- Email: [email protected]
MEDICAL REDUCTION OF STONE RISK IN A NETWORK OF TREATMENT CENTERS COMPARED TO A RESEARCH CLINIC
~~22-5347/98/1605-1629$03.00/0
THE JOURNAL
Vol. 160, 1629-1634, November 1998 Printed in U.S.A.
OF UROLOGY
Copyright 0 1998 by AMERICAN UROLOGICAL ASSOCIATION, INC.
MEDICAL REDUCTION OF STONE RISK IN A NETWORK OF TREATMENT CENTERS COMPARED TO A RESEARCH CLINIC J. LINGEMAN, H. MARDIS, R. KAHNOSKI, D. S. GOLDFARB, S. LACY,* M. GRASSO, S. J. SCHEINMAN, J. H. PARKS,-t>$J. R. ASPLIN'F,$ AND F. L. COEt*$ From the Renal Section, University of Chicago, Chicago, Illinois, Institute for Kidney Stone Research, The Methodist Hospital, Indianapolis, Indiana, The Michigan Medical P. C., Grand Rapids, Michigan, The Urology Center, P. C., Omaha, Urology P. C., Lincoln, Nebraska, and Nephrology Division, Department of Medicine, SUNY Health Science Center, Syracuse, the New York University School of Medicine and the New York Department of Veterans Affairs Medical Center, New York, New York
ABSTRACT
Purpose: We determined whether a network of 7 comprehensive kidney stone treatment centers supported by specialized stone management software and laboratory resources could achieve reductions in urine supersaturation comparable to those in a single research clinic devoted to metabolic stone prevention. Materials and Methods: Supersaturation values for calcium oxalate, calcium phosphate and uric acid in 24-hour urine samples were calculated from a set of kidney stone risk factor measurements made at a central laboratory site for the network and research laboratory for the clinic. Individual results and group outcomes were presented to each center in time sequential table graphics. The decrease in supersaturation with treatment was compared in the network and clinic using analysis of variance. Results: Supersaturation was effectively reduced in the network and clinic, and the reduction was proportional to the initial supersaturation value and increase in urine volume. The clinic achieved a greater supersaturation reduction, higher fraction of patient followup and greater increase in urine volume but the treatment effects in the network were, nevertheless, substantial and significant. Conclusions: Given proper software and laboratory support, a network of treatment centers can rival but not quite match results in a dedicated metabolic stone research and prevention clinic. Therefore, large scale stone prevention in a network system appears feasible and effective. KEY WORDS:kidney calculi, treatment outcome. urine
We1 and othersz-5 have reported excellent treatment results from the execution of modern and well established6 medical stone prevention measures in a dedicated research clinic environment. All of the prevention measures to date rely on decreasing urine supersaturation with respect to the main stone forming salts, which are calcium oxalate, calcium phosphates and uric acid. These measures consist of changes in fluid intake and diet, use of medications, such as thiazide diuretic agents, potassium alkali salts and allopurinol, and cure or control of systemic diseases that increase supersaturation, such as primary hyperparathyroidism, sarcoidosis and enteric hyperoxaluria. Given that stone disease affects 3 to 5% of adults,7 significant prevention requires that the benefits of a dedicated clinic be duplicated widely, and in practices that cannot and need not focus all attention on prevention of stone recurrence. Perhaps in a n ideal setting prevention and urological management of stones would be integrated, so as to provide a comprehensive stone management program. Obstacles to such an ideal setting have always included difficulties with obtaining the needed and complex 24-hour urine measurements required to estimate supersaturation. In addition, it is difficult to compile such large arrays of measurements into a practical and clinically instructive format that includes changes with time. What may be most important is securing
good followup through a record system that promotes continuity of care, and provides a longitudinal view of urine and blood measurements before and during treatment. We have attempted to solve some of these problems through a network of treatment centers that are supported by computer software and management tools needed for continuity, and a laboratory specialized for production of urine supersaturation and other stone risk measurements. In addition, the network setting permits pooling of data across centers, and so normative guidelines for reasonable treatment can be determined. Finally, the network includes 1well established6.8.9 stone research clinic that should, in principle, be a benchmark for what can be done through specialization of resources. We present initial supersaturation treatment responses in a network of prevention centers and contrast them to those of a time parallel cohort treated at the specialized stone research clinic. Overall, results are broadly comparable and excellent, suggesting that this specific kind of network system can be a vehicle for treatment of large numbers of patients. METHODS
Beginning on March 15,1995 patients were enrolled at 1of 7 treatment centers, and a cohort of patients were enrolled at the University of Chicago Kidney Stone Treatment and Evaluation Clinic. We include all results for the 7 centers (the network) and the clinic as of December 21, 1997. Each center evolved a particular style of patient care but subscribed to medical prevention as broadly outlined in conventional sources.6.10 Patients were selected for metabolic evaluation
Accepted for publication June 12, 1998. Financial interest and/or other relationship with Merck. I Financial interest and/or other relationship with Litholink COW. $ Financial interest and/or other relationship with National Institutes of Health and National Institute for Diabetes and Digestive and Kidney Diseases. 1629
T
1630
MEDICAL PREVENTION OF NEPHROLITHIASIS
and long-term treatment based on the clinical judgment of the physicians, and not via a specific protocol. The clinic follows a clinical and laboratory protocol detailed elsewhere that arises from the same general consensus guidelines for metabolic prevention.10-13 This report comprises only patients in the clinic and network with initial and followup laboratory information as our intent is to provide initial results of treatment on supersaturation levels. The 24-hour urine collections were used to assess stone risk.14.15 In general, before treatment 2 urine samples were collected at the centers, 3 a t the clinic and 1 at both sites after some weeks to a few months to assess response. Thereafter repeat samples were obtained at yearly intervals to monitor ongoing treatment. Each urine sample was collected when the patient was on the usual diet, with thymol as preservative. At each center in the network urine volume was measured, quality of the collection was assessed by a nurse, drug use was confirmed and collection time was recorded. A 50 ml. aliquot of the urine was sent to the laboratory used by the network. Serum samples were obtained with the urine samples between 7 and 9 a.m. after fasting for 12 hours, and were analyzed at the core laboratory or a convenient alternative. The serum samples (results not shown) were used mainly to detect primary hyperparathyroidism and hypokalemia during treatment. The clinic used an exactly parallel research laboratory to make the same measurements, except that samples were brought to the laboratory. The 2 laboratories exchanged standard unknowns at regular intervals to remain calibrated with each other for all measurements. We have documented previously that both laboratories performed similarly in providing supersaturation values that corresponded well with stone mineral composition measurements made in the same patients.16.17 Urine measurements included calcium, oxalate, citrate, creatinine, magnesium, phosphorus, sodium, ammonium, chloride, uric acid and pH. Because of aging during transit samples from the network were prepared using a protocol that permitted reconstitution of precipitates.17 When tested this protocol allowed us to retrieve an excellent replica of the original urine after deliberate aging in the laboratory for up to 72 hours at 37,4 and 24C. Subsequently we increased this interval to 96 hours. Laboratory methods have been described previously.17 Supersaturation was calculated for calcium oxalate monohydrate, a calcium phosphate phase which was brushite (calcium monohydrogen phosphate) and dihydrogen urate (uric acid) using computer software.'* Results were presented to physicians using a table graphic Briefly, the z score for each urine measurement was calculated against sex matched normals and stone formers to obtain a pair of distances in standard deviation units. The 2 distances were compared and as the distance of a measurement decreased in relation to the distance from normal, the font of the table entry was made larger and darker in 16 stages of gradation, thus, making abnormal values eyecatching and vivid. Every month sites were sent lists of patients needing followup, lost to followup or not yet fully evaluated. In addition, overall results of treatment and initial evaluation were compiled at approximately quarterly intervals, and presented to the sites to aid them in achieving followup, especially of outlier patients. Each site was shown its results compared to all other sites in the network and clinic. Treatment was under the direction of the individual physician but all followed the general outline of correcting established metabolic problems and recognized environmental risk factors. Briefly, idiopathic hypercalciuria was treated with thiazide diuretic agents and reduced sodium intake, in combinations that depended on the patients. Hyperoxaluria was treated with diet change based on diet lists and clinical history. Hyperuricosuria was treated with reduced intake of beef, chicken and fish, and in a few cases (less than 10 for
