- Email: [email protected]
Modulation of calcium oxalate monohydrate crystallization kinetics by urine of preterm neonates
Modulation of Calcium Oxalate Monohydrate Crystallization Kinetics by Urine of Preterm Neonates Eveline A. Schell-Feith, MD, Ivo Que, MD, Dirk J. Kok, PhD, Joana E. Kist-van Holthe, MD, PhD, Esther Ku¨hler, MD, Ronald Brand, PhD, Socrates E. Papapoulos, MD, PhD, and Bert J. van der Heijden, MD, PhD ● Preterm neonates frequently develop nephrocalcinosis (NC). However, the cause has not yet been elucidated. This study focuses on the effects of urine from preterm neonates on crystallization kinetics. Urine samples were collected and renal ultrasound examinations of preterm neonates (gestational age < 32 weeks) were performed during the first weeks of life, at term, and ages 6, 12, and 24 months. The effect of urine on crystallization was determined using a seeded crystal growth system, which measures the square root of solubility product (公Lc), percentage of growth inhibition (GI), and agglomeration inhibition ([tm]) of calcium oxalate crystals. Data for preterm neonates in the first weeks of life (n ⴝ 19) were compared with those for full-term neonates (n ⴝ 17) and healthy adults. Moreover, the correlation between [tm] and urinary (U)citrate level was studied. Mean 公Lc (0.27 ⴞ 0.1 versus 0.36 ⴞ 0.08 mmol/L) and mean [tm] (81 ⴞ 32 versus 143 ⴞ 97 minutes) were lower and mean Ucalcium-creatinine (2.20 ⴞ 1.74 versus 0.46 ⴞ 0.73 mol/mol) and Uoxalate-creatinine ratios (0.39 ⴞ 0.21 versus 0.16 ⴞ 0.09 mol/mol) were greater in preterm neonates in the first weeks of life compared with full-term neonates (p < 0.05). Furthermore, [tm] was less than the lower limit for healthy adults for all but one preterm neonate; [tm] increased and Ucalcium-creatinine and Uoxalate-creatinine ratios decreased with age (p < 0.005). There was a correlation between [tm] and citrate excretion (coefficient of 38; P < 0.001). Patients with and without NC at term did not differ statistically in mean 公Lc, percentage of GI, or [tm]. In conclusion, urine from preterm neonates in the first weeks of life is highly supersaturated and has a defective ability to inhibit calcium oxalate crystal agglomeration. This ability improves with age and is citrate mediated. We suggest that both the high level of supersaturation and defective ability to inhibit calcium oxalate crystal agglomeration contribute to the high incidence of NC. © 2001 by the National Kidney Foundation, Inc. INDEX WORDS: Nephrocalcinosis (NC); preterm neonates; crystallization kinetics; calcium; oxalate; citrate.
P
RETERM NEONATES have a high risk for nephrocalcinosis (NC) in the first months of life. The incidence varies between 17% and 64%, depending on study population, ultrasonographic (US) criteria, and US equipment.1-6 The cause of NC is multifactorial and has not yet been elucidated. Short-term complications include nephrolithiasis with obstruction of the urinary tract and urinary tract infection.2 The long-term outcome of NC in preterm neonates has not been defined, but small-scale studies suggest a decrease in renal function.7-9 In the last two decades, significant progress has been made in the identification and quantification of physicochemical processes involved in calcium oxalate urinary stone formation. The physicochemical processes of crystallization can be divided into thermodynamic, including supersaturation, and kinetic, including nucleation, growth, and agglomeration processes. In adults, the ability of urine to inhibit crystal agglomeration is an important protective mechanism against stone formation. In patients with calcium oxalate stone formation, this ability is decreased, and the extent of this decrease is
directly related to the frequency of stone formation.10,11 The ability to inhibit crystal agglomeration is a citrate-mediated process. Correction of low citrate excretion by citrate treatment increases this ability,10 and citrate treatment is an effective preventive therapy for adults with stone formation.12,13 We recently described a large population of preterm neonates and found that neonates with NC had a lower urinary (U)citratecalcium ratio than those without NC.14 This raised the question of whether the ability of urine to modify crystallization processes is disturbed in preterm neonates. From the Departments of Pediatrics and Endocrinology, Leiden University Medical Center, Leiden; and the Departments of Pediatric Urology and Pediatrics, Sophia Children’s Hospital, Rotterdam, The Netherlands. Received January 9, 2001; accepted in revised form July 6, 2001. Supported in part by grant no. NSN C97.1672 from the Dutch Kidney Foundation. Address reprint requests to Eveline A. Schell-Feith, MD, Van der Waalsstraat 42, 2313 VD, Leiden, The Netherlands. E-mail: schell_ [email protected] © 2001 by the National Kidney Foundation, Inc. 0272-6386/01/3806-0010$35.00/0 doi:10.1053/ajkd.2001.29218
American Journal of Kidney Diseases, Vol 38, No 6 (December), 2001: pp 1229-1234
1229
1230
SCHELL-FEITH ET AL
We investigated the ability of urine from preterm neonates to modulate the kinetics of calcium oxalate monohydrate (COM) crystallization during the first weeks of life, during the onset of NC, and at later times in the first 2 years, when NC tends to diminish. Data for preterm neonates was compared with values for full-term neonates and healthy adults. We investigated whether urine from preterm neonates with NC is less effective in inhibiting the separate processes of crystal formation compared with urine from neonates without NC. Moreover, we studied the correlation between crystal agglomeration and Ucitrate excretion among these subjects. PATIENTS AND METHODS
