BACKGROUND:Primary tumor (PT) sidedness is an established prognostic marker in metastatic colorectal cancer (mCRC) and has a predictive impact on the efficacy of anti-epidermal growth factor receptor (anti-EGFR) antibody [monoclonal antibody (mAb)] in patients with RAS wild-type mCRC. This investigation focuses on patients with BRAFV600E-mutated (BRAFmt) mCRC and examines the efficacy of anti-EGFR mAbs in relation to primary tumor sidedness (PTS). PATIENT AND METHODS:This pooled analysis was carried out using individual patient data from five randomized studies in the first-line setting of mCRC. The population of interest was limited to patients with BRAFmt mCRC and known PTS. For analysis, treatment was stratified into two groups: those treated with anti-EGFR mAbs and those without. Dichotomous variables, such as overall response rate and objective response rate (ORR), were compared using chi-square or Fisher's exact test. Time-to-event endpoints [progression-free survival (PFS) and overall survival (OS)] were analyzed using the Kaplan-Meier method, log-rank test, and Cox regression. An interaction test was carried out via Cox regression. RESULTS:A total of 102 patients with BRAFmt mCRC were identified. The type of targeted therapy (anti-EGFR-based versus non-anti-EGFR) did not significantly impact the outcome. However, in patients with left-sided primary tumors, anti-EGFR mAb-based treatment, compared with non-anti-EGFR, was associated with a higher ORR (58% versus 34%; P < 0.01), trended toward improved PFS [hazard ratio (HR) 0.62; 95% confidence interval (CI) 0.34-1.13; P = 0.12], and demonstrated prolonged OS (HR 0.38; 95% CI 0.20-0.72; P < 0.01). In patients with right-sided primary tumors, anti-EGFR-based therapy had no effect on ORR (33% versus 36%; P > 0.99), induced inferior PFS (HR 1.97; 95% CI 1.12-3.47; P = 0.02), and trended toward a worse OS (HR 1.76; 95% CI 0.99-3.13; P = 0.05). CONCLUSION:This analysis suggests that PTS has predictive value for the efficacy of anti-EGFR mAb in the first-line treatment of BRAFmt mCRC.

Purpose: Platinum-fluoropyrimidine combinations are standard of care for treatment of metastatic esophagogastric adenocarcinoma. The optimal duration of first-line chemotherapy is unknown, however, and maintenance strategies have not yet been established. Design: MATEO is an international randomized phase II trial exploring efficacy and safety of S-1 maintenance therapy in human epidermal growth factor receptor 2 (HER2)-negative advanced esophagogastric adenocarcinoma. After 3 months of first-line platinum-fluoropyrimidine-based induction therapy, patients without progression were randomized in a 2 : 1 allocation to receive S-1 monotherapy (arm A) or to continue combination chemotherapy (arm B). The primary objective was to show non-inferiority of overall survival in the S-1 maintenance group. Progression-free survival, adverse events, and quality of life were secondary endpoints.Results: From 2014 to 2019, 110 and 55 patients were randomized in arm A and arm B, respectively (recruitment closed prematurely). Median overall survival from randomization was 13.4 months for arm A and 11.4 months for arm B [hazard ratio 0.97 (80% confidence interval 0.76-1.23), P = 0.86]. Median progression-free survival from randomization was 4.3 and 6.1 months for arm A versus arm B, respectively [hazard ratio 1.10 (80% confidence interval 0.86-1.39), P = 0.62]. Patients in arm A had numerically fewer treatment-related adverse events (84.9% versus 93.9%) and significantly less peripheral sensory polyneuropathy & GE;grade 2 (9.4% versus 36.7%).Conclusions: S-1 maintenance following platinum-based induction therapy leads to non-inferior survival outcomes compared with the continuation of platinum-based combination. Toxicity patterns favor a fluoropyrimidine maintenance strategy. These data challenge the continued use of platinum combination chemotherapy after response to 3 months induction therapy in patients with advanced human epidermal growth factor receptor 2-negative esophagogastric adenocarcinoma.

