Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications
Next-generation sequencing technology is being used on Phase III trial data.
On August 11, Amgen reported that adding Vectibix (panitumumab) to platinum-based chemotherapy did not create a statistically significant improvement in overall survival in a Phase III trial. The antibody was being tested as a first-line treatment in patients with recurrent and/or resistant metastatic squamous cell head and neck cancer.
An additional approval for Vectibix in head and neck cancer would have significantly expanded the market for the drug; about 400,000 new head and neck cancer cases are reported annually worldwide. The drug was sanctioned in 2006 for EGFR-positive metastatic colorectal cancer (mCRC) after disease progression occurred in patients being treated with fluoropyrimidine-, oxaliplatin-, and irinotecan-containing chemotherapy regiments. In November 2009, the company also did not succeed in Phase III testing to expand the treatment into the first-line setting for mCRC.
Vectibix competes with Erbitux (cetuximab), marketed by Eli Lilly and Merck KGaA, which is approved for both EGFR-expressing mCRC as well as head and neck cancer. Erbitux is a chimeric mouse/human mAb, while Vectibix an entirely human mAb.
Eli Lilly recognized $390.8 million in revenue from U.S. sales of Erbitux in mCRC and head and neck cancer during 2009. Outside the U.S. in 2009, sales of Erbitux increased 23% to €697 million (about $910.6 million), according to Merck. Sales of Vectibix in mCRC totaled $233 million in 2009 versus $153 million in 2008; Vectibix makes up roughly 6% of Amgen’s sales revenue.
While both Vectibix and Erbitux require intravenous administration, Erbitux is administered weekly and Vectibix, every two weeks, a feature that allows Amgen to sell the drug for about 20% less than Erbitux. But Vectibix is still pricey: An infusion every two weeks costs $4,000, or more than $100,000 for a year’s treatment.
KRAS: A Marker for Lack of Response
The availability of an unassailable predictive biomarker for treatment of colorectal cancer patients with Vectibix and Erbitux has made testing for a specific gene mutation a treatment prerequisite. This has given physicians a new comfort level in identifying candidates for these expensive treatments. To be sure, there is no convincing evidence suggesting that KRAS mutations are independent prognostic factors for CRC. But KRAS mutation status has been established as a predictive marker for treatment with EGFR inhibitors. Specifically, KRAS mutations serve as a biomarker for lack of response to anti-EGFR treatment.
The KRAS gene encodes a small GTPase protein on the EGFR pathway. Anti-EGFR antibody therapies work by blocking the activation of EGFR to prevent downstream events that lead to malignant signaling. It is thought, however, that in patients with tumors harboring a mutated KRAS gene, the KRAS-encoded GTPase protein is always turned on, regardless of whether EGFR is activated or therapeutically inhibited.
An economic analysis presented at the “Gastrointestinal Cancers Symposium” last year found that up-front testing for KRAS status and limiting anti-EGFR therapy to patients with wild-type KRAS tumors can result in drug cost savings of $604 million for Erbitux alone, according to lead author Veena Shankaran, M.D., from the Northwestern University Feinberg School of Medicine.
Moving beyond mCRC
Vectibix’ approval in mCRC was based in no small part on retrospective analysis of clinical trial data indicating that only patients with the wild-type KRAS gene, that is, without mutations in the gene, could respond to Vectibix.
In the head and neck cancer trial, reported last month, Vectibix prolonged patients’ median survival period to 11.1 months compared to 9 months with chemotherapy alone, but the improvement in overall survival was not statistically significant. Vectibix improved progression-free survival (5.8 months versus 4.6 months) and objective response rate (36% versus 25%), both secondary endpoints, but these were not tested for statistical significance.
In November 2009, Amgen reported that the Phase III PRIME 203 trial evaluating Vectibix administered in combination with FOLFOX as a first-line treatment of patients with KRAS wild-type mCRC failed to meet a secondary endpoint of overall survival. Originally designed to compare the treatment effect in the overall patient population, the study was amended to analyze outcomes with respect to the presence or absence of activating mutations in KRAS in the tumor itself. Tumor KRAS status was ascertained in more than 90% of the 1,183 patients enrolled in the trial.
Hunting for More Biomarkers
Finding reliable biomarkers to predict drug responses requires a tremendous commitment of resources on the part of drug developers. “Head and neck cancer is an area where there is a major need for biomarkers,” David Reese, M.D., executive medical director in medical sciences and therapeutic head for oncology early development at Amgen, told GEN.
“Amgen has a strong commitment to biomarker development. We incorporate it very early in the drug development process in oncology and other areas. The biomarker team is formed well before the drug is administered to humans and remains in existence throughout the development life cycle of the drug.
“Amgen’s biggest biomarker program success has been in colorectal cancer with some of the work we did with Vectibix. We had performed the first analysis of a Phase III trial, the 408 study, the original monotherapy registration trial for Vectibix. We found only patients with KRAS wild-type tumors appeared to respond to the drug. That led to conditional approval of the drug in Europe and to updates to labels in the United States.”
In Amgen’s most recent work tumor samples from 288 patients were analyzed for mutations in nine genes: KRAS (exon 3), NRAS, BRAF, PIK3CA, PTEN, AKT1, EGFR, beta-catenin (CINN1B), and TP53. All nine genes were either direct or indirect components of the EGFR-signaling pathway.
“Using next-gen sequencing technology, we could simultaneously analyze multiple genes, about a dozen, in the tumor samples from the 408 registration study. The use of the new technology was, to our knowledge, the first time it has been employed in a set of tumor samples from a Phase III trial and allowed a rapid analysis of multiple biomarkers.
“We found another RAS gene, NRAS, which when mutated also seemed to predict resistance to Vectibix,” Dr. Reese told GEN. A higher than expected rate of simultaneous mutation of KRAS and either BRAF or NRAS was observed.
“None of the genes that we analyzed were clear predictors of response or resistance. We viewed this work as a proof of principle that you could employ this technology in this setting and are doing additional work to confirm the associations we found in the study. We believe these technologies will become incorporated into clinical medicine rapidly.” But, Dr. Reese cautioned, “The identification of appropriate biomarkers remains one of the core challenges in the field.”
Patricia F. Dimond, Ph.D. ([email protected]), is a principal at BioInsight Consulting.
