Unlike most mammals, humans lack the enzyme uricase, which breaks down uric acid (UA) to allantoin.1 Chronically elevated UA can lead to gout, which affects approximately 3.8% of the US population.2 Animal models3,4 and human studies5,6 suggest that chronic hyperuricemia also has a deleterious effect on renal function, although the impact of urate-lowering therapy on renal function in humans is unclear. Currently, some clinicians use urate-lowering medications, such as allopurinol, “off-label” in patients with very high UA levels, irrespective of gout history.7 Although allopurinol is well tolerated in a majority of patients, potentially lethal hypersensitivity reactions may occur, particularly in patients with chronic kidney disease, and so the benefits of urate-lowering therapy must be weighed against these risks.8 The objective of this study is to determine the effect of allopurinol on kidney function in a male veteran population.
MATERIALS AND METHODS
This is a retrospective cohort study of patients enrolled at the Veterans Administration New York Harbor Healthcare System, identified by pharmacy and laboratory data. The treatment group consisted of 50 consecutive patients with hyperuricemia defined as serum UA greater than 7 mg/dL, newly started on allopurinol for any reason and with evidence of treatment compliance. Treatment compliance was determined by review of medication refills, provider notes, and significant decrease in UA. Most patients were started on allopurinol for gout (mainly clinical diagnosis), but 4 patients received allopurinol for urate nephrolithiasis, and 1 patient received allopurinol for asymptomatic hyperuricemia. Control subjects were patients with hyperuricemia who were not treated with allopurinol. Thirty-three of the control subjects had asymptomatic hyperuricemia. The remaining patients had a history of gout but were not exposed to allopurinol prior to or during the observation period, per primary care. Control patients were matched to the treated cases by race, sex, initial age, and estimated glomerular filtration rate (EGFR). Patients on hemodialysis and with history of prior UA–lowering therapy, and acute kidney failure during the observation period were excluded. Two patients were excluded in the treatment group: one because of deteriorating renal function from gastrointestinal bleed, the other from nephrolithiasis. Of the control subjects, 3 patients were excluded because of nephrolithiasis and dehydration, and 1 patient with unclear cause of acute renal failure. Analysis was from October 2000 to November 2006, at which time there was a change in analyzers in the laboratory, making further comparisons inappropriate. Estimated glomerular filtration rate was calculated using the Modification of Diet in Renal Disease formula, which accounts for age, race, sex, and serum creatinine. t Test was used to compare means between groups, and Fisher exact test and Spearman correlation coefficients were used to test for association, as appropriate.
Posttreatment EGFR, creatinine, and UA were compared between the treatment and the control groups using analysis of covariance, adjusting for the corresponding pretreatment measurement and age. The 2 outcomes were analyzed separately. Robust regression with M-estimation and bi-square weighting was used to adjust for outliers. Interaction term (treatment × pretreatment level) was considered for each of the outcomes to test whether treatment effect depended on pretreatment levels. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary NC). P < 0.05 was considered statistically significant.
There was no significant difference between the control and treatment groups in terms of means of age, initial blood pressure, initial serum UA, or initial EGFR (EGFRi), although hypertension was diagnosed more often in the allopurinol group (Table 1).
Treatment with a mean 221 (SD, 95.9) mg/d dose of allopurinol resulted in a mean UA of 6.4 (SD, 1.4) mg/dL, compared with control mean UA 8.9 (SD, 1.7) mg/dL (Fig. 1). The difference in final UA (UAf) between the treatment and control groups was 2.5 mg/dL (95% confidence interval [CI], 2.0–3.1 mg/dL; P < 0.0001), after adjusting for initial UA (UAi) and age. The mean final creatinine (Crf) levels were 1.36 (SD, 1.01) mg/dL and 1.34 (SD, 0.63) mg/dL in the treatment and control groups, respectively (Fig. 2). The allopurinol-treated patients had a 0.10 mg/dL lower Crf level (95% CI, 0.003–0.20 mg/dL; P = 0.04) than did the control subjects, adjusted for initial creatinine and age. The mean final EGFR (EGFRf) levels were 83.2 mL/min (SD, 36.8) in the treatment group and 74.2 mL/min (SD, 31.0) in the control group (Fig. 3). The allopurinol-treated patients achieved on average 11.9 mL/min higher GFR (95% CI, 4.8–11.9 mL/min; P = 0.01) than did the control group, after adjusting for EGFRi and age. Treatment effect depended on the EGFRi, as indicated by the significant treatment × pretreatment EGFR interaction (P = 0.004). Treatment effect on EGFRf was estimated at 3 different levels of EGFRi and was most significant for higher EGFRi. At EGFRi 90 mL/min, allopurinol-treated patients had a mean improvement of 21.7 mL/min (95% CI, 11.8–31.5 mL/min; P < 0.0001), whereas at EGFRi 75 mL/min, the mean improvement was 14.8 mL/min (95% CI, 7.3–22.2 mL/min; P = 0.0002). At EGFRi 45 mL/min, the treatment effect was not significant, with a mean increase of 1 mL/min (95% CI, −9.3 to 11.3 mL/min; P = 0.84).
Table 2 separates the allopurinol-treated group and control subjects by initial stage of kidney function. The Spearman correlation coefficient for EGFRf and UAf was not significant in the allopurinol group (r = −0.12; P = 0.41), and there was a negative weak correlation in the control group (r = −0.29; P = 0.048). The Spearman correlation for the change in EGFR and the change in UA was not significant in the allopurinol group (r = 0.11; P = 0.48), but had a moderate negative correlation in the control group (r = −0.45; P = 0.002). Thus, the improvement in the allopurinol-treated EGFR had no significant correlation with the drop in UA, and the higher the EGFR, the lower the UA in the control group.
The average length of follow-up was 3.4 (SD, 1.6) years. Adverse events were as follows in the treatment group: 1 episode of elevated liver function tests, diarrhea in 1 patient, nausea in 1 patient, and gout attacks in 2 patients.
There is controversy over whether hyperuricemia is simply a byproduct of reduced glomerular filtration or whether it has a pathogenic role in kidney dysfunction.9 Prior to the development of urate-lowering treatment, renal impairment was reported in up to 40% of gout patients, and death from renal failure occurred in up to 25% of gout patients.10,11 “Gouty nephropathy” was seen in nearly all gout patients on autopsy, which included arteriolar and glomerular sclerosis, interstitial fibrosis, and medullary urate deposits.10 Many large cohort epidemiologic studies5,9,12 have shown an association between elevated UA and worsening of renal function, in patients with normal kidney function and those with chronic kidney disease.
There are many hypotheses for induction of renal failure by UA, including the direct association between hyperuricemia and future development of hypertension and renal urate deposition.13 In the National Health and Nutrition Evaluation Study, serum UA greater than as low as 5.5 mg/dL was associated with a higher risk of hypertension.14 Interestingly, urate nephropathy glomerular arteriolosclerosis is indistinguishable from hypertensive renal disease. When hyperuricemia is induced in study rats, it causes renal injury via renal vasoconstriction and activation of the renal angiotensin system followed by preglomerular smooth muscle proliferation, interstitial fibrosis, and glomerular hypertension.3,4 Even after adequate blood pressure control, these rats continued to develop classic arteriosclerosis renal lesions, suggesting that UA may cause microvascular disease independent of hypertension.3 In our study of patients with mainly mild to moderate kidney disease, treatment with allopurinol resulted in better renal outcomes for patients with higher GFR. There was no correlation between the lowering of UA and the improvement of kidney function, suggesting that allopurinol may help improve renal function via other mechanisms aside from its effects on UA. Perhaps treatment with allopurinol of hyperuricemic patients prior to renal disease leads to better outcomes, because once renovascular and glomerular changes such as fibrosis have occurred, renal damage may progress independent of UA levels and blood pressure control.
