The effect of sex and BMI on outcomes in patients with metastatic non-small cell lung cancer treated with immunotherapy.

Abstract

e21085 Background: Recent studies suggest that among non-small cell lung cancer (NSCLC) patients treated with immunotherapy (IT), those who are male and/or have higher body mass index (BMI) benefit most; however, the role of other factors such as pretreatment weight loss is not clear. We conducted a retrospective study to further characterize the relationship between sex, BMI and response to IT in NSCLC. Methods: Patients with stage IV NSCLC treated with IT between 2017 and 2019 at UPMC Hillman Cancer Center were included. Demographic and clinical data were obtained from medical records. Chi-square test was used to compare baseline patient characteristics, best response (CR, PR and SD vs. PD), and presence of immune-related adverse events (iRAEs) between BMI and sex categories. Cox proportional hazards models were used to assess the effect of BMI and sex on progression free survival (PFS) and overall survival (OS). Analyses were conducted overall as well as stratified by treatment regime (1st line monotherapy, non-1st line monotherapy, and concurrent chemotherapy). Results: The study population consisted of 297 patients; 50.2% female (N=149), 87.8% white (N=261), and mean age at IT initiation 68 yrs (range: 36-91 yrs). Median follow-up time: 21 months. At IT initiation, 27 patients were underweight (BMI <18.5), 107 normal weight (BMI 18.5-24.9), 96 overweight (BMI 25-29.9), and 67 obese (BMI ≥30). Among underweight patients, weight loss pretreatment (≥10 lbs) was significantly more common ( P=0.02), and response to IT significantly worse (33% vs 61% good response; P=0.005) compared to those with BMI ≥18.5. No significant difference in response was observed between normal, overweight and obese patients, nor between men and women. The presence of iRAEs did not differ by BMI or sex. Females had better OS than males [HR (95%CI): 0.65 (0.47-0.90)] but PFS was similar. In stratified analyses, better OS among females was limited to the concurrent chemotherapy group [0.52 (0.30-0.92)]. Overall, underweight patients had worse OS than those with BMI ≥18.5 [1.71 (1.01-2.92)]; this was not significant after adjusting for pretreatment weight loss [1.48 (0.87-2.53)]. No difference was observed in OS and PFS between normal, overweight and obese patients. In stratified analyses, underweight individuals had worse OS [4.12 (1.55-10.94)] and PFS [3.87 (1.44-10.38)] than those with BMI ≥18.5 when treated with 1st line monotherapy. Weight loss pretreatment was independently associated with worse OS [2.20 (1.51-3.20)] and PFS [1.47 (1.05-2.05)]. Conclusions: In contrast to prior reports, NSCLC patients receiving IT did not benefit from higher BMI or male sex. Females treated with concurrent chemotherapy had improved OS, and pretreatment weight loss was an indicator of poor prognosis. Further study is required to understand the pathobiology behind these predictors.