NSCLC is a disease of the elderly, but frail and older patients (pts) are underrepresented in clinical trials with scarce data to guide treatment. DURATION (NCT03345810) was a phase 2 trial conducted in 31 German centers between 12/17-01/22. 201 older or frail stage IV NSCLC pts (age ≥70 years and/or Charlson Comorbidity Index > 1 and/or performance status (PS) ECOG ≥ 2) were stratified to receive mCTX or dCTX according to the modified Cancer Aging Research Group score (mCARG >3 or <3, respectively), and subsequently randomized between 4 cycles of CTX and 2 cycles of CTX followed by D. Primary endpoint was the rate of treatment-related grade III/IV adverse events (AEs). The median follow-up was 36.8 months (mo) with 17% of pts censored. Median age was 76 years (range 56-90), 59 pts (30%) had an ECOG PS ≥2 and 112 pts (56%) a CCI >1. The PD-L1 tumor proportion score was 0%, 1-49% and >50% in 42%, 50% and 8% of pts, respectively. The rate of grade III/IV AEs was 34%, 34%, 24% and 29% for dCTX , dCTX-D, mCTX-D and mCTX respectively (p=0.68). Treatment-related deaths were reported in 1% (n=2, dCTX-D and mCTX). 2nd-line treatment (proportion anti-PD-1/PD-L1) was administered in 57% (43%), 42% (12%), 31% (16%), 38% (22%) cases after dCTX, dCTX-D, mCTX-D and mCTX, respectively. In the modified per-protocol (PP) population (including pts treated until cycle 3, excluding pts with major protocol deviations), the median overall survival (OS) was 14.3 vs 11.1 mo for dCTX and dCTX-D (HR 1.2, p=0.46) and 10.2 vs 8.3 mo for mCTX-D and mCTX (HR 0.7, p=0.25). Pts with mCARG ≤3 had significantly longer OS compared to patients with mCARG >3 in the PP population (11.0 mo vs. 5.3 mo in median, HR 2.10, p<0.001). The mCARG score stratified older or frail pts with advanced NSCLC for safe treatment conductance on dCTX, resulting in longer OS without increased toxicity compared to mCTX. The rate of grade III/IV-AEs did not exceed 34% and treatment-related mortality was ≤ 1% in both strata. Immature data show a numerical OS advantage for mCTX-D compared to mCTX, which may be related to the low rate of 2nd-line treatment < 40% in these most frail cases.

4008 Background: Chemotherapy (CT) is the standard of care in nonresectable locally advanced pancreatic cancer. The CONKO-007 trial studied the role of sequential CT and chemoradiotherapy (CRT). Methods: In this randomized multicenter phase III trial resectability was judged by an independent surgical board. Patients (pts) received induction chemotherapy (IC) for 3 months (3 cycles gemcitabine (Gem, 1000 mg/m² d1, 8, 15, q4w) or FOLFIRINOX (6 cycles, q2w)). After IC pts without progression were randomized to either continuing CT for another 3 months or receiving CRT (cumulative dose of 50.4Gy, single dose 1.8Gy + Gem 300 mg/m² weekly, followed by 1 cycle of Gem 1000 mg/m² at d1, 8, 15). The primary endpoint of the study was overall survival (OS) since the begin of IC. Determination of sample size calculated 590 pts to be randomized. Due to the exclusion of pts with progressive disease after IC a total of 830 pts should be enrolled. Due to delayed patient accrual the primary endpoint was changed to R0 resection rate resulting in an estimated sample size of 525 pts. Results: Between 04/2013 and 02/2021 a total of 525 pts were enrolled in 47 sites. 402 pts received IC with FOLFIRINOX and 93 pts with Gem. After IC 190 pts were excluded due to progression or toxicity, 335 were randomized, their median FU was 16 months. Hematological toxicities were significantly increased in the CRT arm, non-hematological toxicities were comparable. R0 CRM- resection rate and pCR rate was significantly higher in the CRT arm. R1-resections occurred significantly more often in the CT arm. Median progression-free survival (PFS) (HR 0.919, 95% CI 0.702-1.203, p=0.540) and OS (HR 0.964, 95% CI 0.760-1.225, p=0.766) did not differ significantly in both arms, whereas the PFS rate tended to be higher in the CRT arm after 2 years. OS rates for CRM- R0 surgery with 87.5. ± 0.05% (1y) and 67.2 ± 0.05% (2y) were significantly higher (p<0.01) than for CRM+ R0 surgery with 66.7 ± 0.15% (1y) and 41.2 ± 0.1% (2y) as well as for patients without or incomplete surgery with 68.5 ± 0.03% (1y) and 26.4 ± 0.03% (2y). Conclusions: The addition of radiotherapy after IC improves the R0 CRM - resection and pCR rate without significant change in R0 resection rate (primary endpoint). Pts with R0 CRM - resections had a better prognosis compared to patients with either R0 CRM+ or incomplete or without surgery. However, this effect on resectability did not translate into a statistically significant PFS or OS benefit in the whole cohort. Clinical trial information: NCT01827553. [Table: see text]

