Baltimore-based non-profit Center for Medical Technology Policy released guidelines last week outlining the types of evidence a molecular diagnostic developer should gather on a test to show that it is useful in patient care and to convince payors to reimburse for the diagnostic.
Industry players often cite the lack of consistent reimbursement standards for molecular diagnostics as a major barrier to the advancement of personalized medicine. CMTP released the latest guidelines in an effort to more clearly lay out the types of data payors are looking for in order to determine whether a molecular diagnostic is clinically valid (whether the test accurately gauges patients' clinical status) and is clinically useful (whether testing improves patient outcomes).
Over the years, various advisory and scientific groups have "called for stakeholder-informed, clearly defined evidentiary standards for clinical utility, and the CMTP guidance is the only set of such recommendations that have been developed to date," Pat Deverka, senior research director at CMTP and lead author of the guidance, told PGx Reporter. CMTP aims to inform the design and implementation of comparative effectiveness that patients, clinicians, and payors can use in treatment and policy decisions. The organization engages stakeholders to discuss research questions and design studies, and fosters partnerships to conduct the studies
For example, the CMTP recommends that test developers, when possible, should conduct randomized, controlled trials to evaluate the effectiveness of a test compared to controls for patients that harbor the biomarkers of interest and those who don't. However, CMTP also acknowledges in the guidelines instances when a randomized, controlled trial is not possible and discusses the alternative strategies sponsors can pursue.
"There is increased recognition by payors that study designs other than randomized, controlled trials will in some cases be adequate to demonstrate clinical utility, given the expense and time required to conduct such studies," said Deverka. "Stakeholders that participated in the process of developing [these guidelines], including clinicians, patient advocates, researchers and payors, recognized the need to include a range of potential study designs to demonstrate the clinical utility of MDx tests. Payors, in particular, expressed frustration with the quality of studies they were being asked to evaluate and indicated a willingness to develop a shared set of standards for evaluating clinical validity and clinical utility that could be used by test developers and researchers for new MDx tests."
CMTP developed these guidelines with input from the Centers for Medicare & Medicaid Services, private payors, diagnostics companies, and drug developers. As such, although the clinical utility guidelines don't come from a government agency per se, CMTP believes that its consensus-driven recommendations will be well-received and adhered to by payors and industry.
In order to define the evidence problems with clinical utility studies currently conducted by test developers and come up with recommendations for designing better trials, CMTP experts interviewed CMS officials, representatives from five major private payors, and executives from the drug and diagnostic industries. "Increasingly, we anticipate that payors will cite these [guidelines] as consistent with their evidence expectations, which will further incentivize [industry] considerations during product development," said Deverka. "Following these guidelines enables test developers and payors to have a common framework for evaluating evidence of clinical validity and clinical utility of MDx tests."
Molecular diagnostics firms seeking to launch new tests on the market have long complained about the lack of predictability in the types of evidence standards payors seek for a positive reimbursement decision. The problem is particularly acute in the personalized medicine field, where new molecular tests are coming to market at a rapid pace.
Different attempts at bringing clarity to the reimbursement process for molecular diagnostics have not proven successful. A few years ago, the CMS charged Medicare contractor Palmetto GBA with establishing a standardized process for collecting information on the analytical validity, clinical validity, and clinical utility of molecular tests through the so-called MolDx program, and then, eventually setting pricing based on the value a test provides to patient care. However, the test pricing released by Palmetto thus far has been criticized by industry as woefully low, often falling below the cost of performing the tests. Furthermore, industry players have described CMS's and contractors' process of coding and pricing tests as opaque (PGx Reporter 2/6/2013).
The CMTP recommendations are quite specific in terms of the types of studies payors want to see when establishing the clinical utility of molecular tests. The guidelines, for example, discuss when a prospective-retrospective study may be appropriate; the conditions under which payors will accept data from a single-armed trial; the use of registries to conduct prospective observational studies on tests; when to employ analytical models to gain information on the potential impact of a test on patient outcomes; as well as the types of analyses that won't be sufficient for coverage.
CMTP included process measures – studies showing that a physician changed a treatment strategy for a patient using the results of a test – as an example of the types of data that may not be sufficient to garner reimbursement for a diagnostic unless the study shows that the molecularly guided strategy improves patient outcomes. "Process measures, such as changes in physician behavior, are typically insufficient to qualify as study endpoints, unless there exists a separate, robust body of credible evidence (as determined by widely accepted evidence review standards) linking clinical management decisions with relevant health outcomes," the guidance states. "Studies designed to report intended care plans following an MDx test are insufficient for demonstrating clinical utility."
However, test developers often conduct studies tracking the treatment decisions of physicians in an attempt to show that the diagnostic has an impact on patient care. For example, after only one validation study was published on Genomic Health's Oncotype Dx test for gauging colon cancer recurrence, the company reported data from a study last year that around 30 percent of physicians changed their treatment strategy for colon cancer patients based on results from the test (PGx Reporter 1/25/2012).
"These new results support our efforts to gain broad reimbursement coverage and increase patient access, and highlight the independent value beyond conventional measures that the test provides to physicians, patients, and payors," Steve Shak, Genomic Health's chief medical officer, said in a statement when the data were released. At the time, Palmetto had already extended coverage for the colon cancer test, but Genomic Health was focused on growing reimbursement for the diagnostic among private payors.
Similarly, GenomeDx presented data recently showing that after assessing patients with its prostate cancer prognostic test Decipher, physicians changed their treatment recommendations 43 percent of the time and altered their recommendation from aggressive treatment to observation in 31 percent of the cases. GenomeDx did not assess long-term outcomes in that study, but plans to in future studies (PGx Reporter 2/20/2013 ).
These examples "reflect the current lack of consensus around what type of evidence should be accepted for clinical use and for coverage," said Deverka. "An important objective for CMTP in creating a forum for multi-stakeholder dialogue around these issues is to reduce the uncertainty about what evidence should be generated at each point in the development program, and thereby ensure that useful tests are developed, adopted, and paid for quickly, while those that provide no benefit, or may cause harm, are identified early."
In the guidance, CMTP suggests test developers use a “flow diagram” or a decision tree before conducting clinical utility studies for a diagnostic in order to hypothesize how physician use of the test might impact patient outcomes. "Ideally, the creation of the flow diagram would occur immediately after there was early evidence of the clinical validity of the test," Deverka said.
CMTP is working with groups such as the American Society for Clinical Oncology, as well as private payors, to gain their endorsement of the MDx clinical utility guidelines. "As the [guidelines] gain further visibility and acceptance … we expect that researchers, test developers, and payors will gain experience using the guidelines to both plan and evaluate studies," Deverka noted.
CMTP also plans to publish a shorter version of the MDx clinical utility guidelines in a peer-reviewed journal "so that research funders such as the NCI and PCORI can reference these guidelines in future request for proposals and the larger scientific community has an opportunity to cite this work," Deverka said.