Obesity Drug Outcome Measures: Results of a Multi-Stakeholder Critical Dialogue
- First Online:
- 718 Downloads
There has been great interest in what role drugs should play in the treatment of obesity. Given the complex and multifactorial nature of obesity, drugs represent a valuable adjunct to traditional obesity treatments. However, obesity drugs present specific regulatory challenges, due to the sheer number of individuals affected, the heterogeneity of how obesity affects individuals differently, and the potential for drugs to be misused, especially by those seeking or profiting from cosmetic weight loss. There is a need to refine the characterization of the individuals most at risk from obesity to better assess the balance of benefits versus risks and determine appropriate candidates for drug therapy. As obesity affects feeling, functioning, and survival, outcome measures that reflect improvement in these domains should be explicitly incorporated into regulatory guidance. Regulators could benefit from mechanisms that limit or closely monitor off-label use and thereby allow approval of drugs for narrowly indicated populations.
KeywordsObesity Weight Weight-loss Drugs Pharmaceuticals Food and drug administration FDA Regulatory Regulation Outcome measures Feeling Function Survival
The Zulu word indaba essentially means to gather under a tree and discuss whatever needs discussing. During a 10-month period in 2011–2012, the authors led an indaba among key members of the obesity, regulatory, research, advocacy, and pharmaceutical communities to better understand what the role of medications for the treatment of obesity should be – and how to get there.
This opportunity took shape in the wake of four negative regulatory decisions regarding obesity medications in 2010–2011. In a span of less than 6 months, the U.S. Food and Drug Administration (FDA) requested sibutramine (Meridia) be taken off the market and declined approval for phentermine/topiramate CR (now named Qsymia), lorcaserin (Belviq), and bupropion/naltrexone CR (Contrave) citing need for further study. These decisions were particularly controversial for several reasons:
First, evidence continues to mount that existing obesity treatments are insufficient. While lifestyle intervention, such as various forms of diet and exercise, is the cornerstone of treatment, most individuals do not sufficiently respond to or sustain lifestyle changes. Bariatric surgery has been shown to cause durable weight loss and improved comorbidities in a majority of patients, but it is inaccessible, undesired, or clinically inappropriate for the vast majority of individuals with obesity. No other medical condition affects so many lives yet has such limited clinical treatment options. In 2011, there was only one medication approved for long-term treatment of obesity – orlistat, which is not widely used – and there had been a gap of nearly 13 years since the last medication for obesity had been approved.1
Second, several of the medications submitted to the FDA were already approved in various forms and in wide use. Phentermine has been FDA approved since 1959 and is the most commonly prescribed medication in the US for short-term obesity treatment; topiramate was approved in 1996 and is used for treatment of neurologic conditions, such as epilepsy and migraines; bupropion is approved for treatment of depression and smoking cessation; and naltrexone is approved for treatment of alcohol abuse and opioid addiction.
Third, these medications sufficiently achieved the primary endpoint benchmarks as defined by FDA in its Guidance to Industry for weight loss medications.
Finally, in the case of Contrave, the FDA denied approval even though its Endocrine and Metabolic Drugs Advisory Committee (EMDAC) voted to approve the drug by a vote of 13–7. It is rare for FDA to disregard a recommendation of approval from its advisory committees .
With these and other issues eliciting extreme frustration among affected patients, physicians, and industry, we sought to gather key stakeholders for a series of roundtable discussions to explore the rationale, appropriate use, and outcome measures for obesity drugs. Our primary objectives were to begin a dialogue with relevant stakeholders to understand why – despite epidemic growth of obesity and repeated failure of conservative measures, no obesity drug had been approved since 1999, identify barriers to regulatory approval, and come to consensus about where drug therapy fits in obesity management.
