Shi, Qian

Abstract Background OSis the gold standard primary efficacy endpoint for evaluating treatment strategies for DLBCL. Because current standard of care first-line regimens with rituximab plus chemotherapy produce median OS > 8 years, an endpoint assessed earlier would accelerate the evaluation of new therapies and potentially allow patients (pts) to more rapidly benefit from study results. The SEAL group formally evaluated PFS as a surrogate endpoint for OS in first-line therapy for DLBCL. Methods Through a systematic literature search and input from lymphoma experts on the SEAL committee, we identified multicenter, randomized controlled clinical trials that evaluated active treatment and enrolled at least 100 pts with previously untreated DLBCL or aggressive non Hodgkin lymphoma by WHO/REAL classification, published after January 1, 1995. Studies which enrolled only early stage (I or II) pts and studies that could not provide individual pt data (IPD) were excluded from surrogacy analysis. Potential surrogate candidates for OS included: 1) PFS, defined as the time from initiation of induction treatment to progressive disease, relapse or death due to any cause, and 2) PFS rate at 24 months after initiation of induction treatment (PFS24). These endpoints were derived consistently for all studies based on IPD. Surrogacy was evaluated by the correlation between surrogate candidates and OS. Pt-level correlation was measured by the rank correlation coefficient (ρ) and global odds ratio for PFS and PFS24, respectively, estimated by the Copula model. Trial-level correlation was evaluated by the correlation between trial-specific treatment effects on surrogate candidates and OS with two commonly used trial-level R2s: R2WLS (Sargent et al JCO 2005) and R2Copula (Burzykowski et al J R Stat Soc C 2001 and Burzykowski et al J R Stat Soc A 2004). For the R2 measures, a value closer to 1 indicates a stronger correlation. We pre-specified that if either trial-level R2 was >=0.80 and its 95% confidence interval (CI) excluded 0.60, the candidate would meet criteria for surrogacy. Results Thirteen of 14 studies collected to date in the SEAL database met all inclusion criteria. IPD for 7507 pts (median age 63 years, 54% male, 46% ECOG performance status 0, 8% IPI 0, 50% IPI 1-2, 64% stage III/IV, median follow-up time of 52 months) were analyzed. Among 13 studies, one tested maintenance treatment only and two tested both induction and maintenance regimens. One study included four treatment arms. Thus at the trial level, 17 two-arm comparisons were formed: 14 comparisons comparing induction regimens and 3 comparing maintenance; 15 comparisons involved rituximab. Longer PFS was highly correlated with longer OS at the pt level (ρ= 0.85, 95% CI 0.84-0.86), as was PFS24 (Global odds ratio = 61.1, 95% CI 52.6-69.6). The Table shows trial-level correlations, and the Figure shows trial-level association between estimated treatment effects (log(HR) [hazard ratio]) on PFS and OS. The trial-level correlation between PFS and OS was strong and met the pre-specified surrogacy criteria, with R2WLS of 0.83 (95% CI 0.58-0.94) and R2Copula of 0.85 (95% CI 0.72-0.99). Sensitivity analyses including only comparisons including rituximab or only induction regimens showed consistent results. The overall trial-level correlation between PFS24 and OS was reduced slightly, with R2WLS of 0.77 (95% CI 0.49-0.92) and R2Copula of 0.78 (95% CI 0.59-0.96). Performance of PFS24 was improved when restricting analysis to only including induction studies (see table). Conclusion This large IPD pooled-analysis demonstrates that PFS duration met the pre-specified criteria as a robust surrogate endpoint for OS, and can be considered as a surrogate endpoint for OS for frontline DLBCL trials evaluating immunochemotherapy. The earlier endpoint of PFS24 showed promise as a potential surrogate and further validation is warranted. Figure Figure. Disclosures Shi: Mayo Clinic: Employment. Flowers:Millenium/Takeda: Research Funding; TG Therapeutics: Research Funding; Pharmacyclics, LLC, an AbbVie Company: Research Funding; Gilead: Consultancy, Research Funding; NIH: Research Funding; Mayo Clinic: Research Funding; ECOG: Research Funding; Acerta: Research Funding; Infinity: Research Funding; AbbVie: Research Funding; Roche: Consultancy, Research Funding; Genentech: Consultancy, Research Funding. Ou:Mayo Clinic: Employment. Cunningham:Merrimack: Research Funding; Celgene: Research Funding; Bayer: Research Funding; Astra-Zeneca: Research Funding; Amgen: Research Funding; Medimmune: Research Funding. Pfreundschuh:Amgen: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees; Roche: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Spectrum: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding. Seymour:AbbVie Inc.: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Genentech: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Roche: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Gilead: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Janssen: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Takeda: Honoraria, Membership on an entity's Board of Directors or advisory committees. Jaeger:Roche: Honoraria, Research Funding; Celgene: Honoraria, Research Funding. Haioun:Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees; Sandoz: Honoraria, Membership on an entity's Board of Directors or advisory committees; Roche: Honoraria, Membership on an entity's Board of Directors or advisory committees. Ghesquieres:Mundipharma: Consultancy; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees; Roche France: Research Funding. Merli:Teva Pharmaceuticals Industries: Membership on an entity's Board of Directors or advisory committees; Roche: Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene: Membership on an entity's Board of Directors or advisory committees. Herbrecht:Cell Therapeutics, Inc: Membership on an entity's Board of Directors or advisory committees; Gilead: Membership on an entity's Board of Directors or advisory committees; Servier: Membership on an entity's Board of Directors or advisory committees. Flament:Celgene: Employment, Equity Ownership. Fu:Celgene: Employment, Equity Ownership. Coiffier:MorphoSys: Consultancy, Membership on an entity's Board of Directors or advisory committees; Gilead: Consultancy, Membership on an entity's Board of Directors or advisory committees; Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees; Celltrion: Consultancy, Membership on an entity's Board of Directors or advisory committees; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees. Sargent:Celgene: Research Funding.

Abstract Background: Limited data exist to describe the clinical features and outcomes for elderly pts with FL. The FLASH group performed a prospectively planned pooled analysis of IPD from first-line RCTs and examined associations between age (<= 70 vs. > 70 years), clinical characteristics and FL outcomes. Methods: We identified 18 randomized controlled multicenter clinical trials in the FLASH database which enrolled elderly pts (> 70 yrs). From these 18 studies, 5922 previously untreated FL pts were included for this analysis. Complete response (CR) at 24 and 30 months were defined as whether the pt achieved and remains in CR at 24 months and 30 months after initiation of induction treatment. PFS24 is defined as proportion of pts progression-free and alive at 24 months post-randomization. Time to progression (TTP) and overall survival (OS) were defined as time from randomization to the date of progression (censoring for death without progression) and date of death due to any cause, respectively. Progression-free survival (PFS) was defined as time from randomization to the date of progression or death, due to any cause, whichever comes first. Non-lymphoma death was defined as death without prior progression. Early disease outcomes, i.e. CR24, CR30, and PFS24 were primary outcomes; secondary outcomes were TTP, OS, and PFS. For time-to-event outcomes, Kaplan-Meier method and log-rank test were used for univariate estimation and comparison; stratified Cox proportional hazard modeling was used for multivariable analyses. Cumulative incidence methods were used to model PFS while treating disease progression as the primary event of interest and non-lymphoma death as a competing risk. For binary outcomes, multivariable logistic regression was used to estimate the association between age groups and outcomes. Variable