During the late 1970s, under the radiation oncologists to advance the field through clinical trials. leadership of Michael Perry and Philip Schein, CALGB By that time, the GastrointestinalTumor Study Group had launched a series of phase II to III trials in gastrointestinal disbanded and the CALGB members who had been active in cancer to address the research interests of investigators that group lacked a platform for continued clinical investiga- managing these common diseases, complementing the efforts tion in gastrointestinalcancer treatments. of the GastrointestinalTumor Study Group. The nascent In 1987, the Group Chairman, EmilFrei III, conveyed to the Committee also afforded CALGB members the opportunity to CALGB Board of Directors the NationalCancer Institute’s desire enroll patients in the gastrointestinal studies of other cooper- to ensure adequate accrualfor selectedhigh-priority random- ative oncology groups. These efforts led to several publications ized phase III trials, among them several colorectal cancer (1–3). However, patient accruallaggedbehind projections, and adjuvant protocols. Robert J. Mayer represented CALGB in following a major reorganization of CALGB in 1980, research developing an adjuvant colon cancer study to assess the efficacy in gastrointestinalcancer was suspended. of 5-FU and leucovorin at an organizational meeting convened The 1986 publication in New England Journal of Medicine of by the NationalCancer Institute with participants from the the results of the Gastrointestinal Tumor Study Group rectal Eastern Cooperative Oncology Group, North Central Cancer adjuvant trialshowed the valueof postoperative chemo- Treatment Group, and Southwest Oncology Group. That radiation therapy and the potentialto realizemeaningful spring, the plenary presentation of the CALGB meeting was devoted to a review of ‘‘the status of adjuvant therapy for colorectal cancer.’’ The CALGB Board of Directors subsequently Authors’ Affiliations:1University of North Carolina at Chapel Hill, Chapel Hill, created a planning committee to develop clinical trials in North Carolina; 2Duke University, Durham, North Carolina; 3Dana-Farber Cancer gastrointestinalcancers. Dr. Mayer chaired this working group Institute, Boston, Massachusetts; and 4Wake Forest University, Winston-Salem, and was joined by Richard Schilsky, Glenn Steele, and Joel E. North Carolina Tepper. In 1988, CALGB activated Intergroup adjuvant trials in Requests for reprints: Richard M. Goldberg, University of North Carolina at Chapel Hill, CB 7305, 3009 Old Clinic Building, Chapel Hill, NC 27599. colon cancer (CALGB 8896 and Intergroup 0089) and rectal F 2006 American Association for Cancer Research. cancer (CALGB 8894 and North CentralCancer Treatment doi:10.1158/1078-0432.CCR-06-9004 Group 86-17-51).

