Amgen Response to ICER’s Proposed Adaptation of the ICER Value Framework for the Assessment of Treatments for Ultra-Rare Conditions

Executive Summary

Amgen appreciates the opportunity to comment on ICER’s “Proposed adaptation of the ICER value framework for the assessment of treatments for ultra-rare conditions” and we hope that ICER will consider and incorporate our recommendations.

Amgen is concerned that ICER’s proposed framework adaptation runs counter to the U.S. Orphan Drug Act (ODA) of 1983, the legislation designed to incentivize innovation and protect orphan disease patients.  ICER has taken a positive step in its acknowledgement that orphan/ultra orphan diseases require special consideration.  However, ICER’s approach is contrary to both the ODA and the U.S. Federal Drug Administration (FDA) regulatory framework, which has special provisions for orphan diseases.  This legislation provides incentives to address the significant societal burden and high-unmet need of the 25-30 million American patients with rare, disabling, burdensome, and almost uniformly fatal conditions.^{1, 2, 3} Half of those with orphan diseases are children; an estimated 30 percent of children who have an orphan disease will not survive beyond their fifth birthday and only one in 10 orphan disease patients has a viable drug  for their condition.⁴

ICER’s proposed framework to assess the value of orphan drugs (through the common framework) and ‘ultra-orphan’ drugs (through the adapted framework) undervalues patients’ suffering from often life-threatening orphan diseases and the medicines developed to treat them.  Contrary to the Orphan Drug Act, ICER’s proposed ‘ultra-orphan’ adapted framework ignores the majority of orphan diseases, subjecting them to ICER’s common framework. The proposed adapted framework fails to capture the significant burden these conditions place on this population.  It excludes the costs of patients, caregivers, employers and society, undervaluing the ability of new treatments to offset the significant burden of ‘ultra-orphan’ disease; it does the same to orphan diseases by defaulting to ICER’s common framework. Moreover, in practice, no drug ICER has assessed to date, or plans to assess, would qualify as an ICER-defined ‘ultra-orphan’ drug.

ICER proposes a limit on how much should be spent on the health of orphan disease patients.  ICER recommends a willingness-to-pay (WTP) threshold for orphan disease (by inappropriately applying the common disease framework) and for ‘ultra-orphan’ disease (through this adapted framework) in the U.S.  This chosen WTP limit is not informed by what U.S. citizens or the government would want to spend.  This will likely have consequences in slowing the pace of scientific innovation necessary to improve quality of life and potentially find cures for all ODA-defined diseases.  Price thresholds similar to those seen in other countries is not only inappropriate for the U.S. system, but also well studied.^{5, 6} With 17 ODA-defined orphan drugs approved per year at the current pace, it would take nearly 400 years for researchers to find drugs for the millions of remaining patients.⁷  This is in the absence of ICER’s proposed approaches to the assessment of orphan and ‘ultra-orphan’ disease.  However, the wide-reaching effects of this proposed value-framework for ODA-defined orphan drugs, if implemented, could add a further century to the pace of orphan drug development.

The value of orphan and ‘ultra-orphan’ disease drug treatments is not in question nor do they drive healthcare costs.  The total annual cost of treating patients with orphan disease could be conservatively estimated at two trillion dollars with orphan drug costs making up only 4% of this cost.⁸  Although orphan drugs make up $68.7 billion in spend per year in the U.S., their ability to alleviate the grave effects of orphan disease is significant.^{9, 10} Orphan diseases account for $324 billion in lost productivity costs to patients: these are costs that will not be captured in ICER’s proposed narrow approach to assessing ‘ultra-orphan’ and orphan drugs.^{11, 12} For the 1 in 10 patients who have access, orphan drugs can considerably offset total costs that would have been incurred without an available drug, as measured in healthier Americans living more productive lives.  Moreover, orphan drugs contribute $149 billion to the economy and for every dollar spent in the development of an orphan drug, there is a return of 2.4 dollars to the economy every year.¹³  Not only do orphan drugs have a net positive effect on costs and economic return to the economy, but the claim that orphan drugs are putting drug budgets in peril does not hold up against the available evidence.^{14, 15, 16, 17, 18}  With an estimated $750 billion to one trillion in healthcare spend wasted on areas such as unnecessary services and administration costs, ICER has the opportunity to provide insights in areas transformative for U.S. healthcare.^{19, 20}

