Abstract
Objective. Our study evaluated the impaired health status of clinical trial patients with systemic lupus erythematosus (SLE) and explored the relationship between changes in fatigue and pain and their effect on overall health status.
Methods. Pooled treatment and placebo data from a phase Ib clinical trial of adults with moderate/severe SLE were analyzed. Measures included patient-reported Medical Outcome Study Short Form-36 Survey, Version 2 (SF-36v2), Fatigue Severity Scale, and numeric rating scales (NRS) for pain and global health assessment and clinician-reported global assessment of disease activity (MDGA). Disease burden was compared to the US general population. Health status of responders and nonresponders on pain or fatigue were compared.
Results. The sample included 161 patients with SLE, predominantly female (96%) and white (72%), with average age of 43 ± 11 years. Mean SF-36v2 component summary scores reflected overall problems with physical [physical component summary (PCS); 35.2 ± 9.7] and mental health (mental component summary; 40.9 ± 12.9). Patients with SLE had worse health status on all SF-36v2 subscales than the US general population and comparable age and sex norms (effect size −0.51 to −2.15). Pain and fatigue responders had greater improvements on SF-36v2 scores (bodily pain, physical functioning, social functioning, PCS), patient global health assessment NRS, and MDGA than nonresponders. There was moderate agreement in responder status, based on global assessments by patients and clinicians (68.1%), with some discrepancy between patients who were MDGA responders but patient assessment nonresponders (27.7%).
Conclusion. Improvements in patient-reported pain or fatigue correlated with improvements in overall health. Patient assessments offer a unique perspective on treatment outcomes. Patient-reported outcomes add value in understanding clinical trial treatment benefits.
Systemic lupus erythematosus (SLE) is a chronic, autoimmune disease that can affect almost any organ system, with heterogeneous clinical manifestations that can vary over time. Commonly used measures of SLE disease activity assess manifestations in constitutional, mucocutaneous, musculoskeletal, neurological, pulmonary, cardiovascular, abdominal, and renal systems1. The disease affects primarily women in their reproductive years, and has changed in recent years from life-threatening with high rates of early mortality to a chronic disease with longer life expectancy2,3.
Studies have demonstrated a multidimensional effect of SLE on health-related quality of life (HRQOL)4 and posited that the overall effect of SLE may actually be as much or more severe than other more prevalent and better-recognized chronic diseases (e.g., hypertension, diabetes, myocardial infarction)5. In comparing health status of patients with SLE to those with rheumatoid arthritis and other noninflammatory conditions, patients with SLE had better physical function but were more impaired in general health (GH), vitality (VT), social function, role-emotional, and mental health (MH) domains6. Also, HRQOL was predicted by organ damage, comorbidity, income, education, and age6,7,8. Similarly, a review highlighted the importance of HRQOL as an outcome measure in SLE, concluding that age, fatigue, and other psychosocial factors influence HRQOL in a complex way9.
Fatigue, the most common constitutional symptom associated with SLE, affects over 50% of patients with SLE10,11,12. Although common, fatigue is often not associated with SLE disease activity levels or with chronic organ damage. A study investigating relationships between psychosocial factors and fatigue in SLE found that pain and perceived social support, but not disease activity, predicted fatigue levels13. Medications, lifestyle habits, concomitant fibromyalgia, or affective disorders have also been associated with fatigue14. Although chronic fatigue may not be related to SLE disease activity or organ damage, presence of fatigue influences HRQOL, even in the absence of comorbid fibromyalgia15. A conceptual model of HRQOL for SLE based on qualitative research highlighted the relevance of pain, fatigue/tiredness, and skin problems, the relationships between symptoms and burden, and the broad influence of SLE on patient HRQOL from the patient perspective16. Research has suggested that SLE patients with fatigue may be physically compromised11, and that fatigue may be the most debilitating symptom in patients with SLE17,18.
