To evaluate the proposed relationship between persistent reduction of serum urate into the subsaturating range and reduction in the frequency of acute gouty attacks.
To evaluate the proposed relationship between persistent reduction of serum urate into the subsaturating range and reduction in the frequency of acute gouty attacks.
We retrospectively examined data derived from 267 patients who had experienced at least 1 gouty attack before their first visit to our clinic. Serum urate concentration, history of recurrent gouty attacks, and information about antihyperuricemic drug use were collected on each visit for up to 3 years from the first visit of each patient. Data derived from visits >1 year after study entry were subjected to statistical analysis.
When adjusted for baseline serum urate level and the number of gouty attacks prior to study entry, reduction of followup serum urate concentration and antihyperuricemic drug use were each significantly associated with a reduced risk of gouty attacks (odds ratio [OR] 0.42, 95% confidence interval [95% CI] 0.31–0.57; OR 0.22, 95% CI 0.10–0.47, respectively).
The data indicate that reduction of serum urate concentrations to 6 mg/dl or lower will eventually result in a reduced frequency or prevention of future gouty attacks.
Acute gouty arthritis, the most common manifestation of gout, is clearly associated with hyperuricemia (1), which is best defined as extracellular fluid urate supersaturation. Hyperuricemia reflects an enlarged body pool of uric acid and increases the risk for precipitation of monosodium urate crystals in joints, connective tissue, and parenchymal organs, including the kidneys (2). Reduction of the uric acid pool would thus appear most crucial in the management of gout. For this reason, antihyperuricemic drugs, such as allopurinol and uricosuric agents, have been widely prescribed for patients with gout. These drugs effectively reduce serum urate levels, and their use has been shown to improve long-term prognosis in gout (3). Although antihyperuricemic drugs are also widely regarded as useful in reducing the frequency of or preventing acute gouty attacks, only a few reports have supported this contention and even fewer have suggested a target serum urate level for achieving this outcome (4, 5). In this study, we retrospectively analyzed the incidence of recurrent gouty arthritis beginning 1 year after initial evaluation in 267 gout patients. The purpose of this study was to evaluate the proposed relationship between persistent reduction of serum urate into the subsaturating range and reduction in the frequency of acute gouty attacks.
We studied 267 patients with gout who first visited the Institute of Rheumatology, Tokyo Women's Medical University for gout treatment between January 1, 1997 and June 30, 1998 and who attended the clinic for >1 year. By the time of the first visit, all patients had experienced at least 1 attack of gouty arthritis and none were then receiving antihyperuricemic medications. The diagnosis of gout was based on the 1977 criteria proposed by the American College of Rheumatology (formerly American Rheumatism Association) (6). Where possible, diagnosis was definitively confirmed by joint or tophus aspiration and demonstration of monosodium urate crystals by polarized light microscopy. The number of patients with gouty tophi was quite few, as is typical in Japan. Among the 267 patients, 35 patients did not receive antihyperuricemic medication for at least 1 year after the first clinic visit. These patients comprised a no-medication group. The remaining 232 patients received antihyperuricemic medication and comprised a medication group. Decisions regarding treatment with antihyperuricemic medications were made on an individual basis by the physicians responsible for each patient.
The choice of antihyperuricemic medication prescribed for an individual patient was based primarily on daily urinary uric acid excretion, status of renal function, and the presence or absence of urolithiasis, following recommendations previously described (7). Allopurinol (95 patients), benzbromarone (136 patients) (8), or both (1 patient) were prescribed for patients in the medication group. After initiation of treatment, benzbromarone was replaced by allopurinol in 10 patients, probenecid in 2 patients, and sulfinpyrazone in 1 patient and allopurinol was replaced by benzbromarone in 2 patients because of adverse reactions. Drug therapy was started at minimal dosages (allopurinol 100 mg/day; benzbromarone 25 mg/day) in an effort to minimize precipitation of acute gouty arthritis, an adverse effect of antihyperuricemic treatment that is especially common early in treatment if serum urate levels are reduced rapidly (7). Dosing of antihyperuricemic drugs was titrated with serum urate levels and creatinine clearances. Therapeutic target urate levels were not explicitly settled because of the retrospective design of this study. Our consensus of the target levels, however, was 6 mg/dl or below. Ninety-eight percent of patients were prescribed up to 300 mg/day of allopurinol or up to 50 mg/day of benzbromarone. Colchicine or nonsteroidal antiinflammatory agents were prescribed for the purpose of treatment but not of prophylaxis of acute gouty attacks, as is historically typical in Japan. Patients were requested to visit our clinic monthly, at which time the attending physician measured serum urate concentration and recorded the occurrence of gouty attacks in the preceding month.
The characteristics of the patients in each group, including sex, age, interval from the first gouty attack, number of gouty attacks before the first visit, serum urate level, frequency of recurrent gouty attacks, and drugs prescribed and taken, were collected at the first and each subsequent visit. Data were collected for up to 3 years after the first visit, except in 12 patients in the no-medication group who were subsequently treated and whose data collected after initiation of antihyperuricemic medication were excluded from analysis.
