Revisiting Risk and Benefit in Early Oncology Trials in the Era of Precision Medicine: A Systematic Review and Meta-Analysis of Phase I Trials of Targeted Single-Agent Anticancer Therapies.

Abstract

Phase I trials are a crucial step in the evaluation of new cancer therapies. Historically, low rates of response (5%) and comparably high rates of death from toxicities (0.5%) have contributed to debates on the ethics and orientation of these trials. With the introduction of novel targeted therapies, a contemporary estimate is needed. We systematically searched PubMed, Embase, and ClinicalTrials.gov for reports of phase I oncology trials of single-agent targeted immunomodulators, molecularly targeted therapies, and antiangiogenic agents, published between January 2015 and July 2018. Adult and pediatric trials of solid and hematological malignancies were eligible. Treatment-related adverse events (grades 3, 4, and 5) and response rates (objective, complete, and partial) were extracted and analyzed. One hundred and fifty-eight trial reports, covering 6,707 patients, were included. The rate of treatment-related deaths was 0.0% (95% CI, 0.0 to 0.1), while 13.2% of patients (9.5 to 17.3) experienced a grade 3 or 4 treatment-related toxicity. The combined objective response rate was 6.4% (4.6 to 8.5). Among trials using tumor biomarkers as eligibility criteria, the objective response rate was higher (12.0% [7.3 to 17.6] compared to 4.9% [2.5 to 5.7], P value < .01). The same was true of trials focusing on a single tumor type (13.4% [8.2 to 19.4]) compared to multiple tumor types (3.8% [2.5 to 5.3], P value < .01). Reduced grade 5 risk and improved benefit appears to exist in modern phase I oncology trials, particularly in trials that target single tumor types and integrate biomarkers as eligibility criteria. These findings provide information to support informed consent discussions, highlight the need for improved reporting of phase I oncology trials, and provide direction for optimizing their design.