METHODS

Blood was obtained from 50 infants and children (23 girls, 27 boys, aged 0.2 to 17 years; Table 1) with normal renal function, after obtaining informed consent at the Children's Memorial Hospital in Chicago, Illinois, USA (N = 15) and the University Children's Hospital in Hamburg, Germany (N = 35). Infants and children with known renal diseases, intestinal disorders, and medications possibly interfering with oxalate or citrate metabolism were not eligible for the study. Dietary intakes were not restricted.

Table 1 - TABLE 1. Selected characteristics of study participants.

Full table (25K)

To hinder the in vitro neogenesis of oxalate, lithium-heparinized blood (3 ml) was placed directly on ice and immediately centrifuged at 1000 g for five minutes at 4°C, within 10 minutes of venipuncture. Plasma was then ultrafiltered at 1500 g for 20 minutes at 4°C using a Centrisart I ultrafiltration vial (Sartorius Co., NY, USA). Plasma was placed in the outer chamber and 40 l of 1 M HCl per ml plasma was added in the inner chamber to ensure simultaneous acidification of the ultrafiltrate (pH < 1.8)⁵.

We tested the necessity of simultaneous acidification by splitting four samples. Instead of the second ultracentrifugation step mentioned above, plasma from these samples was split, placed on ice and was only acidified with concentrated HCl to a pH below 2.7.

All plasma samples were stored at -20°C until analysis. Plasma analysis was performed at the Children's Memorial Hospital in Chicago, and therefore all samples from Germany were packed on dry ice and sent via express mail to the analyzing lab. All samples included in the study were adequately frozen at arrival.

For measurement of plasma oxalate (P_ox), citrate (P_cit), sulfate (P_sulf) and phosphate, plasma was diluted with 0.3 mM H₃BO₃ (1:5 or 1:10) and injected automatically onto an ion chromatography system (DX-500; Dionex Corp., Sunnyvale, CA, USA), equipped with an analytical column (AS11) and a guard-column (AG11) as the stationary phase. The mobile phase was NaOH; the 50% solution (J.T. Baker, Phillipsburg, NJ, USA) was diluted with water (> 17.5 megaohm resistance) to concentrations of 5 mM and 100 mM. The NaOH was continuously degassed with N₂, and run at 5 mM through 10 minutes, and then as a linear gradient from 5 mM to 52.5 mM through 21 minutes. The eluent background conductivity was suppressed with an anion self-regenerating suppressor (ASRS-I; Dionex Corp.) to a level below 3 S at the highest sodium hydroxide concentration. Computer-based software (Peaknet; Dionex Corp.) was used to calculate the concentration of the measured anion peaks Figure 1). Aqueous calibration standards (0.625 to 10 M for oxalate, 0.625 to 10 M for citrate and 1.25 to 20 M for sulfate and phosphate) were run daily. Aqueous controls (2.5 and 7.5 M) of each analyte were run before and after analysis of the plasma samples.

Figure 1.

Typical ion chromatogram for plasma oxalate, citrate and sulfate determination. Phosphate was run as an internal control; peaks without standard calibrations are not labeled.

Full figure and legend (21K)

Since our calibrations and controls were aqueous, we measured oxalate, citrate and sulfate repetitively in a single plasma sample, and this was viewed as a baseline value. To that sample, we added known quantities of oxalate, citrate and sulfate (+10, +20, +50 mol/liter) and remeasured these specimens multiple times (N = 7).

To establish the reliability and reproducibility of the method, multiple (3) determinations of oxalate, citrate, and sulfate (as well as phosphate, whose levels served as internal patient controls) were performed from the first 29 plasma samples obtained.

In addition to the plasma anion concentrations, serum levels of total calcium, phosphorus, magnesium, sodium, potassium, chloride, bicarbonate and the free-flowing venous pH were measured in each participant by standard laboratory procedure⁶. Serum creatinine was determined using a modified Jaffé method and glomerular filtration rate was calculated using the formula of Schwartz et al⁷.

Results of plasma anion levels are expressed as mol/liter SD. Linear regression analysis was performed between plasma anion concentrations and all other parameters.

The study was approved by the Institutional Review Board of the Children's Memorial Hospital in Chicago, Illinois, USA.

