The Times Are Changing
Recently, sanofi-aventis was forced to reduce the cost of its new cancer drug Zaltrap by 50% in response to Memorial Sloan-Kettering Cancer Center’s decision to not use the drug. The cancer center stated that the new drug was twice as expensive as Avastin (manufactured by Genentech) despite similar efficacy. Both drugs improved median survival by 1.4 months.1 In the New York Times article that reported the story, a senior vice president for oncology at UnitedHealthcare stated that “It was the first time physicians have stood up and said, ‘Enough is enough.’” Many believe that this small story in the New York Times was a watershed moment. “It is shifting the fulcrum of the return on investment in oncology innovation,” remarked Dr Clifford Goodman, PhD, senior vice president at The Lewin Group, at the beginning of his keynote address entitled Health Care Reform: What Now for Evidence Requirements, Personalized Care, and Innovation in Oncology?
So what are the marketplace trends affecting innovation in oncology? Dr Goodman outlined 14 such trends:
1. Scientific and technical advances (eg, lower costs for genome sequencing)
2. Diminished blockbuster opportunities
3. Restructuring and consolidating life science industry
4. Changing research and development roles of big and small companies
5. Shifting relative magnitudes of international markets for oncology interventions
6. Varying and unpredictable international intellectual property protection
7. Global effects of reference pricing and other downward price pressure by national and regional health authorities
8. Complex, variable relationships between the evidence requirements of regulators and payers, and greater reliance on health technology assessment
9. Greater demand for evidence of comparative effectiveness of new interventions (not just absolute efficacy versus placebo)
10. Greater emphasis on discerning heterogeneity of treatment effects
11. Increasing patient centeredness, including greater emphasis on patient-reported outcomes
12. Need to validate companion diagnostics (eg, for personalized medicine), including evidence of their impact on treatment decisions and outcomes
13. Increasingly broad and sensitive postmarket detection of adverse events
14. Increasing yet varying demands of payers for demonstrating costeffectiveness and/or acceptable budget impact
Innovation Isn’t Getting Any Easier
Dr Goodman highlighted why industry faces greater challenges to adding a new oncology drug or procedure into the marketplace. Gaining market approval by the US Food and Drug Administration (FDA) is not the guaranteed ticket to market success. Innovators must pass the scrutiny of regulatory and payment gatekeepers, and the 2
groups often have different evidence requirements (.2 According to Dr Goodman, although payer evidence requirements do not substitute for those of regulators, they are increasingly mediating adoption and use in the marketplace. For example, the main question of regulators continues to be: “Does the drug work?” For payers, it is more likely: “How well does the drug work in the community or general practice compared with standard care?” The answers to these questions often require different types of evidence, which can complicate new therapy development and validation. The evidence could be generated using practical (pragmatic) clinical trials, other nonrandomized controlled trials, adaptive clinical trials, and other trial designs (eg, randomized consent, regression discontinuity, combined single- subject [“n of 1”] trials). Prospective and retrospective observational studies should also be considered, such as population-based longitudinal cohort studies, patient registries, claims databases, clinical data networks, electronic health record data analyses, and passive and active postmarketing surveillance. Dr Goodman also noted that payers may draw on syntheses of existing evidence (eg, systematic reviews, meta-analyses, and modeling). These additional types of studies can supplement payers’ understanding of how therapies work in more diverse and representative populations. Dr Goodman emphasized that “This does make it more difficult to innovate. It is more demanding of evidence. It is more demanding of continued data generation, collection, analysis, and feedback.”
What to Measure?
Dr Goodman stated that innovation in oncology increasingly calls on researchers to use multiple study designs (as listed above). To illustrate the types of evidence questions posed by payers, he showed a typical analytic framework for diagnostics and therapeutics ().3 Dr Goodman stressed that test accuracy is often not sufficient for validating tests in clinical use, and that clinicians, payers, and other decision makers also seek direct or indirect evidence linking tests to clinical decisions and health outcomes.
End Points and Biomarkers
The most sought end point in oncology studies typically is overall survival, defined as the time from study randomization until death from any cause; it is best measured in the intent-to-treat population. However, the FDA and other parties will analyze other end points, including
• Tumor assessments:
º Disease-free survival: time from randomization until recurrence of tumor or death from any cause
º Objective response rate: proportion of patients with tumor size reduction of a predefined amount and for a minimum time period
º Complete response: no detectable evidence of tumor
º Time to progression (TTP): time from randomization until objective tumor progression; does not include death
º Progression-free survival: time from randomization until objective tumor progression or death
• Symptom assessment:
º Patient perspective of direct clinical benefit
º Patient symptom assessments and/or physical signs representing symptomatic improvement (eg, weight gain, decreased effusion)
º Time to progression of symptoms: Similar to TTP, a direct measure of clinical benefit rather than a potential surrogate4
Important work on identifying and validating surrogate end points for regulatory purposes and clinical decisionmaking continues. A recent example is the 2012 FDA guidance on using pathologic complete response as a surrogate end point in neoadjuvant treatment of high-risk early-stage breast cancer to support accelerated FDA approval that would be linked to requirements for subsequent data collection to demonstrate the relationship of the surrogate to health outcomes.5
Dr Goodman noted that the validity of surrogates must be continually scrutinized. For example, although increased bone density has been used as a surrogate end point for osteoporosis medications, it is not necessarily a reliable predictor of decreased fracture rates. Similarly, despite widespread use of prostate-specific antigen levels for clinical decision making, it is often not a reliable predictor of mortality and other outcomes in prostate cancer management.
