Oncology · Medical Literature
Use of circulating tumour DNA to prospectively guide a switch from targeted to immune therapy in BRAF mutant advanced melanoma: the randomised phase II CAcTUS trial.
Abstract
Checkpoint inhibitor immunotherapy (CPI) for BRAF mutant advanced melanoma first-line results in a better long-term survival compared to targeted therapy (TT), however TT induction may benefit poor prognosis groups. The parallel-arm, randomised phase II, multicentre, feasibility CAcTUS trial (Clinicaltrials.gov NCT03808441) randomised 21 patients to receive standard of care investigators choice TT or CPI, switching to the alternative upon progression (n = 10), or commencing TT and switching to CPI upon an ≥80% reduction of BRAF variant allele frequency (VAF) in circulating tumour DNA (ctDNA; n = 11). The study achieved its primary endpoints with 100% (95% confidence interval [CI]: 94-100%) of critical results provided within 7 days to inform a decision to switch and 100% of patients commencing TT achieving an ≥80% reduction of BRAF VAF (95% CI: 80-100%). Secondary outcomes included progression-free survival and overall survival. No new safety signals were observed for TT/CPI. Post-hoc analysis of clinical features, circulating cytokines and chemokines at ctDNA nadir following TT induction suggested a more favourable profile prior to CPI initiation. Longitudinal ctDNA dynamics revealed ctDNA provided an early signal of CPI benefit and that rechallenge with TT following CPI progression resulted in a further ctDNA response. These data support the utility of ctDNA to guide treatment decision-making within a clinically relevant timeframe to optimise treatment scheduling strategies.
Clinical relevance
In patients with BRAF-mutant advanced melanoma, is it feasible to use a ≥80% reduction in BRAF variant allele frequency (VAF) measured by circulating tumour DNA (ctDNA) as a prospective, real-time trigger to switch from targeted therapy (dabrafenib + trametinib) to checkpoint inhibitor immunotherapy (nivolumab + ipilimumab), compared with standard-of-care sequential therapy switched at radiological progression? __SNAPSHOT__ {"studyDesign":"Parallel-arm, randomised phase II, multicentre feasibility trial (CAcTUS; NCT03808441)","setting":"Multicentre; United Kingdom (institution type not individually reported in abstract)","sampleSize":"n=21 total (Control Arm A: n=10; Experimental Arm B: n=11)","population":"Adults with BRAF-mutant advanced cutaneous melanoma (unresectable Stage III or Stage IV), baseline ctDNA BRAF VAF ≥1.5%, ECOG PS 0–2, no symptomatic brain metastases, no prior adjuvant nivolumab + ipilimumab; prior adjuvant PD-1 monotherapy permitted if relapse >6 months post-discontinuation","intervention":"Experimental Arm B: dabrafenib (150 mg BD) + trametinib (2 mg OD) as TT induction, with serial ctDNA BRAF VAF monitoring (q2-weekly × 4 weeks, then q4-weekly); switch to nivolumab + ipilimumab (N 1 mg/kg + I 3 mg/kg q3-weekly × 4 doses, then N 480 mg q4-weekly) upon achieving ≥80% VAF reduction from baseline; resumption of TT permitted upon subsequent progression on CPI","comparator":"Control Arm A (standard of care): investigator's choice of first-line TT (dabrafenib + trametinib) or CPI (nivolumab + ipilimumab), with crossover to the alternative treatment at radiological progression","primaryOutcome":"Feasibility endpoints: (1) proportion of critical ctDNA results returned within 7 days to enable a treatment switch decision; (2) proportion of Arm B patients commencing TT who achieved ≥80% reduction in BRAF VAF. Secondary outcomes included progression-free survival (PFS) and overall survival (OS).","followUp":"Not reported in abstract","fundingSource":"Not reported in abstract"}
Key findings
The trial met both co-primary feasibility endpoints. First, 100% of critical ctDNA results were delivered within 7 days (95% CI: 94–100%), confirming operational viability of real-time ctDNA-guided decision-making. Second, 100% of Arm B patients who commenced TT achieved the ≥80% BRAF VAF reduction threshold that triggered the switch to CPI (95% CI: 80–100%). No new or unexpected safety signals were identified for either the TT (dabrafenib + trametinib) or the CPI (nivolumab + ipilimumab) sequence in either arm. A post-hoc analysis of clinical and immunological parameters at the ctDNA nadir following TT induction in Arm B patients showed improvements in LDH, absolute neutrophil count, neutrophil-to-lymphocyte ratio, lymphocyte count, C-reactive protein, platelet count, and ctDNA VAF at the time of CPI initiation compared to baseline, as well as favourable changes in circulating cytokine and chemokine profiles—findings interpreted by the authors as suggesting a more immunologically permissive state at the point of the ctDNA-guided switch. PFS and OS data were reported as secondary outcomes but no numerical estimates, hazard ratios, confidence intervals, or p-values for survival endpoints are provided in the available abstract. NNT/NNH cannot be calculated from the reported data.
