Search

Background The combination drug TAS-102 is a novel oral nucleoside antitumor agent containing trifluridine and tipiracil hydrochloride, which prevents the degradation of trifluridine. The global phase III RECOURSE trial (Study of TAS-102 in Patients With Metastatic Colorectal Cancer Refractory to Standard Chemotherapies) demonstrated that TAS-102 prolonged the survival of patients with metastatic colorectal cancer (mCRC) whose disease progressed after standard therapies. TAS-102 was first approved in Japan in March 2014, and little is known about its safety and efficacy in clinical practice, especially for mCRC patients with previous regorafenib treatment. Patients and Methods We investigated the safety and efficacy of TAS-102 monotherapy in clinical practice for patients with mCRC refractory to standard therapies who were treated from May 2014 to January 2015. Results A total of 55 patients received TAS-102. The Eastern Cooperative Oncology Group performance status was 0, 1, and 2 in 41.8%, 47.3%, and 10.9% of patients. Of the 55 patients, 32 (58.2%) had been treated with regorafenib before receiving TAS-102. The median progression-free survival and overall survival was 2.0 months and 5.3 months, respectively. Emergency hospitalization was required for 23.6% of the patients during TAS-102 treatment, although most of the events (76.9%) were disease-related. The most common grade 3 or 4 adverse events were neutropenia (41.8%), leukopenia (27.2%), anemia (23.6%), febrile neutropenia (5.5%), and fatigue (3.6%). The frequency of grade ≥ 3 events was not significantly increased among the patients who had compared with those who had not received regorafenib. The progression-free survival (median 2.1 vs. 2.0 months) and overall survival (median 6.2 vs. 4.7 months) were similar for the 2 subgroups. Conclusion The safety and efficacy of TAS-102 monotherapy in clinical practice were maintained, irrespective of previous regorafenib treatment.

57 Background: FOLFOXIRI plus bevacizumab (BEV) is regarded as the standard of care for selected patients (pts) with metastatic colorectal cancer (mCRC), despite the high incidence of neutropenia and diarrhea. The AXEPT phase III study showed that the modified capecitabine (CAP) + irinotecan (IRI) + BEV (CAPIRI+BEV) [CAP 1600 mg/m2, IRI 200 mg/m2, and BEV 7.5 mg/kg q3wk] treatment was non-inferior to FOLFIRI+BEV, with a lower incidence of hematologic toxicity. We hypothesized that the modified CAPIRI combined with oxaliplatin (OX) and BEV (CAPOXIRI+BEV) would be more feasible than FOLFOXIRI+BEV, without compromising efficacy. Methods: The QUATTRO-II study is an open-label, multicenter, randomized phase II study. In Step 1, the recommended doses (RD) of OX and IRI were investigated as a safety lead-in. In Step 2, pts are randomized to either the RD of CAPOXIRI+BEV or FOLFOXIRI+BEV. In Step 1, four dose levels of CAPOXIRI (fixed dose of CAP 1600 mg/m2 and BEV 7.5 mg/kg plus escalated or de-escalated doses of OX and IRI, q3wk) were investigated in a 3+3 manner. A dose level of ≤ 2/6 of dose-limiting toxicity (DLT) cases was expected as the RD. Results: A total of 9 pts (3 at Level 0, 6 at Level 1) were included in Step 1. The baseline characteristics were as follows: the median age was 62 years; 6 were male; 6 presented with a left-sided tumor; 8 had a performance status of 0; all wild type/ RAS mutant/ BRAF V600E mutant were 8/1/0; and UGT1A1 wild type/*6 single hetero/*28 single hetero were 7/0/2. In Level 0 (IRI 200 mg/m2, OX 100 mg/m2), one grade 4 neutropenia and one grade 3 anorexia were observed, but without DLT. In Level 1 (IRI 200 mg/m2, OX 130 mg/m2), two grade 4 neutropenia and one grade 3 colitis were observed, with 1 DLT (febrile neutropenia) case, fully recovered without G-CSF administration. No treatment-related deaths were observed. Although dose modifications were needed in 4 of the 6 pts, no further safety concerns related to treatment continuity were observed in the 2nd or subsequent cycles. Thus, we determined that the dose administered in Level 1 is the RD for Step 2. According to the preliminary efficacy results at 8 weeks after initiating study treatment, 6 pts achieved a partial response (2 in Level 0 and 4 in Level 1). Conclusions: The RD of CAPOXIRI+BEV was 200 mg/m2 IRI, 130 mg/m2 OX, 1600 mg/m2 CAP, and 7.5mg/kg BEV. The randomized phase II Step (Step 2) of QUATTRO-II is ongoing. Clinical trial information: NCT04097444.

Only a small percentage of patients afflicted with gastric cancer (GC) respond to immune checkpoint blockade (ICB). To study the mechanisms underlying this resistance, we examined the immune landscape of GC. A subset of these tumors was characterized by high frequencies of regulatory T (Treg) cells and low numbers of effector T cells. Genomic analyses revealed that these tumors bore mutations in RHOA that are known to drive tumor progression. RHOA mutations in cancer cells activated the PI3K-AKT-mTOR signaling pathway, increasing production of free fatty acids that are more effectively consumed by Treg cells than effector T cells. RHOA mutant tumors were resistant to PD-1 blockade but responded to combination of PD-1 blockade with inhibitors of the PI3K pathway or therapies targeting Treg cells. We propose that the metabolic advantage conferred by RHOA mutations enables Treg cell accumulation within GC tumors, generating an immunosuppressive TME that underlies resistance to ICB., ファイル公開：2021-07-14

e16550 Background: An anti-PD-1 inhibitor pembrolizumab showed a durable response for patients with microsatellite instability high (MSI-H) or mismatch repair deficient (dMMR) advanced solid tumors. However, clinical and molecular features of responders to PD-1 blockade for MSI-H/dMMR advanced gastrointestinal (GI) tumors are not well understood. Methods: Patients with MSI-H/dMMR GI tumors who received anti‒PD-1 monotherapy were enrolled in this study. PD-L1 expression in tumor cells or immune cells as well as combined positive score (CPS) were assessed using the PDL1 IHC 22C3 pharmDx assay. Cancer genome alterations were evaluated by a next generation sequencing-based panel with Oncomine Comprehensive Assay for tumor tissue samples or Guardant360 ctDNA Assay for plasma samples. Results: A total of 43 patients with gastric cancer (n = 19), colorectal cancer (n = 16), cholangiocarcinoma (n = 4), small intestine cancer (n = 2) and pancreatic cancer (n = 2) were analyzed in this study. All patients had measurable lesions and objective response was observed in 20 patients (46%). Objective response rate (ORR) was significantly higher in patients with ECOG PS of 0 than in those with PS of 1 or 2 (64% vs. 22%, p = 0.01) and PD-L1+ in tumor cells than in those with PD-L1- in tumor cells (67% vs. 32%, p = 0.03). ORR was similar according to PD-L1 CPS; 45% in CPS < 1, 47% in CPS ≥1 and 56% in CPS ≥10, respectively. ORR was numerically lower in patients with PTEN mutation than in those with PTEN wild type (29% vs. 56%). Also, durable response rate (objective response lasting continuously ≥6 months) was significantly lower in patients with PTEN mutation than in those with PTEN wild type (0% vs. 52%, p = 0.03). In the overall population, median progression-free survival (PFS) was 10.0 months (95% CI, 4.0-not reached). In univariate analysis, PS (1 or 2 vs. 0) was the strongest clinical factor associated with PFS (3.9 months vs. not reached, HR 3.9; 95% CI 1.7-9.2, p = 0.002). After adjusted by PS, PTEN mutation was only molecular factor associated with shorter PFS (3.4 vs. 24.5 months, HR 3.0; 95% CI 1.1-8.4, p = 0.03). Conclusions: Some clinical and molecular factors were associated with response to PD-1 blockade for MSI-H/dMMR advanced GI tumors. Updated biomarker analysis with whole exome sequencing and multiplex IHC will be presented.

3637 Background: At ESMO2019, we reported the primary results of a randomized controlled trial (FAEISS study) investigating the efficacy of topical corticosteroid treatment to facial acneiform rash (AR) by EGFR inhibitors comparing groups starting with a very strong topical corticosteroid and the standard weak topical corticosteroid. As an ancillary analysis, we investigated the association between AR and the pre-treatment skin types, as well as between the skin types and therapeutic effects of EGFR inhibitors on the primary disease. Methods: Utilizing pre-treatment clinical photos of the face taken according to the method determined by FAEISS study protocol, we divided the skin types into categories including enlarged pore, oiliness, xerosis, wrinkles, skin color/redness, and allocated the score (1-3) by central review. The severity of AR occurred during the study was graded and was evaluated the association with the specific skin type by Fisher’s exact test. We also investigated the association between the skin types and the best overall response (RECISTv1.1) to EGFR inhibitor therapy on the primary disease using the Cochran-Armitage trend test. Results: Of the registered 172 cases of RAS wild-type metastatic colorectal cancer [104 men and 68 women, median age = 68 (26-79)], omitting the cases with unevaluable data, finally we analyzed 146 cases for associations between the skin types and AR and 147 cases for best overall response. Interestingly, AR developed 13.6% of enlarged pore score 1, 29% of score 2 and 45.8% of score 3, and patients with enlarged pore tended to have more AR (p = 0.058). Surprisingly, the response(CR/PR/SD) of the primary disease were 59.1% of the enlarged pore score 1, 70.6% of score 2 and 87.0% of score 3, and showed statistically significant trend(p < 0.038). Conclusions: This study suggested that a skin type (enlarged pore) is a possible marker predicting AR risk in EGFR inhibitor therapy for RAS wild-type metastatic colorectal cancer, and better therapeutic effects of EGFR inhibitors.

