Howard A. Fink, MD, MPH; Timothy J. Wilt, MD, MPH; Keith E. Eidman, DO; Pranav S. Garimella, MD, MPH; Roderick MacDonald, MS; Indulis R. Rutks, BS; Michelle Brasure, PhD, MSPH, MLIS; Robert L. Kane, MD; Jeannine Ouellette; Manoj Monga, MD
Disclaimer: This report is based on research conducted by the Minnesota Evidence-based Practice Center under contract to AHRQ, Rockville, Maryland. The findings and conclusions in this document are those of the authors, who are responsible for its content, and do not necessarily represent the views of AHRQ. No statement in this report should be construed as an official position of AHRQ, the U.S. Department of Health and Human Services, or the U.S. Department of Veterans Affairs.
Acknowledgment: The authors thank Marilyn Eells and Maureen Carlyle for technical editing support.
Grant Support: By contract HHSA 290 2007 10064 1 from AHRQ to the Minnesota Evidence-based Practice Center.
Potential Conflicts of Interest: Dr. Fink: Grant (money to institution): AHRQ; Payment for writing or reviewing the manuscript: American College of Physicians; Travel/accommodations/meeting expenses unrelated to activities listed: USPSTF. Dr. Wilt: Grant (money to author and institution): AHRQ. Dr. Eidman: Grant (money to institution): AHRQ. Dr. Brasure: Grant (money to institution): AHRQ. Dr. Monga: Grant (money to institution): AHRQ; Payment for lectures, including service on speakers bureaus: Mission Pharmacal. All other authors have no disclosures. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M12-0883.
Requests for Single Reprints: Howard A. Fink, MD, MPH, Minneapolis Veterans Affairs Medical Center (11-G), One Veterans Drive, Minneapolis, MN 55417; e-mail, firstname.lastname@example.org.
Current Author Addresses: Dr. Fink: VA Medical Center (11-G), One Veterans Drive, Minneapolis, MN 55417.
Dr. Wilt, Mr. MacDonald, and Mr. Rutks: VA Medical Center (111-O), One Veterans Drive, Minneapolis, MN 55417.
Dr. Eidman: Hennepin County Medical Center, Division of Nephrology, 701 Park Avenue, Minneapolis, MN 55415.
Dr. Garimella: Tufts Medical Center, 800 Washington Street, Box 391, Boston, MA 02111.
Drs. Brasure and Kane and Ms. Ouellette: Division of Health Policy and Management, University of Minnesota School of Public Health, D351 Mayo (MMC 197), 420 Delaware Street SE, Minneapolis, MN 55455.
Dr. Monga: Department of Urology, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, 9500 Euclid Avenue, Cleveland, OH 44195.
Author Contributions: Conception and design: H.A. Fink, T.J. Wilt, K.E. Eidman, M. Monga.
Analysis and interpretation of the data: H.A. Fink, T.J. Wilt, K.E. Eidman, R. MacDonald, M. Brasure, M. Monga.
Drafting of the article: H.A. Fink, K.E. Eidman, M. Brasure, R.L. Kane, J. Ouellette, M. Monga.
Critical revision of the article for important intellectual content: H.A. Fink, T.J. Wilt, K.E. Eidman, P.S. Garimella, R. MacDonald, R.L. Kane, J. Ouellette, M. Monga.
Final approval of the article: H.A. Fink, T.J. Wilt, K.E. Eidman, P.S. Garimella, R. MacDonald, I.R. Rutks, R.L. Kane, M. Monga.
Provision of study materials or patients: I.R. Rutks, M. Brasure.
Statistical expertise: T.J. Wilt, R. MacDonald.
Obtaining of funding: H.A. Fink, T.J. Wilt, R.L. Kane, M. Monga.
Administrative, technical, or logistic support: H.A. Fink, T.J. Wilt, I.R. Rutks, M. Brasure, R.L. Kane.
Collection and assembly of data: H.A. Fink, K.E. Eidman, P.S. Garimella, R. MacDonald, I.R. Rutks, M. Brasure, M. Monga.
Fink H., Wilt T., Eidman K., Garimella P., MacDonald R., Rutks I., Brasure M., Kane R., Ouellette J., Monga M.; Medical Management to Prevent Recurrent Nephrolithiasis in Adults: A Systematic Review for an American College of Physicians Clinical Guideline. Ann Intern Med. 2013;158:535-543. doi: 10.7326/0003-4819-158-7-201304020-00005
Download citation file:
Published: Ann Intern Med. 2013;158(7):535-543.