RESULTS

Specimens from 50 participants (23 girls, 27 boys) with normal renal function were eligible for analyses (Table 1). An additional 8 samples were not preserved correctly at the time of initial venipuncture, or were thawed long before analysis. Measurements made in the excluded samples showed P_ox values to be 17 to 28% higher than the established upper normal level from samples correctly processed (data not shown). P_ox obtained in samples preserved without the second step of centrifugation was considerably higher compared to the adequately preserved samples (8.6 vs. 5.0 mol/liter; 10.8 vs. 7.3 mol/liter; 8.8 vs. 7.5 mol/liter; 6.6 vs. 5.9 mol/liter). A difference in plasma citrate and sulfate levels was not observed. All values presented represent measurements made on plasma samples that had the ultrafiltrate acidified.

Multiple determinations of individual plasma anion levels revealed a high degree of reproducibility for each and a low variation for all plasma anions (Table 2). Phosphorus levels, which were determined as internal control of validity, were within the normal age related values for each child (data not shown)⁸; as a group, values decreased from infancy to adolescence. Plasma phosphorus levels determined by ion chromatography were comparable to those values obtained by a phosphomolybdate reaction monitored spectrophotometrically (r² = 0.92, P < 0.03).

Table 2 - TABLE 2. Coefficients of variation for plasma analytes.

Full table (14K)

Multiple determinations of plasma controls and added concentrations (+10 mol, + 20 mol, + 50 mol) of either oxalate, citrate and sulfate yielded complete concordance with added and measured concentrations (Table 3).

Table 3 - TABLE 3. Repeated determination (N = 7) of plasma controls and simultaneously added known quantities of oxalate, citrate and sulfate (+ 10, + 20 and + 50 mol).

Full table (21K)

The mean levels for plasma oxalate, for plasma citrate and for plasma sulfate did not differ based on age or gender (Table 4 and Figure 2).

Figure 2.

Plasma oxalate (A), citrate (B) and sulfate (C) concentrations (mol/liter) as a function of age in 50 children. There is a lack of significant correlation: in A, r = 0.01; in B, r = 0.06; in C, r = 0.11. The long dash represents the 95% confidence interval and the medium dash is the prediction interval.

Full figure and legend (11K)

Table 4 - TABLE 4. Normative data for plasma oxalate, citrate and sulfate levels in children.

Full table (16K)

In normal children, plasma citrate levels correlated with both venous pH (r = 0.42, P < 0.03) and serum bicarbonate (r = 0.40, P < 0.03). No correlation was found between serum potassium and plasma citrate (r = 0.05). There was no correlation between plasma sulfate and any parameter studied. Venous pH (range 7.34 to 7.41), as well as serum bicarbonate levels (range 21.0 to 26.4 mmol/liter) were normal for age in each participant.

DISCUSSION

We established a precise and reproducible method for the determination of relevant plasma anions involved in mineral metabolism and in the pathogenesis of nephrolithiasis/nephrocalcinosis. We demonstrated that plasma preservation and adequate storage are crucial for plasma oxalate determination. We derived normal values for plasma oxalate, citrate and sulfate that are independent of age and gender.

Using previous methods these plasma anions were not easily measurable. Plasma oxalate determinations have been found to be very unreliable, especially in the very low normal ranges reported here. Experiments with plasma oxalate determination have shown increases in apparent oxalate concentration secondary to oxalate neogenesis. Thus, in 1980 Akcay and Rose found P_ox values ranging from 1 to 16 mol/liter in inadequately preserved samples, possibly due to the oxidation of glyoxylate to oxalate⁹. Values clearly decreased (0.0 to 5.4 mol/liter) when inhibitors of this reaction were added. We also observed that plasma has to be very carefully preserved to ensure a reliable plasma oxalate determination. P_ox levels obtained in thawed plasma samples were clearly elevated. Additionally, samples without a second centrifugation step for simultaneous acidification showed notably higher P_ox values.

