Umbrella and basket trials in oncology: ethical challenges

Abstract

Background: Novel precision oncology trial designs, such as basket and umbrella trials, are designed to test new anticancer agents in more effective and affordable ways. However, they present some ethical concerns referred to scientific validity, risk-benefit balance and informed consent. Our aim is to discuss these issues in basket and umbrella trials, giving examples of two ongoing cancer trials: NCI-MATCH (National Cancer Institute - Molecular Analysis for Therapy Choice) and Lung-MAP (Lung Cancer Master Protocol) study.

Main body: We discuss three ethical requirements for clinical trials which may be challenged in basket and umbrella trial designs. Firstly, we consider scientific validity. Thanks to the new trial designs, patients with rare malignancies have the opportunity to be enrolled and benefit from the trial, but due to insufficient accrual, the trial may generate clinically insignificant findings. Inadequate sample size in study arms and the use of surrogate endpoints may result in a drug approval without confirmed efficacy. Moreover, complexity, limited quality and availability of tumor samples may not only introduce bias and result in unreliable and unrepresentative findings, but also can potentially harm patients and assign them to an inappropriate therapy arm. Secondly, we refer to benefits and risks. Novel clinical trials can gain important knowledge on the variety of tumors, which can be used in future trials to develop effective therapies. However, they offer limited direct benefits to patients. All potential participants must wait about 2 weeks for the results of the genetic screening, which may be stressful and produce anxiety. The enrollment of patients whose tumors harbor multiple mutations in treatments matching a single mutation may be controversial. As to informed consent - the third requirement we discuss, the excessive use of phrases like "personalized medicine", "tailored therapy" or "precision oncology" might be misleading and cause personal convictions that the study protocol is designed to fulfill the individual health-related needs of participants.

Conclusions: We suggest that further approaches should be implemented to enhance scientific validity, reduce misunderstandings and risks, thus maximizing the benefits to society and to trial participants.

Keywords: Basket trial; Ethics; Informed consent; Lung-MAP; Master protocol; NCI-MATCH; Risk-benefit balance; Scientific validity; Umbrella trial.

Fig. 1

General schema dividing master protocol into a basket or an umbrella trial. A basket trial enrolls patients with different cancer types but sharing one common molecular alteration. They receive one treatment. An umbrella trial enrolls patients sharing the same cancer type but different molecular alteration. The treatment is adjusted to the specific target

Fig. 2

The schema of NCI-MATCH study design. The biopsy material derived from registered patients is characterized for specific pre-defined mutations via genetic sequencing. If an actionable mutation is detected, patients are assigned to 1 of 30 treatment sub-protocols. Those experiencing disease progression or serious adverse events undergo review of their previous biopsy results or undergo another biopsy to search for alternative treatment [10]

Fig. 3

The schema of the Lung-MAP study design with five initial sub-studies. Adult patients with recurrent or metastatic squamous cell carcinoma (SCC) after progression on first-line platinum-based chemotherapy could be eligible to register in the Lung-MAP trial. After signing an informed consent, their archival or fresh tumor biopsy sample was screened for genetic aberrations. Results of genomic testing were returned within 16 days. Then, all patients were originally assigned to one of five sub-studies and they were randomized to receive either standard of care or a specific agent tailored to their alteration. One of these sub-trials was called a “non-match” sub-study and it enrolled patients whose tumors did not harbor any of the genetic aberrations tested in other sub-studies [15, 16]