The multisystemic nature of SLE is often a challenge for evaluating treatment benefit because a unidimensional approach may not be appropriate. SLE treatments have been evaluated based on their effects on disease activity, organ damage, and quality of life19. Outcome measures in clinical trials have included disease activity indices, HRQOL measures, global assessments (patient- and physician-assessed), and measures of organ damage and responder indices20. Evaluation of clinical trial treatment benefit has traditionally focused on clinical endpoints, because patient-reported outcome (PRO) measures are often not correlated with disease activity, but rather reflect different aspects of the disease21. The benefit of using PRO measures is that they can provide valuable information on the patient experience and perspective that is different from and supplements traditional clinical outcomes22,23. Patient report of fatigue has been accepted as an important concept to evaluate benefits of SLE treatments22.
Exploring PRO data from SLE trials can help us better understand disease burden and effect on HRQOL and functioning. This information can identify unmet medical needs and help develop strategies for PRO data collection that ensure better assessment of disease activity and detection of change, from the patients’ perspective. The objectives of our study were to evaluate HRQOL of patients with SLE in a clinical trial by comparing their health status to the US population, and exploring relationships between changes in pain and fatigue with changes in overall health status over time.
MATERIALS AND METHODS
Clinical trial study design
Data came from a phase Ib, randomized, double-blind, placebo-controlled, dose-escalation study (n = 161) evaluating safety and tolerability of multiple intravenous doses of sifalimumab (formerly MEDI-545), a fully human anti-interferon (IFN)-α monoclonal antibody, in adult patients with SLE (MI-CP152; clinical trial no. NCT00482989)24. The clinical trial started in March 2008 and was completed in July 2010. Thirty rheumatology centers in 5 countries (Argentina, Brazil, Canada, Chile, and United States) participated.
Study design and subject disposition are described elsewhere24. Subjects were enrolled if they were aged ≥ 18 years with moderate to severe SLE and met at least 4 of the 11 revised American College of Rheumatology classification criteria for SLE25. Eligible subjects also had a positive antinuclear antibody test (at least 1:80 serum dilution in the past or at screening), and met at least 1 of the 2 criteria for disease activity: Safety of Estrogens in Lupus Erythematosus (SELENA)–Systematic Lupus Erythematosus Disease Activity Index (SLEDAI)26 score ≥ 6, or at least 1 system with a score of A, or 2 systems with a score of B on the British Isles Lupus Assessment Group (BILAG) index27 at screening. Patients with infections or recent high (> 20 mg/day) or fluctuating doses of oral corticosteroids, antimalarials, or immunosuppressives; B-cell depleting therapies within the past 12 months, leflunomide in the past 6 months or any other biologic agent in the past 30 days were excluded.
The study was conducted in accordance with the International Conference on Harmonization Guidance for Good Clinical Practice and Declaration of Helsinki. Recruitment procedures were approved by an institutional review board/independent ethics committee prior to study initiation and were performed in accordance with US Health Insurance Portability and Accountability Act requirements and all applicable state and federal laws and regulations. Written informed consent was obtained from subjects before study entry or any study-specific activities.
SF-36v2
The Medical Outcome Study Short Form-36 Health Survey, Version 2 (SF-36v2) is a generic health status measure shown to be reliable and valid in diverse patient groups and healthy populations28,29,30,31. It is one of the most widely used HRQOL instruments among many different disease populations28,29,32,33. The standard SF-36v2 with a 4-week recall period was used.
SF-36v2 contains 8 subscale and 2 summary scores. The subscales are Bodily Pain (BP), GH, MH, Physical Functioning (PF), Role-Emotional (RE), Role-Physical (RP), Social Functioning (SF), and Vitality (VT). Subscale scores range from 0–100, with higher scores reflecting better health status. Mental Component Summary (MCS) and Physical Component Summary (PCS) scores are generated using a weighted combination of the subscale scores. Scores were generated using norm-based methods that standardize scores to have a mean = 50 and SD = 10 in the normative general US population, where scores > 50 are above average and < 50 are below average31. Responder criteria values proposed for SF-36 subscales range from 3.2 (RP) to 5.7 (GH), and 3.1 (PCS) and 3.8 (MCS) for component scores34. Disease-specific estimates of meaningful change in SF-36 scores have also been developed; improvement of 5–10 points for individual SF-36 domains and 2.5–5 points for PCS and MCS summary scores have been considered indicative of important improvement20,35,36. In addition, there is also evidence that minimally important differences vary for improvement versus deterioration, with lower thresholds for worsening, as evidenced by decreases ranging from 1.7 to −14.7 for domains and −0.8 to −2.1 for summary scores seen among SLE clinical trial patients35.