4044 Background: Regulatory T cells (Tregs) potentially induce the resistance of anti-PD1/PD-L1 inhibitors (A-PD1). TAS-116, a novel HSP90 inhibitor, enhanced antitumor immunity via reducing Tregs in vitro and in vivo. Combination of TAS-116 plus A-PD1 showed a superior tumor growth suppression compared with either treatment alone in vivo. Based on the above, we investigated safety and efficacy of TAS-116 in combination with nivolumab in patients with solid tumors. Methods: Enrolled patients received TAS-116 plus nivolumab in a dose-finding part to estimate the maximum tolerated dose and the recommended phase 2 dose (RP2D). Additional patients were enrolled in a dose-expansion part. TAS-116 monotherapy (orally once daily, 80mg on level 1, 120mg on level 2, and 160mg on level 3) was administrated for 2 weeks followed by the combination with nivolumab (intravenously every 2 weeks, 3 mg/kg). The primary endpoint was dose-limiting toxicities (DLTs) during the first cycle (4 weeks). PD-L1 combined positive score (CPS) and tumor mutation burden (TMB) were assessed. We also conducted biomarker research using paired samples from repeated tumor biopsies and blood collections. Results: A total of 44 patients with colorectal cancer (CRC, n = 29), gastric cancer (GC, n = 8), sarcoma (n = 5), non-small cell lung cancer (NSCLC, n = 1) and melanoma (n = 1) after standard of cares were enrolled. One patient had MSI-H CRC, but all other patients had MSS tumors. No DLTs were observed at all levels and TAS-116 160 mg was determined as RP2D. The common grade 3 or worse treatment-related adverse included AST/ALT increased (7%), creatinine increased (5%) and platelet count decreased (5%). Objective tumor response was observed in 6 patients including 4 MSS CRC, 1 MSI-H CRC and 1 sarcoma, resulting in objective response rate (ORR) of 16% in MSS CRC without prior A-PD-1. PD-L1 CPS and TMB could be evaluated in 18 and 17 MSS CRC without prior A-PD-1, respectively. ORR was 27% in patients with CPS ≥1 and 0% in patients with CPS < 1. ORR was 33% with TMB-high (median as the cut-off) and 12% with TMB-low. Analysis of tumor-infiltrating lymphocytes before treatment and after TAS-116 monotherapy demonstrated reduction of FoxP3hiCD45RA−Tregs fraction in the tumor microenvironment. Conclusions: The combination of TAS-116 160mg plus nivolumab had manageable safety profiles and anti-tumor activity especially for MSS CRC patients, which warrants further investigations in a large cohort. Clinical trial information: UMIN000032801 .

322 Background: Nanoparticle albumin-bound (nab)- paclitaxel (PTX) was non-inferiority to solvent-based paclitaxel as second-line for advanced gastric cancer (AGC) with fewer infusion-related reactions and a trend toward improved overall survival (OS) in patients (pts) with peritoneal metastasis (PM) or ascites in ABSOLUTE trial (Shitara K et al. Lancet Gastroenterol Hepatol. 2017). Furthermore, safety and efficacies of nab-PTX plus ramucirumab (RAM) was reported in a phase II trial (Bando H, et al. EJC 2018), although no randomised trial with PTX plus RAM is reported so far. Methods: We retrospectively reviewed consecutive pts with AGC receiving nab-PTX plus RAM or PTX plus RAM as second-line chemotherapy from June 2015 to January 2019 at the National Cancer Center Hospital East. Adverse events were evaluated using Common Terminology Criteria for Adverse Events (CTCAE) v4.0. Progression-free survival (PFS) was calculated using the Kaplan-Meier method, and the differences were evaluated using the log-rank test. Results: A total of 257 pts were included for analysis with 118 pts treated with nab-PTX plus RAM and 139 pts with PTX plus RAM. 151 pts (59%) had peritoneal metastasis and 76 pts (30%) were associated with moderate or massive amounts of ascites. Objective response rates were similar between two groups (nab-PTX plus RAM 34.1% vs. PTX plus RAM 28.0%, p = 0.36). There were no significant differences in PFS (median 3.9 vs. 3.9 months, log-rank p = 0.34; hazard ratio [HR] = 1.14). HR of PFS was 0.96 in pts with PM and 0.79 in pts with moderate or massive ascites. The major grade 3 or higher adverse events were neutropenia (nab-PTX plus RAM 55.1% vs. PTX plus RAM 55.4%), leucopenia (28.8 vs. 35.3%), thrombocytopenia (5.1 vs. 2.9%), and febrile neutropenia (5.1 vs 9.4%). Conclusions: Efficacy and safety of nab-PTX plus RAM were comparable to those of PTX plus RAM in pts with AGC in clinical practice. nab-PTX plus RAM is a treatment option as second-line treatment in pts with AGC.

229 Background: HER2 dual blockade showed the promise in clinical trials in patients (pts) with HER2-positive (HER2-Pos) metastatic colorectal cancer (mCRC). A phase I/II study of a novel anti-HER2 antibody-drug conjugate (HER2-ADC) showed promising activities across different cancer subtypes including HER2-Pos mCRC, and also showed a hint of activity for low-expressing (HER2-L) mCRC, regardless of RAS mutation. However, molecular landscape and prognostic value of HER2-L on mCRC are unclear. Methods: The eligibility included pts with mCRC who had undergone surgical resection of primary tumor. Using the specimen, HER2 expression was evaluated by immunohistochemistry (IHC) and FISH. Definition of HER2 positivity was as follows; HER2-Pos as IHC 3+ or IHC 2+/FISH positive: HER2-L as IHC 2+/FISH negative or IHC 1+: Negative (HER2-Neg) as IHC 0+. NGS tests were also done. Results: Between 2005 and 2015, a total of 370 pts were analyzed; 60% were male, with a median age of 64 y/o. Fifteen pts (4%) with HER2-Pos, 21 (6%) with HER2-L, and 334 (90%) with HER2-Neg were identified. Proportion of left-sided primary tumor was similar among groups ranging from 72% to 78%. No difference was observed in terms of major clinicopathological characteristics. The proportion of co-altered RAS mutations in HER2-L was significantly higher than HER2-Pos ( p = 0.037) and similar to HER2-Neg (27% in HER2-Pos vs. 62% in HER2-L vs. 56% in HER2-Neg). As of median follow-up of 24.8 months, median overall survival (mOS) in HER2-L was significantly better than that in HER2- Pos ( p = 0.029, 18.2 months in HER2-Pos vs. 33.3 in HER2-L vs. 27.9 in HER2-Neg). In 58 pts harboring RAS wild-type who received an anti-EGFR monoclonal antibody therapy, median progression-free survival (mPFS) in HER2-L tended to be better than that in HER2- Pos, with 2.2 months in HER2-Pos, 7.8 in HER2-L, and 5.5 in HER2-Neg ( p = 0.099). Conclusions: Since HER2-L mCRC had the high prevalence of co-altered RAS mutations but showed a better prognosis and might benefit more from an anti-EGFR therapy than HES2-Pos, the HER2-L mCRC seems to have a different biological behavior from HER2-Pos in terms of molecular landscape and prognostic value on mCRC.