This article has been corrected. The original version (PDF) is appended to this article as a supplement.
Optimum management to prevent recurrent kidney stones is uncertain.
To evaluate the benefits and harms of interventions to prevent recurrent kidney stones.
MEDLINE, Cochrane, and other databases through September 2012 and reference lists of systematic reviews and randomized, controlled trials (RCTs).
28 English-language RCTs that studied treatments to prevent recurrent kidney stones and reported stone outcomes.
One reviewer extracted data, a second checked accuracy, and 2 independently rated quality and graded strength of evidence.
In patients with 1 past calcium stone, low-strength evidence showed that increased fluid intake halved recurrent composite stone risk compared with no treatment (relative risk [RR], 0.45 [95% CI, 0.24 to 0.84]). Low-strength evidence showed that reducing soft-drink consumption decreased recurrent symptomatic stone risk (RR, 0.83 [CI, 0.71 to 0.98]). In patients with multiple past calcium stones, most of whom were receiving increased fluid intake, moderate-strength evidence showed that thiazides (RR, 0.52 [CI, 0.39 to 0.69]), citrates (RR, 0.25 [CI, 0.14 to 0.44]), and allopurinol (RR, 0.59 [CI, 0.42 to 0.84]) each further reduced composite stone recurrence risk compared with placebo or control, although the benefit from allopurinol seemed limited to patients with baseline hyperuricemia or hyperuricosuria. Other baseline biochemistry measures did not allow prediction of treatment efficacy. Low-strength evidence showed that neither citrate nor allopurinol combined with thiazide was superior to thiazide alone. There were few withdrawals among patients with increased fluid intake, many among those with other dietary interventions and more among those who received thiazide and citrate than among control patients. Reporting of adverse events was poor.
Most trial participants had idiopathic calcium stones. Nearly all studies reported a composite (including asymptomatic) stone recurrence outcome.
In patients with 1 past calcium stone, increased fluid intake reduced recurrence risk. In patients with multiple past calcium stones, addition of thiazide, citrate, or allopurinol further reduced risk.
Agency for Healthcare Research and Quality.
Lifetime incidence of kidney stones is approximately 13% for men and 7% for women (1, 2). Although often asymptomatic—incidental stones are identified in approximately 5% of individuals who have abdominal ultrasonography or computed tomography imaging (3, 4)—stones may cause renal colic, urinary tract obstruction, and procedure-related illness. In patients with asymptomatic stones who are followed with serial radiography, 11% to 32% develop symptoms or undergo a procedure within 3 to 4 years (5–7). After a symptomatic stone event, the 5-year recurrence rate is 35% to 50% without specific treatment (8). Annual direct costs in the United States may exceed $4.5 billion (1, 9).
About 80% of kidney stones are composed of calcium oxalate, calcium phosphate, or both; uric acid and struvite stones are less common (10). Many patients with stones have low urine volume or biochemical abnormalities (for example, hypercalciuria, hypocitraturia, hyperoxaluria, hyperuricosuria, or abnormal urine pH) (11, 12). Although low fluid or calcium intake increases stone risk, evidence for many other dietary factors is mixed (13–17). Risk is also increased by certain medical conditions, including primary hyperparathyroidism (18), obesity (19), diabetes (20), gout (21), intestinal malabsorption (22), and anatomical abnormalities.
Given these associations and current understanding of kidney stone physiology, treatments aim to prevent stone recurrence by improving the urinary balance between crystal-forming and crystal-inhibiting substances. Systematic reviews report reduced recurrence with increased fluid intake (23), thiazides (24–26), and citrate pharmacotherapy (26, 27), but evidence is insufficient for efficacy of other pharmacologic treatments (24, 26, 28, 29). However, these reviews did not include more recent randomized, controlled trials (RCTs); compare active treatments, including combination regimens; or evaluate the effect of patient factors on treatment outcomes.
Benefits and harms of treatments to prevent recurrent kidney stones are unclear, as are the effects of patient and stone characteristics and biochemistry measures on treatment outcomes. We conducted this systematic review to evaluate the evidence on these questions and to guide the American College of Physicians (ACP) clinical guideline on medical management to prevent recurrent nephrolithiasis in adults.