Fatigue Severity Scale
The Fatigue Severity Scale (FSS) assesses severity of fatigue and interference with daily activities, physical function, exercise, work, family, or social life in the past week37,38. FSS includes 9 items rated on a 1 (strongly disagree) to 7 (strongly agree) scale. The total score is calculated as the mean of the items and ranges from 1–7, where higher scores indicate greater fatigue. In this study, if > 2 item responses were missing, FSS was not calculated and the score was set to missing.
Clinical fatigue has been defined as an FSS total score ≥ 439. Important improvement in fatigue has been described as a 10% decrease (95% CI 4.9–14.6) based on the patient perspective40 and a 15% decrease on FSS based on clinical experts41. These decrease estimates using patient40 and clinician41 input correspond closely to a 1-point change in total FSS score, which was used to assess potentially relevant improvement in fatigue.
Patient numeric rating scales for pain and global health assessment
Numeric rating scales (NRS) assessed pain and global health. Each included a single question on a 0–100 scale in 5-point increments. The pain item asked “How much pain have you had because of your illness in the past week?” with response anchors 0 (no pain) and 100 (pain as bad as it can be). The patient global health assessment asked “Considering all of the ways that your illness affects you, rate how you are doing;” with response anchors 0 (very well) and 100 (very poorly).
A 2-point change in 0–10 point pain NRS scores has been considered clinically meaningful42,43. Multiple anchor-based analyses have also confirmed that patients considered pain intensity reductions on NRS or visual analog scales (VAS) of 1 point or 10–20% as minimally important change, and reduction of at least 2 points or 30% to be moderately clinically meaningful44,45. In addition, change thresholds of up to 30% have been used in SLE as a criterion for meaningful change in patient global assessment46. A 2-point change on a 0–10 scale42,43 equates to a 20% change on the 0–100 NRS for pain and global health adopted in this study, where higher scores represent more pain or worse health, respectively. For each measure, a ≥ 20% increase in NRS score was considered worsening.
Physician global assessment of disease
Clinicians rated SLE disease activity using the physician global assessment of disease (MDGA), a VAS ranging from 0–3.0 measured to 1 decimal place. MDGA score could be interpreted as 0 = none, 1 = mild, 2 = moderate, and 3 = severe. No imputation was performed for missing data. For MDGA, an increase from baseline ≥ 0.3 was considered worsening20,47.
Analytical sample
This was an exploratory analysis evaluating health status and PRO measures in patients with SLE within the context of a controlled clinical trial. The modified intent-to-treat (mITT) population comprised 161 randomized subjects who received at least 1 dose of study medication; 122 of them (75.8%) completed the trial. The mITT population was pooled across treatment groups and analyses were conducted blinded to treatment; the objective of this study was to describe overall burden of SLE and explore relationships between fatigue and other outcomes, regardless of treatment assignment or efficacy. Treatment effects are described elsewhere24.
Baseline and end of study (EOS) clinical trial data were used in this analysis. Baseline for PRO was defined as the last valid assessment prior to first administration of study drug; this timepoint for outcome measures evaluated in the current study coincided with Study Day 0. EOS was identified as Study Day 196.
SAS statistical software version 9.2 (SAS Institute Inc.) was used to conduct analyses. All statistical tests were 2-sided with a significance level of 0.05. No adjustments for multiple comparisons were performed.
