Toxicities associated with checkpoint inhibitor immunotherapy: The Karmanos Cancer Center experience.

Abstract

e14575 Background: PD-1 inhibitors work by re-instating the natural anti-cancer immune-mediated cytotoxicity. Although this new class of therapy offers hope for many advanced stage cancer patients (pts), data on pts who are more likely to experience adverse events (AE) are not available. Methods: Data from pts who received at least one dose of PD-1 inhibitors before August 31, 2016, were captured from our institution’s pharmacy database. AE of anemia, neutropenia, thrombocytopenia, acute kidney injury, hypothyroidism, transaminitis, pneumonitis, colitis and fatigue were recorded and graded based on CTCAEv4. Baseline pt characteristics were statistically compared between subjects who received PD-1 inhibitors on a clinical trial vs as part of standard of care (on vs. off-study) by Fisher’s exact test for categorical variables and Kruskal-Wallis test for continuous variables. Univariable and multivariable logistic regression models were fit to assess associations between toxicities (Grade0-1 vs. Grade2+) and predictors (age, agent, primary tumor, and trial status). Results: 231 pts received at least one dose of PD-1 inhibitors prior to data cut-off. Median age was 65 (24-92) off trial (n = 125) and 59. 5 (25-79) on trial (n = 106). There were 117 (51%) non-small cell lung cancer (NSCLC), 41 (18%) renal cell carcinoma (RCC) and 73 (32%) others. 24 (10%) melanoma, 18 (8%) Hodgkin’s lymphoma (HL), 9 (4%) head and neck squamous cell carcinoma (HNSCC), 8 (4%) small cell lung cancer (SCLC) were included in others. 129 (56%) had nivolumab alone, 50 (22%) had pembrolizumab alone and the rest 52 (23%) received combination treatment. Pts participating in a clinical trial had higher odds to experience grade 2 or more pneumonitis than those not on a trial (unadjusted p = 0.038). On trial pts had lower odds to experience grade 2 or more fatigue than those receiving the agent as standard of care (unadjusted p = 0.026) which remained significant after adjusting age, agent (single agent vs combination) and primary tumor (adjusted p = 0.004). Conclusions: In our observation, pneumonitis was seen more on clinical trials whereas fatigue was seen less.