We followed a protocol developed with stakeholder input. Appendix Figure 1 shows the analytic framework and key questions used to guide this review. The full technical report, which incorporated peer review and public comments, is available at www.effectivehealthcare.ahrq.gov/search-for-guides-reviews-and-reports/?pageaction=displayproduct&productID=1035.
Analytic framework and key questions.
KQ = key question; OTC = over-the-counter.
We searched MEDLINE and the Cochrane Library (both through September 2012), Google Scholar, ClinicalTrials.gov, and Web of Science databases. We reviewed reference lists of eligible RCTs and relevant systematic reviews as well as articles suggested by experts. Appendix Table 1 shows our complete search strategy.
Appendix Table 1. Literature Search Strategies
We included English-language RCTs that involved dietary or pharmacologic treatment to prevent recurrent kidney stones in adults and reported clinical outcomes (including symptomatic, radiographic, or composite [symptomatic or radiographic] stone recurrence or change in stone size) or harms. Appendix Table 2 shows our detailed eligibility criteria. Two independent reviewers examined titles, abstracts, and full articles for eligibility and resolved discrepancies by discussion and consensus.
Appendix Table 2. Study Eligibility Criteria
For each article, 1 reviewer extracted details on study design, participant characteristics, outcomes, and adverse events, and a second reviewer checked accuracy. Using criteria developed by the Cochrane Collaboration (30), 2 reviewers rated individual study quality as good, fair, or poor on the basis of adequacy of allocation concealment (31), blinding, reporting reasons for attrition, and how analyses accounted for incomplete data (Appendix Table 3). Following methods developed by the Agency for Healthcare Research and Quality (AHRQ)'s Effective Health Care Program (32), 2 reviewers graded the strength of evidence (SOE) for the efficacy of each treatment comparison in preventing stone recurrence on the basis of risk of bias, consistency, directness, and precision (Appendix Tables 4, 5, and 6). We resolved discrepancies in quality ratings and SOE grades by discussion and consensus.
Appendix Table 3. Individual Study Quality
Appendix Table 4. Strength of Evidence Grades and Definitions
Appendix Table 5. Strength of Evidence for Prevention of Stone Recurrence: Dietary Intervention Trials
Appendix Table 6. Strength of Evidence for Prevention of Stone Recurrence: Pharmacologic Intervention Trials
We pooled results if clinical heterogeneity of patient populations, interventions, and outcomes was minimal. Data were analyzed in Review Manager, version 5.1 (The Nordic Cochrane Center, Copenhagen, Denmark). We used random-effects models to generate pooled estimates of relative risks (RRs) and 95% CIs, and we summarized statistical heterogeneity using the I2 statistic (33). When there were few RCTs for a given treatment and no overlap of reported outcomes, we synthesized data qualitatively. For analyses of pharmacologic treatments, we evaluated results by drug class and individual agent. Where data allowed, we explored treatment efficacy according to patient characteristics, stone characteristics, baseline and follow-up biochemical measures, and study duration.
This review was nominated to the AHRQ by the American Urological Association and was funded by the AHRQ. AHRQ staff and a technical expert panel that included members of the ACP Clinical Guidelines Committee and the American Urological Association helped to develop and refine the scope and reviewed the draft AHRQ report. Members of the ACP Clinical Guidelines Committee provided support for manuscript preparation and reviewed drafts of this manuscript. The authors are solely responsible for its content.
We included a total of 28 RCTs (8 dietary and 20 pharmacologic) (Appendix Figure 2). Among these, 23 included only participants with calcium stones, 3 were limited to those with struvite stones, and 2 included those with any stone type. Nearly all studies excluded participants known to have conditions associated with kidney stones. Six trials reported symptomatic stone recurrence, 8 reported radiographic recurrence, and 18 reported a composite recurrence outcome. Treatment duration was 1 to 5 years. We rated 2 trials as good quality (34, 35), 2 as poor quality (36, 37), and the remainder as fair quality (Appendix Table 3). The most common limitations in study quality, which were present in approximately two thirds of trials, were unclear descriptions of allocation concealment and failure to report outcomes according to intention-to-treat principles.
Summary of evidence search and selection.
RCT = randomized, controlled trial; SWL = shock wave lithotripsy.
