SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

It is commonly accepted that the target serum urate level in patients receiving urate-lowering therapy for dissolution of urate crystals in hyperuricemia of gout is <6 mg/dl, and that patients with gout should continue urate-lowering therapy for the rest of their lives. This study was undertaken to reevaluate whether this stringent therapeutic target to dissolve crystals must be maintained lifelong to prevent new crystal formation.

Methods

In a prospective cohort of 211 patients with gout, urate-lowering therapy was withdrawn after 5 years if no tophus was present at baseline, or 5 years after resolution of the last tophus. Data on recurrence of gout and on serum urate levels and other potentially associated variables were analyzed.

Results

Multivariate regression analysis of time to crystal-proven recurrence of gout showed that serum urate levels during urate-lowering treatment and after its withdrawal were independently related to gout recurrence. None of the patients who had average serum urate levels of <7 mg/dl after urate-lowering therapy withdrawal developed a crystal-proven recurrence of gout. Post hoc analysis showed that weight loss and use of drugs that lower serum urate, such as losartan or fenofibrate, were associated with serum urate levels of <7 mg/dl during followup after urate-lowering therapy withdrawal; use of diuretics was associated with failure to achieve serum urate levels of <7 mg/dl during followup.

Conclusion

Our data support the hypothesis that after appropriate long-term treatment of hyperuricemia in gout with urate crystal dissolution being the therapeutic target, lifelong treatment can be targeted to achieve serum urate levels just below the threshold for saturation to avoid new crystal formation, similar to cleaning a dirty dish: more is required to get it clean than to keep it clean.

In gout, the major aim of treatment is to reduce serum urate levels to below the threshold for urate saturation and to maintain this reduction in the long term, thus favoring the dissolution of previously formed monosodium urate monohydrate (MSU) crystals and avoiding the formation of new ones (1). According to current recommendations for the treatment of hyperuricemia in gout, the target serum urate level is ≤6 mg/dl. This recommendation is for all gout patients, but for patients with severe gout the target could be lower, in order to promote more rapid dissolution of urate crystals in tissue (2). Indeed, sustained control of serum urate levels below this target is associated with the disappearance of urate crystals in synovial fluid (3). Once urate crystals have dissolved completely during long-term urate-lowering therapy, the next question for clinicians treating patients with gout could be: What is the level of serum urate that should be targeted going forward throughout the patient's life? Is it as low as the level required in order for crystals to dissolve, or could it be higher, i.e., just below the threshold for avoiding the formation of new crystals? The clinical implication is clear: in very-long-term followup after a prolonged period of urate-lowering therapy that has resulted in the complete dissolution of previously formed MSU crystals, targeting just below the threshold for formation of new crystals would imply that lower doses of urate-lowering drugs would be needed.

In the present study, we assessed the clinical value of a 2-stage approach to targeting of serum urate levels. In the first stage, the target serum level would be the level required for crystal dissolution (<6 mg/dl). In the second stage, the target would be the level required for avoidance of new crystal formation (6–7 mg/dl).

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

The study population consisted of a prospective cohort of patients in whom urate-lowering therapy has been withdrawn. The study has been approved by the institutional clinical investigation ethics committee, and a preliminary analysis of this cohort was published previously (4). Patients in the cohort are recruited on an ongoing basis. Criteria for enrollment are as follows: diagnosis of gout based on observation of MSU crystals from a tophus or synovial fluid aspiration, treatment with urate-lowering therapy with average serum urate levels of <7 mg/dl during the entire period of therapy, and compliance with urate-lowering therapy given for a 5-year period in patients who had nontophaceous gout prior to starting urate-lowering therapy, or for 5 years after clinical disappearance of all subcutaneous tophi confirmed by physical examination in patients who exhibited tophaceous gout prior to starting urate-lowering therapy. Compliance was defined as a serum urate level of <6 mg/dl on ≥80% of the occasions it was tested or was based on findings of serum urate levels of <6 mg/dl once stable and effective dosing of urate-lowering drugs was achieved. As of December 2010, there were 229 patients in the cohort.

The following variables were analyzed: age, sex, duration of gout, number of joints ever involved, presence of tophi, urate-lowering drugs prescribed, hypertension, hyperlipidemia, diabetes, use of diuretics, and predicted glomerular filtration rate (determined using the Cockcroft-Gault equation) at withdrawal of urate-lowering therapy and at last observation. Serum urate was measured at baseline and at the time of urate-lowering therapy withdrawal, and the average serum urate level during the observation period after urate-lowering therapy withdrawal was ascertained.