Patients In a prospective study, 57 preterm neonates with a mean gestational age of 28.1 ⫾ 1.9 (SD) weeks and mean birth weight of 1,047 ⫾ 313 g admitted to Leiden University Medical Center (Leiden, The Netherlands) between January 1997 and November 1998 were investigated. This study was part of a larger study of the cause of NC in preterm neonates.6,14 The ethics committee approved the study protocol. Informed consent was obtained after oral and written information had been given. Patients from outside the usual referral region were excluded from the study because of logistic problems. Patients with obvious underlying renal diseases, such as renal tubular acidosis, were excluded from the study. Failure to collect data occurred when patients died or were transferred to another hospital or when urine collection was not successful. Urine aliquots during the first 4 weeks of life and at term were collected in plastic bags. At term was defined as 38 to 42 weeks postconceptional age. Renal untrasonography was performed at 4 weeks of age and at term. If NC was present at term and the parents agreed to participate in the follow-up study, urine collection and renal untrasonography were repeated at ages 6, 12, and 24 months: 89 aliquots (from 57 patients at different times) were collected. When more than one urine sample was collected from a patient in the first 4 weeks, the average of urinary values was used in the analysis, providing the best estimate of urinary level for that patient during the entire period. This procedure avoids overrepresentation of patients with many samples in the analysis. Urine samples from 17 healthy full-term neonates with a gestational age of 38 to 43 weeks (mean, 41 weeks) were collected in the first week of life. Crystallization data for preterm neonates in the first weeks of life (n ⫽ 19) were compared with data for full-term neonates (n ⫽ 17) and healthy adults (n ⫽ 36), as described by Kok et al.10
Physicochemical Methods The effect of urine on COM crystallization was determined by means of a seeded crystal growth system, which has been described in detail.15-18 In this system, the solubil-
ity, growth, and agglomeration of calcium oxalate crystals are measured as three separate and system-independent parameters. The solubility of COM in urine is given by the solubility product, Lc. Lc is the concentration product of calcium and oxalate at equilibrium, when growth and dissolution of crystals in the solution are in equilibrium. It is expressed as the square root of Lc (公Lc; in millimolar). The difference between the Lc and actual concentration product of calcium and oxalate determines the degree of supersaturation. Growth of calcium oxalate crystals in urine is expressed as growth inhibition (GI). The growth constant of crystal growth with an additive, which is urine in this experiment, is less than the growth constant without an additive. GI is expressed as the growth constant with additive as a percentage of the growth constant without additive. Assessment of the agglomeration of COM crystals in urine is given by agglomeration inhibition, [tm] (in minutes). Agglomeration of crystals reduces the rate at which mineral components are taken up because it causes a decrease in total accessible crystal surface. It therefore is expressed in units of time. [tm] by additives is reflected as an increase in [tm]. In summary, a seeded crystal growth system is used in which crystallization processes are followed by measuring the uptake of calcium 45 (45Ca) into the crystal mass. By using aged and pregrown COM crystals and standardizing crystal handling, reproducible constant conditions for each experiment are ensured. Two experiments were performed. In the first, calcium and oxalate concentrations were increased to find the lowest concentration product at which 45Ca uptake caused by crystal growth occurs. This is the Lc for calcium oxalate in the sample tested. In the second experiment, a single calcium oxalate concentration product in the metastable range was used, and time varied and 45Ca uptake were measured. By means of a technique derived from previously described COM crystallization kinetics,18 parameters for crystal GI (percentage of GI) and [tm] can be calculated from the curves obtained.
Biochemical Methods Calcium, creatinine, and citrate were measured using a Hitachi analyzer (Roche Diagnostics, Almere, The Netherlands). Ucalcium levels less than 0.5 mmol/L are not detectable in our laboratory. When Ucalcium was not detectable, we used 0.5 mmol/L in our calculations. The pH of samples for oxalate determinations was adjusted to 2 before samples were frozen. Oxalate was measured manually by a coloric photometric method (Instruchemie, Hilversum, The Netherlands). Ratios of Ucalcium to creatinine, Uoxalate to creatinine, Ucitrate to creatinine, and Ucitrate to calcium were calculated.
Renal Ultrasonography To detect NC, one pediatric radiologist performed all US examinations, using state-of-the-art equipment with a 7- to 7.5-MHz small-part transducer (Toshiba, Zoetermeer, The Netherlands). Transverse and longitudinal images were made of both kidneys. The radiologist was not informed about the patient’s history or treatment. NC was defined as the pres-
CRYSTALLIZATION KINETICS IN PRETERM NEONATES Table 1.
1231
Effects of Urine From Full-Term Neonates in the First Week of Life and Preterm Neonates in the First 4 Weeks of Life on COM Crystallization Kinetics Full-Term Neonates (n ⫽ 17)
公Lc (mmol/L) GI (%) [tm] (min) Uca-cr (mol/mol) Uox-cr (mol/mol) Uci-cr (mol/mol) Uci-ca (mol/mol)
0.36 ⫾ 0.08 58 ⫾ 17 143 ⫾ 97 0.46 ⫾ 0.73 0.16 ⫾ 0.09 0.66 ⫾ 0.98 3.93 ⫾ 4.69
(0.25-0.5) (30-85) (45-330) (0.03-2.4) (0.06-0.39) (0.04-4.33) (0.33-18.1)
Preterm Neonates (n ⫽ 19)
0.27 ⫾ 0.1 64 ⫾ 21 81 ⫾ 32 2.20 ⫾ 1.74 0.39 ⫾ 0.21 0.52 ⫾ 0.45 0.28 ⫾ 0.16
(0.17-0.52) (13-90) (38-147) (0.47-6.44) (0.16-0.96) (0.08-1.7) (0.08-0.62)
P
⬍0.01 NS ⬍0.05 ⬍0.005 ⬍0.001 NS ⬍0.001
NOTE. Values expressed as mean ⫾ SD (range). Abbreviations: Uca-cr, urinary calcium-creatinine ratio; Uox-cr, urinary oxalate-creatinine ratio; Uci-cr, urinary citratecreatinine ratio; Uci-ca, urinary citrate-calcium ratio; NS, not significant.
ence of bright reflections in the medulla or cortex that were reproducible in both the transverse and longitudinal directions with or without acoustic shadowing. Reflections varied from small flecks 1 to 2 mm across to white dots larger than 2 mm to completely echodense pyramids. Two tiny echogenic parallel stripes were considered to be the arcuate or branch arteries.