adjusted in the regression models are FLIPI score without the age component, ECOG performance status (>= 2 vs < 2), and rituximab use. Results: Among 5922 previously untreated FL pts from 18 RCTs, 63% (n = 3728) were <= 60, 28% (n = 1652) 61-70, and 9% (n = 542) > 70 yrs. Pts age > 70 (vs. <= 70) more commonly had elevated LDH (42% vs 36%, p = 0.02), hemoglobin < 12 g/dL (27% vs 19%, p < 0.0001), ECOG PS >= 2 (8.8% vs 5.0%, p = 0.0004), and elevated β2-microglobulin (68% vs 49%, p < 0.001), less often had > 4 lymph nodes involved (54% vs 65%, p < 0.001), and had similar FLIPI scores without age component (p = 0.17). There were no significant differences between groups in: Ann Arbor stage (94% vs 95% stage III/IV) or rituximab use (62% vs 58%). Median survival follow-up was 5.6 yrs. Pts > 70 yrs did not differ from pts <= 70 in rates of CR24 (27% vs 30%, p=0.32), CR30 (29% vs 31%, p=0.55) or PFS24 (63% vs. 67%, p=0.10). Median TTP was 3.8 and 4.0 years for pts > 70 and <= 70, respectively (HR=1.08, 95% CI 0.96-1.23, p=0.20). With a median OS of 14.6 years for all pts, median OS was 7.4 and 15.7 years for pts > 70 and <= 70, respectively (HR=2.35, 95% CI=2.03-2.73, p<0.001). PFS was statistically significantly different but clinically similar at median 3.1 and 3.8 years for pts > 70 and <= 70, respectively (HR=1.26, 95% CI =1.12-1.41, p<0.001). Further investigation showed the difference in PFS was due to higher incidence of non-lymphoma death in pts > 70 with no difference in disease progression (Figure). In regression models, adjusting for rituximab use, ECOG PS >= 2, and FLIPI score without the age component, age > 70 was a significant predictor of OS and PFS (due to higher incidence of non-lymphoma death), but not PFS24, CR24 or CR30 (Table). Conclusions: FL pts > 70 yrs treated on trials have similar early disease outcomes to younger patients. There is no disease-specific outcome difference between age groups. Age alone should not disqualify patients from standard treatments or RCTs. Disclosures Flowers: Genentech: Consultancy, Research Funding; NIH: Research Funding; Mayo Clinic: Research Funding; TG Therapeutics: Research Funding; Millenium/Takeda: Research Funding; ECOG: Research Funding; Acerta: Research Funding; Pharmacyclics, LLC, an AbbVie Company: Research Funding; Roche: Consultancy, Research Funding; AbbVie: Research Funding; Infinity: Research Funding; Gilead: Consultancy, Research Funding. Ou:Mayo Clinic: Employment. Shi:Mayo Clinic: Employment. Hochster:Genentech: Consultancy, Other: Consultancy fees for advisory boards for GI cancer and bevacizumab with amounts within Yale University de minims guidelines. Brice:Roche: Honoraria; Takeda Pharmaceuticals International Co.: Honoraria, Research Funding; Bristol Myers-Squibb: Honoraria; Gilead: Honoraria; Seattle Genetics: Research Funding. Hiddemann:Roche: Other: Grants; Roche: Membership on an entity's Board of Directors or advisory committees; Genentech: Other: Grants. Marcus:Takeda: Consultancy, Speakers Bureau; Roche: Consultancy, Speakers Bureau. Kimby:Abbvie: Consultancy, Honoraria; Jansen: Consultancy, Honoraria; Celgene: Honoraria; Baxalta: Consultancy; Gilead: Honoraria. Herold:Gilead: Other: Personal fees from member advisory board; Celgene: Honoraria; Genentech: Other: Grants; Roche: Honoraria, Other: Grants. De Bedout:Celgene: Employment. Nielsen:Hoffmann-La Roche: Employment. Morschhauser:Roche: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Servier: Consultancy, Honoraria; Gilead Sciences: Consultancy, Honoraria; Janssen: Honoraria. Rummel:Mundipharma GmbH: Other: Personal fees, Research Funding; Roche Pharma AG: Other: Personal fees, Research Funding. Vitolo:Gilead: Other: Honoraria for lectures; Takeda: Other: Honoraria for lectures; Roche: Membership on an entity's Board of Directors or advisory committees, Other: Honoraria for lectures; Janssen: Membership on an entity's Board of Directors or advisory committees, Other: Honoraria for lectures. Salles:Celgene: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Gilead: Honoraria, Research Funding; Janssen: Consultancy, Honoraria; Roche/Genentech: Consultancy, Honoraria, Research Funding; Mundipharma: Honoraria. Sargent:Celgene: Research Funding.

