0022-5347/97/1586-2069$03.00/0

Copyright Q 1997 by AMERICAN UROLOGICAL ASS~CIATION, INC. JOURNAL OF UROLOGY Vol. 158,2069-2073, December 1997

Printed in U.S.A.

POTASSIUM-MAGNESIUM CITRATE IS AN EFFECTIVE PROPHYLAXIS AGAINST RECURRENT CALCIUM OXALATE NEPHROLITHIASIS

BRUCE ETTINGER,* CHARLES Y. C. PAK, JOHN T. CITRON, CARL THOMAS, BEVERLEY ADAMS-HIJET AND ARLINE VANGESSEL

From the Diuision of Research, Kaiser Permanente Medical Care Program, Oakland, California, the Department of Mineral Metabolism, Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas,

Department of Medicine, Kaiser Permanente Medical Center, Walnut Creek, California, Department of Urology, Kaiser Permanente Medical Center, San Francisco, California, and Kaiser Foundation Research Institute, Kaiser Foundation Hospitals, Oakland, California

ABSTRACT

Purpose: We examined the efficacy of potassium-magnesium citrate in preventing recurrent calcium oxalate kidney calculi.

Materials and Methods: We conducted a prospective double-blind study of 64 patients who were randomly assigned to receive placebo or potassium-magnesium citrate (42 mEq. potassium, 21 mEq. magnesium, and 63 mEq. citrate) daily for up to 3 years.

Results. New calculi formed in 63.6% of subjects receiving placebo and in 12.9% of subjects receiving potassium-magnesium citrate. When compared with placebo, the relative risk of treat- ment failure for potassium-magnesium citrate was 0.16 (95% confidence interval 0.05 to 0.46). potassium-magnesium citrate had a statistically significant effect (relative risk 0.10,95% confi- dence interval 0.03 to 0.36) even after adjustment for possible confounders, including age, pretreatment calculous event rate and urinary biochemical abnormalities.

Conclusions: Potassium-magnesium citrate effectively prevents recurrent calcium oxalate stones, and this treatment given for up to 3 years reduces risk of recurrence by 85%.

KEY WORDS: calcium oxalate, citrates, kidney calculi, magnesium, potassium

Potassium citrate effectively reduces recurrence of calcium oxalate nephrolithiasis.' The salutary effect of citrate is at- tributed to its ability to form a soluble complex with urinary calcium and to correct excess urinary acidity.* Additionally, citrate enriched urine has been shown experimentally to have reduced tendency for nucleation, growth, and agglom- eration of calcium oxalate crystals.*%3 However, potassium citrate shares with other potassium salts the tendency to irritate gastric mucosa, and this problem can limit patient acceptability. In clinical trials of potassium citrate, incidence of gastrointestinal adverse events ranged from 9 to 17%.1-4

Dietary magnesium deficiency has been suggested to cause increased risk of calcium calculi,6 and magnesium supple- mentation has been recommended as prophylaxis.6 By en- riching urine with magnesium, the rate of calcium oxalate Crystal growth in vitro can be slowed.7 However, clinical trials of magnesium supplementation did not consistently show therapeutic effects,8-10 probably because supplementa- tion fails to produce sufficiently high urinary magnesium concentration. Isotope studies have indicated poor intestinal absorption of magnesium sdts.11 In a clinical trial using magnesium hydroxide, only 4% of administered magnesium was excreted in urine within 24 hours.10 High dosages of magnesium salts are impractical because they cause diar- rhea.

Recently a new formulation of citrate containing potassium and magnesium at a lower (4:l) ratio was developed by in-

uests for reprints: Division of Research, Kaiser PeFanente

5714. Supported by United States Public Health Service Research

Grants Pol-DK20543 and FD-R-000636. Editor's Note: This article is the third of 5 published in this

h u e for which category 1 CME c-ts cap be earnd. In- h c t i o n s for obtained d t a are gwen anth the queshons On Pages 2274 and 2275.

