NEW YORK (GenomeWeb) – Researchers have been exploring droplet digital PCR (ddPCR) technology as a promising tool for finding and following cell-free circulating tumor DNA (ctDNA) profiles associated with immunotherapy response, including delayed treatment responses in individuals who initially experience tumor stability or growth, known as pseudoprogression.
Numerous tumor and immune cell features have been proposed as potential markers for response to immune checkpoint inhibitors. But the need for validated markers often continues after physicians have selected a treatment strategy, particularly when trying to decide whether to press on or abandon treatment schemes that do not seem to be yielding ideal results in routine tumor scans and related "Response Evaluation Criteria in Solid Tumors" (RECIST) scores.
Liquid biopsies have gained attention as a means of following treatment response and tumor evolution in general. A handful of recent studies suggest ctDNA dynamics may also offer an avenue for tracking and potentially tailoring immunotherapy treatments — an approach that relies on the ability to reliably detect ctDNA, when present, in serial liquid biopsy samples.
That detection can be achieved with ddPCR assays that target notorious driver mutations — the BRAF V600E mutation in melanoma, for example — but such digital assays could also be more personalized to a given individual's tumor, explained George Karlin-Neumann, Bio-Rad Laboratories' director of scientific affairs. Bio-Rad offers a number of validated ddPCR assays.
"Depending how much was known about the tumor, there could be patient-specific markers that could be learned during, say, a sequencing profiling that then might be converted into a digital assay for serial monitoring," Karlin-Neumann said. He noted that the speed, cost, and sensitivity of ddPCR offers an advantage in situations where multiple samples per person are analyzed.
In a Clinical Cancer Research study published early last year, for example, National Cancer Institute researcher Mark Raffeld and colleagues used the BRAF V600E mutation as a marker to follow ctDNA with competitive allele-specific TaqMan PCR in longitudinal blood samples from four dozen individuals with metastatic melanoma who were treated with adoptive T-cell immunotherapy.
Their results revealed an early jump in cell-free tumor DNA in the blood of individuals with the most pronounced response to adoptive, activate autologous tumor-infiltrating lymphocyte treatment: those with complete responses one or two years after treatment had an initial rise in ctDNA, followed by rapidly declining ctDNA levels in the V600E-mutation positive metastatic melanoma patients.
"Responding patients generally do not have significant changes in non-cutaneous RECIST targets before 30 to 60 days following [tumor-infiltrating lymphocyte] infusion, and complete responses are often not confirmed for [one to two] years," the authors of that study wrote, noting that there is "critical need for a biomarker that can provide early information regarding the likelihood and duration of a response to enable rational decisions about altering therapy."
The analyses also highlighted ctDNA kinetic profiles associated with other response types, including partial response and non-response, suggesting that ctDNA dynamics may provide early opportunities to tweak or fine-tune an individual's treatment, noted Karlin-Neumann, who was not involved in the research. He called the study "quite exciting, if it's reproducible."
Researchers from the University Medical Center Groningen reported a similar early spike in ctDNA levels among individuals with advanced non-small cell lung cancer who ultimately responded to the anti-PD-1 drug nivolumab (Bristol-Myers Squibb's Opdivo) in a poster presented at the American Society of Clinical Oncology annual meeting in Chicago last month.
That team focused on a specific KRAS mutation identified in six lung cancer patients' primary tumors for its prospective analysis, using ddPCR to profile ctDNA in blood samples collected prior to treatment and one, two, four, and six weeks after treatment began. After six weeks of treatment, samples were collected every three months or so. The four responding patients had rising ctDNA levels one week into treatment, followed by waning ctDNA detection with KRAS mutation-focused ddPCR.
If such studies pan out, some researchers suspect that ctDNA dynamics may help in determining when an individual's immunotherapy treatment is on track — even when tumor imaging seems to say otherwise.
An Australian team took a slightly different tack, presenting results in a poster at the ASCO conference: Macquarie University senior lecturer Jenny Lee and her colleagues the University of Sydney, the Melanoma Institute Australia, and elsewhere used ddPCR to try to tease out ctDNA profiles characterizing metastatic melanoma patients who had pseudoprogression — and eventual treatment response — on anti-PD-1 checkpoint blockade immunotherapy.
"We compared those patients who had a radiological pseudoprogression, which is defined by initial growth in the tumor followed by shrinkage, to true progression — those patients who had growing lesions that continued to grow on subsequent scans," Lee said.
In a group of 29 metastatic melanoma patients who received a PD-1 inhibitor, the researchers identified nine individuals with pseudoprogression, characterized by clinical progression or by RECIST scores that were indicative of progressive disease after treatment, followed by treatment response.
Seven of the individuals with pseudoprogression had ongoing response after an average of nearly two years of follow up, while the remaining two individuals had subsequent disease progression after seven or 18 months, the team reported.
With the help of ddPCR testing for patient-specific BRAF or NRAS mutations initially identified in tumor tissue, the researchers followed ctDNA in blood samples collected over the first three months of treatment, usually six to eight weeks into treatment and again at week 12.
From those data, they found that pseudo-progressing patients typically had low or dramatically declining ctDNA levels at baseline or over the first few months of treatment, while individuals with true disease progression showed stable or enhanced ctDNA levels three months into the treatment.
While such findings might seem to be at odds with the early ctDNA rise in immunotherapy responders reported by other groups, Lee explained that most of the blood samples included in the analysis were collected several weeks into the anti-PD-1 treatment. But an apparent rise in ctDNA was present in two pseudo-progressing patients who had blood tests just one week into treatment.
"I think for a lot of our patients we actually missed that initial spike," she said. "But regardless, we felt that a week six blood draw was more accurate in predicting overall response in our patients, as well as survival."
Although the phenomenon is relatively rare overall, the existence of pseudoprogression suggests tumor imaging alone may not provide enough information to decide whether a treatment strategy is on the mark. Consequently, the ability to use ctDNA as another window into predicted patient outcomes may be beneficial, especially for treatment types where pseudoprogression is more common, such as melanoma.
"It's really important to be able to accurately differentiate these patients from patients with true progression," Lee said. "It really does not only give us prognostic information, but also changes patients' management."
The pseudoprogression analysis presented in the ASCO poster stemmed from a larger ctDNA study on more than 100 individuals, who were tested at baseline and during treatment with anti-PD-1 therapy alone or in combination, Lee explained.
She and her colleagues are also using ddPCR to follow ctDNA patterns in individuals with brain metastases, in patients with earlier stage cancers, and in individuals receiving different treatment types.
"By using [ctDNA] in different applications, we're trying to understand the biology of this marker but also, in the process, trying to push for using this technique as a biomarker in the clinical setting and not just as a research tool," Lee said.