Evidence for Both Abnormal Set Point of PTH Stimulation by Calcium and Adaptation to Serum Calcium in Hemodialysis Patients with Hyperparathyroidism

Evidence for Both Abnormal Set Point of PTH Stimulationby Calcium and Adaptation to Serum Calcium inHemodialysis Patients with Hyperparathyroidism

MARIANO RODRIGUEZ,4 FRANCISCO CARAVACA,1 ELVIRA FERNANDEZ,2 MARIA J. BORREGO,4

VICTOR LORENZO,3 JUAN CUBERO,1 ALEJANDRO MARTIN-MALO,4 ANGELS BETRIU,2

AURELIO P. RODRIGUEZ,3 and ARNOLD J. FELSENFELD5*

ABSTRACT

In vitro studies of parathyroid glands removed from dialysis patients with secondary hyperparathyroidism andhypercalcemia have demonstrated the presence of an increased set point of parathyroid hormone (PTH) stimu-lation by calcium (set point [PTHstim]), suggesting an intrinsic abnormality of the hyperplastic parathyroid cell.However, clinical studies on dialysis patients have not observed a correlation between the set point (PTHstim) andthe magnitude of hyperparathyroidism. In the present study, 58 hemodialysis patients with moderate to severehyperparathyroidism (mean PTH 780 6 377 pg/ml) were evaluated both before and after calcitriol treatment toestablish the relationship among PTH, serum calcium, and the set point (PTHstim) and to determine whetherchanges in the serum calcium, as induced by calcitriol treatment, modified these relationships. Calcitriol treat-ment decreased serum PTH levels and increased the serum calcium and the setpoint (PTHstim); however, theincrease in serum calcium was greater than the increase in the setpoint (PTHstim). Before treatment with calcitriol,the correlation between the set point (PTHstim) and the serum calcium was r 5 0.82, p < 0.001, and between theset point (PTHstim) and PTH was r 5 0.39, p 5 0.002. After treatment with calcitriol, the correlation between theset point (PTHstim) and the serum calcium remained significant (r5 0.70, p< 0.001), but the correlation betweenthe set point (PTHstim) and PTH was no longer significant (r 5 0.09); moreover, a significant correlation waspresent between the change in the set point (PTHstim) and the change in serum calcium that resulted fromcalcitriol treatment (r 5 0.73, p < 0.001). The correlation between the residual values (deviation from theregression line) of the set point (PTHstim), derived from the correlation between PTH and the set point (PTHstim),and serum calcium was r5 0.77, p< 0.001 before calcitriol and r5 0.72, p< 0.001 after calcitriol. In conclusion,the set point (PTHstim) increased after a sustained increase in the serum calcium, suggesting an adaptation of theset point to the existing serum calcium; the increase in serum calcium resulting from calcitriol treatment wasgreater than the increase in the set point (PTHstim); the set point (PTHstim) was greater in hemodialysis patientswith higher serum PTH levels; and the correlation between PTH and the set point (PTHstim) may be obscuredbecause the serum calcium directly modifies the set point (PTHstim). (J Bone Miner Res 1997;12:347–355)

INTRODUCTION

IN IN VITRO STUDIES, parathyroid glands removed from hy-percalcemic dialysis patients have been shown to have an

increased set point of parathyroid hormone (PTH) stimu-lation by calcium (set point [PTHstim]),

(1–4) which at least intwo studies has been associated with an increase in intra-cellular calcium.(3,4) Moreover, it also has been shown indialysis patients that control of PTH release was moreabnormal in nodular than adjacent parathyroid tissue.(3,5)

In dialysis patients, nodular hyperplasia of the parathyroid*Work performed as Visiting Professor of Medicine, Facultad de

Medicina, Cordoba, Spain.

1Nephrology Services from Hospital Infanta Cristina, Badajoz, Spain2Hospital Universitario Arnau de Vilanova, Lleida, Spain3Hospital Universitario, Tenerife, Spain4Hospital Universitario Reina Sofia, Cordoba, Spain and the Departments of Medicine, University of Lleida, Lleida, Spain5West Los Angeles VA Medical Center and UCLA, Los Angeles, California, U.S.A.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 12, Number 3, 1997Blackwell Science, Inc.q 1997 American Society for Bone and Mineral Research

347

gland is associated with greater severity of secondary hy-perparathyroidism,(6,7) greater parathyroid gland weightthan with diffuse hyperplasia,(7–9) greater percent of cells inthe proliferative S phase of the cell cycle,(10) and a higherincidence of recurrent hyperparathyroidism after parathy-roidectomy.(10,11) In addition, several studies have shownthat the magnitude of PTH secretion in dialysis patients isassociated with the parathyroid gland mass.(6,7,12,13) Thus, itwould not be unreasonable to expect that dialysis patientswith more severe hyperparathyroidism would have a greaterparathyroid gland mass with more nodular hyperplasia anda greater set point (PTHstim). However, in studies in hemo-dialysis patients, the set point (PTHstim) often was notincreased, did not correspond to the magnitude of hyper-parathyroidism,(14) and moreover, did not decrease after acalcitriol-induced reduction in PTH.(15–19) As a result ofthese studies in dialysis patients, there has been consider-able confusion about the clinical significance of the set point(PTHstim) in patients with secondary hyperparathyroidism.It is possible that in the in vivo setting, factors other than

a specific intrinsic property of the parathyroid gland affectthe set point (PTHstim). In an earlier study,

(20) we reportedthat the set point (PTHstim) was greater in patients withosteitis fibrosa than adynamic bone with respective meanPTH levels of 1075 and 94 pg/ml. However, a more re-cent study failed to observe a difference in the set point(PTHstim) between dialysis patients with osteitis fibrosa andthose with adynamic bone, and in that study, PTH levels didnot correlate with the set point (PTHstim).