Abstract Introduction: A recent study revealed an antiproliferative and apoptotic effect of propranolol on multiple myeloma (MM) cells. Our previous small matched case-control study showed longer survival in patients with propranolol and other beta-blockers (BB) intake than those without. This larger scale study was conducted to confirm the positive association of BB and MM survival. Methods: We identified 1971 newly diagnosed pts seen at Mayo Clinic between 1995 and 2010. Cardiac medication usage after diagnosis of MM was extracted from patient records and categorized based on BB intake. Cause of death was collected with death due to MM as the primary interest event and death due to cardiac disease or other reasons as competing risk events. The primary outcomes were MM disease-specific survival (DSS) and overall survival (OS). Cumulative incidence functions and Kaplan-Meier method were used to estimate the 5-year cumulative incidence rate (CIR) of MM death and OS rate, respectively. DSS and OS were compared by Gray's test and log-rank test, respectively. Multivarable Cox proportional hazard models were used to estimate the adjusted cause-specific HR (HRCSadj.) and hazard ratio (HRadj.) for DSS and OS, respectively, adjusting for demographics, disease characteristics, diagnosis year, and various chemotherapies. Results: 930 (47.2%) of MM patients had no intake of any cardiac medications; 260 (13.2%) had BB only; 343 (17.4%) used both BB / non-BB cardiac medications; and 438 patients (22.2%) had non-BB cardiac drugs. Five-year CIR of MM death and OS rate were shown in table. Superior MM DSS was observed for BB only users, compared to patients without any cardiac drugs (HRCSadj., .53, 95% confidence interval [CI], .42-.67, padj.<.0001) and non-BB cardiac drugs users (HRCSadj., .49, 95% CI, .38-.63, padj.<.0001). Patients received both BB and other cardiac drugs also showed superior MM DSS than non-cardiac drugs users (HRCSadj.., .54, 95% CI, .44-.67, padj.<.0001) and non-BB cardiac drug users. (HRCSadj., .50, 95% CI, .40-.62, padj.<.0001). MM DSS does not differ between BB users with and without other cardiac drugs (padj.=0.90). Multivariable analysis showed the same pattern for OS. None of the MM therapies impacted the differences in DSS and OS among BB intake groups (interaction padj.>.60). Conclusion: MM patients with BB intake showed reduced risk of death due to MM and overall mortality compared to patients who used non-BB cardiac or never used cardiac drugs. The result warrants further investigation for anti-cancer effect of BB in MM. Disclosures Shi: Mayo Clinic: Employment. Kumar:Onyx: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; Array BioPharma: Consultancy, Research Funding; Sanofi: Consultancy, Research Funding; Skyline: Honoraria, Membership on an entity's Board of Directors or advisory committees; AbbVie: Research Funding; Glycomimetics: Consultancy; Janssen: Consultancy, Research Funding; Noxxon Pharma: Consultancy, Research Funding; Millennium: Consultancy, Research Funding; BMS: Consultancy; Kesios: Consultancy. Gertz:NCI Frederick: Honoraria; Celgene: Honoraria; Med Learning Group: Honoraria, Speakers Bureau; Research to Practice: Honoraria, Speakers Bureau; Alnylam Pharmaceuticals: Research Funding; Novartis: Research Funding; Prothena Therapeutics: Research Funding; Ionis: Research Funding; Annexon Biosciences: Research Funding; GSK: Honoraria; Sandoz Inc: Honoraria. Kapoor:Celgene: Research Funding; Amgen: Research Funding; Takeda: Research Funding. Dispenzieri:pfizer: Research Funding; Celgene: Research Funding; Alnylam: Research Funding; Jannsen: Research Funding; GSK: Membership on an entity's Board of Directors or advisory committees; Prothena: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees, Research Funding.

3583 Background: The Microsatellite instability (MSI) or deficient mismatch repair (dMMR) phenotype is usually taken as a single biological entity whereas no data are available concerning prognosis and response to chemotherapy between sporadic and familial dMMR cases. Methods: Resected KRAS exon 2 wild-type (WT) tumor stage III colon cancers (N = 4596) from patients (pts) randomly assigned to FOLFOX +/- cetuximab in two adjuvant large phase III trials were prospectively analyzed for MSI status and dMMR mechanism (sporadic vs familial). Stratified Cox models were used to assess prognostic and predictive values of dMMR mechanism by treatment arms, adjusting for age, gender, tumor grade, ECOG PS, pT/pN stage and primary tumor location. Results: Among dMMR patients with complete data for dMMR mechanism analysis (N = 354), there were 255 (72%) sporadic ( BRAF mutated or WT with MLH1 methylation) and 99 (28%) familial (loss of MSH2 or MSH6, or loss MLH1 with BRAF WT and unmethylated MLH1) cases. A large proportion of dMMR sporadic cases were mutated for BRAF (n = 200; 80%). In pts treated with FOLFOX, the disease-free survival (DFS) was not statistically different by dMMR mechanism, while for pts treated with FOLFOX + cetuximab, the sporadic cases did worse than familial cases (DFS; adjusted (adj) HR, 2.69; 95% CI, 1.02-7.08; P= 0.04). Considering the predictive value, a deleterious effect of adding cetuximab to FOLFOX was observed in sporadic (DFS; adjHR, 1.68; 95% CI, 1.01-2.79; P= 0.04) but not in familial dMMR pts (interaction P value regarding treatment effect = 0.03). Furthermore, a non-significant trend to a deleterious effect of adding cetuximab to FOLFOX was observed in BRAF mutant (DFS; adjHR, 1.66; 95% CI, 0.95-2.92; P= 0.07) but not in BRAF WT pts. Conclusions: The addition of cetuximab to FOLFOX was associated with reduced DFS in patients with sporadic dMMR cases. Further studies including the methylator phenotype (CIMP) analysis are needed to validate these results. Clinical trial information: NCT00265811 and NCT00079274.

