Anesth Analg 2004;99:1199-1204
© 2004 International Anesthesia Research Society
doi: 10.1213/01.ANE.0000132780.87036.F5
CRITICAL CARE AND TRAUMA
Parathyroid Hormone Secretion During Citrate Anticoagulated Hemodialysis in Acutely Ill Maintenance Hemodialysis Patients
Robert Apsner, MD*,
Diego Gruber
,
Walter H. Hörl, MD PhD, FRCP*, and
Gere Sunder-Plassmann, MD*
*Department of Medicine III, Division of Nephrology and Dialysis, and
Department of Medical Statistics, University of Vienna, Vienna, Austria
Address correspondence and reprint requests to Robert Apsner, MD, Clinique St. Thérèse, L-2763 Luxembourg. Address e-mail to robert.apsner{at}akh-wien.ac.at
 |
Abstract
|
|---|
Regional citrate anticoagulation during extracorporeal treatment is used in patients at risk for hemorrhage. We conducted a prospective clinical trial on the effect of large- versus small-dose calcium supplementation during citrate anticoagulated hemodialysis on ionized calcium and intact parathyroid hormone (iPTH). Twenty-five treatments were studied in 25 patients with active bleeding or at risk for hemorrhage. Sixteen patients received large-dose calcium (15 mmol/h), and 9 received small-dose calcium (5 mmol/h) substitution during treatment. Ionized calcium increased in 13 of 16 patients in the large-dose calcium group and decreased in 8 of 9 patients in the small-dose calcium group. Intact PTH decreased by 25% in the large-dose group and increased by 121% in the small-dose group (P = 0.0007 for 
; P = 0.007 for %). In the 14 patients in whom ionized calcium increased, iPTH decreased. In 10 of 11 patients in whom ionized calcium decreased, iPTH increased. The increase or decrease of ionized calcium was more predictive for changes in iPTH than was the calcium-substitution rate (R2 = 0.5526 versus 0.3962, respectively). We conclude that the behavior of iPTH can be influenced in a predictive manner by adjusting the calcium-substitution rate during treatment.
IMPLICATIONS: Bleeding complications during extracorporeal detoxification procedures can be reduced if regional citrate anticoagulation is used instead of heparin. Citrate anticoagulation can induce important changes in the secretion of parathyroid hormone.
|
Introduction
|
|---|
In patients with overt bleeding or those at increased risk for hemorrhage, the use of regional citrate is recommended for anticoagulation if hemodialysis or hemo(dia)filtration is performed (1–4). By continuous administration of trisodium citrate into the arterial port of the extracorporeal circuit, calcium ions are bound; thus, the coagulation cascade is interrupted in the circuit. To prevent symptomatic hypocalcemia, calcium chloride is infused into the venous return (1–8). A decrease of ionized calcium during treatment to 0.8 mmol/L may be acceptable, as long as symptoms of hypocalcemia—such as QT prolongation, paresthesia, or tetanic cramps—do not occur (9). However, even small decreases of blood ionized calcium levels stimulate the secretion of parathyroid hormone (PTH) (10,11). Low ionized calcium and high PTH levels have been associated with increased mortality in critically ill surgical patients (12,13). Moreover, decreasing of blood ionized calcium to less than 0.8 mmol/L can compromise myocardial contractility (14). Despite the known effect of citrate administration on calcium homeostasis (9,15), the pattern of PTH secretion during citrate anticoagulated hemodialysis has not yet been studied. Because we are convinced that, during acute illness, efforts should be continued to control hyperparathyroidism in maintenance hemodialysis patients, we tested calcium supplementation strategies for citrate anticoagulation that allow adequate control of the activity of the parathyroid gland. In this study, we investigated the acute changes of ionized calcium and intact PTH (iPTH) during citrate anticoagulated hemodialysis by using either a small- or a large-dose calcium supplementation regimen. Our hypothesis was that lower calcium-substitution rates would result in stimulation of the parathyroid gland, whereas higher calcium-substitution rates would suppress PTH secretion.