(21) Moreover,in a recent study in which hemodialysis patients with simi-lar predialysis PTH levels were separated according tothe serum calcium, we observed that patients with a lowpredialysis serum calcium also had a lower set point(PTHstim).

(22) As a result, we began to question whether theset point (PTHstim) was modified by the actual serum cal-cium and whether this could obscure a correlation betweenthe set point (PTHstim) and the magnitude of hyperpara-thyroidism.(1–4) Thus, if during calcitriol treatment the in-crease in serum calcium were responsible for the increase inthe set point (PTHstim), this would help to explain how theset point (PTHstim) could increase even though PTH secre-tion decreased. Nevertheless, it is important to establishwhether, as suggested by in vitro studies,(1–4) the set point(PTHstim) reflects the magnitude of hyperparathyroidism.The goal of the present study in 58 hemodialysis patients

with moderate to severe hyperparathyroidism was to eval-uate the relationship among PTH, the set point (PTHstim),and the predialysis serum calcium in order to establishwhether a correlation is present between PTH and the setpoint (PTHstim) and to determine whether changes in theserum calcium, as induced by treatment with calcitriol,modified the set point (PTHstim) and the correlation be-tween the set point (PTHstim) and PTH.

MATERIALS AND METHODS

Fifty-eight hemodialysis patients from four different he-modialysis units in Spain were studied. The patients werefrom hemodialysis units in Badajoz, Tenerife, Lleida, and

Cordoba. The mean age of the patients was 51.9 6 15.7years; 24 patients were male and 34 patients were female.The mean duration of hemodialysis was 83.1 6 53.1months. The primary phosphate binder was calcium carbon-ate; in some patients, aluminum hydroxide was added if theserum phosphorus was not adequately controlled. Predialy-sis PTH levels ranged from 252 to 1761 pg/ml. Patients hadnever received calcitriol (n 5 24) or had not receivedcalcitriol for at least 3 months before entry into the study(n 5 34).The dynamics of PTH secretion were evaluated before

and after receiving calcitriol. The criteria for receiving cal-citriol was a predialysis PTH level .250 pg/ml. Thirty-ninepatients received thrice weekly bolus intravenous calcitriol,12 patients received thrice weekly bolus oral calcitriol, and7 patients received daily oral calcitriol. The dose of bolusintravenous calcitriol ranged from 3–6 mg/week; the dose ofbolus oral calcitriol ranged from 3–7.5 mg/week; and thedose of oral calcitriol was 6 mg/week. The duration ofcalcitriol treatment before restudy was 2–3 months in 49patients and 6 months in 9 patients. While the patientsreceived calcitriol, the dialysate calcium concentration wasreduced to 2.5 meq/l. Before calcitriol treatment, the dia-lysate calcium was 3.0 or 3.5 meq/l. Patient consent wasobtained, a consent form signed, and the study was ap-proved by the local ethics committees.As described previously,(15,16,20,22) to determine maximal

PTH secretion and suppression, a low calcium (0.5 mM)hemodialysis and a high calcium (2 mM) hemodialysis wereperformed on separate dialysis days within 1 week of eachother; these studies were performed before the start ofcalcitriol and at the end of calcitriol treatment. During thelow and high calcium studies, blood samples for PTH andionized calcium were obtained at 15- to 30-minutes inter-vals. From the data obtained during dialysis-induced hypo-and hypercalcemia, the following terms were defined: (1)basal PTH was the predialysis PTH level; (2) maximal PTHwas the highest PTH level observed in response to hypocal-cemia and an additional reduction of the serum calcium didnot further increase the PTH value; (3) minimal PTH wasthe lowest PTH level during suppression by hypercalcemiaand a further increase in the serum calcium did not result inany additional decrease in PTH; (4) the ratio of basal tomaximal PTH was the basal PTH divided by the maximalPTH and this fraction was multiplied by 100 to provide apercentage; in normal volunteers, this ratio is 25%(23); (5)the set point of PTH stimulation by calcium (set point[PTHstim]) was defined as we have done previous-ly(15,16,20,22) as the serum calcium concentration at whichmaximal PTH secretion was reduced by 50%. Moreover, ashas been done in other studies,(14,21) the set point (PTHstim)was also calculated by the method of Brown(1) in which theset point (PTHstim) is the serum calcium at the midrangebetween the minimal and maximal PTH; and (6) the basalserum calcium was the serum calcium concentration at thebasal (predialysis) PTH. (7) The serum calcium at maximalPTH (CAmax) was the serum calcium concentration atwhich the PTH level was first observed to be maximal orwithin 10% of the maximal PTH; this definition was usedbecause the PTH-calcium curve is sigmoidal, and as the

