Originally published on July 10.
NEW YORK (GenomeWeb) – Participation in Novartis' biomarker based signal-finding trials, known as the Signature Program, is progressing at an encouraging pace, according to researchers involved in the effort.
Novartis last year launched the Signature Program, a drug trial series in which patients are enrolled into mutation-specific study protocols that are tissue agnostic. Currently, cancer drugs are studied and approved according to a specific tumor site or histology (ie. breast cancer, lung cancer, etc). However, with advances in genomic knowledge, life sciences experts have often discussed future scenarios where therapies will be studied in and approved for the specific molecular pathways they interrogate. Novartis is taking the first step toward exploring such a vision through the Signature Program.
Oregon Health & Science University is among the institutions that have molecularly characterized tumors of patients who have been enrolled in Novartis' program. The drugmaker doesn't pay for the initial genetic screening of patients to see if they are right for one of the Signature protocols. Patients are pre-screened according to their doctors' recommendations before coming on the trial. But this strategy of "flipping the traditional clinical trial paradigm on its head," is actually driving molecular testing in the community healthcare setting, said Christopher Corless, chief medical officer at OHSU's Knight Diagnostics Laboratories.
Two years ago OHSU implemented next-generation sequencing panels to personalize care for its cancer patients. The university then began branching out and offering testing services to community practices interested in genetic testing for their cancer patients. OHSU has worked particularly closely with a group of practices called Cancer Clinics of Excellence (CCE). "What has really driven their interest [in molecular testing] is the Novartis Signature Program," Corless told PGx Reporter.
"If we didn't have that trial, then I would say that the uptake of testing would have been much, much lower, both in Oregon and elsewhere in the country," Corless said. "If you don't have access to these drugs in a trial then your option is to do compassionate use or off-label use. And the opportunities to do that are becoming more and more limited by the day."
Over the past year, since the Signature Program has been enrolling patients, Novartis has been successful at getting participants into study protocols, according to Corless. The CCE practices that OHSU genetically tests patients for have been among the top organizations identifying and enrolling patients to the project. "We have a happy collision of a laboratory that's doing state of the art testing, a pharmaceutical company making available their latest compounds in a very novel trial, and a group of practices that are keen to do clinical trials," Corless said.
He explained that some cancer patients molecularly tested through OHSU's lab get on trials other than the Signature Program, but still, Novartis' project has provided encouragement to large cancer centers and community practices alike to incorporate genetic information in the care of their patients.
At the Personalized Medicine World Conference in January, Steven Stein, Novartis' head of US clinical development and medical affairs, estimated that there were between 30 and 50 patients enrolled across five mutation-specific drug trial protocols within Signature.
At the American Society of Clinical Oncology's annual meeting in June, researchers from Novartis, University of Texas MD Anderson Cancer Center, OHSU's Knight Diagnostic Laboratories and elsewhere, reported that 73 patients had been consented into one of the Signature protocols at 46 network, academic, and community sites. According to Novartis, as of June 25, 83 unique sites have started one or more Signature protocols and 153 patients are enrolled and dosed.
"Based on our experience to date, the Signature clinical trial model is working well," Stein told PGx Reporter earlier this month. "The Signature protocols are a rapid and cost-effective method to evaluate the safety and potential efficacy of the drugs being studied."
The five protocols currently open for patient enrollment include studies for buparlisib for PI3K-activated tumors; dovitinib for tumors with mutations or translocations of FGFR, PDGFR, VEGF, cKIT, FLT3, CSFR1, Trk and RET; binimetinib for RAS/RAF/MEK-activated tumors; encorafenib for patients with BRAF V600 mutated tumors; and sonidegib in those with hedgehog pathway mutated solid tumors or hematologic malignancies. Novartis is planning to open additional protocols later this year for other investigational inhibitors, as well as protocols for combination agents.
This week, Stein told PGx Reporter that based on feedback Novartis has received, academic and community practices enrolling patients in the Signature protocols seem to particularly appreciate the ability to participate in early drug development research, that their patients are able to stay with their treating physicians while taking part in the trial, the rapid startup process for getting a site activated, and that the protocols enable access to investigational agents for patients who wouldn't otherwise get on drug studies.
Though one of the aims of the Signature Program is to increase patients' access to clinical trials, it also may enable Novartis to learn which pharmacogenomic indications to further investigate for a particular therapeutic agent.
Exploring early signals
Novartis is hoping that the Signature Program will enable it to quickly glean whether its investigational drugs have an efficacy signal in small molecularly defined patient subsets, while learning more about the various cancers that the patients have and exploring the therapeutic potential of a particular compound. The primary endpoint in the Signature studies is to gauge the clinical benefit of each therapeutic agent.
Using a Bayesian approach, patients will be clustered into similar indications. If a drug appears to have efficacy in a particular type of cancer in a specific molecularly defined group, then Novartis will conduct a larger, formal investigation to explore this signal further. If a drug in a particular indication and patient subset doesn't appear to have a signal, then Novartis will drop that effort.
