Sargent, Daniel J.

3515 Background: Patients with metastatic colorectal cancer (mCRC) treated with 5Fluorouracil plus leucovorin (5FU/LV) have a median life expectancy (LE) of approximately12 months. The addition of irinotecan, oxaliplatin, bevacizumab and cetuximab has increased LE significantly. Although they have been widely adopted, little is known about the financial impact of these new drugs. Methods: Using published reports and aggregate data from NCCTG 9741, we developed a Markov Model which assumes forward progression through up to three lines of therapy compared to 5FU/LV alone. Patients who do not die of toxicity (tox) transition through supportive care prior to death. State changes are based on progression and dose-limiting toxicity. Drug costs are based on Average Sales Price. No other direct or indirect costs are included. Dose modifications for toxicity are defined as 80% of standard doses. Sensitivity analyses (SA) were performed on key variables. Results: LE, total drug costs, and cost-effectiveness (CE) ratios compared to single line-5FU/LV and FOLFOX for 9 commonly used treatement sequences are presented below. SA show that results are sensitive to progression rates,drug costs and length of time on supportive care. Changes in 1st line tox rates have a greater impact on final results than changes in 2nd or 3rd line therapy tox rates. Conclusions: Using drug costs alone, this model shows sequential combination therapy including all available agents to cost $2000-$2800K/week life gained ($100-$145K/year) compared to both 5FU/LV and FOLFOX. For clarity only 9 strategies are presented, but the model presented will contain multiple sequences consisting of 1–3 lines of therapy. Refined tox data, associated costs and quality of life adjustments are needed for realistic comparisons among specific combinations. Combination therapy may have CE ratios similar to other currently accepted intensive medical interventions. These data can help inform discussions of how the cost of care impacts patients, providers, and societies. [Table: see text] [Table: see text]

3503 Background: The association between race & toxicity or response to systemic chemotherapy is poorly understood. We correlated relevent polymorphisms with toxicity & outcome data in AA & C patients (Pts) enrolled in N9741, a large randomized MCRC trial. Methods: Toxicity, response rate (RR), time to progression (TTP), overall survival (OS), & dose intensity (DI) were examined as a function of race in 1412 Pts enrolled on IFL, both full and reduced dose, FOLFOX, or IROX including 486 Pts with pharmacogenomic data. Accrual by race permitted comparison of C (n=1297) to AA (n=112). Results: C & AA were similar in disease extent, PS, gender, prior adjuvant treatment, and DI at cycles 1, 3, 6, or 12. The impact of race on RR, toxicity & the distribution of four relevent drug metabolism genotypes are in the table . AA had lower RR overall & in each treatment arm. Lower toxicity rates were seen in AA Pts, mainly due to less diarrhea, as measured by both CTC V2 & patient reported outcome data. These relationships were maintained in multivariate models adjusting for arm, age, gender, & PS. In these multivariate models, the odds ratio between RR and race was 1.72 (p=0.012) (29% AA vs 41% C) & for severe toxicity the odds ratio was 1.76 (p=0.006) (34% AA vs 48% C). Overall, race was not a significant predictor of TTP or OS (p > 0.50 for each). The hazard ratio comparing FOLFOX to IFL for C was 0.67 and for AA was 0.55. TTP for FOLFOX was superior to both IFL and IROX for both C & AA. Significant differences in pharmacogenetic variants were observed between C & AA, although the sample size for investigating the three-way combination of Rx, race, & pharmacogenomics is too small (36 AA) to draw definitive conclusions. Conclusions: It appears that AA have less toxicity & a lower RR than C when treated with standard MCRC regimens. Different frequencies of polymorphisms in genes relevent to drug metabolism may help explain these clinical differences. [Table: see text] [Table: see text]

21046 Background: The Epidermal Growth Factor Receptor (EGFR) is a transmembrane receptor tyrosine kinase that is important for colon cancer cell growth and represents a therapeutic target. We studied EGFR expression in relation to rates of apoptosis and cell proliferation. Methods: TNM stage II and III colon carcinomas (n= 361) from patients treated in a 5-FU-based adjuvant trial were studied. EGFR expression (EGFR pharmDx™) and extent were analyzed and their relationship to rates of cell turnover, histologic grade, mismatch repair (MMR) status (defective if instability at BAT 26 locus plus loss of hMLH1, hMSH2 or hMSH6 proteins) and clinical outcome were determined. Results: EGFR plasma membrane staining was detected in 214 (59.3 %) colon carcinomas. Increased EGFR intensity (2+, 3+) was associated with a higher number of caspase-3- (p= 0.02) and Ki-67- (p< 0.001)positive tumor cells. Increased EGFR intensity (p= 0.02) and extent (p= 0.03) were associated with poor differentiation, and did not depend upon MMR status (Breslow-Day test p ≥ 0.46). Diffuse (vs. focal) EGFR extent was increased in tumors with defective MMR [22 of 32 (69%)] compared to those with intact MMR [136 of 271 (50%)] (p= 0.047). EGFR intensity was unrelated to MMR status. Neither EGFR intensity nor extent was significantly associated with patient survival. Conclusions: Stage II and III colon cancers with EGFR overexpression show adverse prognostic variables, i.e., increased rates of cell turnover and poor histologic grade, suggesting that EGFR may be an important target for adjuvant therapy. *Chi-square test for categorical data, Wilcoxin rank-sum test for continuous data [Table: see text] No significant financial relationships to disclose.

