Untargeting Xalkori? Hints Of Efficacy In ALK-Negative Patients Must Be Addressed, FDA Says

FDA approved Pfizer’s lung cancer drug for patients positive for the gene translocation – but review documents show that agency reviewers focused on the efficacy seen in the few ALK-negative patients in the pivotal trial, and want Pfizer to follow up on that signal post-marketing. Regulatory expectations for disproving effect in the non-selected population will be an important issue as more personalized medicines reach the agency.

Pfizer Inc.’s ALK-positive non-small cell lung cancer therapy Xalkori (crizotinib) and its companion diagnostic may be standard bearers for the new era of personalized medicine, but FDA is hedging its bets with a post-marketing study that will evaluate signals of efficacy in the broader NSCLC population that is negative for the anaplastic lymphoma kinase gene translocation biomarker.

Only about 4% of NSCLC patients are positive for the ALK translocation, a biomarker only identified in 2007. Pfizer built on strong indications of efficacy in the ALK-positive subgroup to propel the drug, and its diagnostic, rapidly through development and review. The drug and diagnostic were approved Aug. 26, 2011, ahead of the six-month priority review timeframe (Also see "Synchronized Regulation: Tandem Approvals Of Xalkori, Diagnostic Required Considerable Coordination" - Pink Sheet, 1 Mar, 2012.).

A small cohort of ALK-negative NSCLC patients were enrolled in the Xalkori pivotal trials, but they had comparable response to the ALK-positive patients.

Personalized medicine and drug/diagnostic co-development are certainly dominant trends in R&D – especially in cancer. “With the discovery of molecular targets and targeted therapies, new treatment options for these patients are evolving,” clinical reviewer Shakun Malik stated in her Aug. 12, 2011 review. Development strategies and regulatory expectations also are adapting; FDA has issued policy on drug/diagnostic co-development, and possible modifications to regulatory standards for targeted therapies have been part of discussions of “adaptive licensing” approaches (Also see "Expedited Approval Pathway Concept Draws Strong Support But Different Proposals" - Pink Sheet, 1 Dec, 2011.).

Malik pointed to the potential for biomarkers to identify the most appropriate patients for older lung cancer agents like the anti-VEGF-A monoclonal antibody Avastin (bevacizumab) and epidermal growth factor receptor tyrosine kinase inhibitor Tarceva (erlotinib). “Although both these targeted agents approved for NSCLC do not require demonstration of specific molecular abnormalities in patient’s tumor tissue, there is an increasing awareness of the importance of identifying specific NSCLC molecular drivers to appropriately direct targeted agents to patient populations.”

The literature is tantalizing, but not yet conclusive, Malik acknowledged. “In clinical trials, when compared to chemotherapy, EGFR tyrosine kinase inhibitors are associated with a high response rate (70-80%) in NSCLC patients whose tumor harbors EGFR favorable mutations … associated with improved [progression-free survival],” she reported. “However, no overall survival advantage has been demonstrated so far.”

As the Xalkori post-marketing commitment suggests, some uncertainty remains about the practical applications and utility of biomarkers.

Amid the enthusiasm for personalized medicine, a few skeptical voices have been raised. A panel report published in Lancet Oncology Sept. 26, 2011, for example, questioned the evidence of association between biomarkers and disease risk. “Despite its promise over the past decade, the science has not yet lived up to expectations,” the report states (Also see "Quality Of Biomarkers Questioned Again; Lancet Report Calls For Better Evidence" - Pink Sheet, 1 Oct, 2011.). A Tufts Center for the Study of Drug Development survey earlier in the year found skepticism among payers over the clinical usefulness and lack of hard evidence of health outcomes for some pharmacogenetic tests and companion diagnostics (Also see "Payers Seek Missing Link Between Diagnostic Tests And Clinical Outcomes" - Pink Sheet, 25 Jul, 2011.).

Similar Response Rate

The overall response rate in the targeted ALK-positive patient population was 50% in Study A and 61% in Study B, higher than the rates seen with approved chemotherapy for NSCLC. But FDA reviewers also honed in on the indication that the non-selected group may have responded as well.

Twenty-three patients with ALK-negative NSCLC, as determined by the Vysis diagnostic, were treated with crizotinib in the registration trials. Eight of the ALK-negative patients had not received prior chemotherapy for metastatic disease, Ellen Maher, the cross-disciplinary team leader, observed in a July 12, 2011 review. Five of 19 ALK-negative patients met the investigator response criteria and another two had a single assessment of response, meaning that the ALK-negative group could have a 35% response rate.

