Initiation of Allopurinol at First Medical Contact for Acute Attacks of Gout: A Randomized Clinical Trial

BRIEF OBSERVATION Initiation of Allopurinol at First Medical Contact for Acute Attacks of Gout: A Randomized Clinical Trial Thomas H. Taylor, MD, a,b,c John N. Mecchella, DO, b,c Robin J. Larson, MD, a,b Kevin D. Kerin, MD, a,b Todd A. MacKenzie, PhD b a White River Junction VA Regional Medical Center, White River Junction, Vt; b Dartmouth Medical School, Lebanon, NH; c Dartmouth Hitchcock Medical Center, Lebanon, NH. ABSTRACT OBJECTIVE: Streamlining the initiation of allopurinol could result in a cost benefit for a common medical problem and obviate the perception that no treatment is required once acute attacks have resolved. Our objective was to test the hypothesis that there is no difference in patient daily pain or subsequent attacks with early versus delayed initiation of allopurinol for an acute gout attack. METHODS: A total of 57 men with crystal-proven gout were randomized to allopurinol 300 mg daily or matching placebo for 10 days. All subjects received indomethacin 50 mg 3 times per day for 10 days, a prophylactic dose of colchicine 0.6 mg 2 times per day for 90 days, and open-label allopurinol starting at day 11. Primary outcome measures were pain on visual analogue scale (VAS) for the primary joint on days 1 to 10 and self-reported flares in any joint through day 30. RESULTS: On the basis of 51 evaluable subjects (allopurinol in 26, placebo in 25), mean daily VAS pain scores did not differ significantly between study groups at any point between days 1 and 10. Initial VAS pain scores for allopurinol and placebo arms were 6.72 versus 6.28 (P.37), declining to 0.18 versus 0.27 (P.54) at day 10, with neither group consistently having more daily pain. Subsequent flares occurred in 2 subjects taking allopurinol and 3 subjects taking placebo (P.60). Although urate levels decreased rapidly in the allopurinol group (from 7.8 mg/dl at baseline to 5.9 mg/dl at day 3), sedimentation rates and C-reactive protein levels did not differ between groups at any point. CONCLUSIONS: Allopurinol initiation during an acute gout attack caused no significant difference in daily pain, recurrent flares, or inflammatory markers. Published by Elsevier Inc. The American Journal of Medicine (2012) 125, 1126-1134 KEYWORDS: Allopurinol; Gout; Gout outcomes Funding: none. Conflict of Interest: None. The views expressed herein do not necessarily represent the views of the Department of Veterans Affairs or the US Government. Authorship: All authors had access to the data and played a role in writing this manuscript. Reproducible Research Statement: Research study protocol, STATA statistical code, and data set, without personal identifiers, available from Thomas H. Taylor (e-mail: Tom.Taylor@va.gov). This study was reviewed by the Veterans Affairs Research and Development Committee and received approval from the Dartmouth Medical School Institutional Review Board. ClinicalTrials.gov registration number: NCT01310673. Requests for reprints should be addressed to Thomas H. Taylor, MD, White River Jct VA Regional Medical Center, Department of Medicine, Dartmouth Medical School, 215 N. Main St, White River Junction, VT 05009. E-mail address: Tom.Taylor@va.gov. Medical teaching suggests that allopurinol should not be initiated in the setting of an acute gout attack, because rapid lowering of serum urate may exacerbate the attack. Delayed initiation may come at a price, as many patients never start definitive urate-lowering therapy and are skeptical of chronic therapy after acute symptoms resolve. Since the introduction of allopurinol in 1964, reviews have attributed acute attacks of gout or worsening of ongoing attacks to the initiation of allopurinol. 1 Inference may be implied, because gout attacks continue to occur during the first few months after allopurinol is started 2 and are proportional to the rate of uric acid lowering. 3,4 Theories to explain allopurinol-induced exacerbation of gout are controversial and lack evidence, but they generally implicate urate concentration flux and remodeling of microscopic tophi. 3,5 0002-9343/$ -see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.amjmed.2012.05.025