|
Methods
|
|---|
In the setting of an acute dialysis/intensive care unit (ICU), we studied, in a prospective manner, 28 acutely ill maintenance hemodialysis patients. They were scheduled for acute hemodialysis anticoagulated by citrate because of active bleeding or a high risk for hemorrhage. The inclusion criteria for the study were age older than 19 yr, no signs of hepatic failure, and hemoglobin values of at least 10 g/dL. They were assigned to receive either large-dose or small-dose calcium supplementation during treatment. The study protocol was approved by the Ethical Committee of the General Hospital and Medical School of Vienna, and written, informed consent was obtained from each participant enrolled.
Twenty-five treatments in 25 patients (16 large-dose and 9 small-dose; age, 19–87 yr; mean ± SD, 56 ± 14.6 yr; median, 56 yr) were evaluated. Because of randomization errors in the night shift, 16 patients (9 male and 7 female) were included in the large-dose group and 12 patients (9 male and 3 female) into the small-dose group. One patient of the small-dose group was excluded from the study after initiation of dialysis because his parathyroid gland had been resected years ago and his autograft was not functioning, which was not known at the time of enrollment. Two other patients of the small-dose calcium group were excluded because essential laboratory data were missing. The reasons for terminal renal failure were diabetes mellitus (n = 9), shrunken kidneys (n = 6), biopsy-proven glomerulonephritis (n = 3), polycystic kidney disease (n = 2), drug toxicity (n = 3), acute renal failure (n = 1), and reflux nephropathy (n = 1). Four patients had active bleeding, 15 had had surgery the day of or the day before treatment, and 6 patients were at high risk for hemorrhage for other reasons (central venous catheter insertion and recent gastrointestinal hemorrhage).
We performed 4 h of hemodialysis at a mean blood flow rate of 223 ± 22 mL/h (150–250 mL/h), and we used a commercially available calcium- and magnesium-free dialysate (FWI; B. Braun Ges.m.b.H., Maria Enzersdorf, Austria) with a reduced bicarbonate (26 mmol/L) and sodium (135 mmol/L) concentration (13). We set the dialysate flow at 500 mL/min. For anticoagulation, trisodium citrate 50 mmol/h (500 mmol/L; 100 mL/h; Mayrhofer Pharmazeutica Ges.m.b.H, Linz, Austria) was infused into the arterial port of the extracorporeal circuit (6). Calcium was substituted by infusing calcium chloride (500 mmol/L; Mayrhofer Pharmazeutica) (at two different infusion rates) into the venous return. Low-flux cellulose triacetate dialyzers (Nissho Corporation, Osaka, Japan) were used in 21 patients (13 in Group 1 and 8 in Group 2), and low-flux polysulfone filters (Fresenius Medical Care, Bad Homburg, Germany) were used in the remaining 4 patients (3 in Group 1 and 1 Group 2). The mean ultrafiltration rates were 2200 and 1889 mL in Groups 1 and 2, respectively (P = 0.55).
After reviewing the charts of routine citrate anticoagulated hemodialysis treatments, we defined two starting doses for calcium substitution at which we expected a modest increase (large dose) or decrease (small dose) of systemic ionized calcium. For the large-dose calcium group, we chose an initial calcium infusion rate of 15 mmol/h (30 mL/h) instead of our usual infusion rate of 10 mmol/h (20 mL/h). The initial calcium infusion rate in the small-dose calcium group was set at 5 mmol/h (10 mL/h). If ionized calcium levels increased by more than 20% of baseline or more than 1.5 mmol/L (which was the case in four treatments), we reduced the infusion rate by steps of 2.5 mmol/h (5 mL/h). If ionized calcium decreased by more than 20% of baseline or less than 0.8 mmol/L, which occurred during four dialyses, we increased the calcium supplementation by steps of 2.5 mmol/h (5 mL/h). The resulting mean calcium-substitution rates were 14.1 ± 1.5 mmol/h (28.3 ± 3.0 mL/h) and 6.0 ± 1.1 mmol/h (11.9 ± 2.2 mL/h) in the large- and small-dose groups, respectively (P < 0.0001).
