The epidermal growth factor receptor (EGFR), one of the many receptor tyrosine kinases that are targeted in cancer treatment, is frequently overexpressed or mutated in cancer, resulting in cell proliferation, survival, invasion, metastasis, and tumor-induced neoangiogenesis.1 The approach to targeting EGFR in cancer has included the use of small molecule tyrosine kinase inhibitors (TKIs), such as gefitinib, erlotinib, lapatinib, and afatinib, and monocloncal antibodies such as cetuximab, panitumumab, and matuzumab (Figure).2
A session titled “Targeting EGFR: The Next 10 Years,” held May 31, 2014, on the second day of the 50th Annual Meeting of the American Society of Clinical Oncology (ASCO), provided a progress report on the successful targeting of the protein in non-small cell lung cancer (NSCLC).
The opening talk by Pasi A. Janne, MD, PhD, director of the Lowe Center for Thoracic Oncology at the Dana-Farber Cancer Institute and professor of medicine, Harvard Medical School, was titled “Clinical Activity of the Mutant-Selective EGFR Inhibitor AZD9291 in Patients With EGFR Inhibitor-Resistant NSCLC.” Janne presented results from a phase 1, open-label, multicenter clinical trial with a selective, third-generation EGFR-TKI by AstraZeneca, AZD9291, that has been found to be effective against both EGFR-TKI—sensitizing and resistance (T790M) mutations in preclinical models.
The trial, conducted in 199 EGFR mutant (T790M+) NSCLC patients with acquired resistance to EGFR-TKIs, conducted a dose escalation and expansion study of oral AZD9291. Among all the evaluable patients, the confirmed+unconfirmed overall response rate (c+uORR) was 51% (91/177). Response Evaluation Criteria in Solid Tumors (RECIST) responses were observed at all dose levels and in brain metastases. Of 132 patients with centrally confirmed T790M, the c+uORR in 89 EGFR T790M+ patients was 64% (95% CI: 53%-74%), and in 43 EGFR T790M patients was 23% (95% CI: 12%-39%). The overall disease control rate in T790M+ patients was 96% (85 of 89). Among the 60 patients with a confirmed response, 97% (58 of 60) were ongoing at data cutoff and the longest duration of response was greater than 8 months. AZD9291 did not present any dose-limiting toxicities, maximum tolerated dose (MTD) was not defined, and the most common adverse events (AEs) were diarrhea, rash, and nausea. Very few grade 3 or higher AEs were observed in any of the cohorts. The overall response rate (ORR) was 53% and the overall disease control rate was 83%. A progression-free survival (PFS) benefit was observed in T790M+ patients. The results of this study have helped establish the dose for the phase 2 trials.
The phase 1 study demonstrated that T790M+ NSCLC patients with acquired resistance to EGFR-TKIs respond very well to AZD9291 and have a higher ORR compared with the EGFR T790M patients. Further clinical developments directed at the T790M+ patient population are ongoing.
The subsequent presentation was a dose-finding study for Clovis Oncology’s CO-1686, a molecule that targets the EGFR T790M mutation in NSCLC patients, which recently received the FDA’s “breakthrough therapy” designation.3 Designed to target mutant EGFR, CO-1686 does not bind wild-type EGFR, thus increasing specificity and reducing adverse effects, which makes it eligible as first-line therapy in patients with activating EGFR mutations.4 Clovis wants to develop CO-1686 as an oral treatment for NSCLC patients who have failed EGFR-directed therapy, such as Tarceva (Genentech) or Iressa (AstraZeneca), and have also developed the T790M mutation, which is the dominant resistance mechanism to Tarceva/Iressa.4
The results were presented by Lecia V. Sequist, MD, MPH, Massachusetts General Hospital. The first-in-human dose-finding study was conducted in 88 patients with EGFR-mutated advanced NSCLC. The formulation, a hydrogen bromide salt form of CO-1686, was administered to patients who were previously treated with EGFR TKI and had a tumor biopsy in screening for central EGFR genotyping. The patients had previously received a median of 3 therapies, with 40% having had more than 1 prior line of EGFR TKI. Related AEs, which were mild and observed in about 10% of patients, included nausea, fatigue, and hyperglycemia (managed with oral hypoglycemic agents such as metformin, with or without dose reduction). A recommended phase 2 dose of 750 mg twice a day was selected.
ORR was 58% in biopsy-proven, heavily pretreated, centrally confirmed T790M+ patients, and the responses deepened over time with longer follow-up. Median PFS was not reached at the time the data was presented, but was projected at more than 12 months. The trial also observed central nervous system responses with the drug.