xhole series) allopurinol when diet failed and stones were ;urgically active. Low urine citrate was treated with 40 to 60 mEq. potassium citrate salts daily, divided into 2 or 3 doses. Hypokalemia from thiazide was treated with potassium citrate, and not chloride. If potassium wasting was severe, 5 to 15 mg. amiloride were used each morning and diet sodium reduction was attempted in all cases. In all patients greater than 2 1. urine volume was sought with considerable clinical emphasis. Our study did not include cystinuria, struvite stones or primary hyperoxaluria. Less than 10 patients had enteric hyperoxaluria treated with cholestyramine and special diet changes. We did not attempt to reduce the mass of treatments, intervals, dosing and drug types into comparisons between and among sites as the result would be too extensive for 1 presentation. Instead, we show the general supersaturation outcome of these treatments. For analysis we pooled data from the sites and compared results to those from the reference clinic using analysis of variance (ANOVA), regression and t tests executed with conventional computer software. For each patient we calculated the mean values for all urine analytes before and during treatment and supersaturation estimates. Therefore, each patient is represented by 1 pair of values for each analyte, and only once in grand means comparisons between the centers and clinic. RESULTS
Patient characteristics and followup. At the closing date there were 145 men and 86 women enrolled in the clinic cohort, and 780 men and 386 women in the network. Mean patient ages plus or minus standard error of the mean were 45 5 1 and 41 ? 1, and 49 2 1 and 42 5 1 years for men and women in the clinic compared to network, respectively (p not significant). The types of stones in the network were greater than 50% calcium oxalate in 66% of patients, greater than 50%calcium phosphate in 8%,uric acid in any stone in 16%, and greater than 20% but less than 50%calcium phosphate in 7%.17 In the clinic percentages were 68, 7, 16 and 8% of patients, respectively.16 Laboratory followup urine measurements during treatment were made in 102 men (70%)and 58 women (67%)in the clinic compared t o 330 men (42%)and 163 women (42%) in the network (chi-square 38 and 17, men and women, respectively, p <0.001 for both sexes). Patients with followup were in the program longer as expected (clinic 19.6 compared to 13.4 months with and without followup, respectively, p <0.001 and network 14.2 compared to 9.3 months, respectively, p <0.001). The significant differences in percentage of patients with followup in the clinic compared to network reflected in part the shorter duration in the program. The fraction of patients with followup was strongly correlated with time in the program (F-21, p <0.001). By ANOVA the slopes of followup fraction compared to time in program were not homogeneous, that is they differed between the clinic and network (p <0.001 for comparison). Even so, allowing for the interaction of slope and site (slope in the clinic compared to network), and covariance of followup with time in the program, the corrected fractions of patients with followup were 63 2 4 compared to 42 t 1 (p <0.001).In other words, the network did not do as well as the clinic in achieving followup, even after allowing as much as possible for the fact that it more recently enrolled more patients and, therefore, followup time was shorter. Thus, the fraction of patients with followup was higher in the clinic than network. Supersaturation response to treatment. Calcium Oxalate: Although initial values of supersaturation were lower in the network than in the clinic, mean decrease in supersaturation was lower in the network, and so treatment supersaturation values did not differ (table 1). The decrease in supersat-
1631
MEDICAL PREVENTION OF NEPHROLITHIASIS TABLE1. Univariate comparisons of principal values between sites Clinic Mean Before Treatment
2
Network Mean
SD
During Treatment
Change
Before Treatment
2
SD
During Treatment
Change
Supersaturation calcium 9.6 i 0.2 6.8 2 0.2 -2.8 2 0.2 8.8 5 0.147.2 + 0.21 -1.6 t 0.2t oxalate monohydrate -0.07 t 0.05 Supersaturation brushite 1.5 i 0.05 1.25 t 0.04 -0.23 + 0.06 1.3 t 0.04 1.28 2 0.06 -0.48 ? 0.05* 0.76 i 0.04i Supersaturation uric acid 1.22 5 0.04 0.55 2 0.03 -0.66 5 0.04 1.24 5 0.03 0.34 t 0.041 Vol. (1./24 hrs.) 1.76 2 0.04 2.28 + 0.05 0.52 2 0.04 1.78 2 0.03 2.12 2 0.05 Change is mean change between before and after treatment. Significance values are t tests. All values of change are significant against 0 at p <0.001 except for network change in supersaturation brushite, which is not significant at p = 0.10. * For difference from clinic p <0.05. t For difference from clinic p <0.01.
TABLE2. Multivariate adjusted means of supersaturation change uration was proportional to the initial supersaturation value by site (fig. 1).Values for the regression coefficients were significant in the network and clinic (p <0.001 for both) and slopes Clinic Mean t SD Network Mean 2 SD differed when tested for homogeneity (fig. 1). Increase in Change in calcium oxalate -2.2 2 0.2 -1.8 2 0.1 urine volume correlated with decrease in supersaturation monohydrate supersaturation (r = 0.64 and 0.695 for the network and clinic, respectively, -0.12 2 0.06 -0.11 5 0.04 Change in brushite superp <0.001 for both). The increase in urine volume was greater saturation in the clinic than network (table 1).The effect of increase in -0.50 2 0.03* Change in uric acid super-0.63 5 0.04 urine volume can be seen in figure 1,with symbol size scaled saturation to volume increase. Large circles tend t o fall t o the bottom of Values are mean change in supersaturation adjusted for change in urine volume and initial supersaturation using ANOVA. the graph. * For difference from clinic p = 0.019. Given that change in urine volume and initial supersaturation influenced the decrease in supersaturation, and that the sites differed in these covariates as well as in the decrease in supersaturation, we asked if a site difference re- nificant against 0 for the clinic but not the network, and mained once the effects of these 2 covariates were considered decrease in supersaturation did not differ between the clinic (table 2). Least square means after adjustment for change in and network (table 1).Decrease in supersaturation was provolume and initial supersaturation values were no longer portional to initial supersaturation (fig. 2) as with calcium significantly different (p = 0.131). Therefore, we concluded oxalate supersaturation (fig. 11, and the regression coeffithat the better performance of the clinic may well have re- cients for the clinic and network were significant and differed flected a greater increase in urine volume, and patients with significantly (fig. 2). Decrease in supersaturation was highly dependent on change in urine volume (p <0.001 for regresgreater initial supersaturation values. Calcium Phosphate: Initial and treatment values of super- sion at both sites), and large circles predominate at the saturation were not significantly different between the clinic bottom of distributions (fig. 2). When change in volume and and network (table 1).Change in supersaturation was sig- initial supersaturation were considered, adjusted least
B
1
0
10
0
0
0
0
0 0
0
0
L"
10 20 30 Pre Rx CaOx SS
40
0
10 20 30 Pre Rx CaOx SS
40
FIG. 1. Change in calcium oxalate monohydrate (CaOx)supersaturation ( S S )(Yaxis)versus before (Pre)treatment (Rr)supersaturation
(xaxis) for clinic (A) and network (B).Change in supersaturation was strongly inverse to initial supersaturation. Increased urine volume,
shown by size of symbol, was most marked when supersaturation decrease was greatest as shown by clustering of large circles at bottom of distributions. Values for regression coefficients were -0.51 versus -0.66 for network and clinic (p <0.001 each). Homogeneity of slopes was not present (p <0.01 for interaction of site and slope).
1632
MEDICAL PREVENTION OF N E P H R O L I T H M I S
7
0
1
2 3 4 5 Pre Rx CaP SS
0
6
1
2 3 4 5 Pre Rx CaP SS
6
FIG. 2. Change in brushite ( C a p )su ersaturation (SS) (Y axis) versus before (Pre)treatment (Rr)supersaturationjX axis) for clinic (A) and network (B). Plotting is same as in 8gure 1. Regression coefficients were -0.409 versus -0.663 for network and cllmc (p <0.001 for both). Homogeneity of slopes was not present (p <0.001).
square means for change in supersaturation did not differ (p = 0.92) between the network and clinic (table 2). Unlike calcium oxalate supersaturation, decreasing values were not strong in any treatment setting although dependence on urine volume and initial supersaturation values were strong. Uric Acid: Before treatment values of supersaturation were not different between the network and clinic (table 1 and fig. 3). Treatment reduced supersaturation significantly in the network and clinic but mean reduction in supersatu-
ration was greater in the clinic (table 1).Decrease in supersaturation was obviously dependent on the initial value, and individual regressions were highly significant, and not different between the network and clinic (fig. 3). Volume increase was highly correlated with decrease in supersaturation even when initial supersaturation was considered (p values for regression of supersaturation decrease in change of volume were <0.001 at the network and clinic). When decrease in supersaturation was adjusted for initial super-
0-----
-
0
I 1
2 3 4 5 Pre Rx Uric Acid SS
6
0
1
2 3 4 5 Pre Rx Uric Acid SS
6
FIG. 3. Change in uric acid (CIA)supcrsaturation (SS)(Y axis) versus before (Pre)treatment (Rr)supersaturation (X axis) for clinic (A)and network (B). Plotting is same us in figure 1. Regression coefficients were -0.572 versus -0.679 for network and clinic (p <0.001 for both). Slopes wcrc homogrncous.
1633
MEDICAL PREVENTION OF NEPHROLITHIASIS
saturation and change of volume (table 21, an effect of site could still be detected, meaning that the greater decrease in supersaturation by univariate analysis (table 1)was not entirely explained by differences in initial supersaturation and change in urine volume. The decrease occurred in large part from a greater increase in urine pH in the clinic (not shown) as should be expected.
DISCUSSION
Our impression is that significant treatment effects were present and broadly comparable in the clinic and 7 network sites. Treatment led to important reductions in supersaturation with respect to calcium oxalate and uric acid. The decrease in supersaturation depended on the initial value and increase in urine volume, as one would expect. Decreases in calcium phosphate supersaturation were weak in the network and clinic, and actually insignificant in the network. Because supersaturation is the main established force driving crystallization9 and established stone preventions act at least in part via decreased supersaturation, the decreases we observed strongly support the idea that the network will eventually decrease stone recurrence, as has been reported from the clinic,1.14,15,19.20 For calcium oxalate and uric acid supersaturation decreases were excellent in the network, although less vibrant than in the clinic. Weaker network decreases were traceable mainly to a lesser increase in urine volume and to a lower pretreatment supersaturation value for calcium oxalate. That a network of distributed treatment centers can reduce supersaturation effectively is not surprising but has never been demonstrated to our knowledge. Also not surprising is that a dedicated research clinic can do better in reducing supersaturation. However, despite differences, both environments produced treatment supersaturation values for calcium oxalate below 7.5, which is closer to the normal value than to the levels found among stone forming patients,16 and below 1.0 for uric acid, a normal range. Given these results our model for treatment of large numbers of patients seems secure so far. For calcium phosphate supersaturation neither treatment environment was remarkably effective, and they did not differ by simple t tests or after allowing for covariation of supersaturation reduction with urine volume change and initial supersaturation values. However, the change in supersaturation was significant compared to 0 in the clinic and not in the network. This finding points t o a weakness of available treatments or lack of effective strategies for use of treatments that seems universal within our system of treatment centers. In the clinic and network calcium phosphate supersaturation elevation correlated with the amounts of calcium phosphate solid phases in stones,16.17 and so the I failure t o reduce calcium phosphate supersaturation has a theoretical potential to maintain stone formation through calcium phosphate nucleation.g.21 However, this potential has not yet been explored clinically. Although treatment effects in the network rival those of the clinic the fraction of patients who had followup urine samples was significantly smaller in the network. The natural explanation, that entry of new patients is much higher in the network than in the clinic and, therefore, mean time of network patients in the program was shorter, is not sufficient. Even after allowing for time in the program a lower percentage of network patients were followed to assess treatment response. We have no ready mechanism to explain this difference and decline to speculate. However, it is a crucial problem that will require substantial remedy in the long k m . The patients who did not have followup did not contribute t o our data on treatment outcome, and SO the clinic and network report alike patients compliant enough to have
followup. Patients who did not have followup may well have achieved no treatment effects, which is a moot point. Our reliance on supersaturation as a principal index of treatment occurred as a result of a broad consensus that supersaturation is the most direct available expression of crystallization potential that one can obtain in clinical practice.16,22Supersaturation is the usual physical chemical expression of crystallization force.9 Measured in urine it is a strong correlate of stone mineral composition in the clinic16.17 and network.17 Available medical prevention to date acts at least in part by decreasing urine supersaturation, which includes reversal of hypercalciuric states using thiazide,23.2* low sodium diet25 and, when needed, removal of parathyroid glands in primary hyperparathyr~idism.~ Prevention also includes treatment of low urine citrate by potassium citrate,5 acid urine pH with alkali salts,26hyperuricosuria and hyperoxaluria by diet,27 hyperuricosuria with allopurin013~2~.2S and all urine supersaturation values with increased urine flow.29 Supersaturation is not a complete or wholly sufficient statement of stone risk as inhibitorss.9 and promoters, and family history of stones are also important,3O but it is the most compact and powerful present available measurement of stone potential that one can manipulate through treatment. With regard to a final gold standard, one must wait for some years to pass and count the formation of new stones. We shall no doubt do that. However, in the interim it is important and responsible to assure ourselves that supersaturation has decreased, because if it does not one has no a priori reason to believe that treatments will be effective in stone prevention. For example we have shown not only that stone mineral composition is highly predicted by urine supersaturation measurements16 but also that treatments that prevented stones regularly decreased relevant supersaturation values in the clinic. Others have shown similar results.g.31.32 That our results show a decrease in supersaturation gives considerable confidence in the network arrangement. The many difficulties of a central laboratory, sample collection in diverse settings, reporting and maintaining a time sequential grasp of urine chemistry data, even interested in Drevention and willing those of keeDing - - Datients _ to obtain followup measurements such as those we prese; here, offer the kinds of challenge that, when successfully overcome, represent a kind of minor clinical triumph.
REFERENCES
1. Parks, J. H. and Coe, F. L.: An increasing number of calcium oxalate stone events worsens treatment outcome. Kidney Int., 4 5 1722,1994. 2. Ettinger, B., Citron, J. T., Livermore, B. and Dolman, L. I.: Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J. Urol., 1 3 9 679,1988. 3. Ettinger, B., Tang, A,, Citron, J. T., Livermore, B. and Williams, T.: Randomized trial of allopurinol in the prevention of calcium oxalate calculi. New Engl. J. Med., 3 1 5 1386,1986. 4. Pak, C. Y. C., Fuller, C., Sakhaee, K., Preminger, G. M. and Britton, F.: Long-term treatment of calcium nephrolithiasis with potassium citrate. J. Urol., 134: 11, 1985. 5. Barcelo, P., Wuhl, O., Servitge, E., Rousaud, A. and Pak, C. Y. C.: Randomized double-blind study of potassium citrate in idiopathic hypocitraturic calcium nephrolithiasis. J. Urol., 1 5 0 1761,1993. 6. Coe, F. L., Parks, J. H. and Asplin, J. R.: The pathogenesis and treatment of kidney stones. New Engl. J. Med., 327: 1141, 1992. 7. Smith, L. H.: The medical aspects of urolithiasis: an overview. J. Urol., 141: 707, 1989. 8. Asplin, J. R., Parks, J. H. and Coe, F. L.: Dependence of upper limit of metastability on supersaturation in nephrolithiasis. Kidney Int., 5 2 1602,1997. 9. Coe, F. L., Favus, M. J., Pak, C. Y. C., Parks, J. H. and Preminger, G. M.: Kidney Stones: Medical and Surgical Management. Philadelphia: Lippincott-Raven Publishers, 1996.
1634
MEDICAL PREVENTION O F NEPHROLITHIASIS
ogy and treatment of nephrolithiasis: new research venues. 10. Asplin, J. R., Favus, M. J. and Coe. F. L.: Nephrolithias. In: The J. Bone Min. Res., 12: 522, 1997. Kidney, 5th ed. Phladelphia: W. B. Saunders Co., pp. 189323. Coe, F. L.: Treated and untreated recurrent calcium nephrolithi1935, 1996. asis in patients with idiopathic hypercalciuria, hyperuricos11. Coe, F. L. and Favus, M. J.: Disorders of Bone and Mineral uria, or no metabolic disorder. Ann. Intern. Med., 87: 404, Metabolism. New York: Raven Press, 1992. 1977. 12. Favus, M. J., Berelowitz, M. and Coe, F. L.: Effects of somatostatin on intestinal calcium transport in the rat. Amer. J . Physiol. 24. Coe, F. L., Parks, J. H., Bushinsky, D. A., Langman, C. B. and Favus, M. J.: Chlorthalidone promotes mineral retention in 241: G215, 1981. patients with idiopathic hypercalciuria. Kidney Int., 3 3 1140, 13. Consensus conference. Prevention and treatment of kidney stones. J.A.M.A., 260: 977, 1988. 1988. 14. Parks, J. H. and Coe, F. L.: The financial effects of kidney stone 25. Kok, D. J., Iestra, J . A,, Doorenbos, C. J. and Papapoulos, S. E.: The effects of dietary excesses in animal protein and in sodium prevention. k d n e y Int., 50: 1706, 1996. on the composition and the crystallization kinetics of calcium 15. Coe, F. L. and Parks, J. H.: Clinical approach. In: Nephrolithiaoxalate monohydrate in urines ofhealthy men. J. Clin. Endocr. sis: Pathogenesis and Treatment, 2nd ed. Chicago: Year Book Metab., 71: 861, 1990. Medical Publishers, 1988. 16. Parks, J. H., Coward, M. and Coe, F. L.: Correspondence be- 26. Pak, C. Y. C., Sakhaee, K. and Fuller, C.: Successful management of uric acid nephrolithiasis with potassium citrate. Kidtween stone composition and urine supersaturation in nephney Int., 3 0 422, 1986. rolithiasis. Kidney Int., 51: 894, 1997. 17. Asplin, J . R., Parks, J . , Lingeman, J., Kahnowski, R., Mardis, H., 27. Coe, F. L., Moran, E. and Kavalich, A. G.: The contribution of dietary purine over-consumption to hyperpuricosuria in calLacey, S., Goldfarb, D., Grasso, M. and Coe, F.: Supersaturacium oxalate stone formers. J . Chronic Dis., 2 9 793, 1976. tion and stone composition in a network of dispersed treat28. Coe, F. L. and Raisen, L.: Allopurinol treatment of uric-acid ment sites. J . Urol., 159: 1821, 1998. 18. Werness, P. G., Brown, C. M., Smith, L. H. and Finlayson, B.: disorders in calcium-stone formers. Lancet, 1: 129, 1973. EQUIL2: a basic computer program for the calculation of uri- 29. Pak, C. Y. C., Sakhaee, K., Crowther, C. and Brinkley, L.: Evinary saturation. J. Urol., 134: 1242. 1985. dence justifying a high fluid intake in treatment of nephroli19. Strauss. A. L., Coe, F. L., Deutsch, L. and Parks, J . H.: Factors thiasis. Ann. Intern. Med., 9 3 36, 1980. that predict relapse of calcium nephrolithiasis during treat- 30. Curhan, G. C., Willett, W. C., Rimm, E. B. and Stampfer, M. J.: ment: a prospective study. h e r . J . Med., 72: 17, 1982. Family history and risk of kidney stones. J. Amer. SOC.Neph20. Parks, J. H., Coe. F. L. and Strauss, A. L.: Calcium nephrolithirol., 8 1568, 1997. asis and medullary sponge kidney in women. New Engl. 31. Tiselius, H. G.: A simplified estimate of the ion-activity product J. Med., 306: 1088, 1982. of calcium phosphate in urine. Eur. Urol., 1 0 191, 1984. 21. Herring, L. C.: Observations on the analysis of ten thousand 32. n u n , M. J. and Schober, S.: Urolithiasis in Tennessee: a n occuurinary calculi. J. Urol.. 88:545, 1962. pational window into a regional problem. Amer. J. Public 22. Coe, F. L. and Parks. J . H.: New insights into the pathophysiolHealth, 81: 587, 1991.