Although frail and older pts represent a major fraction of NSCLC pts in routine clinical practice, this population is still underrepresented in clinical trials. The DURATION trial is a prospective, stratified, randomized, multicenter phase II study aiming to provide evidence on first line D after two cycles CTX in this population. Here we report the results for the efficacy endpoints within the first year of follow-up. In this trial (NCT03345810), 201 pts diagnosed with stage IV NSCLC and ≥70 years of age and/or Charlson Comorbidity Index > 1 and/or performance status (PS) ECOG ≥ 2 were stratified according to the Cancer and Age Research Group (CARG) score in "fit" (combination CTX, cCTX) for CARG ≤3 and "unfit" (mono CTX, mCTX). After stratification pts were randomized 1:1 to receive 4 cycles of cCTX (n=49) versus 2 cycles of cCTX and D (cCTX-D, n=48) or 4 cycles of mCTX (n=52) versus 2 cycles of mCTX and D (mCTX-D, n=52). The median age was 76 years (range 56-90), 131 pts (65%) were men, 59 pts (30%) had a PS ECOG ≥2. The PD-L1 expression was 0%, 1-49% and ≥50% in 42%, 51% and 7% of pts, respectively. Compared to the "unfit" group, "fit" pts showed longer progression free survival (PFS, HR 0.54, log-rank p<0.001) and overall survival (OS, HR 0.45, log-rank p<0.001). The 1-year PFS rate was 12% vs. 21% for cCTX and cCTX-D and 3% vs. 12% for mCTX and mCTX-D, while the 1-year OS was 53% vs. 41% for cCTX and cCTX-D and 22% vs. 24% for mCTX and mCTX-D, respectively. The median PFS was 6 and 5 months (mo) for cCTX and cCTX-D (log-rank p= 0.921) and 4 and 3 mo for mCTX and mCTX-D (log-rank p= 0.951), respectively. The median OS was not reached for cCTX and 10 mo for cCTX-D (log-rank p=0.142), while it was 6 and 5 mo for mCTX and mCTX-D (log-rank p=0.995), respectively. cCTX or mCTX according to risk stratification (CARG score) with two cycles followed by D showed an antitumor activity comparable to those of 4 cycles of standard CT in the frontline treatment of older and/or frail stage IV NSCLC pts. The results of the safety primary endpoint for this trial are pending.

PLATON is designed to elevate personalized therapy based on genomic tumor profiles in gastrointestinal cancer patients. Hereby, PLATON’s study-design focuses on the patient’s tumor molecular profiling. Within the network a web application will be developed to link clinical investigators and information on study sites, cancer patients and genetic alteration data, as well as available clinical trials at PLATON`s study sites.