Taylor et al Allopurinol in Acute Gout Attacks 1127 Recommendations have come to include complex guidelines for delayed and incremental initiation of definitive treatment, commencing after the acute attack has subsided, for both allopurinol and uricosurics. 6,7 Despite these well-intentioned guidelines, recent studies have implicated poor compliance, deficits in patient and physician CLINICAL SIGNIFICANCE knowledge, and complexity of present regimens requiring patients to return for graded increases of allopurinol as factors impeding better outcomes. 8-13 If initiation of allopurinol could be simplified and administered in an adequate dose of 300 mg at the first medical encounter during an acute attack, then opportunity for education, improved outcomes, and cost containment might be realized. Evidence in support of delayed and stepped increase of allopurinol treatment is poor and supported by 2 studies describing case series. 14,15 No study has evaluated the initiation of allopurinol during attacks in patients with primary gout as they present in primary care settings, concomitantly treated with both indomethacin and colchicine. Prophylactic colchicine has been shown to reduce the frequency of gout flares in patients with interval gout 16 and in patients beginning treatment with uricosurics 17 and allopurinol. 18 Should allopurinol predispose to exacerbation, it is equally conceivable that the best time to initiate allopurinol would be during the acute attack, when patients also receive treatment doses of indomethacin and prophylactic doses of colchicine. We challenge current teaching by testing the hypothesis that there is no difference in patient-reported pain or subsequent attacks with early initiation of allopurinol given during the acute attack, concomitant with indomethacin treatment and a prophylactic dose of colchicine. MATERIALS AND METHODS We designed a randomized, double-blind, placebo-controlled, parallel-arm, single-center, noninferiority study of the early initiation of full-dose allopurinol (300 mg) versus placebo in adults with acute gout, both arms receiving indomethacin treatment and a prophylactic dose of colchicine. The Research and Development Committee at the White River Junction Veteran s Affairs Medical Center and the Committee for the Protection of Human Subjects at Dartmouth College approved the original protocol and reviewed the trial annually. All subjects provided written informed consent. There was no pharmaceutical industry participation or support. Guidelines recommend delayed and graded initiation of allopurinol. Poor outcomes for gout treatment might be improved if allopurinol could be initiated at the initial presentation of acute attacks. No difference in pain by daily visual analogue scale was seen when allopurinol was administered during the acute attack. The percentage of recurrent attacks was similar to that in studies in which febuxostat, allopurinol, or placebo was initiated after the attack. The trial was conducted between 1998 and 2009 at the Veteran s Affairs Medical Center in White River Junction, Vermont. Patients presenting within 7 days of onset of an acute gout attack were evaluated, and American College of Rheumatology criteria for acute arthritis of gout were met, 19 including the presence of monosodium urate crystals on arthrocentesis of the primary joint on the day of study entry. Exclusion criteria included secondary gout (because it is dependent on the treatment of the underlying disease); the presence of tophaceous gout (because of concern that tophi could make evaluation of resolution and exacerbations difficult); a history of congestive heart failure; anticoagulant use; a recent serum creatinine greater than 1.3 mg/dl (because these patients should not receive indomethacin); or the use of steroids, colchicine, allopurinol, uricosuric drugs, chemotherapy, or immunosuppressive therapy in the past 6 months. Although all subjects brought to the attention of the principal investigator were screened consecutively, primary providers also made decisions regarding eligibility and subjects were highly selected by study criteria; thus, information regarding the number and characteristics of those excluded could not be reliably tracked. Randomization and Interventions Fifty-seven eligible subjects were randomized, in a 1:1 ratio without stratification, to receive allopurinol 300 mg daily for 10 days (allopurinol group) or placebo for 10 days (placebo group). Subjects and evaluators had no access to the randomization sequence. The randomization sequence was determined by the study pharmacist using a random number generator and kept in the pharmacy vault. Study drugs were given directly to study patients through the pharmacy, and neither patients nor evaluators were aware of which medication was given. In addition to the 10-day course of allopurinol or placebo, all patients received indomethacin 50 mg 3 times per day for 10 days and colchicine 0.6 mg 2 times per day for 90 days. All patients were started on open-label allopurinol 300 mg daily on day 11 and followed for 30 days. Outcomes and Follow-up Two primary outcomes included pain scores at each of days 1 to 10, as measured by a visual analogue scale (VAS) standardized to 10 cm for the primary affected joint, and self-reported subsequent gout flares in any joint during days 1 to 30. For VAS scoring, subjects were given a diary in

1128 The American Journal of Medicine, Vol 125, No 11, November 2012 which they indicated their daily pain level on lines anchored by no pain at all and worst pain imaginable. For subsequent gout flares, subjects recorded events in their diary and were asked at each follow-up visit whether they had any new or recurrent gout flares since their last visit. Clinical confirmation of such events was not required because of concern for underreporting. Secondary outcomes included erythrocyte sedimentation rates, and C-reactive protein levels were added when they became available in 2005. Complete blood counts, liver function tests, and creatinine levels were followed to monitor for toxicity. Serum urate levels were followed to confirm compliance and to document the expected early rapid decline with allopurinol initiation. Follow-up visits for all outcomes occurred on days 3, 10, and 30 plus or minus 3 days to accommodate weekends or conflicts. Sample Size In noninferiority terms, there is 92% power to detect an inferiority margin of 2.0 cm assuming an actual inferiority margin of 0.5 cm, based on a 5% 1-sided type I error rate. Because there were no studies using the VAS in acute gout, we chose 2.0 cm as a clinically relevant difference. This is consistent with the reproducibility of the VAS in patients with rheumatoid arthritis, 1.5 cm. 20 If this were a superiority trial, 57 subjects provided 90% power to detect a 1.5-cm difference in mean VAS pain scores with a standard deviation of 1.8 cm using a 2-sided P value of.05. Statistical Analysis All outcome measures were prespecified and prospectively collected. The final analysis plan was determined after completion of the trial but before reviewing the data. We compared baseline characteristics between study arms using the t test for continuous variables and the chi-square test for dichotomous variables. Because no minimally important difference in VAS pain scores has been established in acute gout, 21 we asserted that a 2-cm difference in 10-cm VAS pain scores would be clinically relevant in our study and assessed whether the 95% confidence intervals (CIs) around the differences in mean VAS pain scores for days 1 to 10 contained this value. In addition, we assessed for statistically significant differences in VAS pain scores at each of days 1 to 10 and performed a longitudinal analysis in which VAS, measured at days 2 to 10, was compared between the 2 study arms while adjusting for VAS on day 1 using a linear mixed-effects model with fixed effects for study arm, day, baseline VAS (day 1), and random intercept for each subject. This model was run with and without an interaction of study arm and day (ie, different slopes for both study arms). In addition, we considered a model that also had a random slope for each patient. Subgroup and Sensitivity Analysis Because first attacks of gout may respond differently to treatment than subsequent attacks, we performed a post hoc subgroup analysis in which we repeated each of the VAS pain score comparisons according to whether the Figure 1 Flow diagram of enrolled subjects.

Taylor et al Allopurinol in Acute Gout Attacks 1129 subjects were having their first gout attack or had had previous attacks. Because intention-to-treat analysis can bias results toward falsely underestimating treatment effects, which would favor our hypothesis of no difference in daily pain or flares, we chose to report our per-protocol analysis (restricted to randomized subjects who were adherent to study treatment and completed 10-day and 30-day follow-ups) as the primary result. However, to also address the possibility that subjects with incomplete data or follow-up were not missing at random, we performed 3 sensitivity analyses for our VAS pain score outcome an intention-to-treat analysis based on available data, an intention-to-treat analysis substituting missing values with the mean VAS score of the subjects in the same arm who reported the outcome for that day, and an intention-to-treat analysis substituting missing values with the highest VAS score reported by another subject for that day. To evaluate the effect of compliance, we performed a compliance-adjusted analysis, 22 in which we used the mean serum urate level measured at days 3 and 10 (both post-randomization) as a measure of compliance. We derived a compliance-adjusted treatment effect using the study arm as the instrumental variable. 23 RESULTS Of 57 subjects randomized, 31 were allocated to the allopurinol group and 26 were allocated to the placebo group. For the per-protocol analysis, 6 subjects were excluded post-randomization because of failure to adequately comply with the study medication or follow-up visits (Figure 1). The baseline characteristics of the participants were similar between study arms for both the intention-to-treat (Supplemental Table 1) and per-protocol populations (Table 1), neither of whom contained any statistically significant differences. For the per-protocol population, the mean duration of incident attack before entry was 4.2 days for the allopurinol group and 3.9 days for the placebo group. Nine of the patients in the allopurinol group and 4 patients in the placebo group presented with their first attack of gout. Compliance During the placebo-controlled portion of the study (days 1-10), serum urate levels decreased rapidly in the allopurinol group, reaching 6.5 mg/dl by day 10 for all but 1 of the per-protocol subjects (Figure 2). Conversely, urate levels remained elevated in the placebo group until day 11 when a similar rapid decline was observed after initiation of open-label allopurinol in all patients. Pain on Daily Visual Analogue Scale On the basis of per-protocol analysis, initial mean VAS pain scores for the allopurinol and placebo groups were 6.72 versus 6.28 (P.37) decreasing to 0.18 versus 0.27 (P.54) at day 10 (Figure 3). The largest difference in mean VAS pain scores between groups was 0.44 cm on the day of randomization, and the largest value included Table 1 Characteristic Characteristics of Study Participants Allopurinol Group (n 26) Placebo Group (n 25) P Value Mean age (SD), y 57 (14) 61 (11).23 Men, n (%) 26 (100) 25 (100) History of diabetes, n (%) 4 (15) 5 (20).67 Hypertension, n (%) 15 (58) 19 (76).17 Hyperlipidemia, n (%) 17 (65) 14 (56).49 Mean BMI (SD), kg/m 2 32 (5) 32 (6).79 Current alcohol use, n (%) 7 (27) 10 (40).32 Use of diuretics, n (%) 5 (19) 7 (28).46 Mean entry creatinine 1.1 (0.18) 1.1 (0.21).77 (SD), mg/dl Mean entry urate (SD), 7.8 (1.12) 7.6 (1.72).76 mg/dl Mean age of first attack 53.9 (16.9) 54.5 (11.5).87 (SD), y* Attack No., n (%).12 First attack 9 (35) 4 (16) 2-9 attacks 14 (54) 13 (52) 10 attacks 3 (12) 8 (32) Study joint, n (%).41 Joints of the hand 2 (8) 2 (8) Wrist 1 (4) 2 (8) Elbow 2 (8) 0 (0) Knee 5 (19) 5 (20) Ankle 7 (27) 4 (16) MTP 9 (35) 9 (36) Multiple joints 0 (0) 3 (12) BMI body mass index; MTP metatarsophalangeal; SD standard deviation. *n 25 for allopurinol group and n 23 for placebo group. Categoric P values are listed for categoric variables by chi-square analysis. in the 95% CIs was 1.55 cm seen on day 2. Mean VAS pain scores did not statistically significantly differ between study groups at any point between days 1 and 10, and longitudinal analysis found that VAS pain across days 2 to 10, adjusting for baseline VAS and restricting to identical slopes between the allopurinol and placebo study arms, was not different ( 0.16 cm; 95% CI, 0.50 to 0.83; P.62). All 3 intention-to-treat sensitivity analyses showed similar findings, with the maximum difference in mean VAS pain scores between the allopurinol and placebo groups always occurring on the day of randomization and never exceeding 0.62 cm (the value from the extreme assumptions analysis). Likewise, the largest value contained in the 95% CIs was 1.63 cm on day 2 (the value based on extreme assumptions), and there were no statistically significant differences at any individual time point or longitudinally. Compliance-adjusted analysis showed that subjects receiving allopurinol had lower VAS pain scores, but the differences were not significant (P.10). Subgroup analysis