348 RODRIGUEZ ET AL.

PTH value approaches the asymptotic portion of the curve,considerable variation in serum calcium can be observedduring small changes in PTH; similarly, for the same rea-son, (8) the serum calcium at minimal PTH (CAmin) was theserum calcium concentration at which the PTH level wasfirst observed to be minimal or within 10% of the minimalPTH.Intact PTH was measured with an immunoradiometric

assay for parathyroid hormone (Allegro, Nichols Institute,San Juan Capistrano, CA, U.S.A.). Normal values are10–65 pg/ml, and the range of the standard curve is 0–1600pg/ml; appropriate dilutions were performed for highervalues. During the low and high calcium studies, serumionized calcium was measured at bedside with a calciumselective electrode (Ciba Corning c-634 [Ciba Corning, Es-sex, U.K.] in two dialysis units and Radimeter ICA [Rad-imeter, Copenhagen, Denmark] in the other two, both witha coefficient of variation of 1%). At all other times, serumcalcium and phosphorus were measured with standard lab-oratory techniques. In each patient both before and aftercalcitriol, the basal concentration of calcium, phosphorus,and PTH were considered the mean of two predialysis (highand low dialysate calcium) values. The individual values ofPTH and ionized calcium were plotted and the parathyroidfunction curve was hand fitted. Serum aluminum was mon-itored in the four hemodialysis units during the study periodand was consistently less than 40 mg/l in all study patients.For comparisons between the two groups before and

after calcitriol treatment, the paired Student t-test was used.Linear regression analysis was performed to evaluate thepotential association between two variables; the slopes ofthe regression lines were calculated according to themethod of Brace.(24) Multiple and stepwise regression wereperformed to analyze association between several variables.To analyze the dispersion from the regression line, a com-parison was performed between residual values and a thirdvariable which potentially could have affected the residual.Results are expressed as the mean 6 SD, and significancewas defined as a p value ,0.05.

RESULTS

Shown in Table 1 are the before and after calcitriol basalserum calcium, serum phosphorus, set point (PTHstim) ascalculated by both methods, basal PTH, maximal PTH,minimal PTH, the serum calcium at maximal PTH, theserum calcium at minimal PTH, and the ratio of basal/maximal PTH. After calcitriol treatment, the basal serumcalcium, serum phosphorus, the serum calcium at maximalPTH, the serum calcium at minimal PTH, and set point(PTHstim) increased while the basal PTH and the basal/maximal PTH ratio decreased. The minimal and maximalPTH were not different before and after calcitriol treat-ment. The different regimens of calcitriol did not have asignificant effect on the parameters analyzed.Shown in Table 2 are the correlations observed before

and after calcitriol treatment among the basal PTH, maxi-mal PTH, minimal PTH, serum calcium, set point of PTHstimulation by calcium, serum phosphorus, and the basal/maximal PTH ratio. Before treatment with calcitriol, thebasal (predialysis) ionized calcium and the set point(PTHstim) were highly correlated (r 5 0.82, p , 0.001)(Fig. 1A). After treatment with calcitriol, the correlationbetween the basal ionized calcium and the set point re-mained significant (r 5 0.70, p , 0.001) (Fig. 1B). Beforecalcitriol treatment, a wide range of serum ionized calciumconcentrations was observed, and the majority of the pa-tients (n 5 41) had ionized serum calcium concentrationslower than 1.15 mM; however, after calcitriol treatment, theserum ionized calcium increased, and in 16 patients theserum calcium was greater than 1.30 mM. As shown inTable 2, the respective values for these correlations whenthe set point (PTHstim) was calculated by the method ofBrown were similar, r5 0.78, p, 0.001 before calcitriol andr 5 0.68, p , 0.001 after calcitriol. Moreover, as shown inFig. 2, a high degree of correlation was present between thechange (delta) in the set point (PTHstim) and the change(delta) in basal ionized calcium after calcitriol treatment(r 5 0.73, p , 0.001).Before calcitriol treatment, the correlation between the

TABLE 1. BIOCHEMICAL DATA AND PARAMETERS OF THE PTH-CALCIUM CURVE BEFORE AND AFTER CALCITRIOL TREATMENT

Precalcitriol Postcalcitriol P value Normals*

Basal ionized calcium (mM) 1.106 0.07 1.21 6 0.15 ,0.001 1.22 6 0.03Serum phosphorus (mg/dl) 5.736 1.51 6.78 6 2.03 ,0.001 3.30 6 0.51Setpoint-1 (PTHstim) (mM) 1.10 6 0.07 1.17 6 0.07 ,0.001 1.18 6 0.03Setpoint-2 (PTHstim) (mM) (Brown) 1.076 0.07 1.13 6 0.07 ,0.001 1.17 6 0.03Basal PTH (pg/ml) 7806 377 608 6 490 ,0.001 27 6 7Maximal PTH (pg/ml) 15616 762 1369 6 1283 NS 112 6 32CAmax (mM) 0.91 6 0.10 0.99 6 0.09 ,0.001 0.98 6 0.02Minimal PTH (pg/ml) 2766 211 266 6 249 NS 7 6 2CAmin (mM) 1.26 6 0.08 1.30 6 0.09 ,0.01 1.44 6 0.04Basal/maximal PTH (%) 516 15.1 44 6 15.6 ,0.01 25 6 13

Mean 6 SD. Number of patients 5 58 pre- and postcalcitriol. Set point (PTHstim): setpoint of PTH stimulation by calcium. Set point-1:calculated as 50% of PTHmax; set point-2: calculated as (PTHmax 2 PTHmin/2. *Obtained from Brent et al.

(23) The normal values forCAmax and CAmin published by Brent et al.

(23) represent the end of a long asymptotic segment of the curve.