www.aacrjournals.org 3589s Clin Cancer Res 2006;12(11Suppl) June 1,2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. The first meeting of the CALGB gastrointestinalworking Selected AdjuvantTrials in Colon Cancer group in 1988 was attended by 55 members. Thirty-two CALGB surgeons convened in New York in January 1989 CALGB 8896 was entitled ‘‘A prospectively randomized trial to stimulate surgical support for this initiative. By the time of low-dose leucovorin plus 5-FU, high-dose leucovorin plus of the second gastrointestinalworking group meeting that 5-FU, levamisole plus 5-FU, or low-dose leucovorin plus 5-FU Spring, 44 patients had been registered onto the adjuvant plus levamisole following curative resection in selected patients colorectal cancer trials. In recognition of the obvious Group with Duke’s B or C colon cancer.’’ This study, designed in the interest, the Board of Directors advanced the gastrointestinal late 1980s, was originally a comparison of two different dose working committee to full disease committee status, with schedules of adjuvant 5-FU and leucovorin compared with a Drs. Mayer and Tepper appointed chair and vice-chair, nontreatment controlarm. In September 1989, data showing respectively. Their tenure as committee leaders continued the superiority of 5-FU and levamisole to a nontreatment untillate 2004. Under their leadership, the Gastrointestinal controlarm ledto revision of this study to substitute 5-FU and Cancer Committee developed several phase II and III trials levamisole as the control arm (9). A fourth arm was also added, and participated in the design and accrualto many trialsin combining 5-FU, leucovorin, and levamisole. By 1992, a total which multiple oncology cooperative groups collaborated. of 3,561 eligible patients had been registered, 936 (26%) by The leadership of the Committee rotated in 2004 when CALGB members. Richard M. Goldberg (medical oncology), A. William Black- Analysis showed that the four treatments produced similar stock (radiation oncology), and Monica Bertagnolli (surgical disease-free and overall survival, but the toxicity patterns of the oncology) were appointed chair and cochairs, respectively. four treatment regimens varied greatly. In particular, the results Statisticians Donna Niedzwiecki and Donna Hollis as well as of this trialdirectlycomparing the 5-day ‘‘bolus’’ 5-FU/ other members of the statisticaland administrative team leucovorin program (‘‘Mayo regimen’’) with the weekly higher provided continuity through the leadership transition. dose 5-FU/leucovorin program (‘‘Roswell Park regimen’’) Throughout its existence, the Committee has been guided showed the 5-day bolus schedule to produce a significantly by the GastrointestinalCancer Committee members made up higher rate of neutropenia and stomatitis than the weekly of leading gastrointestinal cancer researchers from multiple program, whereas the latter had a slightly higher rate of severe disciplines affiliated with numerous prominent cancer diarrhea. The study also showed conclusively that the addition research centers across the country. The Committee’s studies of levamisole to 5-FU/leucovorin did not add benefit, leading pertain to cancers originating anywhere in the gastrointesti- to the replacement of levamisole by leucovorin as the preferred nal tract. Highlights from selected research studies are modulator of 5-FU. The final results of this study, with a included in this report. median follow-up time in excess of 10 years, were recently published (10). Surgical Trials Because outcomes were identicalamong the four arms of this study, all patients could be pooled permitting a series of Glenn Steele led an Intergroup study coordinated by CALGB secondary analyses. Three of these secondary analyses were led that examined sphincter sparing surgery in patients with distal by Jeffrey Meyerhardt from CALGB (11–13). Obese patients rectal adenocarcinomas (7). This study remains the only multi- who received adjuvant chemotherapy were found to have a institutionaleffort mounted to examine this management significant increase in overall mortality and a borderline approach to our knowledge. The 59 patients with T1 tumors significant increase in disease recurrence. Patients with diabetes underwent local excision with no further treatment and that mellitus who entered this trial were found to have a signi- strategy resulted in one local recurrence. The 51 patients with T2 ficantly higher rate of disease recurrence and mortality. Survival tumors received 5,400 cGy with bolus 5-FU after local excision. in this study was not associated with hospital volume for colon Of these patients, the 7 who recurred locally were all able cancer as had been documented previously for rectal primary to undergo abdominalperinealresection. The failure-free cancers. In an additional analysis led by Charles Fuchs of survivalwas 85% for the T 1 and 78% for the T2 patients at a CALGB, the clinical benefits in African American and Caucasian median of 6 years of follow-up. This trial provided evidence patients were found to be similar as opposed to results sug- that sphincter function could safely be preserved in the gested in other trials, and African Americans overall experi- majority of patients with T1 and T2 distalrectalcancers but enced less toxicity than did Caucasians (14). also showed a surprisingly high local recurrence rate in patients Two additional Intergroup colon adjuvant trials that were with T2 lesions, particularly in the presence of perineural initiated by CALGB have been thus far been reported in abstract involvement. form as their data continue to mature for article submission. CALGB surgeons also participated in the Intergroup trial of Protocol 9581, led by Thomas Colacchio, randomized 1,738 laparoscopic versus open colectomy led by Heidi Nelson from stage II patients after resection of their primary tumors to North CentralCancer Treatment Group. This study randomized monoclonal antibody 17-1A or to no postoperative therapy 872 patients and showed that cancer-related outcomes were no (15). There was little treatment-related toxicity but no different between these approaches (8). Laparoscopically discernable benefit associated with this treatment. CALGB resected patients had modestly shorter hospital stays, dimin- 89803, led by Leonard Saltz, randomized 1,264 stage III ished parenteral and oral narcotic requirements, and smaller patients to weekly bolus 5-FU with leucovorin alone or with incisions, but their surgeries took more operative time. There irinotecan (IFL; ref. 16). In 2004, after a median follow-up time was clear evidence that with experience operative times for of 2 years, the CALGB Data and Safety Monitoring Board laparoscopic surgery decreased to the point of mirroring those released the results due to futility. No difference in failure-free for open colectomy. survival( P = 0.88) or overallsurvival( P =0.92)was

