Abstract
Background Complex innovative trial designs (CIDs) have gained prominence in recent years as a means to enhance efficiency and flexibility in the development of therapies for both hematologic and solid tumor malignancies. By addressing multiple clinical questions within a single framework, CIDs offer a compelling alternative to traditional randomized controlled trials, which have long been the gold standard of cancer research. The rise of CIDs in oncology has been fueled by the growing availability of targeted therapies and biomarker-driven strategies. However, the adoption and utility of CIDs in hematologic malignancies remain less well characterized. In this study, we examine the use of four CID types (adaptive, basket, umbrella, and de-escalation) across hematologic and solid malignancy trials registered between 2015 and 2023, with a focus on sponsor type, trial phase, and temporal trends.
Methods We conducted a cross-sectional analysis of interventional clinical trials involving hematologic and solid malignancies registered on ClinicalTrials.gov between January 1, 2015 and December 31, 2023. The trials were categorized by cancer type (hematologic vs. solid), sponsor (federal, academic/non-profit, or industry) and phase. Four complex trial designs (adaptive, basket, umbrella and de-escalation) were identified using keyword flags. We compared the prevalence of each trial design based on sponsor and phase among hematologic and solid malignancies using chi-square tests in Stata.
Results We analyzed 427 interventional trials involving hematologic malignancies and 325 involving solid tumors registered from 2015 to 2023. Adaptive trial designs were infrequently used in both groups (1.87% in hematologic vs. 1.23% in solid malignancies), with no significant variation based on sponsor type or trial phase.
In contrast, basket trials were more common in both hematologic (6.1%, p<0.001) and solid malignancies (11.4%, p<0.001). Academic or non-profit organizations were the primary sponsors, especially in solid tumors (38.4% vs. 12.8% in hematologic malignancies). Basket designs were predominantly associated with phase 2 trials, particularly in solid malignancies (35.7%, p<0.001) compared to hematologic trials (10.4%, p=0.005).
Umbrella trials were rarely used (0.47% in hematologic vs. 0.62% in solid malignancies), with no significant differences observed by sponsor or trial phase.
De-escalation designs were significantly more common in solid tumor trials (13.5%, p<0.001) compared to hematologic trials (4.9%, p=0.005). Industry was the most frequent sponsor of de-escalation trials in both groups (18.1% in solid vs. 8.2% in hematologic). These designs were more often used in phase 1 and 1/2 trials, with 21.2% of solid and 11.5% of hematologic trials employing them.
Temporal analysis showed that the use of CIDs peaked between 2015 and 2018, followed by a decline in subsequent years across both malignancy types.
Conclusion Despite their potential to accelerate therapeutic development and personalize care, CIDs remain markedly underutilized in clinical trials, particularly in hematologic malignancies. While basket and de-escalation designs show greater adoption, especially in solid malignancies and early-phase trials, adaptive and umbrella designs are strikingly rare. The observed decline in CID use after 2018 highlights missed opportunities in leveraging these efficient, flexible trial frameworks. Expanding the use of CIDs in hematologic cancers could transform trial efficiency, reduce development timelines, and better align research with modern biomarker-driven treatment strategies.