Genetic testing is also subject to greater scrutiny. Evidence of analytic validity (test accuracy for a genotype) must increasingly be accompanied by evidence of clinical validity (test accuracy for identifying a clinical condition or other phenotype) and, further, evidence of clinical utility (impact on clinical decisions and health outcomes). Although evidence is commonly lacking for clinical utility and clinical validity for assessing risk or otherwise managing disease, it can be lacking for analytic validity as well.6 Dr Goodman noted that even when such evidence is available, payers recognize that the data trail in practice from patient indications for a test to test ordering to test result to indicated clinical decision is typically incomplete and otherwise poorly documented.
Dr Goodman noted that “Personalized medicine is not simply about the genomics.” Personalized medicine is intended to tailor medical care to the particular traits (or circumstances or other characteristics) of a patient that influence response to a healthcare intervention. In addition to genetic traits, these may be sociodemographic, clinical, behavioral, economic, environmental, and other personal traits, as well as personal preferences. Dr Goodman emphasized that, for any form of cancer, there may be a great diversity of such traits—including where and how people live and what their incentives are—that can affect individual patient response across the broader at-risk population and, therefore, should be weighed in determining optimal care for individuals. Dr Goodman described how the conduct of subgroup analyses in comparative effectiveness research and patient-centered outcomes research to discern heterogeneity of treatment effects reflects, in part, the need to yield evidence to support personalized medicine.
Evidence for the Medicare Evidence Development & Coverage Advisory Committee
To illustrate what types of evidence are of increasing interest to payers, Dr Goodman referred to his recent experience on the Medicare Evidence Development & Coverage Advisory Committee (MedCAC), which advises the Centers for Medicare & Medicaid Services about the quality of evidence pertaining to Medicare national coverage determinations. The questions placed before the MedCAC emphasize whether there is adequate evidence to determine whether a technology, therapeutic, or diagnostic has an impact on patient outcomes and, if the evidence is adequate, what it demonstrates. When the available evidence shows favorable impacts on patient outcomes, the MedCAC asks whether it also applies to community-based settings and to Medicare beneficiaries in particular.
Dr Goodman gave 2 examples from the MedCAC. One was pharmacogenomic testing for 5 types of cancer, including HER2/neu for patients with breast cancer who are candidates for trastuzumab and K-RAS for patients with metastatic colorectal cancer who are candidates for cetuximab and/or panitumumab. Rather than focusing on the accuracy of these tests as such, the MedCAC questions addressed the evidence for whether each pharmacogenomic test improves outcomes for patients with cancer whose anticancer treatment strategy is guided by the test results and how well that evidence pertains to Medicare patients in community settings.
In the second example, Dr Goodman discussed how MedCAC examined evidence for FDA-approved Provenge (sipuleucel-T), an immunotherapy for the treatment of asymptomatic or minimally symptomatic metastatic castrate-resistant prostate cancer. The committee examined available evidence about whether the therapy would improve overall survival, control disease-related symptoms, and minimize patient burdens associated with anticancer therapy, including access to care, delivery, and side effects. The MedCAC was generally confident about the quality of the evidence demonstrating some improvement in overall survival and minimizing patient burdens, but not for control of diseaserelated symptoms. Further, given that physicians in practice can prescribe FDA-approved therapies for other indications, the MedCAC examined whether the evidence pertaining to the FDAapproved indication was generalizable to 3 off-label uses of Provenge: prostate cancer that has not metastasized; metastatic, castrate-resistant disease with more severe symptoms; and metastatic prostate cancer that has not failed hormonal therapy. The MedCAC expressed low confidence in the evidence supporting all 3 off-label indications. Finally, the MedCAC looked at the generalizability of these findings to patient groups that may have been under-represented in the available clinical trials and community settings. Dr Goodman emphasized that these types of questions are asked as well by public and private sector payers.
Dr Goodman cited Provenge in returning to the theme of innovation and value. Media reports have highlighted the price of Provenge, which was $93,000 per course of therapy at the time of the MedCAC meeting, and several public commenters raised concerns that this would affect the Med- CAC’s deliberation. However, he made it clear that the MedCAC does not address matters of cost, and none of its evidence questions addressed the cost of Provenge. What is apparent, though, is that the cost of new cancer therapies is integral to their value. To illustrate, Dr Goodman referred to the survival curves presented at the MedCAC meeting displaying the difference in overall survival between the Provenge treatment group and the placebo group in the pivotal randomized controlled trial ().7 There was a 4.1-month median difference in overall survival. He stated that, while the Medicare program does not consider cost-effectiveness, many cannot help but wonder whether the space between those 2 curves is worth $93,000. Dr Goodman stated “I will suggest that type of question is being asked more and more, and was probably in the minds of the Memorial Sloan-Kettering folks in the New York Times article I showed you earlier.”