Clin Cancer Res 2006;12(11Suppl) June 1, 2006 3590s www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. CALGB Gastrointestinal Cancer Committee discernable. A total of 18 patients randomized to the selected patients, it is uncertain if the results with HAI treatment irinotecan/5-FU/leucovorin regimen died due to treatment- are superior to those that can be achieved with contemporary associated toxicity in comparison with 6 patients in the 5-FU/ systemic chemotherapy programs. CALGB is currently partici- leucovorin cohort. The trial was important, as many oncologists pating in the NationalSurgicalAdjuvant Breast and Bowel had already begun to prescribe IFL as adjuvant chemotherapy Project C-09 study evaluating systemic capecitabine and for stage III colon cancer based on its superiority to 5-FU/ oxaliplatin with or without HAI in patients with resected liver leucovorin in the metastatic setting. The negative results of metastases, a research question that is at least in part an 89803, although surprising, spared many patients from the outgrowth of the research effort of Dr. Kemeny. excessive toxicity of the IFL regimen. Intergroup study N9741 (CALGB 89804) randomized patients with advanced colorectal cancer to IFL, oxaliplatin plus infused Adjuvant Therapy of Rectal Cancer 5-FU and leucovorin (FOLFOX), or irinotecan and oxaliplatin. Eventually, >1,700 patients enrolled through several trial CALGB 8894 was entitled ‘‘A phase III protocol for surgical modifications. As the study was accruing, CALGB Pharmacology adjuvant therapy of rectal cancer; a controlled evaluation of (a) and ExperimentalTherapeutics Committee Vice Chair Howard protracted infusion of 5-FU as a radiation enhancer and (b) McLeod proposed that it would be an ideal vehicle through 5-FU plus methyl-CCNU chemotherapy.’’ This study enrolled which to examine polymorphisms in candidate genes regulating 666 patients, 13% registered by CALGB, and showed the drug metabolism. Because cohorts of patients were exposed to superiority of infusionalcompared with bolus5-FU as a radia- two of three chemotherapy drugs, comparisons correlating both tion sensitizer, setting a standard of care for management of toxicity and response with common polymorphisms among rectalcancer in the postoperative setting that is current today. It groups treated or not exposed to each agent were possible. also established that the addition of the potentially leukemo- Collection of germ-line DNA was optional on this trial; 520 genic agent methyl-1-(2-chloroethyl)-3-cyclohexyl-L-nitrosourea patients donated blood for this purpose. This is the largest cohort to 5-FU added toxicity without augmenting efficacy (17). of advanced colorectal cancer patients enrolled in a trial to date Optimal chemotherapy for locally advanced rectal cancer for which both clinical data and germ-line DNA are available and was addressed in CALGB 9081/Intergroup 0114, which thus is among the largest populations currently available for compared bolus 5-FU (standard of care) with 5-FU plus correlation of genetic variables with cancer patient treatment leucovorin, levamisole, or both. All 1,792 registered patients outcomes. In addition, the study collected quality-of-life data received two cycles of chemotherapy before and after 5-FU- using the validated European Organization for Research and based chemoradiation. With a median follow-up of 7.4 years, Treatment of Cancer QLQ C30 questionnaire, customized there was no difference in overall survival or disease-free sur- questions to address neuropathy, and the Uniscale as a global vivalbetween the four treatment groups (18). Retrospective assessment periodically. subset analyses of these data resulted in several important The primary activity and toxicity data were reported on 795 observations. Thirty-four percent of patients with tumor recur- patients (23). For patients randomized to FOLFOX, the median rence or progression in a solitary site underwent resection time to progression was 8.7 months, response rate was 45%, and with curative intent (19). The 5-year survivalfor patients median survivaltime was 19.5 months. These were significantly undergoing surgery was 27% versus only 6% for those patients superior to results observed for the standard regimen of IFL for not resected (P < 0.001), underscoring the potentialvalueof all cancer outcomes (6.9 months, 31%, and 15.0 months) and aggressive surgicalresection for patients with limitedmeta- for irinotecan and oxaliplatin (6.5 months, 35%, and 17.4 static disease. Another analysis showed that hospital surgical months). The FOLFOX regimen had significantly lower rates of volume did not significantly affect cancer recurrence or sur- severe nausea, vomiting, diarrhea, febrile neutropenia, and dehy- vival, whereas sphincter-preserving surgery was more com- dration but a higher rate of sensory neuropathy and neutropenia. monly done at high-volume centers (20). Early in the trial, several arms delivered 5-FU as a bolus in The current Intergroup phase III trial is evaluating the role of combination with oxaliplatin or irinotecan, among them then adjuvant oxaliplatin/5-FU chemoradiation for the treatment of standard IFL regimen (24). Each of these regimens was associated locally advanced rectal cancer. The preclinical rationale and with significant toxicity, including early toxic deaths. These phase I/II trial(CALGB 89901) in support of this important findings were a catalyst for the widespread adoption in North study originated, in part, from CALGB investigators (21). America of 5-FU infusion-based therapies, the preferred delivery platform used in most of Europe. Advanced Colorectal Cancer Studies The study also contributed to the identification and subsequently sparked a better understanding of the severe Nancy Kemeny led CALGB 9481, a trial that randomized 135 toxicity, including early treatment-related deaths in a small patients with unresectable metastatic colorectal cancer limited minority of patients treated with irinotecan-based therapies. to the liver to i.v. 5-FU plus leucovorin or hepatic arterial These observations have since been explained, mainly through infusion (HAI) of fluorodeoxyuridine (22). The patients the work of CALGB-associated investigator teams, by poly- receiving HAI had better survival(24 versus 20 months; P = morphisms in the UGT1A1 gene that controls the metabolism 0.003), a higher response rate (47% versus 24%; P = 0.01), and of irinotecan (25). longer time to hepatic progression (10 versus 7 months; P = In N9741, as in other trials, surgery was possible in all arms 0.03). The patients assigned to receive HAI also had less of the study but was more often possible with the oxaliplatin- systemic toxicity and better physicalfunctioning. Although containing regimens. The patients who underwent resection CALGB 9481 unambiguously showed the value of HAI of metastases, 2% of those enrolled, enjoyed a prolonged chemotherapy compared with systemic 5-FU/leucovorin for median survivalof 44 months (26). The achievement of a