ICER’s goal is to create a “more effective, efficient, and just health care system”,²¹ however, ICER’s proposed approach for assessing orphan drugs could unfortunately do more harm than good to orphan disease sufferers in its current form. The ODA was created because there is a recognition that patients with rare and ultra-rare disease need to be protected with special measures. This ‘rule of rescue’ is the social contract made to ensure those with orphan and ‘ultra-orphan’ diseases are protected.  ICER has the potential to make a positive contribution to healthcare decision-making by providing information that helps navigate the complex landscape of clinical outcome data.  ICER can work to ensure the voices of all Orphan Drug Act-defined orphan disease patients are heard and accurately reflected in ICER assessments.  To help reach these goals, Amgen specifically recommends the following changes to ICER’s proposed framework adaptation for orphan disease:

1. Align the framework with the definitions and provisions in place to protect orphan disease patients
2. Ensure the patient voice is heard and patients are put at the center of assessments
3. Do not attempt to set an arbitrary national threshold for orphan drugs
4. Include costs and cost savings resulting from drugs that are relevant to all stakeholders
5. Do not apply the ICER “Final Value Assessment Framework for 2017-2019” to orphan drugs: the methodological concerns applied to common disease would be further amplified in orphan disease

ICER’s focus should be on guidance in helping to navigate the complex landscape of clinical outcomes.  Our major concerns with ICER’s proposed approach and our recommendations are further elaborated below.

1) ICER’s de facto classification of the majority of orphan diseases as common diseases runs counter to the Orphan Drug Act, the U.S. law put in place to protect these patients. 

ICER’s  proposed ‘non-ultra-orphan’ designation for orphan disease drugs with 10,000-200,000 patients is in direct conflict with legislation designed to encourage orphan drug development.

• The Orphan Drug Act (ODA) in 1983 clearly defines an orphan disease as those that affect 200,000 or fewer individuals.²²  ICER’s proposed definition of ultra-rare disease (i.e., no more than 10,000 patients in the United States) essentially redefines the criteria set out in the Orphan Drug Act and will leave many patients with ‘non-ultra orphan’ diseases disadvantaged.^{23, 24, 25}
• The epidemiology assessment of orphan disease should be based on the science and historical probability of success and eliminate potential for subjective bias.²⁶ ICER’s definition of ‘ultra-orphan’ on the basis of a drug that has “little chance of future expansion of indication or population” to above 20,000 individuals, is less data based and more a product of estimation or subjective judgment. ^{27, 28} The value of an ODA-defined orphan indication should be divorced from the number of patients a drug treats in another indication.
• Empirical application of the common drug framework to ODA-defined orphan drugs could be disastrous for patients.  Every past ICER assessment of an orphan condition has used ICER’s common drug framework and this has resulted in a recommendation of ‘low value’.  Although the panel has been able to overturn this in the past, with ICER’s revised value framework for common drugs and ‘ultra-orphan’ drugs, the panel can no longer make these deliberations. [Appendix, Table 1]
• ICER should consider comments on its proposed approach to assessment of ODA-defined orphan drugs before applying it to assessments.  For example, in ICER’s finalized scoping document for CAR-T in R/R B-cell ALL, this severe pediatric disease (for which child patients are faced with a shrinking number of drugs for cancer that is no longer responding to treatment) is classified by ICER as a common disease on the basis that CAR-T may at some time in the future be used to treat other diseases.  ICER also defines hemophilia A with inhibitors as a ‘common disease’ despite there being less than 10,000 patients in the U.S.²⁹  ICER is evaluating treatments based only on estimates of potential treatment population before a drug is approved for that indication.

2) ICER’s proposed framework deprives orphan disease patients of a voice in determining value and access to drug treatments.

The complex nature of measuring value and contextual considerations cannot be captured in ICER’s proposed approach to assessing ‘ultra-orphan’ and orphan drugs.

