Only 15 percent of drug labels claiming drug-biomarker interactions were found to convincingly demonstrate clinical utility, and nearly two-thirds of drug labels did not convincingly prove either clinical validity or utility. Based on this study, authors concluded that it is likely premature to include biomarker testing recommendations in drug labels or standard practice.

Overall, the authors pulled labels identifying 119 drug-biomarker combinations for 107 drugs from FDA’s database and assessed the quality and completeness of the evidence presented as well as whether the label made a clear recommendation. Therapeutic areas identified included: oncology (37/119), neuropsychiatry (33/119), gastroenterology (9/119), infectious disease (9/119), cardiovascular disease (8/119), and others.

Clinical Utility

The authors found that nearly 85 percent of labels included incomplete information on clinical utility. Targeted therapies, which consist mainly of oncology drugs, were found to be much more likely to contain convincing utility evidence (50 percent) than non-targeted therapies (1.2 percent). Even among the oncology products, authors cite that 62 percent of labels were incomplete on clinical utility information, with only about a third considered convincing. Compared to the 62 percent incomplete rate for oncology products, 100 percent of labels for three other therapeutic areas (neuropsychiatry, gastroenterology and cardiovascular disease) were incomplete and/or had insufficient information for utility.

Clinical Validity

Evidence of clinical validity fared somewhat better. About 36 percent of all labels were considered complete, and nearly two-thirds of oncology labels were considered complete. Regardless, the majority of labels still provided incomplete information for neuropsychiatry, gastroenterology and cardiovascular disease (91 percent, 78 percent and 75 percent respectively).

The study reiterates opinions long expressed by payers, including members of the Medicare Evidence Development & Coverage Advisory Committee (MEDCAC). In meetings on pharmacogenomics testing in 2009 and in 2010, the Committee expressed variable confidence in the evidence base for various pharmacogenomic oncology biomarkers and low confidence in the ability of biomarkers to change physician-directed patient management. Although clinical utility data is not necessary for FDA approval, manufacturers should take heed on provider and payer perceptions of the evidentiary strength supporting drug-biomarker products. Avalere recently published a perspective and roadmap on the changing regulatory and reimbursement policy trends affecting market access for Diagnostics.

For more information about clinical utility or clinical validity, please contact Lakshman Ramamurthy at LRamamurthy@Avalere.com.

View the full JAMA Internal Medicine study.