Characteristics and research waste of immunotherapy randomized clinical trials: a cross-sectional analysis

Immunotherapy spans tumor and non-tumor disciplines, yet the surge in randomized controlled trials (RCTs) has not yielded proportional improvements in evidence quality, leading to substantial research waste. In this study, the immunotherapy RCT landscape was characterized, and the prevalence and risk factors for research waste were quantified.

This registry-based cross-sectional study analyzed phase III/IV immunotherapy RCTs from ClinicalTrials.gov. The publication status was verified via PubMed, Scopus, and Google Scholar. Research waste was defined as non-publication, inadequate reporting (CONSORT 2010), or avoidable design flaws (Cochrane Risk of Bias Tool). Associations between study characteristics and outcomes (publication status and waste) were evaluated using logistic regression.

Among the 164 RCTs analyzed, 127 (77.4%) exhibited research waste. With respect to publication status, 74 (45.1%) were published, and 90 (54.9%) were unpublished. Multivariate analysis revealed that non-tumor diseases (adjusted OR = 0.29, 95% CI: 0.09-0.85, P = 0.028) and a quadruple-blind design (adjusted OR = 4.84, 95% CI: 1.71-14.90, P = 0.004) were significantly associated with publication status. Univariable analysis revealed that non-tumor diseases (OR = 4.36, 95% CI: 2.02-9.57, P<0.001) and European principal investigators (PIs) (OR = 3.48, 95% CI: 1.49-8.16, P = 0.004) were positively correlated with waste, whereas international recruitment (OR = 0.16, 95% CI: 0.07-0.36, P<0.001), multicenter design (OR = 0.14, 95% CI: 0.05-0.35, P<0.001), and large sample size (OR = 0.27, 95% CI: 0.12-0.60, P = 0.001) were negatively correlated. After adjustment, non-tumor diseases remained a risk factor (OR = 3.84, 95% CI: 1.18-13.54, P = 0.029), whereas a multicenter design had a strong protective effect (OR = 0.07, 95% CI: 0.00-0.57, P = 0.030).

This study revealed a high incidence of research waste in immunotherapy RCTs, which was predominantly driven by non-tumor studies. These findings underscore the urgent need to prioritize large-scale collaborative frameworks to optimize resource allocation and ensure reliable clinical evidence.