Characteristics of Clinical Trials Evaluating Biosimilars in the Treatment of Cancer: A Systematic Review and Meta-analysis.

IMPORTANCE

Biologics account for almost half of US drug spending but may be subject to competitive pricing pressures by US Food and Drug Administration-approved biosimilars. The extent of the preapproval clinical testing that is needed and how these biosimilars compare with the originator biologic products remain critical issues in establishing a vibrant biosimilar market.

OBJECTIVES

To analyze the design of cancer biosimilar efficacy studies compared with the reference drug pivotal trials and provide summary risk ratio estimates for each cancer type drug subgroup.

DATA SOURCES

A systematic search was performed of articles and abstracts published using Embase, PubMed/MEDLINE, and ClinicalTrials.gov, last updated April 18, 2021.

STUDY SELECTION

All studies or abstracts in English comparing a disease-modifying cancer biologic and its biosimilar and reporting efficacy or surrogate efficacy results were included.

DATA EXTRACTION AND SYNTHESIS

Outcome estimates and study characteristics were extracted from each study. Among biosimilar efficacy studies, random-effects meta-analyses were performed for each cancer type molecule outcome subgroup, calculating pooled relative estimates and 95% CIs.

MAIN OUTCOMES AND MEASURES

Study characteristics, such as population size, blinding, and randomization, were compared between biosimilar trials and those of reference drugs. Risk ratio estimates for relative change to surrogate measures (eg, progression-free survival) were collected for biosimilars and their reference products.

RESULTS

A total of 31 cancer biosimilar studies of 3 reference products involving 12 310 patients were included. In all 7 subgroups, the biosimilars analyzed were indistinguishable from their reference drug on surrogate efficacy. Six reference drug trials were included, involving 1811 patients. On average, biosimilar studies involved more patients than reference drug trials (mean number of patients, 397 vs 302), were more likely to be randomized clinical trials rather than single-group or observational studies (100% [31 of 31] vs 50% [3 of 6]), and were more likely to be double blind rather than open label (84% [26 of 31] vs 17% [1 of 6]).

CONCLUSIONS AND RELEVANCE

This systematic review and meta-analysis found that the biosimilars for the cancer drugs in this sample were subjected to rigorous clinical evaluations, and the results were statistically indistinguishable from those of original products across drugs, cancer types, and outcome measures.