PTH RELEASE IN SECONDARY HYPERPARATHYROIDISM 349

basal PTH and the set point (PTHstim) was significant (r 50.39, p 5 0.002) (Fig. 3); moreover, this correlation wassimilar when the set point (PTHstim) was calculated by themethod of Brown, r 5 0.42, p 5 0.001 (Table 2). Aftercalcitriol treatment, the correlation between the basal PTHand the set point (PTHstim) was no longer significant(Table 2) nor was the correlation between the delta setpoint and the delta basal PTH (r 5 0.09). However, aninverse correlation (r 5 20.57) was observed between theserum calcium and basal/maximal PTH ratio. To determinethe possible effects of the serum calcium, basal PTH, andbasal/maximal PTH ratio on the set point (PTHstim), amultiple regression followed by stepwise regression wasperformed before and after calcitriol treatment with the setpoint (PTHstim) as the dependent variable and the serumcalcium, basal PTH and basal/maximal PTH ratio as inde-pendent variables. Since controversy exists regarding thetwo definitions of the set point (PTHstim), set point 1 andset point 2 were each analyzed. As shown in Table 3, beforeand after calcitriol treatment for both definitions of theset point (PTHstim), the serum calcium and the basal/max-imal PTH ratio were highly correlated with the set point(PTHstim). However, for both the multiple regression andthe stepwise regression, the influence of the serum calciumwas considerably greater than that of the basal/maximalPTH ratio.When a multiple regression was performed between the

set point (PTHstim) (dependent variable) and the serumcalcium and PTH, both before and after calcitriol treat-ment, the addition of PTH modestly but significantly in-creased the correlation. The correlation increased from0.82 to 0.84 before and from 0.70 to 0.74 after calcitriol.The strong correlation between the serum calcium and

the set point (PTHstim) before and after calcitriol treatment

suggested the possibility that the serum calcium was thepredominant influence on the set point (PTHstim) and thuscould obscure any correlation between PTH and the setpoint (PTHstim). To evaluate this possibility, the residualvalues of the set point (PTHstim) from the correlation be-tween the basal PTH and the set point (PTHstim) werecorrelated with the serum calcium. (The residual values forthe set point were derived from the PTH–set point regres-sion line by subtracting for each individual point, the pre-dicted from the actual set point (PTHstim). The predictedset point (PTHstim) is that which corresponds to the actualset point-PTH regression line). As shown in Fig. 4, a sig-nificant correlation was present between the serum calciumand the residual values of the set point both before andafter calcitriol treatment, r 5 0.77, p , 0.001 and r 5 0.72,p , 0.001, respectively. Similar correlations were obtainedwhen the set point (PTHstim) was calculated by the methodof Brown, r 5 0.71, p , 0.001 (before calcitriol) and r 50.70, p , 0.001 (after calcitriol). Thus, these results wouldsuggest that the existing serum calcium strongly influencedthe set point (PTHstim) and distorted the relationship be-tween PTH and the set point (PTHstim).Since in previous publications the set point (PTHstim) has

not always been calculated by the same method, it has beenthe source of some controversy with regards to whether thetwo methods provide similar information.(14–16,20,21) Asshown in Table 2, the correlation between the two set pointswas high with an r value of 0.97 before and 0.95 aftercalcitriol treatment; moreover, the correlations between theset point (PTHstim), as calculated by both methods, and theother parameters shown in Table 2 were similar.The data shown in Table 1 indicate that the maximal

PTH did not change, and while a significant decrease inbasal PTH was observed, the decrement was not large.

TABLE 2. CORRELATIONS BEFORE AND AFTER CALCITRIOL TREATMENT

Serum PTHbasal PTHmax PTHmin Calcium Set Point-1 Set Point-2 Phosphorus

Before CalcitriolPTHbasal —PTHmax r 5 0.84 —PTHmin r 5 0.54 r 5 0.58 —Calcium r 5 0.29 r 5 0.34 r 5 0.48 —Set point-1 r 5 0.39 r 5 0.23 r 5 0.39 r 5 0.82 —Set point-2 (Brown) r 5 0.42 r 5 0.22 r 5 0.27 r 5 0.78 r 5 0.97 —Phosphorus r 5 044 r 5 0.32 r 5 0.12 r 5 0.09 r 5 0.12 r 5 0.15 —PTH basal/max r 5 0.22 r 5 20.29 r 5 20.07 r 5 20.17 r 5 0.20 r 5 0.25 r 5 0.29

After CalcitriolPTHbasal —PTHmax r 5 0.82 —PTHmin r 5 0.71 r 5 0.67 —Calcium r 5 20.24 r 5 0.02 r 5 20.04 —Set point-1 r 5 0.09 r 5 0.14 r 5 0.09 r 5 0.70 —Set point-2 (Brown) r 5 0.07 r 5 0.16 r 5 .05 r 5 0.68 r 5 0.95 —Phosphorus r 5 0.45 r 5 0.33 r 5 0.22 r 5 20.20 r 5 20.08 r 5 20.06 —PTH basal/max r 5 0.41 r 5 20.05 r 5 0.08 r 5 20.57 r 5 20.11 r 5 20.13 r 5 0.26

Number of patients 5 58 pre- and postcalcitriol. Set point-1: calculated as 50% of PTHmax; set point-2: calculated as (PTHmax 2PTHmin)/2. p , 0.05 for r value greater than 0.27; p , 0.01 for r value greater than 0.34. p , 0.001 for r value greater than 0.42.