10529 Background: BAX mutations are associated with defective DNA mismatch repair (MMR) in human colon cancers. However, the impact of BAX inactivation upon tumor cell apoptosis and proliferation in vivo remain unknown. We analyzed and compared caspase-3 and Ki-67 expression in tumors with and without BAX mutations. Methods: TNM stage II and III (n= 377) colon carcinomas were studied from participants in a 5-FU-based adjuvant therapy trial. Archival tumors were analyzed for instability at the BAT26 mononucleotide locus using polymerase chain reaction and hMLH1, hMSH2 and hMSH6 by immunohistochemistry (IHC). Frameshift mutations in a tract of eight deoxyguanosines within BAX were analyzed. Expression of caspase-3 and Ki-67 proteins were analyzed by IHC. Results: Thirty-nine of 377 (10%) tumors showed defective MMR defined as instability at BAT26 and loss of either hMLH1, hMSH2 and/or hMSH6 proteins. BAX mutations were found in 20 of 37 (54%) MMR deficient tumors and in 1 of 50 (2%) tumors with intact MMR. Mean and median number of caspase-3-positive cells were increased in tumors with defective MMR (p= 0.04), but did not differ based upon BAX status [ Table ]. However, tumors with BAX mutations showed higher Ki-67 labeling indices compared to those with wild type BAX (p= 0.01)[ Table ]. Neither BAX mutations nor caspase-3- positive cells were prognostic in a univariate analysis. Tumors with lower Ki-67 extent had improved overall survival (p=0.06), but not DFS (p=0.24). Defective MMR (vs intact) was associated with better DFS in a multivariate analysis (p= 0.03). Conclusion: MMR deficient colon cancers show frequent BAX inactivation, yet have increased apoptotic rates as indicated by increased caspase-3 expressing tumor cells. BAX mutation was associated with hyperproliferation suggesting a growth advantage compared to wild type tumors. [Table: see text] No significant financial relationships to disclose.

6517 Background: In NCCTG CTs a subset of AEs are assessed at each patient (pt) evaluation based on the known safety profile of agents(s). The NCCTG routinely pre-fills the “known” AE list onto CT Case Report Forms (CRFs). Newly identified AEs may expand the AE assessment list for ongoing CTs. Our survey of NCCTG AE data (JCO 2005), demonstrated that 85% of AEs reported were pre-filled on CRFs, of which, 83% did not actually occur (Grade 0). Extending this work, we evaluated the influence of pre-filling AEs on CRFs, relative to the final AE rates reported. Methods: Our non-random sample contains 507,899 AEs collected from 1/99–6/06 on 74 NCCTG CTs, 13 of which had AEs added to the CRF pre-fill list during the CT (2,604 pts, 3 Ph I/II, 8 Ph II, 2 Ph III, 9 investigational agents). Results: An average of 2.8 AEs (range 1–6) were added to CRFs for ongoing CTs, primarily for Oxaliplatin induced AEs. 58% (21/36) of AEs added during a CT were not reported prior to the addition, 22% (8) were not reported afterwards. 5 CTs had significantly higher AE rates (p<0.01) after expanding the AE list, most notably for required blood chemistries (SGOT/alk phos/creatinine/bilirubin 14.2 vs 0.72%). Overall, the same AEs were 4-fold (range 0–25) as likely to be reported when pre-filled on the CRF. Regardless of pre-filling, only 6% of the newly required AEs were Gr 3+. Conclusions: Our data suggest a significant difference between the AE rates reported if included in the CT CRF assessment list. This may significantly influence the reported results of a CT, explaining differential AE rates reported across CTs of the same agent(s). A prospective study is planned to formally evaluate this observation. Supported by NIH Grant CA25224. No significant financial relationships to disclose.