The primary variable studied was the incidence of acute gouty arthritis >1 year after each patient's first visit. As mentioned in greater detail in the Discussion section, attacks during the first year were not included in this assessment.
We evaluated the relationship between average serum urate concentration during the whole investigation period and recurrence of gouty attacks by a logistic regression model. The investigation period of was at least 1 year and up to 3 years. Average serum urate concentration was calculated by the trapezoid method considering the serum urate concentration at each visit and the intervals between consecutive visits. The formula for the calculation was
where SU is the serum urate concentration; dayi is the number of days since the first visit of the i-th visit; observation period is the whole investigation period in days. Baseline serum urate level was calculated using the same formula above, in this case defining the observation period as beginning with the first visit and ending, as appropriate, with initiation of antihyperuricemic medication.
Patients in the medication and no-medication groups were further subdivided into attack and no-attack groups on the basis of the occurrence or nonoccurrence of gouty attacks >1 year after the first visit. The relationship between antihyperuricemic medication and the recurrence of gouty attacks was analyzed using a logistic regression model. The mean average serum urate in each subgroup was also analyzed.
All data were analyzed using SAS 8.02 (SAS Institute, Cary, NC).
Of the 267 patients analyzed in this study, 91 experienced at least 1 recurrence of acute gouty arthritis a year or more after the initial visit (attack subgroup); the remaining 176 patients had no recurrences a year or more after the initial visit (no-attack subgroup; Table 1). As discussed in greater detail below, antihyperuricemic drug-treated patients in the no-attack subgroup had a lower mean average serum urate concentration during the whole investigation period than that of treated patients in the attack subgroup (Table 1). There was no association between recurrence of gouty attacks and the type of antihyperuricemic drugs used (P = 0.549, Fisher's exact test). Twenty-five of 93 patients taking allopurinol (26.9%) and 38 of 123 patients taking benzbromarone (30.9%) had recurrent gouty attacks.
|Characteristics||All patients||Medication group||No medication group|
|Attack||No attack||Attack||No attack||Attack||No attack|
|No. of patients (%)||91 (34.1)||176 (65.9)||69 (29.7)||163 (70.3)||22 (62.9)||13 (37.1)|
|Age, mean (range) years||45.5 (24–80)||50.0 (16–92)||45.0 (24–80)||50.2 (16–92)||47.0 (30–69)||47.2 (34–69)|
|Interval from first attack, mean (range) months||58.9 (0.1–295.0)||51.0 (0.2–299.9)||64.5 (0.1–274.6)||53.6 (0.2–299.9)||41.5 (0.7–295.0)||18.9 (0.3–68.3)|
|No. with attacks before first visit <4 / ≥4||59/32||135/41||41/28||124/39||18/4||11/2|
|Baseline serum urate, mean (SE) mg/dl||7.79 (0.14)||7.45 (0.13)||7.80 (0.18)||7.47 (0.13)||7.76 (0.22)||7.17 (0.49)|
|Serum urate during the whole investigation period, mean (SE) mg/dl||7.20 (0.09)||6.46 (0.07)||7.01 (0.10)||6.36 (0.06)||7.75 (0.18)||7.76 (0.23)|
The number of gouty attacks prior to the observation period in the attack subgroup was significantly higher than that in the no-attack subgroup (P = 0.044, Fisher's exact test). The initial serum urate concentration in the attack subgroup was higher than that in no-attack subgroup, although it showed only a trend toward significance (P = 0.068, 2-sample Wilcoxon's test; Table 1). These 2 parameters appeared to influence the likelihood of recurrence of acute gouty attacks 1 year or more after the initial visit. We therefore included both parameters as covariates in the subsequent logistic regression model.
Recurrence of acute gouty attacks was associated with average serum urate concentration during the whole investigation period with an odds ratio (OR) of 0.42 (P < 0.001, 95% confidence interval [95% CI] 0.31–0.57) when average serum urate concentration was adjusted for the 2 covariates, baseline serum urate, and the number of gouty attacks prior to the observation period. Figure 1 shows the clear relationship displayed between average serum urate concentration and the percentage of patients who experienced at least 1 recurrent gouty attack during the observation period.
Overall, average serum urate concentration in the attack subgroup was 7.20 mg/dl (SE 0.09 mg/dl) and that in the no-attack subgroup was 6.46 mg/dl (SE 0.07 mg/dl; Table 1).
Sixty-nine of the 232 patients (29.7%) in the medication group and 22 of the 35 patients (62.9%) in the no-medication group experienced at least 1 recurrent gouty attack during the observation period (Table 1). Adjusted for the 2 covariates, baseline serum urate and the number of gouty attacks prior to the observation period, antihyperuricemic medication decreased the risk of recurrent attacks with an OR of 0.22 (P < 0.001, 95% CI 0.10–0.47).
Among patients in the medication group, the adjusted means of the average serum urate values in attack and no-attack subgroups were 7.01 mg/dl (SE 0.10 mg/dl) and 6.36 mg/dl (SE 0.06 mg/dl), respectively (Table 1).