• ICER’s proposed approach to evaluating evidence in ‘ultra-orphan’ and orphan drugs (the latter defaulting to their common framework), unreasonably penalizes all new ‘ultra-orphan’ and orphan drugs for lack of evidence at FDA approval.  It disregards orphan disease characteristics such as extremely small populations, softer endpoints (which are harder to measure) and difficulties in running clinical programs. ^{30, 31, 32} Also, ICER does not make sufficient allowances for assessments to capture the important nuances in orphan and ‘ultra-orphan’ disease assessment, including society’s ‘rule of rescue’ as well as patients’ ‘value of hope.^{33, 34} 

ICER’s voting process for orphan drugs marginalizes the independent public appraisal committee’s contribution, the effect of which is new drugs will be undervalued. 

• In the prior process applied to both orphan and non-orphan drugs, the independent public appraisal committee could determine value according to contextual criteria with no quantified quality-adjusted life-year (QALY) range limit.  In ICER’s 2017-2019 framework, which would also apply to orphan drugs, the committee is no longer empowered to do this.³⁵ The inflexibility of this system will, by definition, label any ‘non-ultra-orphan’ intervention that falls $25,000 above the set threshold of $150,000 as ‘low’ value.  This is despite the fact that within 2016 alone all orphan drugs in ICER drug assessments were above $175,000, and six out of eight of the orphan drug regimens tied or received a majority ‘intermediate care’ value by the public panel (See Table 1 in the Appendix).³⁶ 

ICER’s panel composition does not allow patients or caregivers to vote on treatments that directly affect them.

• Less than 1 in 10 committee members that ICER ask to vote on the value of a drug is a patient, patient caregiver or patient advocacy group and it does not adequately take into account the voice of the patient, patient caregivers or patient associations.  It is important for patients to be able to express the grave and individual burden they experience from orphan disease. ^{37, 38}

ICER has assessed ‘ultra-orphan’ disease patients as common disease patients in past evaluations.  

• In the 2015 High Cholesterol ICER assessment, the cost-effectiveness model for homozygous familial hypercholesterolemia (HoFH), an ICER-defined ‘ultra-orphan’ condition, was not modeled separately from heterozygous familial hypercholesterolemia (HeFH) because the “expected number of patients is small [n=300-400 in the U.S.].”  During ICER’s Round Table, participants informed ICER that HoFH is “clearly [an] identified unmet need” and “the discussed criteria could be relaxed”.³⁹  Here, ICER grouped and assessed orphan disease patients together with a different (non-orphan) disease, solely by virtue of small numbers.
• Within ICER’s proposed approach to the assessment of orphan and ‘ultra-orphan’ disease, all 2016 assessed orphan drugs would be classified as ‘low value’.  Although ICER lacks the mandate to set thresholds and assessments of value, if ICER assessments were to inform access decisions, patients could lose access to new treatments, pay higher co-payments and be forced to try inappropriate treatments prior to new treatments.

3) ICER’s proposed willingness-to-pay thresholds are inappropriate to orphan diseases

ICER’s accountability in attempting to re-define orphan disease and set national thresholds for value is subject to substantial contention.

• ICER is assuming an inherent role that exceeds its level of accountability.  ICER’s attempts to re-define orphan disease, its assertion of nationwide immovable orphan disease and ‘ultra-orphan’ disease willingness-to-pay thresholds and stated limits across categories of national healthcare expenditure put it in an untenable position, incongruous with the needs of orphan disease patients. This is in stark comparison to those agencies accountable to the U.S. government such as the FDA.^{40, 41}  ICER’s best role is articulated by Peter Neumann and Joshua Cohen in: “ICER should simply calculate and disseminate cost-effectiveness ratios and let its audiences decide whether an intervention represents reasonable value”.⁴²

Subjecting orphan drugs to a value framework with fixed thresholds like ICER’s, could be devastating for patients with orphan diseases.

• Many health technology assessment (HTA) groups globally have recognized the ethical, equity and social justice challenges of applying a willingness-to-pay threshold to ODA-defined orphan diseases. Applying a cost-benefit ratio for an orphan disease drug is contrary to an egalitarian/utilitarian approach (maximizing equity for individuals).⁴³ This approach would prioritize the least costly patients rather than the sickest of patients who lack a sufficient voice in the healthcare system.⁴⁴
• ICER’s QALY threshold anchors to 1970 treatment decision standards, not the dynamic environment of 2017.^{45, 46} Moreover it suggests that the U.S. spend as little as 6 times less per QALY than in 1970.⁴⁷  Imposing a willingness-to-pay (WTP) threshold for orphan drugs will not address the needs of the complex U.S. Healthcare System and may jeopardize our societal desire for equity and justice in the insurance system.^{48, 49}

ICER’s proposed ‘ultra-orphan’ definition, if ever implemented, would make the U.S. one of the most restrictive places in the world for ODA-defined orphan disease sufferers.

