Challenges Associated with Estimating Minimal Clinically Important Differences in COPD—The NHLBI Perspective

Abstract

Chronic Obstructive Pulmonary Disease (COPD) is a common lung disease that exemplifies the value, as well as the difficulties and challenges, of using minimal clinically important differences (MCID) in clinical research. Development and validation of better endpoints for clinical studies is critical to research progress in COPD. However, the clinical, genetic, and pharmacological heterogeneity of the COPD patient population complicates attempts to define and validate MCIDs for COPD. It is difficult to identify a single measurable outcome that reflects the many components of the COPD patient's health state. Acute exacerbations of symptoms, which COPD patients often experience, present another challenge in the development of MCIDs for this disease. Consequently, the NHLBI does not require the use of MCIDs in clinical research. This allows research on the causes, prevention and diagnosis of COPD and use of endpoints for which an MCID is not yet known. It is important for the scientific community to reach agreement on what is a meaningful MCID in therapeutic trials for COPD. Further research into the concept of the MCID and its application should enable therapeutic trials in COPD to yield knowledge that is more effectively translated into improved public health.

• Lung diseases
• Chronic obstructive pulmonary disease
• Clinical research endpoints

Introduction

The determination of minimal clinically important differences (MCID) is important for evaluating the relative risk-benefit ratios of various disease interventions. MCID is defined as the smallest difference in a measurable clinical parameter that indicates a meaningful change in the condition of the patient for better or for worse, as perceived by the patient, clinician, or the investigator. Since standardized, accepted MCIDs will enable research on interventions that may improve public health, developing MCIDs for heart, lung, and blood diseases and sleep disorders is an appropriate aim for research funded by the National Heart, Lung, and Blood Institute (NHLBI). Furthermore, research that uses MCIDs increases the evidence-base for medical practice, aids health care management and resource allocation, and allows clearer communication of research findings to the public. This paper uses Chronic Obstructive Pulmonary Disease (COPD) to exemplify the value, as well as the difficulties and challenges, of using MCIDs in clinical research.

NHLBI Mission

The mission of the NHLBI is to provide leadership for a national program in diseases of the heart, blood vessels, lung, and blood, blood resources, and sleep disorders. Accordingly, the institute plans, conducts, fosters, and supports an integrated and coordinated program of basic research, clinical investigations and trials, observational studies, and demonstration and education projects that address the causes, prevention, diagnosis, and treatment of heart, blood vessel, lung, and blooddiseases and directs research on the development and evaluation of interventions and devices related to prevention, treatment, and rehabilitation of patients suffering from such diseases and disorders [[1]]. COPD is an important target of NHLBI programs and development and/or refinement of MCIDs useful for clinical studies of COPD falls clearly within the mission of the Institute.

Although MCIDs are important to its mission, NHLBI does not require all investigators to consider MCIDs in their clinical research. This flexible approach allows research on broad topics including the causes, prevention, and diagnosis of disease and endpoints for which an MCID is not yet known. This approach also recognizes the heterogeneity of responses to interventions and the potential for developing individualized treatments.

The Public Health Burden of COPD

The term COPD encompasses a range of clinical conditions associated with airflow limitation, including chronic bronchitis, emphysema, and bronchiolitis. Long-term exposure to airborne particulates—from cigarette smoke, outdoor and indoor air pollution, environmental irritants, or occupational dusts and chemicals—is believed to lead to COPD, but airway infections may also play a role. The most consistent and troubling characteristic of COPD is persistent and progressive airflow limitation, which is only poorly reversed by bronchodilators. COPD has, in addition, a multitude of other manifestations, including lung physiological dysfunction, deterioration of cognitive and psychosocial function, chronic cough with sputum production, malnutrition, dyspnea, weakness of ventilatory and peripheral muscles, sleep disorders, and various co-morbid conditions.

COPD causes over 100,000 deaths per year and accounts for nearly one-fourth of smoking-attributable mortality. COPD is responsible for over $15 billion in direct costs for health care per year in the United States [[2]]. Prevalence, morbidity, and mortality rates for COPD vary by gender, ethnicity, socioeconomic status, geography, lifestyles, and genetics, among many other factors. COPD mortality, which is highest in white males, has remained relatively constant since the early 1980s for this group. During the same period, COPD has gradually increased in black males, doubled in black and white females, and increased substantially in women of various racial and ethnic groups. In fact, more women now die from COPD than men.

The diagnosis and management of COPD are complicated by its demographic and clinical heterogeneity; multiple etiologies; variable patterns of occurrence, progression, severity, and response to medications [[3]]. No single pathogenetic theory seems capable of encompassing all known correlates of the disease [[4]] and there is, as yet, no cure.

