Patterns of co-first and co-last authorship in cancer phase 3 trials.

Abstract

50 Background: Phase 3 randomized clinical trials (RCTs) are the gold standard for efficacy in cancer treatment, and the resulting manuscripts carry great weight for clinical practice and for the authors themselves. First/last authorship on high-impact publications is a frequently used metric for promotion and identification of leadership of subsequent studies. Co-first/last (joint) authorship has become more common in many fields; the patterns of joint authorship in cancer phase 3 RCT publications are relatively unexamined. Methods: We identified all phase 3 RCTs published between 1998-2023 in the HemOnc knowledge base as of April 25, 2023. Joint authorship was identified through a combination of parsing PubMed structured data and scraping publisher websites for common phrases e.g., “contributed equally”. The identified manuscripts were matched by year and reference type (primary efficacy reports vs all others) 1:1 to those without joint authorship. Results: Out of N=3911 manuscripts meeting criteria, we identified 382 (10%) with joint authorship (Table). The incidence of joint authorship increased from 1/98 manuscripts published in 1998 to 34/194 (17.5%) in 2022. Manuscripts with joint authorship had a median (range) of 3 (3-13) co-first/last authors, with 54 (14%) having both co-first and co-last authors. Jointly-authored manuscripts had a median (IQR) of 21 authors (17-29) vs 20 (15-23) for matched manuscripts, p<0.001. Of 970 unique authors who had ³1 joint authorship, the most common country of affiliation was China (254, 26%) followed by US (195, 20%); in the matched manuscripts, 243/672 (36%) first/last authors were based in the US, followed by France (65, 10%). Joint authors with known gender were more likely to be women vs matched non-joint first/last authors (OR 1.28, 95% CI 0.99-1.66). Joint authors in an ordinal position other than first or last had a less extensive publication record as of 2023, median (IQR) of 5 (2-11) phase 3 publications vs 4 (2-9), p<0.001. Conclusions: Joint authorship in phase 3 cancer RCTs is uncommon but has increased over the past 25 years. Based on the less extensive publication record of non-first/last joint authors, promotion of relatively junior investigators may play a role in joint authorship decisions. Societal factors may also play a role, as evidenced by the discrepancy between US-based joint and non-joint authors. While not statistically significant, a trend towards more women in joint authorship roles is notable. Further examination of the impact of joint authorship on career trajectory and subsequent trial leadership opportunities is warranted. [Table: see text]