Accepted for publication June 20, 1997.

M* edica "4 Care Program, 3505 Broadway, Oakland, Califorrua 94611-

vestigators at the University of Texas Southwestern Medical Center at Dallas, Texas.l2 Compared with placebo, short- term (1-week) treatment using potassium-magnesium citrate combined with constant dietary regimen increased urinary citrate levels by 611, urinary magnesium levels by 43%, and urinary pH by 0.6 unit.13 UMary saturation of calcium ox- alate decreased by 31%, inhibition of calcium oxalate crys- tallization increased by 46%, and inhibition of calcium phos- phate crystallization increased by 90%. These effects were more prominent than those achieved using potassium citrate of equivalent potassium c0ntent.1~

We hypothesized that long-term use of this formulation might combine the beneficial urinary physicochemical effects of citrate and magnesium, thereby reducing the risk of kid- ney stone formation. Containing less than the usual amount of potassium, this citrate salt could reduce incidence of gas- trointestinal side effects. Therefore, we tested tolerability and effectiveness of potassium-magnesium citrate in patients diagnosed as having recurrent, active calcium oxalate neph- rolithiasis.

METHODS

Subjects. From among calculus analyses done at Kaiser Permanente Medical Centers in Northern California, we se- lected all reports of calculi containing at least 50% or more calcium oxalate. After reviewing the medical records, we selected patients who had active, recurrent calculous disease and no secondary cause for nephrolithiasis. All subjects had had 2 or more calculi within the previous 5 years and at least 1 calculus within the previous 2 years. Excluded were sub- jects who had obstructive uropathy, chronic urosepsis, renal failure (serum creatinine greater than 1.8 mgJdl., normal 1.5 or less), renal tubular acidosis, or lithotripsy treatment within the previous 6 months. Patients were advised of the study design and gave informed consent as specified by the

2069

2070 POTASSIUM-MAGNESIUM CITRATE PREVENTS NEPHROLITHIASIS

Northern California Kaiser Permanente Institutional Re- view Board.

Outcomes. Study participants were given simple metabolic assessment while following an unrestricted diet. Automated chemistry panels (SMA-24) and 24-hour urine collections were obtained at the beginning of the study to determine baseline values, at 5 months, and every 4 months thereafter for as long as 37 months. We measured urine for total vol- ume, creatinine level, pH, potassium, calcium, magnesium, sodium, phosphate, chloride, sulfate, ammonium, oxalate and citrate. Results of these tests were used to calculate urinary saturation of stone forming substances (calcium ox- alate, sodium urate, brushite LCaHPO, * 2waterl) and the amount of undissociated uric acid.14 Urine could not be ana- lyzed for inhibition of either calcium oxalate crystallization or calcium phosphate crystallization13 because the fresh urine samples required could not be shipped to the Univer- sity of Texas Southwestern Medical Center for Mineral Me- tabolism and Clinical Research in Dallas, where the analysis was done, quickly enough h m San Francisco. On the basis of %hour urinary excretion, we defined low urine volume as 1200 ml. or greater excreted, hypercalciuria as 300 mg. or greater calcium excreted for men and 250 mg. or greater for women, hyperuricosuria as 800 mg. or greater uric acid ex- creted for men and 750 mg. or greater for women, hyperox- aluria as 40 mg. or greater oxalic acid excreted and hypoci- traturia as 320 mg. or greater citrate excreted.

Pretreatment rates of calculus formation were based on all calculous events that occurred within 3 years before the study. For each subject we obtained a coned view x-ray film of the kidneys a t the start of the study and annually there- after. We calculated an index of obesity, body mass index, by dividing each subject's weight (kg.) by the square of the height (m.). To determine biochemical effects of treatment, we calculated for each subject the mean of all values for all treatment visits for each variable and the percentage changes from baseline values. For each group (potassium- magnesium citrate or placebo) we calculated mean baseline, treatment and percentage change values for each biochemi- cal variable.