3539 Background: Previous studies suggest a possible gender-specific response to Bev in mCRC, showing a benefit in males, while the effect in females is less significant. Therefore, we evaluated response to Bev according to gender. Methods: Data from 3369 mCRC patients enrolled on 4 first-line randomized trials testing Bev (2000-2007) were pooled. Association between gender and progression-free survival (PFS)/overall survival (OS) was evaluated by stratified Cox regression model, adjusted for potential confounders. Predictive value was evaluated by interaction (inter.) effect between gender and treatment. In a pre-planned secondary analysis, analyses were stratified using an age cut-point of 60 years (commonly used in breast cancer trials) to evaluate the possible role of menopausal-related effects. Results: OS was not statistically different between males and females (median OS [mOS], 18.8 vs. 17.6 months [mo], adjusted hazard ratio [HRadj], 0.93, 95% confidence interval [CI], 0.84-1.03, p, 0.15) in the overall population. Bev was associated with an improved mOS in males and females, with a 2.3 and 0.6 mo benefit, respectively, as well as an improved PFS. There was no statistically significant interaction effect between gender and treatment (see table). Further stratified by age (< vs. ≥ 60 years), Bev resulted in improved PFS and OS in both genders, at all ages, except for the effect in young females which did not reach statistical significance (see table). Conclusions: Our results confirmed the mOS benefit from addition of Bev to first-line chemotherapy in mCRC in both genders, although the benefit in females was < 1 month. For females under the age of 60, there are uncertainties for mOS benefit from addition of Bev and further evaluation is needed. [Table: see text]

3525 Background: Microsatellite instable/deficient mismatch repair (MSI) metastatic colorectal cancers have been reported to be of poor prognosis. The interaction between MSI and BRAFV600E mutation complicates the picture. Methods: Patients with resected stage III CC from 7 studies with disease recurrence and data available for MSI and BRAFV600E status were analyzed. The primary endpoint was survival after recurrence (SAR) to assess the prognostic roles of MSI and BRAFV600E, respectively. Associations of markers with SAR were analyzed using Cox proportional hazards models adjusted for clinicopathologic features (data collected 12/1998 to 11/2009). Results: Among 2630 patients with cancer recurrence (1491 men [56.7%], mean age, 58.5 [19-85] years), multivariable analysis revealed that patients with MSI tumors (n = 220) had significantly better SAR (adjusted hazard ratio [aHR], 0.82; 95% CI, 0.69-0.98; P = .029) than patients with microsatellite stable /proficient MMR (MSS) tumors (n = 1766). This was also observed when looking at patients treated by the standard FOLFOX adjuvant regimen only (aHR, 0.76; 0.58-1.00; P = .048). Same trends were observed when looking at MSI/dMMR patients outcome in BRAFV600E wild-type (aHR, 0.84; P = .10) and mutant (aHR, 0.88; P = .43) subgroups separately, without reaching statistical significance. As previously described poor SAR was observed in BRAFV600E mutants vs wild type patients (n = 244; aHR, 2.06; 95% CI, 1.73-2.46; P < .0001) and this was also true in BRAFV600E mutants MSI/dMMR patients (n = 77, aHR, 2.65 ; 95% CI, 1.67-4.21; p < .0001). Other factors associated with a poor SAR were : olderage, male gender, T4/N2, proximal primary tumor location, poorly differentiated adenocarcinoma, and early recurrence (by 1y increase). Conclusions: In stage III colon cancer patients recurring after adjuvant chemotherapy and before the era of immuno-oncologic agents, MSI/dMMR was associated with a better survival compared to MSS. BRAFV600E mutation seems to be a poor prognostic factor for both MSI/dMMR and MSS/pMMR patients.