The meetings were particularly notable for the participation of a breadth of stakeholders with wide-ranging backgrounds, training, beliefs, and knowledge about obesity. Project participants included high-ranking officials from FDA, NIH, and CDC, along with an invited group composed of clinicians and researchers who specialize in obesity, public health experts, consumer advocates, and representatives of the pharmaceutical industry. During the course of four 8-hour meetings and numerous emails, conference calls, and informal discussions bridging the live meetings, we discussed the intricacies and uncertainties of obesity medications and the FDA approval process. It was clear that there are unique challenges regarding approval of obesity drugs and several limitations, including implicit assumptions and prejudices about obesity, that need to be overcome in order to facilitate an objective and meaningful path toward availability of increased treatment options.
By the end of this nearly year-long indaba, we reached consensus on numerous central points, and we conjecture that this process played a valuable role in FDA’s approvals of Qsymia and Belviq, both approved within weeks of our final meeting. The summary and webinar describing the process are available on the George Washington University School of Public Health and Health Services website . We describe below several central findings and recommendations from the process.
Understanding Obesity as a Complex Biological Condition
Despite the number of affected individuals and the age-old history of this disorder, obesity is poorly understood as a clinical and physiologic entity – even by most physicians – and is far more complex than generally appreciated. Though obesity and weight regulation are typically relegated to a simple and static balance of eating and exercising, we discussed the complex physiological, psychological and social aspects of this unique condition. Obesity is a distinct phenotype that is multifactorial, developing from an interplay of biological, genetic, environmental, behavioral, social, economic, and policy factors. Obesity is a heterogenous disorder, affecting different individuals in different ways. The obesity state is dynamic and affects nearly every organ of the body; for example, it is only recently appreciated that adipose tissue acts as an active endocrine organ, producing numerous hormones that communicate with the brain and other organs, and contribute to the development or worsening of many illnesses . Further, while theoretically reversible, numerous physiologic responses to weight loss promote weight regain [3•]. Chronic calorie overload may also alter key neurologic areas that regulate energy balance . When weight is lost, the body may compensate by altering the central nervous system and peripheral body systems to lower metabolism and increase hormones that stimulate appetite. Combined, these mechanisms counter sustained weight loss and promote weight regain. This research begins to explain why lasting weight loss can be elusive and why obesity defies behavioral treatment in many affected individuals. In addition to the physiological underpinnings, obesity has many other contributing factors, making treatment of those with obesity even more challenging.
These and other recent advances in our collective understanding of obesity biology continue to counteract the age-old adage to “just eat less, exercise more” and support the need for continued research and additional safe and effective treatment options. While no treatment options are “silver bullets,” pharmacotherapy is a valuable adjunct to behavior modification to achieve weight loss, facilitate or reinforce healthier lifestyle habits, and potentially mitigate biological responses that promote weight regain.
Balancing Benefit and Risk
Obesity poses a unique clinical and public health scenario. In excess of 100 million individuals in the U.S. alone are affected, with a substantial number actively seeking treatment. Obesity is associated with immense morbidity, mortality, and healthcare costs, including association with more than 60 chronic diseases and health conditions, physical symptoms (e.g., joint pain, urinary incontinence), functional limitations (e.g., impaired mobility), and psychosocial problems (e.g., body image disorders, bullying) [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]. In addition, existing treatment options are lacking, either due to marginal effectiveness (lifestyle intervention) or marginal access (bariatric surgery). These facts suggest that expanded and effective treatment options could lead to substantial public health benefits and healthcare cost savings.
Still, due to the sheer number affected, even small relative risks can be amplified many-fold when treatments are delivered to large populations, creating a difficult benefit-risk balance. This scenario is not qualitatively different from other chronic diseases; however, the magnitude of those affected and nuances and uniqueness of obesity, which includes varying misperceptions, stigmas, and prejudices, have made a rational approach to obesity treatments difficult to navigate for regulators, industry, practitioners, and patients.
In assessing and balancing benefit versus risk, several considerations were centrally discussed, including how to effectively characterize the population to improve risk stratification, how to nurture access and innovation while limiting medically inappropriate use, how to streamline approval of medications when appropriate, and how to uniquely gauge patient benefits and benefit-risk tradeoff willingness.