THE JOURNAL
Vol. 160, 1629-1634, November 1998 Printed in U.S.A.
OF UROLOGY
Copyright 0 1998 by AMERICAN UROLOGICAL ASSOCIATION, INC.
MEDICAL REDUCTION OF STONE RISK IN A NETWORK OF TREATMENT CENTERS COMPARED TO A RESEARCH CLINIC J. LINGEMAN, H. MARDIS, R. KAHNOSKI, D. S. GOLDFARB, S. LACY,* M. GRASSO, S. J. SCHEINMAN, J. H. PARKS,-t>$J. R. ASPLIN'F,$ AND F. L. COEt*$ From the Renal Section, University of Chicago, Chicago, Illinois, Institute for Kidney Stone Research, The Methodist Hospital, Indianapolis, Indiana, The Michigan Medical P. C., Grand Rapids, Michigan, The Urology Center, P. C., Omaha, Urology P. C., Lincoln, Nebraska, and Nephrology Division, Department of Medicine, SUNY Health Science Center, Syracuse, the New York University School of Medicine and the New York Department of Veterans Affairs Medical Center, New York, New York
ABSTRACT
Purpose: We determined whether a network of 7 comprehensive kidney stone treatment centers supported by specialized stone management software and laboratory resources could achieve reductions in urine supersaturation comparable to those in a single research clinic devoted to metabolic stone prevention. Materials and Methods: Supersaturation values for calcium oxalate, calcium phosphate and uric acid in 24-hour urine samples were calculated from a set of kidney stone risk factor measurements made at a central laboratory site for the network and research laboratory for the clinic. Individual results and group outcomes were presented to each center in time sequential table graphics. The decrease in supersaturation with treatment was compared in the network and clinic using analysis of variance. Results: Supersaturation was effectively reduced in the network and clinic, and the reduction was proportional to the initial supersaturation value and increase in urine volume. The clinic achieved a greater supersaturation reduction, higher fraction of patient followup and greater increase in urine volume but the treatment effects in the network were, nevertheless, substantial and significant. Conclusions: Given proper software and laboratory support, a network of treatment centers can rival but not quite match results in a dedicated metabolic stone research and prevention clinic. Therefore, large scale stone prevention in a network system appears feasible and effective. KEY WORDS:kidney calculi, treatment outcome. urine
We1 and othersz-5 have reported excellent treatment results from the execution of modern and well established6 medical stone prevention measures in a dedicated research clinic environment. All of the prevention measures to date rely on decreasing urine supersaturation with respect to the main stone forming salts, which are calcium oxalate, calcium phosphates and uric acid. These measures consist of changes in fluid intake and diet, use of medications, such as thiazide diuretic agents, potassium alkali salts and allopurinol, and cure or control of systemic diseases that increase supersaturation, such as primary hyperparathyroidism, sarcoidosis and enteric hyperoxaluria. Given that stone disease affects 3 to 5% of adults,7 significant prevention requires that the benefits of a dedicated clinic be duplicated widely, and in practices that cannot and need not focus all attention on prevention of stone recurrence. Perhaps in a n ideal setting prevention and urological management of stones would be integrated, so as to provide a comprehensive stone management program. Obstacles to such an ideal setting have always included difficulties with obtaining the needed and complex 24-hour urine measurements required to estimate supersaturation. In addition, it is difficult to compile such large arrays of measurements into a practical and clinically instructive format that includes changes with time. What may be most important is securing
good followup through a record system that promotes continuity of care, and provides a longitudinal view of urine and blood measurements before and during treatment. We have attempted to solve some of these problems through a network of treatment centers that are supported by computer software and management tools needed for continuity, and a laboratory specialized for production of urine supersaturation and other stone risk measurements. In addition, the network setting permits pooling of data across centers, and so normative guidelines for reasonable treatment can be determined. Finally, the network includes 1well established6.8.9 stone research clinic that should, in principle, be a benchmark for what can be done through specialization of resources. We present initial supersaturation treatment responses in a network of prevention centers and contrast them to those of a time parallel cohort treated at the specialized stone research clinic. Overall, results are broadly comparable and excellent, suggesting that this specific kind of network system can be a vehicle for treatment of large numbers of patients. METHODS
Beginning on March 15,1995 patients were enrolled at 1of 7 treatment centers, and a cohort of patients were enrolled at the University of Chicago Kidney Stone Treatment and Evaluation Clinic. We include all results for the 7 centers (the network) and the clinic as of December 21, 1997. Each center evolved a particular style of patient care but subscribed to medical prevention as broadly outlined in conventional sources.6.10 Patients were selected for metabolic evaluation
Accepted for publication June 12, 1998. Financial interest and/or other relationship with Merck. I Financial interest and/or other relationship with Litholink COW. $ Financial interest and/or other relationship with National Institutes of Health and National Institute for Diabetes and Digestive and Kidney Diseases. 1629
T
1630
MEDICAL PREVENTION OF NEPHROLITHIASIS
and long-term treatment based on the clinical judgment of the physicians, and not via a specific protocol. The clinic follows a clinical and laboratory protocol detailed elsewhere that arises from the same general consensus guidelines for metabolic prevention.10-13 This report comprises only patients in the clinic and network with initial and followup laboratory information as our intent is to provide initial results of treatment on supersaturation levels. The 24-hour urine collections were used to assess stone risk.14.15 In general, before treatment 2 urine samples were collected at the centers, 3 a t the clinic and 1 at both sites after some weeks to a few months to assess response. Thereafter repeat samples were obtained at yearly intervals to monitor ongoing treatment. Each urine sample was collected when the patient was on the usual diet, with thymol as preservative. At each center in the network urine volume was measured, quality of the collection was assessed by a nurse, drug use was confirmed and collection time was recorded. A 50 ml. aliquot of the urine was sent to the laboratory used by the network. Serum samples were obtained with the urine samples between 7 and 9 a.m. after fasting for 12 hours, and were analyzed at the core laboratory or a convenient alternative. The serum samples (results not shown) were used mainly to detect primary hyperparathyroidism and hypokalemia during treatment. The clinic used an exactly parallel research laboratory to make the same measurements, except that samples were brought to the laboratory. The 2 laboratories exchanged standard unknowns at regular intervals to remain calibrated with each other for all measurements. We have documented previously that both laboratories performed similarly in providing supersaturation values that corresponded well with stone mineral composition measurements made in the same patients.16.17 Urine measurements included calcium, oxalate, citrate, creatinine, magnesium, phosphorus, sodium, ammonium, chloride, uric acid and pH. Because of aging during transit samples from the network were prepared using a protocol that permitted reconstitution of precipitates.17 When tested this protocol allowed us to retrieve an excellent replica of the original urine after deliberate aging in the laboratory for up to 72 hours at 37,4 and 24C. Subsequently we increased this interval to 96 hours. Laboratory methods have been described previously.17 Supersaturation was calculated for calcium oxalate monohydrate, a calcium phosphate phase which was brushite (calcium monohydrogen phosphate) and dihydrogen urate (uric acid) using computer software.'* Results were presented to physicians using a table graphic Briefly, the z score for each urine measurement was calculated against sex matched normals and stone formers to obtain a pair of distances in standard deviation units. The 2 distances were compared and as the distance of a measurement decreased in relation to the distance from normal, the font of the table entry was made larger and darker in 16 stages of gradation, thus, making abnormal values eyecatching and vivid. Every month sites were sent lists of patients needing followup, lost to followup or not yet fully evaluated. In addition, overall results of treatment and initial evaluation were compiled at approximately quarterly intervals, and presented to the sites to aid them in achieving followup, especially of outlier patients. Each site was shown its results compared to all other sites in the network and clinic. Treatment was under the direction of the individual physician but all followed the general outline of correcting established metabolic problems and recognized environmental risk factors. Briefly, idiopathic hypercalciuria was treated with thiazide diuretic agents and reduced sodium intake, in combinations that depended on the patients. Hyperoxaluria was treated with diet change based on diet lists and clinical history. Hyperuricosuria was treated with reduced intake of beef, chicken and fish, and in a few cases (less than 10 for
xhole series) allopurinol when diet failed and stones were ;urgically active. Low urine citrate was treated with 40 to 60 mEq. potassium citrate salts daily, divided into 2 or 3 doses. Hypokalemia from thiazide was treated with potassium citrate, and not chloride. If potassium wasting was severe, 5 to 15 mg. amiloride were used each morning and diet sodium reduction was attempted in all cases. In all patients greater than 2 1. urine volume was sought with considerable clinical emphasis. Our study did not include cystinuria, struvite stones or primary hyperoxaluria. Less than 10 patients had enteric hyperoxaluria treated with cholestyramine and special diet changes. We did not attempt to reduce the mass of treatments, intervals, dosing and drug types into comparisons between and among sites as the result would be too extensive for 1 presentation. Instead, we show the general supersaturation outcome of these treatments. For analysis we pooled data from the sites and compared results to those from the reference clinic using analysis of variance (ANOVA), regression and t tests executed with conventional computer software. For each patient we calculated the mean values for all urine analytes before and during treatment and supersaturation estimates. Therefore, each patient is represented by 1 pair of values for each analyte, and only once in grand means comparisons between the centers and clinic. RESULTS
Patient characteristics and followup. At the closing date there were 145 men and 86 women enrolled in the clinic cohort, and 780 men and 386 women in the network. Mean patient ages plus or minus standard error of the mean were 45 5 1 and 41 ? 1, and 49 2 1 and 42 5 1 years for men and women in the clinic compared to network, respectively (p not significant). The types of stones in the network were greater than 50% calcium oxalate in 66% of patients, greater than 50%calcium phosphate in 8%,uric acid in any stone in 16%, and greater than 20% but less than 50%calcium phosphate in 7%.17 In the clinic percentages were 68, 7, 16 and 8% of patients, respectively.16 Laboratory followup urine measurements during treatment were made in 102 men (70%)and 58 women (67%)in the clinic compared t o 330 men (42%)and 163 women (42%) in the network (chi-square 38 and 17, men and women, respectively, p <0.001 for both sexes). Patients with followup were in the program longer as expected (clinic 19.6 compared to 13.4 months with and without followup, respectively, p <0.001 and network 14.2 compared to 9.3 months, respectively, p <0.001). The significant differences in percentage of patients with followup in the clinic compared to network reflected in part the shorter duration in the program. The fraction of patients with followup was strongly correlated with time in the program (F-21, p <0.001). By ANOVA the slopes of followup fraction compared to time in program were not homogeneous, that is they differed between the clinic and network (p <0.001 for comparison). Even so, allowing for the interaction of slope and site (slope in the clinic compared to network), and covariance of followup with time in the program, the corrected fractions of patients with followup were 63 2 4 compared to 42 t 1 (p <0.001).In other words, the network did not do as well as the clinic in achieving followup, even after allowing as much as possible for the fact that it more recently enrolled more patients and, therefore, followup time was shorter. Thus, the fraction of patients with followup was higher in the clinic than network. Supersaturation response to treatment. Calcium Oxalate: Although initial values of supersaturation were lower in the network than in the clinic, mean decrease in supersaturation was lower in the network, and so treatment supersaturation values did not differ (table 1). The decrease in supersat-
1631
MEDICAL PREVENTION OF NEPHROLITHIASIS TABLE1. Univariate comparisons of principal values between sites Clinic Mean Before Treatment
2
Network Mean
SD
During Treatment
Change
Before Treatment
2
SD
During Treatment
Change
Supersaturation calcium 9.6 i 0.2 6.8 2 0.2 -2.8 2 0.2 8.8 5 0.147.2 + 0.21 -1.6 t 0.2t oxalate monohydrate -0.07 t 0.05 Supersaturation brushite 1.5 i 0.05 1.25 t 0.04 -0.23 + 0.06 1.3 t 0.04 1.28 2 0.06 -0.48 ? 0.05* 0.76 i 0.04i Supersaturation uric acid 1.22 5 0.04 0.55 2 0.03 -0.66 5 0.04 1.24 5 0.03 0.34 t 0.041 Vol. (1./24 hrs.) 1.76 2 0.04 2.28 + 0.05 0.52 2 0.04 1.78 2 0.03 2.12 2 0.05 Change is mean change between before and after treatment. Significance values are t tests. All values of change are significant against 0 at p <0.001 except for network change in supersaturation brushite, which is not significant at p = 0.10. * For difference from clinic p <0.05. t For difference from clinic p <0.01.
TABLE2. Multivariate adjusted means of supersaturation change uration was proportional to the initial supersaturation value by site (fig. 1).Values for the regression coefficients were significant in the network and clinic (p <0.001 for both) and slopes Clinic Mean t SD Network Mean 2 SD differed when tested for homogeneity (fig. 1). Increase in Change in calcium oxalate -2.2 2 0.2 -1.8 2 0.1 urine volume correlated with decrease in supersaturation monohydrate supersaturation (r = 0.64 and 0.695 for the network and clinic, respectively, -0.12 2 0.06 -0.11 5 0.04 Change in brushite superp <0.001 for both). The increase in urine volume was greater saturation in the clinic than network (table 1).The effect of increase in -0.50 2 0.03* Change in uric acid super-0.63 5 0.04 urine volume can be seen in figure 1,with symbol size scaled saturation to volume increase. Large circles tend t o fall t o the bottom of Values are mean change in supersaturation adjusted for change in urine volume and initial supersaturation using ANOVA. the graph. * For difference from clinic p = 0.019. Given that change in urine volume and initial supersaturation influenced the decrease in supersaturation, and that the sites differed in these covariates as well as in the decrease in supersaturation, we asked if a site difference re- nificant against 0 for the clinic but not the network, and mained once the effects of these 2 covariates were considered decrease in supersaturation did not differ between the clinic (table 2). Least square means after adjustment for change in and network (table 1).Decrease in supersaturation was provolume and initial supersaturation values were no longer portional to initial supersaturation (fig. 2) as with calcium significantly different (p = 0.131). Therefore, we concluded oxalate supersaturation (fig. 11, and the regression coeffithat the better performance of the clinic may well have re- cients for the clinic and network were significant and differed flected a greater increase in urine volume, and patients with significantly (fig. 2). Decrease in supersaturation was highly dependent on change in urine volume (p <0.001 for regresgreater initial supersaturation values. Calcium Phosphate: Initial and treatment values of super- sion at both sites), and large circles predominate at the saturation were not significantly different between the clinic bottom of distributions (fig. 2). When change in volume and and network (table 1).Change in supersaturation was sig- initial supersaturation were considered, adjusted least
B
1
0
10
0
0
0
0
0 0
0
0
L"
10 20 30 Pre Rx CaOx SS
40
0
10 20 30 Pre Rx CaOx SS
40
FIG. 1. Change in calcium oxalate monohydrate (CaOx)supersaturation ( S S )(Yaxis)versus before (Pre)treatment (Rr)supersaturation
(xaxis) for clinic (A) and network (B).Change in supersaturation was strongly inverse to initial supersaturation. Increased urine volume,
shown by size of symbol, was most marked when supersaturation decrease was greatest as shown by clustering of large circles at bottom of distributions. Values for regression coefficients were -0.51 versus -0.66 for network and clinic (p <0.001 each). Homogeneity of slopes was not present (p <0.01 for interaction of site and slope).