• ICER’s proposed definition for ‘ultra-orphan’ disease of less than or equal to 10,000 patients equates to a prevalence of three per 100,000 or less.  This excludes any special provisions for orphan disease assessment outside of ICER’s ‘ultra-orphan’ definition.  Because if this, if implemented this would make the U.S. the second most stringent country in the world for ODA-defined orphan drugs.⁵⁰  Only England’s threshold (two per 100,000) is more restrictive than ICER’s proposed definition.  The impact of implementing this in England has significantly restricted reimbursement, specifically, less than half of approved orphan drugs are reimbursed,⁵¹ and orphan disease patients wait for drugs, on average, over two years.^{52, 53, 54}

The level of certainty around a single threshold is meaningless in orphan disease.

• ODA-defined Orphan diseases typically have very few patients compared to common diseases.  Further complicating this, new drugs are often approved at a much earlier stage than other drugs (Phase II).  This means that any attempt to set one static price threshold that applies to what is typically a heterogeneous population in orphan disease for which very little is known, would be meaningless at best and at worst could inadvertently harm patients.

ICER’s proposed use of willingness-to-pay thresholds based on the quality adjusted life year pose methodological and ethical challenges in orphan disease.

• ICER assessments use a disease outcome measure that can be used across different diseases, the quality adjusted life-year.  On technical grounds, QALYs suffer significant shortfalls if applied to orphan disease including (1) they cannot address the heterogeneity in drug options (2) they cannot be derived for very young or very old populations (3) Caregiver QALYs usually are not considered (4) Patients with lower QALYs whose lives are extended will have overall higher/unfavorable incremental cost per QALYs than patients with mild disease.
• ICER’s Updated Framework that also applies to all orphan drugs, sets a willingness-to-pay (WTP) threshold of $100,000-$150,000 per QALY.  This QALY threshold does not have a scientific foundation.  ICER reaches this threshold primarily through two sources which have been extensively criticized: (1) the arbitrary and misapplied WHO “benchmark” of 1-3 times GDP and (2) recent UK work undertaken to set the threshold on the true opportunity cost at the margin of health spending, extensively criticized for its empirical shortcomings.^{55, 56, 57, 58, 59, 60}
• The use of willingness-to-pay price thresholds is not correlated with improved health outcomes and require more research before implementation in the U.S.⁶¹

4) ICER’s proposed assessment approach for ‘ultra-orphan’ and orphan drugs excludes patient, caregiver and employer costs.

ICER’s limitation of costs to the health system undervalues the impact of orphan drugs, and will lead to decisions that shift costs to orphan disease patients, their caregivers, employers and society.

• ICER’s choice of costs to include in its proposed value framework goes against the body of academic research in health technology assessment.^{62, 63, 64, 65} ICER states that their goal of taking a population perspective is to “analyze evidence in a way that supports population-level decisions and policies”.⁶⁶ However, ICER’s empirical choice of perspective is that of the medical system not the wider population.  ICER’s inclusion of some costs (medical costs as incurred by insurers) and exclusion of other costs (all other costs including patient costs) marginalizes the financial burden of diseases that patients experience, which could be alleviated by new drugs.

This approach obscures the fact that these costs are not ‘saved’ but simply incurred by another stakeholder.  

• Specific patient groups, their caregivers and employers are likely to be penalized because of this choice of perspective.  For many orphan diseases the costs patients and society incur may approach or exceed the health insurer costs.  For example, these costs as a percent of medical costs are as high as 94% for cystic fibrosis; 47% for hemophilia, and 216% for Scleroderma patients respectively.^{67, 68} Out of pocket expenditure in Fragile X in the U.S. have been reported to be as high as $17,476 per patient.^{69, 70}

5) ICER will apply the “Final Value Assessment Framework for 2017-2019” to all orphan drugs that expand beyond the 10,000  patient definition of ‘ultra-orphan’ drugs.