These multiple systemic manifestations make the identification of reliable and valid indices of the overall physical, physiological, and psychological health status of the COPD patient difficult when evaluating treatment options.

The Challenges

Clinical, Genetic, and Pharmacological Heterogeneity of the COPD Patient Population

Because of the clinical heterogeneity of COPD patients and the fact that responses to treatments are neither uniform nor predictable, there is considerable interest in recognition of patient subgroups that are responsive to particular treatments and in the development of individualized therapy. Stratification of patients for therapies and in clinical studies may need to take into account genotype and biomarkers in addition to symptoms and functional measures. The striking clinical, genetic, and pharmacological heterogeneity of the COPD patient population has important implications with regard to definition and validation of MCIDs. This issue also can limit the applicability of MCIDs, validated in one study population, to other groups.

Slow Progression of COPD

The physiologic and pathologic abnormalities in COPD usually progress over many years. They typically begin with mucus hypersecretion and ciliary dysfunction. Chronic cough and sputum production are followed, often years later, by other symptoms and physiological abnormalities. In advanced COPD, peripheral airway obstruction, parenchymal destruction, and pulmonary vascular abnormalities compromise the lung's capacity for gas exchange, producing hypoxemia and, in some, hypercapnia. Cardiovascular problems (pulmonary hypertension and cor pulmonale), cognitive difficulties, depression, weakness, and weight loss may also develop in severe COPD [[4]].

Spirometric measures of expiratory flow limitation, a hallmark of COPD and highly reproducible, may not correlate well with other symptoms or show changes over the time course of most clinical studies [[4]]. Statistically detectable differences in pulmonary physiology do not necessarily indicate important symptomatic benefits [[5]]. On the other hand, treatments that induce physiological changes that are slow and small in magnitude may nonetheless be associated with important improvements in symptoms of patients with severe COPD. Relative changes, rather than absolute changes, may be more meaningful in patients with poor lung function and low exercise capacity. Furthermore, small changes in the rate of decline in lung function that would be undetectable in most clinical studies might have large effects on life expectancy. Hence, the slow time course of COPD development and progression justifies some caution regarding MCIDs that are determined in studies that last a few years or less.

Multiplicity of Intervention Outcome Measures

COPD involves a range of physiological, biochemical, physical, and psychological abnormalities. The challenge associated with measuring the impact of an intervention lies in limiting the battery of measures so as to not overburden thepatient or investigator while collecting enough data to assess the breadth of impacts associated with COPD. There are currently no established, clinically meaningful relationships between functional capacity, performance, and other measures of dysfunction. However, there are patient reported outcome measures, all with good psychometric properties, which assess: 1) functional impairment; 2) attributes such as depression; and 3) well-being. Questionnaires, commonly used instruments that elicit responses such as “a little bit better,” are often incapable of detecting small but meaningful differences in symptoms.

Such empirical estimates of MCID work best with highly motivated participants and necessarily focus on patients' perspectives and cannot be broadly extended to patients with psychiatric and neurologic complications [[5]]. Since the COPD patient's health status and quality of life are compromised at many levels, selected measures of clinical improvement capture only a portion of the disease, and not the sum total of the patient's actual or perceived state of health, even if they are statistically significant. It is difficult to identify a single measurable outcome that reflects the many components of the patient's health state impacted by COPD. As a NHLBI Workshop Report noted, “an important challenge in COPD is the development of more powerful, multi-variate methods for predicting individual outcome and individual responsiveness to particular therapies on the basis of clinical and laboratory characteristics” [[6]].

COPD Patients Experience Frequent Acute Exacerbations

Many COPD patients experience acute exacerbations of varying severity that may require hospitalization and emergency treatment [[3]]. In fact, nearly half of all exacerbations are not reported to physicians. Those requiring hospitalization are associated with 3–4% mortality [[7]]. Since exacerbations usually involve worsening of dyspnea and increased sputum volume and purulence, they introduce variability in measurements of disease severity presenting another challenge in the development of MCIDs for this disease.

Exacerbations are manifested heterogeneously across patients; there are several other conditions that can mimic the appearance of an exacerbation, including pneumonia, congestive heart failure, pneumothorax, pleural effusion, pulmonary embolism, and arrhythmia [[3]]. It is a challenge to measure the beneficial effects of interventions directed at amelioration of exacerbations, and assessment of the clinical importance of these beneficial effects is especially difficult.