Based on previous experience which showed that the start of effective therapy may be marked by washout of small calculi or gravel, we included in the study design a 1-month grace period. During this time, any passage of calculi was not considered to indicate failure of therapy and did not cause removal of any subject from the study. Thereafter any pas- sage of new calculi, radiographic appearance of new calculi, or enlargement of preexisting calculi seen radiographically was considered to indicate treatment failure and caused the patient to be removed from the study. An impartial referee identified calculous events by judging clinical information and x-ray films without knowing treatment status. To avoid further bias the referee compared x-ray films without know- ing the temporal sequence.

Treatment assignment. Because previous researchers re- ported correlation between calculus burden and risk of recur- rent formation of calculi,11 we stratified study subjects on the basis of calculus absence or presence. Within each stratum, using computer-generated random numbers, we randomly assigned subjects to receive visually identical tablets of ei- ther placebo or potassium-magnesium citrate. All subjects were instructed in a diet that recommended restricting salt, refined sugar, foods rich in oxalate and animal protein but allowed up to 2 servings of dairy foods per day. Subjects were further instructed to take 2 tablets of the study drug 3 times daily. This provided a daily dose of 42 mEq. potassium, 21 mEq. magnesium, and 63 mEq. citrate to subjects who re- ceived active drug. Compliance with medication regimen was confirmed by counting tablets regularly, but compliance with dietary instruction was neither encouraged nor assessed.

Adverse events. Gastrointestinal symptoms were assessed

a t baseline and at each followup visit by using a Likert score (I-none, 2-slight, 3-some, 4-considerable). We deter- mined that subjects had a study related gastrointestinal symptom if the symptom score was greater a t any followup visit than it was at the start of the study and if it had a value of a t least 3 on the Likert scale. At each visit subjects were also asked if they found the study medication disagreeable. These analyses included subjects who left the study because of adverse effects.

Statistical analysis. Descriptive statistics were computed for baseline variables in each group. For adverse events we used Fisher's exact test to compare differences in proportions between the 2 groups.

Repeated measures analysis of variance models were used to assess urinary biochemical parameters. A statistically sig- nificant interaction between group and time factors indicated intergroup differences in treatment response. Results were tabulated using baseline and mean followup values and were compared within each treatment group using paired t tests. Statistical significance levels as reported were not adjusted for multiple testing.

Cox proportional hazards models were used to estimate relative risk (potassium-magnesium citrate versus placebo) of having a calculous event. Covariates assessed in the model included gender, age, rate of calculus formation within 3 years before the study, baseline calculus burden and pres- ence of urinary biochemical abnormalities. A similar analysis was done by adding as treatment failures the 6 patients (5 who received potassium-magnesium citrate, 1 who received placebo) who did not have a stone event but who left the study after having an adverse reaction to the study drug.

Statistical significance level was a = 0.05. Statistical anal- ysis was done using computer statistical software.

RESULTS

Baseline clinical variables were similar for potassium- magnesium citrate and placebo groups (table 1). Mean uri- nary biochemical parameters as well as prevalence of defined biochemical parameters and prevalence of defined biochem- ical abnormalities also did not differ significantly between the 2 groups. Prevalence of multiple urinary metabolic de- rangements was similar between groups. Among subjects who received placebo 84.8% had 1 or more, 51.5% had 2 or

TABLE 1. Baseline clinical characteristics of nephrotithzasis studv m O U D

~

Placebo Potassium-Magnesium Citrate

No. pts. Age (yrs.) M e n omen Baseline No. calculi/pt. 9% pts. with calculi (No.) No. calcuWpt. in last 3 yrs. Urinary:

PH Total vol., (1Jday) Calcium (mg./day) Phosphorus mgJday Oxalate mglday Sodium mEqJday Potassium mEqJday Magnesium mgJday Ammonium mEqlday Citrate mgJday Uric acid mglday