Before and after dialysis, samples for determination of iPTH, calcitonin, osteocalcin, 25-hydroxyvitamin D3, and 1,25-dihydroxyvitamin D3 were drawn from the vascular access (arteriovenous fistula or central venous catheter) into 8-mL tubes containing a clot activator (C. A. Greiner und Söhne Ges.m.b.H., Kremsmünster, Austria). Two hours after the start of dialysis (T = 120), an additional sample was drawn from the inlet blood line (before the dialyzer and before citrate infusion).
Intact PTH, calcitonin, and osteocalcin were determined by using immunoradiometric assays (ELSA-PTH, ELSA-hct, and ELSA-OSTEO; CIS Bio International, Gif/Yvette, France). 25-Hydroxyvitamin D3 and 1,25-dihydroxyvitamin D3 were determined by extraction and purification of vitamin D metabolites, followed by competitive radioimmunoassays (125I RIA Kits for 25-Hydroxyvitamin D and 1,25-Hydroxyvitamin; DiaSorin, Stillwater, MN).
Samples for determination of blood urea nitrogen (BUN) and electrolytes were drawn before (T = 0) and 2 min after the end of treatment (T = 242) from the vascular access into 8-mL tubes containing a clot activator (C. A. Greiner und Söhne Ges.m.b.H.) and were processed in the routine laboratory with standard methods. To estimate the amount of detoxification, the urea reduction rate (URR) was calculated [URR = 1 – (C1/C0), where C1 indicates BUN after dialysis and C0 indicates BUN before dialysis].
At 0 and 242 min, samples for determination of whole-blood ionized calcium and blood gas analysis were drawn from the vascular access into heparinized syringes (Arterial Blood Sampling Kit; SIMS Portex Inc., Keene, NH). Additional samples were drawn from the inlet blood line (before the dialyzer and before citrate infusion) 15 min after the start of dialysis and at 120 min. The analyses were performed by point-of-care testing with ion-selective electrodes (BGElectrolytes Analyzer; Instrumentation Laboratory, Lexington, KY).
At the end of each dialysis, the arterial and venous drip chambers and the filter were inspected for visible signs of coagulation. Corresponding to the standard of care at our department, a semiquantitative score with a range from 0 (no signs of clotting) to 4 (complete occlusion) was used for evaluation of clotting in the extracorporeal circuit. Values of 0 or 1 were considered to indicate sufficient anticoagulation. Eventual citrate-associated adverse effects were detected by systemic ionized calcium measurements, continuous electrocardiograph tracing, noninvasive automatic arterial blood pressure recordings, and clinical observation for signs of hypocalcemia.
Laboratory variables at the beginning and at the end of dialysis, as well as changes of these variables in the two groups (large dose versus small dose), were compared by using Student’s t-tests. We used Holm’s method for adjusting for multiplicity to minimize the possibility of differences arising by chance alone. Analysis of covariance was used to determine whether the course of ionized calcium during treatment more accurately predicted the observed changes in iPTH than assignment to either the large-dose or the small-dose calcium group. The clotting scores were compared between the groups by using Fisher’s exact test and Wilcoxon’s two-sample test.