Serum urate levels were measured during followup at least twice during the first year after withdrawal of urate-lowering therapy and at least once yearly thereafter. Average serum urate levels during followup were calculated using a trapezoidal method to balance time between serum urate determinations (5). Patients in whom urate-lowering therapy was reinitiated or low-dose colchicine or nonsteroidal antiinflammatory drugs were initiated for any reason while they were asymptomatic after urate-lowering therapy withdrawal were analyzed to the last observation prior to starting the medication. Recurrence was defined as a clinical event suggestive of a gout flare that was confirmed by a finding of MSU crystals.

For continuous variables, the mean ± SD, median, and interquartile range (IQR) were calculated. Survival analysis was done with Kaplan-Maier curves and estimates, and survival curves were compared by log rank test. Multivariate Cox proportional risk hazard models were analyzed to assess the association of variables with recurrence of gout. Analysis time was the time elapsed from urate-lowering therapy withdrawal to confirmed recurrence or to cutoff (last recorded visit), whichever occurred first. Log rank tests were used to select variables possibly associated with the outcome, with continuous variables transformed into discrete variables following quartile distribution. Variables showing P values of less than 0.2 were selected for further stepwise multivariate Cox models. Proportionality was assessed with plots of residuals.

For analysis purposes, patients were divided into 4 groups according to serum urate levels after urate-lowering therapy withdrawal: group 0 (6.00–6.99 mg/dl), group 1 (7.00–8.20 mg/dl), group 2 (8.21–9.32 mg/dl), and group 3 (9.33–12.40 mg/dl) (the 3 latter groups coming from the tertile distribution of levels ≥7.00 mg/dl). Time to recurrence was analyzed by group. The IBM SPSS version 18.0 statistical package was used for analysis. Post hoc analysis with chi-square test, Fisher's exact test, or t-test as applicable was performed to explore variables that might potentially be associated with serum urate levels <7 mg/dl during followup after withdrawal of urate-lowering therapy.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Two hundred eleven patients were included in the current analysis. There were 202 men (95.7%) and 9 women. The mean ± SD age at onset of clinical gout was 59.9 ± 11.6 years (median 55.5 [IQR 48–65]), and the time from onset of gout to first visit was 6.7 ± 6.3 years (median 5 [IQR 2–10]). The remaining 18 patients in the cohort of 229 were excluded from the present analysis for the following reasons: 8 were lost to followup prior to the first scheduled followup visit, 3 had not yet had the first scheduled followup visit, 1 had average serum urate levels of <6 mg/dl (0.36 mmoles/liter), and 6 did not provide consent for joint aspiration

Fifty-two of the 211 patients (24.6%) had subcutaneous tophi at baseline, and 55 (26.1%) had experienced gout flares in >4 joints since the onset of clinical gout. Mean ± SD serum urate levels were 8.8 ± 1.3 mg/dl (median 8.5 [IQR 8.0–9.4]), 4.9 ± 0.9 mg/dl (median 4.9 [IQR 4.4–5.4]), and 8.5 ± 1.4 mg/dl (median 8.4 [IQR 7.5–9.5]) at baseline, during urate-lowering therapy, and after withdrawal of urate-lowering therapy, respectively. The mean duration of urate-lowering therapy was 66 ± 2.6 months (median 60.2 [IQR 60–72]). Patients with nontophaceous gout received urate-lowering therapy for 60 ± 1.4 months (median 60 [IQR 60–60]), and those with tophaceous gout for 80.9 ± 8.7 months (median 81 [IQR 73–85]).

The mean ± SD number of serum urate determinations per year during followup after withdrawal of urate-lowering treatment was 1.9 ± 0.6 (median 2 [IQR 1–6]). Flares were observed during the first 2 years of urate-lowering therapy, with the last flare observed in any patient occurring at 23 months after initiation of the therapy. The followup time after withdrawal of urate-lowering therapy was 33.1 ± 22.6 months (median 27.5 [IQR 16.0—47.5]).

Eighty-two patients (38.9%) had a crystal-proven recurrence of gout during the followup observation period; in 37 of the 82 (45.1%), the recurrence occurred in a first metatarsophalangeal joint. None of the 27 patients in group 0 had recurrence of gout, while 13 of 61 (21.3%), 31 of 61 (50.8%), and 38 of 62 (61.3%) of the patients in groups 1, 2, and 3, respectively, experienced recurrence (Table 1 and Figure 1). In 8 patients a possible event occurred but urate crystals could not be retrieved after joint aspiration, and they were not considered recurrences according to the protocol definition.

