Usage Examples
- The basket trial opened a thirteenth cohort this quarter, so the protocol amendment went in before the first patient was screened.
- Every cohort in the basket trial carries its own futility boundary, and two of them closed at the interim.
- We cannot claim the pan-tumour result: the basket trial's per-histology analysis was exploratory, not pre-specified.
What is Basket Trial?
A basket trial is a master-protocol design that enrolls patients into one study by shared molecular target rather than by disease, testing a single therapy across multiple histologies as separate, independently analyzed cohorts.
Basket trials exist because a targeted therapy is matched to a molecular alteration, not to an organ. When the same fusion or mutation turns up in a dozen rare tumour types, running a dozen separate trials is slower than the biology justifies and frequently impossible to enrol. A basket trial answers that by opening one protocol and letting the biomarker, rather than the tumour site, define eligibility.
A basket trial covers several disease cohorts receiving one investigational therapy under a single master protocol, with shared eligibility logic, shared safety monitoring, and one statistical framework. A basket trial does not cover the reverse arrangement, many therapies tested inside one disease, which is an umbrella trial. Nor is a basket trial a retrospective pooling of separately conducted studies; the cohorts must be planned together, in advance.
A basket trial is run in practice as parallel cohorts, each with a pre-specified response endpoint and futility rule, opened and closed by protocol amendment under one IND. Because ICH E9 treats subgroup findings as exploratory unless they were planned as part of the confirmatory analysis, a basket trial that leaves its per-cohort analysis undefined until the data arrive will struggle to turn cohort results into label language.
Not to be confused with
- Umbrella Trial
- an umbrella trial holds the disease constant and varies the therapy across molecularly defined arms; a basket trial holds the therapy constant and varies the disease. Identical master-protocol machinery, opposite axis of variation.
- Master Protocol
- a master protocol is the governing document that lets several sub-studies run under one framework. A basket trial is one design you can build with that document; umbrella and platform trials are the others.
- Phase 3
- a phase designation describes the evidentiary stage of a study, while "basket trial" describes only its enrolment logic. A basket trial can sit at any phase, so calling a study a basket tells you nothing about whether its evidence is confirmatory.
- Orphan Drug
- orphan status is a regulatory designation granted for a rare disease. A basket trial is a study design that may generate evidence across several rare indications, but the design itself confers no designation and no exclusivity.
No regulation names the basket trial as a design. A basket trial's obligations come from the IND rules the study runs under and the statistical principles its cohort analyses have to satisfy.
What you must do
- 1File a protocol amendment before conducting any study not already covered by a protocol in the IND; this is the mechanism by which a sponsor opens, expands, or closes a basket cohort21 CFR 312.30(a)
- 2State in the protocol the criteria for patient selection and for exclusion of patients and an estimate of the number of patients to be studied, at cohort level and not only in aggregate21 CFR 312.23(a)(6)(iii)(c)
- 3Set out in the protocol any subgroup or cohort effect that forms an integral part of the planned analysis, instead of defining per-histology comparisons once the data are visibleICH E9 section 5.7
- 4Identify any multiplicity remaining in a confirmatory analysis in the protocol, and either specify the adjustment procedure or record in the analysis plan why adjustment is not considered necessaryICH E9 section 5.6
- 5Notify FDA and all participating investigators of qualifying safety information no later than 15 calendar days after the sponsor determines it qualifies for reporting, aggregating exposure across every cohort in the basket21 CFR 312.32(c)(1)
Common mistakes
Reporting per-histology response rates the protocol never planned to test
ICH E9 section 5.7 is explicit that subgroup analyses are exploratory in most cases, and that "any conclusion of treatment efficacy (or lack thereof) or safety based solely on exploratory subgroup analyses are unlikely to be accepted." A basket that publishes twelve unplanned cohort response rates converts none of them into label language, and the sponsor has funded a trial that supports no indication.
Opening a cohort administratively instead of by protocol amendment
21 CFR 312.30(a) requires a protocol amendment whenever a sponsor intends to conduct a study not covered by a protocol already in the IND. A cohort added by memo or by a site-level decision is an unauthorised clinical investigation; the data are exposed to exclusion at review, and the observation attaches to the sponsor, not the site.
Sizing the basket in aggregate rather than per cohort
21 CFR 312.23(a)(6)(iii)(c) requires an estimate of the number of patients to be studied alongside the selection and exclusion criteria. Sponsors who plan a total enrolment figure without a per-cohort estimate and futility boundary keep non-responding histologies open, spending enrolment and diluting the safety database on cohorts that were never powered to answer anything.
When This Matters
- The basket trial opened a thirteenth cohort this quarter, so the protocol amendment went in before the first patient was screened.
- Every cohort in the basket trial carries its own futility boundary, and two of them closed at the interim.
- We cannot claim the pan-tumour result: the basket trial's per-histology analysis was exploratory, not pre-specified.
Frequently Asked Questions
A basket trial holds the therapy constant and varies the disease; an umbrella trial holds the disease constant and varies the therapy across molecularly defined arms. Both run sub-studies under one master protocol, so the machinery is the same and only the axis of variation differs.
Related Use Cases
- Pharma Use Cases
Cut NDA and sNDA prep time by 60% with AI-assisted drafting and automated readiness checks
- Biotech Use Cases
Compress IND prep from 8-12 weeks to under 3 weeks with AI-assisted drafting and validation
- Clinical Workflows
Track clinical trial regulations across global markets
- Regulatory Affairs Workflows
Cut regulatory intelligence tracking from 10+ hours/week to automated, real-time alerts
