Health care treatments are becoming increasingly tailored to individual patients in terms of patient preferences and individual clinical factors. However, these tailored treatments often have a high price tag. Frameworks or algorithms that are able to assess the value of a specific therapy for a specific patient are needed in order to preserve access to these therapies and better integrate them into the health system from a reimbursement standpoint. Effective value frameworks must incorporate patient preferences, clinical trial subgroup assessments, and post-trial analyses into their design.
The advent of personalized medicine means that therapies are being increasingly tailored to the clinical characteristics and preferences of patients. Obviously, this is a world we all want, where the drugs match who we are and what will work for us, but these therapies are typically costly. With rising concerns about health care costs, there has been increasing interest in assessing the value of drugs and other health care interventions.
Most recently, in May 2018, the US Department of Health and Human Services released a Request for Information on the HHS Blueprint to Lower Drug Prices and Reduce Out-of-Pocket Costs.1 Among other topics, the Administration sought input on value-based arrangements and price reporting, saying that value-based transformation of the entire health care system is one of the top 4 HHS priorities.2
Personalized medicine and driving for value are key directions in health care today. They can support one another; in fact, by narrowing a medicine’s use to when it is most effective as a wise use of resources. There are a range of issues to consider when attempting to successfully integrate personalized medicine and value frameworks though.
Current Value Frameworks
Many organizations, including the Institute for Clinical and Economic Review (ICER), the National Comprehensive Cancer Network (NCCN), and the American Society for Clinical Oncology, have developed value frameworks.3 But, when it comes to weighing the value of a medication, value assessments incorporate the perspectives and costs that are most relevant to health systems and payers—not the consequences that can impact patients and their families. Value is much broader than cost.
Clinical guidelines or evidence-based treatment protocols historically focused just on the clinical characteristics of treatment decisions, ie, benefits and harms. More recently, the NCCN has developed “Frameworks for Resource Stratification of NCCN Guidelines.”4 These frameworks consider available economic resources as well as benefits and harms. Combining clinical with economic considerations, the NCCN frameworks suggest a limited set of treatment options for countries with modest financial means and a broader set of treatments for countries with greater ability to fund health care. The NCCN frameworks have substantial clinical detail and clinically defined subgroups. However, they have less information about patient preferences and quality of life. As the frameworks are increasingly used world-wide, it will be important that they incorporate patient-derived personalized medicine elements.
If the application of the framework is not sufficiently personalized, then the value determination applies only to the average patient, not specific ones. Ideally, a framework should assess value based upon what matters to each patient. For example, one patient may treasure survival at the utmost, while another prefers a more tolerable regimen and sustained quality of life. In assessing value, the same drug regimen would likely have different value for these different patients. One use of a patient-specific value framework would be to assist in the physician-patient dialogue as they consider various therapeutic options. This type of shared decision-making approach is both feasible and desirable.
Patient-specific value determination would likely not be so straightforward for a payer. A payer needs to determine what regimens to approve and what patient financial copayments are required for all patients with a particular oncologic condition. Having a different value-based approach for each patient would be infeasible. Although each patient has unique characteristics, there are likely a finite number of preference subtypes (eg, those patients who might accept difficult treatment regimens in the pursuit of long-term survival, those patients who wish to maximize their current activities and lifestyle). A modest number of value determinations could reflect these key differences. In an analogous fashion, there are proposals to price a drug differently by clinical indication. A drug that can successfully treat one tumor type might have a higher price associated with it than the use of this same drug for a less responsive tumor. In a similar way, value that is personalized to patient preference subtype might be considered in discussions of price.
