Abstract
Objective. To systematically review the evidence on the efficacy, safety, and cost-effectiveness of urate-lowering therapy for gout: xanthine oxidase inhibitors (allopurinol and febuxostat), uricosuric medications (benzbromarone, probenecid and sulfinpyrazone), and uricases (pegloticase and rasburicase).
Methods. A systematic review was performed as part of the 3e (Evidence, Expertise, Exchange) Initiative on Gout. The primary efficacy outcomes were frequency of acute gout attacks, study participant withdrawal due to adverse events, and cost-effectiveness. Serum urate-lowering was a secondary outcome and was the most commonly reported outcome in the included trials.
Results. The search identified 17 articles for efficacy, 31 for safety, and 3 for cost-effectiveness. The main outcome described in these studies was serum urate-lowering. Allopurinol, febuxostat, and pegloticase are all effective at lowering serum urate compared to placebo and febuxostat (≥ 80 mg) was more effective at lowering serum urate than allopurinol. Compared to probenecid, benzbromarone was more effective at lowering serum urate. Regarding acute gout attacks, pegloticase and febuxostat (≥ 120 mg) resulted in more acute attacks than placebo. Regarding the primary safety outcome, more withdrawals due to adverse events were seen only when pegloticase was compared to placebo. The two trials of cost-effectiveness were inconclusive.
Conclusion. There is currently moderate quality data supporting the efficacy and safety of allopurinol, febuxostat, benzbromarone, and probenecid in gout. Pegloticase, while efficacious, is associated with more withdrawals due to adverse events and infusion reactions. There is insufficient evidence currently with respect to the cost-effectiveness or the most optimal sequencing of urate-lowering therapy.
- GOUT
- TREATMENT
- ALLOPURINOL
- FEBUXOSTAT
- URICOSURIC AGENTS
- PEGLOTICASE
This article is part of the 3e (Evidence, Expertise, Exchange) Initiative on Diagnosis and Management of Gout1. The objective of the current work was to systematically review the available literature concerning the following one of 10 selected questions as an evidence base for generating the recommendations: What is the efficacy, cost efficacy and safety for urate-lowering therapy (allopurinol but also febuxostat, peguricase, benzbromarone, and probenecid) in the treatment of gout? Which sequence of urate-lowering drugs or combinations should be recommended?
This article includes a shortened version of 2 Cochrane reviews evaluating the efficacy and safety of allopurinol and uricosuric medications for chronic gout based upon randomized controlled trials (RCT) and controlled clinical trials (CCT)2,3,4,5. It also includes a safety review based on observational studies and a cost-effectiveness review.
MATERIALS AND METHODS
The systematic reviews performed to address the 3e research question proposed in this article followed the guidelines of the Cochrane Handbook6.
Rephrasing the research question
The question formed by the expert panel was translated into epidemiologic terms using the PICO (Population, Intervention, Comparator, Outcome) format6. Separate PICO were created for efficacy, safety, and cost-effectiveness. The population was any adult (age ≥ 18 yrs) with gout. Interventions were xanthine oxidase inhibitors (allopurinol and febuxostat), uricosuric medications (benzbromarone, probenecid, and sulfinpyrazone), and uricases (pegloticase and rasburicase). Comparators were placebo, no treatment, nonpharmacologic treatment, and other urate-lowering therapy. Trials investigating combinations of 2 or more urate-lowering therapies or comparing different doses of the same medication were also included. Outcomes were based on those recommended by Outcome Measures in Rheumatology (OMERACT)7,8. For efficacy, the primary outcome was frequency of acute gout attacks. Secondary efficacy outcomes were pain reduction, health related quality of life, serum urate normalization, function (i.e., activity limitation), and tophus regression. The primary safety outcome was participant withdrawal due to adverse events (WAE) while secondary outcomes were total adverse events (AE) and serious adverse events (SAE). The primary outcome for the economic analysis was the incremental cost-effectiveness ratio (ICER) as cost per quality-adjusted life-years. Only RCT and CCT (i.e., where the method of allocation is quasi-random) were considered for efficacy, while for safety we also included cohort studies, case-control studies, and case series (> 20 patients). For cost-effectiveness we included full and partial economic evaluations, RCT that included economic data, and technology assessments.
