Research Backs Anti-EGFR Therapy in mCRC With Wild-Type KRAS

New research on predictive and prognostic biomarkers for therapy targeting the epidermal growth factor receptor (EGFR) strongly suggests that panitumumab should be reserved for metastatic colorectal cancer (mCRC) tumors with wild-type KRAS. The investigators stated that panitumumab should not be used in tumors with KRAS mutations, even those with specific G12 and G13 mutations previously suggested to confer a favorable response to anti-EGFR therapy.

“The take-home message from our study is that patients with mCRC tumors that harbor any of the most common codon 12 or 13 mutant KRAS alleles are unlikely to benefit from panitumumab therapy. Currently, only mCRC patients with wild-type KRAS tumors should be treated with EGFR antibodies,” stated Marc Peeters, MD, PhD, Antwerp University Hospital, Belgium.

“In general, patients with mutant tumors do not benefit from EGFR inhibitors. The discussion [about mutations] is still open from a scientific point of view,” Peeters said. “More research is needed.”

The study explored the predictive and prognostic impact of the seven most common codon 12 and 13 KRAS mutations on progression-free survival (PFS), overall survival (OS), and overall response rate (ORR) in patients with mCRC treated with panitumumab or control therapy.

The patient population was derived from three phase III placebo-controlled studies, all showing a significant advantage for panitumumab in wild-type KRAS mCRC tumors. The first study compared panitumumab plus FOLFOX4 versus FOLFOX4 alone in the first-line setting; the second study compared panitumumab plus FOLFIRI versus FOLFIRI alone as second-line therapy; and the third study compared panitumumab plus best supportive care versus best supportive care alone.

The distribution of the seven common mutations in the study population was compared with a large mutation database of more than 9000 patients with mCRC and found to be consistent with the public database, Peeters said.

The investigators then looked at the distribution of the seven mutations in the study population according to demographic factors, and no imbalance in the distribution of mutations was observed in the various demographic categories.

Looking at the impact of mutations in the control arms of the three phase III studies, a negative prognostic effect was observed for G12C related to FOLFOX4 for PFS but not for OS.

In the wild-type population, for PFS and OS, G12V significantly favored the panitumumab arm (P = .037), and G13D significantly favored the control arm (P = .002).

In a pooled analysis of all three panitumumab arms from the phase III studies, only G12A showed a relationship that favored the control arm.

“For the prognostic effect, there was no consistent association for any allele with PFS or OS outcomes in control arm-treated patients. For the predictive effect, there was no consistent evidence that any individual KRAS allele compared to the remaining mutant KRAS alleles or the entire mutant KRAS group differentially impacted response or survival,” Peeters stated. “In the pooled analysis of three phase III trials, G12A was the only mutant allele significantly associated with a negative panitumumab treatment effect on OS.”

Historical studies of another anti-EGFR therapy, cetuximab (Erbitux), were based on more heterogeneous patient populations, and it is difficult to draw conclusions about the prognostic and/ or predictive value of mutations in those studies, Peeters explained.

Evidence regarding the impact of other mutations in CRC suggests that their impact is low, he continued. “We don’t have analysis in a standard setting of other mutations outside the codon 12 and 13 regions,” according to Peeters.

Weighing in on the discussion, Sabine Tejpar, MD, UZ Leuven, Belgium, said: “To study that, you need extra-large randomized studies. We are not there yet.”

Peeters M, Douillard J, Van Custem E, et al. Mutant (MT) KRAS codon 12 and 13 alleles in patients (pts) with metastatic colorectal cancer (mCRC): assessment as prognostic and predictive biomarkers of response to panitumumab (pmab). J Clin Oncol. 2012;30: (suppl 4; abstr 383).