374 Background: Pembrolizumab, anti–PD-1 antibody, provides response rates of around 15% in patients (pts) with PD-L1-positive advanced gastric cancer (AGC). Lenvatinib, a multikinase inhibitor of VEGF receptors and other receptor tyrosine kinases, substantially decreased the tumor-associated macrophages and increased infiltration of CD8-positive T cells and enhanced anti-tumor activity of PD-1 inhibitors in vivo model. This phase 2 study has been conducted to evaluate efficacy and safety of the combination of lenvatinib plus pembrolizumab in pts with AGC. Methods: Eligible pts were with AGC having measurable lesions according to RECIST ver. 1.1. Pts could be enrolled regardless of PD-L1 status. Pts received 20 mg oral lenvatinib daily plus 200 mg intravenous pembrolizumab every 3 weeks. Primary endpoint was objective response rate (ORR). Planned sample size was 29 pts based on Simon’s optimal two-stage design with one-sided ɑ = 5% and power = 80%. The threshold and expected ORRs were 10% and 30%. PD-L1 combined positive score (CPS) was assessed using the anti–PD-L1 22C3 antibody. Results: From October 2018 to March 2019, 29 pts (27 MSS and 2 MSI-H) were enrolled and assessed for anti-tumor response. Fourteen pts received the study treatment as first-line and 15 pts as second-line. ORR was 69% (95% CI 49 to 85). The disease control rate was 100%. ORR in MSS pts was 70%. ORR was numerically higher in pts with CPS≥1 (n=19, ORR 84%) than that of pts with CPS

43 Background: Regorafenib (REG) and trifluridine/tipiracil hydrochloride (FTD/TPI) demonstrated overall survival (OS) benefit in patients (pts) with metastatic colorectal cancer (mCRC) in the CORRECT and RECOURSE phase III trials. In Japan, REG and FTD/TPI have been approved in 2013 and 2014, respectively. However, little is known about survival impact on these agents in the real-world setting. Therefore, the aim of this retrospective study was to evaluate the effects of REG and FTD/TPI in pts with mCRC. Methods: We collected medical records from consecutive 1142 pts who had been initiated with first-line chemotherapy for mCRC from 2008 to 2016 at National Cancer Center Hospital East. The survival outcomes were compared between pts from 2008 to 2011 (cohort A) and those from 2012 to 2016 (cohort B). This study excluded pts who have not been refractory or intolerant to standard chemotherapy including fluoropyrimidine, oxaliplatin, irinotecan, and anti-EGFR antibody if KRAS exon 2/ RAS wild-type tumors. Results: A total of 590 pts were analyzed (cohort A; N = 236 and cohort B; N = 354). More patients received at least one of REG or FTD/TPI in cohort B (16.1% vs. 59.9%, p < 0.001). The time from initiation to end of standard chemotherapy was comparable between the two cohort (20.0 vs. 17.5 months, p = 0.266). With a median follow-up period of 34.9 months, salvage-line OS (sOS) after standard chemotherapy was significantly longer in cohort B (4.8 vs. 6.6 months, p = 0.001), while there was only a favorable trend in cohort B in terms of OS from start of first-line treatment (27.3 vs. 28.5 months, p = 0.516). In cohort B, pts who sequentially received both of REG and FTD/TPI showed longest sOS (median, both of REG and FTD/TPI; 11.3 months, either REG or FTD/TPI; 7.0 months, neither REG nor FTD/TPI; 3.1 months). Conclusions: Our study showed that REG and FTD/TPI led to prolongation of sOS in the real-world setting, indicating the importance of strategies which make all active agents available to pts with mCRC.

382 Background: The KEYNOTE-059 study showed the preliminary antitumor activity and tolerability of chemotherapy with pembrolizumab (P) for advanced gastric cancer (AGC). In Japan, S-1 + platinum regimen is a standard chemotherapy for AGC. The KEYNOTE-659 study (NCT03382600) investigated the efficacy and safety of S-1 + oxaliplatin (SOX; cohort 1) or cisplatin (SP; cohort 2) with P as the first line treatment in patients (pts) with human epidermal growth factor receptor 2 (HER2)-negative, programmed death-ligand 1 (PD-L1)-positive AGC. Here, we report the results of cohort 1. Methods: The key inclusion criteria were as follows: age ≥18 to ≤75 years; an ECOG performance status of 0 or 1; and chemotherapy-naïve, HER2-negative and PD-L1-positive AGC. PD-L1 positivity was defined as a combined positive score of ≥1 using the IHC 22C3 PharmDx assay. An S-1 dose of 40-60 mg per dose was orally administered, twice daily, for the first 2 weeks of a 3-week cycle. P (200 mg) and oxaliplatin (OX; 130 mg/m2) were administered on day 1 of each cycle. The primary endpoint was overall response rate (ORR) that was assessed by a blinded independent central review (BICR). The secondary endpoints were progression-free survival (PFS), overall survival (OS), disease control rate (DCR), duration of response (DOR), and safety. Results: From April to September 2018, 54 pts were enrolled at 25 sites in Japan. The median follow-up time was 10.1 months. The median number of P doses and cycles in SOX were 9 (range, 2-18) and 6 (range, 2-13), respectively. The relative dose intensities of S-1 and OX were 73% and 60%, respectively. The ORR and DCR assessed by BICR were 72.2% (95% CI 58.4-83.5) and 96.3% (95% CI 87.3-99.5), respectively. The median PFS was 9.4 months (95% CI 6.6-NR). Median DOR and OS were not reached. Grade ≥3 adverse events (AEs) were reported in 31 pts (57.4%). The most common treatment-related AEs of grade ≥3 were thrombocytopenia (14.8%), neutropenia (13.0%), colitis (7.4%), and adrenal insufficiency (5.6%). There were no treatment-related deaths. Conclusions: This study showed the encouraging efficacy and manageable safety of SOX with P therapy as a first line in pts with HER2-negative, PD-L1-positive AGC. Clinical trial information: NCT03382600.

107 Background: The anti–PD-1 antibody pembrolizumab (P) provides response rates of 28-57% in patients (pts) with MSI-H metastatic colorectal cancer (mCRC) vs 0% in those with non-MSI-H cancers. STAT3 has been previously reported as a potential key driver of immune evasion. This study investigates efficacy and safety for the combination of BBI608 (napabucasin), which blocks phosphorylated STAT3 and downregulates IDO1 and PD-L1, with P, in pts with mCRC. BBI608 480 mg BID with P was determined as the recommended phase II dose in phase I. Methods: Phase II included Cohorts A (MSI-H) and B (non-MSI-H). Pts with mCRC not responding to or intolerant of standard chemotherapies were enrolled. The primary endpoint was immune-related objective response rate (irORR), according to irRECIST. The sample size for Cohort A (10 pts) was derived in an exploratory manner. In Cohort B, assuming null and alternative hypotheses of irORR = 5% and 20% led to an estimated required sample size of 40 pts, with a 1-sided alpha of 5% and power of 90%. Genomic profiles and the consensus molecular subtypes (CMS) of colorectal cancer were determined by whole exome sequencing and RNA sequencing as previously described. Results: From Feb/2017 to Jun/2018, 10 pts were enrolled in Cohort A and 40 in Cohort B. The irORR was 50% (5 of 10 pts) in Cohort A and 10% (95% CI 2.8 to 23.7) (4 of 40 pts) in Cohort B. Of evaluable 19 pts for CMS classification in Cohort B, CMS1, CMS2, CMS3, and CMS4 were detected in 3, 6, 4, and 6 cases, respectively. The irORR was 33% (1 of 3 pts), 0% (0 of 6 pts), 25% (1 of 4 pts), 33% (2 of 6 pts) in CMS1, CMS2, CMS3, and CMS4, respectively. One CMS3 patient with partial response had POLE mutation, while 1 CMS1 and 2 CMS4 pts with partial response did not. The most common grade 3 or higher treatment-related adverse events included fever (10%) in Cohort A, and diarrhea (5%) and appetite loss (7.5%) in Cohort B, without unexpected safety signals. No treatment-related deaths occurred. Conclusions: BBI608 with P showed encouraging anti-tumor activity with acceptable toxicity for non-MSI-H mCRC pts as well as MSI-H mCRC pts. Impact of CMS on the efficacies of this combination warrants further investigation in the additional cohort of this study. Clinical trial information: NCT02851004.

135 Background: In the phase 1 REGONIVO study, regorafenib of 80 mg/day plus nivolumab showed manageable safety profiles and encouraging anti-tumor activity for advanced colorectal cancer (CRC) or gastric cancer (GC) with objective response rate (ORR) of 36% in CRC and 44% in GC (Fukuoka, et al. ASCO 2019). Updated efficacy results are presented. Methods: Enrolled patients (pts) received regorafenib plus nivolumab in a dose-finding phase to estimate the maximum tolerated dose (MTD). Additional pts were enrolled in a dose-expansion phase. Regorafenib of 80 to 160 mg was administered once daily for 21 on 7 days off with nivolumab 3 mg/kg every 2 weeks. The primary endpoint was dose-limiting toxicity (DLT) during cycle one to estimate the MTD and the recommended dose. PD-L1 combined positive score (CPS) was assessed using the anti–PD-L1 28-8 antibody. Tumor mutation burden (TMB) was measured using Oncomine tumor mutation load assay. Results: Fifty pts were enrolled (25 CRC; 25 GC) until October 2018 with median prior treatment line of 3. Efficacy results were updated as of September 1st 2019. One CRC pt was with MSI-high but all other pts were with MSS or MMR-proficient. Among the 20 pts (9 CRC and 11 GC) with objective response (40%), responses are still ongoing in 13 pts (7 CRC and 6 GC) and the median duration of response was not reached (NR). Median progression free survival (PFS) was 7.8 months in CRC (95% CI, 2.8- NR) and 5.5 months (95% CI, 2.6-10.2 months) in GC. One-year PFS rate was 41.7% in CRC and 22.4% in GC. Median overall survival (OS) was not reached in CRC (95% CI, 9.7-NR) and 12.1 months (95% CI, 5.2-NR) in GC. One-year OS rate was 68% in CRC and 55.3% in GC. No significant difference of PFS and ORR was observed in CRC according to PD-L1 and TMB. Conclusions: Encouraging anti-tumor activity of the combination of regorafenib plus nivolumab had been maintained with long-term follow-up. A randomized study for MSS CRC is under planning. Clinical trial information: NCT03406871.