Exploring the HRQOL burden of SLE
To characterize the overall burden of SLE, mean baseline SF-36v2 subscale scores for the clinical trial sample were compared to US general population means using groups stratified by age and sex31. In the absence of individual patient level data for general population norms that would allow us to precisely estimate the norms by matching with the age and sex distribution of the trial sample, we used an indirect approach to evaluate burden. We compared mean trial sample scores by age and sex to the appropriate general population age and sex cohorts for whom normative scores have been reported. One-sample t tests evaluated the hypothesis that mean baseline study sample scores are equal to published reference scores. Effect sizes (ES) were computed as the difference between study sample mean and the reference mean, divided by the pooled SD for both samples (i.e., Cohen’s d). ES can be interpreted as small (0.20), moderate (0.50), or large (0.80)48.
The clinical trial sample was predominantly female with an average age across treatment groups of 40–45 years; SF-36v2 age and sex norms most similar to the study sample were chosen. SF-36v2 baseline scores in the study sample were compared to 1998 US general population norms for (a) females, ages 35–44 years (n = 820); (b) males and females, ages 35–44 years (n = 1520); and (c) males and females, ages 18–96 years (n = 6742)31. The full study sample was used for these comparisons; analyses were not stratified by age or sex of SLE patients because of small sample size (n = 161) and few males (n = 7).
Relationships between changes in pain and fatigue and changes in health status and global assessments
Patients were classified as responders or nonresponders based on change in pain and fatigue scores. Responder status was defined using change from baseline to EOS, computed as change score = endpoint score–baseline score; a negative change score represents improvement. Responder groups were expected to demonstrate more change over time on outcome measures. FSS change score ≤ −1.0 from baseline to EOS was defined as an FSS responder and change > −1.0 as an FSS nonresponder. Similarly, a ≥ 20% increase in NRS pain score from baseline to endpoint was considered worsening and assigned as a non-responder. Patients with < 20% increase in NRS pain score were considered pain responders, with improvement or no substantial increase in pain level.
Changes from baseline to EOS based on SF-36v2 and patient and physician global assessments were compared between pain and fatigue responder and nonresponder groups. ES were computed for responders and nonresponders as the PRO change score, divided by SD of baseline PRO score.
Relationship between responders based on changes in physician and patient global assessments
Concordance between responders defined based on patient assessments of global health and physician assessments of disease activity was examined to explore the relationship between physician and patient-defined assessments. Frequencies (n, %) of responders and nonresponders based on patient-reported global health assessment NRS and clinician-reported MDGA were compared for study completers with data for both measures at baseline and Day 196.
Responder status was defined as score change from baseline to EOS in patient global health assessment NRS and clinician-reported MDGA, where negative change scores represented improvement. Patient assessment responders were defined as < 20% increase in patient global health NRS score, and physician assessment responders as < 0.3-point increase in clinician-assessed MDGA VAS. Fisher’s exact test was conducted to analyze the 2 × 2 contingency table.
RESULTS
Sample characteristics
Demographic and clinical characteristics for the 161 SLE trial patients are presented in ⇓Table 1. Patients were predominantly female (95.7%), white (72.0%), and not Hispanic or Latino (64.0%), with an average age of 43 ± 11 years (range: 18–71 yrs). The majority (71.4%) came from North America (US: n = 114, 70.8%; Canada: n = 1, 0.6%), with fewer patients from South America (Argentina: n = 28, 17.4%; Brazil: n = 13, 8.1%; Chile: n = 5, 3.1%). Baseline disease activity indicated moderate/severe SLE despite standard of care treatment.
Sociodemographic characteristics of the SLE clinical trial participants.
Disease characteristics of the SLE clinical trial participants.
Burden of SLE and health status
On average at baseline, patients reported compromised health status across all SF-36v2 subscales and summary scores (Table 2). Subscale means ranged from 34.5 ± 9.6 for GH to 42.1 ± 13.2 for MH. Summary scores also indicated problems with overall physical (PCS; mean ± SD = 35.2 ± 9.7) and mental (MCS; mean ± SD = 40.9 ± 12.9) components of health, relative to means of 50 in the normative US general population31.
