Assessing New Medications Using Evidence-Based Value

St. Louis—The US Food and Drug Administration (FDA) must approve new medications, but not all reviews are supported by evidence-based research and data supporting validation of performance in a clinical setting. Likewise, there is variability in the comprehensiveness and reliability of national compendia for options for treatment of conditions such as cancer, creating possible confusion in the healthcare system and the creation of differing levels of acceptable evidence supporting the value of a medication, making informed decision-making difficult.

At a managed care essentials session at the AMCP meeting, attendees heard discussions on what constitutes high-quality scientific evidence, how to apply evidence in determining overall value of a medication, and how to identify inconsistencies in national compendia recognized by the Centers for Medicare & Medicaid Services (CMS). The session, Evidence-Based Value Analysis of New Medication—2010 Update, focused on oncology and biotechnology/specialty medications. The first speaker was Helen Sherman, PharmD, chief pharmacy officer, RegenceRx. She began with an overview of 6 new oncology medications, noting that only 3 of the 6 have had randomized controlled trials. The 3 medications with randomized trials are everolimus, pazopanib, and cabazitaxel; the others are pralatrexate, ofatumumab, and romidepsin. She continued with information on biosimilars, medications that require fewer data than interchangeable biologics for FDA approval. She said that other than in 2009, there has been little innovation in new molecular entities/biologics since 1999, but there has been a steady increase in the number of new drug applications and biologic license applications approved by the FDA for new formulations of existing medications, new combinations, new indications, or over-the-counter products.

Data on the reliability of scientific studies, based on a RegenceRx appraisal using CONSORT standards, found that 8% of studies conducted from 2006 through the second quarter of 2010 (n=1165) were reliable. Reliable studies contain 4 essential components: randomization, a meaningful end point, blinding, and power (sample size). Issues associated with randomization include >5% of participants unaccounted in final analysis, >5% difference in loss to follow-up between ≥2 arms of the study, and >10% but <20% loss to follow-up within a study. Any of these issues lead to uncertainty and the study can “fall apart,” Dr. Sherman said.. Issues with the end point include end points with uncertain clinical significance or relevance, nonstandard or nonvalidated end point for the disease being studied, and a subjective end point that may introduce potential for bias. Issues associated with blinding include an open-label study with a subjective end point, a situation Dr. Sherman termed “particularly troublesome.” Dr. Sherman concluded her presentation with the top opportunities to improve the quality of scientific studies: (1) analyze all randomized subjects, (2) ensure solid blinding and concealment of allocation, (3) select a robust sample to support good power for the study, (4) establish clinically valid and objective end points, and (5) decrease the attrition rate throughout the study.

Guidelines and Compendia

The session continued with a presentation from Lynn Nishida, BS, RPh, director of clinical pharmacy services at RegenceRx. She began by listing issues surrounding the reliability of various compendia and practice guidelines, citing a critical appraisal of 431 clinical practice guidelines that found 88% were unclear as to whether a systematic literature search had been performed, 82% lacked details of application of explicit evidence grading, and 67% were missing descriptions of the individuals involved in the development of the guideline. A 2007 Agency for Healthcare Research and Quality Tech Assessment Agency appraisal of compendia found potentially outdated information, use of inconsistent methods to critique studies and report levels of evidence supporting or refuting off-label use, a lack of transparent decision-making processes and methodology for including scientific information, variation in the level or extent of reporting safety information, and potential conflicts of interest or financial ties.

Ms. Nishida continued by outlining potential impacts of health plans using guidelines and compendia, using the example of cancer medications. Plans should compare compendia differences for medically accepted uses based on CMS definitions, apply compendia to the plan’s claims experience, and determine how many would have met or not met National Comprehensive Cancer Network criteria. In conclusion, she listed considerations and steps health plans should apply to guideline/compendia use. Practical applications include prescriber responsibility to provide evidence-based medicine documentation, consideration of resources, and assessment of the impact on the member, prescriber, and plan. A general philosophy would include recognition that compendia are not the ultimate source for coverage determination, an understanding of the variation of evidence accepted by the compendia, and weighing the strength of the evidence-based medicine used in the creation of the compendia.

New Medications

The session concluded with a presentation from Laurie Wesolowicz, PharmD, director of pharmacy services–clinical, BlueCross BlueShield of Michigan. She provided a list of selected, recently approved medication classes, including biologic/specialty medications to treat inflammation and pulmonary arterial hypertension, “firsts” approved by the FDA (a drug to treat infantile spasms and one to increase walking speed in multiple sclerosis [MS]), as well as a medication to treat advanced/metastatic renal cell carcinoma.

The 2 biologics most recently approved by the FDA for inflammatory conditions are tocilizumab and ustekinumab. Dr. Wesolowicz said that ustekinumab has raised concerns at the FDA about a possible increase in infections, but it has a unique mechanism of action (MOA) and will be closely monitored for efficacy and safety. Tocilizumab also has a unique MOA and may be effective when other options prove ineffective, she said. There is no cure for MS, but medications can decrease the intensity of the exacerbations of the disease, Dr. Wesolowicz said. New options include fingolimod and dalfampridine. Fingolimod is used to treat remitting/relapsing MS (RRMS), and is the first oral disease-modifying option for RRMS. Dalfampridine is an oral medication approved to increase walking speed. Dr. Wesolowicz said that it may be a useful add-on treatment as a disease-modifying agent. The 3 classes of treatment options for pulmonary arterial hypertension treatment include endothelin receptor antagonists, phosphodiesterase type 5 (PDE-5) inhibitors, and prostanoids.

There are quality clinical data available on at least 1 drug in each of the 3 classes, Dr. Wesolowicz said: sildenafil (PDE-5 inhibitor), ambrisentan (endothelin receptor antagonist), and iloprost (prostanoid). She advised that formularies should include at least 1 oral, 1 infusion, and 1 inhalation drug for pulmonary arterial hypertension. Vigabatrin has been approved by the FDA for treatment of infantile spasms, a rare, severe seizure condition in infants. The approval has been 20 years in coming, because of concerns of serious visual toxicity. The approval includes a requirement for a risk evaluation and mitigation strategy to address those concerns. The last medication Dr. Wesolowicz discussed was for treatment of advanced/metastatic renal cell carcinoma, a condition with highly variable survival rates, making it difficult to assess treatment effectiveness of new therapies. There have been 2 new products approved by the FDA since 2007: temsirolimus and everolimus. The evidence on the value of the treatments is uncertain, according to Dr. Wesolowicz. In conclusion, she acknowledged the increasing complexity of assessing the value of new medications, particularly oncology and biotechnology products. She said that official compendia contain sometimes conflicting information and may not “tell the whole story” about a new drug.—Tori Socha