1130 The American Journal of Medicine, Vol 125, No 11, November 2012 Figure 2 Mean serum urate over study period. Error bars represent 95% CIs. according to whether the subject was having a first gout attack versus having had prior attacks revealed similar small, nonsignificant differences (Supplemental Appendix). Gout Flares The rate of new or recurrent gout flares between days 1 and 30 was 2 of 26 (7.7%) in the allopurinol group and 3 of 25 (12.0%) in the placebo group (P.61). Stratified by the timing of the events, a single gout flare occurred between days 1 and 10 in a subject in the allopurinol group at day 8, whereas 4 gout flares occurred between days 11 and 30, 1 in a subject in the allopurinol group at day 30 and 3 in subjects in the placebo group at days 16, 20, and 30. All flares were at least 5 days removed from the initiation of allopurinol, and none involved exacerbation of the index joint. Only 1 flare resulted in the patient seeking care. Subgroup analysis was not performed because all flares occurred in subjects with prior attacks. Secondary End Points Mean erythrocyte sedimentation rate and C-reactive protein levels declined over the study period in both groups and were not statistically significantly different at any point (Figure 4). Adverse Events Among per-protocol subjects, elevation of serum creatinine 1.5 mg/dl occurred in 1 subject from each study arm. Colchicine reductions due to gastrointestinal symptoms occurred in 8 subjects (31%) in the allopurinol group and 12 subjects (48%) in the placebo group. There was 1 unexpected death in an 80-year-old subject in the allopurinol group who developed gastroenteritis, pneumonia, fever, dehydration, and acute renal failure. Support was withdrawn in accordance with his advanced directives. He had taken 4 doses of study medication and is not included in the per-protocol analysis. Another patient in the placebo group had a hypersensitivity reaction with rash, fever, and mild transaminitis, leading to discontinuation of allopurinol at day 30. All data for this subject were collected and included in both per-protocol and intention-to-treat analyses. DISCUSSION This first randomized, double-blind, placebo-controlled study of patients started on an adequate dose (300 mg) of allopurinol during the acute gout attack, while simultaneously treated with indomethacin and colchicine, produced surprisingly similar declines in VAS and selfdetermined gout flares, with narrow CIs. Although an adequate dose of allopurinol is whatever dose is required to achieve a serum urate below the goal of 6.0 mg/ dl, 24,25 a starting dose of 300 mg of allopurinol avoids the complex stepped increments recommended by guidelines. 6,7 Secondary outcomes, rate of erythrocyte sedi-