•  Amgen has concerns on ICER’s Final Value Assessment Framework for 2017-2019 that will also apply to ODA-defined orphan conditions above 10,000 patients.
• ICER's legitimacy and accountability in setting national arbitrary thresholds in the Final Value Assessment Framework for 2017-2019 will continue to be subject to substantial contention.  ICER's potential future value lies in a role centered on guidance.  ICER can inform decisions on value based on key pillars of evidence and its strengths, robust analytics and the identification of areas of uncertainty; and do so with flexibility, inclusiveness, scientific integrity, transparency and patient centricity, in the absence of 'one size fits all' absolutes and thresholds.  This would ultimately allow each budget-holder and decision-maker to leverage ICER's insights in making their decisions on value.
• ICER's proposed willingness-to-pay QALY threshold lacks scientific merit and specificity to the complex US healthcare system.  ICER's attempt to anchor their lower QALY threshold at $100,000 is irrelevant to the current dynamic U.S. environment.  Of note, treatments that have an incremental cost-effectiveness above $175K per QALY will automatically be judged 'low value'.  This departs from flexible thresholds that leading health economists recommend AND past assessments whereby many drugs that have had incremental cost-effectiveness ratios above $175K were still judged by the panel as of 'high' or 'intermediate' value.
• ICER's 2017-2019 Value Framework does not address contextual considerations in patient treatment populations.  A challenge with the prior version of the Framework that has not been addressed in the current version is that important determinates of value were buried into contextual considerations with less visibility and inability to influence the quantitative analysis.
• In defining and voting on value, ICER's 2017-2019 Value Framework leaves out the patient perspective and that of other key stakeholders in ICER's public appraisal committees.  There is the opportunity to further democratize the public appraisal committees (not just the policy roundtable) so that they are composed of more patients, caregivers, patient advocates, clinical experts and manufacturers to help determine the value of new treatments.
• ICER's limitation of costs to the health system will undervalue treatments and lead to decisions that shift costs to patients, their caregivers, employers and society.  Payer-borne and monetary costs are only one aspect of healthcare burden.
• ICER's decision-making based on its proposed budget impact model continues to be detrimental to the health needs of patients.  Rather than setting national budget impact thresholds that lose relevance in the current multi-payer context, if instead it defined its role as one that is advisory, ICER could help articulate to decision-makers what types of elements should be included in their specific budget impact calculations.
• ICER models used for analysis continue to lack transparency, availability and replicability.  There is the opportunity to make ICER's and their academic partners' research methods, assumptions, data inputs, and equations available in a completely transparent manner, such that results are fully reproducible by third parties and are reviewed by known experts.

Conclusion

At 25-30 million, the number of patients suffering from orphan disease is a third of the total population with cardiovascular disease, about the same size as the population in the U.S. who are currently living with diabetes, and two times more than those living beyond a cancer diagnosis (many of which are orphan diseases themselves).^{71, 72, 73} Yet despite these large numbers, the drug development constellation is infinitely more complex due to lack of data and information in a field where most of what we have learned about orphan disease in the last five years completely eclipses everything we have learned over the last five centuries.  Although we agree with ICER’s mission to create a “more effective, efficient, and just health care system”,⁷⁴ a one size fits all, inflexible approach, such as the one ICER has proposed for orphan drugs, could negatively affect the one in every 10 Americans who are affected by an orphan disease.⁷⁵ These patients most need hope, dignity, and respect enhanced by access to valuable drugs for an equal chance at achieving a healthy life as everyone who does not suffer from an orphan disease. ICER often refers to European approaches in value assessment and there may be elements that ICER could adopt.  Namely, time, stakeholder engagement, extensive research and caution in value framework development.  It has taken over 2 years for the European Working Group for Value Assessment and Funding Processes in Rare Diseases to develop principals for assessment.⁷⁶  ICER should consider waiting until more research and insights are available to inform an appropriate methodology for the assessment of treatments for orphan disease conditions. In the meantime, ICER should refocus on aligning its role with its stated mission to enable more objective and robust dialogues that inform decisions on 'value' by putting the patient at the center of each assessment.

Appendix

Table 1: 2016 ICER Orphan Condition Assessments With Care Value Votes⁷⁷

References & Endnotes: