Gout, a common form of inflammatory arthritis, is REVIEW

Transcription

1 Annals of Internal Medicine REVIEW Management of Gout: A Systematic Review in Support of an American College of Physicians Clinical Practice Guideline Paul G. Shekelle, MD, PhD; Sydne J. Newberry, PhD; John D. FitzGerald, MD, PhD; Aneesa Motala, BA; Claire E. O Hanlon, MPP; Abdul Tariq, BS; Adeyemi Okunogbe, MD; Dan Han, MPA; and Roberta Shanman, MLS Background: Gout is a common type of inflammatory arthritis in patients seen by primary care physicians. Purpose: To review evidence about treatment of acute gout attacks, management of hyperuricemia to prevent attacks, and discontinuation of medications for chronic gout in adults. Data Sources: Multiple electronic databases from January 2010 to March 2016, reference mining, and pharmaceutical manufacturers. Study Selection: Studies of drugs approved by the U.S. Food and Drug Administration and commonly prescribed by primary care physicians, randomized trials for effectiveness, and trials and observational studies for adverse events. Data Extraction: Data extraction was performed by one reviewer and checked by a second reviewer. Study quality was assessed by 2 independent reviewers. Strength-of-evidence assessment was done by group discussion. Data Synthesis: High-strength evidence from 28 trials (only 3 of which were placebo-controlled) shows that colchicine, nonsteroidal anti-inflammatory drugs (NSAIDs), and corticosteroids reduce pain in patients with acute gout. Moderate-strength evidence suggests that low-dose colchicine is as effective as highdose colchicine and causes fewer gastrointestinal adverse events. Moderate-strength evidence suggests that uratelowering therapy (allopurinol or febuxostat) reduces long-term risk for acute gout attacks after 1 year or more. High-strength evidence shows that prophylaxis with daily colchicine or NSAIDs reduces the risk for acute gout attacks by at least half in patients starting urate-lowering therapy, and moderate-strength evidence indicates that duration of prophylaxis should be longer than 8 weeks. Although lower urate levels reduce risk for recurrent acute attacks, treatment to a specific target level has not been tested. Limitation: Few studies of acute gout treatments, no placebocontrolled trials of management of hyperuricemia lasting longer than 6 months, and few studies in primary care populations. Conclusion: Colchicine, NSAIDs, and corticosteroids relieve pain in adults with acute gout. Urate-lowering therapy decreases serum urate levels and reduces risk for acute gout attacks. Primary Funding Source: Agency for Healthcare Research and Quality. (Protocol registration: ahrq.gov/ehc/products/564/1992/gout-managment-protocol pdf) Ann Intern Med. 2017;166: doi: /m For author affiliations, see end of text. This article was published at on 1 November Gout, a common form of inflammatory arthritis, is characterized by acute intermittent episodes of synovitis that cause joint swelling and pain. Approximately 8 million persons in the United States have gout (1). It occurs when excess urate in the body crystalizes (as monosodium urate) in joint fluid, cartilage, bones, tendons, bursas, or other sites. The crystals can directly initiate an acute inflammatory attack. In some patients, acute gout attacks become progressively more frequent, protracted, and severe and may progress to a chronic inflammatory condition. In addition, some patients develop tophi, which are deposits of urate crystals at the surface of joints or in skin or cartilage. This systematic review was proposed by the American College of Physicians (ACP) to support the development of a clinical practice guideline that would aid primary care practitioners in the management of adult patients with gout. METHODS We developed a protocol ( ahrq.gov/ehc/products/564/1992/gout-managment-protocol pdf); followed PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines for reporting systematic reviews (2); and detailed search and selection processes, inclusion criteria, and evidence tables in an evidence report (3). Key Questions Key questions proposed by ACP were revised on the basis of input from a group of key informants, a technical expert panel, and the public. Treatment questions addressed benefits and harms of pharmacologic and dietary therapies for adults with acute gout attacks and the intermediate and long-term benefits and harms of therapies for adults with gout and hyperuricemia. Additional management questions addressed monitoring of serum urate levels and whether criteria exist to identify patients who are candidates for discontinuation of urate-lowering therapy or anti-inflammatory prophylaxis. Data Sources and Searches We searched (without language restrictions) for systematic reviews and original research studies in PubMed, EMBASE, the Cochrane Collaboration, and See also: Related articles....27,52,58 Editorial comments...71, American College of Physicians 37

2 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout the Web of Science, using the terms gout and gouty and terms for tophi. The start date of the searches was 1 January 2010, which was at least 1 year before the search dates for the most recent systematic reviews, and the end date was 1 March Relevant references were obtained from 29 recent systematic reviews. We searched ClinicalTrials.gov from inception to 1 March 2016 and the Web of Science from 1 January 2010 through 1 March 2016, and we contacted manufacturers of prescription medications used to treat gout for recently completed studies and unpublished or non peer-reviewed study findings in July Appendix Table 1 (available at provides detailed search methods. Study Selection Two reviewers independently screened records (titles, abstracts, and articles) to identify reviews and studies that reported on the benefits (randomized trials) or harms (observational studies and trials) of treatment and management strategies for gout. We examined only medications approved by the U.S. Food and Drug Administration (FDA), except for nonsteroidal anti-inflammatory drugs (NSAIDs), which are commonly used to treat gout. As suggested by the ACP Clinical Guidelines Committee, we excluded pegloticase and lesinurad, which primary care physicians are unlikely to prescribe. Studies that enrolled participants with no formal gout diagnosis (based on either synovial fluid analysis or a clinical diagnosis) were excluded. Data Extraction and Quality Assessment Study-level details were abstracted by one reviewer and checked by a second reviewer, with reconciliation of disagreements by group discussion. Risk of bias of individual studies was assessed independently by 2 reviewers using an adapted Cochrane risk-of-bias tool (4), with reconciliation of disagreements by the project lead. A modified AMSTAR (A Measurement Tool to Assess Systematic Reviews) tool was used to assess the quality of systematic reviews (5). Data Synthesis and Grading We deemed studies to be too few in number and too heterogeneous to support new meta-analysis. We assessed the overall strength of evidence as high, moderate, low, or insufficient for each conclusion by using guidance suggested by the Effective Health Care Program (6). We also applied criteria proposed by Bradford Hill for causality when judging strength of evidence (7), including the strength, consistency, and specificity of the association; the temporal relationship; the biologic gradient or dose response curve; the biologic plausibility; and coherence. Role of the Funding Source This topic was proposed by the ACP to the Agency for Healthcare Research and Quality (AHRQ) and funded by AHRQ. Staff at AHRQ and ACP helped to develop and refine the scope of the study and reviewed the draft report. RESULTS The Figure shows the search and selection process that identified 155 articles meeting the inclusion criteria. Of these, 22 evaluated dietary therapy or traditional Chinese medicine; details about the inconclusive evidence from those studies are in the evidence report (3). Pharmacologic Treatment for Acute Gout We identified evidence for several pharmacologic treatments for acute gout: colchicine, NSAIDs, corticosteroids, and animal-derived corticotropin formulation (8 35). None of the studies assessed differences in effectiveness by patient characteristics, such as age, sex, duration of the episode, history, genetic profile, baseline serum urate level, or presence of comorbidities. Colchicine Although colchicine was used to treat gout in the ninth century or earlier (36), its use has been evaluated in only 2 randomized, placebo-controlled trials (12, 14). These trials enrolled 184 and 43 patients, respectively, who had crystal-proven gout or met the American College of Rheumatology criteria and had mean duration of symptoms of 38 hours or less. Both studies found that patients treated with colchicine had better pain relief than placebo recipients (38% vs. 16% achieved a 50% decrease in target joint pain at 24 hours, and 41% vs. 9% achieved a 50% decrease in baseline pain score at 24 hours). The earlier trial, published in 1987, used an initial colchicine dose of 1 mg followed by 0.5 mg every 2 hours until symptom relief or adverse effects occurred (14). All of the colchicine-treated patients had gastrointestinal adverse effects by 24 hours, and 91% reported adverse effects before achieving a 50% reduction in pain intensity. The later trial compared lowdose (an initial dose of 1.2 mg followed by 0.6 mg 1 hour later) versus high-dose (an initial dose of 1.2 mg followed by 0.6 mg per hour for 6 hours) colchicine (12). Low-dose colchicine was as effective as high-dose colchicine compared with placebo and was much better tolerated; for example, 23% versus 77% of patients receiving low- and high-dose therapy reported diarrhea, and 0% versus 19% of these patients reported severe diarrhea (Table 1). NSAIDs One low-quality, placebo-controlled trial assessed NSAIDs in patients with acute gout (30). This small study assessed the effect of tenoxicam (40 mg once daily) in 30 patients and found that a greater proportion of those receiving tenoxicam than those receiving placebo reported at least 50% improvement at 24 hours. Sixteen randomized trials compared one NSAID with another (8, 9, 11, 15 19, 21 23, 27 29, 31, 32). Most studies found no statistically significant differences in outcomes between treatments, although only 3 studies enrolled more than 100 patients (9, 22, 23), meaning that most studies were small and had limited 38 Annals of Internal Medicine Vol. 166 No. 1 3 January

3 Systematic Review in Support of ACP Guideline on Management of Gout REVIEW Figure. Literature flow diagram. Abstracts identified for dual review (n = 7928) References identified from previous systematic reviews or guidelines: 233 Titles identified from RAND library searches: 7403 Gray literature: 0 ClinicalTrials.gov: 292 Abstracts excluded (n = 7226) Not human subjects: 295 Not about gout or hyperuricemia associated with gout: 1630 Not about gout diagnosis or management or did not address key question: 3855 Study of risk factors for gout did not test possible treatment: 89 No original data or non systematic review background: 508 Case reports: 287 Population not of interest: 75 No abstract: 199 Gout diagnosis only: 104 Biologics outside scope of review: 133 Duplicate data: 51 Total articles identified for full-text review (n = 702) Full-text articles excluded (n = 547) Not human subjects: 2 Not about gout or hyperuricemia associated with gout: 26 Not about gout diagnosis or management or did not address key question: 156 Study of risk factors for gout that did not test possible treatment: 18 No original data or non systematic review background: 97 Study design: 66 Case reports: 51 Population not of interest: 8 Gout diagnosis only: 6 Biologics outside scope of review: 64 No outcomes of interest: 11 No interventions of interest: 2 Duplicate data: 37 Article not found: 3 Articles included for data synthesis* Acute gout (n = 49) Included in this analysis: 28 Included only in evidence report: 21 Diet and lifestyle management (n = 22) Treatment for hyperuricemia (n = 62) Included in this analysis: 27 Included only in evidence report: 35 Monitoring (n = 27) Included in this analysis: 9 Included only in evidence report: 18 Criteria for discontinuation (n = 3) Included in this analysis: 3 * The number of included studies for this article differs from the number of included studies in the evidence report (3) because this review did not address all of the key questions addressed in the report. Results for this question are not included in this article. See the evidence report (3). power to detect differences (Appendix Table 2, available at Corticosteroids No published placebo-controlled trials assessed oral corticosteroids for acute gout. Six randomized trials compared corticosteroids (3 oral, 2 intramuscular, and 1 intravenous) versus NSAIDs. These studies, which enrolled 27 (20), 60 (13), 90 (24), 92 (33), 120 (10), and 416 (34) patients, found few differences in effectiveness or adverse events between treatments. In one study of oral corticosteroids, 90 patients presenting to the Annals of Internal Medicine Vol. 166 No. 1 3 January

4 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout Table 1. Key Trials of Treatments for Acute Gout Study, Year (Reference) Ahern et al, 1987 (14) Terkeltaub et al, 2010 (12) Alloway et al, 1993 (20) Zhang et al, 2014 (13) Man et al, 2007 (24) Study Population Patients with acute gout proven by crystals in synovial fluid analysis Patients with confirmed previous diagnosis of gout with 2 gout attacks in the prior 12 mo; 9% had tophi All patients seen by the rheumatology section for acute gout during a 9 month period were asked to participate Patients who met ACR criteria for acute gout Patients with acute arthritis suggestive of gout who presented to the emergency department Average Age, y Male, % Symptom Duration Participants, n Average: 38 h Colchicine, 1 mg followed by 0.5 mg every 2 h until relief or intolerable adverse effects 2. Placebo Current symptoms: <12 h Comparison Duration Results Colchicine, 1.2 mg followed by 0.6 mg 1 h later ( low-dose ) 2. Colchicine, 1.2 mg followed by 0.6 mg every hour up to 6 h ( high-dose ) 3. Placebo Average: 3.2 d Triamcinolone acetonide, 60 mg intramuscularly 2. Indomethacin, 50 mg 3 times daily for 2 d <24 h Betamethasone, 7mg intramuscularly 2. Diclofenac, 75 mg twice daily for 7 d Average: 2.3 d Diclofenac, 75 mg intramuscularly; indomethacin, 50 mg orally; acetaminophen, 1 g orally; and a 5-d prescription for an indomethacin taper 2. Prednisolone, 30 mg orally; acetaminophen, 1g;and prednisolone, 30 mg/d for 5 d 48 h Proportion of patients with 50% pain relief: At 24 h: Colchicine: 41% Placebo: 9% At 48 h: Colchicine: 73% Placebo: 36% Nausea and vomiting at 24 h: Colchicine: 100% Placebo: 24% (nausea only) 24 h Proportion of patients with 50% pain relief at 24 h: Low-dose colchicine: 38% High-dose colchicine: 33% Placebo: 16% Gastrointestinal adverse events: Low-dose colchicine: 26% High-dose colchicine: 77% Placebo: 20% 30 d No significant difference in any outcome. No adverse events in either group. 14 d Proportion of patients reporting severe or extreme pain at day 2: Betamethasone: 10% Diclofenac: 30% All time points beyond 3dhad no statistically significant differences between groups. Adverse events related to the study drug: Betamethasone: 10% Diclofenac: 20% 12 d Statistically significant difference of 1-mm faster improvement on a 10-cm pain scale, favoring prednisolone, of questionable clinical significance. Adverse effects: Indomethacin: 63% Prednisolone: 27% Continued on following page 40 Annals of Internal Medicine Vol. 166 No. 1 3 January

5 Systematic Review in Support of ACP Guideline on Management of Gout REVIEW Table 1 Continued Study, Year (Reference) Janssens et al, 2008 (10) Siegel et al, 1994 (26) Axelrod and Preston, 1988 (25) Liu et al, 2015 (33) Rainer et al, 2016 (34) Study Population Patients presenting to family physicians with acute monoarthritis who then had crystal-proven gout All patients presenting to the rheumatology service with acute gout were asked to participate Patients presenting with acute onset of pain and crystal-proven gout Patients who met 1997 ACR criteria for acute gout Patients presenting to the emergency departments of 4 Hong Kong hospitals who had symptoms lasting 3 d and were considered to have gout Average Age, y ACR = American College of Rheumatology. Male, % Symptom Duration Participants, n d Prednisolone, 35 mg/d for 5 d 2. Naproxen, 500 mg twice daily for5d Average: 28 d Corticotropin, 40 IU intramuscularly 2. Triamcinolone acetonide, 60 mg intramuscularly Duration of pain: <24 h Comparison Duration Results Indomethacin, 50 mg 4 times daily until pain abated 2. Corticotropin, 40 IU intramuscularly d Colchicine, 0.5 mg twice daily, and intravenous dexamethasone, 2.5 mg for 3 d 2. Colchicine, 1 mg orally then 0.5 mg every 1 2 h for 24 h then 1 mg twice daily for3d Average: 2.8 d Indomethacin, 50 mg 3 times daily for 2 d then 25 mg 3 times daily for 3 d 2. Prednisolone, 30 mg/d for 5 d 5 d No significant difference in outcomes between groups. No significant differences in reported adverse events between groups. 30 d More patients in the corticotropin group than the triamcinolone acetonide group required reinjection due to inadequate symptom relief (9 vs. 5; P = 0.11). Equivalent duration to 100% resolution for both groups (7.9 vs. 7.6 d). 1 y Time to pain relief was faster in the corticotropin group than in the indomethacin group (3 vs. 24 h). Adverse effects were more common in indomethacintreated patients. 1 mo Proportion of patients reporting treatment as markedly effective: At6h: Colchicine and dexamethasone: 70% Colchicine: 28% At 24 h: Colchicine and dexamethasone: 85% Colchicine: 70% Adverse effects: Colchicine and dexamethasone: 0% Colchicine: 76% (nausea only) 14 d No statistically significant differences in pain scores at any time point except in pain at rest on days 1 5. No difference in overall adverse effects. 11.1% of indomethacintreated patients and 5.8% of prednisolone-treated patients had abdominal pain. emergency department with acute arthritis suggestive of gout were randomly assigned to receive either prednisolone, 30 mg/d for 5 days, or indomethacin, 50 mg 3 times daily for 2 days and 25 mg/d for the next 3 days; all patients also received paracetamol. No statistically or clinically significant between-group differences were seen at any evaluation point (up to 14 days). More patients in the indomethacin group than the prednisolone Annals of Internal Medicine Vol. 166 No. 1 3 January

6 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout group reported adverse events (63% vs. 27%) (24). In another study of oral corticosteroids, 120 patients who presented to family physicians with acute monoarthritis and had monosodium urate crystals on synovial fluid examination were randomly assigned to prednisolone, 35 mg/d, or naproxen, 500 mg twice daily. Over the subsequent 4 days, no statistically significant differences in effectiveness outcomes were observed between groups. Equal proportions of patients in both groups reported an adverse event (66% vs. 63%) (10). The third study of oral corticosteroids randomly assigned 416 patients with a clinical diagnosis of gout with symptoms lasting less than 3 days to indomethacin or prednisolone for 5 days. There were no important statistically or clinically significant differences in pain outcomes or overall adverse events. Gastrointestinal adverse events were more common in the indomethacin-treated patients, and skin rash was more common in the prednisolone-treated patients (34) (Table 1). Corticotropin No published placebo-controlled trials tested corticotropin for acute gout. Two randomized, controlled trials compared intramuscular corticotropin versus other active agents. One evaluated corticotropin, 40 IU, versus triamcinolone, 60 mg, both given intramuscularly only once (26); the other evaluated corticotropin, 40 IU once, versus indomethacin, 50 mg 4 times a day until pain abated (25). In the former study, patients treated with triamcinolone required fewer reinjections for inadequate pain relief than those treated with corticotropin at up to 30 days of follow-up (5 vs. 9 patients; P = 0.11). In the latter study, pain relief was faster with corticotropin than with oral indomethacin (3 vs. 24 hours to total pain relief), and at the dose of 50 mg 4 times a day, 55% of indomethacin-treated patients reported gastrointestinal adverse events (no adverse events were reported by patients receiving corticotropin) (Table 1). Summary Despite the small number of placebo-controlled trials, we judged the strength of evidence as high that colchicine, NSAIDs, and corticosteroids relieve pain in patients with acute gout, based on the known physiology of gout, the known mechanism of action of these drugs, the proven effectiveness of these medications in other painful or inflammatory conditions, and evidence of equivalence in head-to-head trials of patients with acute gout. We judged the evidence for use of corticotropin to be of moderate strength because of high risk of bias in the only 2 head-to-head randomized trials. Pharmacologic Management of Hyperuricemia in Patients With Gout We found 11 systematic reviews (37 47), 1 metaanalysis (48), 1 new abstract (49), 5 secondary analyses (50 54) of trials already included in the systematic reviews, 9 new trials assessing effectiveness (55 63), and 26 studies of adverse effects that addressed this question (64 89). Four large randomized trials of febuxostat provided most of the effectiveness evidence: APEX (Allopurinol- and Placebo-Controlled Efficacy Study of Febuxostat) (90), FACT (Febuxostat versus Allopurinol Controlled Trial) (91), the CONFIRMS (Confirmation of Febuxostat in Reducing and Maintaining Serum Urate) trial (92), and the open-label EXCEL (Febuxostat/ Allopurinol Comparative Extension Long-Term) study (93). The details of these studies are presented in Table 2. More than 90% of enrolled participants were men; mean age was approximately 52 years, and mean baseline serum urate level was approximately 583 μmol/l (9.8 mg/dl). Across the studies, patients were randomly assigned to various doses of febuxostat (40, 80, 120, and/or 240 mg/d; doses >80 mg/d are not FDAapproved), allopurinol, or placebo, with adjustment for renal insufficiency. Patients in all studies received prophylaxis with colchicine or naproxen when starting urate-lowering therapy. The findings were as follows. First, all active therapies decreased serum urate levels compared with placebo. Second, no important differences were seen between treatment groups or between active treatment and placebo in the frequency of acute gout attacks. Third, in the long-term extension study (EXCEL), patients achieving a serum urate level less than 357 μmol/l (<6 mg/dl) had progressive decreases in their risk for acute gout attacks (to about 5% at 12 months and near zero at 32 months), regardless of choice of urate-lowering therapy. Fourth, febuxostat, 80 mg/d, was more effective than allopurinol, 300 mg/d, at decreasing serum urate levels. Fifth, no major or important differences were seen in outcomes or total adverse events between allopurinol, 300 mg/d, and febuxostat, 40 mg/d. Sixth, discontinuation of colchicine or naproxen prophylaxis after 8 weeks in both APEX and FACT was associated with a spike in acute gout attacks not seen in CONFIRMS, which maintained prophylaxis throughout the 6-month duration of the study. Finally, studies assessing comparative effectiveness in subgroups were sparse but generally found no differences in clinical outcomes, although they had limited power to detect such differences (3). Several smaller studies had findings similar to those of the larger randomized, controlled trials (49, 55, 63, 94, 95). Appendix Table 3 (available at presents selected adverse events seen in the randomized, controlled trials of urate-lowering therapy. Differences between active treatments and between active treatments and placebo were mostly small and not statistically significant. Although skin rash was not statistically significantly more likely in allopurinol-treated patients in these trials, observational evidence suggested that allopurinol is a cause of DRESS (drug rash with eosinophilia and systemic symptoms) syndrome (65, 68 70, 72) and is more common in patients with the HLA-B*5801 allele (41, 62, 66, 74, 79, 86, 88). In summary, high-strength evidence suggests that urate-lowering therapy reduces serum urate levels. However, it does not reduce the risk for acute gout 42 Annals of Internal Medicine Vol. 166 No. 1 3 January

7 Systematic Review in Support of ACP Guideline on Management of Gout REVIEW Table 2. Key Trials of Urate-Lowering Therapy in Patients With Gout Study, Year (Reference) APEX; Schumacher et al, 2008 (90) FACT; Becker et al, 2005 (91) CONFIRMS; Becker et al, 2010 (92) Study Population Adults with gout and hyperuricemia; 28% had tophi Adults with gout and hyperuricemia; 24% had tophi Adults with gout and hyperuricemia Average Age, y Male, % Mean Baseline Serum Urate Level, mol/l (mg/dl) Participants, n (9.8) Placebo 2. Allopurinol, 300 mg/d 3. Febuxostat, 80 mg/d 4. Febuxostat, 120 mg/d 5. Febuxostat, 240 mg/d All patients were treated with colchicine or naproxen for prophylaxis for 8 wk (9.8) Allopurinol, 300 mg/d 2. Febuxostat, 80 mg/d 3. Febuxostat, 120 mg/d All patients were treated with colchicine or naproxen for prophylaxis for 8 wk (9.6) Allopurinol, 300 mg/d (200 mg/d in patients with renal impairment) 2. Febuxostat, 40 mg/d 3. Febuxostat, 80 mg/d All patients were treated with colchicine or naproxen for prophylaxis for 6 mo Comparison Duration Results 28 wk All treatments reduced serum urate levels more than placebo. Proportion of participants achieving serum urate levels <357 μmol/l (<6 mg/dl): Allopurinol, 300 mg/d: 41% Febuxostat, 80 mg/d: 76% No difference in acute gout attacks between febuxostat, 80 mg/d, and allopurinol, 300 mg/d. No differences in any adverse event. 52 wk Proportion of participants achieving serum urate level <357 μmol/l (<6 mg/dl): Allopurinol, 300 mg/d: 21% Febuxostat, 80 mg/d: 53% No difference in acute gout attacks between allopurinol, 300 mg/d, and febuxostat, 80 mg/d (64% each). No differences in any treatment related adverse events. 6 mo Proportion of participants achieving serum urate level <357 μmol/l (<6 mg/dl): Allopurinol, 300 or 200 mg/d: 42% Febuxostat, 40 mg/d: 45% Febuxostat, 80 mg/d: 67% No difference in acute gout attacks between groups (10% 15% at beginning of treatment, 5% 8% at end of trial). No differences between groups in total participants with an adverse event. Continued on following page Annals of Internal Medicine Vol. 166 No. 1 3 January

