ctDNA Predicted Early Relapse in Resected Stage II/III Melanoma

November 18, 2017

Levels of circulating tumor DNA predicted worse outcomes including relapse and survival in patients with resected stage II/III melanoma, according to the results of a study.

Levels of circulating tumor DNA (ctDNA) predicted worse outcomes including relapse and survival in patients with resected stage II/III melanoma, according to the results of a study published in Annals of Oncology.

Using droplet digital polymerase chain reaction, Rebecca J. Lee, of Cancer Research UK Manchester Institute, and colleagues detected for BRAF and NRAS mutations in plasma taken from 161 patients with high-risk stage II/III melanoma after surgery. They found that patients with higher levels of ctDNA had significantly worse disease-free interval, distant metastasis–free interval, and 5-year overall survival (OS).

“The majority of patients with detectable ctDNA relapsed within 1 year of surgery suggesting that ctDNA in the plasma appears to reveal occult metastatic disease that is not evident on radiological imaging,” the researchers wrote. “Notably, we were able to identify melanoma patients at high risk of both distant metastatic relapse and local recurrence, which is consistent with studies showing that ctDNA can signal micrometastatic disease after neoadjuvant chemotherapy post surgical resection in breast cancer, and following surgery for stage II colorectal cancer.”

Lee and colleagues collected tumor samples as part of the AVAST-M trial, which compared bevacizumab and placebo in 1,343 patients with high-risk resected stage II/III melanoma. In this retrospective analysis, they analyzed ctDNA carrying either BRAF or NRAS mutation in the baseline resected tumor. BRAF mutation was found in 132 samples and NRAS in 29.

The researchers detected ctDNA in 19 (12%) of the plasma samples. Of the positive samples, 15 had a BRAF mutation and 4 had an NRAS mutation.

Those patients with detectable ctDNA had a significantly decreased disease-free interval (hazard ratio [HR], 3.12; 95% CI, 1.79–5.47; < .0001) and distant metastasis–free interval (HR, 3.22; 95% CI, 1.80–5.79; < .0001) compared with those samples with undetectable levels of ctDNA. The median disease-free interval was 0.3 years for detectable ctDNA compared with 4.2 years for undetectable ctDNA. Median distant metastasis–free interval was 0.6 years for detectable ctDNA compared with not yet reached at 5 years follow-up for undetectable ctDNA. These differences remained significant even after researchers adjusted for performance status and disease stage.

In addition, patients with detectable ctDNA had a 5-year OS of only 33% compared with 65% in those with undetectable ctDNA. OS was significantly worse for detectable ctDNA (HR, 2.63; 95% CI, 1.40–4.96; = .003) even after adjustment for performance status (HR, 2.50; 95% CI, 1.32–4.74; = .005). Median OS was 2.9 years for detectable ctDNA compared with not reached at 5 years follow-up for undetectable ctDNA.

“The ability to predict progression to stage IV disease is extremely important in light of recent findings that immune checkpoint inhibition improves OS in stage III melanoma,” the researchers wrote. “Detection of ctDNA allows identification of a subgroup of patients at high risk of early relapse and inferior survival, allowing stratification of patients to adjuvant regimens associated with higher toxicity but greater potential for efficacy.”