6511 Background: In March 1998, the Common Toxicity Criteria (CTC) version 2.0 introduced the collection of attribution of adverse events (AE) to study drug. Collection of attribution adds time and cost to the clinical trial process, and reporting in the literature has varied. At ASCO 2006 results were presented from NCCTG trial 97–24–51 that showed ∼ 50% of events on the placebo arm were reported as attributed to study drug and 25% of the time the same event was not consistently attributed. We sought to validate these results in a second placebo controlled trial. Methods: Patients on the placebo arm of the phase III ACOSOG trial Z9001 (imatinib mesylate vs. placebo after resection of primary gastrointestinal stromal tumors) were evaluated since the reported AEs could not be due to study drug. Attribution was collected per CTCv3.0 as not related, unlikely, possibly, probably, or definitely related, and categorized as “unrelated” (not related or unlikely) and “related” (possible, probable, or definite). Patterns were evaluated using generalized estimating equations adjusting for multiple events per patient. Results: 202 patients and 2871 AEs were included. 52% of AEs were reported as “related”. Known imatinib AEs were more likely to be reported as “related”. The reported rate of AE relatedness varied by AE type. No patterns were observed by gender, age, PS, severity or visit of AE. When the same event was reported on the same patient on multiple visits, the attribution category changed at least once 33% of the time with 19% changing from “related” to “unrelated” or vice versa. Conclusion: 52% of events on the placebo arm of Z9001 were reported as attributed to study drug and 19% of the time the same event was not consistently attributed. These data provide further evidence that attribution is difficult to determine, unreliable, and of questionable value in interpreting randomized clinical trials. No significant financial relationships to disclose. [Table: see text]

4080 Background: Waterfall plots have been used to capture the anti-tumor efficacy of both biologic agents (Ratain JCO 2006), which mainly exhibit a cytostatic effect, but also of chemotherapeutic agents. We are not aware of an application to therapeutic studies in CRC. We conducted waterfall plot analyses on the data of N9741 which compared IFL, FOLFOX4, and IROX as first-line therapy in ACRC (Goldberg JCO 2004). Methods: 634 (IFL 215, FOLFOX 215, IROX 204) of the 795 pts presented in the published analysis had measurable disease, and are included. The percentage change in tumor measurements from baseline to 12 wks was calculated, allowing for a time window of assessment from 6 to 20 wks for those without a 12 wk measurement. Pts with progression before the 12 wk assessment were assigned a 100% increase. The average reduction in tumor burden (ARTB) was calculated as difference in waterfall AUC divided by the number of pts. ARTB captures the average percentage reduction in tumor measurements at 12 wks for a patient entering the trial. Results: Waterfall plot analysis demonstrated some tumor growth at 12 wks in 23.7% (IFL), 15.8% (FOLFOX), and 26.9% (IROX) of pts. In contrast, tumor shrinkage was observed in 67.9% (IFL), 71.6% (FOLFOX), and 61.4% (IROX) of pts. Total differential AUC was -34.4 (IFL), -61.6 (FOLFOX), and -30.4 (IROX) resulting in an ARTB of 16.0% (95% Confidence Interval (CI) (23.7%, 8.2%)) for IFL, 28.7% (35.1%, 22.2%) for FOLFOX, and 14.9% (22.5%, 7.3%) for IROX. Conclusions: Waterfall plot analyses are feasible in this and likely other CRC trials using conventional chemotherapy. Our analysis confirms the published results of N9741, but suggests a larger degree of benefit for individuals based on anti-tumor activity than captured by conventional response analysis. Waterfall plots provide a potentially useful new efficacy metric, the average reduction of tumor burden per pt, which should be validated in future trials. No significant financial relationships to disclose.

4067 Background: N9741 randomized 1691 patients (pts) to seven 5FU, Ox and Iri containing regimens for MCRC. After 20.4 months median follow-up, FOLFOX and IROX improved time to progression (TTP) and overall survival (OS) compared to IFL. Here we update OS and TTP, and report an analysis of factors prognostic of pt outcome. Methods: 5 yr OS and TTP were calculated by Kaplan-Meier per treatment (rx) arm. Pt factors [rx arm, age, gender, prior adjuvant rx, body mass index (BMI), performance status (PS), number of disease sites (#sites), neutrophil count (ANC), hemoglobin, platelets, bilirubin, alkaline phosphatase (ALK), aspartate aminotransferase (AST)] were assessed for univariate association with OS, TTP, response rate (RR), and any grade =3 toxicity. Associated factors were included in multivariate Cox and logistic regression models. Results: After a median follow-up of 4.3 yrs, FOLFOX treated pts were more likely to live 5 yrs, 9.2% versus 5.4% for IROX (p=.016) and 3.8% for IFL (p<.001). Median TTP was also significantly longer in the FOLFOX arm, 9.2 mo vs 6.5 mo for IROX (p=<.001) and 6.0 mo for IFL (p=<.001). The prognostic model found higher PS, ANC and ALK, and more disease sites were prognostic of worse OS (see table ). Age = 70 was associated with poorer survival (HR death 1.33, p=.03), but not TTP, RR, or grade =3 toxicity. FOLFOX rx was the most powerful prognostic factor for TTP and OS. The odds of grade =3 toxicity were significantly higher in FOLFOX treated pts (OR 1.65, p<.001) and lower in men (OR .63, p<.001). No significant interactions between rx arm and pt factors were present, suggesting no factor examined is predictive of rx-specific outcome. Conclusions: 9% 5 yr OS in MCRC pts treated with first-line FOLFOX sets a new benchmark. We confirmed pre-rx PS, WBC, and ALK are prognostic for OS, though FOLFOX rx was the strongest prognostic factor. No factor was associated with differential outcome or toxicity by rx arm, thus these factors cannot be used to guide rx selection. [Table: see text] [Table: see text]