Skewed data of the primary end-point are reported as the median and the minimum and maximum values (iPTH and osteocalcin). All other data are expressed as mean ± SD. Relative changes are reported as the mean percentage change (%). A P value <0.05 was considered to indicate statistical significance.
|
Results
|
|---|
In the large-dose calcium group (Group 1), ionized calcium increased slightly from 1.10 ± 0.16 mmol/L to 1.17 ± 0.10 mmol/L (increase in 13 of 16 patients and decrease in 3) (P = 0.12). The 3 patients in whom a decrease of ionized calcium was observed were those with the highest baseline ionized calcium levels (1.50, 1.27, and 1.25 mmol/L). In the small-dose calcium group (Group 2), a significant decrease of ionized calcium from 1.12 ± 0.13 mmol/L to 0.89 ± 0.10 mmol/L (P = 0.0012) was observed (decrease in 8 of 9 patients). The patient in whom ionized calcium increased despite the low substitution rate was the one with the lowest baseline value (0.87 mmol/L). The values (mean of : 0.07 ± 0.18 mmol/L in Group 1 and –0.23 ± 0.14 mmol/L in Group 2) were significantly different between groups (P = 0.0002). Details on the behavior of ionized calcium during treatment are shown in Figure 1 and Table 1.
View this table:
[in this window]
[in a new window]
|
Table 1. Demographic Data, Dialysis-Related information, and Results of Ionized Calcium and Intact Pass Thyroid Hormone (iPTH) During Citrate Anticoagulated Hemodialysis
|
|
The median iPTH decreased from 132.5 pg/mL (range, 29.4–489 pg/mL) to 54.9 pg/mL (range, 4.1–308 pg/mL) in the large-dose calcium group (decrease in 13 of 16 patients; P = 0.01), whereas in the small-dose group, iPTH increased from 220 pg/mL (range, 64.5–1370 pg/mL) to 406 pg/mL (range, 158–1710 pg/mL; increase in 7 of 9 patients; P = 0.045). The difference between groups was significant (P = 0.0007). The mean of the relative changes was a 25.4% decrease in Group 1 and a 121.4% increase in Group 2 (P = 0.007 for %). In the 14 patients in whom ionized calcium increased during the procedure, a decrease of iPTH was observed (100%), whereas in 10 of 11 patients in whom ionized calcium decreased, an increase of iPTH occurred (91%) (Table 1). Analysis of covariance confirmed that the increase or decrease of ionized calcium during treatment was more predictive of changes in iPTH than was the initial calcium-substitution rate (R2 = 0.5526 versus 0.3962, respectively; Fig. 2). When the variables "increase or decrease of ionized calcium" and "group" were combined, a modest increase of the predictive value was observed (R2 = 0.6311). Absolute values of ionized calcium did not predict the behavior of iPTH during treatment.
Median osteocalcin decreased from 60.1 ng/mL (range, 4.9–422 ng/mL) to 41.5 ng/mL (range, 3.7–351 ng/mL) in Group 1 (P = 0.0014) and from 65.2 ng/mL (range, 12–610 ng/mL) to 53.3 ng/mL (range, 7.2–648 ng/mL) in Group 2 (P = 0.21). Significant differences were not observed between groups (P = 0.51 for the absolute differences; P = 0.83 for %). We observed no significant changes for calcitonin, 25-hydroxyvitamin D3, or 1,25- dihydroxyvitamin D3.
Citrate-associated adverse events were not observed. The URR was similar in the two groups (0.60 ± 0.08 versus 0.60 ± 0.11; P = 0.88). During dialysis, the overall mean pH increased from 7.36 to 7.39 (P = 0.005). Significant differences between groups for pH were not observed. The control of sodium, potassium, magnesium, and phosphate was adequate (Table 2). In the large-dose calcium group, the mean clotting scores for the dialyzer and the arterial and venous drip chambers were 0.31 (median, 0; range, 0–2), 0.06 (median, 0; range, 0–1), and 0.19 (median, 0; range, 0–1), respectively. In the small-dose calcium group, the clotting scores for the dialyzer and the drip chambers were 0. The difference between groups was not statistically significant (P = 0.13, P = 0.51, and P = 0.20).