Systematic literature search
For each review we performed searches from inception to October 2011 in MEDLINE, EMBASE, and Cochrane Central using a comprehensive search strategy developed in collaboration with an experienced librarian. For details on search strategies see the online Appendix available from www.3egout.com. Abstracts from the 2010 and 2011 American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR) conferences and the reference lists from retrieved articles and systematic reviews were also hand searched for additional studies. The original searches have been updated since the 3e recommendations were formulated and are discussed further in the associated Cochrane articles4,5 updated May 13, 20134, and January 14, 20145; the updated searches did not retrieve new information that would alter the 3e recommendations.
Selection of articles
Two reviewers (AK, RS) independently screened titles and abstracts retrieved from the searches and the full text as necessary to identify eligible studies. Included articles could be from any one of the languages spoken by the 3e bibliographic authors including English, French, Dutch, German, Spanish, or Portuguese (Cochrane reviews were unrestricted by language).
Data extraction and assessment of risk of bias
The same 2 reviewers independently extracted data of included studies into a standardized template, including trial characteristics, country and source of funding; participant characteristics and inclusion/exclusion criteria; and description of the interventions and outcomes reported. Raw data (means and standard deviations for continuous outcomes and number of events or participants for dichotomous outcomes) were extracted for outcomes of interest. Risk of bias (ROB) for RCT and CCT was assessed using the Cochrane ROB Tool9, for cohort studies using Hayden, et al10, and for case-control studies using the Newcastle-Ottawa Scale (NOS)11.
Data analysis
For continuous data, results were analyzed as mean difference (MD) between the intervention and comparator group with 95% confidence intervals. For dichotomous data, a relative risk (RR) with corresponding 95% confidence intervals was calculated. Only data from studies considered sufficiently clinically homogeneous were pooled. Statistical heterogeneity was assessed using the I2 statistic.
RESULTS
The literature searches identified 1594 efficacy abstracts, 2486 safety abstracts, and 244 cost-effectiveness abstracts (Figure 1). Of these, 17 trials met inclusion criteria for efficacy outcomes, 31 for safety outcomes, and 3 full-text articles (describing 2 studies) for cost-effectiveness outcomes. No additional studies were included from conference abstracts as none had been accepted for publication by our review deadline (February 1, 2012). Characteristics of included studies and study participants are summarized in Tables 1–3.
Allopurinol
Nine studies examined allopurinol efficacy (6 RCT12 – 16,17 and 3 CCT18,19,20) and 20 examined allopurinol safety outcomes (Tables 1, 2). Gout prophylaxis was provided in all studies with either colchicine14,15,20 or a choice of colchicine or nonsteroidal antiinflammatory drugs (NSAID)12,13,16,17,18,19. Prophylaxis duration varied between 1 month and duration of the study (up to 6 months). All but 3 studies14,18,20 provided safety data. An additional 2 cohort studies21,22, 1 open label extension23, 1 case-control study24, and 10 case series or other observational studies25 – 29,30 – 34 also provided safety data.
Febuxostat
There were 4 RCT that assessed efficacy12,13,17,35 (Tables 1, 2). As detailed above, 2 RCT compared febuxostat to allopurinol12,13, and 1 compared febuxostat to both allopurinol and placebo17. A fourth study compared febuxostat to placebo35. Allopurinol was given in a dose of 300 mg daily13 or 300 mg daily in participants with normal renal function and 100 mg or 200 mg daily in participants with impaired renal function12,17. Flare prophylaxis was with colchicine35 or either colchicine or NSAID12,13,17. Safety data were provided by all 4 RCT as well as 2 open-label extension studies23,36, 2 posthoc analyses37,38, and 1 observational study39.
Uricosurics
There were 4 trials that assessed uricosuric efficacy15,16,20,40 (Tables 1, 2). Two RCT compared benzbromarone to allopurinol15,16, 1 RCT compared benzbromarone to probenecid in participants who had failed allopurinol40, and 1 CCT compared probenecid to allopurinol therapy20. No studies of sulfinpyrazone met inclusion criteria. Acute gout prophylaxis was used in all studies with either colchicine15,20 or a choice of colchicine or NSAID16,40. Where compared, allopurinol was used at doses up to 600 mg/day in 2 studies16,20 and 300 mg/day in 1 study15. Safety data were provided by all trials and 2 additional observational studies21,34.