Taylor et al Allopurinol in Acute Gout Attacks 1131 Figure 3 Mean VAS scores on days 1 to 10 for allopurinol versus placebo. Error bars represent 95% CI. The longitudinal analysis found that the VAS in the allopurinol arm was lower by 0.16 cm (95% CI, 0.50 to 0.83; P.62) across days 2 to 10, adjusting for baseline VAS and restricting to identical slopes between the 2 study arms. CI confidence interval; VAS visual analogue scale. mentation rate and C-reactive protein decline, did not significantly differ between immediate versus delayed groups. All 26 patients in the immediate allopurinol treatment group demonstrated a rapid decrease in serum urate level, testament to a high degree of compliance and adequate urate flux to test the hypothesis. Per-protocol analysis was chosen over intention-totreat analysis, although we report both. Because a rapid decrease in serum urate is considered to be an explanation for allopurinol-induced precipitation of gout attacks, we evaluated patients who actually took allopurinol and returned for 10- and 30-day visits. Flux in serum urate levels, as occurred in recent febuxostat studies, when a potent inhibitor of xanthine oxidase caused a precipitous decrease in uric acid, exemplifies the association of rapidly lowering urate with more frequent attacks of gout. 26 Urate flux as a sole explanation seems unlikely, because many patients with asymptomatic hyperuricemia never develop gout. 27 Dialysis rapidly decreases serum urate levels but is not associated with attacks of gout. Varying concentrations of urate crystals can be injected into murin knee joints without causing gout. 28 Urate crystals are seen in asymptomatic intercritical gout joints and other asymptomatic joints. 29,30 Urate or calcium pyrophosphate crystals can be mixed with monocytes or macrophages, in vitro, without induction of cytokines; inflammatory cytokine induction in this model first requires cell priming to activate the inflammasome platform. 28 Thus, urate flux, as an instigator of the acute gout attack, requires adjunct events, 31 possibly modifiable with adjunct therapy. The number of self-reported acute attacks was low and similar in both groups. All attacks were in joints other than the primary joint, and only 1 was severe enough to seek treatment. The rate of recurrent gout attack was 10% over 1 month for the entire group, 8% in the allopurinol group, and 12% in the placebo group. This 1-month attack rate compares with that in recent febuxostat studies: Febuxostat Versus Allopurinol Control Trial (FACT) trial, 3 Allopurinol and Placebo-Controlled, Efficacy Study of Febuxostat (APEX) trial, 32 Efficacy and Safety of Oral Febuxostat in Participants With Gout (CONFIRMS) trial, 33 and a phase II dose study 26 (Table 2). All patients in these comparison groups also were taking prophylactic colchicine or nonsteroidal anti-inflammatory drugs. Our first month 8% recurrent attack rate after allopurinol administration during the acute attack compared favorably to that in the allopurinol groups in previous studies 3,26,32,33 (11.5%-13%) that did

1132 The American Journal of Medicine, Vol 125, No 11, November 2012 Figure 4 Mean erythrocyte sedimentation rates over the study period. Error bars represent 95% CIs. ESR erythrocyte sedimentation rate. Table 2 Percentage of Acute Attacks in the First Month of Urate-Lowering Therapy Study Drug and Daily Dose Gout Flairs/ First Month FACT trial Febuxostat 80 mg 15% 756 patients Febuxostat 120 mg 22% Allopurinol 300 mg 13% APEX trial Febuxostat 80 mg 14% 799 patients Febuxostat 120 mg 18% Allopurinol 300/100 mg* 11.5% Placebo 10% CONFIRMS trial Febuxostat 40 mg 12.5% 2268 patients Febuxostat 80 mg 13.5% Allopurinol 300/200 mg* 11.5% Phase II dose study Febuxostat 40 mg 8% 153 patients Febuxostat 80 mg 8% Febuxostat 120 mg 13% Placebo 11% All patients were taking prophylaxis with colchicine or nonsteroidal anti-inflammatory drug, and none were started on allopurinol during the acute attack. *Dose adjusted for renal failure. Reported 8-week incidence divided by 2. not receive the drug during an acute attack. The recurrence rates in those studies also compared favorably with their own placebo groups (10%-11%), which corroborates our results. Given the low incidence and severity of subsequent attacks, and results aligned with prior studies, a larger study designed to detect smaller statistical differences is not likely to show clinical significance. Study Limitations Limitations include a study population recruited from a single Veterans Affairs Medical Center, all male, and possibly not generalizable to a non Veterans Affairs population. Exclusion criteria limiting comorbidities and treatment of primary gout on presentation mitigate possible differences between populations. We studied patients with relatively newly diagnosed gout, as seen in the primary care setting. All cases of gout were crystal proven on the day of entry, some for the first time, but this may not be similar to patients with long-standing gout. Although the average number of prior attacks on entry was 3.6, allopurinol was started in some patients on their first attack of gout. This may be controversial, but a spontaneous attack, without a predisposing condition