Platinum-fluoropyrimidine-based polychemotherapy is standard for advanced Her-2 negative esophagogastric adenocarcinoma. The optimal duration of treatment is unknown; prolonged application of platinum compounds leads to side effects with questionable overall survival (OS) benefit. MATEO is an international, multicenter, randomized phase II trial exploring the role of fluoropyrimidine maintenance therapy with S-1 (tegafur/gimeracil/oteracil) compared with prolonged polychemotherapy. Patients without progression after 12 weeks of platinum-based induction therapy were randomized in a 2:1 allocation to receive S-1 (Arm A) or to continue with the same polychemotherapy regimen (Arm B) until tumor progression or limiting toxicity. The primary endpoint was OS, secondary endpoints included progression-free survival (PFS), response rate (ORR), safety/toxicity and quality of life. Between November 2014 and April 2019, 217 patients were registered and 165 patients were randomized following 12 weeks of induction therapy. Recruitment was stopped prematurely due to delayed accrual and emerging new therapeutical options. Oxaliplatin-fluoropyrimidine doublet and triplet regimens were the preferred induction regimens (FLO/FOLFOX 47.0%; FLOT 44.7%). Median OS from randomization for patients in Arm A was 13.3 months (95% CI 10.6-15.5) versus 11.4 months (95% CI 9.0-16.5) in Arm B (Hazard Ratio (HR) 1.02 [95% CI 0.70-1.49], p=0.91). Median PFS in Arm A and B was 4.8 months (95% CI 4.0-6.6) and 5.9 months (95% CI 3.8-9.0) respectively (HR 0.93 [95% CI 0.65-1.34], p=0.70). During maintenance phase treatment-related grade 3-5 toxicity was observed in 24.5% (Arm A) versus 27.3% (Arm B), treatment-related peripheral sensory neuropathy (grade 1,2,3) was observed in 11.8%/7.3%/2.7% (Arm A) vs. 7.3%/30.9%/9.1% (Arm B). S-1 as maintenance therapy showed comparable activity in comparison to a prolonged platinum-based therapy with a favorable toxicity pattern. Updated results will be presented at the meeting.

Background: Elderly patients represent a major population of non-small cell lung cancer (NSCLC) patients in routine clinical practice, yet are underrepresented in clinical trials. In particular, data regarding efficacy and safety in frail or elderly patients with respect to immunotherapy is lacking. Importantly, immunosenescence in elderly patients can result in altered activities of immune-modulating drugs such as PD-1/PD-L1 inhibitors. Thus, there is an urgent need to assess safety and efficiency of such inhibitors in this group. Trial design: In this prospective, open label, treatment stratified, and randomized phase II study, 200 patients with stage IV NSCLC, >70 years old and/or "frail" (Charlson Comorbidity Index >1) or restricted performance score (ECOG >1), who are amenable for at least chemotherapy with gemcitabine or vinorelbine, will be included. Patients are stratified after a modified CARG Score (a predicting chemotherapy toxicity score): "Fit" patients are treated with doublet chemotherapy (carboplatin/nab-paclitaxel), while "unfit" patients receive monochemotherapy (gemcitabine or vinorelbine). Patients are then 1:1 randomized and receive either 4 cycles of chemotherapy + follow-up every 8 weeks (Arm A/D) or 2 cycles of chemotherapy followed by 2 cycles of durvalumab and subsequent maintenance durvalumab every 4 weeks (Arm B/C). The primary endpoint is the rate of treatment related grade III/IV adverse events (CTCAE V4.03). As secondary endpoints progression-free survival according to RECIST 1.1, overall survival, descriptive subgroup analyses according to PD-L1 expression, and quality of life will be addressed. Geriatric screening assessments and functional tests will be performed to complete the description of a potential "frail" and "elderly" patient cohort (G8-questionnaire, Timed up & go test, 6MWT). Further, a biomarker profiling program will analyse immune-related effects and potentially identify novel response predictors. The DURATION trial will prospectively investigate the safety and tolerability of PD-L1 treatment with durvalumab after chemotherapy in elderly and frail patients and thereby provide new insights into the effect of PD-L1 blockade and the impact of immunosenescence in this important cohort. Clinical trial identification: EudraCT: 2016-003963-20. Legal entity responsible for the study: AIO-Studien-gGmbH. Funding: AstraZeneca, Celgene. Disclosure: J. Kuon: Honoraria: AstraZeneca, Pfizer. M. Serke: Honoraria: BMS, Celgene, Lilly, Roche. AD boards: AstraZeneca; BMS, Boehringer, Celgene, Hexal, Lilly, Merck, MSD, Pfizer, Roche, Teva, Abbvie, AIO. M. Faehling: Honoraria: AstraZeneca, BMS, MSD, Roche. M. Wermke: Honoraria: BMS, Novartis, Roche, Bayer, Glenmark, AstraZeneca; Travel cost reimbursements: AstraZeneca, BMS, MSC, Novartis, Glenmark; Research funding: Novartis, Pfizer. M. Thomas: Speaker honoraria: Lilly; BMS, MSD, Roche, Pfizer, AstraZeneca; Advisory boards: Lilly, BMS, MSD, Roche, Pfizer, AstraZeneca, Celgene; Mediolanum Scientific projects: AstraZeneca, BMS, Celgene. All other authors have declared no conflict of interest.