Uricases
Three RCT of pegloticase (described in 2 reports41,42 and 1 posthoc analysis43) met inclusion criteria for efficacy, while 2 additional phase 1 studies of pegloticase were included in the safety assessment44,45. No studies met inclusion criteria for rasburicase. Two replication RCT assessed the efficacy and safety of pegloticase 8 mg vs placebo for 6 months; pegloticase was given either every 2 weeks or every month41. A third study randomized participants to different pegloticase doses (between 4 mg and 12 mg) given either every 2 weeks or every month42. Colchicine or NSAID prophylaxis was provided for the duration of the replication RCT41 but was provided at the discretion of the investigator in the earlier Phase 2 study42.
Economic search results
Two full or partial economic analyses (3 reports, but 2 of them covered the same study) met inclusion criteria (Table 3).
Risk of Bias
Risk of bias of included studies is shown in Tables 1 and 2; and further details are reported in the Cochrane reviews4,5. For allopurinol the 9 RCT and CCT had unclear12,13,16,17 or high risk of bias14,15,18,19,20; the 14 observational studies had moderate22,24 or high risk of bias21,23,25 – 29,30,31,32,33. For febuxostat, the 4 RCT had unclear risk of bias12,13,17,35, and the 5 observational studies had unclear37 and high23,36,38,39 risk of bias. For uricosurics, 6 RCT of uricosuric efficacy had unclear16,40 or high risk of bias15,20,21,34. The 2 replication RCT of pegloticase41 and the posthoc analysis of these studies43 had low risk of bias. The other Phase 2 and Phase 1 studies had high risk of bias42,44,45.
Efficacy of Interventions
For full results of the data analyses see online appendix, available from www.3egout.com.
Allopurinol vs placebo
In the first 2 months of therapy, while all patients were taking acute gout prophylaxis, there was no statistically significant difference in the frequency of acute gout attacks between allopurinol at the studied doses (up to 300 mg/day) and placebo (1 trial; unclear ROB). Significantly more participants taking allopurinol achieved a serum urate level < 6.0 mg/dl (0.36 mmol/l) (1 trial; unclear ROB; RR 49.3; 95% CI 7.0 to 349.0; absolute risk difference (ARD) 38% more achieved target)5. There was no between-group difference in number of tophi following therapy (but results were not presented)17.
Febuxostat versus placebo
Frequency of acute gout attacks was similar for placebo and febuxostat 40 mg (1 trial; unclear ROB) and 80 mg/day (2 trials; unclear ROB), while higher doses of febuxostat were more likely than placebo to be associated with acute gout attacks: 120 mg febuxostat/day (2 trials; unclear ROB; pooled RR 1.7; 95% CI 1.3 to 2.3; ARD 16% more likely) and 240 mg febuxostat/day (1 trial; unclear ROB; RR 2.6; 95% CI 1.8 to 3.7; ARD 31% more likely). All doses of febuxostat studied were more likely than placebo to achieve a serum urate level of < 6.0 mg/dl (0.36 mmol/l): febuxostat 40 mg/day (1 trial; unclear ROB; RR 44.1; 95% CI 2.8 to 702.9; ARD 57% more achieved target), febuxostat 80 mg/day (2 trials; unclear ROB; pooled RR 78.8; 95% CI 16.0 to 388.2; ARD 72% more achieved target), febuxostat 120 mg/day (2 trials; unclear ROB; pooled RR 90.1; 95% CI 18.3 to 443.4; ARD 86% more achieved target), and febuxostat 240 mg/day (1 trial; unclear ROB; RR 116.9; 95% CI 16.6 to 824.2; ARD 91% more achieved target).
Allopurinol versus febuxostat
During the first 8 weeks of therapy, while taking gout prophylaxis, no significant difference in acute gout attacks was detected comparing allopurinol at studied doses (up to 300 mg/day) and febuxostat 80 mg/day (2 trials; unclear ROB)5. Allopurinol was less likely to be associated with an acute gout attack when compared with higher doses of febuxostat: 120 mg/day (2 trials; unclear ROB; pooled RR 0.6; 95% CI 0.5 to 0.8; ARD 14% less likely) or 240 mg/day (1 trial; unclear ROB; RR 0.4; 95% CI 0.3 to 0.6; ARD 29% less likely). There was no significant difference between allopurinol and febuxostat 40 mg/day with respect to serum urate normalization (1 trial; unclear ROB). Compared with higher doses of febuxostat, however, allopurinol was less likely to achieve the serum urate target of < 6.0 mg/dl (0.36 mmol/l): febuxostat 80 mg/day (3 trials; unclear ROB; pooled RR 0.6; 95% CI 0.5 to 0.7; ARD 32% fewer achieved target), febuxostat 120 mg/day (2 trials; unclear ROB; pooled RR 0.5; 95% CI 0.4 to 0.5; ARD 42% fewer achieved target), or febuxostat 240 mg/day (1 trial; unclear ROB; RR 0.4; 95% CI 0.4 to 0.5; ARD 53% fewer achieved target)5.