Statistics To evaluate the effect of age on 公Lc, percentage of GI, [tm], and excretion of calcium and oxalate, we performed an analysis of variance, accounting for within-subject repeated measurements. The same analysis was performed to evaluate the effect of citrate on [tm]; this analysis was corrected for the effect of age. Groups were compared by means of Student’s t-test for normally distributed data and Mann-Whitney test for nonparametric data. Results are given as mean ⫾ SD.
were included in the figure. [tm] increased significantly with age (P ⬍ 0.005). 公Lc and percentage of GI did not change significantly with age, but Ucalcium-creatinine and Uoxalate-creatinine ratios decreased significantly with age (P ⬍ 0.01). Urine samples collected in the first weeks of life and at term were obtained from an unselected group of patients; however, sequential studies were obtained only from preterm neo-
RESULTS
Table 1 shows that mean 公Lc and mean [tm] were significantly lower in urine from preterm neonates in the first weeks of life compared with full-term neonates (P ⬍ 0.05). Mean Ucalciumcreatinine and Uoxalate-creatinine ratios were significantly greater for preterm neonates in the first weeks of life compared with full-term neonates, whereas the mean Ucitrate-calcium ratio was significantly lower (P ⬍ 0.005). Figure 1 shows the [tm] for preterm neonates in the first 4 weeks of life compared with full-term neonates and healthy adults.10 All but 1 ([tm] ⫽ 147 minutes) of the 19 preterm neonates measured in the first 4 weeks of life had a [tm] less than the lower limit for healthy adults (range, 145 to 584 minutes; mean ⫾ SEM, 275 ⫾ 23).10 Figure 2 shows repeated measurements of [tm] for preterm neonates in the first 2 years of life. Only patients with sequential data points
Fig 1. [tm] of calcium oxalate crystals in urine of preterm neonates, full-term neonates, and adults. (䊐) Preterm neonates in the first 4 weeks (n ⴝ 19) have a lower mean [tm] (81 ⴞ 32 minutes) compared with (Œ) full-term neonates in the first week (n ⴝ 17; 143 ⴞ 97 minutes; P < 0.05). The lower limit measured for healthy adults is indicated as a dashed line.10
1232
SCHELL-FEITH ET AL
Fig 2. Lines show repeated measurements of [tm] in preterm neonates in the first 2 years of life, which increase significantly with age (P < 0.005).
nates in whom NC was diagnosed at term. Therefore, a selected population was studied. Patients with (n ⫽ 8) and without NC (n ⫽ 16) at term did not differ statistically in mean 公Lc (0.34 ⫾ 0.18 versus 0.27 ⫾ 0.08 mmol/L), mean percentage of GI (63% ⫾ 20% versus 56 ⫾ 16), or mean [tm] (106 ⫾ 55 versus 121 ⫾ 48 minutes). Finally, [tm] and excretion of citrate correlated positively after correction for the effect of age, with a coefficient of 38 (95% confidence interval, 20 to 56; P ⬍ 0.001; Fig 3).
determine the degree of supersaturation. We found a lower Lc and greater excretion of calcium and oxalate for preterm compared with full-term neonates. Consequently, urine from preterm neonates in the first weeks will have a greater level of supersaturation than urine from full-term neonates. The following kinetic events occur during stone formation: nucleation, growth, and agglomeration. The process of nucleation cannot be measured directly because particles formed are too small. In vitro measurement instruments, described here, measure not only the previously mentioned effect of Lc, but the effect of growth and agglomeration on the crystallization process separately. It is well known that small particles can be formed and excreted in urine without leading to the formation of renal stones.19 In adults, urine from patients with stone formation can be clearly distinguished from that from healthy subjects only in its effect on crystal agglomeration. Kok et al10 found that adult stone formers lack a defensive mechanism (the ability of urine to inhibit crystal agglomeration) that ensures that the size of particles formed is restricted. This finding was confirmed by Erwin et al,11 who showed that [tm] decreases with increasing stone frequency, making it a useful tool to measure the risk for stone recurrence. Crystal percentage of GI did not change with age and was not different for preterm compared
DISCUSSION
Preterm neonates run a high risk for developing NC in the first months of life. In most patients, NC tends to diminish spontaneously in months to years,6 but the presence of crystals in the kidney at the very beginning of life might result in a decrease in renal function.7-9 The cause of NC is multifactorial and has not yet been elucidated fully.14 For adults, more knowledge is present about stone formation; however, the formation of Ucalcium oxalate stones is considered a complex and incompletely understood process. The physicochemical processes of stone formation can be divided into thermodynamic and kinetic processes. The thermodynamic process requires free energy and is determined by the level of supersaturation in urine. The Lc and concentrations of calcium and oxalate in urine
Fig 3. Correlation of calcium oxalate crystal agglomeration and Ucitrate concentration from preterm neonates. [tm] and citrate excretion positively correlate after correction for the effect of age (coefficient of 38; 95% confidence interval, 20 to 56; P < 0.001).
CRYSTALLIZATION KINETICS IN PRETERM NEONATES
with full-term neonates. The growth of crystals is a very slow process. Therefore, we did not expect to find an important role of this factor in the development of NC because the time urine spends in different nephron segments is relatively short.10,20-22 Conversely, agglomeration is a rapid process, and large particles can be formed within a short period.23,24 Our results show that urine from preterm neonates in the first weeks of life is less able to inhibit calcium oxalate crystal agglomeration. Compared with full-term neonates who are not at risk for developing NC, the mean [tm] in the first weeks of life was significantly less for preterm neonates. Furthermore, [tm] was less than the lowest value found by Kok et al10 for healthy adults for all except one preterm neonate. In preterm neonates, NC develops in the first weeks of life and subsequently gradually disappears in most patients in months to years. Consistent with this is our finding that [tm] increases with age, and the degree of supersaturation appears to decrease with age (although solubility did not change with age, both calcium and oxalate excretion decreased with age). Premature neonates excrete urine with a high level of supersaturation and diminished ability to inhibit agglomeration, but as they mature, saturation level and this ability improve. In a hyperoxaluric rat model, deposition of calcium oxalate crystals also disappeared when oxaluric pressure was reduced. It appears that the rat kidney carries out processes for the active removal of calcium deposits. These findings were confirmed in human renal material.25,26 It should be mentioned that the group of patients studied between the age of 6 and 24 months were a selected study population because we only followed up neonates with NC at term. However, it is theoretically unlikely that these results would not be reproducible for an unselected group of patients. Because preterm neonates without NC at term have an older gestational age at birth and greater birth weight than those with NC at term,14 those patients not studied most likely would have greater [tm] and lower Ucalcium-creatinine and Uoxalate-creatinine ratios in the period of 6 to 24 months. Thus, an unselected group of patients would have yielded, if anything, an even more striking increase in [tm] and decrease in calcium and ox-
1233
alate excretion with age. Furthermore, when Ucalcium was not detectable in our laboratory, we used 0.5 mmol/L in the calculations. We presume that this assumption did not influence our results considerably because Ucalcium-creatinine ratios were similar to the values of Karlen et al27 for preterm and full-term neonates in the first 2 weeks. Healthy full-term children hardly ever form stones. Recently, we described a theoretical model that makes it plausible that this low prevalence in children is caused by the following mechanism. Nephron segments in children are shorter than those in adults, which results in a shorter transit time for urine in the segments and lower concentration values in the loops of Henle of children.28 Therefore, although [tm] is often lower than in healthy adults, full-term neonates have a low risk for forming stones. We could not show a significant difference in effect on crystallization kinetics between preterm neonates who developed NC and those who did not. However, preterm neonates as a whole showed an increased level of supersaturation and decreased ability to inhibit crystal agglomeration. We hypothesize that all preterm neonates are at risk for developing NC, but that in some preterm neonates, the balance between the various stone-inhibiting and stone-promoting factors is more negative than in others. An important and clinically relevant question is which factors influence the crystallization processes. Kok et al10 showed that citrate, both as a pure compound18 and in urine from adults,10 is a very potent inhibitor of crystal agglomeration. This study shows a correlation of citrate and [tm] in preterm neonates. In addition, preterm neonates have a low Ucitrate-calcium ratio compared with full-term neonates.29 Similar to the mechanism in adults, citrate seems to be a major modulator of the agglomeration process in preterm neonates as well. In adults, correction of low citrate excretion by citrate treatment increases the ability to inhibit agglomeration,10 whereas an increase in [tm] reflects a decrease in stone-formation rate in citrate-treated patients.13 Our results suggest that citrate treatment of preterm neonates may influence positively the balance of stone-inhibiting and stone-stimulating factors and thus may potentially be effective preventive therapy.
1234
SCHELL-FEITH ET AL
We conclude that urine from preterm neonates is highly supersaturated, with a defective ability to inhibit calcium oxalate crystal agglomeration in the first weeks of life. This ability improves with age and, similar to adults, is citrate mediated. We suggest that both the high level of supersaturation and defective ability to inhibit calcium oxalate crystal agglomeration contribute to the high incidence of NC in preterm neonates. REFERENCES 1. Jacinto JS, Modanlou HD, Crade M, Strauss AA, Bosu SK: Renal calcification incidence in very low birth weight infants. Pediatrics 81:31-35, 1988 2. Downing GJ, Egelhoff JC, Daily DK, Alon U: Furosemide-related renal calcifications in the premature infant. A longitudinal ultrasonographic study. Pediatr Radiol 21:563565, 1991 3. Short A, Cooke RW: The incidence of renal calcification in preterm infants. Arch Dis Child 66:412-417, 1991 4. Katz ME, Karlowicz MG, Adelman RD, Werner AL, Solhaug MJ: Nephrocalcinosis in very low birth weight neonates: Sonographic patterns, histologic characteristics, and clinical risk factors. J Ultrasound Med 13:777-782, 1994 5. Saarela T, Vaarala A, Lanning P, Koivisto M: Incidence, ultrasonic patterns and resolution of nephrocalcinosis in very low birthweight infants. Acta Paediatr 88:655-660, 1999 6. Schell-Feith EA, Holscher HC, Zonderland HM, Kistvan Holthe JE, Conneman HN, van Zwieten PHT, Brand R, van der Heijden AJ: Ultrasonic features of nephrocalcinosis in preterm neonates. Br J Radiol 73:1185-1191, 2000 7. Ezzedeen F, Adelman RD, Ahlfors CE: Renal calcification in preterm infants: Pathophysiology and long-term sequelae. J Pediatr 113:532-539, 1988 8. Downing GJ, Egelhoff JC, Daily DK, Thomas MK, Alon U: Kidney function in very low birth weight infants with furosemide-related renal calcifications at ages 1 to 2 years. J Pediatr 120:599-604, 1992 9. Jones CA, King S, Shaw NJ, Judd BA: Renal calcification in preterm infants: Follow up at 4-5 years. Arch Dis Child Fetal Neonatal Ed 76:F185-F189, 1997 10. Kok DJ, Papapoulos SE, Bijvoet OL: Crystal agglomeration is a major element in calcium oxalate urinary stone formation. Kidney Int 37:51-56, 1990 11. Erwin DT, Kok DJ, Alam J, Vaughn J, Coker O, Carriere BT, Lindberg J, Husserl FE, Fuselier H Jr, Cole FE: Calcium oxalate stone agglomeration reflects stone-forming activity: Citrate inhibition depends on macromolecules larger than 30 kilodalton. Am J Kidney Dis 24:893-900, 1994 12. Pak CY, Fuller C: Idiopathic hypocitraturic calciumoxalate nephrolithiasis successfully treated with potassium citrate. Ann Intern Med 104:33-37, 1986 13. Fuselier HA, Moore K, Lindberg J, Husserl FE, Cole
FE, Kok DJ, Whitehead D, Galliano DJ, Erwin DT: Agglomeration inhibition reflected stone-forming activity during long-term potassium citrate therapy in calcium stone formers. Urology 52:988-994, 1998 14. Schell-Feith EA, Kist-van Holthe JE, Conneman N, Van Zwieten PH, Holscher HC, Zonderland HM, Brand R, Van Der Heijden BJ: Etiology of nephrocalcinosis in preterm neonates: Association of nutritional intake and urinary parameters. Kidney Int 58:2102-2110, 2000 15. Kok DJ, Papapoulos SE, Blomen LJ, Bijvoet OL: Modulation of calcium oxalate monohydrate crystallization kinetics in vitro. Kidney Int 34:346-350, 1988 16. Will EJ, Bijvoet OLM, Blomen LJMJ, van de Linden H: Growth kinetics of calcium oxalate monohydrate 1. J Cryst Growth 64:297-305, 1983 17. Blomen LJMJ, Will EJ, Bijvoet OLM, van de Linden H: Growth kinetics of calcium oxalate monohydrate 2. J Cryst Growth 64:306-315, 1983 18. Bijvoet OLM, Blomen LJMJ, Will EJ, van de Linden H: Growth kinetics of calcium oxalate monohydrate 3. J Cryst Growth 64:316-325, 1983 19. Werness PG, Bergert JH, Smith LH: Crystalluria. J Crystal Growth 53:166-181, 1981 20. Finlayson B: Where and how does urinary stone disease start?, in Van Reen R (ed): Idiopathic Urinary Bladder Stone Disease. Bethesda, MD, National Institutes of Health, 1977, pp 7-31 21. Burns JR, Finlayson B, Gauthier J: Calcium oxalate retention in subjects with crystalluria. Urol Int 39:36-39, 1984 22. Kok DJ, Khan SR: Calcium oxalate nephrolithiasis, a free or fixed particle disease. Kidney Int 46:847-854, 1994 23. Blomen LJMJ: Growth and agglomeration of calcium oxalate mono-hydrate. Doctoral Thesis, University of Leiden, The Netherlands, 1982 24. Kok DJ: Crystallization and stone formation inside the nephron. Scanning Microsc 10:471-484, 1996 25. de Water R, Noordermeer C, van der Kwast TH, Nizze H, Boeve ER, Kok DJ, Schroder FH: Calcium oxalate nephrolithiasis: Effect of renal crystal deposition on the cellular composition of the renal interstitium. Am J Kidney Dis 33:761-771, 1999 26. de Water R, Noordermeer C, Houtsmuller AB, Nigg AL, Stijnen T, Schroder FH, Kok DJ: Role of macrophages in nephrolithiasis in rats: An analysis of the renal interstitium. Am J Kidney Dis 36:615-625, 2000 27. Karlen J, Aperia A, Zetterstrom R: Renal excretion of calcium and phosphate in preterm and term infants. J Pediatr 106:814-819, 1985 28. Kok DJ, Schell-Feith EA: Risk factors for crystallization in the nephron: The role of renal development. J Am Soc Nephrol 10:S364-S370, 1999 (suppl 14) 29. Nikkila M, Koivula T, Jokela H: Urinary citrate excretion in patients with urolithiasis and normal subjects. Eur Urol 16:382-385, 1989
P
RETERM NEONATES have a high risk for nephrocalcinosis (NC) in the first months of life. The incidence varies between 17% and 64%, depending on study population, ultrasonographic (US) criteria, and US equipment.1-6 The cause of NC is multifactorial and has not yet been elucidated. Short-term complications include nephrolithiasis with obstruction of the urinary tract and urinary tract infection.2 The long-term outcome of NC in preterm neonates has not been defined, but small-scale studies suggest a decrease in renal function.7-9 In the last two decades, significant progress has been made in the identification and quantification of physicochemical processes involved in calcium oxalate urinary stone formation. The physicochemical processes of crystallization can be divided into thermodynamic, including supersaturation, and kinetic, including nucleation, growth, and agglomeration processes. In adults, the ability of urine to inhibit crystal agglomeration is an important protective mechanism against stone formation. In patients with calcium oxalate stone formation, this ability is decreased, and the extent of this decrease is