Table 1. Recurrence of gout and time to recurrence, by average serum urate levels during followup
Group, average serum urate level (n)Events, no. %Time to recurrence, months*
Mean (95% CI)Median (95% CI)
  • *

    95% CI = 95% confidence interval.

0, 6.00–6.99 mg/dl (27)0
1, 7.00–8.20 mg/dl (61)13 (21.3)87 (70–103)124 (–)
2, 8.21–9.32 mg/dl (61)31 (50.8)51 (43–60)46 (41–50)
3, 9.33–12.40 mg/dl (62)38 (61.3)28 (24–33)25 (22–28)
   Total (211)82 (38.9)56 (48–64)47 (43–51)
thumbnail image

Figure 1. Survival function plot by serum urate (Sur) levels after withdrawal of urate-lowering therapy.

Download figure to PowerPoint

The estimated median time to recurrence was 47 months (95% confidence interval [95% CI] 43–51) (Table 1). The cumulative recurrence rate at 1, 2, 3, and 4 years was 6.6%, 11.4%, 20.4%, and 29.4%, respectively.

Kaplan-Meier survival analysis showed possible associations of recurrence events with several variables. Higher risk of recurrence was associated with higher serum urate levels prior to initiation of urate-lowering therapy, higher serum urate levels during followup after withdrawal of urate-lowering therapy, presence of tophi prior to initiation of urate-lowering therapy, and polyarticular involvement. With the P value cutoff of 0.2 used for the initial bivariate analysis, lower risk of recurrence was associated with higher estimated creatinine clearance at last observation, lower serum urate levels during urate-lowering therapy, and treatment with benzbromarone. Cox multivariate survival analysis showed that higher serum urate levels during treatment (hazard risk 1.57 per mg/dl [95% CI 1.18–2.08]) and higher serum urate levels during followup after withdrawal of urate-lowering therapy (hazard risk 2.29 per mg/dl [95% CI 1.91–2.74]) were the only variables independently associated with crystal-proven recurrence (Table 2).

Table 2. Survival analysis of variables associated with crystal-proven recurrence of gout among patients in whom urate-lowering therapy had been withdrawn*
VariableBivariate analysis (Kaplan-Meier log rank)Multivariate analysis (Cox proportional hazards)
HR95% CIPHR95% CIP
  • *

    HR = hazard ratio; 95% CI = 95% confidence interval; ClCr = creatinine clearance.

Serum urate level (mg/dl) at baseline1.421.18–1.700.001.030.84–1.260.74
Serum urate level (mg/dl) during treatment1.731.31–2.290.001.571.18–2.080.00
Serum urate level (mg/dl) after withdrawal of treatment2.351.97–2.810.002.291.91–2.740.00
ClCr (ml/minute/1.73 m2) at last observation0.990.98–1.000.190.990.99–1.000.17
Tophi1.781.13–2.810.011.300.78–2.170.31
Polyarticular involvement1.991.26–3.140.001.190.65–2.180.57
Benzbromarone use0.740.47–1.140.170.770.47–1.180.31

Factors related to achievement of serum urate levels ranging from 6 to 6.99 mg/dl after withdrawal of urate-lowering therapy were analyzed. This analysis revealed associations of the following variables with achievement of serum urate levels of <7 mg/dl after treatment withdrawal: reduction of weight at last followup visit compared to baseline weight at the time of enrollment in the cohort (mean ± SD weight change −3.33 ± 4.64 kg in patients with serum urate levels <7 mg/dl versus 1.24 ± 4.67 kg in those with levels ≥7 mg/dl; P < 0.01), treatment with losartan (5 of 27 patients with serum urate levels <7 mg/dl [18.5%] versus 9 of 184 with levels ≥7 mg/dl [4.9%]; P = 0.02), and treatment with fenofibrate (4 of 27 [14.8%] versus 5 of 184 [2.7%]; P = 0.02); treatment with diuretics was associated with lower likelihood of achieving serum urate levels of <7 mg/ml (1 of 27 [3.7%] versus 32 of 184 [17.4%]; P < 0.01). Although associated with lower serum urate levels, these variables were not independently associated with lower risk of recurrence.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

The main goal in the treatment of gout is to achieve complete dissolution of already-formed MSU crystals and to avoid further formation of new ones. There is compelling evidence that targeting and long-term sustaining of serum urate levels below the threshold for urate saturation is associated with improved clinical outcomes (6), and such an approach was included in the European League Against Rheumatism Task Force for Gout recommendations for the management of gout (2). Recurrence of gout is related to increasing serum urate levels after urate-lowering therapy is withdrawn (4), and lifelong maintenance of urate-lowering therapy is advised to avoid new MSU crystal formation after complete dissolution of MSU crystals with appropriate treatment (7).