Comparison of Medical Outcome Study Short Form-36 Survey, Version 2 (SF-36v2) subscale scores for SLE clinical trial participants and US general population norms.
Comparison to US general population
Mean SF-36v2 subscale scores for the trial sample and the 3 US general population norms are presented in Figure 1. Compared with the US general population and age and sex norms, SLE trial patients had significantly worse health status across multiple subscales, with differences of moderate to large ES magnitudes (p < 0.0001 for all subscales, ES = −0.51 to −2.15).
Comparison to US population norms of mean baseline Medical Outcome Study Short Form 36-item Survey, Version 2 (SF-36v2) subscale scores in a systemic lupus erythematosus (SLE) clinical trial.
Comparison of changes in pain and fatigue to changes in global assessments
Baseline, endpoint, and change scores for PRO (SF-36v2; patient global health assessment NRS) and clinician-reported MDGA are reported by pain responder status (Table 3). Pain responders, identified based on pain NRS scores, reported large improvements on SF-36v2 BP (mean change = 8.8, ES = 0.8). Generally, pain responders experienced larger amounts of change, representing improvement, in SF-36v2 subscales over time than pain nonresponders. Pain responders reported large improvements on SF (mean change = 8.7, ES = 0.7) and PF (mean change = 7.0, ES = 0.6) subscales, and PCS (mean change = 7.4, ES = 0.7).
Change in health status by pain responder statusa in the systemic lupus erythematosus clinical trial participants.
Changes in patient global health assessment (mean change = −26.0, ES = −1.0) and clinician-reported MDGA (mean change = −0.7, ES = −1.5), suggested that pain response was associated with dramatic improvement in overall health status at EOS, as perceived by both patient and physician.
Fatigue responders
Baseline, endpoint, and change scores for PRO and clinician-reported MDGA are presented by fatigue responder status (Table 4).
Change in health status by fatigue responder statusa in the SLE clinical trial participants.
Fatigue responders and nonresponders reported improvement in health status on the SF-36v2; however, larger changes were observed among responders, particularly in SF-36v2 BP (mean change = 6.1, ES = 0.6), PF (mean change = 6.8, ES = 0.6), and SF (mean change = 7.6, ES = 0.6) subscales, and PCS (mean change = 6.0, ES = 0.7). Patients experienced less improvement in domains related to mental and emotional aspects (e.g., MH and RE subscales and MCS).
Fatigue responders also experienced better overall health status than nonresponders. ES estimates suggested that fatigue responders experienced more improvement, particularly in patient global health (ES = −0.7) and clinician-reported MDGA (ES = −1.6).
Relationship between responders based on changes in patient and physician global assessments
Responder status was defined based on change scores on patient-assessed health status and physician-assessed disease activity. There was moderate agreement in responder status, with 68.1% of patients (n = 81) identified as responders based on patient global health assessment NRS and physician-assessed MDGA. Nearly one-third of patients (27.7%; n = 33) were identified as responders using clinician-assessed MDGA, but nonresponders based on the patient global health, indicating discrepancy between physician and patient evaluations. Five subjects (4.2%) were identified as nonresponders based on both patient and physician assessments. No patients were classified as responders by patient global health assessment, but as nonresponders based on clinician ratings. Fisher’s test for the 2 × 2 contingency table was significant (p = 0.0027), indicating a relationship between responders defined based on patient self-evaluation and physician assessment.