Activity product: Calcium oxdate m.2 x lo-* Brushite m.* x

33 47.45 t 9.05

2W5 1.70 t 1.91 72.7 (24) 1.7 2 0.8

5.96 t 0.41 1.74 t 0.70 275 t 131

36 _f 12 172 2 87 58 5 20

121 t_ 47 42 t 19

549 t 280 695 2 227

1.77 2 0.87 5.54 2 4.68

1051 2 299

31 48.6 I 11.22

22J9 1.68 5 1.92

71.0 (22) 1.7 5 0.8

6.01 t 0.46 1.98 t 1.15 237 f 120 983 5 329

37 f 13 164 5 66 56 5 25

116 t 44 38 I 10

587 2 374 722 f 229

1.40 2 0.76 4.10 -t 5.43

Sodium urate m.' x lo-+ 1.28 t 0.85 0.99 -t 0.65 Undissociated uric acid mg./day 152 t 101 162 f 111 Data presented as mean 2 SD. No statistically significant differences between placebo and potassium-

magnesium citrate groups.

POTASSIUM-MAGNESIUM CITRATE PREVENTS NEPHROLITHIASIS 207 1

more and 24.2% had 3 or more abnormalities, whereas among subjects receiving potassium-magnesium citrate, 73.3% had 1 or more, 40.0% had 2 or more and 16.7% had 3 or more abnormalities or greater.

F i W n subjects left the potassium-magnesium citrate arm ofthe study, and 8 subjects left the placebo arm. About half ofthese subjects left the study during its first 6 months. The =st left throughout the remaining period. Adverse reactions ascribed to study medication accounted for 5 (16.1%) subjects who stopped taking potassium-magnesium citrate (3 stopped because of gas, bloating or nausea, 1 stopped because of difficulty swallowing the tablet, 1 stopped because of skin rash). Difficulty swallowing tablets caused 1 (3.0%) to stop taking placebo. We saw no statistically significant differences in individual gastrointestinal symptoms between the placebo and active drug groups. About 11% of all subjects had diffi- culty swallowing the rather large pills. During the study 39.4% of the placebo group and 41.9% of the potassium- magnesium citrate group had new or worsening baseline symptoms of eructation, bloating or gaseousness. However, only 15.2% of the placebo group and 25.8% of the potassium- magnesium citrate group reported more than slight symp- toms. Diarrhea was not reported by placebo users but was reported by 11.5% of potassium-magnesium citrate users. However, only 1 subject (3.9%) noted this problem at more than 1 followup visit. Overall 21.4% of patients using potassium-magnesium citrate and 3.58% of subjects taking placebo found the study drug more than slightly disagreeable 88 reported at more than 1 followup visit or left the study because of it (p = 0.052).

Compliance with study drug protocol was excellent in both groups. Median compliance was 86.9% for placebo and 89.0% for potassium-magnesium citrate. More than 70% compli- ance was achieved by 73.1% of patients receiving placebo and 80.8% of those receiving potassium-magnesium citrate.

Table 2 shows the effect of treatment on urinary biochem- &try and on saturation of calculus forming salts. In the placebo group, no statistically significant change was seen in my urinary biochemical or saturation parameter. In the potassium-magnesium citrate group, statistically significant hcreases occurred in urinary magnesium, potassium and citrate levels. Marginal increase in urinary pH (p = 0.08) and statistically significant decline in urinary ammonium levels were seen, as was a modest statistically significant increase

urinary oxalate levels. We found no statistically signifi- 9 t change in urinary saturation of calcium oxalate, brush- ?t" or monosodium urate (as assessed by activity product) or 1l1 uric acid (as assessed by daily excretion rate of undissoci- ated uric acid).

