Background The International Gastric Cancer Linkage Consortium (IGCLC) consensus guideline advises prophylactic gastrectomy in early adulthood to prevent gastric cancer development in CDH1 germline mutation carriers; psychosocial reasons may postpone gastrectomy. We analyzed the yield of signet-ring cell carcinoma (SRCC) during surveillance gastroscopy in CDH1 mutation carriers. Methods A retrospective analysis on surveillance gastroscopies in CDH1 mutation carriers was performed. The yield of SRCC in both targeted and random biopsies was studied. Endoscopic (biopsy) results were compared with the histopathologic outcomes in gastrectomy specimens. Results 42 CDH1 mutation carriers (18 men; mean age 43, range 20–82 years) underwent 96 surveillance gastroscopies. SRCC lesions were identified on surveillance gastroscopy in 21 patients (50 %), by either targeted biopsies only (n = 11), random biopsies only (n = 3), or both random and targeted biopsies (n = 7). SRCC was detected in 41 /377 targeted biopsies (11 %), whereas random biopsies revealed SRCC in 14/1563 biopsies (0.9 %). At least one SRCC lesion was found in 26 of 30 gastrectomy specimens. In 18 of these 26 specimens (69 %), SRCC had been identified by endoscopic biopsies. Missed lesions were all small superficial SRCC foci, mainly in the body of the stomach. Conclusion In our cohort of CDH1 mutation carriers, SRCC lesions were identified by an extensive endoscopic surveillance protocol in 69 % of SRCC-positive patients who underwent a gastric resection. The low number of SRCC detected through random sampling demands a critical reappraisal of random biopsy sampling in the IGCLC guideline.

4059 Background: Immune checkpoint blockade improves clinical outcomes for patients with gastric and gastro-esophageal junction (GEJ) cancers, but its efficacy and impact on the tumor microenvironment in non-metastatic, resectable disease remains largely unknown. Peri-operative FLOT, the current standard-of-care, leads to pathologic complete responses (pCR) and major pathologic responses (MPR) in 16% and 37% of patients, respectively. An important open question is whether PDL-1 blockade monotherapy can prime the tumor microenvironment in a favorable manner, prior to combination with chemotherapy. Methods: We report results from the phase 2 PANDA trial (NCT03448835) of neoadjuvant atezolizumab (anti-PDL-1) plus docetaxel, oxaliplatin, and capecitabine (DOC) in patients with resectable gastric or GEJ adenocarcinoma. Patients received a single cycle of atezolizumab monotherapy, followed by 4 cycles of atezolizumab+DOC. Tumor tissue was collected at baseline, after atezolizumab monotherapy, the first atezolizumab+DOC, and at resection. The primary endpoints were safety and feasibility in 20 patients, and secondary endpoints included MPR (<10% viable tumor rest) and disease-free survival. Results: Twenty patients, of which 18 with mismatch repair (MMR) proficient and two with MMR-deficient tumors, were evaluable for safety and efficacy analyses. MPR was observed in 14/20 patients (70%; 95% CI 46–88%), including 9 pCR (45%; 95% CI 23-68%). Among patients with intestinal type adenocarcinoma, 12/15 (80%; 95% CI 52-96%) had an MPR, with 9/15 (60%; 95% CI 32-84%) pCR. Treatment was well tolerated, with two patients (10%) experiencing a grade 3 immune adverse event. At a median follow-up of 29 months (IQR 16-34), 15 patients (75%) were alive and disease-free. None of the patients with an MPR recurred. All patients underwent resections without treatment-related delays and no unexpected surgical complications were documented. Translational analyses, including baseline PDL-1 CPS score and whole exome sequencing (WES), plus CD8 T-cell infiltration and RNA sequencing at 4 timepoints will be presented at the meeting. Conclusions: Our data show that the addition of atezolizumab to neoadjuvant chemotherapy leads to promising pathologic responses in gastric/GEJ adenocarcinoma, which appears to be higher than in historical controls, with no recurrences in responders. These data should be validated in a large randomized controlled trial. Clinical trial information: NCT03448835.

No abstract