8 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout Table 2 Continued Study, Year (Reference) EXCEL; Becker et al, 2009 (93) Study Population Open-label extension trial enrolling patients from FACT and APEX Average Age, y Male, % Mean Baseline Serum Urate Level, mol/l (mg/dl) Participants, n (9.8) Allopurinol, 300 mg/d (100 mg/d in patients with renal impairment) (n = 145) 2. Febuxostat, 80 mg/d (n = 649) 3. Febuxostat, 120 mg/d (n = 292) Comparison Duration Results mo (664 participants [61%] completed the study) Maintenance of serum urate level <357 μmol/l (<6 mg/dl) resulted in progressive reduction in proportion of participants with acute gout attacks, to near zero by 32 mo. APEX = Allopurinol- and Placebo-Controlled Efficacy Study of Febuxostat; CONFIRMS = Confirmation of Febuxostat in Reducing and Maintaining Serum Urate; EXCEL = Febuxostat/Allopurinol Comparative Extension-Long Term; FACT = Febuxostat versus Allopurinol Controlled Trial. attacks in the first 6 months. Long-term extension studies show that patients continuing urate-lowering therapy who achieve serum urate reductions have fewer acute gout attacks, which supports the moderatestrength conclusion that urate-lowering therapy reduces the risk for such attacks after about 1 year. Prophylaxis Against Acute Gout Attacks at Initiation of Urate-Lowering Therapy Initiation of therapy to decrease serum urate levels is associated with an increased frequency of acute gout attacks (96). This is a likely explanation for the observation that use of urate-lowering therapy does not reduce the frequency of acute gout attacks compared with placebo for the first 6 months. More than 30 years ago, investigators tested colchicine as prophylaxis against acute attacks for patients starting uricosuric therapy (97, 98). However, the first randomized, placebocontrolled trial of colchicine prophylaxis at initiation of allopurinol therapy was not published until 2004 (57). In this study, investigators randomly assigned 51 patients to colchicine, 0.6 mg twice daily, or placebo when starting allopurinol therapy at 100 mg once daily and titrating upward, with a target serum urate level of 387 μmol/l (6.5 mg/dl). Occurrence of gout attacks was recorded by patient recall at 3- and 6-month visits. Among 43 patients who completed the trial (mean age, 63 years; >50% male; >60% with tophi; approximately 10% with chronic renal insufficiency), 77% of placebo recipients versus 33% of colchicine-treated patients had gout attacks (P = 0.008). During the first 3 months of treatment, placebo recipients averaged about 2 attacks and colchicine-treated patients averaged about 0.5 attack. From months 3 to 6, this advantage diminished, with about 1 attack per patient in the placebo group and almost no attacks in the colchicine group. Diarrhea was much more common in colchicine-treated patients (43%) than in placebo recipients (approximately 4%). The use of prophylactic therapy concomitant with the initiation of urate-lowering therapy has been the recommended standard of care for many years (99, 100). All 3 of the recent large trials of urate-lowering therapy (FACT, APEX, and CONFIRMS) used prophylaxis with colchicine or NSAIDs (90 92), even though no randomized trials had assessed NSAIDs as a prophylactic therapy for urate-lowering therapy. In both FACT and APEX, the number of acute attacks spiked after discontinuation of prophylaxis at 8 weeks, with an approximate doubling of the proportion of patients reporting an attack (from 20% before to 40% after discontinuation). CONFIRMS, which continued prophylaxis for the entire 6 months of the trial, saw no spike in attacks. One randomized trial with high risk of bias compared different durations of colchicine prophylaxis in patients with gout initiating allopurinol therapy (59). However, the outcome measure was any evidence of recurrence of gouty arthritis, and the criteria for this clinical event and loss to follow-up were not specified. Wortmann and colleagues collected adverse event data from all 3 of the major trials of febuxostat and allopurinol and pooled data from FACT and APEX (56). None of these trials randomly assigned patients to different prophylaxis regimens; rather, selection bias was potentially present because assignment was at the discretion of the treating physician. In all 3 studies, upper respiratory infection was the most frequently reported adverse event (8% to 9% in each group, with no statistically significant difference). Overall, adverse events were higher with colchicine prophylaxis than with naproxen prophylaxis (55% vs. 44%). Diarrhea was about 3 times more common with colchicine than with naproxen (8.4% vs. 2.7%). In CONFIRMS, no statistically significant difference was seen in overall adverse events reported (about 55% in both groups), but gastrointestinal and abdominal pains were about 3 times more frequent in naproxen-treated patients (3.2% vs. 1.2%). Headache was more commonly reported in colchicine-treated patients (2.8% vs. 0.9%). In summary, high-strength evidence suggests that prophylaxis with either colchicine or NSAIDs reduces the risk for acute gout attacks in patients initiating 44 Annals of Internal Medicine Vol. 166 No. 1 3 January

9 Systematic Review in Support of ACP Guideline on Management of Gout REVIEW urate-lowering therapy. The optimal duration of such prophylactic therapy is unknown, but moderatestrength evidence suggests that it should be longer than 8 weeks. Treatment Monitoring of Patients With Gout A large body of evidence (which includes the fact that urate is soluble up to a concentration of about 404 μmol/l [6.8 mg/dl], above which precipitation may begin [101]) supports the hypothesis that lower serum urate levels are causally associated with a lower rate of acute gout attacks. A post hoc analysis that combined data from 3 large trials (ULT, FACT, and APEX) provided the best clinical data about the relationship between achieved urate levels and risk for acute gout attacks (102). That analysis included more than 1800 patients with gout and a baseline serum urate level of 476 μmol/l (8.0 mg/dl) or greater. The achieved urate level was 1 of 3 variables (along with baseline presence of tophi and percentage change in serum urate level from baseline) that were associated with acute gout attacks requiring treatment (adjusted odds ratio, 1.42 [95% CI, 1.16 to 1.73] and 2.70 [CI, 1.72 to 4.22] at 6 and 12 months after initiation of therapy, respectively). Patients who achieved a serum urate level less than 357 μmol/l (<6.0 mg/dl) at the end of 1 year had a risk for acute gout attacks of approximately 5%, whereas patients with levels at or above 357 μmol/l ( 6.0 mg/dl) had risks of about 10% to 15%. Other reports, including analyses from several retrospective cohort studies, support a relationship between achieved lower serum urate levels and reduced risk for acute gout attacks ( ). One study that reported outcomes for a cohort of 158 patients with incident gout who were followed for a mean of 13 years found that 70% had a flare during the extended follow-up and that higher urate levels were predictive of a subsequent acute flare (odds ratio, 1.35 [CI, 1.2 to 1.5]) (109). There are potential nonarticular impacts of hyperuricemia in patients with gout that may be affected by serum urate levels. One retrospective observational study of U.S. veterans with gout used the U.S. Department of Veterans Affairs data warehouse to follow 2116 patients for a mean of 6.5 years. Comparing patients with high versus low serum urate levels, the investigators reported about a 2-fold difference in