350 RODRIGUEZ ET AL.

However, if we only consider the patients who responded tocalcitriol treatment, defined as a 40% decrease in basalPTH or a basal PTH of less than 300 pg/ml, the maximalPTH decreased in this group (n 5 27) from 1229 to 666pg/ml ( p , 0.001); conversely, in the nonresponder group(n 5 31), the maximal PTH was greater than in the re-sponder group (1229 vs. 1850 pg/ml; p 5 0.001) and did notchange after calcitriol (1850 vs. 1982 pg/ml). In addition,when the comparison between delta ionized calcium andthe delta set point (PTHstim), shown in Fig. 2, was separatedaccording to responders and nonresponders to calcitriol,the correlation remained significant and was similar for thetwo groups, r 5 0.83, p , 0.001 and r 5 0.72, p , 0.001,respectively.

DISCUSSION

The results of the present study demonstrate that inhemodialysis patients the set point (PTHstim) was highlycorrelated with the serum calcium concentration and, more-over, the calcitriol-induced modification of the serum cal-

cium was associated with parallel changes in the set point(PTHstim), suggesting that the set point (PTHstim) was mod-ified by the change in the serum calcium. Before calcitrioltreatment, a modest but significant correlation was ob-served between PTH and the set point (PTHstim). However,when the residual values of the set point (PTHstim) derivedfrom the PTH–set point regression analysis were correlatedwith the serum calcium, a high degree of correlation waspresent; this finding would suggest that the actual serumcalcium concentration obscured the correlation betweenPTH and the set point (PTHstim).Both before and after calcitriol treatment, the parameter

that best correlated with the set point (PTHstim) was thepredialysis (basal) serum calcium. Moreover, supportingthe idea that the set point (PTHstim) is strongly dependenton the serum calcium was the high correlation between the

FIG. 1. The correlation between the serum ionized cal-cium and the set point of PTH stimulation by calcium (setpoint [PTHstim]) is shown before (A) and after (B) treat-ment with calcitriol. CTR, calcitriol.

FIG. 2. The correlation between the calcitriol-inducedchange (delta) in ionized calcium and the calcitriol-inducedchange (delta) in the set point of PTH stimulation bycalcium (set point [PTHstim]) is shown.

FIG. 3. The correlation between the basal PTH and theset point of PTH stimulation by calcium (set point[PTHstim]) is shown before treatment with calcitriol. CTR,calcitriol.

PTH RELEASE IN SECONDARY HYPERPARATHYROIDISM 351

delta set point (PTHstim) and the delta serum calcium. Thefact that a significant correlation was also observed betweenthe set point (PTHstim) and the basal/maximal PTH ratiomay not indicate a cause-and-effect relationship. Sincechanges in serum calcium produced parallel effects on theset point (PTHstim) and the basal/maximal PTH ratio, thiscould have resulted in colinearity between the serum cal-cium and the basal/maximal PTH ratio.The finding that the set point (PTHstim) changes in as-

sociation with the change in serum calcium is also sup-ported by the results of several recent studies in hemodial-ysis patients(16,19,25,26) and the study by Keaton et al.(27) inwhich, for similar PTH levels, the PTH–calcium curve inhypercalcemic, neonatal calves was shifted to the right ofthe PTH-calcium curve in older, normocalcemic calves.Furthermore, since in the present study calcitriol treatmentsimultaneously decreased PTH as it increased the serumcalcium, it is not surprising that a significant correlation was

not present between the basal PTH and the set point(PTHstim) after calcitriol.It should be noted that the slope for the correlation

between the delta calcium and the delta set point (PTHstim)was 0.67 (Fig. 2); this indicates that the increase in serumcalcium was greater than the corresponding increase in theset point (PTHstim). Thus, these data would suggest that theincrease in serum calcium functions to suppress simulta-neously PTH secretion even as the PTH–calcium curvemoves to the right accompanying the new existing serumcalcium concentration. The basal–maximal PTH ratio be-fore calcitriol was 51%; this indicates that as compared withnormal parathyroid glands in which the basal–maximalPTH ratio is approximately 25%,(23) parathyroid gland se-cretion is stimulated at the basal state. After calcitriol treat-ment, the basal–maximal PTH ratio decreased to 44%,which was closer to normal than before calcitriol therapy.The decrease in the basal–maximal PTH ratio would not

TABLE 3. MULTIPLE REGRESSION

PrecalcitriolSet point-1 (50% of PTH) (dependent variable) vs. serum calcium, basal PTH, basal/max PTH (independent variables)

t value p value R2 (Seq) R2 (Simple)

Stepwise

t value p value

Ca 13.20 ,0.001 0.68 0.68 14.5 ,0.001PTH 0.95 50.34 0.70 0.15 0.9 50.34Basal/Max PTH 5.24 ,0.001 0.80 0.03 5.8 ,0.001

Set point-2 ([PTHmax 2 PTHmin]/2) (dependent variable) vs. serum calcium, basal PTH, basal/max PTH (independentvariables)

t value p value R2 (Seq) R2 (Simple)

Stepwise

t value p value

Ca 11.35 ,0.001 0.60 0.60 12.7 ,0.001PTH 0.78 50.43 0.64 0.17 0.8 50.43Basal/Max PTH 5.54 ,0.001 0.77 0.06 6.1 ,0.001

PostcalcitriolSet point-1 (50% of PTH) (dependent variable) vs. serum calcium, basal PTH, basal/max PTH (independent variables)

t value p value R2 (Seq) R2 (Simple)