Characterizing the Population
Obesity is not a defining characteristic. The phenotype can be achieved by numerous pathways and can exhibit varying health complications. The population of individuals with obesity is not homogeneous; rather, obesity varies in severity, onset, and expression of symptoms and comorbid conditions. Current characterization of obesity, however, is crudely based on size, as determined by the body mass index (BMI) scale, regardless of the wide interindividual differences of persons with similar BMI. For example, some individuals affected by obesity have no immediate health consequences or impairments, whereas others of the same BMI may have significant health or quality of life limitations. As such, the current characterization of individuals by BMI does not clearly delineate the population that could most benefit from access to treatment, making BMI a limited clinical tool.
Obese and “Well”: Individuals who carry excess weight but do not have comorbidities or risk factors for comorbid conditions and do not experience impairments in daily feeling or functioning
Obese with Risk Factors: Individuals who carry excess weight who do not yet have overt health consequences, but have measurable risk factors for comorbid conditions and/or impairments in daily feeling or functioning
Obese and “Sick”: Individuals who carry excess weight and exhibit one or more obesity-attributable comorbidities or impairments in daily feeling or functioning
By characterizing individuals by level of health risk, rather than simply by size, regulators could better assess the relative benefits and risks of a given treatment across the obesity spectrum. One recently published example of this categorization framework is the Edmonton Obesity Staging System (EOSS), which takes into account a number of variables beyond BMI, including the presence of physical symptoms, risk factors, chronic disease diagnoses, and end-stage organ disease . While several categorization frameworks are possible, EOSS has already been shown to improve the identification of health and mortality risks associated with excess weight, and it should be strongly considered for use to improve risk stratification beyond BMI, thereby improving the ability to balance benefits and risks [19, 20].
Limiting or Mitigating the Risk of Medically Inappropriate Use of Obesity Drugs
While the FDA can approve drugs for specific populations deemed in medical need, the agency is mandated with protecting the public’s health and has few mechanisms to restrict medically inappropriate use. In the case of drugs that affect weight, the FDA is legitimately concerned with potential serious side effects and/or medically inappropriate use by both non-indicated and indicated populations. Non-indicated use, including some cases of off-label prescribing, is particularly challenging in the case of obesity medications, as many individuals without clinical obesity would likely seek access for cosmesis, even though risk of treatment likely outweighs any cosmetic benefit. Even in those for whom obesity medications are theoretically indicated – i.e., BMI > 30 kg/m2 or BMI > 27 kg/m2 with comorbidities NHLBI – risks of drug therapy may outweigh benefit. The situation may be further complicated by direct-to-consumer drug advertising, which may misguide patient expectations and contribute to pressuring of physician prescribing, and a widespread lack of physician training in obesity medicine.
The FDA should explore ways, either through existing or novel risk management mechanisms, to limit inappropriate prescribing and medically inappropriate use, including close monitoring of off-label prescribing, while fostering availability of drug treatments that benefit indicated populations. The Risk Evaluation and Mitigation Strategies (REMS) program is the FDA’s current mechanism for limiting risks associated with the use of certain drugs. REMS was instituted in 2007 to afford timely access to drugs that would otherwise not likely have been approvable . REMS have been used to prevent medically inappropriate use through a provision known as Elements to Assure Safe Use (ETASU), which can be used to restrict distribution and/or require provider training . However, REMS have several limitations and were not intended to address off-label prescribing.
One possible alternative to REMS is the creation of a class of drugs “not to be prescribed off-label.” This idea is not without merit, as France recently instituted a similar regulatory mechanism and within the United States, advocacy groups such as the Infectious Disease Society of America (IDSA) have called for the FDA to adopt a similar mechanism to address antimicrobial resistance [23, 24]. The efficacy of such a mechanism is not guaranteed to successfully limit inappropriate or off-label use, and may have unintended consequences, but could create an alternative pathway for drugs with very limited indications, where misuse or overuse use is a prohibiting factor.