The latest American College of Rheumatology guidelines recommend urate-lowering therapy to achieve a minimum serum urate of less than 6 mg/dL, for gout patients with any of the following: tophus/tophi, more than 2 attacks of gout per year, past kidney stones, or stage 2 or worse chronic kidney disease.15 Prior “off-label” recommendations included treating hyperuricemic patients with at least 12 or 13 mg/dL, irrespective of gout history.7 In this study, treatment of patients with even lower mean UA levels resulted in an improvement of renal function, most significantly in those with an overall maintained renal function. It remains to be seen if more aggressive treatment with allopurinol would further substantiate renal function outcomes.
The possible benefit of allopurinol on renal function must be weighed against the risk of Allopurinol hypersensitivity syndrome, a potentially lethal reaction to allopurinol estimated to occur in 0.1% of treated patients.8 There were no significant adverse reactions to allopurinol in this group, despite observation of a population with multiple comorbidities. This suggests that benefits may outweigh risks of allopurinol treatment in certain subsets of patients.
This study has limitations. The retrospective case-control design can introduce sampling error, which was mitigated by using consecutive patients from pharmacy and laboratory data. The proxy for allopurinol adherence was reduction in serum UA and review of medication refills, similar to the medication possession ratio used in other retrospective studies. Patients were not matched for medication use or medical history, although there was no significant difference between the 2 groups in terms of common medical conditions that can affect renal function such as diabetes and hypertension. The results are from a male veteran population at 1 institution. Further studies are needed to ascertain whether the results are generalizable to other populations. The strength of the study is that it reflects outcomes in a real clinical setting with usual care.
1. Ejaz AA, Mu W, Kang DH, et al. Could uric acid have a role in acute renal failure? Clin J Am Soc Nephrol
2. Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia
in the US general population: the National Health and Nutrition Examination Survey 2007–2008. Arthritis Rheum
3. Kang DH, Nakagawa T, Feng L, et al. A role for uric acid in the progression of renal disease. J Am Soc Nephrol
4. Sánchez-Lozada LG, Tapia E, Santamaría J. Mild hyperuricemia
induces vasoconstriction and maintains glomerular hypertension in normal and remnant kidney rats. Kidney Int
5. Iseki K, Oshiro S, Tozawa M, et al. Significance of hyperuricemia
on the early detection of renal failure in a cohort of screened subjects. Hypertens Res
6. Iseki K, Ikemiya Y, Inoue T, et al. Significance of hyperuricemia
as a risk factor for developing ESRD in a screened cohort. Am J Kidney Dis
7. Dincer HE, Dincer AP, Levinson DJ. Asymptomatic hyperuricemia
: to treat or not to treat. Cleve Clin J Med
8. Dalbeth N, Stamp L. Allopurinol
dosing in renal impairment: walking the tightrope between adequate urate lowering and adverse events. Semin Dial
9. Chonchol M, Shlipak MG, Katz R, et al. Relationship of uric acid with progression of kidney disease
. Am J Kidney Dis
10. Talbott JH, Terplan KL. The kidney in gout. Medicine (Baltimore)
11. Berger L, Yü TF. Renal function in gout. IV. An analysis of 524 gouty subjects including long-term follow-up studies. Am J Med
12. Obermayr RP, Temml C, Knechtelsdorfer M, et al. Predictors of new-onset decline in kidney function in a general Middle-European population. Nephrol Dial Transplant
13. Feig DI, Kang DH, Johnson RJ. Uric acid and cardiovascular risk. N Engl J Med
14. Goldstein HS, Manowitz P. Relation between serum uric acid and blood pressure in adolescents. Ann Hum Biol
15. Khanna D, Fitzgerald JD, Khanna PP, et al. 2012 American College of Rheumatology guidelines for management of gout. Part 1: systematic nonpharmacologic and pharmacologic therapeutic approaches to hyperuricemia
. Arthritis Care Res (Hoboken)