Stepwise

t value p value

Ca 19.39 ,0.001 0.46 0.46 9.4 ,0.001PTH 1.73 50.08 0.55 0.007 1.7 50.08Basal/Max PTH 3.46 ,0.001 0.63 0.01 3.5 50.001

Set point-1 ([PTHmax 2 PTHmin]/2) (dependent variable) vs. serum calcium, basal PTH, basal/max PTH (independentvariables)

t value p value R2 (Seq) R2 (Simple)

Stepwise

t value p value

Ca 8.33 ,0.001 0.46 0.46 12.7 ,0.001PTH 1.71 50.09 0.53 0.005 0.8 50.43Basal/Max PTH 2.81 50.007 0.59 0.02 2.8 50.007

352 RODRIGUEZ ET AL.

have been possible if the increment in serum calciumafter calcitriol was not greater than that of the set point(PTHstim). This concept is graphically depicted in Fig. 5.Previous in vitro studies have noted that the set point

(PTHstim) was greater in parathyroid glands removed fromdialysis patients than in normal parathyroid glands.(1–4)

Interestingly, in one of these studies,(2) the set point(PTHstim) was not increased in the only hypocalcemic pa-tient. However, recent studies have reported conflictingresults on whether the set point (PTHstim) is greater inpatients with more severe hyperparathyroidism.(14,20,21)

Moreover, a reduction of hyperparathyroidism after calcit-riol treatment has not always been associated with a de-crease in the set point (PTHstim).

(15–19,28–30) The results ofthe present study would suggest that the association be-tween the set point (PTHstim) and the existing serum cal-cium is stronger than that between the set point (PTHstim)and PTH. Moreover, it should be recognized that the serumcalcium is also modifying PTH so that, independent of thedegree of hyperparathyroidism, hypocalcemia increases andhypercalcemia decreases the actual PTH level. Thus, if the

serum calcium is directly modifying the set point (PTHstim)(direct relationship) and in addition also affects PTH se-cretion (inverse relationship), it would be difficult to ob-serve, even when present, a correlation between PTH andthe set point (PTHstim). The fact that a significant correla-tion was observed between PTH and the set point (PTHstim)despite the obscuring effect of calcium would suggest thatthe correlation is real and would be stronger if the effect ofthe serum calcium were removed. This conclusion is sup-ported by the demonstration that the residual values of theset point for the correlation between PTH and the set pointcorrelated well with the serum calcium.In an earlier study in patients with low bone turnover,(20)

we observed a correlation between the serum calcium andthe set point (PTHstim), a result that was similar to ourfindings in the present study. We also observed that patientswith adynamic bone had a lower set point (PTHstim) thanpatients with osteitis fibrosa.(20) This latter result was likelydue to the fact that our patients with adynamic bone notonly had a lower PTH but also a lower serum calcium,which according to our present data is a compelling reasonto have a lower set point (PTHstim). In the continuousambulatory peritoneal dialysis (CAPD) patients with ady-namic bone of Sanchez et al.,(21) the serum calcium and theset point (PTHstim) were reported to be similar to patientswith osteitis fibrosa. Thus, since the buffering capacity ofthe bone is decreased in adynamic bone(31) and their CAPDpatients were on a 3.5 meq/l calcium dialysate which resultsin an increased transfer of calcium,(32) it was not surprisingthat these patients had a similar serum calcium and setpoint (PTHstim) as patients with osteitis fibrosa.In a recent study, Goodman et al. failed to find a corre-

FIG. 4. The correlation between the serum calcium andthe residual values of the set point of PTH stimulation bycalcium (set point [PTHstim]), obtained for the relationshipbetween the PTH and the set point (PTHstim), is shownbefore (A) and after (B) calcitriol treatment. CTR, cal-citriol.

FIG. 5. A schematic representation of the effect that asustained increase in the serum basal calcium has on PTHsecretion is shown. The sustained increase in the serumcalcium results in a shift of the PTH–calcium curve to theright with an increase in the set point of PTH stimulation bycalcium (set point [PTHstim]). However, since the set point(PTHstim) increases less than the basal serum calcium, theincrease in serum calcium still has a suppressive effect onPTH secretion.

PTH RELEASE IN SECONDARY HYPERPARATHYROIDISM 353

lation between the basal PTH and the set point (PTHstim)and concluded that abnormalities in the set point (PTHstim)do not account for increased PTH secretion in patients withchronic renal failure.(14) The results of the present studycontradict this conclusion. There are several possible rea-sons for our different conclusions. One is that the magni-tude of hyperparathyroidism may have been greater in ourstudy in which the number of patients with maximal PTHlevels greater than 1000 were 44/58 (76%) and the numberwith maximal PTH levels greater than 1500 pg/ml were25/58 (43%). This compares with the study of Goodman etal.(14) in which only 11/26 (42%) had maximal PTH levelsgreater than 1000 pg/ml and 5/26 (19%) had maximal PTHlevels greater than 1500 pg/ml. A second reason is that thecurrent study was performed in hemodialysis patients andthe other study in CAPD patients receiving a dialysate witha high calcium concentration (3.5 meq/l). Moreover, it ispossible that the calcium loading in these CAPD pa-tients(31) produced a situation analogous to our patientstreated with calcitriol in whom the serum calcium in-creased. A final consideration is that the number of patientsin the present study was considerably greater (58 vs. 26patients) and this would increase the power to determine adifference for any applied statistical analysis.The two different methods that have been advocated for

the calculation of the set point (PTHstim) have engenderedconsiderable confusion and controversy in the recent liter-ature because each gives a different value for the set point(PTHstim).