directly related to the frequency of stone formation.10,11 The ability to inhibit crystal agglomeration is a citrate-mediated process. Correction of low citrate excretion by citrate treatment increases this ability,10 and citrate treatment is an effective preventive therapy for adults with stone formation.12,13 We recently described a large population of preterm neonates and found that neonates with NC had a lower urinary (U)citratecalcium ratio than those without NC.14 This raised the question of whether the ability of urine to modify crystallization processes is disturbed in preterm neonates. From the Departments of Pediatrics and Endocrinology, Leiden University Medical Center, Leiden; and the Departments of Pediatric Urology and Pediatrics, Sophia Children’s Hospital, Rotterdam, The Netherlands. Received January 9, 2001; accepted in revised form July 6, 2001. Supported in part by grant no. NSN C97.1672 from the Dutch Kidney Foundation. Address reprint requests to Eveline A. Schell-Feith, MD, Van der Waalsstraat 42, 2313 VD, Leiden, The Netherlands. E-mail: schell_ [email protected] © 2001 by the National Kidney Foundation, Inc. 0272-6386/01/3806-0010$35.00/0 doi:10.1053/ajkd.2001.29218
American Journal of Kidney Diseases, Vol 38, No 6 (December), 2001: pp 1229-1234
1229
1230
SCHELL-FEITH ET AL
We investigated the ability of urine from preterm neonates to modulate the kinetics of calcium oxalate monohydrate (COM) crystallization during the first weeks of life, during the onset of NC, and at later times in the first 2 years, when NC tends to diminish. Data for preterm neonates was compared with values for full-term neonates and healthy adults. We investigated whether urine from preterm neonates with NC is less effective in inhibiting the separate processes of crystal formation compared with urine from neonates without NC. Moreover, we studied the correlation between crystal agglomeration and Ucitrate excretion among these subjects. PATIENTS AND METHODS
Patients In a prospective study, 57 preterm neonates with a mean gestational age of 28.1 ⫾ 1.9 (SD) weeks and mean birth weight of 1,047 ⫾ 313 g admitted to Leiden University Medical Center (Leiden, The Netherlands) between January 1997 and November 1998 were investigated. This study was part of a larger study of the cause of NC in preterm neonates.6,14 The ethics committee approved the study protocol. Informed consent was obtained after oral and written information had been given. Patients from outside the usual referral region were excluded from the study because of logistic problems. Patients with obvious underlying renal diseases, such as renal tubular acidosis, were excluded from the study. Failure to collect data occurred when patients died or were transferred to another hospital or when urine collection was not successful. Urine aliquots during the first 4 weeks of life and at term were collected in plastic bags. At term was defined as 38 to 42 weeks postconceptional age. Renal untrasonography was performed at 4 weeks of age and at term. If NC was present at term and the parents agreed to participate in the follow-up study, urine collection and renal untrasonography were repeated at ages 6, 12, and 24 months: 89 aliquots (from 57 patients at different times) were collected. When more than one urine sample was collected from a patient in the first 4 weeks, the average of urinary values was used in the analysis, providing the best estimate of urinary level for that patient during the entire period. This procedure avoids overrepresentation of patients with many samples in the analysis. Urine samples from 17 healthy full-term neonates with a gestational age of 38 to 43 weeks (mean, 41 weeks) were collected in the first week of life. Crystallization data for preterm neonates in the first weeks of life (n ⫽ 19) were compared with data for full-term neonates (n ⫽ 17) and healthy adults (n ⫽ 36), as described by Kok et al.10
Physicochemical Methods The effect of urine on COM crystallization was determined by means of a seeded crystal growth system, which has been described in detail.15-18 In this system, the solubil-
ity, growth, and agglomeration of calcium oxalate crystals are measured as three separate and system-independent parameters. The solubility of COM in urine is given by the solubility product, Lc. Lc is the concentration product of calcium and oxalate at equilibrium, when growth and dissolution of crystals in the solution are in equilibrium. It is expressed as the square root of Lc (公Lc; in millimolar). The difference between the Lc and actual concentration product of calcium and oxalate determines the degree of supersaturation. Growth of calcium oxalate crystals in urine is expressed as growth inhibition (GI). The growth constant of crystal growth with an additive, which is urine in this experiment, is less than the growth constant without an additive. GI is expressed as the growth constant with additive as a percentage of the growth constant without additive. Assessment of the agglomeration of COM crystals in urine is given by agglomeration inhibition, [tm] (in minutes). Agglomeration of crystals reduces the rate at which mineral components are taken up because it causes a decrease in total accessible crystal surface. It therefore is expressed in units of time. [tm] by additives is reflected as an increase in [tm]. In summary, a seeded crystal growth system is used in which crystallization processes are followed by measuring the uptake of calcium 45 (45Ca) into the crystal mass. By using aged and pregrown COM crystals and standardizing crystal handling, reproducible constant conditions for each experiment are ensured. Two experiments were performed. In the first, calcium and oxalate concentrations were increased to find the lowest concentration product at which 45Ca uptake caused by crystal growth occurs. This is the Lc for calcium oxalate in the sample tested. In the second experiment, a single calcium oxalate concentration product in the metastable range was used, and time varied and 45Ca uptake were measured. By means of a technique derived from previously described COM crystallization kinetics,18 parameters for crystal GI (percentage of GI) and [tm] can be calculated from the curves obtained.