www.aacrjournals.org 3591s Clin Cancer Res 2006;12(11Suppl) June 1,2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. complete remission in the absence of surgery also led to To extend this observation further, CALGB protocol9865 comparably long remissions in the 4% of patients classified in collected tumor blocks from the CALGB patients enrolled on that category (27). Intergroup 0089 to determine whether tumors with MSI exhibited unique molecular characteristics that would further Esophageal and Gastric Cancer Trials distinguish them from tumors without this genetic pattern. Extracted DNA was analyzed to identify those tumors with MSI Until recently, the standard therapy for patients with locally as well as those showing the alternative pathway to colon advanced gastric/gastroesophagealcancer was surgicalresec- carcinogenesis of chromosomalinstability. The study found that tion, a strategy leading to 20% to 30% survival. Neither one third of sporadic colorectal cancers do not exhibit defining adjuvant chemotherapy alone nor radiation alone improved characteristics of either chromosomalinstability or MSI, indicat- this outcome. The Southwest Oncology Group initiated a trial ing that pathogenesis of colon cancer may not be fully defined by (Intergroup 0116/CALGB 9195), which randomized 556 either chromosomalinstability or MSI pathways (31). Additional patients, of which 98 were contributed by CALGB members studies of tumors with high levels of MSI showed that f70% to receive either surgery or surgery followed by 5-FU/leucovorin also had methylation leading to silencing or inactivation of the and radiation therapy. The median relapse-free survival was 19 hMLH1 promoter. This finding suggests that this epigenetic event months in the surgery-only group versus 30 months in the is a significant contributor to the MSI phenotype of sporadic chemoradiation group (P < 0.001; ref. 28). The median overall colorectalcancer (32). Additionalwork showed that the tumor survivalwas 27 months for the surgery-onlygroup versus 36 suppressor genes, PTEN, RUNX3, and APC, were frequently months for combination therapy (P = 0.005). There was also a inactivated by promoter hypermethylation in sporadic colorectal reduction in the rate of local failure in the combined modality cancer with high levels of MSI (33–35). arm when compared with the surgery only (19% versus 29%). This combination therapy is now standard and serves as the Predicting Imatinib Response in Gastrointestinal controlarm for CALGB 80101, an ongoing Intergroup phase III Stromal Tumors trial evaluating adjuvant epirubicin, cisplatin, and infusional 5- FU chemotherapy plus chemoradiation. Gastrointestinalstromaltumors (GIST) are rare sarcomas CALGB initiated protocol9781 in 1997 to definitively (f0.2% of allgastrointestinalmalignancies; ref. 36). In determine whether the administration of preoperative 5-FU/ 1998, Hirota et al. identified gain-of-function mutations of cisplatin and radiation therapy improved the outcome over the KIT proto-oncogene in the majority of GIST (37). This surgery alone in patients with esophageal cancer. The study enabled development of the first effective drug for GIST, the closed due to slow accrual after enrolling only 59 patients. small-molecule inhibitor imatinib mesylate (imatinib). In Nevertheless, a recent analysis indicates that patients treated 2002, it was reported that patients with metastatic GIST with trimodality therapy had a median survival of 4.5 years treated with imatinib had a 54% response rate, stable disease compared with 1.8 years (P = 0.02) for those treated with in 28%, and disease progression in 14% (38). Imatinib was surgery alone and 5-year overall survival of 39% versus 16% approved for treatment of metastatic or unresectable GIST in (29). Complication rates in the postoperative setting were not February 2001 (39). significantly different for the trimodality and surgery only Intergroup study S0033 evaluated imatinib dose response cohorts. (400 versus 800 mg) in metastatic or unresectable GIST. Correlative science studies for this trial coordinated by CALGB Correlative Science in Gastrointestinal (protocol 150105) characterized the molecular and cytogenet- Malignancies ic variables associated with KIT oncogene function to determine associations with response (40). The relationship In the early 1990s, CALGB began correlative science studies between mutations in KIT or PDGFRA tyrosine kinases and in gastrointestinalcancers. The goalwas to identify tumor- or drug response was examined in 324 patients with KIT-positive patient-specific characteristics that predicted disease outcome GIST. KIT mutations were present in 86% and PDGFRA or treatment toxicity. The challenge at that time was a lack of mutations in 1% for a mutation frequency of 87%. Patients tissue specimens linked to clinical outcomes, as this was before whose tumor expressed an exon 11 KIT mutant isoform were collection of tumor blocks and peripheral blood samples from more likely to have a clinical response to imatinib (67%) than study participants had become a routine practice. those with tumors expressing KIT exon 9 mutant isoforms (40%) or no kinase mutations (39%; P = 0.0022). There was Defining Subtypes a trend toward increased survivalfor patients with KIT exon 11 mutant GISTs compared with any other KIT mutation or The first CALGB gastrointestinaltissue resource linkedto an wild-type GIST, but this difference was not statistically ongoing clinical trial was developed from colorectal cancer significant. The investigators concluded that activating muta- patients treated on Intergroup protocol0089, the previously tions of KIT or PDGFRA are found in the vast majority of KIT- described study of patients with stage II or III colon cancer positive GISTs and that KIT exon 11 tumor genotype correlates treated with 5-FU with or without leucovorin/levamisole (11). with favorable response and time to progression. These results An assessment of a cohort of 431 patients enrolled in this study confirmed that GIST kinase genotype is prognostic of the showed that the presence of microsatellite instability (MSI) likelihood and duration of response to imatinib therapy. In predicted for a higher probability of disease-free survival at addition, this work indicated that new therapeutic strategies 5 years (64% versus 49%; P = 0.02) but no difference in overall are needed for patients with GISTs not expressing an exon 11 survivalat 5 years (68% versus 56%; P = 0.20; ref. 30). mutant KIT isoform.

