Accepted Article Article type : Original Manuscript Interpreting the Effectiveness of Opioids and Pregabalin for Pain Severity, Pain Interference, and Fatigue in Fibromyalgia Patients 1 Fred Davis, Mark Gostine,2 Bradley A. Roberts,1 Rebecca Risko,1 Joseph C. Cappelleri,3 Alesia Sadosky4 1 ProCare Systems, Inc; 2Michigan Pain Consultants; 3Pfizer Inc. Groton, CT; 4Pfizer Inc., New York, NY Target journal: Pain Practice Corresponding author: Rebecca Risko, RN, BSN ProCare Pain Solutions 61 Commerce Ave SW Grand Rapids, MI 49503 Tel: 616-940-3504 Email: firstname.lastname@example.org Keywords: Fibromyalgia, Pain, Fatigue , Opioids, Patient-reported outcomes widespread pain , Sensitization , Outcome Assessment , chronic This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/papr.12651 This article is protected by copyright. All rights reserved. Disclosure Accepted Article This study was sponsored by Pfizer. Joseph C. Cappelleri, and Alesia Sadosky are employees of Pfizer, the sponsor of this study. Fred Davis, Bradley Roberts, and Rebecca Risko are employees of ProCare Systems, Inc., which received funding from Pfizer for conducting the analyses. Mark Gostine collaborated with Pfizer on this project; he was not financially compensated for his participation or the development of this manuscript. Editorial support was provided by E. Jay Bienen and was funded by Pfizer. Abstract Objective: To evaluate the effectiveness of opioids and/or pregabalin on patient-reported outcomes among fibromyalgia (FM) patients based on levels of improvement. Methods: A total of 1,421 FM patients were identified with 3,082 observational periods of opioids with or without pregabalin use between April 2008 and February 2015. Patients were categorized by opioids, and pregabalin with and without opioids; opioids were designated by morphine equivalent dose (MED) of ≤20 (low MED), >20 to <100 (moderate MED), ≥100 (high MED), and pregabalin dosages ≤150mg, 151- 300mg, and 301-450mg. Proportions of patients meeting clinically relevant thresholds of ≥30% and ≥50% improvement for pain interference (Ability to enjoy life; Activity; Mood; Relationships; Sleep), pain severity, and fatigue were compared among treatments, and area-under-the-curve (AUC) for improvement and worsening of effects was determined, enabling ranking of treatments. Further analysis compared pregabalin dosages. Results: Pregabalin without opioids resulted in the highest proportions of patients with ≥30% improvement on all pain items and pain interference with “Ability to enjoy life,” “Activity” “Mood,” and “Sleep.” For the ≥50% threshold, pregabalin alone was highest for all pain interference items and for “Average pain” and “Worst pain.” Pregabalin was consistently This article is protected by copyright. All rights reserved. lowest across thresholds for fatigue, but showed better results combined with moderate MED Accepted Article opioids. Pregabalin dosages recommended for treatment of FM (151-450mg) generally resulted in the highest proportion of patients achieving thresholds relative to opioids. The AUC results were consistent with thresholds; pregabalin without opioids resulted in the greatest benefits with regard to improvement, with the highest ranking for overall improvement and overall effects. Conclusion: Pregabalin without opioids provided the most favorable outcomes overall based on ≥30% and ≥50% improvement thresholds and AUC, with support for moderate MED opioids+pregabalin in patients suffering from fatigue. While most patients took less than recommended pregabalin doses, higher doses may lead to improved outcomes. INTRODUCTION Fibromyalgia (FM) is a chronic disorder that is thought to result from a dysfunction of central pain processing pathways,1, 2 which results in the hallmark symptoms of widespread musculoskeletal pain, joint stiffness, and fatigue. In addition to these symptoms, FM is characterized by sleep disturbances, depression, and cognitive difficulties.3, 4 As FM severity increases from mild to incapacitating, the disease inflicts a greater burden on patients as shown by its associations with reductions in daily function and quality-of-life, lost productivity, and higher healthcare resource use.5-7 Recommendations for the management of FM include use of pharmacologic agents from a variety of therapeutic classes including antidepressants and anticonvulsants, but specifically recommend against the use of opioids except for weak opioids such as tramadol.8-11 In the United States, the Food and Drug Administration has approved three drugs for the treatment of FM, two antidepressants (duloxetine and milnacipran), and an anticonvulsant (pregabalin). This article is protected by copyright. All rights reserved. However, treatment of FM remains challenging, and studies have shown low or ineffective Accepted Article use of approved or recommended therapies, with opioids consistently reported as the most frequently prescribed pharmacologic agent despite little evidence for their efficacy.2, 12-15 Indeed, studies have suggested that opioid use may be associated with higher healthcare resource utilization and costs among patients with FM.15, 16 Nevertheless, an older survey among patients with FM found that these patients perceived opioids to be among the most effective therapies.17 Thus, since there remains a need for evaluating the effectiveness of opioids, the objective of this study is to characterize and compare the effectiveness of opioids with pregabalin both in the absence and presence of concomitant opioids use across a range of patient-reported outcomes relevant to pain, daily function, and fatigue using two complementary methods that can enrich the interpretation of effects. METHODS