Biochemical Methods Calcium, creatinine, and citrate were measured using a Hitachi analyzer (Roche Diagnostics, Almere, The Netherlands). Ucalcium levels less than 0.5 mmol/L are not detectable in our laboratory. When Ucalcium was not detectable, we used 0.5 mmol/L in our calculations. The pH of samples for oxalate determinations was adjusted to 2 before samples were frozen. Oxalate was measured manually by a coloric photometric method (Instruchemie, Hilversum, The Netherlands). Ratios of Ucalcium to creatinine, Uoxalate to creatinine, Ucitrate to creatinine, and Ucitrate to calcium were calculated.
Renal Ultrasonography To detect NC, one pediatric radiologist performed all US examinations, using state-of-the-art equipment with a 7- to 7.5-MHz small-part transducer (Toshiba, Zoetermeer, The Netherlands). Transverse and longitudinal images were made of both kidneys. The radiologist was not informed about the patient’s history or treatment. NC was defined as the pres-
CRYSTALLIZATION KINETICS IN PRETERM NEONATES Table 1.
1231
Effects of Urine From Full-Term Neonates in the First Week of Life and Preterm Neonates in the First 4 Weeks of Life on COM Crystallization Kinetics Full-Term Neonates (n ⫽ 17)
公Lc (mmol/L) GI (%) [tm] (min) Uca-cr (mol/mol) Uox-cr (mol/mol) Uci-cr (mol/mol) Uci-ca (mol/mol)
0.36 ⫾ 0.08 58 ⫾ 17 143 ⫾ 97 0.46 ⫾ 0.73 0.16 ⫾ 0.09 0.66 ⫾ 0.98 3.93 ⫾ 4.69
(0.25-0.5) (30-85) (45-330) (0.03-2.4) (0.06-0.39) (0.04-4.33) (0.33-18.1)
Preterm Neonates (n ⫽ 19)
0.27 ⫾ 0.1 64 ⫾ 21 81 ⫾ 32 2.20 ⫾ 1.74 0.39 ⫾ 0.21 0.52 ⫾ 0.45 0.28 ⫾ 0.16
(0.17-0.52) (13-90) (38-147) (0.47-6.44) (0.16-0.96) (0.08-1.7) (0.08-0.62)
P
⬍0.01 NS ⬍0.05 ⬍0.005 ⬍0.001 NS ⬍0.001
NOTE. Values expressed as mean ⫾ SD (range). Abbreviations: Uca-cr, urinary calcium-creatinine ratio; Uox-cr, urinary oxalate-creatinine ratio; Uci-cr, urinary citratecreatinine ratio; Uci-ca, urinary citrate-calcium ratio; NS, not significant.
ence of bright reflections in the medulla or cortex that were reproducible in both the transverse and longitudinal directions with or without acoustic shadowing. Reflections varied from small flecks 1 to 2 mm across to white dots larger than 2 mm to completely echodense pyramids. Two tiny echogenic parallel stripes were considered to be the arcuate or branch arteries.
Statistics To evaluate the effect of age on 公Lc, percentage of GI, [tm], and excretion of calcium and oxalate, we performed an analysis of variance, accounting for within-subject repeated measurements. The same analysis was performed to evaluate the effect of citrate on [tm]; this analysis was corrected for the effect of age. Groups were compared by means of Student’s t-test for normally distributed data and Mann-Whitney test for nonparametric data. Results are given as mean ⫾ SD.
were included in the figure. [tm] increased significantly with age (P ⬍ 0.005). 公Lc and percentage of GI did not change significantly with age, but Ucalcium-creatinine and Uoxalate-creatinine ratios decreased significantly with age (P ⬍ 0.01). Urine samples collected in the first weeks of life and at term were obtained from an unselected group of patients; however, sequential studies were obtained only from preterm neo-
RESULTS
Table 1 shows that mean 公Lc and mean [tm] were significantly lower in urine from preterm neonates in the first weeks of life compared with full-term neonates (P ⬍ 0.05). Mean Ucalciumcreatinine and Uoxalate-creatinine ratios were significantly greater for preterm neonates in the first weeks of life compared with full-term neonates, whereas the mean Ucitrate-calcium ratio was significantly lower (P ⬍ 0.005). Figure 1 shows the [tm] for preterm neonates in the first 4 weeks of life compared with full-term neonates and healthy adults.10 All but 1 ([tm] ⫽ 147 minutes) of the 19 preterm neonates measured in the first 4 weeks of life had a [tm] less than the lower limit for healthy adults (range, 145 to 584 minutes; mean ⫾ SEM, 275 ⫾ 23).10 Figure 2 shows repeated measurements of [tm] for preterm neonates in the first 2 years of life. Only patients with sequential data points
Fig 1. [tm] of calcium oxalate crystals in urine of preterm neonates, full-term neonates, and adults. (䊐) Preterm neonates in the first 4 weeks (n ⴝ 19) have a lower mean [tm] (81 ⴞ 32 minutes) compared with (Œ) full-term neonates in the first week (n ⴝ 17; 143 ⴞ 97 minutes; P < 0.05). The lower limit measured for healthy adults is indicated as a dashed line.10
1232
SCHELL-FEITH ET AL
Fig 2. Lines show repeated measurements of [tm] in preterm neonates in the first 2 years of life, which increase significantly with age (P < 0.005).