new diagnoses of kidney disease (4% vs. 2% at year 1 and 9% vs. 5% at year 3, respectively) (110). A second study used U.S. claims data to identify pairs of patients with gout who were or were not treated with urate-lowering therapy and were followed for slightly more than a year. Cardiovascular events were common in both groups (incidence rates were 24% and 21%), but there were no statistically significant differences in the composite outcome or its components between groups (111). In summary, the only way to assess whether uratelowering therapy in patients with gout is having the desired effect on serum urate levels is by monitoring, but no direct evidence supports or refutes the value of such monitoring. The notion that decreasing serum urate levels will reduce the risk for acute gout attacks is supported by physiology and observational evidence. The evidence base for use of serum urate level as a target value for treatment is limited by the lack of any trial that has based treatment decisions on different specific targets (such as a target of 416 vs. 357 μmol/l [7.0 vs. 6.0 mg/dl]) or any target as opposed to treating symptoms. Treating to a target necessarily means increasing doses of medication in patients who may be asymptomatic. The value of such a strategy has yet to be proved, and examples exist from other studies using intermediary biomarkers (such as elevated blood pressure or blood glucose level or low hemoglobin level), in which treating to a target resulted in more adverse effects than benefits. Thus, despite the strong biologic appeal of such a strategy and its advocacy by major specialty society guidelines (99, 100, 112), we judged the strength of evidence for monitoring to be low. Discontinuation of Pharmaceutical Management for Patients Receiving Medications for Chronic Gout Three prospective observational cohort studies found that some patients who were currently being treated remained asymptomatic for up to 2 or 3 years after urate-lowering therapy was stopped and that serum urate levels in the presence or absence of treatment were strong predictors of which patients remained asymptomatic ( ). Because a strategy of treatment discontinuation has never been tested in a clinical trial, we judged the strength of evidence to be insufficient that criteria exist to identify patients for whom urate-lowering therapy can be safely discontinued. DISCUSSION A principal finding of this review is that highstrength evidence suggests that colchicine, NSAIDs, and corticosteroids help reduce symptoms in patients with acute gout (Table 3). Despite the limited number of placebo-controlled trials, we were able to reach strong conclusions due to specific features of gout, namely that symptoms result from an inflammatory reaction to the deposition of urate crystals, which occurs when the level of urate increases above its saturation point in the blood. Hence, in an era that predated the widespread practice of placebo-controlled trial testing of therapies, clinicians prescribed anti-inflammatory medications to treat symptoms of acute gout. Corticosteroids are among the most powerful and effective anti-inflammatory medications available. Although no randomized, placebo-controlled trials have tested their use in acute gout, corticosteroids have proven efficacy in other inflammatory conditions, which increases our confidence that they are effective in treating the inflammatory reaction in acute gout. With regard to the management of hyperuricemia in patients with gout, we likewise used direct and indirect evidence to reach conclusions about the effectiveness of urate-lowering therapy at reducing the risk for acute Annals of Internal Medicine Vol. 166 No. 1 3 January

10 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout Table 3. Summary of Prior Knowledge, Findings From the Systematic Review, and Strength of Evidence, by Key Question Key Question Prior Knowledge Used in Determining Strength of Evidence Sources of Evidence Included in This Systematic Review Acute gout treatment Colchicine reduces pain NA 2 placebo-controlled RCTs (n =45and 184), both with low risk of bias Low-dose colchicine is as effective as higher-dose colchicine for reducing pain, with fewer adverse effects NSAIDs reduce gout pain NA Biologic rationale (anti-inflammatory action) Placebo-controlled RCT evidence that NSAIDs provide temporary pain relief for many conditions 1 head-to-head RCT with low risk of bias (n = 184) Strength of Evidence High Moderate 1 placebo-controlled RCT with high risk High of bias (n = 30) High-strength observational data (NSAID use as prophylaxis against gout flare; see Management of hyperuricemia in patients with gout section in this table) 16 head-to-head RCTs Moderate No difference between NSAIDs in effectiveness Equivalence in effectiveness among NSAIDs in many other conditions Systemic corticosteroids reduce pain Biologic rationale (anti-inflammatory action) No placebo-controlled RCTs Equivalence to NSAIDs in 6 RCTs (n = 27, 90, 120, 60, 95, and 416); 4 of 6 RCTs had low risk of bias Animal-derived corticotropin formulation reduces pain Biologic rationale (anti-inflammatory action) No placebo-controlled RCTs Approximate equivalence to NSAIDs and intramuscular corticosteroids in RCTs (1 RCT of each, n = 76 and 31, both with high risk of bias) High Moderate Management of hyperuricemia in patients with gout ULT does not reduce the risk for acute gout attacks within the first 6 mo NA 2 placebo-controlled RCTs with low risk of bias (n = 1072 and 57) ULT reduces serum urate levels NA 4 placebo-controlled RCTs, all with low risk of bias (n = 1072, 96, 153, and 57) ULT reduces the risk for acute gout attacks after 1 y Prophylactic therapy with low-dose colchicine or low-dose NSAIDs reduces the risk for acute gout attacks in patients initiating ULT Longer durations of prophylaxis with colchicine or NSAIDs (>8 wk) are more effective than shorter durations in patients initiating ULT Acute gout attacks are caused by elevated serum urate concentrations NA NA No placebo-controlled RCTs assess long-term risk for acute gout attacks RCTs with low risk of bias show that ULT reduces serum urate levels Open-label extension study of a ULT RCT shows reduced risk for acute gout attacks over time (<5% at approximately 1 y) 1 placebo-controlled RCT of colchicine with low risk of bias (n = 43) Strong observational evidence across 3 RCTs with low risk of bias that included different durations of prophylaxis (n = 762, 2269, and 1072) Indirect evidence from comparisons across 3 RCTs of differing durations of prophylaxis 1 RCT with high risk of bias (n = 190) High High Moderate High Moderate Treatment monitoring Monitoring serum urate levels improves outcomes Treating to a specific target serum urate level reduces the risk for gout attacks NA Lower serum urate levels are associated with reduced risk for gout attacks No direct evidence An argument can be made indirectly on the basis of the evidence that elevated serum urate levels cause gout No RCT evidence Variable targets proposed or assessed in the literature NA = not applicable; NSAID = nonsteroidal anti-inflammatory drug; RCT = randomized, controlled trial; ULT = urate-lowering therapy. Insufficient Low gout attacks over time, despite the fact that this outcome has not been studied in any placebo-controlled trial lasting longer than a few months (Table 3). We based our rating of moderate strength of evidence on the high-strength evidence that urate-lowering therapy reduces serum urate levels, that serum urate level is a strong predictor of risk for acute gout attacks, and that the open-label extension studies of urate-lowering therapy trials have shown a graded relationship between the serum urate level achieved and the risk for acute gout attacks. We concluded that after the initial period of increased risk for acute gout attacks associ- 46 Annals of Internal Medicine Vol. 166 No. 1 3 January