" I a c 0 0.6 1 3j

--, -,

n Relalive risk (KMgCitrate/Placebo) - 0.16 95% Conlldence Interval (0.05 to 0 46)

0.0 0 4 8 12 16 20 24 28 32 36

Time (months) Proportion of patients remaining free of stones during treatment

wit2t.assium-magnesium citrate (Mllg Citrate) or placebo. Unad- jus relative risk is shown.

group and in 12.9% of the potassium-magnesium citrate group (see figure). These calculous events consisted mainly of new stones seen on x-ray films or passed. Such events ac- counted for 75% of treatment failures in the placebo group and 81% of treatment failures in the potassium-magnesium citrate group. Of the 8 stones passed, 6 were analyzed and all contained predominantly calcium oxalate. Compared with placebo, risk of treatment failure using potassium- magnesium citrate was 0.16 (95% confidence interval 0.05 to 0.46). Adjustment for possible confounders including age, pretreatment calculous event rate, low urine volume and hyperoxaluria also indicated a substantially reduced risk of failure (relative risk 0.10, 95% confidence interval 0.03 to 0.36). Adding as treatment failures the 6 subjects who left the study because of adverse drug events still demonstrated efficacy of potassium-magnesium citrate compared with pla- cebo, given that the intent to treat analysis yielded a relative risk of 0.35 (95% confidence interval 0.16 to 0.76).

DISCUSSION

We have found that treatment using potassium- magnesium citrate markedly reduces risk of recurrent cal- cium oxalate nephrolithiasis. In most clinical trials of kidney stone prophylaxis, subjects receiving placebo have a 15 to 20% annual rate of stone recurrence,1o a rate somewhat New calculous events occurred in 63.6% of the placebo

TABLE 2 . Urinary biochemistry and saturation after treatment using potassium-magnesium citrate for recurrent rrephrolithiasis Placebo Gmup Potassium-Magnesium Citrate Group

Pretreatment Posttreatment % Change Pretreatment Posttreatment 5 Change

Magnesium mgJday 121 : 47 105 ? 32 5.7 I 51.9 116 c 44 142 C 42' 29.9 I 42.8

Citrate mgJday 549 f 280 548 ? 265 11.1 I 62.5 587 2 37.4 769 c 289* 50.4 5 73.7 PH 5.96 -+ 0.41 6.02 2 0.32 0.6 2 8.7 6.01 : 0.46 6.29 2 0.58 4.5 2 10.4 brnonium mEqJday 41 2 19 37 2 16 -1.5 I 44.1 38 ? 10 28 C l o t -21.5 f 28.0

Potassium mEqJday 58 2 20 58: 17 10.6 2 36.9 56 : 25 89 ? 27t 69.1 z 49.1

Calcium mgJday 275 2 131 261 I 123 9.3 z 38.2 237 z 120 225 f 92 0.5 z 36.2

Phosphorus mglday 1051 f 299 1014 2 325 995 r 340 6.4 f 20.5 0.4 f 21.1 983 f 329 uric acid mgJday 695 2 227 694 2 194 726 z 210 2.3 z 19.7 2.0 c 23.5 722 z 229 Total vol. IJday 1.74 Z 0.70 1.79 ? 0.84 -0.6 ? 29.2 1.98 Z 1.15 2.01 f 1.00 7.8 I 29.2 Activity product:

&alate mgJday 36 Z 12 39Z 10 15.1 Z 36.4 37 5 13 44 2 21' 18.7 2 28.1

Calcium o d a t e , m,' x 1.77 ? 0.87 1.72 2 0.53 27.6 ? 64.2 1.40 Z 0.76 1.48 f 7.75 7.9 t 49.0 Brushite, m.* x 1 0 - ~ 5.54 ? 4.68 4.82 Z 4.02 21.6 2 71.9 4.10 Z 5.43 5.75 z 3.54 106.4 f 146.1 sodium urate, m? x lo-' 1.28 2 0.85 1.32 2 1.05 55.3 c 128.7 9.89 2 6.48 1.24 c 7.44 36.2 z 73.2 - Undhwiated uric acid mgJday 152 Z 101 185 Z 124 80.5 I 206.9 162 ? 111 105 ? 136 58.6 f 340.7

Data presented as mean z SD. P 10.05 versus pretreatment.

t P <0.001 versue pretreatment.