Clin Cancer Res 2006;12(11Suppl) June 1, 2006 3592s www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. CALGB Gastrointestinal Cancer Committee

Ta b l e 1. CALGB Gastrointestinal Cancer Committee portfolio: currently active trials (treatment and correlative science), protocols in development (approved by Ethics Committee), and concepts in development (trials endorsed by CALGB through the Cancer Trials Support Unit)

Trial number Status Description Accrual current/target Gastric C80101 Active Adjuvant chemoradiation postresection 216/540 C150205 Active (C80101) Tumor:TS, ERCC-1,MSI, E-cadherin, epidermal growth factor 127/5 40 receptor, p27, cyclo-oxygenase-2, c-erbB-2; serum: insulin-growth factor-I, insulin-growth factor-II, insulin- growth factor ^ binding protein-3 Advanced pancreatic C80303 Active Gemcitabine + placebo vs gemcitabine + bevacizumab 533/590 C150405 Active (C80303) Serum: vascular endothelial growth factor and related 412/5 9 0 growth factors, markers of coagulation and endothelial cell activation; matrix-assisted laser desorption ionization- time of flight profiling S0205* Active Gemcitabine F cetuximab 702/704 C80401 In development Chemotherapy/radiotherapy + biological, nonmetastatic ö/90 CALGB Concept (letter of Phase IIgemcitabine, erlotinib and sunitinib; phase IIsunitinib ö/53; ö/64 investigation under review) GIST Z9001* Active Adjuvant STI571 (Gleevec) vs placebo 549/732 Colorectal C80405 Active Chemotherapy + bevacizumab vs chemotherapy + cetuximab vs 44/2,289 chemotherapy + combination; metastatic colon C150506 Active (C80405) Analysis of multiple potentially prognostic or predictive genetic and 25/2,289 molecular markers C80402 Active Phase IIsecond-line liposome-encapsulated SN38 (metabolite 0/50 of irinotecan) metastatic oxaliplatin failures E5202* Active Adjuvant FOLFOX F bevacizumab high-risk stage II 33/3,610 N0147* Active Adjuvant oxaliplatin + 5-FU F cetuximab 564/2,300 C-08* Active FOLFOX6 F bevacizumab resected stage II/III 1,650/2,632 R-04* Active Preoperative radiotherapy + capecitabine F oxaliplatin vs preoperative 250/1,606 radiotherapy + capecitabine + i.v. 5-FU F oxaliplatin; operable rectal cancer C-09* Active Capecitabine/oxaliplatin F HAI resected/ablated liver metastasis 0/400 C80501 In development Second-line AZD2171 + irinotecan; metastatic colon ö/130 C80404 In development Pharmacogenomics of high-dose irinotecan + cetuximab; metastatic colon ö/244 ACCENT Meta-analysis Adjuvant outcomes for Caucasian vs African ancestry 22,154 ACCENT Meta-analysis Adjuvant outcomes for North Americans, Canadians, and Europeans 22,154 Esophageal C80302 Active Multimodality for advanced 0/88 C80403 In development Combinations with cetuximab for advanced ö/216 Neuroendocrine C80602 Concept (currently under AMG706 F bevacizumab ö/136 review)

*Endorsed by CALGB through the CancerTrials Support Unit.