Taylor et al Allopurinol in Acute Gout Attacks 1133 (surgery, diuretic initiation, trauma), is cause for definitive urate-lowering therapy. Treatment of the first spontaneous attack is based on the fact that 60% of individuals with an initial attack experience a second flare within 1 year. 34,35 The cost associated with gout flares and uncontrolled hyperuricemia is double the cost of gout flares with treated hyperuricemia 6.0 mg/dl. 36 Early streamlined treatment with allopurinol might mollify side effects from more extensive use of nonsteroidal anti-inflammatory drugs, therapeutic doses of colchicine, and new expensive treatments with febuxostat, pegloticase, anakinra, rilonacept, and canakinumab, meant for advanced cases of gout. New treatment strategies should suggest future research directions. There is a need to study the initiation of allopurinol in chronic tophaceous gout and secondary gout. Patients unable to take indomethacin because of renal failure, heart failure, or anticoagulation are often treated with alternative strategies: oral colchicine, prednisone, intra-articular steroid, and interleukin-1 antagonist anakinra. Simplified allopurinol administration in a more complex population with gout might better address issues of compliance and cost-effectiveness. There are epidemiologic suggestions that hyperuricemia is a mediator of a variety of cardiovascular conditions. 37 Early initiation of allopurinol, for the first gout attack, might improve cardiovascular outcomes. CONCLUSIONS In uncomplicated gout, all 3 drugs, nonsteroidal anti-inflammatory drug of choice, a prophylactic dose of colchicine 0.6 mg once daily, and adequately dosed allopurinol 300 mg once daily, may be started during the acute attack. Graded incremental dosing of allopurinol was not used because we wanted to show that the initiation of treatment can be done to advantage with an adequate 300 mg dose, saving unnecessary visits, expense, chronic underdosing, and complexity of therapy. References 1. Neogi T. Clinical practice. Gout. N Engl J Med. 2011;364:443-452. 2. Grahame R, Scott JT. Clinical survey of 354 patients with gout. Ann Rheum Dis. 1970;29:461-468. 3. Becker MA, Schumacher HR Jr, Wortmann RL, et al. Febuxostat compared with allopurinol in patients with hyperuricemia and gout: FACT trial. N Engl J Med. 2005;353:2450-2461. 4. Sundy JS, Becker MA, Baraf HS, et al. Reduction of plasma urate levels following treatment with multiple doses of pegloticase (polyethylene glycol-conjugated uricase) in patients with treatment-failure gout: results of a phase II randomized study. Arthritis Rheum. 2008; 58:2882-2891. 5. Terkeltaub RA. Clinical practice. Gout. N Engl J Med. 2003;349:1647-1655. 6. Zhang W, Doherty M, Bardin T, et al. EULAR evidence based recommendations for gout. part II: Management. report of a task force of the EULAR standing committee for international clinical studies including therapeutics (ESCISIT). Ann Rheum Dis. 2006;65:1312-1324. 7. Jordan KM, Cameron JS, Snaith M, et al. British society for rheumatology and British health professionals in rheumatology guideline for the management of gout. Rheumatology (Oxford). 2007;46:1372-1374. 8. Mikuls TR, Farrar JT, Bilker WB, Fernandes S, Saag KG. Suboptimal physician adherence to quality indicators for the management of gout and asymptomatic hyperuricaemia: results from the UK General Practice Research Database (GPRD). Rheumatology (Oxford). 2005;44: 1038-1042. 9. Barber C, Thompson K, Hanly JG. Impact of a rheumatology consultation service on the diagnostic accuracy and management of gout in hospitalized patients. J Rheumatol. 2009;36:1699-1704. 10. Cannella AC, Mikuls TR. Understanding treatments for gout. Am J Manag Care. 2005;11(15 Suppl):S451-S458; quiz S465-S468. 11. Taylor WJ, Schumacher HR Jr, Singh JA, Grainger R, Dalbeth N. Assessment of outcome in clinical trials of gout a review of current measures. Rheumatology (Oxford). 2007;46:1751-1756. 12. Singh JA, Hodges JS, Toscano JP, Asch SM. Quality of care for gout in the US needs improvement. Arthritis Rheum. 2007;57:822-829. 13. Sarawate CA, Brewer KK, Yang W, et al. Gout medication treatment patterns and adherence to standards of care from a managed care perspective. Mayo Clin Proc. 2006;81:925-934. 14. Yue TF, Gutman AB. Effect of allopurinol (4-hydroxypyrazolo-(3,4- D)pyrimidine) on serum and urinary uric acid in primary and secondary gout. Am J Med. 1964;37:885-898. 15. Excess of amphicillin rashes associated with allopurinol or hyperuricemia. A report from the Boston collaborative drug surveillance program, Boston University Medical Center. N Engl J Med. 1972;286: 505-507. 16. Yu T. The efficacy of colchicine prophylaxis in articular gout a reappraisal after 20 years. Semin Arthritis Rheum. 1982;12:256-264. 17. Hollingworth P, Reardon JA, Scott JT. Acute gout during hypouricaemic therapy: prophylaxis with colchicine. Ann Rheum Dis. 1980;39: 529. 18. Borstad GC, Bryant LR, Abel MP, Scroggie DA, Harris MD, Alloway JA. Colchicine for prophylaxis of acute flares when initiating allopurinol for chronic gouty arthritis. J Rheumatol. 2004;31:2429-2432. 19. Wallace SL, Robinson H, Masi AT, Decker JL, McCarty DJ, Yu TF. Preliminary criteria for the classification of the acute arthritis of primary gout. Arthritis Rheum. 1977;20:895-900. 20. Massy-Westropp N, Ahern M, Krishnan J. A visual analogue scale for assessment of the impact of rheumatoid arthritis in the hand: validity and repeatability. J Hand Ther. 2005;18:30-33. 21. Grainger R, Taylor WJ, Dalbeth N, et al. Progress in measurement instruments for acute and chronic gout studies. J Rheumatol. 2009;36: 2346-2355. 22. Little RJ, Long Q, Lin X. A comparison of methods for estimating the causal effect of a treatment in randomized clinical trials subject to noncompliance. Biometrics. 2009;65:640-649. 23. Fischer K, Goetghebeur E, Vrijens B, White IR. A structural mean model to allow for noncompliance in a randomized trial comparing 2 active treatments. Biostatistics. 2011;12:247-257. 24. Perez-Ruiz F. Treating to target: a strategy to cure gout. Rheumatology (Oxford). 2009;48(Suppl 2):ii9-ii14. 25. Stamp LK, O Donnell JL, Zhang M, et al. Using allopurinol above the dose based on creatinine clearance is effective and safe in patients with chronic gout, including those with renal impairment. Arthritis Rheum. 2011;63:412-421. 26. Becker MA, Schumacher HR Jr, Wortmann RL, et al. Febuxostat, a novel nonpurine selective inhibitor of xanthine oxidase: a twentyeight-day, multicenter, phase II, randomized, double-blind, placebocontrolled, dose-response clinical trial examining safety and efficacy in patients with gout. Arthritis Rheum. 2005;52:916-923. 27. Chen LX, Schumacher HR. Gout: can we create an evidence-based systematic approach to diagnosis and management? Best Pract Res Clin Rheumatol. 2006;20:673-684. 28. Joosten LA, Netea MG, Mylona E, et al. Engagement of fatty acids with toll-like receptor 2 drives interleukin-1beta production via the

