Abstract
BACKGROUND: While evidence of hyperprogressive disease (HPD) continues to grow, the lack of a consensual definition obscures a proper characterization of HPD incidence. We examined how HPD incidence varies by the tumor type or the type of definition used. METHODS: We searched PubMed, Embase, the Cochrane Library of Systematic Reviews, and Web of Science from database inception to June 21, 2022. Observational studies reporting HPD incidence, in patients diagnosed with solid malignant tumors and treated with immune checkpoint inhibitors (ICI), were included. Random-effects meta-analyses were performed, and all statistical tests were 2-sided. RESULTS: HPD incidence was 12.4% (95% CI 10.2%-15.0%) with evidence of heterogeneity (Q = 119.32, p < 0.001). Meta-regression showed that the risk of developing HPD was higher in patients with advanced gastric cancer (adjusted odds ratio [OR], 10.83; 95% CI, 2.14-54.65; p < 0.001), hepatocellular carcinoma (adjusted OR, 7.99; 95% CI, 1.68-38.13; p = 0.006), non-small cell lung cancer (adjusted OR, 7.14; 95% CI, 1.58-32.29; p = 0.005), and mixed or other types (adjusted OR, 5.09; 95% CI, 1.12-23.14, p = 0.018) than in patients with renal cell carcinoma. Across definitions, HPD defined as a tumor growth kinetics ratio ≥ 2 (adjusted OR, 1.82; 95% CI, 1.08-3.07; p = 0.025) based on the Response Evaluation Criteria in Solid Tumors (RECIST) reported higher incidence than when HPD was defined as RECIST-defined progressive disease and a change in the tumor growth rate (TGR) exceeding 50% (∆TGR > 50). CONCLUSIONS: The incidence of immunotherapy-related HPD may vary across tumor types and definitions used, supporting the argument for a uniform and improved method of HPD evaluation for informed clinical decision-making.