Figure 2 shows the incidence of recurrent gouty attacks from the initial visit through the observation period in the medication and no-medication groups. Although the incidence of acute gouty arthritis was greater during the first year in the medication than in the no-medication group, the percentage of patients with gouty attacks gradually decreased in the medication group to a level exceeded by that in the no-medication group. An underestimation in later incidence rates among members of the no-medication group seems likely due to the small number of subjects followed, in part because 12 members of this group were withdrawn to recieve treatment of recurrent gouty attacks.
This study demonstrated that the lower the serum urate levels, the less the likelihood of recurrent acute gouty attacks. Recurrence of acute gouty attacks was significantly associated with average serum urate concentration during the whole investigation period. We used the recurrence of acute gouty arthritis during an observation period beginning 1 year after the first visit as our primary measure of the clinical efficacy of antihyperuricemic medication because use of these agents is frequently associated with gouty attacks in the early weeks and months of treatment (7, 9), particularly if prophylactic colchicine or nonsteroidal antiinflammatory drugs are not prescribed (10). We confirmed this phenomenon in our study (Figure 2).
Because the primary aim of treating gout patients with antihyperuricemic drugs is to maintain subsaturating levels for a period sufficient to normalize body uric acid pools, we chose the average serum urate concentration during a lengthy investigation period as an explanatory variable. During long-term administration of these agents, patient compliance with drug administration is often an issue in an asymptomatic individual, and, as with others before us (11), we assumed that poorer compliance was associated with higher serum urate concentrations in this study. We therefore considered average serum urate concentration to be an appropriate explanatory variable in evaluating the relationship of antihyperuricemic treatment and reduction in the recurrence of gouty attacks, regardless of patient compliance with drug. Our results showed that 25% of patients returned to our clinic on a monthly basis and 50% of patients came to our clinic at least once every-other month, on the average. As we assumed, the longer average visit interval trended toward the higher average serum urate level.
In this study, antihyperuricemic drugs effectively reduced the risk of recurrent gouty attacks during the observation period by significantly reducing the average serum urate concentration, with a significant OR of 0.22. The patients in the medication group had experienced more gouty attacks before starting medication and had longer duration of gouty arthritis than patients in the no-medication group. Although these differences were not statistically significant, we believe the group assignments reflected the standard therapeutic strategy of antihyperuricemic medication in gout patients who have experienced frequent versus infrequent gouty attacks. That is, patients who have experienced fewer attacks are less likely to be treated with antihyperuricemic medication. Although the patients in the no-medication group were considered to be at lower risk at the beginning of the investigation period, they nevertheless experienced a higher incidence of recurrent gouty attacks than patients in the medication group. This result demonstrated that maintaining serum urate at subsaturating levels was effective in gout patients and that antihyperuricemic drugs were clearly useful for this purpose.
A target serum urate level to be achieved with antihyperuricemic drug therapy has been controversial (5, 10, 11). In theory, urate levels below ∼7 mg/dl may be considered appropriate because at the sodium concentrations prevailing in extracellular fluids, serum at 37°C is supersaturated for monosodium urate at concentrations >6.8 mg/dl (12). As shown in Table 1, however, the mean average serum urate concentration in the patients in the medication group who experienced recurrent gouty attacks was only 7.01 mg/dl, whereas those in the no-attack subgroup averaged 6.36 mg/dl. This result suggests that 7 mg/dl is not a suitable target level. The logistic analysis shown in Figure 1 demonstrated that the lower the serum urate level, the lower the incidence of recurrent gouty attacks. On the other hand, larger doses of antihyperuricemic drugs are generally required to achieve lower serum urate levels but may increase the risk of adverse events. Excessive doses of allopurinol in patients with renal insufficiency clearly increase the risk of adverse reactions (13). Also, excessive hypouricemia caused by inappropriate usage of uricosuric drugs, such as benzbromarone, may provoke hyperuricosuria, which may, in turn, increase the risk of urinary stone formation in the first 4–8 weeks after the initiation of therapy (14).
Among 81 patients in this study whose average serum urate concentrations were <6.0 mg/dl, all in the medication group, 71 patients (86%) had no recurrent gouty attacks during the observation period. Also, Li-Yu et al (15) recently reported that maintaining serum urate levels <6 mg/dl for several years effectively reduced urate crystal stores in the knee joint synovial fluid (15). In view of these results, we recommend a target serum urate level ≤6.0 mg/dl as an aim of antihyperuricemic therapy for purposes of reducing the incidence of acute arthritic attacks in patients with gout. Our results also suggest that higher serum urate levels may be the risk of the incidence of recurrent gouty attack. Patients who have had high serum urate levels for long duration and have more gouty attack history could be regarded as high-risk patients and should be treated with antihyperuricemic drug.
This recommendation should be further evaluated in prospective clinical studies.
We wish to thank Professor Michael A. Becker (University of Chicago, Chicago, IL) and Dr. Nancy Joseph-Ridge (Lake Forest, IL) for their thoughtful review of the manuscript and valuable suggestions.