Abstract

Immune checkpoint blockade improves clinical outcomes for patients with G/GEJ cancers, but its efficacy and impact on the tumor microenvironment (TME) in the neoadjuvant setting remains largely unknown. Peri-operative FLOT chemotherapy, currently 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 anti-PD-L1 monotherapy can prime the TME prior to combination with chemotherapy. In the phase II PANDA study (NCT03448835), 20 patients with resectable G/GEJ adenocarcinoma were neoadjuvantly treated with a single cycle of atezolizumab (anti-PD-L1), followed by 4 cycles of atezolizumab plus docetaxel, oxaliplatin and capecitabine (A-DOC) and subsequent surgery. Tumor tissue was obtained at baseline, after atezolizumab monotherapy, after first A-DOC and at surgery. The primary endpoint was safety. Disease-free survival, pCR rate and biomarker assessments were prespecified exploratory endpoints. Treatment was well-tolerated and safe, and all patients underwent timely surgery. MPR (≤10% viable tumor rest) was observed in 14/20 patients (70%), including 9 pCR (45%). At a median follow-up of 29 months, 15 patients (75%) were disease-free, including all patients with an MPR. Using gene expression analysis, PD-1 was significantly higher at baseline in responders vs nonresponders, while neutrophil signature was significantly higher in nonresponders vs responders. TMB and PD-L1 CPS were not associated with response. Immune activation was already observed after one cycle of atezolizumab monotherapy, as evidenced by increases in CD4+ and CD8+ T cell infiltration, PD-1, CXCL13 and IFN-γ signatures compared to baseline. This was significant only in responders. Neoadjuvant atezolizumab plus chemotherapy leads to promising pathologic responses in G/GEJ cancer. Baseline PD-1 and neutrophil signature were associated with response. Biomarker analyses after a single dose of atezolizumab showed immune activation in responders and may be used as an early indicator of response. These data should be validated in a large randomized controlled trial.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call