Data source and population The data source for this retrospective analysis was clinic-level data from the ProCare Systems network of chronic pain clinics in Michigan between April 2008 and February 2015. All patient data were de-identified and fully compliant with the Health Insurance Portability and Accountability Act (HIPAA). Patients with FM were identified using traditional selection criteria and predictive modeling as previously described.18 All patients were required to have Class 1 FM, i.e., regional FM with classic symptoms, according to the four identified classes that reflect symptom severity and comorbidities, pain regions, and treatment intensity.19 Patients were also required to have been prescribed opioids and/or pregabalin and have completed baseline and follow-up PRISM™ Pain Health Assessment (PHA). The PHA is a self-report measure that contains This article is protected by copyright. All rights reserved. core outcome domains for evaluating efficacy and effectiveness of treatments on disease Accepted Article presence, pain characteristics, physical function, and psychosocial function. Patients were assigned to groups stratified by their average morphine equivalent dose (MED: low, MED ≤20; moderate, MED >20 to <100; and high, MED ≥100), and subgroups to investigate outcomes at varying levels of pregabalin dosage including ≤ 150 mg, 151- 300 mg, and 301-450 mg; the recommended dosages of pregabalin for the treatment of FM are 300-450 mg.20 Calculations of MED were performed using common guidelines.21 Observational time (i.e., days between baseline and follow-up PHAs) was limited to 56-365 days. Outcomes Outcomes consisted of questions from the PHA, including five items relating to pain interference (Ability to enjoy life, Activity in general, Mood, Relationships with others, and Sleep), three for measuring pain severity (Average pain in the last four weeks, Pain right now while completing your PHA, and Worst pain in the last four weeks), and two for fatigue (Do you feel tired? Do you feel worn out?). The questions on pain severity and interference in the PHA are closely aligned with the Brief Pain Inventory, a validated measure of pain severity and interference using a 0-10 numerical rating scale (NRS; 0 = no pain or interference, 10 = worst possible pain or complete interference).22 Questions that address fatigue are similar to those that comprise the “Vitality” scale of the validated 36-item Short Form Health Survey SF-36.23 This article is protected by copyright. All rights reserved. These outcomes are consistent with general recommendations of the Initiative on Methods, Accepted Article Measurement, and Pain Assessment in Clinical Trials (IMMPACT)24 as well as more specific recommendations for FM by the Outcome Measures in Rheumatology (OMERACT) group.25, 26 Analysis A response was considered any improvement from baseline (change from baseline ≥ 0) on the assessed outcome. The proportions of responders who reported improvement of at least 30% and 50% from baseline were compared between treatment arms. These thresholds are consistent with IMMPACT recommendations for clinically relevant improvements in pain, with the 30% and 50% thresholds reflecting moderate and substantial improvements, respectively.27 Statistical significance for all comparisons were based on 95% confidence intervals calculated (bias corrected and accelerated method) using bootstrapping with 10,000 iterations. Additionally, efficacy was evaluated using an area-under-the-curve (AUC) adaptation of a cumulative proportion of responder’s analysis (CPRA). The CPRA is a method that represents the cumulative proportion of patients who reach a specific response, calculated across the range of possible responses for improvement from baseline.28 As has been previously proposed,29, 30 the application of AUC to a responder analysis enriches the interpretation of improvement in selected outcomes and encompass all relevant responder information. Graphically, the change in the outcome of interest is depicted on the x-axis with the proportion of patients on the y-axis. All possible cutoff points of improvement can be incrementally represented, allowing for a visual representation of change scores. Extending this concept, multiple groups can be included within the same graph and subsequently This article is protected by copyright. All rights reserved. compared for patient responses to treatment. For each outcome, the AUC was estimated Accepted Article based on percent improvement on the outcome. In the current analysis, the AUC for each item was derived using Simpson’s rule.31 Differences in the derived AUCs were compared between treatments, with statistical significance based on calculated 95% confidence intervals (95% CI) according to a bias corrected and accelerated method using bootstrapping with 10,000 iterations. RESULTS Population characteristics A total of 1,421 unique FM patients taking opioids and/or pregabalin were identified, with 3,082 pre-/post-observational periods. Of these observations, 2,396 (77.7%) were female with a mean (SD) age of 50.5 (12.7) years (Table 1). The time under assessment was approximately 180 days and was similar among the treatments. Because there were some gaps in opioid and pregabalin prescriptions during patients’ assessment periods, the mean percent time on opioids and pregabalin represent the fraction of time in days that patient prescriptions could be accounted for over the entire assessment period, and ranged from 50% to 99%. The number of opioid prescriptions tended to be higher across increasing levels of MED (Table 1). The mean MED for all patients taking opioids was 60.6 111.4, and the mean pregabalin dosage was 159.7 