Correlating programmed death ligand 1 (PD-L1) expression, mismatch repair deficiency, and outcomes across tumor types: implications for immunotherapy

Oncotarget. 2017 Aug 24;8(44):77415-77423. doi: 10.18632/oncotarget.20492. eCollection 2017 Sep 29.

Abstract

The identification of biomarkers associated with response to therapeutic agents is central to optimizing patient outcomes. Expression of the immune checkpoint proteins PD-1/L1, and DNA mismatch repair deficiency (dMMR) status may be predictive response biomarkers for immunotherapies, but their overlap requires further study. We prospectively conducted PD-L1 and MMR immunohistochemistry (IHC) on 430 consecutive patients with advanced gastrointestinal (GI) cancers, genitourinary (GU) cancers or rare cancers between June 2012 and March 2016. Overall 393/430 (91.4%) patients were evaluable for PD-L1 expression by IHC. The frequency of tumor PD-L1 positivity (PD-L1+) was 16.5% (65/393). Among anatomic tumor sites PD-L1+ was 28.6% in melanoma, 22.2% in GC, 20.9% in CRC, 12.5% in BTC, 7.1% in GU cancer, 6.7% in HCC, 0% in pancreatic cancer and 0% in sarcoma. Among the 394 evaluable for MLH1/MSH2 expression cases, 18 patients (4.5%) had dMMR tumors. The dMMR was most common in GC (7.1%) followed by 6.7% in HCC, 4.4% in CRC, and 2.7% in sarcoma. Of the 365 patients evaluable for both PD-L1 and MLH1/MSH2 expression, there was a significant association between the PD-L1 expression and MLH1/MSH2 loss (P = 0.01), but not with overall survival within tumor types. PD-L1 status and dMMR are overlapping putative response biomarkers in immunoncology. Clinical trials with biomarker enrichment restricted to PD-L1+ or dMMR may be inadequate to capture the subset of patients who may benefit from immune mediated therapies. More robust immunotherapy biomarkers and careful clinical trial design are warranted.

Keywords: biomarker; gastrointestinal cancer; immunotherapy; mismatch repair deficiency (dMMR); programmed death-ligand 1 (PD-L1).