Developing Methods to Evaluate Micrometastatic to manifest tumor progression. The application of better Disease: Sentinel Lymph Node Studies in techniques to detect these high-risk cases would identify Colorectal Cancer patients who may benefit from adjuvant chemotherapy. The term, micrometastatic disease (MMD), broadly describes The resected lymph nodes in 35% to 45% of patients evidence of tumor metastases within regionallymphnodes that undergoing potentially curative surgery will be found to be are not scored ‘‘positive’’ by conventionalhistopathologic tumor free (41). However, 12% to 25% of these stage II patients criteria. Examples of MMD include small foci of tumor cells experience post-treatment recurrence, indicating that current visualized by H&E stain or individual tumor cells identified histopathologic staging methods fail to identify those destined only after application of immunohistochemistry for tumor

www.aacrjournals.org 3593s Clin Cancer Res 2006;12(11Suppl) June 1,2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. antigens, such as carcinoembryonic antigen or cytokeratins. For relative science studies initiated over the past 15 years are colon cancer, the clinical significance of regional MMD is just now yielding data. Prospective correlative science studies unknown. Retrospective studies employing a variety of disease were initiated for patients treated on stage II (CALGB 9581) definitions do not consistently show a relationship between and stage III (CALGB 89803) colorectal cancer trials. These presence of MMD and tumor recurrence. Prospective studies of databases are nearly mature, and the results of large, MMD are problematic due to the many subjects required, the adequately powered correlative science studies examining substantial time and expense involved in the methodologies to p53 expression and genotype, thymidylate synthase expres- identify micrometastases, and the lack of a clear definition of sion, microvesseldensity, p27 expression, and MSI willbe this entity. reported soon. CALGB 80001 aimed to determine whether sentinellymph Additionalcorrelative science studies in gastrointestinal node sampling could identify a subset of lymph nodes that tumors will examine predictive markers for response to predicted the status of the nodalbasin for resected coloncancer adjuvant treatment of gastric cancer (CALGB 150205), the and therefore could be extensively evaluated for the presence of correlation of angiogenesis and inflammation markers with micrometastases. Patients enrolled underwent sentinel lymph outcome in patients with advanced pancreatic cancer (CALGB node sampling following injection of 1% isosulfan blue, and 150405), and a broad panelof tumor- and serum-based both sentineland nonsentinelnodes obtained during primary markers of response to treatment with chemotherapy plus tumor resection were sectioned at multiple levels and stained anti–vascular endothelial growth factor and anti–epidermal using anti–carcinoembryonic antigen and anti-cytokeratin growth factor receptor therapy in advanced colorectal cancer antibodies. Using standard histopathology, sentinel nodes (CALGB 150506). failed to predict the presence of nodal disease in 13 of 24 (54%) of node-positive cases (42). Immunostains were done Summary for cases with negative nodes by standard histopathology. Depending on the immunohistochemicalcriteria used to assign At the 50th anniversary of CALGB, the Gastrointestinal lymph node positivity, sentinel node exam resulted in either an Cancer Committee can celebrate the development of and unacceptably high false-positive rate (20%) or a low sensitivity participation in several important clinical trials that have for detection of MMD (40%; ref. 43). By examining both changed clinical practice. Table 1 details the treatment studies sentineland nonsentinelnodes, this multi-institutionalstudy that are currently active or in development through the CALGB showed that sentinel nodes from resectable colon cancers did GastrointestinalCancer Committee. Over the past 15 years, we not accurately predict the presence of either conventionally and others have embarked on research programs aimed at the defined nodal metastases or MMD. As a result, sentinel lymph molecular classification of gastrointestinal tumors. Additional- node sampling is not a useful technique for the study of MMD ly, a focus on individualized medicine through pharmacoge- in patients with colon cancer. nomics has become integralto the Committee’s research agenda. As these treatment and translational studies mature Correlative Science Works in Progress in the next few years, we hope to improve standards of care and to set the stage for a time when clinicians will be able to make Because of the time required for building prospective treatment decisions based on tumor- and patient-specific outcome-linked tissue resources, most gastrointestinal cor- molecular staging.

