Liquid Biopsy Detects KRAS Mutations in Plasma DNA in Nonresectable Pancreatic Cancer, Can Predict Patient Outcomes

May 2015, Vol 6, No 4

Philadelphia, PA—High levels of KRAS mutations in plasma circulating tumor DNA (ctDNA) predict worse overall survival (OS), whereas low levels of KRAS mutations in plasma ctDNA indicate improved OS in patients with advanced pancreatic cancer.

Because the majority of pancreatic tumors harbor KRAS mutations, analyzing plasma ctDNA levels of these mutations can help to predict patient outcomes, according to results of a prospective-retrospective study presented at the 2015 American Association for Cancer Research annual meeting.

The study also showed that combining baseline KRAS levels in plasma with KRAS levels after 2 cycles of chemotherapy improved the predictive ability of the liquid biopsy analysis compared with baseline KRAS alone.

“Monitoring KRAS levels in plasma and then combining baseline levels with longitudinal levels is a powerful predictor of overall survival,” stated coinvestigator Vlada Melnikova, MD, PhD, Vice President, Research and Development, Trovagene, San Diego, CA. (Lead investigator Julia S. Johansen, MD, DMSc, Clinical Professor, University of Copenhagen, Denmark, did not attend the meeting.)

Liquid biopsies, such as the plasma biopsy, have several advantages over tissue biopsies. Liquid biopsies are easier to obtain and are less costly, because they do not involve invasive procedures. In addition, tissue biopsies can degrade, and repeated biopsy is required in some cases, and the tissue may not be available for biopsy.

Furthermore, biopsies can be required at different stages as a cancer progresses, and it would be more attractive to physicians, patients, and payers to be able to use serial liquid biopsies rather than tissue biopsies.

The prospectively collected samples were retrospectively analyzed and were from patients participating in the Danish BIOPAC study, Dr Melnikova explained. The study was based on archived plasma samples from 182 patients (85 females and 97 males) with advanced or metastatic pancreatic cancer who were treated with gemcitabine or with FOLFIRINOX (leucovorin, 5-fluorouracil, irinotecan, and oxaliplatin), the standard of care when the study was initiated.

The assay used to assess the 640 plasma samples from the 182 patients at baseline and after 2 cycles of chemotherapy was developed by Trovagene, and is highly sensitive for the detection of KRAS mutations in plasma ctDNA. Of 182 patients, 176 had evaluable plasma samples and baseline.

Predicting Survival

Using a cutoff point of 5.5 copies per 100,000 for low and high KRAS levels in plasma ctDNA, a statistically significant negative association was found between KRAS levels and OS (P <.001). Among the patients who received gemcitabine, a median survival of 296 days for those with low baseline KRAS levels compared with 148 days for those with high baseline KRAS levels.

Among the patients who received FOLFIRINOX, the median survival was 499 days for those with low baseline KRAS levels and 210 days for those with high baseline KRAS levels.

The investigators analyzed KRAS mutational load in 176 patients with serially collected plasma samples during treatment, and a similar association was observed. High KRAS counts during treatment were significantly associated with lower OS (P <.001). The strength of this association was greater for postbaseline KRAS counts than for the baseline KRAS counts.

Combining baseline KRAS and longitudinal KRAS data after 2 cycles of chemotherapy was an even better predictor of survival than baseline KRAS alone.

The patients were divided into 4 groups—low baseline KRAS and low longitudinal KRAS (responders), low baseline KRAS and high longitudinal KRAS (nonresponders), high baseline KRAS and high longitudinal KRAS (nonresponders), and high baseline KRAS and low longitudinal KRAS (responders).

Based on combining KRAS levels in plasma ctDNA at baseline and after 2 cycles of chemotherapy, the median survival for each of the 4 groups was 336 days, 252 days, 224 days, and 134 days, respectively.

Commercializing the Test

The company is in the process of introducing the plasma test, but there is no indication yet whether Medicare will reimburse for this test under the not otherwise specified code. The test will be priced at approximately $1300, but insurers may pay less for it.

Other companies are planning to market or are already marketing plasma tests (ie, “liquid biopsies”) to identify genetic mutations.

The company is also rolling out an assay to detect KRAS mutations in urine. A pilot study also presented at the meeting showed that the urinary and plasma assays to detect KRAS mutations in ctDNA were concordant with tissue biopsies of patients with metastatic colorectal cancer, and thus can detect minimal residual disease.

The next step will be to study whether liquid biopsies are prognostic for survival in this group of patients.

Related Articles