High EGFR ligand expression predicts panitumumab benefit

Last Updated: 2016-02-16

By Will Boggs MD

NEW YORK (Reuters Health) - Expression levels of the epidermal growth factor receptor (EGFR) ligands epiregulin (EREG) and amphiregulin (AREG) predict which patients with RAS wild-type advanced colorectal cancer are likely to benefit from panitumumab therapy, researchers from the UK report.

"Most clinicians are used to the concept that advanced colorectal cancer (CRC) patients whose cancers are fully RAS wild-type should, at some stage in their disease course, be offered an anti-EGFR monoclonal antibody (mAb) and have a high chance of benefiting," Dr. Matthew T. Seymour, from St. James's University Hospital, Leeds, told Reuters Health by email.

"However, what we see here is that the 55% of RAS-wt patients who have low ligand expression obtain no benefit; all the benefit of panitumumab comes in the 45% of RAS-wt patients with high ligand expression," Dr. Seymour said.

Previous studies have suggested that AREG or EREG may identify patients who will or will not benefit from anti-EGFR therapy, but none has validated their prognostic role.

Dr. Seymour and colleagues used data from the PICCOLO trial to examine the interactions between ligand expression and treatment benefit of panitumumab in patients with advanced CRC.

Their study included 323 patients with sufficient data, 220 (68.1%) of which were wild-type across all KRAS and NRAS codons (RAS wt).

Among the 99 RAS wt patients with high expression of either AREG or EREG, panitumumab therapy had a significant beneficial effect on progression-free survival and response rate, whereas panitumumab was ineffective among the 121 RAS wt patients with low expression of either ligand.

Patients with RAS or BRAF mutations gained no panitumumab therapy benefit by any end point, according to the February 11 JAMA Oncology online report.

Separately, AREG expression predicted the beneficial effect of panitumumab on progression-free survival, whereas EREG expression did not. Neither ligand predicted a benefit of panitumumab on overall survival.

Using only AREG expression, however, would mean withholding treatment from a small group of patients with isolated high EREG for whom treatment is uncertain from these data.

"The main message is that we can do a lot better than use RAS testing alone to select patients for anti-EGFR therapy in CRC," Dr. Seymour said. "The parallel is with anti-HER2 drugs in breast and stomach cancer, where you can select just 15-25% patients who will get worthwhile benefit, and avoid treating patients who will not. In advanced CRC, similarly, around 20% of patients have RAS-wt, high-ligand cancer."

"PICCOLO was a big phase 3 trial but even so, by the time we had homed in on RAS-wt patients randomized to +/- panitumumab with available tissue for analysis we were down to quite a small number (220) for the key primary analysis," he said. "So we really need for everyone who is sitting on a large (randomized trials) biobank where patients were randomized to +/- anti-EGFR mAb to go back and analyze those samples for AREG and EREG and see if they get the same result."

"It is likely that AREG/EREG is not the end of the story," Dr. Seymour added. "We presented some data at ASCO (American Society of Clinical Oncology) last year looking at HER3, which is highly promising as a way of refining the algorithm even further."

"How the results of future confirmatory work could be interpreted may depend on the context," write Dr. David Cunningham and colleagues from Royal Marsden Hospital, London, UK, in a related editorial. "In the United Kingdom, discretionary funding for the use of anti-EGFR therapies outside the first-line setting has recently been withdrawn. However, if a robust predictive marker could identify a smaller group of patients for therapy leading to a significant survival benefit, this could conceivably result in a successful funding application."

"Conversely," they note, "in non-resource-constrained funding environments, it is less likely that AREG and EREG status may be interpreted in the same way and the negative predictive value of the assay may be more valuable."

"Although the current model requires further validation, it represents an important refinement of our understanding of the use of anti-EGFR therapies for patients with advanced CRC," the editorial concludes.

Dr. Alberto Bardelli from Candiolo Cancer Institute, Torino, Italy, who recently reviewed lesion-specific response to targeted therapy in colorectal cancer (http://bit.ly/1SQYnSJ), told Reuters Health by email, "The impact of EGFR ligands on response to EGFR blockade is not novel. The important aspect of the paper is that it may now be possible to use the (polymerase chain reaction) test in the clinical setting."

"Independent prospective assessment of the findings (and the technology to detect the ligands RNA)" is needed before adopting these biomarkers for therapeutic decision making, Dr. Bardelli said.

Cancer Research UK and Yorkshire Cancer Research supported this research. Two coauthors reported disclosures.

SOURCE: http://bit.ly/1Tm9ce4 and http://bit.ly/1XvT5Ly

JAMA Oncol 2016.