2072 POTASSIUM-MAGNESIUM CITRATE PREVENTS NEPHROLITHWIS

lower than the 27% rate observed in our placebo group. In trials of active drug therapy (thiazide diuretics or potassium citrate) the annual recurrence rate is about 5%,1.10 a rate similar to that observed for patients treated with potassium- magnesium citrate in our study. This new drug formulation thus appears to have therapeutic efficacy similar to or better than the calculus prophylaxes usually prescribed.

Previous studies of citrate have focused on hypocitraturia and have recommended citrate treatment on the basis of this finding.' Fewer than 20% of patients in our study had hy- pocitraturia, and the beneficial effect of citrate on calculus recurrence was not limited to this group. Therefore, our re- sults support unselective use of citrate for preventing cal- cium oxalate lithiasis.

Obtaining patient acceptance of long-term drug prophy- laxis is a challenge for all clinicians who attempt to manage calculous disease medically. Nearly half of subjects given potassium-magnesium citrate, an effective prophylaxis, did not take the drug continuously for 3 years. Compliance and continuation obtained from highly motivated research sub- jects who participate in double-blinded studies may not be as good in clinical practice. For any long-term prophylactic reg- imen, ease of use and freedom from side effects are impor- tant. Compared with a small thiazide tablet taken once daily, potassium-magnesium citrate has the disadvantage of being available only as a large tablet that must be taken 2 at a time 3 times daily. Pill size was a problem for about 1 in 9 study subjects. However, we specifically studied drug tolerability a t each clinic visit, and a higher prevalence of specific symp- toms can be expected when these symptoms are elicited from a checklist than when subjects are asked nonleading queries. We therefore believe that our findings that gastrointestinal symptoms in the potassium-magnesium citrate and placebo groups did not differ to any statistically significant degree and that only a few subjects leR the study because of gastro- intestinal problems indicate good tolerability of potassium- magnesium citrate, although clinicians should be aware that potassium salts can irritate the gastrointestinal tract.

Prevention of stone recurrence by potassium-magnesium citrate could probably be ascribed to alkali loading and to citraturic and magnesiuric effects of treatment. As described in a short-term metabolic study,l3 subjects taking potassium- magnesium citrate in a randomized trial had a statistically significant or marginal increase in urinary pH, citrate levels and magnesium levels. These effects probably further inhib- ited crystallization of calcium salts,'3 although for technical reasons the inhibitory action could not be measured directly. Compared with results reported in the short-term metabolic study,13 potassium-magnesium citrate treatment in this trial produced a less marked rise in urinary pH (0.3 unit versus 0.6), in citrate levels (182 versus 389 mg. per day), and in magnesium levels (26 versus 42 mg. per day). Th is attenu- ated effect probably results from a lower dose (63 versus 74 mEq. citrate per day) and from a lower compliance rate (87 to 89% versus 100%). Moreover, whereas urinary oxalate levels did not change significantly during the short-term metabolic study,l3 they increased modestly in this trial, probably be- cause a random diet was used instead of a constant metabolic diet. Thus, urinary saturation of calcium oxalate, which in an earlier metabolic study was shown to decline, did not change in this trial.

In our trial, incremental increases in urinary pH and ci- trate levels achieved with potassium-magnesium citrate were less than those observed when a comparable dose of potassium citrate was given. In an earlier trial of potassium citrate administered a t a dosage of 60 mEq. per day,' urinary pH rose by about 0.5 unit (compared with a rise of 0.3 unit in our trial), and urinary citrate levels increased by about 300 mg. per day (compared with an increase of 182 mg. per day). The reason for this discrepancy is not apparent.