Speaking at the conference in January about the Signature Program, Novartis' Stein noted that this program represents work the drugmaker is doing outside of the traditional clinical development space. "You have to suspend belief a little bit about the way you think of clinical trials," he said. "It's really a way for a company to get rapid proof-of-concept in rare tumors."
Although patients receive better care when they are enrolled in clinical trials than when they are seen in regular practices, Stein noted that only between 2 percent and 4 percent of US patients are enrolled in studies. Meanwhile, as advances in genomics reveal rarer and rarer pharmacogenomic signals – for example, ROS1 rearrangements characterizing 1 percent of non-small cell lung cancers – drugmakers are finding it challenging to enroll sufficient numbers of study participants in traditional clinical trials and garner enough confidence that their therapeutics interrogate these markers.
Beyond designing mutation-specific drug trial protocols, Novartis introduced a number of other requirements for joining the program, hoping to make its signal-finding effort more efficient. For example, when a physician identifies a patient that they want to enroll in one of the Signature protocols based on the patient's tumor profile, they must accept a non-negotiable standard contract, budget, and informed consent, as well as use central IRB.
"The thing that slows us down is contract negotiations," Stein said back in January, adding that sites that have a good clinical practice record with Novartis or with the pharma industry in general can join the Signature program. Although the "take-it-or-leave-it" contract for Signature is a "controversial" aspect of the program, Stein said he was surprised so many practices and institutions have been willing to accept the terms of the project.
In launching Signature, Novartis was initially focusing on getting community practices to enroll patients. Compared to large cancer centers, getting community practices on board with a clinical trial protocol is cheaper and requires less paperwork, Corless acknowledged. As word spread about the Signature Program, however, Novartis "started to get so much interest from universities that they allowed universities to join," he said. "So, in a sense, now there is competition from both the community and academic practices to try to find slots on the trial."
So far, participation in the Signature Program has been strong and preliminary data suggests that there are benefits to its approach. The average start-up time for a new site to join the program – from the time a new site with an eligible patient contacts Novartis to getting the site up and running with patient consent and screening – was four weeks, according to data reported at ASCO.
"We look in real time all the time," Stein said speaking at the PMWC, adding that a site is brought on to the program only when there is a patient ready to be treated. "We never have a non-enrolling site."
OHSU performs a custom Ion Torrent lung cancer panel of 23 genes and a solid tumor panel of 37 genes, on which most patients get tested. Since Novartis is not paying for upfront molecular testing of Signature study participants, that cost is shifted to the labs participating in the research.
OHSU has had fairly good reimbursement from private payors for the NGS testing they've done on cancer patients who are being considered for enrollment in Signature or other trials. However, since Medicare only pays for specific markers gauged by NGS panels, the reimbursement is often not enough to cover lab costs, Corless said.
Driving adoption of molecular testing
As Novartis' Signature Program is ongoing, the National Cancer Institute is also exploring novel clinical trial designs that aim to capture early drug signals by studying how patients with certain molecular cancer markers respond to different therapeutic agents. For example, in June, the NCI announced the launch of the Lung-MAP trial, a collaboration between SWOG Cancer Research, Friends of Cancer Research, the Foundation for the NIH, and five pharmaceutical companies. In the trial, researchers plan to genomically profile squamous cell carcinoma patients and stratify them to different investigational treatments that are designed to target those characteristics.
"This innovative approach to clinical testing should both improve access to promising drugs for patients and ease the significant recruitment and infrastructure burdens on researchers involved in traditional clinical trials," the NCI said in a statement announcing Lung-MAP's launch.
Ultimately, trials like Signature and LungMap help encourage doctors to molecularly characterize their cancer patients early in their disease paradigm and record the markers as part of their medical record. By the time a patient has failed the third line of therapy, it's often too late to explore precision treatment options. "We're seeing how oncologists are starting to do this in a prospective fashion and not waiting till the very end" to do molecular testing, Corless reflected.
Furthermore, discussions between pharma and the US Food and Drug Administration have taken place about how data from tissue-agnostic, adaptive trials, such as Signature, might be used to speed up approval and access to drugs, he noted. However, it might be some time before any type of concrete framework emerges.
For the time being, Novartis' Stein noted that the Signature trials are proof-of-concept studies. "It's important to note that these are exploratory trials," he told PGx Reporter, emphasizing that while data from the trial protocols can supplement a regulatory filing, they cannot be used independently for any filing at this point.
"We have to be careful. The concept that you only need to target a pathway and not pay attention to where the tumor came from is not entirely safe," Corless said. "For example, we know that BRAF is a great target in melanoma. It's probably going to be a pretty good target in lung cancer for the few that have [mutations]. But when you look at the 6 percent of colon cancers that have BRAF mutations, they do not respond to BRAF inhibitors."
As such, the life sciences community "cannot entirely divorce the type of tumor from the mutation" in the context of drug studies, he continued. "We have to pay attention to both."