DISCUSSION
Our study investigated overall HRQOL burden of SLE and explored relationships between changes in certain PRO measures and overall health status using PRO in a clinical trial setting. Overall health status of study patients as measured by SF-36v2 was considerably worse at baseline compared to US general population and comparable age and sex norms. This highlights the multiple effects of SLE on health and functioning that contribute to disease burden. These findings are similar to previous cross-sectional studies4,5,6, despite the more restrictive inclusion criteria for patients with moderately to severely active SLE despite standard of care used in this study. Baseline SF-36 decrements of up to 32 points have been reported for patients with SLE from 5 randomized controlled trials (RCT) with age-matched and sex-matched US norms7. The smaller differences (in the range of 10–15 points) observed in this SLE study sample relative to norms for females ages 35–44 years suggest that patients in our study may have had slightly better health than the RCT patients. However, possible differences in baseline patient characteristics and protocol-defined inclusion criteria used in the studies, as well as the relatively high proportion of patients in several of the RCT with a history of renal disease, may have contributed to the magnitude of decrements compared to norms. Nevertheless, the pattern of low SF-36 scores for patients with SLE across multiple domains relative to norms identified in both studies supports the broad damaging effect that SLE has on patient health status.
The relationships between changes in pain and fatigue, 2 common SLE symptoms, and changes in overall health demonstrated the potential effect of symptom improvement from the patient perspective. Comparisons of change scores between pain and fatigue responders and nonresponders showed that responders consistently had greater improvements than nonresponders. Pain responders and fatigue responders had larger improvements than nonresponders in health status on SF-36v2, particularly for BP, physical and social functioning domains, and overall physical health. Although smaller improvements were observed for the VT domain, fatigue as measured by FSS and VT based on SF-36 may measure different aspects of energy level. Also, pain response was associated with larger improvements in bodily pain domain than VT, but decreases in pain may relate to increases in energy level. The relationship between pain and fatigue was also evident; half (n = 25, 51%) of the 49 pain responders were also fatigue responders. However, this small sample size of “dual responders” did not permit further analysis of this subpopulation, and future research in larger samples will be needed to better understand the relationship between these variables. In addition, the effects of comorbidities such as fibromyalgia and depression or particular tender points or “trigger spots,” on changes in pain and fatigue were not evaluated in this analysis; these data were not systematically collected as part of the clinical trial study. Overall health and disease status based on patient and physician global health assessments also improved more for pain and fatigue responders than nonresponders, suggesting the potential benefits of improving pain and fatigue on other outcomes.
Although there was moderate agreement (68.1%) in patients who were identified as responders on both the patient and physician global assessments, 27.7% of subjects (n = 33) were identified as responders based on clinician-assessed MDGA but were nonresponders based on patient global health assessment. Identifying these discrepancies was an important finding and clearly indicates that these assessments are not redundant and not interchangeable, but rather provide different perspectives — that of a patient and a physician. Clinical improvements in disease activity, as perceived by clinicians, may not necessarily correspond to improvements in health status valued by patients. In turn, changes that affect patients in a meaningful way and improve their quality of life may not be based on the same indicators of improvement that physicians evaluate in clinical assessments.
Another study concluded from the literature that in the course of the disease, HRQOL for patients with SLE may not change dramatically over time, while disease activity can change more rapidly15. Moreover, health status and less- or non-observable symptoms such as pain and fatigue are better predicted by factors other than disease activity13,14. The heterogeneity of SLE manifestations has led to use of composite endpoints in evaluating disease activity, as witnessed in phase III belimumab49,50 and phase II epratuzumab51,52 studies. Interestingly, in the epratuzumab trial, where rapid improvement in disease activity was seen, close correlations were observed between physician and patient global ratings53.
These findings on the burden of SLE, how changes in pain and fatigue affect overall health, and the unique perspectives offered by patient and clinician assessments, improve our understanding of the difficulties that patients experience daily. It provides insight into potential areas of unmet medical need that treatments must target. Physical function, fatigue, and pain were identified as primary needs in SLE that may require attention. Patients may also have other unmet needs; social functioning and mental health were identified as concepts affecting HRQOL and functioning from the patient perspective. Clinical trials have consistently included HRQOL measures20. Although a few trials have demonstrated differences between active and control groups and dosage levels36,54,55,56, changes in HRQOL may not often or always correlate with changes in disease activity. It is also relevant to focus on specific constructs, including pain, fatigue, and physical function, in developing strategies for PRO data collection in the context of evaluating medical treatments. While the SF-36v2, the health status tool used most frequently in clinical trials, includes a domain for “vitality” and aspects of physical function, it does not specifically measure physical components of fatigue.
