"Liquid Biopsy" Can Detect Colorectal Cancer Resistance Noninvasively

KRAS mutations contribute to acquired resistance of colorectal cancers (CRCs) treated with targeted agents against epidermal growth factor receptor (EGFR). Two studies published this week in Nature characterize acquired KRAS mutations in colorectal cancer and identify some of the mechanisms by which mutations arise as a result of exposure to anti-EGFR antibodies. These results will help researchers make better decisions about future combination therapies to test for KRAS-mutated cancers.

Both studies demonstrate that a “liquid biopsy,” using plasma or blood samples, can be used to detect acquired resistance and potentially measure disease progression. The samples are used to detect and analyze circulating tumor DNA (ctDNA), which have previously been shown to be a reliable way to monitor CRC tumor dynamics. “We believe ctDNA does correlate with tumor burden.  Using highly sensitive digital approaches, ctDNA can be quantitated and tracks perfectly with tumor burden,” said Luis Diaz, of the Ludwig Center for Cancer Genetics and Therapeutics at Johns Hopkins Medical Center, and lead author of one of the studies.

“The collection of plasma can be done in any hospital,” said Alberto Bardelli, lead author of the other Nature study and professor at the University of Turin in Italy. While Bardelli highlighted that the molecular analysis of the samples is complicated, he believes that large oncology hospitals will rapidly take up the technique. “I suspect this will happen very quickly now.” Importantly, these results demonstrate that resistance can be detected earlier and noninvasively-very important in CRC as tumor samples are difficult to obtain.

The data suggests that resistance to EGFR inhibitors can be detected much earlier than the current approach of radiological evaluation which is typically 9 to 18 months after initiation of therapy. Earlier resistance detection would allow patients to receive a new therapy through a clinical trial, lower cost of expensive treatments, and provide more time for a metastatic CRC patient to be treated after tumor relapse is detected.

The first study, from Bardelli and colleagues shows acquired resistance to cetuximab (Erbitux) in colorectal cancers is frequently driven by KRAS mutations. Based on the results, the researchers, from the University of Turin in Italy, suggest taking systematic tumor samples from patients on anti-EGFR clinical trials to assess the actual prevalence of KRAS mutations and mechanisms of acquired anti-EGFR resistance.

The authors are able to detect the KRAS mutations using patient blood samples-a noninvasive way to monitor patient response and progression to treatment. The researchers also demonstrate that simultaneously targeting MEK, downstream of EGFR, and EGFR in an EGFR-resistant cell line results in cell death, shows promise for testing the combination of MEK and EGFR inhibitor upon early detection of EGFR-resistance.

A second study, from Diaz and colleagues shows mutations acquired as a result of anti-EGFR treatment can be detected noninvasively, by measuring ctDNA. 

Cetuximab is a chimeric monoclonal antibody against epidermal growth factor receptor. Cetuximab was approved for colorectal cancer back in 2004 under the US Food and Drug Administration’s (FDA) accelerated approval program. Cetuxiumab is able to block EGFR activity in the presence of a wild-type KRAS protein, a GTPase. If the CRC tumor harbors an activating mutation in KRAS, blocking EGFR will not prevent signaling through this pathway as KRAS is downstream of EGFR. In 2009, the FDA updated the label for cetuximab as well as panitumumab, another EGFR antibody approved in 2006 for CRC, dictating genetic testing of patients to check for the mutational status of the KRAS gene. Patients with a mutated KRAS gene-specifically in the 12th and 13th codon-do not respond to anti-EGFR treatment.

The study from the Italian group used both in vitro cell line data and sequencing analyses of patient tumor samples to demonstrate that tumor cells treated specifically with an EGFR-antibody, but not with chemotherapy, acquire resistance to the treatment via KRAS mutations. The researchers also found drug resistance occurs as a result of selection of already-present KRAS mutant cells. Both amplification and point mutations of the KRAS gene is causally responsible for cetuximab resistance using the in vitro models.

Sequencing data from patients showed KRAS mutations were detected in 0.4% to 1% of resistant tumors. One possible explanation is only a small fraction of tumor cells post-resistance, harbor a KRAS mutation. Bardelli and coauthors highlight a previous study showing unlike more sensitive techniques, Sanger sequencing could not detect a KRAS mutations in CRC patient samples that do not respond to anti-EGFR treatment, suggesting clinical drug resistance can occur if only a subset of tumor cells harbor a drug-resistant mutation.

Bardelli is currently testing two possible explanations why only a small fraction of the tumor cells harbor the KRAS-resistant mutation, both of which may end up being correct. “Resistant tumors are heterogeneous and multiple mechanisms coexist, which would explain the low percentage of KRAS mutant cells,” said Bardelli. “The second [explanation] is that the mutant resistant cells act in a paracrine fashion to support the surrounding cells.”

In the second research report, Diaz and colleagues used a ctDNA assay to measure KRAS mutations from 28 metastatic CRC patients who received treatment with panitumumab. Of 24 patients, 9 had detectable KRAS mutations (38%) post-EGFR treatment but wild-type KRAS prior to therapy with the EGFR inhibitor. In three of these patients, the KRAS mutations were detectable by ctDNA prior to radiographic evidence of disease progression-similar results to the Italian study. The time from detection of the KRAS mutations using ctDNA to evidence of disease progression was about 21 weeks.

A major reason for the lag of data on acquired resistance mutations in CRC is the inability to obtain substantial tumor samples post-treatment. “We found that most clinicians did not believe that resistance to cetuximab might have a genetic basis, and so it was hard to gather the samples,” said Dr. Bardelli. Furthermore, obtaining pre- and post-treatment tumor samples, and especially primary tumor samples from CRC patients is often difficult.

Analysis of blood samples to analyze ctDNA is promising. “Querying the ctDNA can provide a snapshot into the comprehensive genetic status of a patient with metastatic disease, whereas a biopsy only provides insight into the tumor tissue being sequenced,” said Diaz. “To reproduce the time course of our study with biopsies, it would have required over eleven hundred biopsies to interrogate every patient’s individual metastases over five to seven months.”

Bardelli and colleagues are working on identifying other mechanisms of resistance as KRAS mutations only explain 40% to 50% of relapses among CRC patients. The researchers are also working on overcoming resistance with combination therapies by targeting EGFR and MEK, simultaneously, early in the course of relapse, using the liquid biopsy assay as a guide.