Paraphrasing a former defense secretary, Dr Goodman said that when it comes to newly approved, and even newly reimbursed, drugs, “We often don’t know what we don’t know.” As noted above, upon regulatory approval, there is often insufficient evidence for how well a drug will work for more diverse patients in real practice, as well as for rare or delayed adverse events. In US and international markets, payment authorities and life sciences companies have been negotiating various forms of pay-for-performance, risk-sharing, and other types of managed entry for these drugs. For example, in Germany, Roche offered a payfor- performance arrangement on its drug Avastin for several types of cancer. Payers and cancer care providers get refunds if patients do not respond. As with most deals of this nature, various concerns arise about provider incentives and implications for patient access and health. In any case, such arrangements can enable some market access and payment where there might have been more resistance by payers and other gatekeepers.
Similar deals have been struck for other drugs in various countries. For example, in the United Kingdom, AstraZeneca provides Iressa for free for the first 2 cycles of first-line use in patients with EGFR mutation-positive non-small-cell lung cancer. If the patient responds to treatment, the National Health Service will pay a fixed per-patient price of £12,200, regardless of how long the drug is used.8
Dr Goodman stated that the nature of demand for innovation in healthcare is changing. Beyond new or different, and even in addition to more effective, there is increasing, explicit demand for innovation of value, that is, achieving desirable or acceptable improvement in outcomes per incremental expenditure. Payers and more clinicians and patients want to know “Is it worth it?” Of particular note in healthcare systems undergoing change are “disruptive innovations” that can change healthcare regulation, payment, delivery, professional training, and other market characteristics. Examples have included diagnostic imaging that replaced exploratory surgery and gene therapy that is replacing some drugs and biologics.
Concluding Remarks: Moving Toward Breakthrough Innovations
Dr Goodman concluded with considerations for pursuing breakthrough innovation. Among these are to identify objectives for meeting specific unmet health needs as well as the target users and relevant decision makers for such a breakthrough. This involves identifying the attributes that will distinguish the drug for those target groups, whether this is a clinically meaningful treatment effect, a reduction in adverse events, increased adherence, or a reduction in healthcare utilization or costs. Innovators must anticipate and specify the evidence that will be required to validate these attributes and the feasibility of doing so, including costs, timeline, and other practical aspects. Recognizing that market conditions will continue to evolve, innovators must make decisions under uncertainty whether to pursue, adjust, or shelve the innovation, and reallocate resources accordingly. Dr Goodman made it clear that no one said this would be easy, but that innovation has to reach the higher bar in today’s marketplace.
1. Pollack A. Sanofi halves price of cancer drug Zaltrap after Sloan-Kettering rejection. New York Times. November 8, 2012. http://www.nytimes.com/2012/11/09/business/sanofi-halvesprice-of-drug-after-sloan-kettering-balks-at-paying-it.html?_r=0 . Accessed November 20, 2012.
2. Schumock GT, Pickard AS. Comparative effectiveness research: relevance and applications to pharmacy. Am J Health Syst Pharm. 2009;66(14):1278-1286.
3. Harris RP, Helfand M, Woolf SH, et al; Methods Work Group, Third US Preventive Services Task Force. Current methods of the US Preventive Services Task Force: a review of the process. AmJ Prev Med. 2001;20(3 suppl):21-35.
4. US Food and Drug Administration. Guidance for Industry: Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics. http://www.fda.gov/downloads/Drugs/Guidance-ComplianceRegulatoryInformation/Guidances/ucm071590.pdf. Published May 2007. AccessedDecember 10, 2012.
5. US Food and Drug Administration. Guidance for Industry: Pathologic Complete Response in Neoadjuvant Treatment of High-Risk Early-Stage Breast Cancer: Use as an Endpoint to Support Accelerated Approval. Draft Guidance. http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm305501.pdf. Published May 2012. Accessed November 20, 2012.
6. Little J, Wilson B, Carter R, et al. Multigene Panels in Prostate Cancer Risk Assessment. Evidence Report No. 209. (Prepared by McMaster University Evidence-based Practice Center under Contract No. 290-2007-10060-1.) AHRQ Publication. No. 12-E020-EF. Rockville, MD: Agency for Healthcare Research and Quality; July 2012.
7. Kantoff PW, Higano CS, Shore ND, et al; IMPACT Study Investigators. Sipuleucel-T immunotherapy for castration-resistant prostate cancer. N Engl J Med. 2010;363(5):411-422.
8. Senior M. Risk-sharing in oncology: Astra- Zeneca’s single payment access scheme shows where it works. May 2, 2012. http://realendpoints.com/research/risk-sharing-in-oncologyastrazenecas-single-payment-access-schemeshows- where-it-works/. Accessed November 20, 2012.