Allopurinol versus uricosurics
There was no significant difference between 4 months of allopurinol and benzbromarone in frequency of acute gout attacks (1 trial; unclear ROB) or 2 months of probenecid (1 trial; high ROB). Four months of allopurinol or benzbromarone with dose escalation resulted in no significant difference in whether patients achieved a target serum urate level (2 trials; unclear and high ROB)4.
Uricosuric comparison
Benzbromarone demonstrated no statistically significant difference in the frequency of acute gout attacks when compared with 2 months of probenecid (1 trial; unclear ROB). Benzbromarone, however, was more likely to achieve a target serum urate of < 0.3 mmol/l (5 mg/dl) (1 trial; unclear ROB; RR 1.4; 95% CI 1.0 to 2.0; ARD 24% more achieved target)4.
Uricases versus placebo
Pegloticase, given either biweekly or monthly, was associated with more acute gout attacks in the first 3 months of therapy than placebo (2 replication trials; low ROB; biweekly pooled RR 1.4; 95% CI 1.0 to 1.9 and monthly pooled RR 1.5; 95% CI 1.1 to 2.0; ARD 22% and 27% more likely, respectively). Both biweekly and monthly pegloticase administration were more likely to achieve a target serum urate level of < 6 mg/dl (< 0.36 mmol/l) than placebo (2 replication trials; low ROB; biweekly pooled RR 37.4; 95% CI 2.4 to 594.3 and monthly pooled RR 30.5; 95% CI 1.9 to 488.1; ARD 42% and 35% more achieved target, respectively). Pegloticase was associated with greater improvement in disability (measured by the HAQ-DI) when compared to placebo: monthly (2 replication trials; low ROB; mean difference −0.2; 95% CI −0.4 to −0.1), and biweekly (2 replication trials; low ROB; mean difference −0.2; 95% CI −0.4 to −0.1). Resolution of 1 or more tophi was more likely to occur after 6 months of therapy with either monthly pegloticase (2 replication trials; low ROB; pooled RR 2.9; 95% CI 0.7 to 12.0) or biweekly pegloticase (2 replication trials; low ROB; pooled RR 5.5; 95% CI 1.4 to 21.6) when compared to placebo. Pegloticase given biweekly was also associated with greater improvement in participant pain compared with placebo (2 replication trials; low ROB; mean difference −14.2; 95% CI −24.4 to −4.0 on a 0–100 VAS). Monthly pegloticase was not associated with greater improvement in pain compared with placebo.
Safety of interventions
There were no differences in withdrawals due to AE between allopurinol, placebo, or febuxostat. However, allopurinol was associated with more AE than 80 mg (3 trials; unclear ROB; pooled RR 1.1; 95% CI 1.0 to 1.1; ARD 4% more individuals experienced adverse events) and 120 mg/day febuxostat (2 trials; unclear ROB; pooled RR 1.1; 95% CI 1.1 to 1.2; ARD 9% more individuals).
There were no differences between the 4-month comparison of allopurinol and benzbromarone therapy or 2-month comparison of benzbromarone and probenecid with respect to number of withdrawals due to AE or total number of AE.
Compared with placebo, participants treated with pegloticase were more likely to withdraw due to AE (2 replication trials; low ROB; biweekly RR 7.6; 95% CI 1.0 to 55.6 and monthly RR 8.2; 95% CI 1.1 to 59.7; ARD 15% and 17% more withdrew, respectively), but there was no significant difference between pegloticase and placebo with respect to total AE. Pegloticase was noted to have a significantly higher number of infusion reactions than placebo: biweekly 26% (p = 0.002), monthly 42% (p < 0.0001), and placebo 5%41. Based on postmarketing surveillance, an ACR Hotline 201246 indicated that pegloticase should not be used concurrently with urate-lowering medications as there may be a higher risk of infusion reactions.