Value determinations lead to formulary decisions that can affect patient access to various drugs and therapies. For example, the US Department of Veterans Affairs announced its intent to use the ICER value assessment framework for pricing negotiations.5 Without personalized value determinations, then the decision whether to make a medicine available will not be sensitive and may hinder access for patients who would find significant value.6
Merging Value Assessment and Personalized Medicine
It is possible to combine value assessment with personalized medicine. In some cases, it can be done with ease; in others, it is more difficult. In the easiest instance, a clinical trial to gain drug approval a priori has a biologic test or genetic test. Only those patients with the right profile get the medicine in the trials, and the Food and Drug Administration (FDA) approves the medicine with the test associated with it. Clearly ICER or another value group can use that data and have an assessment that is personalized-medicine savvy—associated only with that biology or genetic profile, as in the case of some of the anaplastic lymphoma kinase lung cancer treatments. An increasing number of new oncology drugs have companion diagnostic tests making this approach increasingly feasible.7
In many cases, however, medicines are not tested for FDA approval in that manner, but post-trial data reveals that certain patient groups did better than others. These data are not as ideal as doing it up front, but it is of sufficient quality to be used for value assessments. KRAS genetic testing and the effect of cetuximab on patients with metastatic colon cancer can be used as an example: if a value determination was done before KRAS was discovered, then the clinical data would suggest that the treatment had only modest value. Once the KRAS test was applied, then there would be 2 groups of patients, one for whom the test was positive and one for whom it was not. For one of those groups, the value would be low; for the other, very high.
Value assessment becomes more challenging when there is no clinical trial and no hard biologic test that shows which patients do better or worse. A review of data after the trial might reveal softer characteristics of patients who do better or worse on the medicines, for example, in more severe cases—patients who have been failed by multiple other therapies.
Value is most difficult to demonstrate when there are no clinical trial data, but there is clear information on patient preferences. Treatments will have different value to different patients, even if an average value determination is made. As mentioned above, one patient may place greatest importance on survival, while another prefers a regimen that allows a maintained quality of life. One of the newest value frameworks for non-small cell lung caner,8 by the Innovation and Value Initiative (IVI), uses multicriteria decision analysis9 that incorporates these type of preference factors such as the difficulty of a regimen. This approach leaves room for patient preferences and could reflect value for individual patients.
Conclusion
In order to truly realize the potential of value frameworks in personalized medicine, framework developers need to incorporate subgroup assessments from clinical trials as well as relevant post-trial analyses. They also need to both consider and incorporate patient preferences. Preferences are the most challenging to handle, but groups like IVI have shown that it can be done. Other organizations working on their own frameworks should incorporate similar patient-centric and personalized medicine approaches.
Read the counterpoint to this perspective here.
References
1. US Health and Human Services Department. HHS blueprint to lower drug prices and reduce out-of-pocket costs, request for more information. Federal Register. 2018;83(95):22692-22700.
2. Azar AM II. Value-based transformation of America’s healthcare system [speech transcript]. Hhs.gov website. https://www.hhs.gov/about/leadership/secretary/speeches
2018-speeches/value-based-transformation-of-americas-healthcare-system.html. Published March 8, 2018. Accessed November 27, 2018.
3. Slomiany M, Madhavan P, Kuehn M, Richardson S. Value frameworks in oncology: comparative analysis and implications to the pharmaceutical industry. Am Health Drug Benefits. 2017;10(5):253-260.
4. National Comprehensive Cancer Network (NCCN). NCCN Framework for Resource Stratification of NCCN Guidelines (NCCN Framework™). nccn.org website. https://www.nccn.org/framework/. Accessed November 27, 2018.
5. Brennan Z. ICER to work with VA on drug price negotiations. Regulatory Focus. July 3, 2017. https://www.raps.org/regulatory-focus%E2%84%A2/news-articles/2017/7/icer-to-work-with-va-on-drug-price-negotiations. Accessed November 27, 2018.
6. Dubois RW. Caution warranted as VA incorporates ICER value assessments into formulary management process [blog]. Health Affairs. September 18, 2017. https://www.healthaffairs.org/do/10.1377/hblog20170918.062021/full/. Accessed November 27, 2018.
7. Serluco JT, Oyekan E. Companion diagnostics and the future of oncology clinical trial design. Clinical Leader. July 30, 2018. https://www.clinicalleader.com/doc/companion-
diagnostics-and-the-future-of-oncology-clinical-trial-design-0001. Accessed November 27, 2018.
8. Innovation and Value Initiative (IVI). Application in the real world. Thevalueinitiative.org website. https://www.thevalueinitiative.org/application-real-world/. Accessed November 27, 2018.
9. Angelis A, Kanavos P. Multiple criteria decision analysis (MCDA) for evaluating new medicines in health technology assessment and beyond: the advance value framework. Social Sci Med. 2017;188:137-156.