Background and aim: In metastatic colorectal cancer (mCRC) available systemic treatment options substantially increased in the last decades. Nowadays, overall survival in mCRC patients ranges from 25 to 35 months as recent studies report. We compared treatment modalities and survival in mCRC patients who were treated at our center in two different periods.Patients and methods: Within two sequential monocentric analyses patients with mCRC treated at our Comprehensive Cancer Center (CCC) between 07/1994 and 10/2007 (cohort 1) and from 11/2007 to 05/2010 (cohort 2) were evaluated for applied treatment, for best response to treatment and for survival (OS). For statistical analysis the Kaplan-Meier estimator was used.Results: Both patient cohorts showed comparable characteristics regarding median age (63 vs. 64 yrs), localization of primary tumor (colon 60% vs. rectum 40%) and number and site of distant metastasis (1 site [75%] vs. >= 2 site [25%]; liver-only metastasis [55%]). About half of all patients in each cohort received at least three consecutive chemotherapy regimens. In cohort 1, treatment mainly consisted of chemotherapy alone (>80%), whereas in cohort 2 chemotherapy was combined with a monoclonal antibody in nearly 70%. Rate of surgical resection of metastasis increased over time (8% vs. 17%). Median OS was 27.3 months (cohort 1) vs. 39.4 months (cohort 2).Conclusion: The increasing availability of effective substances including monoclonal antibodies and individual approaches including secondary surgery of distant metastasis might explain that survival in pts with mCRC has substantially improved over the last decades.

Our study investigated the impact of specific KRAS mutations and BRAF mutation on progression-free survival (PFS) and overall survival (OS) in patients with metastatic colorectal cancer (mCRC) treated within the AIO KRK-0104-trial as first-line therapy. In total, 146 (of 185) patients were included in this analysis. Seventy-nine patients presented with KRAS/BRAF wild-type (wt), 41 patients with a KRAS codon 12 and nine patients with a KRAS codon 13 mutation. Seventeen patients presented a BRAF-mutated tumor. The patients of our study were treated with CAPIRI/CAPOX plus cetuximab. Major differences regarding PFS and OS were observed depending on the mutation of the tumor. PFS was 8 months in patients with wt-tumors, 5.8 months with codon 12-mutated, 9.9 months with codon 13-mutated and 4.2 months with BRAF-mutated tumors. OS was 23.5 months in patients with wt-tumors, 18.9 months with codon 12-mutated, 26.2 months with codon 13-mutated and 13.0 months with BRAF-mutated tumors. Although the conventional separation of patients with KRAS wild-type versus KRAS mutant tumors did not have a significant impact on outcome parameters in the AIO KRK 0104-trial, this analysis demonstrates that markedly differing results are obtained when subtypes of KRAS and BRAF mutation are taken into account.