Two RCTs randomly assigned participants with 1 past calcium stone either to increase fluid intake to maintain urine output of greater than 2 or 2.5 L/d or to receive no treatment (Appendix Table 7). In 1 poor-quality trial, participants who were assigned to increased fluids had a reduced risk for composite stone recurrence compared with no treatment (RR, 0.45 [95% CI, 0.24 to 0.84]) (36); SOE for this outcome was low due to the trial's poor quality and consequent high risk of bias. A fair-quality trial reported reduced risk for recurrent radiographic stones with increased fluid intake, but results did not statistically significantly differ between treatment groups (RR, 0.15 [CI, 0.02 to 1.07]) (38); SOE for this outcome was insufficient because of the small number of stone events. Both trials reported few withdrawals, and neither reported data on adverse events.
Appendix Table 7. Summary of Evidence: Dietary Interventions to Prevent Stone Recurrence
One large fair-quality RCT randomly assigned men with more than 1 past kidney stone of any type and soft-drink consumption greater than 160 mL/d to reduced soft-drink intake or no treatment (39) (Appendix Table 7). Although the intervention significantly reduced the risk for symptomatic recurrent stones (RR, 0.83 [CI, 0.71 to 0.98]), SOE for this outcome was low because results were from only 1 trial. In a subgroup analysis not reported as prespecified, the benefit seemed to be limited to participants whose most frequently consumed soft drink at baseline was acidified solely by phosphoric acid (P = 0.02 for interaction). Total fluid intake was similar in both groups, suggesting that results were explained by type and not amount of fluid intake. The study reported few withdrawals and almost no adverse event data.
One fair-quality RCT of participants with more than 1 past calcium stone provided low-strength evidence that, compared with a control diet of greater than 2 L of water intake plus 800 to 1000 mg of calcium per day, neither increased dietary fiber intake (RR, 1.18 [CI, 0.66 to 2.12]) nor decreased animal protein intake (RR, 1.00 [CI, 0.52 to 1.91]) statistically significantly reduced the risk for recurrent stones (40) (Appendix Table 7). However, because withdrawals exceeded 50% in all groups, robust conclusions cannot be drawn from this trial. Strength of evidence was low for both of these interventions because results could not preclude the exclusion of clinically meaningful benefits or harms. The study reported no data on adverse events.
Three RCTs (1 good-, 1 fair-, and 1 poor-quality) randomly assigned participants with 1 (37, 41) or more (35) past calcium stones to different multicomponent diets or a control diet (Appendix Table 7). The good-quality RCT, in which participants with hypercalciuria were advised to increase fluid intake and avoid excess oxalate, found that those randomly assigned to a diet that included normal to high calcium (1200 mg/d), low animal protein, and low sodium intake had a reduced risk for composite stone recurrence compared with those assigned to a low-calcium diet (400 mg/d) (RR, 0.52 [CI, 0.29 to 0.95]) (35). The fair-quality RCT reported that compared with participants assigned to increased fluid intake and 500 to 600 mg of dairy or supplemental calcium per day, those also assigned to low animal protein, high fiber, increased bran, and low dietary purine intake had an increased risk for composite stone recurrence (RR, 5.88 [CI, 1.39 to 24.92]) (41). Results for the second trial may have been affected by a low recurrence rate in the control group. However, considered with negative results from 1 trial of low animal protein intake (40), these results raise questions about whether the reduced recurrence in the first multicomponent diet trial should be attributed in any way to reduced intake of animal protein. Any treatment benefit may have instead been attributable to reduced dietary sodium or avoidance of low dietary calcium. The poor-quality RCT found that compared with a limited biochemical evaluation and uniform diet recommendations, an extensive evaluation and diet recommendations tailored to the biochemical findings of the participant reduced the risk for composite stone recurrence (RR, 0.32 [CI, 0.14 to 0.74]) (37). This trial did not report separate results for any biochemical abnormality subgroup or tailored diet type. Despite individual study quality differences, SOE for these 3 treatment comparisons was low because, in each case, evidence was available from only 1 trial. There were fewer overall withdrawals (35, 41) and withdrawals due to adverse events (35) in the intervention groups in the trials that reported these results separately by treatment group. None of the trials reported data on specific adverse events.