In addition to managing patient risks through mechanisms such as those described above, it may also be helpful for the FDA to employ mechanisms to help patients manage expectations of obesity treatment. One possible tool for influencing prescriber and patient attitudes toward appropriate medical use of drugs lies in the label. Patients should understand that currently approved and pending obesity medications typically result in an average of 5–10 percent of body weight loss over several months, followed by a plateau (or even partial regain) of weight despite continued use. Such weight loss is not likely to produce dramatic changes in physical appearance but may lead to significant improvements in health and function in appropriately chosen patients. Obesity medication labels should clearly specify that these are drugs with potentially serious adverse effects intended for the clinical treatment of chronic obesity, not for “miracle” weight loss, and may require long-term use.
Patient-Centeredness in the Assessment of Benefits and Risks
To better account for the varying patient risk profiles across the obesity spectrum, the FDA, as well as clinicians, could use a more patient-centered approach in evaluating the benefits and risks of obesity drugs. Patient centeredness describes actively including patients in the process of medical decision-making, including regulatory processes, by respecting and including patients’ wants, needs, and preferences. The level of risk a patient may be willing to accept will likely vary with many factors, such as how severely their obesity impacts their overall health, daily feeling and functioning, and quality of life. By applying objectively measured patient preferences and values to the drug approval process, FDA may find that certain drugs warrant approval for some populations but not others. For example, FDA may determine that for those who are obese and “sick,” the benefits of taking a particular drug outweigh its risks, whereas this may not be true for those who are obese and “well." By further refining this framework, the FDA could potentially make drugs with possible serious adverse side effects available only to those with significant or unmet medical need, while limiting or restricting use among other patients, without having to deny approval across the entire population. To do so, the FDA would not only need to consider different risk mitigation strategies to ensure medically appropriate access, but also improved ways of defining and quantifying benefits.
One such area of importance to include in the benefit-risk calculus is improvements in feeling and function associated with obesity treatment. Because obesity can manifest in a wide variety of ways, treatment may result in many health benefits, reductions in symptoms, and improvements in quality of life, in addition to weight loss. The FDA regularly approves drugs, including those with serious adverse effects, that demonstrate improvement in one or more domains of feeling and functioning when it is determined that the benefits outweigh the risks. However, current FDA Guidance for obesity medications is based on primary outcomes of weight loss and changes in cardiometabolic parameters. Less attention is given to alleviation of other health consequences associated with obesity, which may be more salient to patients. Alleviation or mitigation of symptoms of obesity (for example, chronic pain, incontinence, sleep disturbance) and functional limitations (for example, decreased mobility) have not been sufficiently emphasized as critical areas affected by obesity, which, in and of themselves, might warrant approval.
Further, there is a lack of tools for measuring the impact of obesity on quality of life and other patient reported outcomes (PROs). While the Impact of Weight on Quality of Life (IWQoL) scale is a widely used scientific measure for changes in quality of life, as currently constructed it has not been approved by the FDA for regulatory purposes. Other known measures of quality of life that assess more global influences do not yet sufficiently capture important nuances of obesity’s impact on patients’ lives. A tool that can effectively capture patient perspectives on feeling and functioning is essential.