Background Clinicopathological and molecular features of responders to nivolumab for advanced gastric cancer (AGC) are not well understood. Methods Patients (pts) with AGC who were treated with nivolumab after two or more chemotherapy regimens in a single institution from September 2017 to May 2018 were enrolled in this study. PD-L1 expression in tumor cells (TC) and mismatch repair (MMR) were analyzed by immunohistochemistry. Epstein-Barr virus (EBV) was detected by in situ hybridization. Cancer genome alterations were evaluated by a next-generation sequencing-based panel. High tumor mutation burden (TMB) was defined as more than 10 mutations/megabase. Results A total of 80 pts were analyzed in this study. Tumor response was evaluated in 72 pts with measurable lesions and 14 pts (19%) had an objective response. Overall response rate (ORR) was significantly higher in pts with ECOGPS 0 in those with PS 1 or 2, MMR-deficient (MMR-D) in those with MMR-proficient (MMR-P), PD-L1+ in TC in those with PD-L1- in TC and PIK3CA mutation in those with PIK3CA wild-type. ORR was 31% in pts with at least one of the following factors; MMR-D, high TMB, EBV+ and PD-L1+ in TC vs. 0% in those without these factors. Progression-free survival was significantly longer in pts with PS 0 than in those with PS 1 or 2, MMR-D than in those with MMR-P, and PD-L1+ in TC than in those with PD-L1- in TC. Conclusions Some features were associated with favorable response to nivolumab for AGC. Combining these features might be useful to predict efficacy. Electronic supplementary material The online version of this article (10.1186/s40425-019-0514-3) contains supplementary material, which is available to authorized users.

Recently, the U.S. Food and Drug Administration approved pembrolizumab for patients (pts) with PD-L1-positive metastatic gastric cancer (MGC) based on 22C3 immunohistochemistry (IHC) assay. However, little is known about detailed clinicopathological features of 22C3 PD-L1 expression in MGC. Pts with histologically confirmed MGC were eligible for this prospective observational study. PD-L1 expression (22C3) on tumor cell (TC) or immune cell (IC) and mismatch repair (MMR) were analyzed by IHC. Epstein–Barr virus (EBV) was detected by in situ hybridization. The expressions of tyrosine kinase receptors (RTKs) and cancer genome alterations were evaluated by IHC or next-generation sequencing. A total of 225 pts were analyzed in this study. PD-L1 expression on TC, PD-L1 on IC, MMR-deficient (D-MMR), and EBV positivity were identified in 8.4, 65.3, 6.2, and 6.2% cases, respectively. PD-L1 expression in TC was more frequently observed in pts with D-MMR (P

Background High deoxyuridine triphosphatase (dUTPase) in tumor tissue, as a gatekeeper enzyme for 5-fluorouracil (5-FU), is associated with 5-FU resistance. TAS-114 is an oral dUTPase inhibitor which enhance antitumor activity with 5-FU or fluoropyrimidines. Phase I study of TAS-114 in combination with S-1 showed its tolerability and preliminary antitumor signals for patients (pts) with non-small cell lung cancer and advanced gastric cancer (AGC). This phase II study has been conducted to evaluate efficacy and safety of TAS-114 and S-1 combination in pts with AGC. Methods The main eligibility criteria is pts with AGC after two or more previous chemotherapy regimens containing fluoropyrimidines, platinum agents, and taxanes or irinotecan. The primary endpoint is objective response rate (ORR) by investigators’ judgement. Using Simon’s optimal two-stage design, 29 pts were required, with one-sided a = 5% and power=80%. The threshold and expected ORRs were 5% and 25%. Protein expression levels of dUTPase and BRCA1 in tumor samples were determined by immunohistochemistry. Results The accrual was terminated in June 2018. Of all 20 enrolled pts, the ORR and disease control rate were 5.0% (95% confidence interval [CI], 0.1-24.9%) and 70.0% (95% CI, 45.7-88.1%), respectively. The median progression-free survival (PFS) and overall survival were 2.4 months (95% CI, 1.2-3.3 months) and 7.1 months (95% CI, 5.2-9.4 months), respectively. The median PFS in the groups with high and low dUTPase protein expression in the cytoplasm was 2.8 months (95% CI, 1.4–3.9) and 1.6 months (95% CI, 0.6–2.4), respectively (hazard ratio, 0.40 [95% CI, 0.16–1.04], P = 0.047). Grade 3 or higher treatment-related adverse events were anemia (20%), leucopenia (15%), neutropenia (10%), rash (10%), thrombocytopenia (5%), and Iymphopenia (5%). Conclusions TAS-114 with S-1 showed modest antitumor activity with acceptable safety profiles for heavily pretreated pts with AGC. Legal entity responsible for the study The authors. Funding Taiho. Disclosure D. Takahari: Research grant / Funding (self): Taiho Pharmaceutical Co; Research grant / Funding (self): Ono Pharmaceutical Co; Honoraria (self): Eli Lilly Japan; Honoraria (self): Bristol-Myers Squibb; Honoraria (self): Yakult Honsha Co; Honoraria (self): Chugai Pharmaceutical Co., Ltd. A. Kawazoe: Honoraria (self), Research grant / Funding (institution): Taiho. K. Shitara: Honoraria (self): Novartis; Honoraria (self): AbbVie; Honoraria (self): Yakult; Advisory / Consultancy: Astellas Pharma; Advisory / Consultancy: Lilly; Advisory / Consultancy: Bristol-Myers Squibb; Advisory / Consultancy: Takeda; Advisory / Consultancy: Pfizer; Advisory / Consultancy: Ono Pharmaceutical; Advisory / Consultancy: MSD; Research grant / Funding (institution): Astellas Pharma; Research grant / Funding (self): Lilly; Research grant / Funding (institution): Ono Pharmaceutical; Research grant / Funding (institution): Dainippon Sumitomo Pharma; Research grant / Funding (institution): Daiichi Sankyo; Research grant / Funding (institution): Taiho Pharmaceutical; Research grant / Funding (institution): Chugai Pharma; Research grant / Funding (institution): MSD; Research grant / Funding (institution): Medi Science. All other authors have declared no conflicts of interest.

Background Molecular subtypes including mismatch repair-deficient (MMR-D) affect the efficacy of immune checkpoint inhibitors in patients (pts) with advanced gastric cancer (AGC). However, the association between the subtypes and efficacy of standard chemotherapy remains unclear. Methods AGC pts received chemotherapy from October 2015 to July 2018 with available molecular features were analyzed in this study. We investigated the efficacy of first- (fluoropyrimidine + platinum ± trastuzumab) or second-line (taxanes ± ramucirumab or irinotecan) chemotherapy according to four subtypes: MMR-D, EBV+, HER2+ and others (all-negative). Results Total 410 pts were analyzed: MMR-D, n = 24 (5.9%); EBV+, n = 16 (3.9%); HER2+, n = 56 (13.7%); all-negative, n = 314 (76.6%) with no overlapping between molecular subgroups. Among 285 pts receiving standard first-line chemotherapy, the median PFS were 4.2, 6.0, 7.5, 7.6 months and the ORR were 31, 62, 62, 51% in MMR-D, EBV+, HER2+, and all-negative groups, respectively. Multivariate analysis showed shorter PFS in MMR-D group vs. all-negative (HR 1.98, 95% confidence interval, 1.13-3.47; p = 0.016), while other comparison between subgroups did not show significant difference. In 262 pts receiving second-line chemotherapy, there was no significant difference in efficacy among the four subtypes: the median PFS were 3.9, 6.6, 3.7, 3.9 months and the ORR were 37, 43, 14, 32% in MMR-D, EBV+, HER2+ and all-negative groups, respectively. Twelve MMR-D pts received subsequent immune checkpoint inhibitors, which showed higher ORR (64%) and longer PFS (median PFS: 13.0 months) compared with earlier line chemotherapy. Conclusion MMR-D might result in shorter PFS with first-line chemotherapy for AGC compared with other subgroups. Meanwhile immune checkpoint inhibitors achieved a longer PFS than standard chemotherapy in most MMR-D pts, supporting earlier use of immune checkpoint inhibitors.

To determine the cumulative incidence, disease-specific mortality, and all-cause mortality of pancreatic cancer (PC) in patients with intraductal papillary mucinous neoplasms (IPMNs) and to identify imaging findings that are associated with these outcomes.This retrospective study had institutional review board approval, and the need to obtain patient consent was waived. Data from an electronic database were analyzed and supplemented by chart reviews for 285 patients with nonresected IPMNs who were periodically followed up with imaging (1273 multidetector computed tomography and 750 magnetic resonance cholangiopancreatography examinations). The Kaplan-Meier method was used to estimate the cumulative development of PC, PC mortality, and all-cause mortality (factors were compared by using the log-rank test).Over a median imaging follow-up period of 39 months, 12 (4.2%) of 285 patients developed PC; the cumulative 5-year PC incidence was 3.9% for branch duct (BD)-IPMNs, 45.5% for main duct (MD)-IPMNs (P.01), 7.7% for cysts 30 mm or larger, and 5.3% for cysts smaller than 30 mm (P = .82). Over a median survival follow-up period of 47.5 months, seven (2.5%) of 285 patients died of PC and 14 (4.9%) patients died of other causes. Cumulative 5-year PC mortality was 2.1% for BD-IPMNs, 18.5% for MD-IPMNs (P.01), 2.6% for cysts 30 mm or larger, and 2.8% for cysts smaller than 30 mm (P = .90). Cumulative 5-year all-cause mortality was 5.5% for BD-IPMNs, 18.5% for MD-IPMNs (P.01), 12.5% for cysts 30 mm or larger, and 5.9% for cysts smaller than 30 mm (P = .89).Five-year PC development, disease-specific mortality, and all-cause mortality were approximately 4%, 2%, and 6% for BD-IPMNs and 46%, 19%, and 19% for MD-IPMNs, respectively. The presence of an MD-IPMN, but not cyst size, was significantly associated with PC development and subsequent mortality.

Background Ramucirumab has recently proved to be effective for advanced or recurrent gastric cancer (AGC). Ascites and peritoneal metastasis are among the most common complications of AGC. However, there are few data on the safety and efficacy of paclitaxel plus ramucirumab in patients with AGC with ascites. The purpose of this retrospective study was to evaluate the safety and efficacy of paclitaxel plus ramucirumab in patients with AGC with ascites. Methods We retrospectively evaluated the safety and efficacy of paclitaxel plus ramucirumab in patients with AGC with ascites in comparison with patients without ascites in a single institution from June 2015 to May 2016. The median progression-free survival (PFS) and overall survival (OS) were calculated using the Kaplan-Meier method, and differences evaluated using the Log-lank test. The differences in baseline characteristics and response rates of each ascites group were calculated for homogeneity by chi-square tests and for trends by Fisher’s exact test. Results Eighty-three patients were analyzed in this study. Ascites was detected in 40 patients, 26 patients (31%) had small to moderate ascites and 14 (17%) had massive ascites. The proportion of patients who started with a reduced dose of paclitaxel was higher for patients with massive ascites than others. The frequencies of any grade 3 or 4 hematological toxicity were 51% in patients without ascites, 77% in patients with small to moderate ascites, and 71% in patients with massive ascites. The frequencies of common ramucirumab-related adverse events were also not significantly different among ascites groups, however one patient had a tumor hemorrhage, and one patient had a gastrointestinal perforation. PFS and OS were shorter in patients with massive ascites than in patients with small or moderate ascites or patients without ascites. Conclusions The use of paclitaxel and ramucirumab in patients with AGC with large amounts of ascites was tolerable with adequate dose modification. However, we should pay attention to the risks of ramucirumab-related toxicity in patients with bleeding tumors or intestinal stenosis.

It has been reported that favorable influences of cytotoxic agents to anti-tumor immune response included immunogenic cell death and suppression of regulatory T cell and myeloid-derived suppressor cell. Some clinical trials showed that the addition of immune checkpoint inhibitor to standard chemotherapy improved efficacy in patients with non-small-cell lung cancer or malignant melanoma in first-line settings. Phase III trials of the combination of immune checkpoint inhibitor and chemotherapy in several malignancies are ongoing.

Background: The use of nivolumab or irinotecan as the third-line treatment for patients with advanced gastric cancer (AGC) remains controversial. Methods: This study analyzed patients with AGC treated with nivolumab or irinotecan (nivolumab group or irinotecan group, respectively) from May 2016 to April 2019 following two or more previous lines of chemotherapy. Univariate survival analysis was conducted to identify the clinical and molecular factors associated with progression-free survival (PFS). Results: A total of 156 patients (74 treated with nivolumab and 82 treated with irinotecan) were analyzed. The median PFS was 1.9 months in both treatment groups. The median overall survival (OS) was 7.2 and 6.2 months in the nivolumab and irinotecan groups, respectively. Eastern Cooperative Oncology Group performance status of 1 or more, liver metastasis, a large tumor size at baseline, and HER2-positive status were associated with a worse PFS in the nivolumab group compared with the irinotecan group. The nivolumab group showed a significantly longer PFS (median 3.1 versus 2.0 months) and OS (median 12.9 versus 7.8 months) than the irinotecan group in patients with 0 or 1 of these factors, whereas the irinotecan group showed a significantly longer PFS (median 1.0 versus 1.8 months) and a trend of longer OS (median 3.9 versus 6.1 months) in patients with ⩾2 of these factors. Conclusions: Some clinical and molecular factors were associated with outcomes following nivolumab or irinotecan as the third- or later-line treatment in patients with AGC. These factors must be considered while selecting an optimal treatment option.

Background Surgery is recommended for patients with high-risk submucosal invasive rectal cancer (SM-RC) after local resection. However, surgery affect patient’s quality of life due to stoma placement and impaired anal function. Thus, alternative treatments to prevent local metastasis are anticipated. This study assessed the safety of adjuvant chemoradiotherapy with capecitabine for patients with high-risk SM-RC after local resection. Methods This single-arm, multicenter, phase II trial enrolled patients with high-risk SM-RC who underwent local resection within 12 weeks prior to enrollment. High-risk SM-RC was defined as the presence of at least one of the following factors: poor differentiation of adenocarcinoma, ≥ 1mm of submucosal invasion, presence of lymphovascular invasion, and grade 2 or 3 of tumor budding. Protocol treatment included 45.0 Gy radiotherapy with conventional fractionation and 825 mg/m2 capecitabine administered twice daily until the completion of radiotherapy. The primary endpoint was treatment completion rate, with an expected rate of 95% and a threshold of 80%. This study was registered with the University Hospital Medical Information Network, number UMIN000016785. Results A total of 29 patients from six institutions were enrolled between May 2015 and February 2018. One patient was ineligible. Twenty-three patients completed treatment with a completion rate of 82% (80% confidence interval; range, 69%–91%). The remaining five patients also completed treatment with protocol deviation from the planned treatment schedule. The mean relative dose intensity of capecitabine was 89% (range, 58%–100%). The most common adverse events were radiation dermatitis (54%), anal pain (39%), and anal mucositis (29%). No grade 3 or higher adverse events were reported. Conclusions Adjuvant chemoradiotherapy using capecitabine demonstrated manageable safety in patients with high-risk SM-RC after local resection. Clinical trial identification UMIN000016785. Legal entity responsible for the study The authors. Funding The National Cancer Center Research and Development Fund (25‐A‐12) to Dr Saito. Disclosure All authors have declared no conflicts of interest.

Background Although several studies suggested that immune checkpoint inhibitors may enhance an efficacy of radiotherapy (RT), it remains unclear in patients (pts) with advanced gastric cancer (AGC). Methods Total 34 AGC pts treated with RT for the primary gastric lesion from April 2013 to March 2019 were analyzed in this study. Pts were divided into two groups according to whether they had received nivolumab before RT or not. Primary tumor responses were assessed according to volumetry on CT scan with tumor response as > 70% decrease in tumor volume. Endoscopic responses were evaluated in pts who underwent endoscopy after RT. Tumor infiltrating lymphocytes (TILs) were also analyzed to examine the tumor microenvironmental immune status in the pts. Results Among 34 pts, 16 pts had received nivolumab before RT. There were no significant differences in baseline characteristics between nivolumab-exposed and nivolumab-naive groups. 70% tumor reductions by RT were observed in 25% (4/16) in nivolumab-exposed group but no patients (0/18) in nivolumab-naive group (P = 0.039). Among pts who received endoscopy after RT, four of seven pts in nivolumab-exposed group experienced partial response, while no response was observed among 10 nivolumab-naive pts (P = 0.015). TILs were evaluated in pts whose tumor samples were available. CD8+ T cells/effector regulatory T cells (eTreg cells) ratio in TILs was markedly increased after nivolumab in pts who achieved response, but not observed in pts who had no tumor reduction. A significant increase of CD8+ T cells/eTreg cells ratio after nivolumab treatment was observed in nivolumab-exposed pts (n = 15), whereas such a tendency was not observed in nivolumab-naive pts (n = 10). Conclusions RT to primary gastric cancer significantly led to tumor response in nivolumab-exposed pts compared with nivolumab-naive pts. Immune profiling suggested that nivolumab might enhance the sensitivity to RT by immunoactivation via increasing CD8+ T cell/eTreg cell ratio.