Six fair-quality RCTs randomly assigned participants with recurrent calcium stones to thiazide or either placebo (42–44) or control (45–47) (Figure and Appendix Table 8). All but 1 trial (45) reported dietary co-interventions in both study groups, most commonly increased fluid and decreased oxalate intake. We found moderate-strength evidence that thiazides decreased risk for composite stone recurrence (RR, 0.52 [CI, 0.39 to 0.69]; 5 trials) (42, 43, 45–47). Although thiazides did not reduce risk for symptomatic recurrence, SOE for this outcome was insufficient due to the small number of recurrent symptomatic stone events and large imprecision of the risk estimate (44). Another trial reported a lower risk for extracorporeal lithotripsy with thiazide than with control (8.0% vs. 26.0%; RR, 0.31 [CI, 0.11 to 0.88]) (45). Results for composite stone recurrence did not seem to differ as a function of study duration (2 vs. ≥3 years) or thiazide type (hydrochlorothiazide, chlorthalidone, or indapamide) or dosing regimen, although no trials studied doses of hydrochlorothiazide or chlorthalidone lower than 50 mg/d and 25 mg/d, respectively, and statistical power to test all of these comparisons was low. Results from 1 trial done in a general practice setting (RR, 0.45 [CI, 0.19 to 1.09]) (43) seemed similar to overall results. Compared with participants in the placebo and control groups, those randomly assigned to receive thiazide were statistically significantly more likely to withdraw for any reason or because of adverse events. Adverse events were inconsistently reported, and no individual event was reported in more than 1 participant in any trial.
Forest plots for risk for composite stone recurrence with pharmacologic treatment versus placebo or control.
MH = Mantel–Haenszel.
Appendix Table 8. Summary of Evidence: Pharmacologic Interventions to Prevent Stone Recurrence
Six RCTs (1 good-quality and 5 fair-quality) randomly assigned participants with predominantly recurrent calcium stones to citrate or either placebo (34, 48) or control (49–52). Four of these trials prescribed increased fluid intake to all study participants (48–50, 52), but no other dietary co-intervention was reported in more than 1 trial. We found moderate-strength evidence that citrates reduced risk for composite stone recurrence (RR, 0.25 [CI, 0.14 to 0.44]; 4 trials) (Figure and Appendix Table 8) (34, 48, 49, 52). Strength of evidence that citrates did not reduce risk for radiographic recurrence was low due to the imprecision of the estimate (50). Although results did not seem to differ as a function of the number of past stone events, citrate type (potassium citrate, potassium–magnesium citrate, or potassium–sodium citrate), or study duration (1 vs. ≥2 years), statistical power to test these comparisons was low (34, 48, 49, 52). Compared with participants in the placebo and control groups, those randomly assigned to receive citrate were statistically significantly more likely to withdraw for any reason or because of adverse events and to report adverse events.
Four fair-quality RCTs randomly assigned participants with recurrent calcium stones to allopurinol or either placebo (53, 54) or control (55, 56) (Figure and Appendix Table 8). Three of these trials prescribed increased fluid intake to all participants (53, 54, 57), and 1 also instructed all participants to use sodium bicarbonate to keep their urine pH above 6.5 (57). Moderate-strength evidence showed that allopurinol reduced risk for composite stone recurrence (RR, 0.59 [CI, 0.42 to 0.84]; 2 trials) (53, 56). Although allopurinol did not statistically significantly reduce risk for either symptomatic or radiographic stone recurrence (53), SOE for these outcomes was low and insufficient, respectively, due to the small number of recurrent stone events and the magnitude of imprecision of the risk estimates (53, 56). Two trials (53, 56) reported that participants randomly assigned to receive allopurinol less frequently withdrew for any reason or because of adverse events. Adverse events were infrequently reported for both treatment groups.
Three fair-quality RCTs randomly assigned participants with recurrent struvite (ammonium–magnesium–phosphate) kidney stones and chronic urinary tract infections to receive acetohydroxamic acid (AHA) or placebo (58–60) (Appendix Table 8). Most participants were considered nonsurgical candidates. Participants in both groups received concomitant required (60) or optional (58, 59) antibiotics, but no trials reported a dietary co-intervention. There was no statistically significant difference between the AHA and placebo groups in risk for symptomatic or radiographic stone recurrence. Strength of evidence for both of these outcomes was insufficient due to the small number of recurrent stone events and large imprecision of the risk estimates (58, 59). Although each trial reported a statistically significant reduction in stone growth with AHA versus placebo (58–60), each trial defined stone growth differently. Withdrawals and adverse events were common and statistically significantly more frequent with AHA, although individual adverse events were inconsistently reported.
