Early ctDNA Changes Predict Survival in Advanced Melanoma Treated With Immunotherapy

Decreases in circulating tumor DNA (ctDNA) within the first month of immune checkpoint inhibitor (ICI) therapy were strongly associated with higher response rates and longer survival in patients with unresectable stage III/IV melanoma, according to a multi-institutional analysis published in the Journal of Clinical Oncology Precision Oncology. In the retrospective cohort of 117 patients treated with anti–PD-1–based regimens, a reduction in ctDNA levels 3 to 4 weeks after treatment initiation was associated with markedly higher odds of objective response and disease control, as well as significantly improved progression-free survival (PFS) and overall survival (OS).

The study evaluated real-world data from 3 US academic centers between August 2021 and August 2024. Eligible patients had histologically confirmed unresectable stage III or IV melanoma and were receiving anti–PD-1–based therapy, either nivolumab (Opdivo) or pembrolizumab (Keytruda) monotherapy or nivolumab in combination with relatlimab (Opdualag) or with ipilimumab (Yervoy). The majority (57.3%) received nivolumab/ipilimumab with 23.1% receiving anti–PD-1 monotherapy and 19.7% receiving nivolumab/relatlimab.

All patients underwent tumor-informed ctDNA testing using the personalized Signatera assay at baseline and again at 3 to 4 weeks, before the second treatment dose. Radiographic response was assessed retrospectively using RECIST version 1.1 criteria.

At a median follow-up of 13.4 months, 65% of the 117 patients demonstrated a decrease in ctDNA at 3 to 4 weeks. Compared with patients whose ctDNA level increased, those with a decrease had significantly higher odds of disease control (OR, 30.56; 95% CI, 10.64-87.74; P < .001) and objective response (OR, 23.54; 95% CI, 8.58-64.57; P < .001).

Survival outcomes were similarly differentiated. A ctDNA decrease was associated with improved PFS (HR, 0.18; 95% CI, 0.11-0.31; P < .001) and OS (HR, 0.28; 95% CI, 0.13-0.56; P < .001). The 12-month PFS was 67.6% in the ctDNA-decrease group vs 23.1% in the ctDNA-increase group. The 12-month OS rate was 77.5% vs 43.5%, respectively.

Patients with ctDNA increase of at least 20% at 3 to 4 weeks had an increased risk of progression or death (HR, 7.25; 95% CI, 2.79-18.82; P < .001) and of death (HR, 7.35; 95% CI, 1.71-31.47; P = .07). Patients who had ctDNA increase early on had a median PFS of just 2.3 months.

Among patients with an early ctDNA decrease, those who ultimately achieved undetectable ctDNA had significantly improved PFS (HR, 0.14; P < .001) and OS (HR, 0.07; P < .001) compared with those without clearance. Among patients with early decreases, those who ultimately achieved a complete ctDNA clearance had a 1-year overall survival exceeding 90%.

Nine of the 41 patients with ctDNA increase later achieved ctDNA clearance; 6 of the 9 received interventions including radiotherapy and palliative resection. Of 16 patients with undetectable baseline ctDNA, 15 remained undetectable and 1 developed detectable ctDNA at 3-4 weeks.

Notably, baseline ctDNA levels alone were not predictive of objective response, underscoring the importance of dynamic change rather than absolute value.

A significant challenge in immunotherapy is the phenomenon of pseudoprogression, where imaging suggests disease growth due to immune-related inflammatory changes rather than actual tumor proliferation. In a subset of patients where imaging indicated progression, a concurrent decrease in ctDNA levels from baseline at 3 to 4 weeks accurately predicted that the radiographic changes were inflammatory. This suggests that ctDNA dynamics can serve as a critical differentiator between true disease progression and immune-mediated responses.

Although radiographic imaging remains the current standard of care for monitoring treatment, ctDNA offers a "window into treatment efficacy" weeks before traditional scans can provide a clear picture. Radiographic response assessment using RECIST 1.1 typically occurs after 2 to 4 months, and immune-related response patterns such as pseudoprogression complicate interpretation.

An early increase in ctDNA could serve as a trigger for prompt treatment escalation, earlier clinical trial enrollment, or more frequent radiographic monitoring. Conversely, early ctDNA clearance or a significant decrease provides a rationale for continuing current therapy, offering reassurance during equivocal imaging results, and potentially informing future de-escalation strategies.

The transition from a prognostic tool to a predictive one that guides active treatment adaptation requires further validation. Although the retrospective data is compelling, prospective clinical trials are essential. These trials will determine if adapting treatment based on ctDNA signals can directly improve patient survival and quality of life. The study was retrospective and included heterogeneous treatment regimens. The median follow-up of 13.4 months limits long-term survival interpretation.