(14–16,20,21) The results of the present studywould indicate that although the set point (PTHstim) wasdifferent for the two methods, this difference is essentiallya constant and thus the information provided should besimilar. As such, the correlation coefficients for the two setpoints were 0.97 before and 0.95 after calcitriol. Moreover,when the set point (PTHstim) was calculated by the twomethods, both before and after calcitriol treatment, thecorrelations between the set point (PTHstim) and the otherparameters in Table 2 were similar, as were the correlationsbetween the serum calcium and the residuals of the setpoint (PTHstim). Thus, it is difficult to accept the claim thatone method for calculating the set point (PTHstim) is betterthan the other.(33–35)

In conclusion, our results indicate that (1) the set point(PTHstim) increased after a sustained increase in the serumcalcium suggesting an adaptation of the set point to theexisting serum calcium; (2) the increase in serum calciumresulting from calcitriol treatment was greater than theincrease in the set point; thus, the increase in serum calciumstill resulted in relative suppression of PTH; (3) as has beensuggested by in vitro studies, the set point of PTH stimula-tion by calcium was greater in hemodialysis patients withhigher PTH levels; and (4) the correlation between PTHand the set point (PTHstim) may be obscured because theserum calcium directly modifies the set point (PTHstim).

ACKNOWLEDGMENTS

The authors thank Dr. Armando Torres for his assistanceand suggestions during the performance of the study and

the writing of the manuscript. This study was supported bygrants from the FIS (grant No. 93/0630 and 93/0560), PlanNacional I 1 D (PB 92-0917) and Fundacion HospitalReina Sofia-CajaSur.

REFERENCES

1. Brown EM 1983 Four-parameter model of the sigmoidal rela-tionship between parathyroid hormone release and extracellu-lar calcium concentration in normal and abnormal parathyroidtissue. J Clin Endocrinol Metab 56:572–581.

2. Brown EM, Wilson RE, Eastman RC, Pallotta J, Marynick SP1982 Abnormal regulation of parathyroid hormone release bycalcium in secondary hyperparathyroidism due to chronic renalfailure. J Clin Endocrinol Metab 54:172–179.

3. Wallfelt C, Larsson R, Gylfe E, Ljunghall S, Rastad J, Aker-strom G 1988 Secretory disturbance in hyperplastic parathy-roid nodules of uremic hyperparathyroidism: Implication forparathyroid autotransplantation. World J Surg 12:431–438.

4. Wallfelt C, Gylfe E, Larsson R, Ljunghall S, Rastad J, Aker-strom G 1988 Relationship between external and cytoplasmiccalcium concentrations, parathyroid hormone release andweight of parathyroid glands in human hyperparathyroidism. JEndocrinol 116:457–464.

5. Rudberg C, Grimelius L, Johansson H, Ljunghall S, OdseliusR, Pertoft H, Rastad J, Wide L, Akerstrom G 1986 Alterationin density, morphology and parathyroid hormone release ofdispersed parathyroid cells from patients with hyperparathy-roidism. Act Path Microbiol Immunol Scand Sect A94:253–261.

6. Mendes V, Jorgetti V, Nemeth J, Lavergne A, LecharpentierY, Dubost C, Cournot-Witmer C, Bourdon R, Bourdeau A,Zingraff J, Drueke T 1983 Secondary hyperparathyroidism inchronic hemodialysis patients: A clinico-pathological study.Proc EDTA 20:731–738.

7. Malmaeus J, Grimelius L, Johansson G, Akerstrom G, Ljung-hall S 1984 Parathyroid pathology in hyperparathyroidism sec-ondary to chronic renal failure. Scand J Urol Nephrol18:157–166.

8. Krause MW, Hedinger CE 1985 Pathologic study of parathy-roid glands in tertiary hyperparathyroidism. Human Pathol16:772–784.

9. Hellman P, Akerstrom G, Ljunghall S, Rastad J 1989 Surgicalfindings and results of subtotal and total parathyroidectomy inhypercalcemic patients with uremic hyperparathyroidism. ActaChir Scand 155:573–582.

10. Tominaga Y, Tanaka Y, Sato K, Numano M, Uchida K, Falk-mer U, Grimelius L, Johansson H, Takagi H 1992 Recurrentrenal hyperparathyroidism and DNA analysis of autograftedparathyroid tissue. World J Surg 16:595–603.

11. Cagne ER, Urena P, Leite-Silva S, Zingraff J, Chevalier A,Sarfati E, Dubost C, Drueke TB 1992 Short and long termefficacy of total parathyroidectomy with immediate autograft-ing compared with subtotal parathyroidectomy in hemodialysispatients. J Am Soc Nephrol 3:1008–1017.

12. McCarron DA, Muther RS, Lenfesty B, Bennett WM 1982Parathyroid function in persistent hyperparathyroidism: Rela-tionship to gland size. Kidney Int 22:662–670.

13. Johnson WJ, McCarthy JT, VanHeerden JA, Sterioff S, GrantCS, Kao PC 1988 Results of subtotal parathyroidectomy inhemodialysis patients. Am J Med 84:23–32.

14. Goodman WG, Belin T, Gales B, Juppner H, Segre GV,Salusky IB 1995 Calcium-regulated parathyroid hormone re-

354 RODRIGUEZ ET AL.

lease in patients with mild or advanced secondary hyperpara-thyroidism. Kidney Int 48:1553–1558.