One fair-quality RCT randomly assigned participants with recurrent calcium stones to receive either thiazide plus citrate or thiazide monotherapy, reported no dietary co-intervention, and found no between-group difference in risk for composite stone recurrence or extracorporeal lithotripsy (45) (Appendix Table 8). Strength of evidence for the former outcome was low due to the imprecision of the risk estimate. This study reported that no participants withdrew but did not report data on adverse events.
Another fair-quality RCT randomly assigned participants with recurrent calcium stones and hypercalciuria to receive either thiazide plus allopurinol or thiazide monotherapy, reported that all participants were instructed to increase fluid intake and make additional dietary changes, and found no between-group difference in risk for composite stone recurrence (46) (Appendix Table 8). Strength of evidence for this outcome was insufficient due to the small number of recurrent stone events and the large imprecision of the risk estimate. Although participants assigned to combination treatment had no statistically significantly decreased risk for withdrawal for any reason or due to adverse events, no data on adverse events were reported for this group.
We could not determine whether the effect of studied interventions differed by stone type because all 3 AHA trials were limited to participants with struvite stones, 2 trials that included participants with any stone type did not report results as a function of stone type, and the remaining 23 trials limited participation to those with calcium stones. Further, no trials evaluated the effect of any interventions in patients with uric acid or cystine stones.
Results were mixed about whether baseline biochemistry measures predicted treatment effectiveness for reducing stone recurrence risk. In 2 RCTs in patients with calcium stones plus hyperuricosuria (uric acid level >4.76 mmol/d [800 mg/d] in men and >4.43 mmol/d [750 mg/d] in women) (53) or hyperuricemia (uric acid level >356.88 µmol/L [6 mg/dL]) (56), those randomly assigned to allopurinol had a significantly lower risk for composite recurrent stones than those in the control group (RR, 0.59 [CI, 0.42 to 0.84]). However, rates of symptomatic stone recurrence did not seem lower with allopurinol than with control in trials of participants unselected for high urinary or serum uric acid levels (54, 55).
In contrast, other baseline biochemistry measures did not seem to predict efficacy of dietary or pharmacologic treatments compared with control for recurrent stone outcomes. More specifically, baseline urinary calcium levels made no statistically significant difference in the efficacy of increased fluid intake, diet, thiazides, citrate, or allopurinol versus control. We based this observation on comparisons of results from trials that included patients with (46), without (38, 48, 53), or unselected for baseline hypercalciuria (34, 36, 42, 43, 46, 47, 49, 56), as well as analyses adjusted for baseline urinary calcium levels (35). Similarly, baseline urine oxalate levels made no statistically significant difference in the efficacy of increased fluid intake, diet, thiazides, or citrate compared with control; this observation was based on comparisons of results among patient groups with (44), without (38, 42, 47), or unselected for hyperoxaluria (36) and results that were adjusted for baseline urinary oxalate levels (34, 35) or baseline hyperoxaluria (34). Efficacy of citrate treatment for recurrent stone outcomes did not differ between patient groups with (48) or unselected for hypocitraturia (34, 49). Moreover, no RCT data addressed whether the effect of any treatment on risk for recurrent stones differs according to urinary magnesium, phosphate, or potassium level; urine pH; or any measure of urine supersaturation at baseline.
Finally, in 1 RCT, participants randomly assigned to an extensive biochemical evaluation plus diet treatment tailored to their individual biochemistry results had a statistically significantly lower risk for recurrent stones than those assigned a limited evaluation plus uniform diet treatment (37). However, because the trial did not report separate results by biochemical abnormality, we could not isolate the effects of any individual baseline biochemistry measure on treatment outcomes.
