This article has been updated to correct the name of the Phase I extension trial to Study 1001. Originally published June 3.
CHICAGO – A small subset of non-small cell lung cancer patients with tumors that have rearrangements in the ROS1 gene have responded so well to Pfizer's Xalkori (crizotinib) in clinical trials that some academic medical oncologists are already incorporating screening for this molecular marker into clinical practice.
In an extension of the Phase I Study 1001, which last year led to Xalkori's approval as a treatment for NSCLC patients with ALK rearrangements, researchers from Pfizer and various academic institutions tracked response to the drug in 15 NSCLC patients with ROS1 rearrangements. As of April 19, there were 14 evaluable study participants – one patient experienced a complete response, seven experienced a partial response, four had stable disease, and two experienced disease progression.
This study was presented by lead author Alice Shaw of the Massachusetts General Hospital Cancer Center at the American Society of Clinical Oncology annual meeting here this week. Following her presentation, Memorial Sloan-Kettering Cancer Center medical oncologist Gregory Riely reviewed the findings and said that based on the current data, if he had NSCLC patients with ROS1 rearrangements, he would treat them with Xalkori. During separate presentations on NSCLC pharmacogenomics at the conference, researchers from several academic cancer centers, such as Vanderbilt University, noted that they are already screening for ROS1 rearrangements in adenocarcenoma patients.
Noting the preliminary nature of the findings, Riely acknowledged that "clearly, [Pfizer] is going to have to test this [finding] in more patients and there will need to be all sorts of [regulatory] steps" ahead of expanding the drug's indication to ROS1-positive NSCLC patients.
Nevertheless, "I don't think anybody can argue with this data," he asserted, characterizing the requirements that Pfizer needs to fulfill in order to formally market the drug in this subpopulation of patients and gain reimbursement from insurers as "just window dressing." Riely disclosed before his talk that he had received research funding from Pfizer and had consulted for a number of other drug developers.
Pfizer told PGx Reporter that it is currently working to better understand the mechanism of response in NSCLC patients with ROS1 rearrangements. The company did not indicate whether, based on these early findings, it would try to expand Xalkori's label to include this patient population.
"The ROS1 study reflected the activity of crizotinib in a small group of patients, and further study is needed to determine the appropriate path forward. Pfizer will continue to enroll patients and evaluate clinical research of crizotinib in patients with ROS1 rearrangements to better understand the compound’s activity in this population," a company spokesperson said.
Although it is currently unclear what Pfizer's regulatory strategy is for Xalkori regarding the ROS1-positive NSCLC patient subset, these early findings appear to be enough for some medical oncologists to move ahead with incorporating analysis of ROS1 fusion markers in the treatment of lung cancer patients. Some presenters at ASCO, who are already convinced of the strength of the Xalkori response in this patient population, highlighted the need for new regulatory models that would enable drug companies to introduce new PGx indications in step with rapidly advancing scientific knowledge for marketed drugs.
"Do we need separate clinical trials for each of the genotypes? Do we always need to compare the efficacy of a targeted therapy in a molecularly defined population to systemic therapy?" Pasi Jänne, a researcher and medical oncologist at Dana Farber Cancer Institute, asked during a separate presentation at the meeting on personalized lung cancer treatments.
Referring to Xalkori's "impressive activity" in NSCLC patients with ROS1 rearrangements, he wondered whether the drug sponsor would need to do a randomized Phase III clinical trial comparing Xalkori to chemotherapy to gain marketing approval in the population. "Or do we accept that [this] level of activity is sufficient to forgo" that kind of evaluation, he posited.
There was similar excitement over how well ALK-positive NSCLC patients responded to Xalkori in preliminary studies. The drug was approved by the FDA last summer on an accelerated time frame -- after a review time of just over three months -- based on data from two single-arm trials involving 255 patients.
FDA officials have previously outlined how flexible the agency's drug and diagnostics divisions were with Xalkori's application in an effort to speed the drug to market (PGx Reporter 2/1/2012). Pfizer has said it is conducting larger studies to confirm the findings from these earlier investigations in ALK-positive patients, and is looking into other response markers and cancer resistance pathways for the drug (PGx Reporter 2/29/2012).
Study Details
In the ROS1 extension cohort, the objective response rate was 57 percent in the eight patients with a complete or partial response. The median duration of treatment was 25.7 weeks. Since "many of the patients in the study are still ongoing [with treatment], this is most certainly an underestimate," said Shaw, who disclosed being a consultant for Pfizer. At eight weeks, 79 percent of patients had disease control with Xalkori treatment, meaning their disease wasn't progressing.
Meanwhile, three patients discontinued participation in the trial because their cancer progressed. Among responsive patients, only one had stopped seeing a benefit from the drug as of the cut-off date for analysis.
Regarding the two patients still in the trial who experienced progressive disease on Xalkori, one patient had an abnormal FISH test result and the researchers believe this patient may not have a ROS1 fusion gene. The second patient had a bowel perforation not related to the study and the researchers stopped Xalkori treatment. However, upon resuming treatment, this patient achieved a 60 percent reduction in tumor burden, Shaw said.