Inadequacy of Current Management Approaches for COPD and the Need for Tailored Multicomponent Management Strategy

Because of the diverse pulmonary, systemic, physiological, neurocognitive, psychological, and nutritional impairments seen during stable and exacerbated COPD, therapeutic and other interventions for COPD need to accomplish multiple aims, including preventing progression, alleviating symptoms, improving exercise tolerance, and enhancing overall quality of life. Both pharmacologic and nonpharmacologic modalities may be indicated, including bronchodilator β-2 agonists, anticholinergics, methylxanthines, corticosteroids, mucolytics, oxygen therapy, nutritional interventions, and pulmonary rehabilitation, among others. Current treatments for moderate COPD are only palliative. They aim at stabilizing symptoms and enhancing quality of life, rather than eliminating the symptoms. There is no single treatment that addresses all symptoms. For example, bronchodilators, central to acute symptomatic management of COPD, do not completely reverse airflow limitation and do nothing to slow the progression of disease. Furthermore, relief in symptoms following bronchodilator administration only weakly correlates with changes in spirometric indices. Alternate markers are needed for assessing the benefits of therapeutic interventions. Therapies for advanced disease, including long-term oxygen therapy and lung volume reduction surgery, yield only modest extensions in life expectancy for selected groups of patients.

Conducting Clinical Trials in COPD in the Background of the Complexity of the Disease and Multiplicity of Outcomes

Clinical trials that test the efficacy of existing and new treatments are essential for the development of more effective therapies for COPD. The NHLBI continues to sponsor a number of clinical trials. Recently, the Institute established a COPD Clinical Research Network to carry out multiple therapeutic trials.

Although many NHLBI clinical trials have been conducted without formal consideration of established MCIDs, the Institute is optimistic that current and future studies will lead to the definition and refinement of MCIDs, as well as assessment of their applicability and limitations. For example, NHLBI's National Emphysema Treatment Trial (NETT) collected many different outcome measures simultaneously, some of which already have established MCIDs. These data provide an opportunity to assess intervention effects across many outcome measures and may lead to testable hypotheses on MCID. NETT gathered very relevant data regarding the 6-minute walk test that is regularly used for evaluation of functional status in patients with lung disease in both clinical practice and clinical trials. “Learning effects” in severely ill patients performing the 6-minute walk test in NETT were smaller than those found in other studies. As a result, the NETT investigators recommended that the 6-minute walk test be administered before and after an intervention with control groups or repeated testing; the tests should be spaced several weeks apart to minimize the “learning effect” [[8]].

NHLBI Perspective on the Use of MCIDS for COPD

The accumulated experience from attempts to develop MCIDs for COPD underscores both the value and limitations of using MCIDs in clinical research. As noted by a panel charged with developing a consensus on clinically important changes in health-related quality of life (HRQoL) for patients with COPD, “establishing clinical changes for HRQoL measures requires both clinical insight into etiology, symptoms, and progression of the diseases, as well as patient insight related to living with chronic disease” [[9]]. With regard to the wide use of decline in post-bronchodilator FEV₁ over time as the gold standard for measuring disease progression in pre-morbid COPD, an NHLBI Workshop Report concluded that “emerging evidence indicates that alternative measures, such as inspiratory capacity, may better reflect the ventilatory dysfunction in COPD, and that reliance on FEV₁ may cause studies to miss beneficial effects of therapies such as increased exercise capacity, quality of life, or cognitive function or lessened dyspnea, cough, sputum production, depression, or frequency or severity of exacerbations” [[6]].

Conclusion

In summary, COPD is a complex disease that requires multiple outcome measurements. It may not be feasible to determine a single MCID for each measure that would be accurate for all subject populations. Based on the current understanding of the disparate modes of initiation and disease progression, heterogeneity of the nature and degree of clinical expression, variation in responses of individual patients to interventions, and consequent need for multiple, tailored approaches to symptom mitigation or alleviation, one might question whether a composite measure may be most appropriate. Progress in this direction has been made recently with description of the BODE index [[10]]. Determination of the MCID for composite measures, such as the BODE Index, may be an important goal for future research. It is important for the scientific community to reach agreement on what is a meaningful MCID in therapeutic trials for COPD. A retrospective evaluation of data from clinical trials in COPD may be of value to that discussion. Results are already available from several NHLBI supported trials, including the Nocturnal Oxygen Treatment Trial (NOTT), the Lung Health Studies (LHS), and NETT. Future studies by the COPD Clinical Research Network will likely provide additional data. There can be little doubt that further research into the concept of the MCID and its application will enable therapeutic trials in COPD to yield knowledge that is more effectively translated into improved public health.