No RCTs compared risk for stone recurrence between treatments according to follow-up biochemistry measures or changes from pretreatment biochemistry values. Although RCTs that involved increased fluid intake (36) and a multicomponent diet (35) reported that treatment reduced both urine supersaturation levels and risk for recurrent stones, neither study formally tested whether these outcomes were associated. In contrast, no pharmacologic RCT reported follow-up urine supersaturation levels. Data from both dietary and pharmacologic RCTs suggest that follow-up urinary calcium levels may have limitations as a predictor of treatment efficacy in preventing stone recurrence. Urinary calcium levels were unchanged from baseline in all diet trials that reported these results (35, 40, 41), including the 1 trial that reported treatment benefit (35). In the 4 thiazide trials that reported follow-up urinary calcium levels, 3 reported statistically significant decreases among participants assigned to thiazide but not to control (42, 44, 61) and 1 reported statistically significant decreases among those in both groups (46). These results suggest that reduction in urinary calcium level may be a sensitive but nonspecific predictor of thiazide efficacy in preventing stone recurrence. We could not determine whether decreases in urinary uric acid level or increases in urine pH predict effectiveness of allopurinol in reducing stone recurrence (53).
Few RCTs examined the effect on stone outcomes of modifying individual dietary components. Increased fluid intake more than halved the risk for composite or radiographic stone recurrence and seemed well-tolerated, although SOE for these findings was limited by study quality and size (36, 38). Reducing intake of soft drinks acidified solely by phosphoric acid in men with high intake at baseline modestly decreased risk for recurrent symptomatic stones (39). However, these subgroup results were based on post hoc analyses, and generalizability to other populations is uncertain. Results were inconsistent about whether other dietary interventions added benefit to increased fluid intake. For example, trials that compared diets that included low animal protein intake (alone or as part of a multicomponent diet) with control diets reported reduced stone recurrence (35), no between-group risk difference (40), and increased risk for stone recurrence (41). Even if these discrepant results could be explained in part by discounting the low-quality negative studies, the benefits observed in the single positive study may not have been attributable to low animal protein. Reduction in recurrent stone risk may have been driven by reduced dietary sodium or avoidance of low dietary calcium, but even here the absence of any other trials that compared different dietary sodium or calcium intakes (alone or as part of a multicomponent intervention) raises uncertainty about the benefit of these dietary interventions.
Among pharmacologic treatments, thiazides, citrate, and allopurinol each decreased risk for recurrent calcium stones more than increased fluid intake alone. Although neither the thiazide nor the citrate results seemed to differ as a function of the number of past stone events, drug type or dose, or study duration, statistical power to evaluate these questions was low. Another caveat is that no trials evaluated the efficacy of lower thiazide doses currently used to treat hypertension, so whether these doses similarly reduce recurrence risk is unknown. Limited evidence from RCTs suggests that AHA does not reduce recurrence of symptomatic or radiographic struvite stones, but these trials only addressed management of patients who were not considered surgical candidates. Data directly comparing active pharmacologic treatments to prevent stone recurrence were extremely limited. No trials directly compared thiazide, citrate, or allopurinol monotherapy, and there was only low-strength evidence that addition of citrate (45) or allopurinol (46) to thiazide resulted in no further reduction in stone recurrence risk.
Evidence from RCTs is limited for whether stone composition predicts treatment efficacy in preventing recurrence. No trials enrolled participants with different stone types and reported recurrence outcomes as a function of type, AHA was evaluated only in patients with struvite stones, and all other trials were limited to patients with calcium stones. In addition, no trials examined the effect of any therapy in reducing risk for recurrent uric acid or cystine stones. Because most patients have calcium stones, increasing fluid intake in all patients with kidney stones with or without adding thiazide or citrate therapy might reduce recurrence risk. However, no trials tested this strategy.