Further explorations into relationships between disease and functioning can aid in identifying relevant endpoints. Information on patient concerns about sleep, body image, work ability, intimacy, and medication side effects need to be evaluated while also assessing treatment benefits. Understanding these relationships can also be useful in designing clinical trials, such as by developing endpoint models relevant to patients that inform the hierarchy of testing used for detecting changes and treatment benefit. This information can also spur development of treatments with the potential to improve clinical condition and HRQOL in patients with SLE. As suggested in recent guidance for developing SLE treatments22,23, collecting PRO data early in exploratory clinical studies provides useful information and helps to gain experience with these outcomes.
Limitations to the study with a bearing on interpretation should be considered. The study had a relatively small sample size; results should be interpreted with caution. There was variable enrollment by country and site; these geographical differences may introduce possible bias. Comparisons of health status were made to available US norms for SF-36; these norms may not be universally representative of health status for each country included in the trial.
Further, although we compared trial SF-36 means with age-specific and sex-specific norms at the mean level, no adjustments to the normative scores were made to reflect the age and sex distribution of our trial sample, because of a lack of individual patient level data for the general population normative sample. While this limits the robustness of our estimates, our reported method of indirect comparisons has previously been used in the absence of individual patient level data.
In addition, the focus of analysis in the current study was change in terms of improvement shown by “responders.” “Nonresponders” could have experienced no change or worsening; it must be acknowledged that clinically important improvements and deterioration may not be symmetrical and equivalent and “nonresponse” may not be the same as deterioration. Because of the nature of the inclusion and exclusion criteria used for entry into the phase Ib study, generalizability of these findings may be limited to clinical settings with similar patient populations.
Finally, heterogeneous baseline clinical and disease characteristics of this population can make between-group comparisons challenging. However, these differences are not uncommon in SLE clinical studies, where patients tend to have highly variable disease characteristics (e.g., number/type of organs affected, severity of organ damage, disease activity, frequency/severity of SLE flares).
SLE is associated with significant HRQOL burden on patients. The disease affects multiple domains of physical and psychological functioning. Pain and fatigue affect the daily lives of patients, and are important symptoms to consider in evaluating patient experience with the disease and treatment. While improvements in pain or fatigue appeared to improve overall patient health, the individual and dual roles of pain and fatigue as confounding or effect-modifying variables in evaluating efficacy requires further exploration. Patient global health and physician assessment of disease activity offer unique perspectives on treatment outcomes.
PRO measures add value in understanding treatment benefit in clinical trials from the patient point of view and provide an important perspective that is generally not observed by clinical assessments of disease activity. Early exploratory clinical studies should include validated PRO tools to collect data on PRO; early findings from experience with PRO measures can help to inform endpoint models for future trials and accurately evaluate benefits of treatments for SLE that affect patients in a meaningful way.
Acknowledgment
The authors thank the following individuals for their contributions to the study: Wen-Hung Chen for analytical design and interpretation of results, Ren Yu for data analysis and statistical support, Shannon Kummer for project support (UBC), Jichao Sun, Rong Ye, and Charles/Chenxiong Le for design and execution of the clinical trial and 2e manuscript (MedImmune).
Footnotes
-
↵* Dr. Ethgen died May 13, 2012.
-
Sponsored by MedImmune LLC, including support for research and manuscript preparation. MP is affiliated with Johns Hopkins University School of Medicine and has consultancy agreements with MedImmune, UCB, Pfizer, Human Genome Sciences, GlaxoSmithKline, TEVA, and Anthera Pharmaceuticals. AKK and AH are employees of United BioSource Corporation. AF and WG are employees of MedImmune LLC. DE was an employee of MedImmune LLC at the time this study was conducted. KG was an employee of MedImmune LLC at the time of study conduct and writing of the manuscript.
- Accepted for publication July 16, 2013.