References 1. NicholsWC, Moertel CG, RubinJ, et al. Phase IItrial of and open colectomy for colon cancer. N Engl J Med of the colon (CALGB 9581). Annu ASCO Proc 2004; indicine N-oxide (INDI) in patients with advanced co- 2004;350:2050^9. 22:250s. lorectal carcinoma. CancerTreat Rep 1981;65:337^ 9. 9. Moertel CG, FlemingTR, Macdonald JS, et al. Fluoro- 16. Saltz LB, Niedzwiecki D, Hollis D, et al. Irinotecan 2. Cullinan SA, Moertel CG, Fleming TR, et al. A com- uracil plus levamisole as effective adjuvant therapy af- plus fluorouracil/leucovorin alone (FL) in stage III co- parison of three chemotherapy regimens in the treat- ter resection of stage III colon carcinoma: a final lon cancer (Intergroup trial CALGB C89803). Ann ment of advanced pancreatic and gastric cancer. report. Ann Intern Med 1995;122:321 ^ 6. ASCO Proc 2004;22:245s. JAMA 1985;253:2061 ^ 6. 10. Haller DJ, Catalano PG, McdonaldJS, et al. Phase III 17. O’Connell MJ, Martenson JA, Wieand HS, et al. Im- 3. Ajani JA, Faintuch JS, McClure RK, et al. Phase II study of fluorouracil, leucovorin, and levamisole in proving adjuvant therapy for rectal cancer by combin- study of spirogermanium in patients with advanced high-risk stage II and III colon cancer: final report of ing protracted-infusion fluorouracil with radiation colorectal cancer. IND1986;4:383 ^ 5. Intergroup 0089. J Clin Oncol 2005;23:8671 ^ 8. therapy after curative surgery. N Engl J Med 1994; 4. Douglass HO, Jr., Moertel CG, Mayer RJ. Survival af- 11. MeyerhardtJA, Catalano PJ, Schrag D, et al. Associ- 331:502^7. ter postoperative combination treatment of rectal can- ation of hospital procedure volume and outcomes in 18. Tepper JE, O’Connell M, Niedzwiecki D, et al. Adju- cer. N Engl J Med 1986;315:1294^ 5. patients with colon cancer at high risk for recurrence. vant therapy in rectal cancer: analysis of stage, sex 5. Poon MA, O’Connell MJ, Moertel CG, et al. Bio- Ann Intern Med 2003;139:649 ^57. and local controlöfinal report of Intergroup 0114/ chemical modulation of fluorouracil: evidence of sig- 12. Meyerhardt JA, Catalano PJ, Haller DG, et al. CALGB 9081. J Clin Oncol 2002;20:1744^ 50. nificant improvement of survival and quality of life in Influence of body mass index on outcomes and 19. Tepper JE, O’Connell MJ, Hollis D, et al. Analysis of patients with advanced colorectal carcinoma. J Clin treatment-related toxicity in patients with colon salvage surgery after failure of primary therapy in rec- Oncol 1989;7:1407 ^ 18. carcinoma. Cancer 2003;98:484^95. tal cancer: results from Intergroup Study 0114. J Clin 6. Laurie JA, Moertel CG, FlemingTR, et al. Surgical ad- 13. Meyerhardt JA, Catalano PJ, Haller DG, et al. Impact Oncol 2003;21:3623 ^ 8. juvant therapy of large-bowel carcinoma: an evalua- of diabetes mellitus on outcomes in patients with co- 20. Meyerhardt JA, Tepper JE, Niedzwiecki D, et al. tion of levamisole and the combination of levamisole lon cancer. J Clin Oncol 2003;21:433 ^ 40. Impact of hospital procedure volume on surgical ope- and fluorouracil. The North Central CancerTreatment 14. McCollum AD, Catalano PJ, Haller DG, et al. Out- ration and long-term outcomes in high-risk curatively- Group and the Mayo Clinic. J Clin Oncol 1989;7: comes and toxicity in African-American and Cauca- resected rectal cancer. J Clin Oncol 2004;22: 1447 ^ 5 6. sian patients in a randomized adjuvant chemotherapy 166 ^ 74. 7. Steele GD, Jr., HerndonJE, Bleday R, et al. Sphincter- trial for colon cancer. J Natl Cancer Inst 2002;94: 21. Ryan DP, Niedzwiecki D, Hollis D, et al. Phase I trial sparing treatment for distal rectal adenocarcinoma. 116 0 ^ 7. of oxaliplatin, 5-FU, radiation therapy as initial therapy Ann Surg 1999;6:433^41. 15. Colacchio TA, Niedzwiecki D, Compton C, et al. for localized T3-4 rectal cancer. J Clin Oncol. In press 8. The Clinical Outcomes of Surgical Therapy Study Phase III trial of adjuvant immunotherapy with MOAB 2006. Group. A comparison of laparoscopically assisted 17-1 j following resection of stage II adenocarcinoma 22. Kemeny N, Niedzwiecki D, Hollis D, et al. Hepatic

Clin Cancer Res 2006;12(11Suppl) June 1, 2006 3594s www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. CALGB Gastrointestinal Cancer Committee