Evidence is limited for whether baseline biochemistry measures predict treatment efficacy in preventing stone recurrence. Hyperuricosuria or hyperuricemia may predict reduced risk for recurrent calcium stones with allopurinol treatment. However, because both thiazides (42, 43, 45) and citrates (49) reduced risk for calcium stone recurrence in trials that included at least some patients with hyperuricosuria, and no trials directly compared allopurinol with these agents in patients with high uric acid levels, we do not know whether allopurinol should be the preferred initial therapy in this subgroup of patients. Patients in 1 trial who were randomly assigned to an extensive biochemical evaluation and tailored diet were less likely to have a recurrent stone than those assigned to a limited biochemical evaluation and uniform treatment (37). However, because the study did not report how biochemistry results were defined as abnormal and reported results only between the 2 treatment groups overall, we could not determine to what extent, if any, the biochemistry testing accounted for the between-group difference in outcomes. It seems likely, on the basis of results from other diet trials, that not all tailored diet subgroups contributed to the overall benefit, and some may have been harmful. We identified limited evidence that efficacy in reducing risk for recurrent stones does not differ for any dietary or pharmacologic treatment compared with control among patient groups with, without, unselected for, or adjusted for baseline hypercalciuria, hyperoxaluria, or hypocitraturia. These results are limited because some RCTs did not report information on baseline biochemistry measures, other trials did not report how biochemical abnormalities were defined, and definitions varied in trials that reported biochemical abnormalities. Because any association between biochemical abnormalities and risk for recurrent stones is not likely to be defined by a single threshold and may be continuous (62), the failure of trials to report results as a function of a standardized series of biochemical thresholds is limiting. Nevertheless, these results raise questions about the necessity of measuring baseline urinary biochemistry values in all patients with initial or even recurrent calcium kidney stones.
Although many RCTs reported results of follow-up biochemistry measures, none reported and compared between-treatment outcomes of stone recurrence completely subsequent to and stratified by follow-up biochemistry levels or by changes in these measures from baseline. Results from 2 diet trials suggested that a decrease from baseline in several measures of urine supersaturation may be associated with a reduction in risk for recurrent stones. Future studies testing these follow-up measures as predictors of treatment efficacy are warranted. Data from both dietary and pharmacologic RCTs suggest that follow-up urinary calcium level may be a sensitive but nonspecific predictor of stone recurrence. Reductions in high baseline urinary calcium levels during treatment may be attributable, at least in part, to regression to the mean (63).
The available data limit this review in several ways. First, few trial data existed for some treatment comparisons, owing to few trials and small sample sizes. In these cases, determining whether insufficient evidence for treatment benefit reflects inadequacy of the treatment or limitations in the data may be impossible. This also raises the question of whether, despite our comprehensive search strategy, the results of this review could be affected by publication bias if some unpublished trials were not identified. Second, most trials reported few data on treatment harms, thus limiting our confidence around risk estimates for these outcomes. Third, nearly all trials enrolled adults with idiopathic calcium stones. Therefore, results may not be generalizable to individuals with conditions predisposing them to kidney stones, those with noncalcium kidney stones, or children. Fourth, few trials reported symptomatic stone recurrence as an isolated outcome. Instead, most results were driven by radiographic stone recurrence—at best, a surrogate outcome that clinicians and patients may consider less relevant to their treatment decisions. Fifth, only 1 trial recruited participants from a primary care setting (43). However, because the benefit with thiazide versus control in this study seemed similar to that in trials done in specialty stone centers, the effect of thiazides, at least, may be insensitive to recruitment source. Sixth, this review was limited by inconsistent reporting and categorization of baseline biochemistry measures between trials. Finally, although efficacy of treatment for preventing stone recurrence did not seem to differ as a function of treatment duration among trials lasting at least 1 year, this review had low statistical power to evaluate this question.
In conclusion, we found that increased fluid intake substantially reduced risk for recurrent calcium stones. In men with high soft-drink consumption, decreasing intake reduced recurrent stone risk, although benefit may be limited to those whose most commonly consumed baseline soft drink is acidified solely by phosphoric acid. Results were mixed for the potential benefit of other dietary interventions. In individuals with multiple past calcium stones, most of whom received increased fluid intake as a co-intervention, thiazides, citrates, and allopurinol each further reduced risk for stone recurrence. Other than uric acid level, baseline biochemistry measures did not predict efficacy of any treatment. Withdrawals were low in trials evaluating increased fluid intake; high in long-term trials evaluating other dietary interventions; and variable in pharmacologic trials, although higher than for control for both thiazide and citrate treatment. Adverse event reporting was consistently poor. Existing gaps in RCT evidence may require clinicians to use other sources of evidence to inform their clinical management of patients with kidney stones. Future studies should be designed a priori to collect long-term data on symptomatic stone recurrence and other clinical outcomes; report these efficacy outcomes as a function of patient characteristics, including comorbid conditions and baseline biochemistry measures; and predefine and systematically report adverse events. Additional trials directly comparing different active treatments and combination treatments are warranted.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Nephrology, Guidelines, Nephrolithiasis, Urological Disorders, Prevention/Screening.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only