The ion chromatographic method described in this study was found to be both reproducible and reliable with low coefficients of variation for all anions (Table 2). We measured plasma phosphorus simultaneously by this method, on the same samples as an internal control, and found values comparable to plasma phosphosphorus levels determined in standard fashion by a phosphomolybdate reaction, and within the normal age-related range for each participant. Therefore, we believe our values for oxalate, citrate and sulfate are valid for normative data. Such plasma analyses may give additional information about the underlying pathophysiology of several diseases which cause nephrolithiasis, or might be helpful in the diagnosis of primary hyperoxaluria or renal tubular derangements with changes in citrate excretion and pH regulation.

There are a number of reported plasma oxalate levels in normal adults, using different analytical methods that include gas chromatography, oxalate oxidase or other types of ion chromatography^{10,11,12,13,14,15,16,17}. P_ox values range from 0.6 to 2.8 mol/liter (oxalate oxidase)¹⁰ up to 6.75 2.62 mol/liter using ion chromatography¹³. Our values of 6.43 1.06 mol/liter would therefore fit in this spectrum, but would nevertheless be located at the upper range of the values achieved by other methods, and in adults.

Clearly, lower values in healthy children (ranging from 0.78 to 3.02 mol/liter with a geometric mean of 1.53 mol/liter) were reported by Barratt et al¹⁸. This discrepancy between their values and those reported here is most likely due to methodological differences, as they employed an enzymatic assay using oxalate oxidase for P_ox determination¹⁸. We can eliminate the concern that oxalate neogenesis induced an increase in P_ox in our samples, since sample preservation started within 10 minutes after blood was withdrawn and equivalent P_ox levels were obtained in 29 samples with multiple determinations over 21 days, once preserved.

Kasidas and Rose reported gender specific differences in P_ox levels, with significantly higher values in females¹⁵, but we and others cannot support their findings^{10,11,12,13,14, 16,17,18}. Also, we could not demonstrate any age related variations, which, in contrast, have been found for urinary oxalate/creatinine ratios in infancy and childhood¹⁹.

Citric acid, a tricarboxylic acid, is a very potent inhibitor of urinary calcium-oxalate and calcium-phosphate crystallization. Much of the plasma citrate is complexed to calcium, magnesium and sodium, while bone is the other major reservoir of (plasma) citrate²⁰. Therefore, the mean P_cit levels reported in adults were relatively low (100 mol/liter)²¹. Increases in plasma citrate levels were observed after exercising, ingestion of an oral citrate load and after prolonged fasting^22,23,24.

Normal values of plasma citrate in infants and children have not previously been published. The mean (free) plasma citrate levels of 79.3 27.4 mol/liter determined in our study are comparable to the values previously obtained in adults²¹. All of our children studied were normal with respect to acid-base homeostasis. Even so, we observed a positive correlation between both venous pH and serum bicarbonate to measured plasma citrate levels. These correlations might not be of great clinical importance, as plasma citrate has relatively little influence on renal tubular citrate handling and therefore urinary citrate excretion^25,26. Ongoing studies will attempt to elucidate the influence of plasma citrate levels on plasma saturation in patients with oxalosis and in patients with end-stage renal disease.

Inorganic plasma sulfate derives from the metabolism of sulfur containing amino acids. Sulfate is normally excreted completely via the kidney, or incorporated into glycosaminoglycans and sulfatides. Plasma levels for inorganic sulfate in childhood were reported by Michalk and Manz²⁷, who found mean plasma sulfate levels of 241 59 mol/liter, which are comparable to the levels found in our study. Cole et al observed higher plasma sulfate levels in neonates during the first weeks of life (471 24 mol/liter)²⁸. An increase in plasma sulfate concentration was found in chronic renal insufficiency²⁹, although its significance remains uncertain at present.

In conclusion, we established a reliable and reproducible ion chromatographic method for the simultaneous determination of plasma anions with minimal blood volume requirements. Normal values for plasma oxalate, citrate and sulfate were determined for healthy infants and children. Using these results, we can now study the role of plasma anions in disorders of oxalate metabolism, or renal tubular derangements with changes in urinary citrate excretion and systemic pH regulation.

References

References

Acknowledgments

B. Hoppe is supported by a grant (Ho 1272/4-1) of the Deutsche Forschungsgemeinschaft. We acknowledge the general support of the Children's Memorial Institute for Education and Research (CMIER).