Table 2 summarizes the safety outcomes from the included observational studies.
Economic outcomes
One of the included articles was an evaluation of the manufacturer of febuxostat’s submission to the National Institute for Health and Clinical Excellence (NICE) single technology appraisal process47,48. Although the NICE technology appraisal concluded that the economic analysis submitted by the manufacturer was full of uncertainties — precluding its use in assessing the cost-effectiveness of febuxostat in the management of hyperuricemia in patients with gout — no model details were presented to enable review47. The other study, by Ferraz, et al49, is a decision analysis model from a 1993 Canadian population analysing the initiation of urate-lowering drug (allopurinol) versus non-urate-lowering drug initiation from a societal perspective regarding costs with a 1-year period. Although this study was judged to be at low ROB (per NICE guideline methodology for economic evaluations50), its assumptions regarding medication adherence, acute attack rate, and effectiveness of different urate-lowering medications are now known to be erroneous, making it difficult to draw any conclusions from this study.
DISCUSSION
The objective of this systematic review was to summarize the evidence for the efficacy, safety, and cost-effectiveness of urate-lowering therapies in patients with gout. We identified 17 studies addressing efficacy, 31 studies addressing safety, and 2 studies addressing cost-effectiveness.
The most well documented outcome was serum urate normalization. Acute gout attacks, the primary outcome of this systematic review, are more difficult to interpret because the incidence of these attacks is higher in the course of urate-lowering therapy (mobilization gout). The short-term nature of the included studies did not allow assessment of the longterm reduction in acute gout attacks. Very few of the studies addressed the other OMERACT gout outcomes, such as pain reduction, health related quality of life, function (i.e., activity limitation), and tophus regression.
There have recently been 2 other systematic reviews of urate-lowering therapy in gout51,52. The first is summarized in the ACR Gout Guidelines51,53. The conclusions of both reviews are broadly consistent with our review although they differed with respect to study inclusion. The review on which the ACR guidelines were based included observational studies40,54,55, while Hamburger, et al included uncontrolled trials34,56, and studies of people with asymptomatic hyperuricemia57. As well Hamburger, et al limited the review to include only studies published from 2005, and excluded uricosurics other than probenecid because these are not currently available in the US14,15,18,19,20,40.
In summary, there is currently moderate quality evidence for the efficacy and safety of a number of urate-lowering therapies including allopurinol, febuxostat, benzbromarone, and probenecid. Pegloticase, while efficacious, results in a greater number of withdrawals due to AE and infusion reactions compared with placebo. It is thought that these infusion reactions may occur more frequently in individuals with higher anti-pegloticase antibody levels41. At the time of this review, there was insufficient economic evidence from full or partial economic evaluations with which to support or refute treatment decisions regarding urate-lowering therapy. Future research would be beneficial in examining higher doses of allopurinol, combination therapy, longer duration of urate-lowering therapy, and economic analyses, as well as reporting other outcome measures such as health related quality of life, pain reduction, and tophus regression.
Acknowledgments
We acknowledge the support of the Cochrane Musculoskeletal Group in publishing the associated Cochrane reviews. We also acknowledge Simon Huang, Ian Tsang, Eugene Agranovich, Caroline van Durme, Francisca Sivera, Arnd Kleyer, and Daniel Aletaha for assistance with translation of potentially relevant studies.
Footnotes
-
The authors acknowledge the work of all members of the 3e scientific committees and all participants in the national meetings. This article is derived from the 3e Gout program, which was sponsored by AbbVie Inc. Margaux Orange, Paris, France, provided logistical and administrative support for the 3e Gout meetings; this work was funded by AbbVie Inc. AbbVie employees were present during the 3e meetings, but did not influence the scientific discussions. AbbVie did not review the content or have influence on this manuscript. This article is also based on 2 Cochrane reviews submitted to the Cochrane Database of Systematic Reviews (CDSR) (see the www.cochranelibrary.com for information). Cochrane reviews are regularly updated as new evidence emerges and in response to feedback, and the CDSR should be consulted for the most recent version of the review.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.
- 32.
- 33.
- 34.
- 35.
- 36.
- 37.
- 38.
- 39.
- 40.
- 41.
- 42.
- 43.
- 44.
- 45.
- 46.
- 47.
- 48.
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.
- 55.
- 56.
- 57.