Background Nivolumab (NIVO) showed a survival benefit compared to placebo in patients (pts) with advanced gastric cancer (AGC) after two or more previous lines of chemotherapy. It has been reported that prior treatment with immune check point inhibitors might improve responses to salvage chemotherapy without unexpected safety signals in non-small cell lung cancer. However, safety and efficacy of irinotecan (IRI) administered after NIVO in pts with AGC remains unclear. Methods Pts with AGC treated with IRI after NIVO (cohort A) or IRI before NIVO (cohort B) from September 2017 to December 2018 were analyzed in this study. Clinical outcomes after IRI were compared between two groups. Results A total of 44 pts (14 in cohort A and 31 in cohort B) were included in this study. Common grade 3 or worse adverse events in cohort A were leukocytopenia (21%), neutropenia (14%), thrombocytopenia (7%) and diarrhea (7%), which were not significantly different in cohort B. No immune related adverse events occurred in cohort A. The ORR-IRI was 21% in cohort A and 16% in cohort B (p = 0.48), and the median PFS-IRI was 2.67 and 2.37 months (HR = 0.87, p = 0.68), respectively. In cohort A, the PFS of IRI was longer than that of NIVO in 9 of 14 pts (64%). Conclusions IRI administered after NIVO was manageable without unexpected adverse events. Although prior NIVO did not seem to improve responses to IRI, it was notable that IRI showed a longer PFS than that of prior NIVO in a substantial number of pts.

Background Nivolumab showed a survival benefit compared to placebo in patients (pts) with advanced gastric cancer (AGC) after two or more previous lines of chemotherapy. However, it remains unclear whether to use nivolumab or irinotecan in AGC pts as a third- or later-line treatment. Meanwhile, ECOG PS or molecular characteristics are associated with outcomes after PD1 blockade for AGC (Shitara K, et al, Lancet 2018; Mishima S, et al, JITC 2019). Methods AGC Pts treated with nivolumab or irinotecan after fluoropyrimidine based regimens and taxanes from May 2016 to November 2018 were analyzed. Subgroup analysis was conducted according to ECOG PS or other clinicopathological characteristics. Results Total 128 pts (58 in the nivolumab group and 70 in the irinotecan group) were included in this study. The median PFS were both 1.9 months (HR 0.85; 95% CI, 0.58-1.27). Nivolumab showed a longer PFS than that of irinotecan in pts with ECOG PS of 0 (N = 66), with median PFS of 3.1 vs. 2.0 months (HR, 0.53; 95% CI, 0.30-0.94), while irinotecan showed a trend of longer PFS in pts with ECOG PS of 1 or more (N = 62; median 1.1 vs. 1.8 months; HR 1.51; 95% CI, 0.87-2.63; p for interaction = 0.009). Interaction between treatment effect and ECOG PS was consistent if it was limited to microsatellite stable AGC. Conclusions ECOG PS was associated with outcomes after nivolumab or irinotecan in AGC pts as a third- or later-line, which warrants further evaluation in a larger cohort. Biomarker analysis will be presented at the meeting.

2522 Background: Immune suppressive cells such as regulatory T cells (Tregs) or tumor-associated macrophages (TAMs) may contribute to resistance to anti-PD-1/PD-L1 inhibitors (A-PD1). Regorafenib, a potent inhibitor of angiogenic and oncogenic kinases, reduced TAMs in tumor models. The combination of regorafenib plus A-PD1 exhibited superior tumor growth suppression compared to either treatment alone in murine models. Methods: In this study, we enrolled patients (pts) with previously treated, advanced GC or CRC. The pts received regorafenib plus nivolumab in a dose-finding phase to estimate the maximum tolerated dose (MTD). Additional pts were enrolled in a dose-expansion phase to further establish the safety and determine the preliminary efficacy. Regorafenib of 80 to 160 mg was administered once daily for 21 on 7 days off with intravenous nivolumab 3 mg/kg every 2 weeks. The primary endpoint was dose-limiting toxicity (DLT) during cycle one (4 weeks) to estimate the MTD and the recommended dose. Results: Fifty pts were enrolled (25 GC; 25 CRC) until October 2018. The median prior treatment line was 3 (range 2-8). During dose-escalation, 3 DLTs were observed with regorafenib 160 mg, including grade (G) 3 maculopapular rash, mucositis and proteinuria, while there was no DLT with 80 or 120 mg. In the dose expansion cohort with regorafenib 120 mg, the dose was reduced to 80 mg owing to frequent G3 skin toxicities. Grade ≥ 3 treatment related adverse events occurred in 17 pts; the common events ( > 5%) being rash (14%), palmar-plantar erythrodysesthesia (10%), and proteinuria (8%). Objective tumor response was observed in 19 pts (38%) including 11 MSS GC, 7 MSS CRC and 1 MSI-H CRC for response rates of 44% in GC and 29% in MSS CRC. Three of the 7 A-PD1 pretreated GC pts achieved a partial response. The pre- and post-treatment tumor samples showed a reduction of FoxP3hiCD45RA-Tregs fraction at the tumor response. Conclusions: The combination of regorafenib 80mg plus nivolumab had a manageable safety profiles and encouraging anti-tumor activity in MSS GC and CRC pts, which warrants further investigations in a larger cohort. Updated biomarker analysis will be presented. Clinical trial information: NCT03406871.

Recent guidelines suggest that imaging surveillance be conducted for 5 years for patients with at most one high-risk feature. If there were no significant changes, surveillance is stopped. We sought to validate this follow-up strategy. In study 1, data were analysed for 392 patients with intraductal papillary mucinous neoplasms (IPMNs) and at most one high-risk feature who were periodically followed up for more than 1 year with imaging tests. In study 2, data were analysed for 159 IPMN patients without worsening high-risk features after 5 years (stop surveillance group). In study 1, pancreatic cancer (PC) was identified in 12 patients (27.3%) in the endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) indication group and none in the non-EUS-FNA indication group (P

Antibodies against programmed death 1 (PD-1) and its ligand programmed death ligand 1 (PD-L1) have recently demonstrated promising results in gastric cancer (GC). PD-L1 expression, the presence of tumor-infiltrating lymphocytes (TILs), and mismatch repair (MMR) deficiency have been proposed as predictive biomarkers for anti-PD-1/PD-L1 antibodies. The aim of this study was to investigate the clinical relevance of PD-L1 expression with TIL, MMR, and Epstein–Barr virus (EBV) status in GC. We performed a tissue microarray analysis in 487 advanced GC patients who underwent gastrectomy. PD-L1 expression on tumor cells (TCs) and tumor-infiltrating immune cells (TIICs), the densities of TILs, and MMR status were evaluated by immunohistochemistry. EBV was detected by in situ hybridization. PD-L1 expression on TCs and TIICs, MMR deficiency, and EBV positivity were identified in 22.8, 61.4, 5.1, and 5.1 % cases respectively. PD-L1 expression was more frequently observed in the elderly (TCs P = 0.002), in males (TCs P = 0.029; TIICs P = 0.043), in patients with poorly differentiated adenocarcinoma with solid-type histological features (TCs P

4106Background: Hyperprogressive disease (HPD) during PD-1 blockade reported in several cancer types is a big hurdle for applying cancer immunotherapy, whereas there is no report in gastric cancer ...

3011Background: Significant advances have been made in cancer immunotherapy with immune check point blockade (ICB). Yet, most patients with various tumor types finally experienced disease progressi...

98 Background: Ramucirumab (RAM), a monoclonal antibody against VEGFR2, was recently approved for the treatment of advanced gastric cancer (GC). However, little is known about the immunological effects in advanced GC patients. Methods: Patients with advanced GC who had been planned to receive chemotherapy containing RAM were prospectively enrolled for this study from January 2016 to September 2016. Paired samples were obtained from peripheral blood mononuclear cells (PBMCs) and tumor infiltrating lymphocytes (TILs) in primary tumors pre- and post-RAM therapy to investigate immune profiles. Results: A total of 59 PBMCs and TILs from 18 patients were collected. Higher frequencies of FOXP3highCD45RA-CD4+ T cells (effector regulatory T cells: eTregs) were detected in TILs compared with PBMCs (19.05±10.48% vs 1.89±1.0%, P < 0.0001), and the presence of high eTreg in TILs was not reflected in PBMCs. eTregs of TILs, but not of PBMC were significantly decreased after RAM-containing therapies (22.67±11.19% vs. 16.33±8.44%, P = 0.034). Expressions of immune checkpoint (IC) molecules (PD-1, CTLA-4, LAG-3, and ICOS) were much higher in TILs than those in PBMCs, and all these molecules exhibited higher expressions on eTregs than on CD8+ T cells. Among them, ICOS was specifically expressed by eTregs in TILs. PD-1+CD8+ T cells in TILs were significantly decreased after RAM-containing therapies (54.99±19.06% vs. 39.23±17.15%, P = 0.0005), but not in PBMC. Patients with partial responses had significantly higher frequency of eTreg in pre-treatment TIL than those with progression disease (32.56±12.11% vs. 14.83±3.18%, P = 0.036). Furthermore, patients with high eTreg frequency had significantly longer progression-free survival than those with low frequency (161 days vs. 76 days, P = 0.024). Conclusions: There was significant difference in immune profile between PBMC and TIL in GC patients. eTregs and PD-1 expression on CD8+T cells in TILs, not in PBMCs were decreased after RAM-containing therapies. This observation suggests the importance of TIL analyses and might be a rationale to use RAM as an immunomodulator in combination with IC inhibitors.