In a previous study 13.3%' of subjects taking 1,500 mg. of

magnesium hydroxide daily left the study because of magne. sium hydroxide related diarrhea, whereas this condition did not affect subjects taking 650 mg. of magnesium hydroxide daily.10 In that study, 24-hour urinary excretion of magne- sium increased 50 mg. when 1,500 mg. per day given, but this treatment was no more effective than placebo in reducing recurrence of calculi. In the current study potassium. magnesium citrate did not cause frequent or troublesome diarrhea, but it raised urinary magnesium levels by a mean of only 26 mg. per day, a smaller increase than we had hoped to achieve. The small incremental rise in urinary magnesium levels may contribute to the effectiveness of potassium- magnesium citrate in preventing calculus formation but is unlikely to be solely responsible for this effectiveness.

Our study had several limitations. The number of subjects was too small to allow definitive assessment of drug effec- tiveness for each type of urinary metabolic abnormality. Sim- ilarly our study observed too few treatment failures for us to analyze reasons for failure. We did not test potassium- magnesium citrate against potassium citrate, so we cannot provide evidence to show whether the new formulation has better tolerability or effectiveness. A large number of pa- tients would be required for the study to have sufficient power to determine whether these similar drugs have clinical differences.

CONCLUSIONS

Potassium-magnesium citrate can substantially reduce the 3-year recurrence rate of calcium oxalate stone formation. This benefit does not depend on the presence of hypocitratu- ria, and the drug can therefore be prescribed for calcium oxalate calculus prophylaxis without extensive metabolic testing. Whether substituting magnesium for potassium in the drug formulation provides additional therapeutic effect is unclear because the incremental rise in urinary magnesium excretion is small. Gastrointestinal tolerability of potassium- magnesium citrate compared with potassium citrate must be proved by appropriate comparative studies.

John R. Poindexter, Center for Mineral Metabolism and Clinical Research, University of Texas Southwestern Medical Center, provided data management and Jan Koska per- formed laboratory analysis of urine. Mission Pharmacal, San Antonio, Texas, supplied the drug study material.

REFERENCES

1. Pak, C. Y. C., Fuller, C., Sakhaee, K, Preminger, G. M., Britton, F. Long-term treatment of calcium nephrolithiasis with potas- sium citrate. J. Urol., 134: 11, 1985.

2. Pak, C. Y. C . Citrate and renal calculi. Miner Electrolyte Metab., 1 3 257, 1987.

3. Kok, D. J., Papapoulos, S. E., Bijvoet, 0. L. M. Excessive crystal agglomeration with low citrate excretion in recurrent stone- formers. Lancet, 1: 1056, 1986.

4. Zerwekh, J. E., Reed, B. Y., Heller, H. J., Gonzalez, G. B., Haussler, M. R. and Pak, C. Y. C. Normal vitamin D receptor concentration and responsiveness to 1,25-dihydroxyvitamin D3 in skin fibroblasts from patients with absorptive hypercal- ciuria. Calcif. Tissue Int., in press.

5. Preminger, G. M., Baker, S., Peterson, R., Poindexter, J. and Pak, C . Y. C. Hypomagnesiuric hypocitraturia: an apparent new entity for calcium nephrolithiasis. J. Lithotripsy Stone Dis., 1: 22, 1989.

6. Moore, C . A. and Bunce, G. E. Reduction in frequency of renal calculus formation by oral magnesium administration: a pre- liminary report. Invest. Urol., 2 7, 1964.

7. Meyer, J. L. and Smith, L. H. Growth of calcium oxalate crystals. I. A model for urinary stone growth. Invest. Urol., 1 3 31, 1975.

8. Johansson, G., Backman, U., Danielson, B. G., Fellstrom, B., Ljunghall, S. and Wikstrom, B. Magnesium metabolism in renal stone formers. Effects of therapy with magnesium hy- droxide. Scand. J. Urol. Nephrol., 53: 125, 1980.