nates in whom NC was diagnosed at term. Therefore, a selected population was studied. Patients with (n ⫽ 8) and without NC (n ⫽ 16) at term did not differ statistically in mean 公Lc (0.34 ⫾ 0.18 versus 0.27 ⫾ 0.08 mmol/L), mean percentage of GI (63% ⫾ 20% versus 56 ⫾ 16), or mean [tm] (106 ⫾ 55 versus 121 ⫾ 48 minutes). Finally, [tm] and excretion of citrate correlated positively after correction for the effect of age, with a coefficient of 38 (95% confidence interval, 20 to 56; P ⬍ 0.001; Fig 3).
determine the degree of supersaturation. We found a lower Lc and greater excretion of calcium and oxalate for preterm compared with full-term neonates. Consequently, urine from preterm neonates in the first weeks will have a greater level of supersaturation than urine from full-term neonates. The following kinetic events occur during stone formation: nucleation, growth, and agglomeration. The process of nucleation cannot be measured directly because particles formed are too small. In vitro measurement instruments, described here, measure not only the previously mentioned effect of Lc, but the effect of growth and agglomeration on the crystallization process separately. It is well known that small particles can be formed and excreted in urine without leading to the formation of renal stones.19 In adults, urine from patients with stone formation can be clearly distinguished from that from healthy subjects only in its effect on crystal agglomeration. Kok et al10 found that adult stone formers lack a defensive mechanism (the ability of urine to inhibit crystal agglomeration) that ensures that the size of particles formed is restricted. This finding was confirmed by Erwin et al,11 who showed that [tm] decreases with increasing stone frequency, making it a useful tool to measure the risk for stone recurrence. Crystal percentage of GI did not change with age and was not different for preterm compared
DISCUSSION
Preterm neonates run a high risk for developing NC in the first months of life. In most patients, NC tends to diminish spontaneously in months to years,6 but the presence of crystals in the kidney at the very beginning of life might result in a decrease in renal function.7-9 The cause of NC is multifactorial and has not yet been elucidated fully.14 For adults, more knowledge is present about stone formation; however, the formation of Ucalcium oxalate stones is considered a complex and incompletely understood process. The physicochemical processes of stone formation can be divided into thermodynamic and kinetic processes. The thermodynamic process requires free energy and is determined by the level of supersaturation in urine. The Lc and concentrations of calcium and oxalate in urine
Fig 3. Correlation of calcium oxalate crystal agglomeration and Ucitrate concentration from preterm neonates. [tm] and citrate excretion positively correlate after correction for the effect of age (coefficient of 38; 95% confidence interval, 20 to 56; P < 0.001).
CRYSTALLIZATION KINETICS IN PRETERM NEONATES
with full-term neonates. The growth of crystals is a very slow process. Therefore, we did not expect to find an important role of this factor in the development of NC because the time urine spends in different nephron segments is relatively short.10,20-22 Conversely, agglomeration is a rapid process, and large particles can be formed within a short period.23,24 Our results show that urine from preterm neonates in the first weeks of life is less able to inhibit calcium oxalate crystal agglomeration. Compared with full-term neonates who are not at risk for developing NC, the mean [tm] in the first weeks of life was significantly less for preterm neonates. Furthermore, [tm] was less than the lowest value found by Kok et al10 for healthy adults for all except one preterm neonate. In preterm neonates, NC develops in the first weeks of life and subsequently gradually disappears in most patients in months to years. Consistent with this is our finding that [tm] increases with age, and the degree of supersaturation appears to decrease with age (although solubility did not change with age, both calcium and oxalate excretion decreased with age). Premature neonates excrete urine with a high level of supersaturation and diminished ability to inhibit agglomeration, but as they mature, saturation level and this ability improve. In a hyperoxaluric rat model, deposition of calcium oxalate crystals also disappeared when oxaluric pressure was reduced. It appears that the rat kidney carries out processes for the active removal of calcium deposits. These findings were confirmed in human renal material.25,26 It should be mentioned that the group of patients studied between the age of 6 and 24 months were a selected study population because we only followed up neonates with NC at term. However, it is theoretically unlikely that these results would not be reproducible for an unselected group of patients. Because preterm neonates without NC at term have an older gestational age at birth and greater birth weight than those with NC at term,14 those patients not studied most likely would have greater [tm] and lower Ucalcium-creatinine and Uoxalate-creatinine ratios in the period of 6 to 24 months. Thus, an unselected group of patients would have yielded, if anything, an even more striking increase in [tm] and decrease in calcium and ox-
1233
alate excretion with age. Furthermore, when Ucalcium was not detectable in our laboratory, we used 0.5 mmol/L in the calculations. We presume that this assumption did not influence our results considerably because Ucalcium-creatinine ratios were similar to the values of Karlen et al27 for preterm and full-term neonates in the first 2 weeks. Healthy full-term children hardly ever form stones. Recently, we described a theoretical model that makes it plausible that this low prevalence in children is caused by the following mechanism. Nephron segments in children are shorter than those in adults, which results in a shorter transit time for urine in the segments and lower concentration values in the loops of Henle of children.28 Therefore, although [tm] is often lower than in healthy adults, full-term neonates have a low risk for forming stones. We could not show a significant difference in effect on crystallization kinetics between preterm neonates who developed NC and those who did not. However, preterm neonates as a whole showed an increased level of supersaturation and decreased ability to inhibit crystal agglomeration. We hypothesize that all preterm neonates are at risk for developing NC, but that in some preterm neonates, the balance between the various stone-inhibiting and stone-promoting factors is more negative than in others. An important and clinically relevant question is which factors influence the crystallization processes. Kok et al10 showed that citrate, both as a pure compound18 and in urine from adults,10 is a very potent inhibitor of crystal agglomeration. This study shows a correlation of citrate and [tm] in preterm neonates. In addition, preterm neonates have a low Ucitrate-calcium ratio compared with full-term neonates.29 Similar to the mechanism in adults, citrate seems to be a major modulator of the agglomeration process in preterm neonates as well. In adults, correction of low citrate excretion by citrate treatment increases the ability to inhibit agglomeration,10 whereas an increase in [tm] reflects a decrease in stone-formation rate in citrate-treated patients.13 Our results suggest that citrate treatment of preterm neonates may influence positively the balance of stone-inhibiting and stone-stimulating factors and thus may potentially be effective preventive therapy.
