We are entering a bold new period for personalized medicine and molecular IVDs, highlighted by the AMA’s approval of more than 100 new genomic test codes for 2012. But there are still some urgent problems in our regulatory and reimbursement policy arenas.
The regulatory problem is the disincentive that our mixed environment of approved IVDs and competing laboratory-developed tests (LDTs) impose on IVD companies that make long-term investments in product development and FDA submissions for companion diagnostics. To meet FDA standards, the pharmaceutical developer and the diagnostic manufacturer enter a years-long resource-intensive collaboration followed by a potentially excruciating review to bring both drug and test to market.
Their hard-won right to sell a complete diagnostic kit is eroded in the marketplace by copy-cat analyte-specific reagents (ASRs) or unregulated “research use only” (RUO) reagents. Copycat products distributed via RUO channels are sold without data on clinical validity or shelf-life testing. The resulting LDTs lack clinical correlations data with actual patient response or FDA-approved control tissue kits. Lost is the quality assured by FDA review and maintained by regulated, but more costly, manufacturing. Copy-cat products can displace both FDA-validated tests for drug selection and cancer-detecting tests to diagnose conditions like bladder cancer. Physicians, patients, and payers are unlikely to know the provenance of the patient’s test.
The reimbursement problem is tied to the 100-plus new genetic Current Procedural Terminology (CPT) codes. The AMA’s policy group for physician services conveyed these new codes to Medicare in 2011 with recommendations to value the codes under a method called the physician fee schedule-and that’s the problem. The physician fee schedule undervalues capital equipment, provides no accounting for our investments in R&D costs or intellectual property, imposes co-pays from which clinical chemistry tests (such as genetic tests) are exempt, and portrays the test result as the personal service of a physician, equivalent to reading a glass-slide breast biopsy. Various scaling reductions in the CMS accounting process yield a Medicare payment that is below the cost of the inputs on the laboratory bench. This tilts the healthcare system to be more accommodating to LDTs than to regulated and validated IVDs. Fortunately, at least so far, CMS has declined to accept the AMA’s recommendation: the agency has deferred use of the 2012 codes until 2013, allowing an extra year for the agency to choose the appropriate price-setting process.
The policy problems described herein have policy solutions. On the regulatory front, FDA continues to exercise enforcement discretion in the regulation of LDTs. While it would be unwieldy or impossible to regulate all LDTs, the proliferation of copycat RUO products that quickly follow the approval of a safe and effective IVD is a risk to patient safety and public health. Cracking down on obvious RUO products that mimic an approved test would be a defensible position for FDA to take. There is a second avenue here, as well: where the FDA label for a drug requires use of an “approved or cleared” companion test, payers might insist on CPT codes, modifiers, or policies that favor unequivocally validated test products. If a few payers issued flagship policies in this direction, others would follow.
The reimbursement issue for the new genetic codes will be aired publicly by CMS this summer. The future of Medicare payment policy will be open to public comment at CMS in July and August, and both IVD manufacturers and trade groups should pay attention and provide comment to CMS. Moving genomic tests to a discounted and copay-based physician fee schedule is a drawback to the industry. If this change in CMS payment policy is approved for human genetic tests, the same pricing approach could later be applied to infectious-disease molecular diagnostics. Medicare’s own regulations (42 CFR 415.130) state that physician pathology services must be surgical pathology or otherwise “require” the personal work of a pathologist. This is true for some molecular tests based on in situ hybridization and physician review at the microscope. But, in general, it is crystal clear that automated IVD molecular tests do not require a licensed physician to personally supervise and interpret each result with a narrative medical report. This summer, the IVD industry should ensure that Medicare authorities are careful in setting the rules for reimbursing genomic tests, as the current direction is creating uneven precedents with unfavorable implications.

Bruce Quinn, MD, PhD, is Senior Health Policy Specialist at Foley Hoag LLP (Boston). He can be reached via e-mail at bquinn@foleyhoag.com.