Xalkori's efficacy in ROS1-positive patients is "remarkably similar to those that were received with the first cohort of ALK-positive patients treated with crizotinib in a Phase I study and reported at ASCO three years ago," Shaw noted. The researchers also saw a safety profile for Xalkori in ROS1-positive patients similar to that of ALK-positive patients.
Based on these findings, Shaw and her colleagues concluded that patients with ROS fusion markers comprise a "distinct" NSCLC patient subpopulation. "This study represents the first clinical validation of ROS as a therapeutic target in cancer," the study authors concluded.
ROS1 rearrangements can occur when the gene partners with a number of other genes. However, these fusions, which occur in approximately 1 percent of NSCLC patients, do not overlap with ALK rearrangements or KRAS or EGFR mutations. This and other studies suggest that ROS1 fusions appear to occur in younger people who are never smokers with adenocarcinoma tumor histology.
During her presentation at the meeting, Shaw highlighted the experience of two patients who represented the "typical response" seen in the extension cohort.
The first patient, a 40-year-old female never smoker, had NSCLC that was advancing despite treatment with an Avastin and chemotherapy regimen. She was treated with Xalkori after genetic testing revealed that she had ROS1-positive tumors. "Soon after [being on the drug] she noted an improvement in her symptoms," Shaw said, noting that this patient experienced a partial response for almost one year.
The second patient, treated at Massachusetts General Hospital, was a never smoker who was diagnosed with advanced NSCLC at age 31. He was very sick by the time Shaw discovered that he had ROS1-positive tumors. Within a few weeks of starting treatment with Xalkori, however, the patient reported improved breathing.
"By two weeks almost all of his disease-related symptoms had resolved," Shaw said. This patient experienced a complete response with Xalkori treatment for over 13 months.
Complex Testing Scenario
In lung cancer, oncologists are rapidly moving away from differentiating patients simply based on tumor histology – as adenocarcinmas, squamous cell carcinomas, and large cell carcinomas – and are increasingly characterizing tumors by their molecular features and using this information to guide treatment decisions. Although around 35 percent of molecular markers associated with NSCLC are unknown, EGFR and KRAS mutations comprise a large portion of the markers implicated in the disease. Additionally, clinicians are also beginning to test for a growing list of point mutations and gene arrangements, such as ROS1, ALK, and RET, in order to personalize treatments for increasingly small NSCLC subpopulations.
The rapid pace of discovery linking molecular markers to lung cancer has made it challenging for practicing oncologists to establish up-to-date testing strategies for lung cancer patients. Riely said that at Memorial Sloan-Kettering, all patients with lung adenocarcinomas get tested for EGFR and HER2 insertions and deletions and are evaluated for a number of point mutations via mass spectrometry-based genotyping. If those tests come up normal, then they receive FISH-based testing for ALK, ROS1, and RET rearrangements.
Although he emphasized the need for routine molecular testing of all NSCLC patients, Riely noted that it is "quickly becoming untenable to do [such assessments] in a piecemeal fashion." For example, Jänne pointed out during his presentation, if a separate FISH test has to be performed for every genomic alteration associated with NSCLC treatment response, oncologists would not have enough tissue samples to analyze and the cost of running multiple tests would become prohibitive. Riely, Jänne, and other presenters at ASCO advocated the use of advanced technology platforms that can simultaneously characterize a variety of molecular changes in patients' tumors.
Memorial Sloan-Kettering, for example, has launched the Integrated Mutational Profiling of Actionable Cancer Targets, or IMPACT, program, in which oncologists are using targeted next-generation sequencing to identify molecular abnormalities -- including insertions, deletions, fusion genes, amplifications, and deletions -- in 230 cancer-related genes. This approach "really gets all the things we think matter with lung cancer patients," Riely said, adding that in this program researchers aren't yet analyzing ROS1 rearrangements.
This highlights the importance of "how we have to build assays that can incorporate new findings in a relatively rapid way," he reflected.
Although there are advantages to using NGS platforms in characterizing multiple genomic alterations from a single patient sample, there are disadvantages, too. The lack of standardization of analytical methods, high costs, and uncertainty around how to garner regulatory approval are all drawbacks of using an NGS platform to guide treatment strategies, Jänne pointed out.
For this reason, most drug developers are largely using whole-genome sequencing and NGS techniques in the research setting, but have yet to advance commercial predictive or prognostic tests based on these technologies.
In the expansion cohort, Shaw and colleagues gauged ROS1 rearrangements with a laboratory-developed break-apart FISH probe test similar to Abbott's ALK companion test approved for Xalkori by the FDA.
Pfizer didn't comment on whether it would stick with Abbott's FISH platform if the company decides to expand Xalkori's label to the ROS1-positive subpopulation. "We are in the exploratory stages with a ROS1 test, but it's too early to disclose commercial plans," an Abbott spokesperson said.