These studies open the door for future clinical investigations to further validate the usefulness of ctDNA as both a surveillance and prognostic tool in gynecologic cancers.
—John A. Martignetti, MD, PhD
A new study has demonstrated for the first time that personalized circulating tumor DNA (ctDNA) biomarkers in gynecologic cancers can detect the presence of residual tumor earlier than currently used serum and imaging studies.1,2 According to the data, undetectable levels of ctDNA at the completion of primary therapy were also an independent predictor of survival.
“Detection of residual disease represents a diagnostic dilemma and potential critical inflection point in precision medicine,” said John A. Martignetti, MD, PhD, of Icahn School of Medicine at Mount Sinai. “Ultimately, these studies open the door for future clinical investigations to further validate the usefulness of ctDNA as both a surveillance and prognostic tool in gynecologic cancers.”
As Dr. Martignetti reported at the 2016 Society of Gynecologic Oncology Annual Meeting on Women’s Cancer, the measurement of ctDNA—the so-called liquid biopsy—represents a powerful emerging technology capable of providing accurate assessment of both tumor behavior and disease burden. Moreover, a personalized test, unique to each patient, can be specifically designed for use.
“This is a noninvasive assessment,” said Dr. Martignetti. “It’s a blood draw. You can monitor patients throughout the disease course and in real time. And if you’re monitoring the mutations, you can monitor the evolution of tumor growth, response to treatment, acquired resistance to treatment, and development of tumor heterogeneity.”
John A. Martignetti, MD, PhD
For this single-institution study led by Dr. Martignetti, Peter Dottino, MD (Director of Gynecologic Oncology, Mount Sinai Health System), and Elena Pereira, MD, the team enrolled 44 cancer patients (24 ovarian, 17 uterine, 2 fallopian tube, and 1 peritoneal carcinoma). Serum and tumor samples were collected at the time of surgery and then throughout the treatment course. Tumor-specific mutations were identified for each patient using either whole-exome or targeted gene sequencing, and ctDNA levels were quantified using droplet digital polymerase chain reaction. These levels were then correlated with clinical disease status, CA-125 values, computed tomography (CT) scan results, surgical findings, and progression-free and overall survival
“We sought to explore the usefulness of ctDNA as both a surveillance and prognostic biomarker and to compare ctDNA efficiency, sensitivity, and lead time against serum CA-125 values and CT scans,” Dr. Martignetti explained. “To be enrolled in the study, each patient had to have blood samples drawn for ctDNA analysis within several weeks of either their CT scans or surgery. This allowed us to directly compare our ctDNA biomarker test results to useful and commonly accepted standards.”
Surveillance and Prognostic Biomarker
Researchers found that the sensitivity and specificity of ctDNA, detected in 93.8% of patients, were highly correlated with CA-125 and CT scanning, with a number of notable exceptions.
“We saw that ctDNA correlates with tumor presence and is in general as sensitive and specific as CA-125,” said Dr. Martignetti. “A powerful distinction in ctDNA as a biomarker was that, whereas only 60% of individuals at the time of surgery or recurrence had abnormal CA-125 levels, almost all patients under these same conditions had detectable ctDNA levels.”
More interesting to the researchers, however, were the six patients with negative CT scans, in whom circulating tumor DNA successfully identified the presence of cancer. “We had six individuals in our study population who had no evidence of disease on CT but positive traces of ctDNA,” said Dr. Martignetti. “All of these patients had disease progression as defined by tumor recurrence within 1 to 11 months of positive ctDNA and negative imaging. In other words, CT scan failed to detect the cancer that the ctDNA did.” On average, circulating tumor DNA had a predictive lead time of 7 months over CT imaging.
Dr. Martignetti also highlighted the potential for using ctDNA as a prognostic biomarker, as undetectable levels of ctDNA at the completion of primary therapy were associated with markedly improved progression-free survival and overall survival (P = .0048 and P = .0194, respectively).
“If we did not detect that patient’s unique ctDNA signature at the completion of primary chemotherapy,” said Dr. Martignetti, “the patients had an outstanding overall survival profile, whereas those with detectable ctDNA levels did not.”
He concluded, “These are relatively small studies. Ultimately, we need to demonstrate the clinical utility, but I think they’re a great starting point.” ■
Disclosure: Dr. Martignetti reported that RainDance Technologies, Inc, and Swift Biosciences provided technical and financial support for reagents used in these studies.
In the discussion session, Douglas A. Levine, MD, Head of Gynecology Research Laboratory at Memorial Sloan Kettering Cancer Center, said that while advanced genomics can help to predict outcomes, stratify treatment, and understand biology, one of the problems of precision medicine is that...