The response rate in ALK-negative patients, if confirmed, would be “similar to the response rate in patients with ALK-positive NSCLC in Study A,” Maher noted.

“It is necessary to further explore the response in patients determined to be ALK negative by the Vysis ALK Break Apart FISH probe assay to determine if crizotinib should be confined to the ALK positive population,” according to FDA’s development template for post-marketing requirements and commitments. The development templates, which FDA must complete for each post-approval investigation, provide a closer look at the agency’s expectations for the research than the brief description in the approval letter.

FDA Envisions A “Phase II-Type” Of Trial In ALK-Negative NSCLC

Maher’s July review makes clear that a study of additional ALK-negative patients in NSCLC would be a post-marketing request. And while most reviewer discussion of the issue also describes the post-approval study as a PMR, it was cleared as a post-marketing commitment. Division Director Robert Justice explained that “since the activity of crizotinib in patients with ALK-negative NSCLC is not a safety issue, it will be a post-marketing commitment.” (The authority to enforce mandatory post-market studies granted under the FDA Amendments Act is limited to safety issues.)

The PMR/PMC development template for the ALK-negative post-marketing commitment explains that the purpose of the trial is to “provide additional information on the safety and efficacy of crizotinib in a comparative setting of ALK-negatives using the drug’s companion diagnostic test.”

The planning document indicates that FDA is expecting an “exploratory/Phase II type” of clinical trial in which the Vysis assay is used to enroll ALK-negative patients, and the patient population should be similar to the registrational trial.

The post-marketing trial should address four questions, according to the PMC planning document:

• “Does the companion test divide the population into groups of patients who will respond better or worse (or not as well) when treated with crizotinib?”
• “Is crizotinib active in patients identified as ALK negative, based on the current established cut-off (<15%) using the Vysis ALK Break Apart FISH probe assay?”
• “In ALK-negative patients, is the activity of crizotinib greater than the activity of standard lines of therapy for this patient population (compared to historical/currently known response rates)?”
• “Should other biomarkers, either in addition to or in combination with ALK, be taken into account when determining if treatment with crizotinib should be considered?”

Crizotinib’s Multilayered Activity

One of the difficult aspects of interpreting the molecular characteristics of cancer is the interconnected complexity of various molecular pathways and characteristics. “Targeting a single molecular defect is not enough,” a 2011 report from the American Society of Clinical Oncology, “Accelerating Progress Against Cancer – ASCO’s Blueprint For Transforming Clinical and Translational Cancer Research,” states. “Most cancers are driven by multiple mutations” (Also see "Smarter, Better, Faster, Smaller: ASCO Pushes For Cancer Trial Improvements" - Pink Sheet, 7 Nov, 2011.).

“Like other approved kinase inhibitors, crizotinib targets several proteins at clinically relevant concentrations, including c-Met/hepatocyte growth factor receptor (HGFR), anaplastic lymphoma kinase (ALK), and Recepteur d’Origine Nantais (RON),” Acting Pharmacology Team Leader Whitney Helms noted in an Aug. 10, 2011 memo.

“Though the drug has its strongest potency against c-Met, the clinical development of crizotinib has focused on patients with tumors expressing ALK gene translocation products,” Helms said. The pharm/tox team constructed labeling language to point out that crizotinib demonstrated “concentration-dependent inhibition of ALK and c-Met phosphorylation in cell-based assays using tumor cell lines and demonstrated antitumor activity in mice.” Pharmacology reviewer Brenda Gehrke recommended adding mention of c-Met to the language on activity against tumor xenografts in mice “to indicate that the drug was studied and had activity in c-Met xenograft models as well.”

Indeed, crizotinib was originally positioned for NSCLC patients expressing c-Met – the Phase I A8081001 study started in 2006, and the ALK fusion biomarker was not identified in 2007, when a Nature article described “an inversion within chromosome 2p resulting in the formation of a fusion gene product associated protein-like 4 (EML4) gene and the ALK gene in NSCLC cell lines and archived clinical specimens,” Malik noted. Pfizer quickly shifted focus to ALK, and approached FDA in 2009 to discuss a registration strategy specifically for ALK-positive NSCLC after noting partial responses in seven of 14 ALK-positive trial patients (Also see "Crizotinib Holds Promise For Lung Cancer, And For Pfizer Oncology" - Pink Sheet, 14 Jun, 2010.).