15. Rodriguez M, Felsenfeld AJ, Williams C, Pederson JA, LlachF 1991 The effect of long-term intravenous calcitriol adminis-tration on parathyroid function in hemodialysis patients. J AmSoc Nephrol 2:1014–1020.

16. Pahl M, Jara A, Bover J, Rodriguez M, Felsenfeld AJ 1996 Theset point of PTH stimulation by calcium and the reduction ofparathyroid hormone in hemodialysis patients. Kidney Int49:226–231.

17. Ali AA, Varghese Z, Moorhead JF, Baillod RA, Sweny P 1993Calcium set point progressively worsens in hemodialysis pa-tients despite conventional oral 1-alpha hydroxycholecalciferolsupplementation. Clin Nephrol 39:205–209.

18. Caravaca F, Cubero JJ, Jimenez F, Lopez JM, Aparicio A, CidMC, Pizarro JL, Liso J, Santos I 1995 Effect of the mode ofcalcitriol administration of PTH-ionized calcium relationshipin uraemic patients with secondary hyperparathyroidism.Nephrol Dial Transplant 10:665–670.

19. Hardy P, Shenouda M, Moriniere P, Legallais C, Elesper N,Boitte F, Brazier M, Fournier A 1995 Shift of calcium set-pointin uremia does not explain PTH hypersecretion nor the 1-alphaOH vitamin D3 suppressive effect on PTH secretion. AbstractBook, XIIIth International Congress of Nephrology, p. 413(abstract).

20. Felsenfeld AJ, Rodriguez M, Dunlay R, Llach F 1991 A com-parison of parathyroid-gland function in haemodialysis pa-tients with different forms of renal osteodystrophy. NephrolDial Transplant 6:244–251.

21. Sanchez CP, Goodman WG, Ramirez JA, Gales B, Belin TR,Segre GV, Salusky IB 1995 Calcium-regulated parathyroidhormone secretion in adynamic renal osteodystrophy. KidneyInt 48:838–843.

22. Felsenfeld AJ, Jara A, Pahl M, Bover J, Rodriguez M 1995Differences in the dynamics of PTH secretion in hemodialysispatients with marked secondary hyperparathyroidism. J AmSoc Nephrol 6:1371–1378.

23. Brent GA, Le Boff MS, Seely EW, Conlin PR, Brown EM 1988Relationship between the concentration and rate of change ofcalcium and serum intact parathyroid hormone levels in nor-mal humans. J Clin Endocrinol Metab 67:944–950.

24. Brace RA 1977 Fitting straight lines to experimental data.Am J Physiol 233:R94–R99.

25. Borrego J, Almaden Y, Martin-Malo A, Rodriguez M, AljamaP 1995 The effect of calcitriol on the set point of the PTH-calcium curve in hemodialysis patients. Abstract Book, XIIIthInternational Congress of Nephrology, p. 470 (abstract).

26. Borrego MJ, Felsenfeld AJ, Martin-Malo A, Alamaden Y,Concepcion MT, Aljama P, Rodriguez M 1996 Evidence foradaptation of the entire PTH-calcium curve to sustainedchanges in the serum calcium in hemodialysis patients. NephrolDial Transplant (in press).

27. Keaton JA, Barto JA, Moore MP, Gruel JB, Mayer GP 1978Aletered parathyroid response to calcium in hypercalcemicneonatal calves. Endocrinology 103:2161–2167.

28. Delmez JA, Tindira C, Grooms P, Dusso A, Windus DW,Slatopolsky E 1989 Parathyroid hormone suppression by intra-venous 1,25-dihydroxyvitamin D. J Clin Invest 83:1349–1355.

29. Malberti F, Surian M, Cosci P 1992 Effect of chronic intrave-nous calcitriol on parathyroid function and set point of PTHstimulation by calcium in dialysis patients with refractory sec-ondary hyperparathyroidism. Nephrol Dial Transplant7:822–828.

30. Dunlay R, Rodriguez M, Felsenfeld A, Llach F 1989 Directinhibitory effect of calcitriol on parathyroid function (sigmoi-dal curve) in dialysis patients. Kidney Int 36:1093–1098.

31. Kurz P, Monier-Faugere MC, Bognar B, Werner E, Roth P,Vlachojannis J, Malluche HH 1994 Evidence for abnormalcalcium homeostasis in patients with adynamic bone disease.Kidney Int 46:855–861.

32. Coburn JW 1993 Mineral metabolism and renal bone disease:Effects of CAPD versus hemodialysis. Kidney Int 43 (Suppl40):S92–S100.

33. Goodman WG, Salusky IB 1996 Parathyroid gland functionand the set point for PTH release: Understanding the availabledata. Nephrol Dial Transplant 11:16–18.

34. Felsenfeld AJ, Rodriguez M 1996 The set point of calcium—another view. Nephrol Dial Transplant 11:1722–1725.

35. Goodman WG, Salusky I 1996 The set point revisited, again.Nephrol Dial Transplant 11:1725–1727.

Address reprint requests to:Mariano Rodriguez, M.D.

Unit of InvestigationHospital Reina SofiaMenedez Pidal S/N

14004 Cordoba, Spain

Received in original form April 5, 1996; in revised form October28, 1996; accepted November 13, 1996.

PTH RELEASE IN SECONDARY HYPERPARATHYROIDISM 355