Phase I solid tumor trials (PHIstt) in the era of precision oncology (PO): A systematic survey.

Abstract

e13603 Background: In the era of PO, systemic therapies for patients (pts) with solid tumors have fundamentally changed from chemotherapy (CT) to targeted therapy (TT) and immunotherapy (IO). No study to the best of our knowledge has described PHIstt in this era. In this systematic survey, we describe the evolving features of PHIstt which began enrollment between 2010-2020, focusing on type of dose escalation scheme (DES) utilized, use of expansion cohorts (ECs) and nature of adverse events (AEs). Methods: The primary aim was to estimate the proportion of PHIstt which used rule-based DES. A literature search was performed by a biomedical librarian to identify PHIstt in adult patients published between January 1, 2000 – December 31, 2020 from a selection of 12 journals that publish PHIstt. However, we included studies enrolling between 2010-2020 to better capture agents in the PO era. We included trials with both single and multiple agents, and single and multiple modalities if the systemic therapy tested was being escalated. We excluded studies which only included intra-tumoral therapy, radiation therapy, or supportive care. Two reviewers assessed studies for inclusion and performed data abstraction; a 3rd reviewer established concordance. Results: Of 10,744 studies retrieved, 10,195 were non-topical or began enrollment prior to 2010; 437 studies in pts aged ≥ 18 years with advanced solid tumors were included. A median of 31 pts were enrolled per study (IQR 19-50), with median age 60 years in escalation cohorts. In studies reporting pt sex, 45.7% were female, and 53.9% were male. The most common classes of therapeutics were TT (47.6%), IO (22%), and CT (6.9%). Only 40.1% of studies reported ECs in the primary paper, and 46.6% of ECs used genomic selection. Rule-based designs (which assign patients to dose levels based upon prespecified rules of actual AE observations) were used in 89% of PHIstt; a 3+3 DES was used in 80.5% of these studies. With regards to grade 3/4 AE reporting, 56.6% of studies reported treatment-related AEs, 21.3% reported treatment-emergent AEs and 10.7% did not specify how they were reported. The most common class of grade 3/4 AEs was hematologic (37.1%). The most common dose-limiting toxicity classes were other (64.6%) and gastrointestinal (16.9%). Of all the drugs tested in PHIstt, 37.5% were tested subsequently in phase II while 10.3% garnered licensure. Conclusions: There is room for improvement in the trial design of PHIstt which may lead to more regulatory success. Some areas we identified include the more regular adoption of genomically selected ECs, more consistent reporting of AEs and diversification beyond rule-based designs to better assess safety/efficacy profiles of non-CT agents. Dose-Escalation Designs. [Table: see text]