CO-1686 proves promising in T790M+ EGFR mutant NSCLC patients, and the phase 1 study provided proof of concept that targeting the mutant form of EGFR can reduce AEs like diarrhea and rash. The phase 2/3 program initiated this year called TIGER (third generation inhibitor of mutant EGFR in lung cancer) is actively recruiting patients for 3 registration studies.
The subsequent presentation by Gideon Michael Blumenthal, MD, Holy Cross Hospital, who is also a clinical team lead with the FDA, was entitled “Overall Response Rate (ORR) as a Potential Surrogate for Progression-Free Survival (PFS): A Meta-Analysis of Metastatic Non-Small Cell Lung Cancer (mNSCLC) Trials Submitted to the FDA.” There is an ongoing debate about the appropriate end points in clinical trials—they hold immense importance in oncology trials—and the FDA has guidelines for same.5 The use of surrogate markers has stemmed from the accelerated drug approval process, whereby drug manufacturers often use surrogate end points to speed up drug launch, to save or extend lives. These surrogate end points include ORR, PFS, disease-free survival (DFS), time to progression (TTP), and patient-reported outcomes.6
Targeted therapies (TT) administered as single agents in molecularly defined mNSCLC subsets are yielding high ORRs. Additionally, large improvements in PFS with favorable benefit-to-risk ratio have served as the basis of drug approval in mNSCLC. However, the relationship of ORR with PFS or OS in mNSCLC is not established. The FDA conducted a meta-analysis using data from 15 mNSCLC trials that were submitted to the agency, which enrolled 12,534 patients, including 3 trials of TT in molecularly enriched populations with high (HR) and OS HR versus the estimated odds ratio (OR) of ORR on the log scale was calculated.
On a trial level, the meta-analysis of randomized, active-controlled trials indicated a strong correlation between ORR and PFS (R2 = .89), especially for head-to-head trials, which means ORR could be an acceptable surrogate for PFS. A correlation between ORR or PFS and OS is not established and may be confounded by crossover in the TT trials. This might either be the result of a weak relation or no relation or a high crossover, under-power, or long post program survival, in the 3 small targeted therapy trials in the enriched population, which may confound the analysis. At the trial level, a TT in a molecularly defined subset of mNSCLC with a large magnitude of effect on ORR will likely have a large effect on PFS.
Lawrence H. Schwartz, MD, a diagnostic radiologist at Columbia University Medical Center and at NewYork-Presbyterian Hospital, concluded the session with his talk titled “Getting the Right Drug to the Right Patient Faster,” wherein he provided his critique on the current treatment protocols that analyze response versus progression of disease.
“It’s critical to distinguish response and progression as 2 distinct events,” said Schwartz. Response rate is objective, measurable, and reproducible. Additionally, there’s a long history—more than 30 years—of standardization and measurement of response. But the definition of response needs to be challenged and modified, he said. RR can be strongly correlated with median survival time, but the question remains—– what is “optimal” RR?
Moving on to analyzing progression, Schwartz asked, “How accurate a gold standard is progression?” Alternate methods to assess response and progression to find effective therapies earlier are needed. To this end, the FNIH-sponsored VOL-PACT trial is under way, which considers an entire spectrum of metrics, not just RR. “The value proposition of these simulations is that combining metrics with imaging would provide much improved results,” said Schwartz.
He concluded the session by saying that there is a need to combine smarter trials and smarter trial end points, including imaging, to derive optimal results and responses.
1. Ciardiello 1, Tortora G. EGFR antagonists in cancer treatment. N Engl J Med. 2008;358(11):1160-1174.
2. Chanprapaph K, Vachiramon V, Rattanakaemakorn P. Epidermal growth factor receptor inhibitors: a review of cutaneous adverse events and management [published online March 2, 2014]. Dermatol Res Pract. doi:10.1155/2014/734249.
3. Clovis Oncology receives breakthrough therapy designation for CO-1686 for the treatment of second-line EGFR mutant non-small cell lung cancer (NSCLC) in patients with the T790M mutation. Business Wire website. http://www.businesswire.com/news/home/20140519006502/en/Clovis-Oncology-Receives-Breakthrough-Therapy-Designation-CO-1686#.U76NarH93Z0. Accessed July 10, 2014.
4. CO-1686. Clovis Oncology website. http://www.clovisoncology.com/products-companion-diagnostics/co-1686/. Accessed May 31, 2014.
5. Guidance for industry clinical trial endpoints for the approval of cancer drugs and biologics. FDA website. http://www.fda.gov/download/drugsGuidanceComplianceRegulatoyInformation/Guidance/UCM071590.pdf. Published May 2007. Accessed May 31, 2014.
6. McCain JA. The ongoing evolution of endpoints in oncology. Manag Care. 2010;19(5 suppl 1). http://www.ptcommunity.com/supplements/1005_CB_endpoint/CB_endpoints.pdf. Accessed May 31, 2014.