157 Background: Gastric neuroendocrine carcinoma (G-NEC) is associated with poor prognosis. Most patients (pts) present with advanced disease and have short survival. Ramucirumab (RAM), a monoclonal antibody for VEGFR-2, has been effective for advanced gastric adenocarcinoma (AC). However, little is known about efficacy of RAM containing chemotherapy in G-NEC. Methods: We retrospectively analyzed and compared the efficacy and safety of RAM containing chemotherapy in pretreated G-NEC and AC pts in our hospital (March 2015- January 2017). G-NEC was defined by NEC features, regardless of the proportion, based on histology and neuroendocrine markers (synaptophysin, chromogranin A or CD56). VEGFR-2 expression in tumor vessels was evaluated in archival primary G-NEC tissues by immunohistochemistry using the same anti-VEGFR2 primary antibody and scoring scheme (vascular VEGFR2 H-score) as in REGARD trial (Fuchs CS, et al. BJC. 2016). Results: Seventeen G-NEC and 173 AC pts were analyzed. Therapy for G-NEC included RAM + paclitaxel (n = 13), RAM + irinotecan (n = 2), and RAM monotherapy (n = 2). The overall response rate was significantly better in pts with G-NEC than AC (59 vs. 22%, p = 0.001). Progression-free survival was significantly longer in pts with G-NEC than AC (median 6.8 vs. 3.2 months, HR 0.38 [95%CI 0.22-0.66]; p < 0.001). Overall survival was significantly longer in pts with G-NEC than AC (median 16.2 vs. 9.6 months, HR 0.4 [95%CI 0.19-0.86]; p = 0.014). No severe or unexpected adverse event occurred. Median vascular VEGFR2 H-score, based on available G-NEC tissues from 12 pts, was 220 (range 150-260) which was markedly higher than that reported on AC tissues from REGARD trial as historical control (median 35, range 0-240). Conclusions: Compared to AC patients, RAM containing chemotherapy showed a promising activity without severe or unexpected safety profile in G-NEC patients. This may in part be explained by higher vascular VEGFR-2 expression in G-NEC tissues. These data, though preliminary, support further clinical evaluation of RAM and biomarker investigation in this subset of gastric cancer.

TPS885 Background: Antiangiogenic treatments are a proven useful tool to improve clinical outcomes in patients (pts) with metastatic colorectal cancer (mCRC). Recently, higher levels of vascular endothelial growth factor-D (VEGF-D) are a potential predictive biomarker for ramucirumab efficacy on overall and progression-free survival in mCRC (Tabernero J, et al. ESMO 2017). However, there are limited data on mCRC associating the efficacy of antiangiogenic therapy with angiogenesis-related mediators, such as VEGF family members and their receptors, and with dynamic transition among these mediators during clinical courses. In addition, immune-related factors such as interferon gamma (IFNγ) and transforming growth factor β1 (TGFβ1) may be associated with the regulation of angiogenesis. Methods: This prospective longitudinal study aims to investigate the association between plasma angiogenesis-related mediators and clinical outcomes in mCRC. The key eligibility criteria include pts with mCRC who will receive one of the following regimens: first-line chemotherapy plus bevacizumab, first-line chemotherapy plus anti-EGFR antibody (either cetuximab or panitumumab), second-line FOLFIRI plus ramucirumab, second-line FOLFIRI plus aflibercept, or second-line chemotherapy plus bevacizumab. Plasma is collected twice in pairs from all pts pre- and post-treatment. Comprehensive measurements of plasma angiogenesis-related mediators, such as placental growth factor (PlGF), hepatocyte growth factor (HGF), IL-6, IL-8, angiopoietin-2, neuropillin-1, tissue inhibitor of metalloproteinase-1 (TIMP-1), soluble intercellular adhesion molecule-1 (sICAM-1), soluble vascular cell adhesion molecule-1 (sVCAM-1), thrombospondin-2 (TSP-2), osteopontin (OPN), sVEGFR-1, sVEGFR-2, sVEGFR-3, VEGF-A, VEGF-D, as well as key targets associated with immunotherapy such as IFNγ and TGFβ1 are analyzed in parallel by the multiplex assay with Luminex® technology. The target sample size is 1,000. This study was initiated in September 2017. Clinical trial information: UMIN000028616.

865 Background: Trifluridine/tipiracil (TAS-102) is an oral combination therapy approved for the treatment of patients with metastatic colorectal cancer (mCRC). Phase 1/2 C-TASK FORCE study of TAS-102 plus bevacizumab for patients with refractory mCRC demonstrated a promising efficacy results with mild toxicity profile (Kuboki Y, et al. Lancet Oncology, 2017). The retrospective single-institutional study is aiming to investigate safety of TAS-102 plus bevacizumab for patients with refractory mCRC in real-world clinical practice. Methods: MCRC patients who were refractory or intolerant to a fluoropyrimidine, irinotecan, oxaliplatin, anti-VEGF therapy, and anti-EGFR therapy (if wild-type RAS) and consecutively received TAS-102 plus bevacizumab between January 2016 and June 2017, were analyzed. Adverse events (AEs) were assessed by the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 4.0. Results: A total of 33 patients were identified for safety analysis. Patient characteristics were as follows; male/female, 19/14; median age, 59.0 (range 38-81); ECOG Performance Status of 0/1, 21/12; and All wild/ RAS mutant/ BRAF mutant/ unknown, 16/15/1/1. Six patients had received prior regorafenib. The relative dose intensities of TAS-102 and bevacizumab were 87% and 79%, respectively. Grade 3 or higher hematological and non-hematological AEs were follows; leucopenia, 33%; neutropenia, 42%; anemia, 9%; and proteinuria, 12%. There were no febrile neutropenia or treatment-related death. G-CSF was given in 8 patients (24%), while intravenous antibiotics was in one patient. There were four patients with emergency admission, and out of these, the two patients had intestinal perforation, and perianal abscess, which were regarded as treatment-associated; ultimately these AEs were fully recovered. Conclusions: Our findings suggested safety for TAS-102 plus bevacizumab combination in real-world clinical practice is consistent with those in clinical trial setting.

Background The mutation in KRAS exon 2 is a validated biomarker of resistance to anti-epidermal growth factor receptor (EGFR) therapy in metastatic colorectal cancer (mCRC). Several reports have confirmed associations of other RAS mutations with resistance to anti-EGFR therapy. However, the impact of BRAF and PIK3CA mutations on the efficacy of anti-EGFR therapy remains controversial. Little is known about the frequencies and clinicopathological features of these mutations, as well as the therapeutic effects of anti-EGFR therapy in mCRC patients with these mutations, especially in the Asian population. Methods In this retrospective observational study, frequencies and clinicopathological features of KRAS, NRAS, BRAF and PIK3CA mutations were evaluated in patients with mCRC. Among patients treated with anti-EGFR therapy, objective response, progression-free survival (PFS), and overall survival (OS) were evaluated according to gene status. Results Among 264 patients, mutations in KRAS exon 2, KRAS exons 3 or 4, NRAS, BRAF and PIK3CA were detected in 34.1%, 3.8%, 4.2%, 5.4% and 6.4%, respectively. Thus, a total of 12.1% of patients without KRAS exon 2 mutations had other RAS mutations. Primary rectal tumors tended to be more frequently observed in RAS mutant tumors. BRAF mutations were more frequently observed with right-sided colon, poorly differentiated or mucinous adenocarcinoma, and peritoneal metastasis. Among the 66 patients with KRAS exon 2 wild-type tumors treated with anti-EGFR agents, PFS (5.8 vs. 2.2 months) and OS (17.7 vs. 5.2 months) were significantly better in patients with all wild-type tumors (n = 56) than in those with any of the mutations (n = 10). The response rate also tended to be better with all wild-type tumors (26.8 vs. 0%). Conclusion Other RAS and BRAF mutations were observed in KRAS exon 2 wild-type tumors, which were associated with some clinicopathological features and resistance to anti-EGFR therapy in our patient cohort. Electronic supplementary material The online version of this article (doi:10.1186/s12885-015-1276-z) contains supplementary material, which is available to authorized users.