Practice implications
The CAcTUS trial establishes proof-of-concept that a ctDNA-guided, biomarker-driven switch from BRAF/MEK-targeted therapy to checkpoint immunotherapy is operationally feasible in BRAF-mutant advanced melanoma: ctDNA can be reliably assayed and results communicated within a clinically actionable 7-day window, and the ≥80% VAF reduction threshold is consistently achieved during TT induction. The trial does NOT establish whether this ctDNA-guided strategy improves PFS, OS, or response rates relative to standard-of-care sequential therapy, as it was powered exclusively for feasibility endpoints. The post-hoc observation that patients exhibit a more favourable immunological and inflammatory milieu at the ctDNA nadir is hypothesis-generating only; these findings are not pre-specified, not adjusted for multiplicity, and derive from a very small exploratory subsample (n=7 for cytokine data). The population studied (n=21 total) is insufficient to draw any efficacy conclusions. Patients with symptomatic brain metastases, poor performance status (ECOG >2), or prior combination CPI adjuvant therapy were excluded, limiting generalizability to those specific groups. NOTE: This summary is based on the published abstract only; the full-text manuscript may contain additional data, survival curves, and subgroup analyses not captured here. __LIMITATIONS__ Extremely small sample size (n=21 total; n=11 experimental arm): severely underpowered for any efficacy or survival inference.,Feasibility-only primary endpoints: no pre-specified hypothesis testing for PFS, OS, or response rate; survival data reported as secondary outcomes without quantitative results in the abstract.,Non-blinded, open-label design with potential for selection and assessment bias.,Early trial stoppage below planned target enrollment (40 patients were planned per ASCO 2021 protocol description), raising concerns about selection and attrition.,Post-hoc, exploratory nature of the immunological and cytokine findings limits any causal interpretation; no multiplicity adjustment reported.,Cytokine/chemokine subsample analysis limited to n=7 patients, markedly restricting statistical and clinical interpretability.,Absence of a radiological response-based comparison for the switch threshold in Arm B: ctDNA ≥80% VAF reduction does not have established concordance with RECIST response milestones in all patients.,Generalizability restricted to patients with detectable ctDNA at baseline (VAF ≥1.5%); patients with low ctDNA-shedding tumours were excluded.,Follow-up duration not specified in the abstract, precluding assessment of long-term durability.,Funding source not reported; potential for undisclosed conflicts of interest cannot be assessed. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ The CAcTUS trial demonstrates that using a ctDNA BRAF VAF ≥80% reduction threshold to prospectively trigger a switch from dabrafenib/trametinib to nivolumab/ipilimumab in BRAF-mutant advanced melanoma is operationally feasible and logistically deliverable within clinically relevant timeframes. It does not demonstrate whether this approach improves patient outcomes compared to standard sequential therapy, and post-hoc immunological signals are preliminary only. A larger, adequately powered randomised trial is required before ctDNA-guided sequencing can be considered for routine clinical practice.