140.8 mg. On average, pregabalin dosage was higher in patients taking a combination opioids + pregabalin relative to pregabalin without opioids and increased across the MED levels (Table 1). This article is protected by copyright. All rights reserved. When pregabalin dosages were stratified across all patients by strata of ≤150 mg, 151-300 Accepted Article mg, 301-450 mg, and 451-600 mg (Table 1), the largest proportions of patients were taking ≤150 mg, ranging from approximately 52% (high MED opioids, pregabalin) to 77% (pregabalin without opioids). Few patients (n = 14; 3.4%) were in the 300-450 mg FDA recommended dosage group. Generally, the proportions of patients taking opioids + pregabalin were higher for all dosage groups than patients taking pregabalin alone. Analysis by response thresholds Pregabalin without opioid use resulted in the highest proportions of patients achieving ≥ 30% improvement across all pain items (Table 2), and was statistically significant relative to low MED opioids + pregabalin for “Pain right now while completing your PHA” (13.2%, 95% CI 0.8%, 27.2%; P < 0.05). While pregabalin without opioids had a low ranking for both of the fatigue items, there were no significant differences for any of the treatment comparisons. On pain interference (Table 2), pregabalin without opioids also resulted in the highest proportions of patients who achieved ≥ 30% improvement except for “Relationships with others,” which was highest with moderate MED opioids + pregabalin. Pairwise comparisons showed that differences between treatments were significant with pregabalin for “Activity in general” relative to low MED opioids (15.0%, 95% CI 0.9%, 30.5%), moderate MED opioids (15.4%, 95% CI 1.1%, 31.1%), and high MED opioids (16.7%, 95% CI 2.3%, 32.4%) (all P < 0.05). The difference was also significant for high MED opioids + pregabalin relative to low MED opioids + pregabalin (16.2%, 95% CI 2.5%, 30.9%; P < 0.04). For “Relationships with others,” pairwise comparisons showed that the differences were significant, with higher proportions reporting ≥ 30% improvement (Table 2), with moderate MED opioids + pregabalin relative to high MED opioids (9.4%, 95% CI 0.7%, 18.7%), low MED opioids This article is protected by copyright. All rights reserved. (9.7%, 95% CI 1.5%, 18.2%), and low MED opioids + pregabalin (13.1%, 95% CI 1.9%, Accepted Article 23.9%) (all P < 0.05). Pregabalin in the absence of opioids resulted in the highest proportions of patients with ≥ 50% improvement in “Average pain” and “Worst pain,” although none of the differences with other groups was significant (Table 2). While pregabalin without opioids was significant versus high MED opioids + pregabalin for “Ability to enjoy life” (14.2%, 95% CI 28.5%, 4.5%; P < 0.05), it had the lowest ranking for ≥ 50% improvement in fatigue items, with differences in “Worn out” that were significant relative to low MED opioids (-9.8%, 95% CI -15.3%, -2.1%) and moderate MED opioids + pregabalin (-10.2%, 95% CI -1.0%, 18.1%) (both P < 0.05). While pregabalin, with or without opioids, at a dosage of 301-450 mg resulted in the highest proportions of patients achieving ≥ 30% improvement on the two psychosocial items (“Mood” and “Relationships with others”), the 151-300 mg dose showed the best results on the other pain interference items (Table 3). Opioids resulted in the lowest proportions except for “Sleep,” for which it ranked next to lowest, with the highest pregabalin dose ranked last. On pain severity (Table 3), the highest pregabalin dose had the best effects on “Average pain” and “Pain right now,” but very low proportions of patients reported ≥ 30% improvement on “Worst pain.” For patients reporting an improvement of at least 50% from baseline, the pregabalin dosages of 151-300 mg and 301-450 mg were found to have the greatest proportion in all ten items, split amongst the three domains, pain interference, pain severity, and fatigue (Table 3). This article is protected by copyright. All rights reserved. Analysis by area-under-the-curve Accepted Article Figure 1 presents sample profiles for the responder and non-responder using the top four treatments for “Activity in general.” This figure visualizes the comparison across the full range of responses and assists in the simultaneous assessment of improvement and worsening, with higher curves for improvement and lower curves for worsening the preferred outcomes. The area under each line represents the AUC, and the area between the two lines expresses the differences in AUC response between treatments. As shown, pregabalin results in consistently better responses across the range of improvement levels, and results in worsening similar to or less than with the comparator opioid groups. Similar results were generally observed for the other outcomes (data not shown). The derived AUCs for percent improvement in pain and fatigue items are shown in Figure 2. While the highest AUCs for pain were observed for pregabalin, the greatest effects on fatigue were observed with moderate MED opioids + pregabalin. These AUCs can be interpreted as if all responders were improved by the same percentage. For example, the AUC of 20.4% with pregabalin for “Pain right now” suggests that pregabalin responders had on average a 20.4% reduction in current pain. Similarly, Figure 2 also presents the AUCs for pain interference, and shows that pregabalin results in the highest AUC for all interference outcomes except for “Relationships,” which was highest with moderate MED opioids + pregabalin. Figure 3 provides a holistic perspective among the treatments by summarizing patient responses along the entire range of responders and non-responders by rank of improvement (Figure 3A), This article is protected by copyright. All rights reserved. rank of worsening (Figure 3B) and the total effect (Figure 3C) as represented by summation Accepted Article of improvement and worsening; the higher the rank, the better the effect (greater improvement and less worsening). The top three treatments as determined by their total effect are pregabalin, high MED opioids + pregabalin, and high MED opioids. Statistical comparison of pairwise differences in percent improvement for the treatment having the greatest response are shown in Figure 4 for pain severity and fatigue, and Figure 5 for pain interference. Treatment groups that included pregabalin, either in the absence of or combined with opioids, showed the greatest differences in effects. Generally, where the differences are statistically significant, patients taking pregabalin without opioids exhibited an improvement of 5%-15% greater than the other treatments (i.e., high MED opioids + pregabalin, high MED opioids, moderate MED opioids, and moderate MED opioids + pregabalin); patients taking high MED opioids + pregabalin reported an improvement of 4%10% greater than the other treatments; patients taking moderate MED opioids + pregabalin improved 1%-5% more than patients taking moderate or high MED opioids alone. Placing the AUC results in a similar context as the main treatment comparisons, Figure 6 shows that pregabalin dosages of 301-450 mg and 151-300 mg were the top two ranked subgroups for overall total effect, improvement, and worsening. Opioids were third ranked based on patients reporting less worsening than pregabalin dosages of ≤150 mg. DISCUSSION The few studies on opioids for the treatment of FM have suggested that while responders to the weak opioid tramadol may have a reduction in pain but not in health-related quality of life,32 opioids may have contributed to increased resource use and costs, or less favorable This article is protected by copyright. All rights reserved. outcomes.15, 16, 33 The current study expands on the evaluated effectiveness of opioids for FM Accepted Article by showing that although opioids may provide benefits in pain and pain interference outcomes, treatment with pregabalin, which is approved for the treatment of FM, appeared to provide more favorable outcomes relative to opioids. Notably, this responder analysis not only used improvement thresholds to show that the benefits represent clinically relevant outcomes that affect patients’ daily lives, but also included an approach that takes into account the entire response profile for patients who reported improvement as well as for those who worsened. Assessment of worsening is an infrequently evaluated concept that provides broader comparative characterization of treatment effects and enables ranking of treatment modalities based on their overall appropriateness. For example, while responders in the moderate MED opioids + pregabalin group exhibited favorable outcomes for improvement, non-responders in this treatment group seemed to worsen to a greater extent than with other treatments, resulting in moderate MED opioids + pregabalin having the lowest ranking for overall worsening relative to other treatments. In contrast, pregabalin in the absence of opioids had the highest ranking for overall improvement as well as for overall effects, with responders not only achieving, on average, improvements in pain that ranged from 11% for “Worst pain” to 20% for “Pain right now,” but average improvements of 18%-26% for pain interference items. Pregabalin dosages of 301-450 mg generally provided the most benefit to patients compared with other pregabalin dosages as well as with patients prescribed only opioids. Interestingly, the mean dosage of pregabalin increased as the opioid MED increased, which may be consistent with a greater need for pain control. These results are in concordance with a recent This article is protected by copyright. All rights reserved. review that reported pregabalin ≥ 300 mg results in major reductions in pain accompanied by Accepted Article ancillary improvements in functional and quality-of-life outcomes.34 However, in the current study, not only was the mean pregabalin dose, 159.7 mg, approximately half of the lowest recommended dose for FM treatment,20 but the largest proportion of patients (60.5%) were taking pregabalin ≤ 150 mg; the 150 mg dose also had the lowest ranking with regard to total effects. While use of low pregabalin doses, i.e., less than the recommended dose, in the clinical setting has consistently been reported,35-37 the reason for such suboptimal dosing is unclear. It should also be noted that very few patients were taking 451-600 mg, which exceeds the indicated dose range for FM.20 The differences in the AUC (Figures 4 and 5) can be interpreted as an indication of the treatment groups that were favored for each comparison across the evaluated outcomes, and the magnitude of the differences in response. Based on these differences, the domains that showed the greatest responses with pregabalin alone as well as when combined with opioids, i.e., pain interference on “Activity in general” and “Sleep”, were also those that are considered of importance by patients.38, 39 The effects on sleep are not surprising given not only the reciprocal relationship between pain and sleep,40-42 but that pregabalin has positive effects on sleep across a range of conditions, with these effects appearing to occur directly on sleep as well as through the reduction in pain including in patients with FM.43, 44 Such effects are especially important in FM, since sleep disturbances are a primary complaint among these patients even though there is little evidence supporting a