1234
SCHELL-FEITH ET AL
We conclude that urine from preterm neonates is highly supersaturated, with a defective ability to inhibit calcium oxalate crystal agglomeration in the first weeks of life. This ability improves with age and, similar to adults, is citrate mediated. We suggest that both the high level of supersaturation and defective ability to inhibit calcium oxalate crystal agglomeration contribute to the high incidence of NC in preterm neonates. REFERENCES 1. Jacinto JS, Modanlou HD, Crade M, Strauss AA, Bosu SK: Renal calcification incidence in very low birth weight infants. Pediatrics 81:31-35, 1988 2. Downing GJ, Egelhoff JC, Daily DK, Alon U: Furosemide-related renal calcifications in the premature infant. A longitudinal ultrasonographic study. Pediatr Radiol 21:563565, 1991 3. Short A, Cooke RW: The incidence of renal calcification in preterm infants. Arch Dis Child 66:412-417, 1991 4. Katz ME, Karlowicz MG, Adelman RD, Werner AL, Solhaug MJ: Nephrocalcinosis in very low birth weight neonates: Sonographic patterns, histologic characteristics, and clinical risk factors. J Ultrasound Med 13:777-782, 1994 5. Saarela T, Vaarala A, Lanning P, Koivisto M: Incidence, ultrasonic patterns and resolution of nephrocalcinosis in very low birthweight infants. Acta Paediatr 88:655-660, 1999 6. Schell-Feith EA, Holscher HC, Zonderland HM, Kistvan Holthe JE, Conneman HN, van Zwieten PHT, Brand R, van der Heijden AJ: Ultrasonic features of nephrocalcinosis in preterm neonates. Br J Radiol 73:1185-1191, 2000 7. Ezzedeen F, Adelman RD, Ahlfors CE: Renal calcification in preterm infants: Pathophysiology and long-term sequelae. J Pediatr 113:532-539, 1988 8. Downing GJ, Egelhoff JC, Daily DK, Thomas MK, Alon U: Kidney function in very low birth weight infants with furosemide-related renal calcifications at ages 1 to 2 years. J Pediatr 120:599-604, 1992 9. Jones CA, King S, Shaw NJ, Judd BA: Renal calcification in preterm infants: Follow up at 4-5 years. Arch Dis Child Fetal Neonatal Ed 76:F185-F189, 1997 10. Kok DJ, Papapoulos SE, Bijvoet OL: Crystal agglomeration is a major element in calcium oxalate urinary stone formation. Kidney Int 37:51-56, 1990 11. Erwin DT, Kok DJ, Alam J, Vaughn J, Coker O, Carriere BT, Lindberg J, Husserl FE, Fuselier H Jr, Cole FE: Calcium oxalate stone agglomeration reflects stone-forming activity: Citrate inhibition depends on macromolecules larger than 30 kilodalton. Am J Kidney Dis 24:893-900, 1994 12. Pak CY, Fuller C: Idiopathic hypocitraturic calciumoxalate nephrolithiasis successfully treated with potassium citrate. Ann Intern Med 104:33-37, 1986 13. Fuselier HA, Moore K, Lindberg J, Husserl FE, Cole
FE, Kok DJ, Whitehead D, Galliano DJ, Erwin DT: Agglomeration inhibition reflected stone-forming activity during long-term potassium citrate therapy in calcium stone formers. Urology 52:988-994, 1998 14. Schell-Feith EA, Kist-van Holthe JE, Conneman N, Van Zwieten PH, Holscher HC, Zonderland HM, Brand R, Van Der Heijden BJ: Etiology of nephrocalcinosis in preterm neonates: Association of nutritional intake and urinary parameters. Kidney Int 58:2102-2110, 2000 15. Kok DJ, Papapoulos SE, Blomen LJ, Bijvoet OL: Modulation of calcium oxalate monohydrate crystallization kinetics in vitro. Kidney Int 34:346-350, 1988 16. Will EJ, Bijvoet OLM, Blomen LJMJ, van de Linden H: Growth kinetics of calcium oxalate monohydrate 1. J Cryst Growth 64:297-305, 1983 17. Blomen LJMJ, Will EJ, Bijvoet OLM, van de Linden H: Growth kinetics of calcium oxalate monohydrate 2. J Cryst Growth 64:306-315, 1983 18. Bijvoet OLM, Blomen LJMJ, Will EJ, van de Linden H: Growth kinetics of calcium oxalate monohydrate 3. J Cryst Growth 64:316-325, 1983 19. Werness PG, Bergert JH, Smith LH: Crystalluria. J Crystal Growth 53:166-181, 1981 20. Finlayson B: Where and how does urinary stone disease start?, in Van Reen R (ed): Idiopathic Urinary Bladder Stone Disease. Bethesda, MD, National Institutes of Health, 1977, pp 7-31 21. Burns JR, Finlayson B, Gauthier J: Calcium oxalate retention in subjects with crystalluria. Urol Int 39:36-39, 1984 22. Kok DJ, Khan SR: Calcium oxalate nephrolithiasis, a free or fixed particle disease. Kidney Int 46:847-854, 1994 23. Blomen LJMJ: Growth and agglomeration of calcium oxalate mono-hydrate. Doctoral Thesis, University of Leiden, The Netherlands, 1982 24. Kok DJ: Crystallization and stone formation inside the nephron. Scanning Microsc 10:471-484, 1996 25. de Water R, Noordermeer C, van der Kwast TH, Nizze H, Boeve ER, Kok DJ, Schroder FH: Calcium oxalate nephrolithiasis: Effect of renal crystal deposition on the cellular composition of the renal interstitium. Am J Kidney Dis 33:761-771, 1999 26. de Water R, Noordermeer C, Houtsmuller AB, Nigg AL, Stijnen T, Schroder FH, Kok DJ: Role of macrophages in nephrolithiasis in rats: An analysis of the renal interstitium. Am J Kidney Dis 36:615-625, 2000 27. Karlen J, Aperia A, Zetterstrom R: Renal excretion of calcium and phosphate in preterm and term infants. J Pediatr 106:814-819, 1985 28. Kok DJ, Schell-Feith EA: Risk factors for crystallization in the nephron: The role of renal development. J Am Soc Nephrol 10:S364-S370, 1999 (suppl 14) 29. Nikkila M, Koivula T, Jokela H: Urinary citrate excretion in patients with urolithiasis and normal subjects. Eur Urol 16:382-385, 1989