11 Systematic Review in Support of ACP Guideline on Management of Gout REVIEW ated with initiation of urate-lowering therapy, such therapy reduces the risk for acute attacks. The increased risk during the initial period can be reduced with concurrent use of anti-inflammatory prophylaxis. Our review reached stronger strength-of-evidence ratings than some prior reviews, particularly Cochrane reviews, that concluded that the strength of evidence is inconclusive with regard to corticosteroids (116), low-quality with regard to colchicine (117), limited with regard to NSAIDs (118), and low-moderate with regard to allopurinol (44). Non-Cochrane reviews tended to reach stronger conclusions (for example, all therapies were found to be effective [119]). The most important question to be answered for the treatment of gout in primary care is whether using urate-lowering therapy to treat all patients to a specific goal, such as less than 357 μmol/l (<6.0 mg/dl), improves patient outcomes. This concept of treat-totarget is supported by indirect evidence but has not been tested. Guidelines and recommendations about treatment target thresholds vary; for example, less than 357 μmol/l (<6.0 mg/dl) is the most commonly suggested threshold, but recent American College of Rheumatology guidelines suggest a threshold of less than 297 μmol/l (<5.0 mg/dl) for some patients to improve signs and symptoms of gout (100). However, for many patients whose gout is clinically wellcontrolled (no flares) during urate-lowering therapy, no direct data support such targets. In fact, the results of one cohort study suggest that once gout has been asymptomatic for 5 years, urate-lowering therapy might be discontinued for many years as long as serum urate levels remain acceptable (for example, <416 μmol/l [<7 mg/dl]) (114). Our review identified several limitations beyond the lack of placebo-controlled studies. For many of the questions of interest, data were not reported on the subgroups or outcomes of interest. Only 10 studies explicitly stated that patients came exclusively from primary care sites or included such patients. We thus judged this evidence to be moderately applicable to primary care. In summary, use of colchicine, NSAIDs, and corticosteroids to relieve symptoms of acute gout attacks is supported by high-strength evidence. In addition, highstrength evidence suggests that urate-lowering therapy with allopurinol or febuxostat decreases serum urate levels, and moderate-strength evidence shows that it reduces the risk for acute gout attacks after several months. In patients initiating urate-lowering therapy, high-strength evidence supports use of prophylactic colchicine or NSAIDs for more than 8 weeks to reduce the risk for acute gout attacks. Use of a treat-to-target strategy for patients with gout and hyperuricemia has logical appeal and is supported by observational evidence but has not been experimentally tested. A longterm trial of a treat-to-target strategy is the most important research need for the management of gout in primary care. From RAND Corporation, Santa Monica, and Veterans Affairs Greater Los Angeles Healthcare System and David Geffen School of Medicine at the University of California, Los Angeles, Los Angeles, California. Disclaimer: The authors of this manuscript are responsible for its content. Statements in the manuscript should not be construed as endorsement by the Agency for Healthcare Research and Quality (AHRQ) or the U.S. Department of Health and Human Services. AHRQ retains a license to display, reproduce, and distribute the data and the report from which this manuscript was derived under the terms of the agency's contract with the author. Acknowledgment: The authors thank Patricia Smith for her administrative assistance on the project and acknowledge the guidance provided by the key informants and technical expert panel members on the evidence report. Financial Support: This project was funded under contract I from the AHRQ, U.S. Department of Health and Human Services. Disclosures: All authors report a grant from AHRQ during the conduct of the study. Dr. Shekelle reports personal fees from ECRI Institute outside the submitted work and royalties from UpToDate. Authors not named here have disclosed no other conflicts of interest. Disclosures can also be viewed at Reproducible Research Statement: Study protocol: Available at /Gout-managment-protocol pdf. Statistical code: Not applicable. Data set: Available in the full report at effectivehealthcare.ahrq.gov/search-for-guides-reviews-andreports/?pageaction=displayproduct&productid=2195. Requests for Single Reprints: Paul G. Shekelle, MD, PhD, RAND Corporation, 1776 Main Street, Santa Monica, CA 90401; , shekelle@rand.org. Current author addresses and author contributions are available at References 1. Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population: the National Health and Nutrition Examination Survey Arthritis Rheum. 2011;63: [PMID: ] doi: /art Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151: [PMID: ] doi: / Shekelle PG, Fitzgerald J, Newberry SJ, Motala A, O Hanlon CE, Okunogbe A, et al. Management of Gout. Comparative Effectiveness Review no (Prepared by the Southern California Evidencebased Practice Center under contract no I.) AHRQ publication no.16-ehc017-ef. Rockville: Agency for Healthcare Research and Quality; Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al; Cochrane Bias Methods Group. The Cochrane Collabora- Annals of Internal Medicine Vol. 166 No. 1 3 January