View this table:
[in this window]
[in a new window]
|
Table 2. Sodium, Phosphorus, and Magnesium Values Before and After Citrate Anticoagulated Hemodialysis with Either Large-Dose or Small-Dose Calcium Supplementation
|
|
|
Discussion
|
|---|
We provide the first evidence that regional citrate anticoagulation during hemodialysis induces profound changes of iPTH. Small-dose calcium supplementation results in decreased ionized calcium concentrations that are associated with a significant increase of iPTH. In contrast, large-dose calcium supplementation increases ionized calcium and decreases iPTH. In our study, the best independent single predictor of the changes in iPTH was the increase or decrease of ionized calcium during treatment. High or low values of ionized calcium per se did not influence the reaction of the parathyroid gland. As a consequence of our findings, we target the calcium substitution during citrate anticoagulated hemodialyses to achieve stable levels of ionized calcium hemodialysis during treatment. In the setting of chronically intermittent hemodialysis, we achieve excellent control of the activity of the parathyroid gland with this strategy.
Citrate is increasingly used for anticoagulation during renal replacement therapies in the ICU (1–8,15–17). Regardless of the treatment modality (hemodialysis or hemodiafiltration; continuous or intermittent), the administration of citrate induces profound changes in calcium homeostasis that require correction by adequate calcium supplementation. In the current literature, calcium replacement is judged sufficient if clinical signs of hypocalcemia—such as paresthesia, tetany, hypotension, or QT prolongation—are absent. In view of our results, one could speculate that more attention should be paid to the behavior of iPTH during (citrate anticoagulated) extracorporeal treatments in the ICU. For maintenance hemodialysis patients, the necessity to control iPTH secretion is undisputed, and it has been shown that subtle variations of the calcium balance during hemodialysis (by manipulating the calcium concentration in the dialysate) have clinically significant long-term effects on PTH secretion and bone metabolism (18). The role of the parathyroid gland in critical illness is unclear, but disturbed calcium homeostasis and secretion of PTH have been associated with poor outcome in the ICU (12,13) and with bone loss during the stay (19–24). Of course, our short-term intervention trial does not provide data on the duration and long-term effects of the citrate-induced changes of serum PTH levels, and it does not clarify the relevance of these changes for the patient in the ICU. However, because of the increasing use of citrate anticoagulation in the ICU, we should become aware of the fact that citrate induces important perturbations of the calcium and PTH homeostasis. Further research should establish the most appropriate calcium-substitution strategies for citrate use in the ICU.
In summary, our results show that variations of ionized calcium during citrate anticoagulation induce acute changes of iPTH secretion. The relevance of these changes for critically ill patients is not known. In view of the increasing use of regional citrate anticoagulation in the ICU (3–5), further research is needed to understand the role of PTH and calcium homeostasis in the critically ill patient undergoing renal replacement therapies.
|
References
|
|---|
- Flanigan MJ, von Brecht J, Freeman RM, Lim VS. Reducing the hemorrhagic complications of hemodialysis: a controlled comparison of low-dose heparin and citrate anticoagulation. Am J Kidney Dis 1987; 9: 147–53.[Web of Science][Medline]
- Pinnik RV, Wiegmann TB, Diederich DA. Regional citrate anticoagulation for haemodialysis in the patient at high risk for bleeding. N Engl J Med 1983; 308: 258–61.[Web of Science][Medline]
- Apsner R, Druml W. More on anticoagulation for continuous haemofiltration [letter]. N Engl J Med 1998; 338: 131–2.[Free Full Text]
- Palsson R, Niles JL. Regional citrate anticoagulation in continuous venovenous hemofiltration in critically ill patients with a high risk of bleeding. Kidney Int 1999; 55: 1991–7.[Medline]
- Kutsogiannis DJ, Mayers I, Chin WD, Gibney RT. Regional citrate anticoagulation in continuous venovenous hemodiafiltration. Am J Kidney Dis 2000; 35: 802–11.[Web of Science][Medline]
- Apsner R, Buchmayer H, Lang T, et al. Simplified citrate anticoagulation for high-flux hemodialysis. Am J Kidney Dis 2001; 38: 979–87.[Web of Science][Medline]
- Morita Y, Johnson RW, Dorn RE, Hill DS. Regional anticoagulation during hemodialysis using citrate. Am J Med Sci 1961; 242: 32–43.[Medline]
- Hocken AG, Hurst PL. Citrate regional anticoagulation in haemodialysis. Clin Nephrol 1992; 37: 36–41.[Medline]
- Ludbrook J, Wynn V. Citrate intoxication. Br Med J 1958; 30: 523–8.