1134 The American Journal of Medicine, Vol 125, No 11, November 2012 ASC/caspase 1 pathway in monosodium urate monohydrate crystal-induced gouty arthritis. Arthritis Rheum. 2010;62:3237-3248. 29. Pascual E, Batlle-Gualda E, Martinez A, Rosas J, Vela P. Synovial fluid analysis for diagnosis of intercritical gout. Ann Intern Med. 1999;131:756-759. 30. Abeles M. Monosodium urate crystals in asymptomatic joints. Arthritis Rheum. 1980;23:124. 31. Dinarello CA. How interleukin-1beta induces gouty arthritis. Arthritis Rheum. 2010;62:3140-3144. 32. Schumacher HR Jr, Becker MA, Wortmann RL, et al. Effects of febuxostat versus allopurinol and placebo in reducing serum urate in subjects with hyperuricemia and gout: a 28-week, phase III, randomized, double-blind, parallel-group trial: APEX trial. Arthritis Rheum. 2008;59:1540-1548. 33. Becker MA, Schumacher HR, Espinoza LR, et al. The urate-lowering efficacy and safety of febuxostat in the treatment of the hyperuricemia of gout: the CONFIRMS trial. Arthritis Res Ther. 2010;12:R63. 34. Brixner DI, Ho MJ. Clinical, humanistic, and economic outcomes of gout. Am J Manag Care. 2005;11(15 Suppl):S459-S464; quiz S465-468 35. Neogi T, Hunter DJ, Chaisson CE, Allensworth-Davies D, Zhang Y. Frequency and predictors of inappropriate management of recurrent gout attacks in a longitudinal study. J Rheumatol. 2006;33:104-109. 36. Halpern R, Fuldeore MJ, Mody RR, Patel PA, Mikuls TR. The effect of serum urate on gout flares and their associated costs: an administrative claims analysis. J Clin Rheumatol. 2009;15:3-7. 37. Feig DI, Kang DH, Johnson RJ. Uric acid and cardiovascular risk. N Engl J Med. 2008;359:1811-1821.

