NEW YORK – A new study in advanced melanoma patients by researchers in Germany has found that tumor mutation burden (TMB), circulating tumor DNA (ctDNA), and cell-free DNA can predict response to combined immunotherapy and overall survival under the treatment.
In particular, treatment failure could be detected as early as three weeks after the start of therapy using ctDNA and cfDNA, which might help doctors and patients decide whether to continue treatment if adverse events occur early.
The study, led by researchers at German molecular diagnostics company CeGaT and University Hospital Tübingen and published last week in the Journal for ImmunoTherapy of Cancer, explored whether TMB, cell-free DNA, and ctDNA can be used to estimate response to a combination of two Bristol-Myers Squibb drugs, ipilimumab (Yervoy), a CTLA-4 antibody, and nivolumab (Opdivo), a PD-1 antibody, in metastatic melanoma.
The researchers included 35 late-stage melanoma patients in their prospective study. They started by analyzing their tumors – in most cases, recently removed metastases – and matched controls using a 710-gene custom sequencing panel of tumor-associated genes covering more than 2 Mb of DNA. From this, they determined both the presence of specific somatic driver mutations and calculated TMB.
They also took blood samples every three to four weeks from the start of treatment and determined both the concentration of cell-free DNA and, using patient-specific digital droplet PCR assays, the amount of ctDNA present. Genes in which they tracked somatic mutations included BRAF, CDK4, GNAQ, JAK2, KRAS, MAP2K1, NF1, NRAS, and STAT1, and the ddPCR assays were either commercially available from Thermo Fisher Scientific or Bio-Rad Laboratories or custom-ordered from these firms.
Based on previous studies, they classified patients into three TMB groups: low TMB with less than 3.3 mutations/Mb, intermediate TMB with 3.3 to 23.1 mutations/Mb, and high TMB above 23.1 mutations/Mb.
Overall, median TMB in responders was significantly higher than in non-responders. Of note, all patients with high TMB happened to be male, and women had worse survival than men that could not be explained completely by the TMB difference.
Also, a 50 percent or greater decrease in cell-free DNA concentration three weeks after treatment initiation was significantly associated with therapy response. In addition, the absence of detectable ctDNA at this first follow-up appointment was more common in responders. Conversely, only four of 18 patients with detectable ctDNA at the first follow-up responded to combined immunotherapy.
Combining the three measurements, the researchers found that no patients with intermediate or low TMB who also had either detectable ctDNA or a cell-free DNA increase of more than 50 percent at the first follow-up responded to combined immunotherapy. "These results offer the possibility of estimating a therapy response at a very early time point, already at first follow-up," the researchers wrote. "In case of severe adverse events early after treatment initiation, this could be helpful in the decision whether to continue immunotherapy or not."
In terms of survival, those with a greater than 50 percent increase in cell-free DNA or detectable ctDNA at the first follow-up had poorer overall survival, and those with high TMB showed prolonged survival. Looking at the three variables in combination, overall survival was worse in patients with intermediate or low TMB who also had either detectable ctDNA or a 50 percent or greater increase in cell-free DNA.
The researchers pointed out that the concentration of cell-free DNA is often easier to determine than the amount of ctDNA, in particular when no patient-specific ddPCR assay is available. In addition, clones with a specific mutation measured by a ddPCR assay can go away under therapy while other clones might keep growing. "Using both markers together can thus help to increase the sensitivity of detecting progress," they wrote.
Their results need to be validated in a larger cohort, they added, noting also that the follow-up time of their cohort was relatively short.