role for sleep physiology in maintaining FM.45 Pregabalin without opioids was also the treatment that resulted in the greatest benefits in reducing pain interference on “Mood.” Despite the benefits of pregabalin on sleep and activity, pregabalin in the absence of opioids had a lower impact on fatigue than on other outcomes and also relative to the other treatment groups. This observation suggests This article is protected by copyright. All rights reserved. that fatigue represents a multidimensional concept in FM that may not necessarily reflect the Accepted Article previously described mediation of fatigue through pain and sleep disturbance.46 Nevertheless, this lack of effect on fatigue by pregabalin is not surprising given previous results suggesting that pregabalin does not reduce fatigue in FM.47 Strengths and limitations A strength of this study is its use of data from clinical practice. However, its retrospective design rather than a randomized controlled trial represents a limitation, as does the fact that the population was from a limited geographic area, reducing generalizability. Another limitation is that longitudinal assessment was not performed in individual patients, since duration of treatment may affect the measured outcomes. Furthermore, it should also be considered that polypharmacy is common in FM patients for the treatment of FM itself as well as for comorbid conditions.2, 48, 49 Thus, since many patients are likely to be taking other medications that are recommended or commonly used for treatment of FM (e.g., antidepressants) in addition to pregabalin and opioids, the positive outcomes may have been impacted in part by medications other than pregabalin and/or opioids. CONCLUSIONS Pregabalin in the absence of opioids generally provided the most favorable outcomes in pain severity and pain interference among responders regardless of whether the results were based on clinically relevant thresholds of improvement from baseline (i.e., ≥ 30% and ≥ 50%) or the entire response profile. However, while most patients took less than the recommended pregabalin doses, the preferential pregabalin dosages for improvement and total effects appears to most likely be between 151 and 450 mg, suggesting that higher doses than currently used by patients in the clinical setting may lead to better outcomes. Some This article is protected by copyright. All rights reserved. consideration for the use of opioids in combination with pregabalin was also suggested, in Accepted Article particular, high MED opioids + pregabalin demonstrated positive outcomes with a favorable response profile. Figure 1. Sample response profiles illustrating the analysis based on area-under-the-curve (AUC). The gray area represents the difference in the AUC between the compared treatments among non-responders (worsening) and responders (improvement); results are shown for “Activity in general” for the top four ranked treatments, with all other variables analyzed in a similar manner. This article is protected by copyright. All rights reserved. Accepted Article This article is protected by copyright. All rights reserved. Figure 2. Percent improvement in pain, fatigue, and pain interference among responders by Accepted Article treatment derived by area under the response curve. This article is protected by copyright. All rights reserved. Figure 3. Area under the curve (AUC) rankings for pain interference, pain Accepted Article severity, and fatigue by treatment. This article is protected by copyright. All rights reserved. Figure 4. Differences (95% confidence intervals [CI]) in area under the response curve Accepted Article among responders for pain interference improvement for the top five ranked treatments. *P < 0.05. Values to the left of the broken vertical line indicate the response favors the comparator treatment listed in the column, while values to the right favor the comparator indicated in the row. This article is protected by copyright. All rights reserved. Figure 5. Differences (95% confidence intervals [CI]) in area under the response curve Accepted Article among responders for pain severity and fatigue improvement for the top five ranked treatments. *P < 0.05. Values to the left of the broken vertical line indicate the response favors the comparator treatment listed in the column, while values to the right favor the comparator indicated in the row. This article is protected by copyright. All rights reserved. Figure 6. Area under the curve (AUC) rankings for pain interference, pain Accepted Article severity, and fatigue by pregabalin dosage and patients prescribed any opioids. This article is protected by copyright. All rights reserved. ccepted Articl Table 1. Demographic characteristics the fibromyalgia patients prescribed opioids and/or pregabalin Variable No. of observations (n) Sex, n (%) Female Male Age, years, mean (SD) No. of assessment days, mean (SD) Opioids Percent time on opioid, mean (SD)* No. of prescriptions, mean (SD) MED, mean (SD)† Pregabalin Percent time on prescription, mean (SD)* No. of prescriptions, mean (SD) Dosage, mg, mean (SD)† Total Pregabalin without opioids Opioids + pregabalin Low MED opioids (≤20) Only opioids Moderate MED opioids (>20, <100) 150 (83) High MED opioids (≥100) Low MED opioids (≤20) Moderate MED opioids (>20, <100) High MED opioids (≥100) 60 (28) 1194 (593) 1057 (491) 459 (195) 3082 (1,421) 52 (40) 110 (78) 2396 (77.7) 686 (22.3) 50.5 (12.7) 179.5 (63.3) 41 (78.8) 11 (21.2) 49.8 (12.4) 180.9 (75.1) 91 (82.7) 19 (17.3) 49.5 (12.3) 185.0 (58.5) 112 (74.7) 38 (33.3) 49.9 (12.0) 179.8 (63.4) 59 (98.3) 1 (1.7) 51.0 (11.1) 170.2 (44.1) 921 (77.1) 273 (22.9) 51.1 (13.7) 179.1 (65.9) 789 (74.6) 268 (25.4) 50.3 (11.9) 177.6 (64.0) 383 (83.4) 76 (16.6) 49.9 (12.3) 184.4 (55.3) 77.9 (31.3) — 58.1 (33.5) 90.5 (17.6) 98.5 (3.5) 56.1 (34.6) 92.9 (15.8) 97.6 (7.7) 6.5 (5.1) — 3.3 (2.4) 7.6 (4.4) 12.7 (5.0) 3.1 (2.3) 7.6 (4.1) 12.3 (5.5) 8.3 (5.4) 47.6 (20.4) 227.6 (140.4) 8.0 (5.7) 47.1 (20.9) 223.3 (199.5) 60.6 (111.4) 65.4 (35.1) 52.9 (37.3) 58.6 (37.0) 71.9 (31.5) 72.4 (33.7) — — — 3.7 (2.4) 2.7 (1.7) 3.5 (2.3) 4.0 (2.8) 4.0 (2.2) — — — 159.7 (140.8) 110.8 (124.3) 141.1 (138.2) 176.2 (131.1) 194.6 (165.6) — — — This article is protected by copyright. All rights reserved. ccepted Articl Variable Total Pregabalin without opioids Opioids + pregabalin Low MED opioids (≤20) Pregabalin dosage, % ≤ 150 mg 60.5 76.9 68.2 151-300 mg 23.1 17.3 15.5 301-450 mg 12.6 1.9 12.7 451-600 mg 3.4 3.9 3.6 MED, morphine equivalent dose; SD, standard deviation Moderate MED opioids (>20, <100) High MED opioids (≥100) Low MED opioids (≤20) Moderate MED opioids (>20, <100) High MED opioids (≥100) 52.7 30.7 14.7 2.0 51.7 23.2 16.7 8.3 — — — — — — — — — — — — *Mean percent time on prescription (days of prescription/days of assessment) †Doses adjusted for time (percent time on prescription x daily dosage) This article is protected by copyright. All rights reserved. Only opioids ccepted Articl Table 2. Proportion of fibromyalgia patients who reported improvement ≥ 30% and ≥ 50% from baseline. Pairs of similar superscripts indicate significant differences (P < 0.05) between the indicated treatments. Values reflect total number of patients in the treatment group with nonmissing data for the item evaluated. Variable ≥ 30% improvement Average pain Pain right now Worst pain Tired Worn out Pain interference with ability to enjoy life Pain interference with activity in general Pain interference with mood Pain interference with relationships Pain interference with sleep Pregabalin 24.4 (11.1, 35.6) 25.0 (13.5, 36.5)a 9.6 (1.9, 17.3) 15.4 (5.8, 25.0) 21.2 (9.6, 32.7) 26.2 (11.9, 38.1) 32.5 (17.5, 45.0)abc 23.8 (9.5, 35.7) 25.6 (10.4, 38.5) 30.8 (15.4, 43.6) Percent of patients (95% confidence interval) Opioids + pregabalin Only opioids Low MED Moderate High MED Low MED Moderate MED opioids MED opioids opioids opioids opioids (≤20) (>20, <100) (≥100) (≤20) (>20, <100) 16.3 (9.2, 23.5) 11.8 (6.4, 18.2)ab 4.5 (0.9, 9.1) 16.5 (10.5, 22.6) 18.7 (12.0, 24.7) 4.7 (1.3, 8.0) 20.4 (9.3, 31.5) 13.6 (7.3, 20.0) 20.0 (12.7, 27.3) 20.5 (11.4, 28.4) 10.9 (4.3, 17.4)d 16.9 (9.0, 24.7)* 13.4 (6.1, 20.7)a 29.3 (19.6, 38.0) 22.0 (15.3, 28.0) 24.0 (17.3, 30.7) 17.9 (11.1, 24.8) 20.6 (13.5, 27.8) 20.5 (13.1, 27.0) 26.5 (18.8, 34.2)abc 24.4 (16.3, 31.7) 16.7 (6.7, 26.7) This article is protected by copyright. All rights reserved. 20.0 (10.0, 30.0) 5.0 (0, 10.0) 21.7 (11.7, 31.7) 14.6 (4.9, 24.4) 27.1 (14.6, 39.6)d 21.4 (9.5, 33.3) 17.5 (5.0, 27.5) 27.1 (14.6, 37.5) High MED opioids (≥100) 17.2 (14.8, 19.5) 19.7 (17.4, 21.9)b 6.6 (5.2, 8.0) 16.6 (14.3, 19.0) 21.0 (17.2, 24.8) 16.8 (14.6, 19.0 17.6 (14.2, 21.1) 6.7 (5.2, 8.2) 6.3 (4.1, 8.5) 20.4 (18.0, 22.6) 24.5 (22.0, 27.0) 14.3 (12.0, 16.6) 17.5 (15.1, 19.8)a 17.3 (14.7, 19.7) 16.8 (14.1, 19.5)b 22.1 (19.4, 24.9) 19.6 (17.1, 21.9) 18.5 (15.0, 22.0) 21.5 (19.0, 23.9) 21.4 (17.4, 25.1) 17.0 (14.5, 19.5) 17.5 (13.8, 21.2) 17.1 (14.6, 19.7)b 15.8 (12.1,19.4)c 17.8 (15.1, 20.4) 16.7 (13.0< 20.2) 18.4 (15.6, 21.1) 17.1 (13.1, 20.8)c 22.5 (19.7, 25.3) 24.7 (20.5, 28.8) ccepted Articl Variable ≥ 50% improvement Average pain Pain right now Worst pain Tired Worn out Pregabalin 13.3 (4.4, 22.2) 11.5 (3.8, 21.2) 5.8 (0, 11.5) 5.8 (0, 11.5) 5.8 (0, 11.5)ab Percent of patients (95% confidence interval) Opioids + pregabalin Only opioids Low MED Moderate High MED Low MED Moderate MED opioids MED opioids opioids opioids opioids (≤20) (>20, <100) (≥100) (≤20) (>20, <100) High MED opioids (≥100) 8.2 (3.1, 13.3) 7.3 (2.7, 11.8) 0.9 (0, 2.7) 6.4 (1.8, 10.9) 11.8 (5.5, 18.2) 9.0 (4.5,14.3) 5.6 (0, 11.1) 7.4 (5.7, 9.0) 7.0 (5.3, 8.7) 8.3 (5.7, 10.8) 10.0 (5.3, 14.7) 2.0 (0, 4.7) 13.3 (8.0, 18.7) 16.0 (10.0, 22.0)a 11.7 (3.3, 20.0) 11.1 (9.4, 12.9)a 1.9 (1.2, 2.8) 11.6 (9.9, 13.5)a 15.6 (13.6, 17.7)b 9.2 (7.5, 11.0) 7.0 (4.8, 9.4)a 2.4 (1.5, 3.3) 11.1 (9.2, 13.0) 2.2 (0.9, 3.7) 8.3 (5.9, 10.9)a 12.8 (10.8, 14.8) 12.6 (9.6, 15.7) 1.7 (0, 5.0) 6.7 (1.7, 13.3) 15.0 (6.7, 23.3) Pain interference with ability to enjoy life 16.7 (4.8, 28.6)a 11.4 (4.5, 18.2) 9.4 (4.3, 14.5) 2.4 (0, 7.3)a 8.8 (7.0, 10.7) 8.6 (6.7, 10.5) 8.1 (5.7, 10.8) Pain interference with activity in general 20.0 (7.5, 32.5) 8.7 (3.3, 15.2) 12.7 (7.1, 18.3) 10.4 (2.1, 18.8) 10.1 (8.1, 12.0) 9.6 (7.6, 11.6) 10.2 (7.3, 13.3) Pain interference with mood 16.7 (4.8, 28.6) 6.7 (2.2, 12.4) 11.5 (5.7, 17.2) 14.3 (4.8, 23.8) 9.8 (7.8, 11.8) 9.2 (7.2, 11.1) 9.7 (7.0, 12.7) This article is protected by copyright. All rights reserved. ccepted Articl Variable Pregabalin Pain interference with relationships 17.9 (5.1, 28.2) Pain interference with sleep 23.1 (10.3, 35.9) Percent of patients (95% confidence interval) Opioids + pregabalin Only opioids Low MED Moderate High MED Low MED Moderate MED opioids MED opioids opioids opioids opioids (≤20) (>20, <100) (≥100) (≤20) (>20, <100) 7.3 (2.4, 13.7 (7.7, 7.5 (0, 15.0) 9.9(7.7, 12.0) 9.5 (7.3, 11.6) 13.4) 19.7) 13.0 (6.5, 19.6) This article is protected by copyright. All rights reserved. 