1632
MEDICAL PREVENTION OF N E P H R O L I T H M I S
7
0
1
2 3 4 5 Pre Rx CaP SS
0
6
1
2 3 4 5 Pre Rx CaP SS
6
FIG. 2. Change in brushite ( C a p )su ersaturation (SS) (Y axis) versus before (Pre)treatment (Rr)supersaturationjX axis) for clinic (A) and network (B). Plotting is same as in 8gure 1. Regression coefficients were -0.409 versus -0.663 for network and cllmc (p <0.001 for both). Homogeneity of slopes was not present (p <0.001).
square means for change in supersaturation did not differ (p = 0.92) between the network and clinic (table 2). Unlike calcium oxalate supersaturation, decreasing values were not strong in any treatment setting although dependence on urine volume and initial supersaturation values were strong. Uric Acid: Before treatment values of supersaturation were not different between the network and clinic (table 1 and fig. 3). Treatment reduced supersaturation significantly in the network and clinic but mean reduction in supersatu-
ration was greater in the clinic (table 1).Decrease in supersaturation was obviously dependent on the initial value, and individual regressions were highly significant, and not different between the network and clinic (fig. 3). Volume increase was highly correlated with decrease in supersaturation even when initial supersaturation was considered (p values for regression of supersaturation decrease in change of volume were <0.001 at the network and clinic). When decrease in supersaturation was adjusted for initial super-
0-----
-
0
I 1
2 3 4 5 Pre Rx Uric Acid SS
6
0
1
2 3 4 5 Pre Rx Uric Acid SS
6
FIG. 3. Change in uric acid (CIA)supcrsaturation (SS)(Y axis) versus before (Pre)treatment (Rr)supersaturation (X axis) for clinic (A)and network (B). Plotting is same us in figure 1. Regression coefficients were -0.572 versus -0.679 for network and clinic (p <0.001 for both). Slopes wcrc homogrncous.
1633
MEDICAL PREVENTION OF NEPHROLITHIASIS
saturation and change of volume (table 21, an effect of site could still be detected, meaning that the greater decrease in supersaturation by univariate analysis (table 1)was not entirely explained by differences in initial supersaturation and change in urine volume. The decrease occurred in large part from a greater increase in urine pH in the clinic (not shown) as should be expected.
DISCUSSION
Our impression is that significant treatment effects were present and broadly comparable in the clinic and 7 network sites. Treatment led to important reductions in supersaturation with respect to calcium oxalate and uric acid. The decrease in supersaturation depended on the initial value and increase in urine volume, as one would expect. Decreases in calcium phosphate supersaturation were weak in the network and clinic, and actually insignificant in the network. Because supersaturation is the main established force driving crystallization9 and established stone preventions act at least in part via decreased supersaturation, the decreases we observed strongly support the idea that the network will eventually decrease stone recurrence, as has been reported from the clinic,1.14,15,19.20 For calcium oxalate and uric acid supersaturation decreases were excellent in the network, although less vibrant than in the clinic. Weaker network decreases were traceable mainly to a lesser increase in urine volume and to a lower pretreatment supersaturation value for calcium oxalate. That a network of distributed treatment centers can reduce supersaturation effectively is not surprising but has never been demonstrated to our knowledge. Also not surprising is that a dedicated research clinic can do better in reducing supersaturation. However, despite differences, both environments produced treatment supersaturation values for calcium oxalate below 7.5, which is closer to the normal value than to the levels found among stone forming patients,16 and below 1.0 for uric acid, a normal range. Given these results our model for treatment of large numbers of patients seems secure so far. For calcium phosphate supersaturation neither treatment environment was remarkably effective, and they did not differ by simple t tests or after allowing for covariation of supersaturation reduction with urine volume change and initial supersaturation values. However, the change in supersaturation was significant compared to 0 in the clinic and not in the network. This finding points t o a weakness of available treatments or lack of effective strategies for use of treatments that seems universal within our system of treatment centers. In the clinic and network calcium phosphate supersaturation elevation correlated with the amounts of calcium phosphate solid phases in stones,16.17 and so the I failure t o reduce calcium phosphate supersaturation has a theoretical potential to maintain stone formation through calcium phosphate nucleation.g.21 However, this potential has not yet been explored clinically. Although treatment effects in the network rival those of the clinic the fraction of patients who had followup urine samples was significantly smaller in the network. The natural explanation, that entry of new patients is much higher in the network than in the clinic and, therefore, mean time of network patients in the program was shorter, is not sufficient. Even after allowing for time in the program a lower percentage of network patients were followed to assess treatment response. We have no ready mechanism to explain this difference and decline to speculate. However, it is a crucial problem that will require substantial remedy in the long k m . The patients who did not have followup did not contribute t o our data on treatment outcome, and SO the clinic and network report alike patients compliant enough to have
followup. Patients who did not have followup may well have achieved no treatment effects, which is a moot point. Our reliance on supersaturation as a principal index of treatment occurred as a result of a broad consensus that supersaturation is the most direct available expression of crystallization potential that one can obtain in clinical practice.16,22Supersaturation is the usual physical chemical expression of crystallization force.9 Measured in urine it is a strong correlate of stone mineral composition in the clinic16.17 and network.17 Available medical prevention to date acts at least in part by decreasing urine supersaturation, which includes reversal of hypercalciuric states using thiazide,23.2* low sodium diet25 and, when needed, removal of parathyroid glands in primary hyperparathyr~idism.~ Prevention also includes treatment of low urine citrate by potassium citrate,5 acid urine pH with alkali salts,26hyperuricosuria and hyperoxaluria by diet,27 hyperuricosuria with allopurin013~2~.2S and all urine supersaturation values with increased urine flow.29 Supersaturation is not a complete or wholly sufficient statement of stone risk as inhibitorss.9 and promoters, and family history of stones are also important,3O but it is the most compact and powerful present available measurement of stone potential that one can manipulate through treatment. With regard to a final gold standard, one must wait for some years to pass and count the formation of new stones. We shall no doubt do that. However, in the interim it is important and responsible to assure ourselves that supersaturation has decreased, because if it does not one has no a priori reason to believe that treatments will be effective in stone prevention. For example we have shown not only that stone mineral composition is highly predicted by urine supersaturation measurements16 but also that treatments that prevented stones regularly decreased relevant supersaturation values in the clinic. Others have shown similar results.g.31.32 That our results show a decrease in supersaturation gives considerable confidence in the network arrangement. The many difficulties of a central laboratory, sample collection in diverse settings, reporting and maintaining a time sequential grasp of urine chemistry data, even interested in Drevention and willing those of keeDing - - Datients _ to obtain followup measurements such as those we prese; here, offer the kinds of challenge that, when successfully overcome, represent a kind of minor clinical triumph.