4057 Background: The prognosis for patients with hepatic metastases from M-CRC can be poor. However, surgery followed by HAI FUDR plus systemic (SYS) 5-FU improves 5-yr survival. Dual therapy OXAL+CAPE has demonstrated activity in advanced CRC. We report final results of an analysis of SYS OXAL+CAPE, alternating with HAI FUDR. The primary endpoint is 2-yr survival (2YS), with 36 of 45 patients surviving 2 yrs as evidence of promising efficacy. Methods: Patients with M-CRC liver lesions amenable to resection ± ablation were eligible. Prior adjuvant chemotherapy for completely resected primary was allowed. HAI+SYS therapy was initiated 21–56 days post-metastasectomy. Four alternating courses of HAI consisted of 0.2 mg/m2/d FUDR and dexamethasone, d1–14 wks 1&2. SYS included 130 mg/m2 OXAL d1, with CAPE at 1,000 mg/m2 po BID, d1–14, wks 4&5. Two additional 3-wk courses of SYS were given. CAPE was reduced to 850 mg/m2/BID after interim review of toxicity. Results: 54 of 73 eligible patients initiated HAI FUDR + SYS. 52% had a solitary met and 24% presented with bilobar mets. Patients completed median of 6 cycles (range 1–6). Reasons for discontinuation included: refusal/toxicity (10), completed per protocol (32), recurrence (4), and medical/other (3). No HAI+SYS related deaths occurred. Median follow-up on the 42 survivors is 28 months (range 7–51). 6 deaths occurred within 2 yrs. 48% (26/54) have recurred; 42% (11/26) with liver involvement. Median time-to-progression is 30 months. The estimated 2YS rate is 88% (95% CI 80–98%) and median overall survival is 46 months (95% CI 41.3-NA). Conclusions: The combination of HAI FUDR and SYS therapy appears to improve outcome following resection of hepatic CRC-M. A 2YS of 88% exceeds our preplanned level of success, supporting the use of this combination in the ongoing NSABP trial C-09. Supported by NIH Grant CA25224–18, Sanofi-Synthelabo, and Roche Laboratories, Inc. No significant financial relationships to disclose.

4017 Background: Published reports suggest that CC pts of A descent have inferior survival compared to W pts. Whether these differences are explained by clinical factors at diagnosis, socioeconomic factors impacting access to care, or intrinsic differences in the biology of the tumors or the response to therapy is unclear. Pts in clinical trials have data collected for important baseline clinical factors and should receive comparable oncologic care regardless of socioeconomic factors. Methods: We analyzed data from 13,435 individual pts on 11 phase III adjuvant CC trials accrued from 1977 to 2002. Analysis was restricted to stage II and III pts, with race reported as black or white. Endpoints were overall survival (OS - time to death), recurrence-free survival (RFS - time to recurrence or death), and recurrence-free interval (RFI - time to recurrence censoring for death). Cox models stratified by study controlled for gender, stage, age, and treatment type (rx) to determine the effect of race. Kaplan-Meier estimates (KM) were adjusted (adj) by the Xie-Liu method for study, gender, stage, age, and rx. Results: A pts (n=1134, 8.4%) were younger than W (median 58 vs 61, p<0.001) and more likely female (55 vs 45%, p<0.001). A pts had poorer OS than W pts ( table ). OS results were consistent in subsets defined by gender, stage, and age. RFS results were attenuated compared to OS, but still favored improved RFS in W pts ( table ). RFI results were further attenuated and not significantly different by race ( table ). Conclusions: Even with identical rx for CC in controlled clinical trials, A pts have poorer OS and RFS than W pts. The OS deficit was consistent across subgroups, and neither deficit was explained by differences in gender, stage, age, or rx. RFI was similar for both races, suggesting that the OS and RFS differences may be largely due to deaths unrelated to CC. [Table: see text] [Table: see text]