18.7 (11.4, 25.2) 18.8 (8.3, 29.2) 13.8 (11.4, 16.1) 15.0 (12.5, 17.3) High MED opioids (≥100) 9.1 (6.4, 12.3) 16.7 (13.1, 20.2) ccepted Articl Table 3. Proportion of fibromyalgia patients reporting improvement ≥ 30% and ≥ 50% from baseline by pregabalin dosages regardless of concomitant opioid use relative to opioids only. Pairs of similar superscripts indicate significant differences (P < 0.05) between the indicated treatments. Variable ≥ 30% improvement Average pain Pain right now Worst pain Tired Worn out Pain interference with ability to enjoy life Pain interference with activity in general Pain interference with mood ≤ 150 mg (n = 225) Percent of patients (95% confidence interval) Pregabalin 151- 300 mg 301-450 mg (n = 86) (n = 47) Opioids (n = 2710) 17.5 (12.5, 22.5) 17.3 (12.0, 22.2) 7.1 (4.0, 10.2)a 17.3 (12.4, 22.2) 20.9 (15.6, 26.2) 17.9 (12.1, 23.7) 19.2 (10.9, 27.4) 15.1 (8.1, 22.1) 2.3 (0, 5.8)a 18.6 (10.5, 26.7) 26.7 (17.4, 36.0) 25.0 (14.1, 34.4) 23.3 (11.6, 34.9) 25.5 (12.8, 36.2) 4.3 (0, 10.6) 14.9 (4.3, 25.5) 23.4 (10.6, 34.0) 21.1 (7.9, 34.2) 17.7 (16.1, 19.2) 18.2 (16.8, 19.7) 6.6 (5.7, 7.6) 19.7 (18.2, 21.2) 22.8 (21.3, 24.4) 16.0 (14.4, 17.5) 17.8 (12.2, 23.3) 22.2 (12.5, 30.6) 20.0 (7.5, 32.5) 17.0 (15.4, 18.6) 18.5 (12.9, 24.2) 19.4 (10.4, 28.4) 28.9 (13.2, 42.1) 17.4 (15.7, 19.0) Pain interference with relationships 20.4 (14.4, 26.3) 22.4 (11.9, 32.8) 24.2 (9.1, 39.4) 17.5 (15.8, 19.2) Pain interference with sleep 28.7 (21.9, 35.4) 29.0 (17.4, 39.1) 22.0 (9.8, 34.1) 22.8 (20.9, 24.5) This article is protected by copyright. All rights reserved. ccepted Articl ≥ 50% improvement Average pain 9.0 (5.5, 13.0) 11.0 (4.1, 17.8) 7.0 (0, 14.0) 7.4 (6.3, 8.5) Pain right now 10.2 (6.2, 14.2) 7.0 (2.3, 12.8) 12.8 (*4.3, 23.4) 9.6 (8.5, 10.7) Worst pain 2.7 (0.9, 4.9) 0 (0, 0) 4.3 (0, 10.6) 2.1 (1.6, 2.7) Tired 7.6 (4.4, 11.1) 15.1 (8.1, 23.3) 6.4 (0, 12.8) 10.8 (9.7, 12.1) Worn out 10.7 (6.7, 14.7)a 19.8 (11.6, 27.9)a 17.0 (6.4, 27.7) 14.0 (12.7, 15.3) Pain interference with ability to enjoy life 10.7 (6.7, 14.7) 12.5 (4.7, 20.3) 7.9 (0, 15.8) 8.6 (7.4, 9.8) Pain interference with activity in general Pain interference with mood 12.2 (7.8, 17.2) 12.5 (5.6, 20.8) 7.5 (0, 15.0) 9.9 (8.6, 11.1) 11.2 (6.7, 15.7) 10.4 (3.0, 17.9) 13.2 (2.6, 23.7) 9.5 (8.3, 10.8) Pain interference with relationships 10.2 (6.0, 15.0) 11.9 (4.5, 19.4) 18.2 (6.1, 30.3) 9.5 (8.3, 10.9) 18.0 (12.4, 23.6) 18.8 (10.1, 27.5) 14.6 (4.9, 24.4) 14.8 (13.2, 16.3) Pain interference with sleep This article is protected by copyright. All rights reserved. References Accepted Article 1. Clauw DJ, Arnold LM, McCarberg BH, FibroCollaborative. The science of fibromyalgia. Mayo Clin Proc. 2011;9:907-911. 2. Halpern R, Shah SN, Cappelleri JC, Masters ET, Clair A. Evaluating Guideline- recommended Pain Medication Use Among Patients with Newly Diagnosed Fibromyalgia. Pain Pract. 2015;16:1027-1039. 3. Bennett RM. Clinical manifestations and diagnosis of fibromyalgia. Rheum Dis Clin North Am. 2009;35:215-232. 4. Kravitz HM, Katz RS. Fibrofog and fibromyalgia: a narrative review and implications for clinical practice. Rheumatol Int. 2015;35:1115-1125. 5. Annemans L, Le Lay K, Taieb C. Societal and patient burden of fibromyalgia syndrome. Pharmacoeconomics. 2009;27:547-559. 6. Schaefer C, Chandran A, Hufstader M, Baik R, McNett M, Goldenberg D, et al. The comparative burden of mild, moderate and severe fibromyalgia: results from a cross-sectional survey in the United States. Health Qual Life Outcomes. 2011;9:71. 7. Chandran A, Schaefer C, Ryan K, Baik R, McNett M, Zlateva G. The comparative economic burden of mild, moderate, and severe fibromyalgia: results from a retrospective chart review and a cross-sectional survey of working-age U.S. adults. J Manag Care Pharm. 2012;18:415-426. 8. Arnold LM, Clauw DJ, Dunegan LJ, Turk DC. A framework for fibromyalgia management for primary care providers. Mayo Clin Proc. 2012;87:488-496. 9. Fitzcharles MA, Ste-Marie PA, Goldenberg DL, Pereira JX, Abbey S, Choiniere M, et al. 2012 Canadian Guidelines for the diagnosis and management of fibromyalgia syndrome: executive summary. Pain Res Manag. 2013;18:119-126. This article is protected by copyright. All rights reserved. 10. American College of Rheumatology. Fibromylagia. Available at: Accepted Article http://www.rheumatology.org/Practice/Clinical/Patients/Diseases_And_Conditions/Fibromya lgia/. Last accessed: December 20, 2016 11. Macfarlane GJ, Kronisch C, Dean LE, Atzeni F, Hauser W, Fluss E, et al. EULAR revised recommendations for the management of fibromyalgia. Ann Rheum Dis. 2016;76:318-328. 12. Ngian GS, Guymer EK, Littlejohn GO. The use of opioids in fibromyalgia. Int J Rheum Dis. 2011;14:6-11. 13. Painter JT, Crofford LJ. Chronic opioid use in fibromyalgia syndrome: a clinical review. J Clin Rheumatol. 2013;19:72-77. 14. Goldenberg DL, Clauw DJ, Palmer RE, Clair AG. Opioid Use in Fibromyalgia: A Cautionary Tale. Mayo Clin Proc. 2016;91:640-648. 15. Margolis J, Masters ET, Cappelleri JC, Smith D, Faulkner S. Evaluating increased resource use in fibromyalgia using electronic health records. Clinicoecon Outcomes Res. 2016;8:675-683. 16. Painter JT, Crofford LJ, Butler JS, Talbert J. Healthcare Costs Associated With Chronic Opioid Use and Fibromayalgia Syndrome. Am J Pharm Benefits. 2014;6:e177-e184. 17. Bennett RM, Jones J, Turk DC, Matallana L. An internet survey of 2,596 people with fibromyalgia. BMC Musculoskelet Disord. 2007;8:27. 18. Gostine M, Davis F, Roberts B, Risko R, Asmus M, Cappelleri JC, et al. Clinical characteristics of fibromyalgia in a chronic pain population. Pain Pract. 2017:Apr 18. [Epub ahead of print] doi: 10.1111/papr.12583. 19. Davis F, Gostine M, Roberts BA, Risko R, Cappelleri JC, Clair A, et al. Classifying Fibromyalgia: Potential for More Precise Treatment Options. Poster presented at the This article is protected by copyright. All rights reserved. American Academy of Pain Medicine 2017 Annual Meeting, Orlando, FL; March 16-19, Accepted Article 2017. 20. Lyrica® [pregabalin] capsules prescribing information. Pfizer, Inc., New York, NY. December 2016. 