REFERENCES
1. Parks, J. H. and Coe, F. L.: An increasing number of calcium oxalate stone events worsens treatment outcome. Kidney Int., 4 5 1722,1994. 2. Ettinger, B., Citron, J. T., Livermore, B. and Dolman, L. I.: Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J. Urol., 1 3 9 679,1988. 3. Ettinger, B., Tang, A,, Citron, J. T., Livermore, B. and Williams, T.: Randomized trial of allopurinol in the prevention of calcium oxalate calculi. New Engl. J. Med., 3 1 5 1386,1986. 4. Pak, C. Y. C., Fuller, C., Sakhaee, K., Preminger, G. M. and Britton, F.: Long-term treatment of calcium nephrolithiasis with potassium citrate. J. Urol., 134: 11, 1985. 5. Barcelo, P., Wuhl, O., Servitge, E., Rousaud, A. and Pak, C. Y. C.: Randomized double-blind study of potassium citrate in idiopathic hypocitraturic calcium nephrolithiasis. J. Urol., 1 5 0 1761,1993. 6. Coe, F. L., Parks, J. H. and Asplin, J. R.: The pathogenesis and treatment of kidney stones. New Engl. J. Med., 327: 1141, 1992. 7. Smith, L. H.: The medical aspects of urolithiasis: an overview. J. Urol., 141: 707, 1989. 8. Asplin, J. R., Parks, J. H. and Coe, F. L.: Dependence of upper limit of metastability on supersaturation in nephrolithiasis. Kidney Int., 5 2 1602,1997. 9. Coe, F. L., Favus, M. J., Pak, C. Y. C., Parks, J. H. and Preminger, G. M.: Kidney Stones: Medical and Surgical Management. Philadelphia: Lippincott-Raven Publishers, 1996.
1634
MEDICAL PREVENTION O F NEPHROLITHIASIS
ogy and treatment of nephrolithiasis: new research venues. 10. Asplin, J. R., Favus, M. J. and Coe. F. L.: Nephrolithias. In: The J. Bone Min. Res., 12: 522, 1997. Kidney, 5th ed. Phladelphia: W. B. Saunders Co., pp. 189323. Coe, F. L.: Treated and untreated recurrent calcium nephrolithi1935, 1996. asis in patients with idiopathic hypercalciuria, hyperuricos11. Coe, F. L. and Favus, M. J.: Disorders of Bone and Mineral uria, or no metabolic disorder. Ann. Intern. Med., 87: 404, Metabolism. New York: Raven Press, 1992. 1977. 12. Favus, M. J., Berelowitz, M. and Coe, F. L.: Effects of somatostatin on intestinal calcium transport in the rat. Amer. J . Physiol. 24. Coe, F. L., Parks, J. H., Bushinsky, D. A., Langman, C. B. and Favus, M. J.: Chlorthalidone promotes mineral retention in 241: G215, 1981. patients with idiopathic hypercalciuria. Kidney Int., 3 3 1140, 13. Consensus conference. Prevention and treatment of kidney stones. J.A.M.A., 260: 977, 1988. 1988. 14. Parks, J. H. and Coe, F. L.: The financial effects of kidney stone 25. Kok, D. J., Iestra, J . A,, Doorenbos, C. J. and Papapoulos, S. E.: The effects of dietary excesses in animal protein and in sodium prevention. k d n e y Int., 50: 1706, 1996. on the composition and the crystallization kinetics of calcium 15. Coe, F. L. and Parks, J. H.: Clinical approach. In: Nephrolithiaoxalate monohydrate in urines ofhealthy men. J. Clin. Endocr. sis: Pathogenesis and Treatment, 2nd ed. Chicago: Year Book Metab., 71: 861, 1990. Medical Publishers, 1988. 16. Parks, J. H., Coward, M. and Coe, F. L.: Correspondence be- 26. Pak, C. Y. C., Sakhaee, K. and Fuller, C.: Successful management of uric acid nephrolithiasis with potassium citrate. Kidtween stone composition and urine supersaturation in nephney Int., 3 0 422, 1986. rolithiasis. Kidney Int., 51: 894, 1997. 17. Asplin, J . R., Parks, J . , Lingeman, J., Kahnowski, R., Mardis, H., 27. Coe, F. L., Moran, E. and Kavalich, A. G.: The contribution of dietary purine over-consumption to hyperpuricosuria in calLacey, S., Goldfarb, D., Grasso, M. and Coe, F.: Supersaturacium oxalate stone formers. J . Chronic Dis., 2 9 793, 1976. tion and stone composition in a network of dispersed treat28. Coe, F. L. and Raisen, L.: Allopurinol treatment of uric-acid ment sites. J . Urol., 159: 1821, 1998. 18. Werness, P. G., Brown, C. M., Smith, L. H. and Finlayson, B.: disorders in calcium-stone formers. Lancet, 1: 129, 1973. EQUIL2: a basic computer program for the calculation of uri- 29. Pak, C. Y. C., Sakhaee, K., Crowther, C. and Brinkley, L.: Evinary saturation. J. Urol., 134: 1242. 1985. dence justifying a high fluid intake in treatment of nephroli19. Strauss. A. L., Coe, F. L., Deutsch, L. and Parks, J . H.: Factors thiasis. Ann. Intern. Med., 9 3 36, 1980. that predict relapse of calcium nephrolithiasis during treat- 30. Curhan, G. C., Willett, W. C., Rimm, E. B. and Stampfer, M. J.: ment: a prospective study. h e r . J . Med., 72: 17, 1982. Family history and risk of kidney stones. J. Amer. SOC.Neph20. Parks, J. H., Coe. F. L. and Strauss, A. L.: Calcium nephrolithirol., 8 1568, 1997. asis and medullary sponge kidney in women. New Engl. 31. Tiselius, H. G.: A simplified estimate of the ion-activity product J. Med., 306: 1088, 1982. of calcium phosphate in urine. Eur. Urol., 1 0 191, 1984. 21. Herring, L. C.: Observations on the analysis of ten thousand 32. n u n , M. J. and Schober, S.: Urolithiasis in Tennessee: a n occuurinary calculi. J. Urol.. 88:545, 1962. pational window into a regional problem. Amer. J. Public 22. Coe, F. L. and Parks. J . H.: New insights into the pathophysiolHealth, 81: 587, 1991.