The AIO KRK-0104 randomised phase II trial investigated the efficacy and safety of two capecitabine-based regimens: combination of capecitabine and irinotecan (CAPIRI) plus cetuximab (CAPIRI-C) and combination of capecitabine with oxaliplatin (CAPOX) plus cetuximab (CAPOX-C) in the first-line treatment of metastatic colorectal cancer (mCRC). Treatment-related skin toxicity (ST) was evaluated separately for capecitabine and cetuximab. The present analysis investigates the correlation of capecitabine-attributed ST (Cape-ST) and parameters of treatment efficacy. Patients with mCRC were randomised to cetuximab (400 mg m−2, day 1, followed by 250 mg m−2 weekly) plus CAPIRI (irinotecan 200 mg m−2, day 1; capecitabine 800 mg m−2, twice daily, days 1–14, every 3 weeks), or cetuximab plus CAPOX (oxaliplatin 130 mg m−2, day 1; capecitabine 1000 mg m−2, twice daily, days 1–14, every 3 weeks). Of 185 recruited patients, 149 (CAPIRI-C, n=78; CAPOX-C, n=71) received study treatment beyond the first tumour assessment and were evaluable for efficacy. Capecitabine-attributed ST, predominantly hand–foot syndrome, was observed in 32.2% of patients. Capecitabine-attributed ST grade 1–3 was associated with a significantly higher disease control rate (DCR) (97.9 vs 86.1%, P=0.038) compared with grade 0 toxicity. Moreover, Cape-ST grade 1–3 related to a markedly longer progression-free survival (PFS) (9.9 vs 5.6 months, P<0.001) and overall survival (OS) (32.8 vs 22.4 months, P=0.008). Separate analyses of treatment arms indicated that the effect of Cape-ST on PFS remained significant for both arms, whereas the effect on OS remained apparent as a strong trend. This analysis supports the hypothesis that for the evaluated regimens, a correlation exists between Cape-ST and treatment efficacy regarding DCR, PFS, and OS.

4067 Background: Response rate captures anti-tumor efficacy and displays the percentage of patients with relevant tumor shrinkage, but does not address the magnitude of response.Therefore, Waterfall plot analyses (WPA) have been used in a single examination for ACRC with conventional chemotherapy (Grothey, ASCO GI ´07). To assess activity of different regimen with ‘targeted‘ drugs, WPA were done on 2 similarly designed rand. 1st line ACRC trials with bevacizumab (B) or cetuximab (C): AIO 0604: XELOX-B vs. XELIRI-B (Schmiegel, ASCO ´07; N=255) and AIO 0104: XELOX-C vs. XELIRI-C (Heinemann, ESMO ´06; N=185). Methods: The percentage change according to RECIST criteria from baseline was evaluated at time of best response. Average reduction in tumor burden (ARTB) was calculated as difference in waterfall area under curve divided by the number of pts. In addition, median reduction in tumor burden (MRTB), which might be preferable as the inherent ceiling effects preclude normal distribution, was calculated. From the 4 arms, pooled analysis was done: XELOX vs. XELIRI (indep. from targeted agent), and B vs. C (indep. from chemotherapy). Results: Data from 199 pts (81% of all pts incl.) of the 0604 trial and from 122 pts (69% of all pts) of the 0104 trial, with measurable disease and at least one confirmation restaging were available. Pooled analysis included all 321 pts: XELOX 156, XELIRI 165; 199 with B, 122 with C. WPA displayed similar tumor growth/ shrinkage rates: 6%/88%(XELOX-B), 8%/87% (XELIRI-B), 4%/93% (XELOX-C), and 8%/86% (XELIRI- C) of pts. Pooled analysis resulted in different ARTB and MRTB (see table). Conclusions: WPA with ARTB and MRTB may provide additional information, compared to rates of tumor reduction. The results suggest a larger benefit for C containing combinations, but should be interpreted with caution due to the non-randomized comparison, limited sample size and un-preplanned character of the analysis. XELIRI XELOX Beva Cetux ARTB (%) 31.2 34.7 29.2 38.9 MRTB (%) 31 33 30 35 p 0.29 0.009 p value: Wilcoxon-Mann-Whitney test, 2-sided Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Expert Testimony Other Remuneration Merck, Pfizer Oncology, Roche, Roche Pharma, sanofi-aventis Merck, Roche, Roche GmbH, sanofi-aventis