9. Wilson, D. R., Straws, A. L. and Manuel, M. A. Comparison of

POTASSIUM-MAGNESIUM CITRATE PREVENTS NEPHROLITHlASIS 2073

medical treatments for the prevention of recurrent calcium nephrolithiasis [abstract]. Urol. Res., 1 2 39, 1984.

10. Ettinger, B., Citron, J. T., Livermore, B. and Dolman, L. I. Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J. Urol., 139: 679, 1988.

11. Graham, L. A., Caesar, J. J. and Burgen, A. S. V. Gastrointes- tinal absorption and excretion of M g 8 in man. Metabolism, 9: 646, 1960.

12. Koenig. K, Padalino, P., Alexandrides, G. and Pak, C. Y. C. Bioavailability of potassium and magnesium, and citraturic response from potassium-magnesium citrate. J. Urol., 14& 330, 1991.

13. Pak, C. Y. C., Koenig, K, Khan, R., Haynes, S. and Padalino. P. Physicochemical action of potassium-magnesium citrate in nephrolithiasis. J. Bone Miner Res., ?: 281, 1992.

14. Pak, C. Y. C., Skurla, C. and Harvey, J. Graphic display of urinary risk factors for renal stone formation. J. Urol., 154: 867, 1985.

EDITORIAL COMMENT

Important advances during the last 25 years have significantly changed the surgical and medical management of stone disease. While shock wave lithotripsy and endourological techniques have basically supplanted the need for open stone removal, debate contin- ues as to the most appropriate metabolic evaluation and the need for medical management to prevent recurrent nephrolithiasis. Previous studies have demonstrated that medical therapy can significantly reduce recurrent stone disease, and suggest that medical prevention may be more cost-effective than repeat noninvasive stone removal procedures. Yet, many physicians continue to avoid medical prophy- laxis of stone disease due to complexity of the metabolic evaluation as well as limited access to various diagnostic tests. Moreover, some have criticized the validity of previous studies that have demon- strated beneficial effects of medical management of stone disease, due to the lack of randomized prospective trials.

The authors have provided exciting new information that ad- dresses many of the problems surrounding medical stone manage

ment. This well performed, randomized, prospective trial investi- gates the nonselective use of a new medimtion. potaasium- magnesium citrate, to prevent recurrent stone formation. In a blinded study the authors have demonstrated significant reduction in stone eventa in the p u p receiving active medication, over those patients who were given the placebo. While the authors acknowledge some limitations of the new medication, potassium-magnesium ci- trate appears to have the significant advantage of preventing recur- rent stones in a noneelective group of calcium stone formers. More- over, the medication was relatively well tolerated and a metabolic evaluation was not required in order to gain the beneficial effects of this new preparation.

This study poses interesting questions for those of us who have long supported the use of selective medical therapy to prevent recur- rent nephmlithiasis. While trained in the 'Dallas School" of stone disease, 1 have long believed in the concept of selective medical management to correct specific underlying physiuxhemical and physiological defecta responsible for recurrent stone disease. Yet, these data suggest that similar success rates in inhibiting new stone disease can be achieved with nonselective administration of potassium-magnesium citrate. In addition, other investigators are looking a t alternative approaches to the medical management of nephrolithiasis, which may also reduce the need for comprehensive metabolic evaluation and extensive medical followup. In the ever- changing medical environment, economics may eventually dictate that nonselective treatment of certain disorders, especially stone disease, may provide a cost-effective alternative to the more classic "physiologically based" treatments of the past.

I t is only with well performed randomized, prospective trials such as the current investigation that we will ultimately be able to deter- mine the most effective, safe and cost efficient approach to prevent recurrent stones. Additional long-term clinical trials are warranted to achieve this goal.

Glenn M. Preminger Division of Urvlqw Duke University Medical Center Durham, North Carolina