3520 Background: Duration of adjuvant FOLFOX or CAPOX for stage III CC is being guided by pt stratification into low (T1-3N1) and high (T4 or N2) risk groups based on the IDEA study. We determined the relative contributions of clinical and molecular features for prediction of time-to-recurrence (TTR) and survival after recurrence (SAR) within each risk group. Methods: Stage III CC (N=5,430) from 2 trials of adjuvant FOLFOX ± cetuximab with similar outcome by study arm [NCCTG N0147 (Alliance), PETACC-8] were used. Tumors were analyzed for mismatch repair (dMMR vs pMMR), mutations in KRAS (exon 2) and BRAFV600E. Median pt follow-up was 83.4 months. Relative contributions to predicting outcome were assessed using χ2 (Harrell’s rms) based on multivariable (MV) Cox models. Results: N (50.8%) and T (31.8%) stage were the top two contributors to prediction of TTR which supports risk grouping. High risk (n=2566) vs low risk (n=2774) pts had poorer TTR (HR 2.7, 95% CI, 2.4-3.0) and SAR [HR 1.6 (1.4-1.9)], both p<.0001. TTR: KRAS contributed the most to predicting TTR among high (58.6%) and low (51.1%) risk pts (Table). Contribution of MMR (16%) to predicting TTR was limited to low risk pts. Contribution of BRAFV600E to TTR was nearly 3-fold increased in high vs low risk pts. SAR: BRAFV600E contributed the most to predicting SAR, especially in high vs low risk pts (2-fold increase). Tumor sidedness and performance status (PS) were key contributors to SAR, but not TTR. MV associations: TTR: low risk, KRAS [HR 1.7 (1.4-2.3], MMR [HR 0.55 (.35-.87), gender (M/F) [HR 1.3 (1.0-1.5)], all p<.04]; high risk: BRAF [HR 1.3 (1.1-1.7)], sidedness (R vs L) [HR 1.14 (1.0-1.3)], KRAS [HR 1.4 (1.2-1.6)], all p<.04]. SAR: BRAF, sidedness, PS (all p<.05). Conclusions: KRAS mutation was the strongest predictor of shorter TTR in both risk groups whereas BRAFV600E was the primary driver of SAR, especially in high risk pts. Support: U10CA180821, U10CA180882, U24CA196171; BMS, Pfizer, Sanofi. NCT00079274.[Table: see text]

3502 Background: DFS with 3 years median follow-up (3yDFS) was validated as a surrogate for OS with 5 years median follow-up (5yOS) in adjuvant chemotherapy CC trials prior. Recent data showed improved survival after recurrence and OS, over time, in pts who received adjuvant FOLFOX. Hence, re-evaluation of the association between DFS and OS, as well as the optimal follow-up of OS to aid its utility in future adjuvant trials is needed. Methods: Individual patient data from 8 randomized adjuvant studies conducted from 1998-2009 were included; 3 trials tested anti-VEGF or anti-EGFR agents. Trial-level surrogacy examining the correlation of treatment effect estimates (i.e. hazard ratios) of 3yDFS and 5y to 8yOS was evaluated using both linear regression (R2WLS) and Copula bivariate (R2Copula) models. For the R2, a value closer to 1 indicates a stronger correlation. Prespecified criteria for surrogacy required either R2WLS or R2Copula ≥ 0.80 and neither < 0.7, with lower-bound 95% Confidence Interval (CI) > 0.60. The rank correlation coefficient (ρ) quantified the individual-level surrogacy. Results: Total of 18,886 pts were analyzed, with median age 60, 54% male, 83% stage III, 59% > 12 nodes examined. Median follow-up for survival ranged from 5 to 10 years across trials. Trial level correlation between 3yDFS and OS remained strong (R2WLS ≥0.74; R2Copula ≥ 0.89) and increased as the median follow-up of OS extended longer (see table). Analyses limited to stage III pts and/or trials tested biologics showed consistent results. Conclusions: 3yDFS remains a validated surrogate endpoint for 5yOS in adjuvant trials in CC pts per prespecified criteria. The correlation was strengthened with more than 6 years of follow-up for OS. [Table: see text]