arterial infusion (HAI) versus systemic therapy for esophageal junction: a phase III Intergroup study 36. Fletcher CD, Berman JJ, Corless C, et al. Diagnosis hepatic metastases from colorectal cancer: a random- 0116/CALGB 9195/SWOG 9008. N Engl J Med of gastrointestinal stromal tumors: a consensus ap- ized trial of efficacy, quality of life (QOL), and molecu- 2001;345:725 ^ 30. proach. Hum Pathol 2002;33:459 ^ 65. lar markers. J Clin Oncol 2006;24:1395 ^ 403. 29. Krasna M,TepperJE, Niedzwiecki D, et al. Trimodal- 37. Hirota S, Isozaki K, MoriyamaY, et al. Gain-of-func- 23. Goldberg RM, Sargent DJ, Morton RF, et al. A ran- ity therapy is superior to surgery alone in esophageal tion mutations of c-kit in human gastrointestinal stro- domized controlled trial of fluorouracil plus leucovorin, cancer: results of CALGB 9781 [abstract 4].Proceed- mal tumors. Science 1998;279:577^ 80. irinotecan, and oxaliplatin combinations in patients ings of the 2006 Gastrointestinal Cancers Sympo- 38. Demetri GD, von Mehren M, Blanke CD, et al. Effi- with previously untreated metastatic colorectal can- sium; 2006 Jan 26-28; San Francisco, CA. cacy and safety of imatinib mesylate in advanced gas- cer. J Clin Oncol 2004;22:23 ^ 30. 30. Watanabe T, Wu T-T, Catalano PJ, et al. Molecular trointestinal stromal tumors. N Engl J Med 2002;347: 24. Delaunoit T, Goldberg RM, Sargent DJ, et al. Mortal- predictors of survival after adjuvant chemotherapy for 472 ^ 80. ity of 5-fluourouracil and leucovorin bolus regimens colon cancer. N Engl J Med 2001;344:1196 ^ 206. 39. Dagher R, Cohen M, Williams G, et al. Approval combined with either irinotecan and oxaliplatin: results 31. Goel A, Arnold CN, Niedzwiecki D, et al. Character- summary: imatinib mesylate in the treatment of from Intergroup trial N9741.Cancer 2004;101:2170^6. ization of sporadic colon cancer by patterns of ge- metastatic and/or unresectable malignant gastroin- 25. InnocentiF,UndeviaSD,IyerL,etal.Geneticvar- nomic instability. Cancer Res 2003;63:1608 ^ 14. testinal stromal tumors. Clin Cancer Res 2002;8: iants in the UDP-glucuronosyltransferase 1A1 gene 32. Arnold CN, Goel A, Compton C, et al. Evaluation of 3034^8. predict the risk of severe neutropenia of irinotecan. J Microsatellite Instability, hMLH1 expression and 40. Heinrich MC, ShoemakerJS, Corless CL, et al. Cor- Clin Oncol 2004;22:1382 ^ 8. hMLH1 promoter hypermethylation in defining the relation of clinical response to imatinib mesylate (IM) 26. Delaunoit T, Alberts SR, Sargent DJ, et al. Chemo- MSI phenotype of colorectal cancer. Cancer Biol Ther and tumor genotype in patients treated with IM for therapy permits resection of metastatic colorectal 2004;3:73 ^ 8. metastatic GI stromal tumors (GISTs). Proc Am Soc cancer: experience from Intergroup N9741. Ann Oncol 33. Goel A, Arnold CN, Niedzwiecki D, et al. Frequent Clin Oncol Annu Meet 2005;23:7. 2005;16:425 ^ 9. inactivation of PTEN by promoter hypermethylation in 41. Jemal A, MurrayT,Ward E, et al. Cancer statistics, 27. Dy GK, Krook J, Green E, et al. Impact of complete microsatellite instability-high (MSI-H) sporadic colo- 2005. CA Cancer J Clin 2005;55:10^ 30. response to chemotherapy on overall survival (OS) in rectal. cancers. Cancer Res 2004;64:3014^ 21. 42. Bertagnolli MM, Miedema B, Redston M, et al. Sen- advanced colorectal cancer (CRC): results from Inter- 34. Goel A, Arnold CN,Tassone P, et al. Epigenetic inac- tinel node staging of resectable colon cancer: results group N9741 [abstract]. American Society of Clinical tivation of RUNX3 in microsatellite unstable sporadic of CALGB 80001. A multicenter study. Ann Surg Oncology 2005 Gastrointestinal Cancers Sympo- colon cancers. Int J Cancer 2004;112:754 ^ 9. 2004;240:624 ^8. sium;2005Jan27-29;Miami,FL.p.A-185. 35. Arnold CN, Goel A, Niedzweicki D, et al. APC pro- 43. Redston M, Compton CC, Miedema BW, et al. 28. Macdonald JS, Smalley SR, Benedetti J, et al. Che- moter hypermethylation contributes to the loss of Analysis of micrometastatic disease in sentinel lymph moradiotherapy after surgery compared with surgery APC expression in colorectal cancers with allelic loss nodes from resectable colon cancer: results of CALGB alone for adenocarcinoma of the stomach or gastro- on 5q. Cancer Biol Ther 2004;3:960 ^ 4. 80001. J Clin Oncol 2006;24:878 ^ 83.

www.aacrjournals.org 3595s Clin Cancer Res 2006;12(11Suppl) June 1,2006 Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. Cancer and Leukemia Group B Gastrointestinal Cancer Committee

Richard M. Goldberg, Donna Niedzwiecki, Monica Bertagnolli, et al.

Clin Cancer Res 2006;12:3589s-3595s.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/12/11/3589s

Cited articles This article cites 39 articles, 15 of which you can access for free at: http://clincancerres.aacrjournals.org/content/12/11/3589s.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/12/11/3589s. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.