Background Ramucirumab (RAM), a monoclonal antibody for vascular endothelial growth factor 2 (VEGFR2), has been effective for advanced gastric adenocarcinoma (AC). However, little is known about the efficacy of RAM-containing chemotherapy (RAM-CTx) in gastric neuroendocrine carcinoma (G-NEC). Methods We retrospectively analysed and compared the clinical outcomes of patients (pts) with G-NEC receiving RAM-CTx, G-NEC receiving CTx without RAM and AC receiving RAM-CTx in our hospital. G-NEC was defined by neuroendocrine carcinoma features, regardless of the proportion, based on histology and neuroendocrine markers (synaptophysin, chromogranin A or CD56). VEGFR2 expression in tumour vessels was evaluated in archival primary G-NEC tissues by immunohistochemistry using the same anti-VEGFR2 primary antibody and scoring scheme (vascular VEGFR2 H-score) as in the REGARD trial. Results Seventeen G-NEC receiving RAM-CTx, 13 G-NEC receiving CTx without RAM and 173 AC pts receiving RAM-CTx were analysed. The overall response rate (59% vs 8 % vs 28%), progression-free survival (median 7.7 vs 1.8 vs 3.3 months) and overall survival (median 16.1 vs 8.6 vs 9.6 months) were significantly better in pts with G-NEC receiving RAM-CTx than G-NEC receiving CTx without RAM or AC receiving RAM-CTx. No severe or unexpected adverse events occurred. The median vascular VEGFR2 H-score, based on available G-NEC tissues from 12 pts receiving RAM-CTx, was 220 (range 150–260), which was markedly higher than that reported on AC tissues from the REGARD trial as historical control (median 35, range 0–240). Conclusions RAM-CTx showed a promising activity without severe or unexpected safety profile in pts with G-NEC. This may in part be explained by higher vascular VEGFR2 expression in G-NEC tissues.

4040 Background: Recently, anti-programmed death 1 (PD-1) or its ligand (PD-L1) antibodies have shown promising activities in metastatic gastric cancer (MGC). However, little is known about detailed clinicopathological features of PD-L1 expression in patients (pts) with MGC. Methods: Pts with histologically confirmed MGC were eligible for this prospective observational study. PD-L1 expression on tumor cell (TC) or inflammatory cell (IC) and mismatch repair (MMR) were analyzed by immunohistochemistry (IHC). Epstein–Barr virus (EBV) was detected by in situ hybridization. The expressions of tyrosine kinase receptors (RTKs) such as HER2, EGFR and MET and cancer genome alterations were also evaluated by IHC or next generation sequencing. Results: A total of 237 pts were enrolled from September 2015 to December 2016. Samples of 199 (84.0%) pts were obtained from biopsy. PD-L1 expression on TC and IC was positive in 27 (11.4%) and 167 pts (70.5%), respectively. One hundred and seventy-one pts (72.2%) had positive PD-L1 expression on either TC or IC. MMR deficient (D-MMR) and EBV was detected in 14 (6.9%, n = 203) and 14 pts (5.9%, n = 237), respectively. PD-L1 expression on TC was more frequently observed in pts with D-MMR (P < 0.001) and KRAS mutation (P < 0.001), while that on IC was more frequently observed in pts with EBV+ (P = 0.045), lymph node metastasis (P = 0.001), and lung metastasis (P = 0.045). In contrast, pts with peritoneal metastasis were associated with less frequent PD-L1 expression in IC (P = 0.003). A significant association was not observed between PD-L1 expression and RTKs expression or presence of other gene alterations. D-MMR was significantly associated with intestinal type (P = 0.026) and absence of liver metastasis (P = 0.022). PD-L1 expression on either TC or IC was not prognostic factor (hazard ratio; 0.92, P = 0.741). Seven pts with D-MMR and seven pts with EBV+ MGC were enrolled in clinical trials of anti-PD-1/PD-L1 antibodies. Conclusions: PD-L1 expression in MGC was associated with some clinicopathological features. Impact of these characteristics on efficacy of anti-PD-1/PD-L1 antibody warrants further evaluation.

TPS3623 Background: Immune checkpoint inhibitor (ICI) was reported to show durable responses in patients with MSI-H (Microsatellite Instability-High) metastatic colorectal cancer (mCRC). On the other hand, for patients with MSS (Microsatellite Stable) mCRC, ICI monotherapy achieved no response. Recently, WNT/β-catenin signaling has been reported to be involved in the elimination of tumor-infiltrating lymphocytes and the resistance of anti-PD-L1 antibodies. CRC is representative cancer with WNT/β-catenin pathway activation. Furthermore, STAT3 has also been reported to be a key driver of this immune evasion. Considering these rationales, the blocking of these signaling pathways with ICI may enhance antitumor immune response. Therefore, we initiated phase I/II study to assess efficacy and safety for the combination of BBI608, which blocks STAT3 and WNT/β-catenin signaling, with pembrolizumab in patients with mCRC. Methods: The eligibility criteria were patients with gastrointestinal cancer not responded to or intolerant of standard chemotherapies (SOC) for phase I part, and MSS mCRC refractory or intolerant to fluoropyrimidine, irinotecan, oxaliplatin, and anti-EGFR antibody (if wild-type RAS) for Cohort B in phase II part. For Cohort A, MSI-H mCRC refractory or intolerant to the SOC, irrespective of anti-EGFR antibody are investigated. Phase I part was designed to determine the recommended phase II dose in a “3+3” cohort-based dose escalation design of BBI608 (240mg BID every day on level 1 and 480mg BID every day on level 2) with pembrolizumab (200mg/body q3w). Primary endpoint of the phase II part is Immune-related objective response rate (irORR) determined by their Response Evaluation Criteria In Solid Tumors (irRECIST). A null hypothesis and alternative hypothesis for cohort B are irORR = 5% and 20%, respectively. Required sample size for Cohort B was 40 with a one-sided alpha of 5% and power of 90%. Required sample size for Cohort A (10 patients) was determined in an exploratory manner. We also investigate biomarker study using paired samples of both tumor biopsy and blood. The enrollment to phase I part began in November 2016. Clinical trial information: NCT02851004. Clinical trial information: NCT02851004.

e15570 Background: Vascular endothelial growth factor-A (VEGF-A) and VEGF recepter-2 (VEGFR-2) axis is known to induce Tregs in tumor bearing mice. Ramucirumab (RAM), a monoclonal antibody for VEGFR-2, was recently approved for patients with AGC. However, little is known about the influence of RAM on Tregs in TILs from AGC patients. Methods: Patients with AGC who had been planned to receive chemotherapy containing RAM were prospectively enrolled for this study from January 2016 to August 2016. Paired samples were obtained from primary tumor at pre- and post-RAM therapy to analyze immune profiles of TILs using flow cytometry assay. Results: A total of 17 patients were analyzed. Median age was 68 years old (range, 46-79). All patients received one or more previous chemotherapy. RAM containing regimens included RAM+paclitaxel (n = 13), RAM monotherapy (n = 3) and RAM+irinotecan (n = 2). FOXP3highCD45RA-CD4+ T cells (effector Tregs: eTregs) in CD4+ T cells was significantly decreased after RAM therapy compared with that in baseline (mean, 19.5% vs.13.6%, P = 0.036). In contrast, no significant change of CD8+ T cells was observed (mean, 36.7% vs. 36.0%, P = 0.86). Expression of programmed cell death-1 (PD-1) on CD8+ T cells was tended to be decreased after RAM therapy. Conclusions: eTregs in TILs was suggested to be decreased after RAM-containing therapy which warrants further evaluation in a larger cohorts. This observation might be a rationale to use RAM as an immunomodulator in combination of immune checkpoint inhibitors for AGC.

2546 Background: TAS-116 is an oral non-ansamycin, non-purine, and non-resorcinol highly selective inhibitor of HSP90α/β. The objective of this FIH study was to determine the MTD and investigate the safety, tolerability, PK, PD (HSP70 protein levels in PBMCs), and antitumor activity of TAS-116. Methods: The study is being conducted in Japan and the UK. Patients with advanced solid tumors received escalating doses of TAS-116 once daily (QD) with an accelerated titration design. After the MTD was determined, safety and tolerability of 5 days on / 2 days off per week administration (QDx5) at the MTD in QD was explored. In parallel, the MTD with every other day administration (QOD) was evaluated by using a 3 + 3 design. Results: As of 20 September 2016, 52 patients were enrolled. TAS-116 was evaluated at doses of 4.8 to 150.5 mg/m2/day in the QD schedule and doses of 107.5 to 295.0 mg/m2/day in the QOD schedule. The MTD was 107.5 mg/m2/day with QD and 210.7 mg/m2/day with QOD. QDx5 at the MTD in QD using a flat dose of 160 mg was evaluated. The most common adverse events in all regimens were gastrointestinal disorders and increased creatinine. DLTs were observed in 4 patients in QD (night blindness, visual disorder, AST/ ALT/gamma-GTP elevations, and anorexia) and in 2 patients in QOD (platelet count decreased, febrile neutropenia, pneumonia, respiratory failure, and septic shock). Reversible eye disorders were observed in all schedules, but those observed in QDx5 were limited to grade 1. The PK level demonstrated dose proportionality without unexpected accumulation under repeated administration. Dose-related HSP70 induction of PBMCs was observed. As of 20 September 2016, three confirmed durable PRs by RECIST were observed (239 days in GIST and 173 days in NSCLC with QD; 293 + days in NSCLC with QOD). PR and SD ≥ 12 weeks were observed in 15 out of 47 patients. Conclusions: TAS-116 had an acceptable safety profile under all schedules, especially QDx5. Preliminary antitumor activity was demonstrated with evidence of target engagement. Dose expansion at the MTD in this phase 1 study and the phase 2 study in patients with GIST are ongoing. Parts of this study will be expanded to the US with an amended study protocol. Clinical trial information: NCT02965885.