Sustaining serum urate levels at <6 mg/dl (therapeutic target) will certainly secure MSU dissolution in the long term and, obviously, avoid new crystal formation. However, once MSU crystals have completely disappeared from tissue, should the targeted level of serum urate remain at <6 mg/dl lifelong, or should a different target, i.e., just below the threshold to avoid new MSU crystal formation (preventive target), be considered for the long term?

The clinical implications for such an approach seem obvious: a lower target for the very-long-term treatment of hyperuricemia in patients with gout may reduce exposure to urate-lowering drugs, as lower doses of these drugs may be needed to achieve this “preventive target” for serum urate, compared to the “therapeutic target.” Thus, a 2-stage treatment approach for the control of serum urate levels in gout could be formulated. A figurative analogy to a dirty dish may help in visualizing the issue: the initial effort to clean the dish (serum urate therapeutic target) would depend on how dirty it is (urate deposition burden) and, once it is clean, light daily wiping may be enough (serum urate preventive target) from then on to avoid dust (new urate crystal) accumulation and keep it clean (no recurrence).

To explore this, we used data from a prospective followup cohort of patients in whom urate-lowering therapy was withdrawn. The number of patients in the cohort at the time of the present study was twice as high as it was at the time of our previous study (4), allowing us to analyze recurrence as an outcome in patients with serum urate levels ranging from 6 to 6.99 mg/dl during followup after withdrawal of urate-lowering therapy. In our previous study only 3 of the 104 patients in the cohort at that point exhibited average serum urate levels of <7 mg/dl, and they were excluded from analysis in that study.

It could be argued that in some patients, the serum urate level could have been maintained in the 6–6.99 mg/dl range simply by reducing the dosage of their urate-lowering drug. However, that approach may have resulted in greater disadvantage than advantage: xanthine-oxidase inhibitors, such as allopurinol, may act as free radical scavengers (8) and have been reported to reduce the frequency of gout flares even in the presence of hyperuricemia (9).

Our study had some limitations. Some patients did not consent to joint aspiration when recurrence was suspected. However, the number of such patients was small, and their exclusion would probably have little impact on the results. Also, the criterion used to classify a patient as having a disease recurrence, i.e., observation of MSU crystals, may be considered very stringent, but MSU yielded from joint aspiration (10) performs better for this purpose than any previously proposed classification criteria set (11).

Data on our cohort of patients in whom urate-lowering therapy was withdrawn consistently show that the only variables associated with crystal-proven recurrence of gout are serum urate levels during urate-lowering therapy and after therapy withdrawal. This can be explained by the fact that patients continued to receive urate-lowering therapy for 5 years after tophi, if present, had resolved, which makes crystal deposition dependent only on serum urate levels and the time of exposure to hyperuricemia in patients with previous gout. Interestingly, although the number of patients was small and the analysis was done on a post hoc basis, concomitant treatments known to have a positive or negative effect on serum urate levels (12) (namely diuretics, fenofibrate, and losartan) as well as weight change (13) during the observation period seem to have a positive or negative impact on maintaining serum urate levels at <7 mg/dl in >10% of our patients in whom specific urate-lowering therapy had been withdrawn during the whole followup period.

Our data indicate that a 2-stage approach to the treatment of hyperuricemia of gout could be plausible in terms of the target for serum urate levels. The first stage would consist of a period of crystal depletion, with the target serum urate level being far below the threshold for saturation of urate (<6 mg/dl) in order to completely dissolve formed crystals; our results show that a 5-year period of this treatment may be enough for patients without tophi. This would be followed by a second, preventive treatment period in which the target is a serum urate level just below that threshold for saturation of urate, to avoid new crystal formation.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Perez-Ruiz had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Perez-Ruiz, Herrero-Beites.

Acquisition of data. Perez-Ruiz.

Analysis and interpretation of data. Perez-Ruiz, Herrero-Beites, Carmona.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We are grateful to Rosario Lopez Santamaria for coordinating the study, and Inmaculada Iriondo and Begoña Balmaseda for their help during years of daily office work, visiting patients, and collecting data. We also thank all of our patients who gave consent for withdrawal of urate-lowering treatment and attended scheduled visits.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES