BESPOKE IO brings ctDNA into the immunotherapy decision loop
Immunotherapy has changed treatment options for many patients with advanced solid tumours, but response assessment still lags behind biology. Conventional imaging can miss early benefit, misclassify delayed responses, and struggle with atypical patterns such as pseudoprogression, leaving oncologists to make difficult decisions with incomplete information. The BESPOKE IO study was designed to test whether serial circulating tumour DNA, or ctDNA, can close that gap by providing a faster, biologically grounded readout of treatment effect.
The study uses a tumour-informed bespoke assay, Signatera, to follow ctDNA dynamics during immunotherapy. Because ctDNA has a short half-life, it can change quickly when tumour burden falls or persists, making it attractive as an early marker of response rather than a late surrogate of radiographic change. In practical terms, a falling ctDNA signal may support continued treatment, while persistent or rising ctDNA may prompt earlier reconsideration of an immunotherapy plan. In a field where imaging can lag behind biology by weeks or months, that difference can alter the entire treatment trajectory.
Serial ctDNA monitoring across a large multicentre cohort
BESPOKE IO is a multicentre, prospective observational study planned to enroll more than 1,500 patients with solid tumours receiving immunotherapy at up to 100 US sites. Patients will be followed for as long as 2 years, with serial ctDNA testing timed to every other treatment cycle. The design is intentionally pragmatic: it is meant to sit alongside routine care rather than replace standard imaging or established clinical assessment.
That structure matters because the study is not asking whether ctDNA can predict biology in isolation. It is asking whether ctDNA can change clinical behavior when used together with conventional response tools. The primary endpoint is the proportion of patients whose treatment regimen changes on the basis of post-treatment bespoke ctDNA results combined with standard response assessment. In other words, the trial is measuring whether molecular information actually reaches the bedside and alters continuation, discontinuation, escalation, or de-escalation decisions.
Serial sampling is especially relevant in immunotherapy, where response kinetics can be unpredictable. Some patients appear to progress on scans before later demonstrating benefit, while others maintain stable radiographic findings despite a molecular signal of resistance. A blood-based assay that reflects tumour turnover in near real time could help clinicians distinguish these patterns earlier than imaging alone.
Why ctDNA may fit the biology of immune response
The scientific rationale rests on the behavior of ctDNA itself. Tumour-derived DNA fragments circulate in the bloodstream and can be detected when active disease sheds material into the circulation. If immunotherapy is controlling tumour growth, the ctDNA burden should drop as viable cancer cells diminish. If treatment is failing, ctDNA may remain detectable or increase even before imaging captures clear progression.
This is particularly appealing in advanced solid tumours, where disease burden is heterogeneous and immune responses can be patchy across lesions. Imaging measures size, shape, and metabolic behavior only indirectly. ctDNA, by contrast, is tied more closely to the presence of living malignant cells. The biological logic is straightforward: less viable tumour should mean less tumour DNA entering the circulation.
The trial’s observational design acknowledges an important reality. Immunotherapy decisions are rarely made on a single data point. Clinicians weigh symptoms, scan results, lab values, organ function, prior response, and toxicity. BESPOKE IO is positioned to show whether ctDNA adds enough incremental information to influence those judgments without displacing standard oncology practice.
Endpoints focused on action, not just correlation
The major secondary endpoints extend beyond decision-making into outcomes: progression-free survival, overall survival, and overall response rate based on ctDNA dynamics. These measures will help determine whether molecular changes track with clinically meaningful benefit rather than simply mirroring radiographic trends.
The choice of endpoints is important. A biomarker can be analytically impressive and still fail at the bedside if it does not connect to patient outcomes. By linking serial ctDNA to response assessment and survival outcomes, the study is trying to answer a more consequential question: does a blood-based signal help identify who is benefiting from immunotherapy early enough to matter?
This is where the study could have real operational value for oncology clinics. If ctDNA-guided decisions consistently identify nonresponders sooner, patients may avoid prolonged exposure to ineffective therapy. If the assay confirms molecular response before scans do, it may support treatment continuation in cases where clinicians might otherwise hesitate. Either way, the result could be more precise use of immunotherapy in a setting where overtreatment and premature discontinuation both carry risk.
Governance, oversight, and what this protocol can and cannot prove
The study received approval from the WCG Institutional Review Board on 22 February 2021. Data protection and privacy regulations will govern the collection, transfer, processing, and storage of patient information. Natera will approve publication of study results in line with site-specific contract requirements.
As a protocol paper, BESPOKE IO does not provide outcome data. It establishes a framework for evaluating clinical utility, not proof of benefit. The observational design also means any future findings will need careful interpretation, since treatment changes will occur within real-world clinical practice rather than a randomized comparison. Even so, the protocol addresses one of the central blind spots in immuno-oncology: radiology alone often cannot capture the pace and complexity of immune-mediated tumour control.
The real test will be whether serial ctDNA can consistently guide action in a way that improves decision quality without adding noise. Until the results emerge, the study stands as a large, methodically built attempt to make immunotherapy monitoring more biologically responsive and less dependent on delayed imaging signals.