21. Centers for Medicare & Medicaid Services/US Department of Health & Human Services. Opioid morphine equivalent conversion factors. Available at: https://www.cms.gov/medicare/prescription-drugcoverage/prescriptiondrugcovcontra/downloads/opioid-morphine-eq-conversion-factorsmarch-2015.pdf. Last accessed: December 20, 2016 22. Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singapore. 1994;23:129-138. 23. Ware J E Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473-483. 24. Dworkin RH, Turk DC, Farrar JT, Haythornthwaite JA, Jensen MP, Katz NP, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain. 2005;113:9-19. 25. Choy EH, Arnold LM, Clauw DJ, Crofford LJ, Glass JM, Simon LS, et al. Content and criterion validity of the preliminary core dataset for clinical trials in fibromyalgia syndrome. J Rheumatol. 2009;36:2330-2334. 26. Mease PJ, Clauw DJ, Christensen R, Crofford LJ, Gendreau RM, Martin SA, et al. Toward development of a fibromyalgia responder index and disease activity score: OMERACT module update. J Rheumatol. 2011;38:1487-1495. 27. Dworkin RH, Turk DC, Wyrwich KW, Beaton D, Cleeland CS, Farrar JT, et al. Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations. J Pain. 2008;9:105-121. This article is protected by copyright. All rights reserved. 28. Farrar JT, Dworkin RH, Max MB. Use of the cumulative proportion of responders Accepted Article analysis graph to present pain data over a range of cut-off points: making clinical trial data more understandable. J Pain Symptom Manag. 2006;31:369-377. 29. Cappelleri JC, Bushmakin AG, Zlateva G, Sadosky A. Pain responder analysis: use of area under the curve to enhance interpretation of clinical trial results. Pain Pract. 2009;9:348353. 30. Bushmakin AG, Cappelleri JC, Zlateva G, Sadosky A. Applying area-under-the-curve analysis to enhance interpretation of response profiles: an application to sleep quality scores in patients with fibromyalgia. Qual Life Res. 2010;20:491-498. 31. Tallarida R, Murray R. Area under a Curve: Trapezoidal & Simpson's rules. Manual of Pharmacologic Calculations: With Computer Programs. New York: Springer; 1987. 7781. 32. Russell IJ, Kamin M, Bennett RM, Schnitzer TJ, Green JA, Katz WA. Efficacy of tramadol in treatment of pain in fibromyalgia. J Clin Rheumatol. 2000;6:250-257. 33. Fitzcharles MA, Faregh N, Ste-Marie PA, Shir Y. Fitzcharles MA, Faregh N, Ste- Marie PA, Shir Y. Opioid use in fibromyalgia is associated with negative health related measures in a prospective cohort study. Pain Res Treat. 2013;2013:898493. 34. Derry S, Cording M, Wiffen PJ, Law S, Phillips T, Moore RA. Pregabalin for pain in fibromyalgia in adults. Cochrane Database Syst Rev. 2016;9:CD011790. 35. Sun P, Zhao Y, Zhao Z, Watson P. Medication dosing patterns associated with duloxetine and pregabalin among patients with fibromyalgia. Curr Med Res Opin. 2011;27:1793-1801. 36. Kim SC, Landon JE, Solomon DH. Clinical characteristics and medication uses among fibromyalgia patients newly prescribed amitriptyline, duloxetine, gabapentin, or pregabalin. Arthritis Care Res (Hoboken). 2013;65:1813-1819. This article is protected by copyright. All rights reserved. 37. Liu Y, Qian C, Yang M. Treatment Patterns Associated with ACR-Recommended Accepted Article Medications in the Management of Fibromyalgia in the United States. J Manag Care Spec Pharm. 2016;22:263-271. 38. Arnold LM, Crofford LJ, Mease PJ, Burgess SM, Palmer SC, Abetz L, et al. Patient perspectives on the impact of fibromyalgia. Patient Educ Couns. 2008;73:114-120. 39. Choy E, Perrot S, Leon T, Kaplan J, Petersel D, Ginovker A, et al. A patient survey of the impact of fibromyalgia and the journey to diagnosis. BMC Health Serv Res. 2010;10:102. 40. Smith MT, Haythornthwaite JA. How do sleep disturbance and chronic pain inter- relate? Insights from the longitudinal and cognitive-behavioral clinical trials literature. Sleep Med Rev. 2004;8:119-132. 41. Finan PH, Goodin BR, Smith MT. The association of sleep and pain: an update and a path forward. J Pain. 2014;14:1539-1552. 42. Koffel E, Kroenke K, Bair MJ, Leverty D, Polusny MA, Krebs EE. The bidirectional relationship between sleep complaints and pain: Analysis of data from a randomized trial. Health Psychol. 2016;35:41-49. 43. Roth T, Arnold LM, Garcia-Borreguero D, Resnick M, Clair AG. A review of the effects of pregabalin on sleep disturbance across multiple clinical conditions. Sleep Med Rev. 2014;18:261-271. 44. Russell IJ, Crofford L, Leon AC, Cappelleri JC, Bushmakin AG, Whalen E, et al. The effects of pregabalin on sleep disturbance symptoms among individuals with fibromyalgia syndrome. Sleep Med. 2009;10:604-610. 45. Diaz-Piedra C, Di Stasi LL, Baldwin CM, Buela-Casal G, Catena A. Sleep disturbances of adult women suffering from fibromyalgia: a systematic review of observational studies. Sleep Med Rev. 2015;21:86-99. This article is protected by copyright. All rights reserved. 46. Nicassio PM, Moxham EG, Schuman CE, Gevirtz RN. The contribution of pain, Accepted Article reported sleep quality, and depressive symptoms to fatigue in fibromyalgia. Pain. 2002;100:271-279. 47. Uceyler N, Sommer C, Walitt B, Hauser W. Anticonvulsants for fibromyalgia. Cochrane Database Syst Rev. 2013:CD010782. 48. Berger A, Martin S, Dukes E, Edelsberg J, Oster G. Patterns of pain-related pharmacotherapy in patients with fibromyalgia. Manag Care Interface. 2008;20:30-38. 49. Vincent A, Whipple MO, McAllister SJ, Aleman KM, St Sauver JL. A cross-sectional assessment of the prevalence of multiple chronic conditions and medication use in a sample of community-dwelling adults with fibromyalgia in Olmsted County, Minnesota. BMJ Open. 2015;5:e006681. This article is protected by copyright. All rights reserved.