Taylor et al Allopurinol in Acute Gout Attacks 1134.e1 Supplemental Table 1 Randomized Characteristics of Study Participants Characteristic Allopurinol Placebo Group (n 31) Group (n 26) P Value Mean age (SD), y 58 (13) 62 (12).23 Men, n (%) 31 (100) 26 (100) History of diabetes, n (%) 4 (13) 5 (19).51 Hypertension, n (%) 17 (55) 20 (77).08 Hyperlipidemia, n (%) 19 (61) 14 (54).57 Mean BMI (SD), kg/m 2 31 (5) 32 (6).7 Current alcohol use, n (%) 9 (29) 11 (42).3 Use of diuretics, n (%) 6 (19) 7 (27).78 Mean entry creatinine (SD), 1.09 (0.18) 1.13 (0.21).44 mg/dl Mean entry urate (SD), mg/dl 7.5 (1.33) 7.7 (1.72).65 Mean age of first attack 54.5 (15.8) 55.6 (12.4).8 (SD), y* Attack No., n (%).15 First attack 10 (32) 4 (16) 2-9 attacks 17 (55) 13 (52) 10 attacks 4 (13) 8 (32) Study joint, n (%).33 Joints of the hand 5 (16) 3 (12) Wrist 1 (3) 2 (8) Elbow 2 (6) 0 (0) Knee 5 (16) 5 (19) Ankle 8 (26) 4 (15) MTP 10 (32) 9 (35) Multiple joints 0 (0) 3 (12) BMI body mass index; MTP metatarsophalangeal; SD standard deviation. *n 29 for allopurinol group and n 24 for placebo group. Categoric P values are listed for categoric variables by chi-square analysis.

1134.e2 The American Journal of Medicine, Vol 125, No 11, November 2012 Supplemental Figure 1 Mean serum urate over study period including all patients randomized (intention to treat analysis). Error bars represent 95% CIs.

Taylor et al Allopurinol in Acute Gout Attacks 1134.e3 Supplemental Figure 2 Mean serum Erythrocyte Sedimentation Rates (ESR) over the study period including all patients randomized (intention to treat analysis). Error bars represent 95% CIs.

1134.e4 The American Journal of Medicine, Vol 125, No 11, November 2012 Supplemental Figure 3 Mean serum C-reactive Protein (CRP) over study period including all patients randomized (intention to treat analysis). Error bars represent 95% CIs.

Taylor et al Allopurinol in Acute Gout Attacks 1134.e5 Supplemental Figure 4 Mean serum creatinine over study period including all patients randomized (intention to treat analysis). Error bars represent 95% CIs.

1134.e6 The American Journal of Medicine, Vol 125, No 11, November 2012 Supplemental Figure 5 Mean VAS Scores on Days 1-10 Allopurinol vs. Placebo for all patients randomized (intention to treat analysis). Supplemental Figure 6 Sensitivity analysis displaying extreme versus average assumptions for Visual Analog Pain Scale (VAS) scores on Days 1-10 Allopurinol vs. Placebo for all patients randomized (intention to treat analysis).

Taylor et al Allopurinol in Acute Gout Attacks 1134.e7 Supplemental Figure 7 Visual Analog Pain Scale (VAS) scores on Days 1-10 Allopurinol vs. Placebo for all patients with their first attack. Supplemental Figure 8 Visual Analog Pain Scale (VAS) scores on Days 1-10 Allopurinol vs. Placebo for all patients whom have had multiple attacks.