Primary biliary cirrhosis (PBC) is a chronic, potentially life threatening liver disease of autoimmune etiology, and is associated with symptoms of fatigue, dry eyes and mouth, pruritus, and abdominal pain. Of PBC patients, 50% to 85% report symptoms of fatigue, although these symptoms are not associated with the severity of liver dysfunction tests or histological stage of the disease.1-4 Given the high prevalence of fatigue in PBC, some researchers have suggested that fatigue be recognized as a component of the disease.5 However, other investigators question the association between PBC and fatigue.4 To resolve this controversy, knowledge of the pathogenic mechanisms responsible for fatigue in PBC would be helpful. There are several hypothesized mechanisms including psychological elements,6 changes in the immune system involving possible effects of cytokines on the nervous system,6, 7 cytokine-mediated hyperactivity,6-8 or altered central serotoninergic neurotransmission.9, 10 Other studies have associated circulating cytokines with patient reported health status, pain, and other symptoms.11, 12
Abbreviations
PBC, primary biliary cirrhosis; HRQL, health-related quality of life; FSS, Fatigue Severity Scale; FIS, Fatigue Impact Scale.
Primary biliary cirrhosis (PBC) is a chronic, potentially life threatening liver disease of autoimmune etiology, and is associated with symptoms of fatigue, dry eyes and mouth, pruritus, and abdominal pain. Of PBC patients, 50% to 85% report symptoms of fatigue, although these symptoms are not associated with the severity of liver dysfunction tests or histological stage of the disease.1-4 Given the high prevalence of fatigue in PBC, some researchers have suggested that fatigue be recognized as a component of the disease.5 However, other investigators question the association between PBC and fatigue.4 To resolve this controversy, knowledge of the pathogenic mechanisms responsible for fatigue in PBC would be helpful. There are several hypothesized mechanisms including psychological elements,6 changes in the immune system involving possible effects of cytokines on the nervous system,6, 7 cytokine-mediated hyperactivity,6-8 or altered central serotoninergic neurotransmission.9, 10 Other studies have associated circulating cytokines with patient reported health status, pain, and other symptoms.11, 12
The randomized crossover clinical trial by Theal et al. published in the June 2005 issue of HEPATOLOGY evaluated the effectiveness of ondansetron, a serotonin receptor antagonist, for the treatment of fatigue in patients with PBC.13 Unfortunately, although the results noted an overall effectiveness of ondansetron in reducing symptoms of fatigue, the findings were inconsistent across the two periods in this crossover study. There was concern whether subjects may have been unblinded to active treatment during the second period due to the high rate of constipation in the ondansetron treated group. In addition, symptoms of fatigue did not return to baseline levels during the one week washout before the start of the second treatment period. The authors suggest that the subjective nature of patients' reports of fatigue may have compromised the assessment of active treatment. The conflicting results observed in this study could be attributed to the selection of a crossover design rather than a randomized parallel group clinical trial, an insufficient washout period, or the patient-reported outcomes.
Patient-reported outcomes, including fatigue, health-related quality of life (HRQL), or other domains, have been increasingly incorporated into studies of chronic liver disease.14, 15 Studies have demonstrated treatment differences in fatigue and HRQL outcomes in chronic hepatitis C, demonstrated improvements associated with sustained viral response, and predicted discontinuation from treatment.14, 16-18 However, few studies in PBC have included measures of fatigue or HRQL.15 Measurement of health-related outcomes from the patient's perspective is important for understanding the impact of disease and its treatment on patient functioning and well-being.19 Patient-reported outcome measures are increasingly incorporated into clinical studies and provide insight into the comprehensive effects of new treatments. The assessment of fatigue requires patient self-reporting and these ratings are considered subjective because fatigue is a subjectively experienced condition. However, just because a report is “subjective” does not mean that it cannot be measured reliably and validly. There is substantial evidence that patient assessment of their own HRQL and symptom severity and frequency can be measured with consistency and validity.20 A a number of fatigue symptom scales have been used in studies of PBC and other chronic liver diseases, including the Fatigue Severity Scale (FSS),16 Fatigue Impact Scale (FIS)21, 22 Fatigue Assessment Instrument,23 Multidimensional Fatigue Inventory,24 and SF-36 vitality scale.25 Other fatigue related measures, such as the FACIT-Fatigue26 and Brief Fatigue Inventory,27 have not been used in PBC or chronic liver disease studies. Several chronic liver disease-specific quality of life instruments also include measures of fatigue, such as the Chronic Liver Disease Questionnaire28 and the Liver Disease Quality of Life Instrument.29 The FSS and FIS were used as outcome assessments in the Theal study,13 and both of these instruments have acceptable reliability and validity in the PBC population.
All of these fatigue symptom, disease-specific quality of life and generic health status measures rely on patient responses and ratings. This is necessary since there is evidence that clinician and patient ratings of symptoms differ; and that often the patient's report is most important for evaluating therapeutic treatment effects.30 In general, the most carefully designed instruments, which were developed based on sound instrument design and psychometrics, have evidence supporting reliability (i.e., internal consistency reliability, test-retest reliability), construct validity and, for several instruments, responsiveness to clinical changes.14, 15 For example, the FSS and SF-36 vitality scales are responsive to changes in sustained viral response17 and have good psychometric qualities in patients with chronic hepatitis C.16 The fatigue symptom and other patient reported outcome measures applied in studies of chronic liver disease have sufficient evidence supporting psychometric characteristics and are likely to be sensitive to changes in fatigue symptoms associated with an effective intervention.
Several methodological challenges associated with patient reported outcome research need to be considered when planning and implementing clinical trials involving fatigue symptoms (and other patient-reported outcomes). First, it is necessary to select an instrument that covers the relevant range of the continuum of the targeted construct, in this case fatigue. For example, Kleinman et al.16 found that the FSS items covered the lower (i.e., most impaired) part of the fatigue continuum, while the SF-36 vitality scale covered the middle of the continuum. A mismatch between the instrument and symptom severity experienced by the patient population might result in a failure to detect treatment effects because of floor and ceiling effects. Second, the psychometric qualities (i.e., reliability, validity) of the selected fatigue measure needs to be acceptable, especially responsiveness, in the patient population.19 An instrument with poor psychometric properties will be less likely to be able to detect treatment effects, if the treatment is actually effective. In addition, practical issues, such as respondent burden and patient acceptance, must be examined.15
Finally, it is important that patients and clinicians be blinded to treatment. Maintaining the patient blind is important for all patient-reported outcome studies, but is most critical for relatively brief studies and interventions (i.e., 2 to 6 weeks), as in the Theal et al. study.13 Although the Theal study was blinded, there is some indication that the blind was less effectively maintained in the second phase of the cross-over design. While blinding is important in all studies, it may result in fewer patient-related biases in reporting symptom and HRQL outcomes for long-term studies, as it is difficult for subjects to maintain reports on improving symptoms over the long term if in fact these changes are not actually being experienced.
In summary, symptoms of fatigue can be assessed reliably and validly in patients with PBC and other chronic liver diseases. The literature of clinical trials in chronic hepatitis C clearly demonstrated the ability of several fatigue measures to differentiate the effects of different treatments and to demonstrate improvements associated with sustained viral response.14 The failure of the Theal et al. study to demonstrate consistent results of ondansetron in treating fatigue in PBC patients may have been attributable to the effects of the randomized, crossover design, unblind of the active treatment, insufficient washout period between treatment periods, or the lack of efficacy of ondansetron. It is unlikely that the results were due to the poor measurement of symptoms of fatigue. Patient-reported outcomes provide useful and meaningful insights into the effectiveness of treatments targeted at those medical conditions, such as fatigue, where no laboratory or physiological measures are available.