- Schwarz P, Sorensen HA, Transbol I, McNair P. Induced hypocalcemia controlled by a citrate clamp technique, and the intact parathyroid hormone response obtained. Scand J Clin Lab Invest 1990; 50: 891–7.[Medline]
- Brown EM. Extracellular calcium, regulation of parathyroid cell function, and role of calcium and other ions as extracellular messengers. Physiol Rev 1991; 71: 371–411.[Free Full Text]
- Carlstedt F, Lind L, Rastad J, et al. Parathyroid hormone and ionized calcium levels are related to the severity of illness and survival in critically ill patients. Eur J Clin Invest 1998; 28: 898–903.[Medline]
- Lind L, Carlstedt F, Rastad J, et al. Hypocalcemia and parathyroid hormone secretion in critically ill patients. Crit Care Med 2000; 28: 93–9.[Web of Science][Medline]
- Olinger GN, Hottenrott C, Mulder DG, et al. Acute clinical hypocalcemic myocardial depression during rapid blood transfusion and postoperative hemodialysis: a preventable complication. J Thorac Cardiovasc Surg 1976; 72: 503–11.[Abstract]
- Janssen MJFM, Huijgens PC, Bouman AA, et al. Citrate anticoagulation and divalent cations in hemodialysis. Blood Purif 1994; 12: 308–16.[Medline]
- Apsner R, Hörl WH, Sunder-Plassmann G. Dalteparin-induced alopecia in hemodialysis patients: reversal by regional citrate anticoagulation [letter]. Blood 2001; 97: 2914–5.[Free Full Text]
- Unver B, Sunder-Plassmann G, Hörl WH, Apsner R. Long-term citrate anticoagulation for high-flux hemodialysis in a patient with heparin induced thrombocytopenia type II. Acta Med Austriaca 2002; 29: 146–8.[Web of Science][Medline]
- Martin KJ, Gonzalez EA. Strategies to minimize bone disease in renal failure. Am J Kidney Dis 2001; 38: 1430–6.[Web of Science][Medline]
- Smith LM, Cuthbertson B, Harvie J, et al. Increased bone resorption in the critically ill: association with sepsis and increased nitric oxide production. Crit Care Med 2002; 30: 837–40.[Medline]
- Aris RM, Lester GE, Ontjes DA. Bone loss physiology in critically ill patients. Chest 1998; 114: 954–5.[Free Full Text]
- Nierman DM, Mechanick JI. Bone hyperresorption is prevalent in chronically critically ill patients. Chest 1998; 114: 1122–8.[Abstract/Free Full Text]
- Lind L, Ljunghall S. Critical care hypercalcemia: a hyperparathyroid state. Exp Clin Endocrinol 1992; 100: 148–51.[Medline]
- Massry SG, Arieff AI, Coburn JW, et al. Divalent ion metabolism in patients with acute renal failure: studies on the mechanism of hypocalcemia. Kidney Int 1974; 5: 437–45.[Medline]
- Zaloga GP. Hypocalcemia in critically ill patients. Crit Care Med 1992; 20: 251–62.[Web of Science][Medline]
Accepted for publication May 4, 2004.
Top
Abstract
Introduction
; n = 16) or small-dose (
; n = 9) calcium substitution. 
P < 0.001 for the 
for ionized calcium during the procedure) on serum concentrations of intact parathyroid hormone (iPTH) (R2 = 0.3962 and 0.5526, respectively).