Consequences of obesity and measures of improvement impacted by weight loss
1. Obesity-associated mortality
a. Obesity-associated mortality
2. Obesity-associated health conditions
a. Diabetes: Measures include diabetes rates (fasting blood sugar, 2-hour post-prandial blood sugar, hemoglobin A1c) or diabetes mortality rates
b. Cardiovascular disease: Measures include cardiac events, myocardial infarction, stroke, and cardiovascular mortality
c. Cancer: Measures include obesity-related cancer incidence, recurrence, and mortality rates
d. Kidney disease: Measures include chronic kidney disease incidence and kidney dysfunction
e. Fertility: Measures include fertility rates in patients with PCOS
“Intermediate outcomes”/risk reduction
3. Measures of intermediate endpoints for obesity associated health problems
a. Diabetes: Measures include changes in diabetes risk biomarkers (such as fasting blood sugar or fasting insulin)
b. Cardiovascular disease: Measures include blood pressure, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, pulse, cardiorespiratory fitness, C-reactive protein, pro-inflammatory cytokines, and pro-thrombotic factors
c. Non-alcoholic fatty liver disease: Measures include changes in liver enzymes, MRI fibrosis scores, or liver biopsy histology in patients with non-alcoholic fatty liver disease
d. Polycystic ovarian syndrome: Measures include changes in ovulatory function and/or insulin sensitivity
e. Chronic kidney disease: Measures include changes in kidney function
f. Cardiorespiratory fitness: Measures include changes in VO2 Max and time to fatigue
4. Measures of obesity-associated symptoms/QoL
a. Medication use: Measures include changes in medication use for diabetes, cardiovascular disease, hypertension, PCOS, osteoarthritis, acid reflux, etc.
b. Sleep apnea: Measures include changes in symptoms of sleep apnea or clinical measures of sleep apnea (such as the apnea-hypopnea index) or reduction in continuous positive airway pressure(CPAP) use
c. Urinary stress incontinence: Measures include changes in symptoms of urinary stress incontinence
d. Polycystic ovarian syndrome: Measures include changes in ovulatory function, insulin sensitivity, and/or fertility in patients with PCOS
e. Osteoarthritis: Measures include changes in symptoms of joint pain in patients with osteoarthritis
f. Acid reflux: Measures include changes in symptoms
g. Hypogonadism: Measures include changes in symptoms or testosterone levels
5. Measures of improvement in physical function
a. Mobility: Measures include changes in symptoms or improvement in 6-minute walk test
b. Functional limitation: Measures include changes in symptoms
6. Measures of improvement in psychosocial and behavioral functioning
a. Overall and categorical quality of life: Measures include changes in symptoms
b. Symptoms/level of depression: Measures include changes in symptoms, improvement in objective measures of depression
c. General self-esteem: Measures include changes in symptoms (e.g., Rosenberg Self-Esteem Questionnaire)
d. Body image (especially evaluative component measures): (e.g., Rosenberg Self-Esteem Questionnaire) changes in symptoms (e.g., Body esteem scale)
While the appropriateness of any individual pharmaceutical agent for any given patient is a highly individualized decision that should be decided between patients and their clinician, most clinicians specializing in the treatment of obesity agree that pharmaceuticals are an important tool in the treatment of obesity. The current FDA framework does not adequately categorize which types of patients with obesity could achieve benefits in feeling, function, and health risk. Nor does it adequately capture the many potential benefits of weight loss (short-term symptomatic, longer-term comorbidities, or effects on quality of life) that may be improved through modest weight loss, aided by pharmacologic treatment. A more comprehensive patient-centered approach in making risk-benefit determinations could help the FDA ensure that safe and effective obesity drugs are available to patient groups for whom the benefits of treatment outweigh the risks associated with a particular drug.
As of the start of this process in mid-2011. In June 2012, Belviq (lorcaserin) was approved by FDA and in July 2012 Qsymia (phentermine/topiramate CR) was approved by FDA for long-term treatment of obesity.
Conflict of Interest
Scott Kahan is on the Scientific Advisory Boards for Vivus and Eisai. He is the Director, Strategies to Overcome and Prevent (STOP) Obesity Alliance; STOP Obesity Alliance has received funding from several companies, including Sanofi, Allergan. His research at George Washington University has been funded by several companies, including Sanofi, Vivus, Orexigen, Takeda, Obesity Action Coalition, American Heart Association, Novo Nordisk, and Eisai.
Christine Ferguson is on the faculty of George Washington University, which has received funding from Sanofi, Allergan USA, Vivus, Orexigen, Takeda, the Obesity Action Coalition, American Heart Association, Novo Nordisk, and Eisai.
Stephanie David declares that she has no conflict of interest.
Lucas Divine declares that he has no conflict of interest.