12 REVIEW Systematic Review in Support of ACP Guideline on Management of Gout tion's tool for assessing risk of bias in randomised trials. BMJ. 2011; 343:d5928. [PMID: ] doi: /bmj.d Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol. 2007;7:10. [PMID: ] 6. Agency for Healthcare Research and Quality. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. AHRQ publication no. 10(14)-EHC063-EF. Rockville: Agency for Healthcare Research and Quality; Hill AB. The environment and disease: association or causation? Proc R Soc Med. 1965;58: [PMID: ] 8. Maccagno A, Di Giorgio E, Romanowicz A. Effectiveness of etodolac ( Lodine ) compared with naproxen in patients with acute gout. Curr Med Res Opin. 1991;12: [PMID: ] 9. Schumacher HR, Berger MF, Li-Yu J, Perez-Ruiz F, Burgos-Vargas R, Li C. Efficacy and tolerability of celecoxib in the treatment of acute gouty arthritis: a randomized controlled trial. J Rheumatol. 2012;39: [PMID: ] doi: /jrheum Janssens HJ, Janssen M, van de Lisdonk EH, van Riel PL, van Weel C. Use of oral prednisolone or naproxen for the treatment of gout arthritis: a double-blind, randomised equivalence trial. Lancet. 2008;371: [PMID: ] doi: /s (08) Li T, Chen SL, Dai Q, Han XH, Li ZG, Wu DH, et al. Etoricoxib versus indomethacin in the treatment of Chinese patients with acute gouty arthritis: a randomized double-blind trial. Chin Med J (Engl). 2013;126: [PMID: ] 12. Terkeltaub RA, Furst DE, Bennett K, Kook KA, Crockett RS, Davis MW. High versus low dosing of oral colchicine for early acute gout flare: twenty-four-hour outcome of the first multicenter, randomized, double-blind, placebo-controlled, parallel-group, dose-comparison colchicine study. Arthritis Rheum. 2010;62: [PMID: ] doi: /art Zhang YK, Yang H, Zhang JY, Song LJ, Fan YC. Comparison of intramuscular compound betamethasone and oral diclofenac sodium in the treatment of acute attacks of gout. Int J Clin Pract. 2014; 68: [PMID: ] doi: /ijcp Ahern MJ, Reid C, Gordon TP, McCredie M, Brooks PM, Jones M. Does colchicine work? The results of the first controlled study in acute gout. Aust N Z J Med. 1987;17: [PMID: ] 15. Cheng TT, Lai HM, Chiu CK, Chem YC. A single-blind, randomized, controlled trial to assess the efficacy and tolerability of rofecoxib, diclofenac sodium, and meloxicam in patients with acute gouty arthritis. Clin Ther. 2004;26: [PMID: ] 16. Shrestha M, Morgan DL, Moreden JM, Singh R, Nelson M, Hayes JE. Randomized double-blind comparison of the analgesic efficacy of intramuscular ketorolac and oral indomethacin in the treatment of acute gouty arthritis. Ann Emerg Med. 1995;26: [PMID: ] 17. Altman RD, Honig S, Levin JM, Lightfoot RW. Ketoprofen versus indomethacin in patients with acute gouty arthritis: a multicenter, double blind comparative study. J Rheumatol. 1988;15: [PMID: ] 18. Eberl R, Dunky A. Meclofenamate sodium in the treatment of acute gout. Results of a double-blind study. Arzneimittelforschung. 1983;33: [PMID: ] 19. Ruotsi A, Vainio U. Treatment of acute gouty arthritis with proquazone and indomethacin. A comparative, double-blind trial. Scand J Rheumatol Suppl. 1978:15-7. [PMID: ] 20. Alloway JA, Moriarty MJ, Hoogland YT, Nashel DJ. Comparison of triamcinolone acetonide with indomethacin in the treatment of acute gouty arthritis. J Rheumatol. 1993;20: [PMID: ] 21. Lederman R. A double-blind comparison of etodolac and high doses of naproxen in the treatment of acute gout. Adv Ther. 1990; 7: Schumacher HR Jr, Boice JA, Daikh DI, Mukhopadhyay S, Malmstrom K, Ng J, et al. Randomised double blind trial of etoricoxib and indomethacin in treatment of acute gouty arthritis. BMJ. 2002;324: [PMID: ] 23. Rubin BR, Burton R, Navarra S, Antigua J, Londoño J, Pryhuber KG, et al. Efficacy and safety profile of treatment with etoricoxib 120 mg once daily compared with indomethacin 50 mg three times daily in acute gout: a randomized controlled trial. Arthritis Rheum. 2004; 50: [PMID: ] 24. Man CY, Cheung IT, Cameron PA, Rainer TH. Comparison of oral prednisolone/paracetamol and oral indomethacin/paracetamol combination therapy in the treatment of acute goutlike arthritis: a double-blind, randomized, controlled trial. Ann Emerg Med. 2007; 49: [PMID: ] 25. Axelrod D, Preston S. Comparison of parenteral adrenocorticotropic hormone with oral indomethacin in the treatment of acute gout. Arthritis Rheum. 1988;31: [PMID: ] 26. Siegel LB, Alloway JA, Nashel DJ. Comparison of adrenocorticotropic hormone and triamcinolone acetonide in the treatment of acute gouty arthritis. J Rheumatol. 1994;21: [PMID: ] 27. Weiner GI, White SR, Weitzner RI, Rubinstein HM. Double-blind study of fenoprofen versus phenylbutazone in acute gouty arthritis. Arthritis Rheum. 1979;22: [PMID: ] 28. Douglas G, Thompson M. A comparison of phenylbutazone and flufenamic acid in the treatment of acute gout. Ann Phys Med. 1970; 10: [PMID: ] 29. Butler RC, Goddard DH, Higgens CS, Hollingworth P, Pease CT, Stodell MA, et al. Double-blind trial of flurbiprofen and phenylbutazone in acute gouty arthritis. Br J Clin Pharmacol. 1985;20: [PMID: ] 30. de la Torre G. A comparative, double-blind, parallel study with tenoxicam vs placebo in acute in acute gouty arthritis. Invest Med Int. 1987;14: Lomen PL, Turner LF, Lamborn KR, Winblad MA, Sack RL, Brinn EL. Flurbiprofen in the treatment of acute gout. A comparison with indomethacin. Am J Med. 1986;80: [PMID: ] 32. Siegmeth W, Placheta P. [Double-blind trial: ketoprofen versus phenylbutazone in acute gouty arthritis (author's transl)]. Wien Klin Wochenschr. 1976;88: [PMID: ] 33. Liu Y, Li ZC, Chen JB, Yu TT, Cui C, Zhang HB. Therapeutic efficacy of small doses of colchicine combined with glucocorticoid for acute gouty arthritis [in Chinese]. Medical Journal of Chinese People's Liberation Army. 2015; Rainer TH, Cheng CH, Janssens HJ, Man CY, Tam LS, Choi YF, et al. Oral prednisolone in the treatment of acute gout: a pragmatic, multicenter, double-blind, randomized trial. Ann Intern Med. 2016; 164: [PMID: ] doi: /m Zhang S, Zhang Y, Liu P, Zhang W, Ma JL, Wang J. Efficacy and safety of etoricoxib compared with NSAIDs in acute gout: a systematic review and a meta-analysis. Clin Rheumatol. 2016;35: [PMID: ] doi: /s Geronikolou SA. Treatment of gout in a recently published 9th century manuscript of Rhazes. Vesalius. 2014;20:95-8. [PMID: ] 37. Manara M, Bortoluzzi A, Favero M, Prevete I, Scirè CA, Bianchi G, et al. Italian Society of Rheumatology recommendations for the management of gout. Reumatismo. 2013;65:4-21. [PMID: ] doi: /reumatismo Tayar JH, Lopez-Olivo MA, Suarez-Almazor ME. Febuxostat for treating chronic gout. Cochrane Database Syst Rev. 2012;11: CD [PMID: ] doi: / cd pub2 39. Faruque LI, Ehteshami-Afshar A, Wiebe N, Tjosvold L, Homik J, Tonelli M. A systematic review and meta-analysis on the safety and efficacy of febuxostat versus allopurinol in chronic gout. Semin Arthritis Rheum. 2013;43: [PMID: ] doi: /j.semarthrit Latourte A, Bardin T, Richette P. Prophylaxis for acute gout flares after initiation of urate-lowering therapy. Rheumatology (Oxford). 2014;53: [PMID: ] doi: /rheumatology /keu Ramasamy SN, Korb-Wells CS, Kannangara DR, Smith MW, Wang N, Roberts DM, et al. Allopurinol hypersensitivity: a systematic 48 Annals of Internal Medicine Vol. 166 No. 1 3 January