Clinical Endocrinology (1992) 37,51 1-519
Circadian rhythm and pulsatility of parathyroid hormone secretion in man Klaus Herfarth', Heinrich Schmidt-Gaykt, Stefanie GraV and Andreas Maiert *Department of Surgery, Div.2.I, University of Heidelberg and tEndocrine Laboratory, Im Breitspiel 15, 6900 Heidelberg, Germany (Received 8 April 1992; returned for revision 21 May 1992; finally revised 19 June 7992; accepted 30 July 1992)
Summary OBJECTIVE We wished to investigate the circadian rhythm and pulsatility of parathyroid hormone (PTH) secretion in man, as conflicting results have been published. DESIGN and PATIENTS To investigate the circadian rhythm during daytime, we sampled (a) peripheral blood at hourly intervals in 12 healthy young men from 0900 h until 1700 h. For observation of pulsatility, we sampled (b) peripheral blood at 1-minute intervals for 1 hour in three healthy men and three healthy women (mean 27.7 years, range 21-56 years) and (c) at 1-minute intervals for 30 minutes in 21 patients with surgically confirmed primary hyperparathyroidism (pHPT). MEASUREMENTS The serum levels of intact PTH were measured by two-site immunoradiometric assay and special care was taken to reduce intra-assay variability, especially at the normal PTH concentration. In series (a), ionized calcium, total calcium and phosphate were also determined. RESULTS A circadian rhythm during daytime was found for intact PTH in healthy men and women with a nadir at 0930 h and a peak in the afternoon. Ionized calcium and total calcium (protein-adjusted)decreased and phosphate increased in the afternoon. These changes were all statistically significant (P < 0.02). Pulsetility of PTH: Statistical cluster analysis of the data showed no pulsatility either in healthy persons or in patients with primary hyperparathyroidism. In two healthy women and one healthy man slight changes of longer duration were discovered, but no complete pulses. In five patients with primary hyperparathyroidism, larger differences between the highest and lowest concentrations of intact PTH were found, but no complete pulses. CONCLUSIONS Our data show a signlflcant clrcadian Correspondence: Professor Dr med. H.Schmidt-Gayk, Im Breitspiel 15, D-6900 Heidelberg, Germany.
rhythm during daytime of intact PTH and only minor changes from minute to minute. The alterations in PTHlevels occurred at longer time intervals in healthy persons. In some patients with primary hyperparathyroidism, decreases of PTH-levels were found. The circadian rhythm of PTH may be due to slight changes in calcium or phosphate concentration.
Circadian rhythm is known in many hormones and pulsatility of secretion is very well established for ACTH and gonadotrophic hormones (Gallagher et al., 1973; Perks et al., 1988; Veldhuis &Johnson, 1988; Veldhuis et al., 1987). The change in normal concentration seems to be important for physiological and pathophysiological effects. Disease-specific patterns of rhythm or pulsatility have been described in various diseases (Kazer et a/., 1987; Linkowski et al., 1987; Liu et al., 1987; Spratt et a/., 1987). The hormonal effect might be changed by different pulsatile pattern of secretion (Gross et al., 1987; Paolisso et al., 1988) and this has also been considered for parathyroid hormone (PTH) secretion (Hesch et al., 1988). For parathyroid hormone several authors have described a circadian variation (Jubiz et al., 1972; LoCascio et al., 1982; Logue et al., 1989, 1990; Markowitz et a/., 1988;Packer et a/., 1990; Radjaipour et al., 1986) or a relation to sleep stage cycles (Kripke e t a / . , 1978) of PTH in serum. In most of these reports the circadian changes of parathyroid hormone during the daytime were not observed as the assays employed were of insufficient sensitivity. Robinson et al. (1982) found no pulsatility in healthy young men and patients with primary hyperparathyroidism, while Fox et al. (1981) had detected pulsatile secretion pattern of PTH in dogs. However, these concentrations were determined with assays for parathyroid hormone fragments. The assays employed were often not sensitive enough to determine PTH correctly in the normal range; in addition, pulsatility may be overlooked due to accumulation of inactive fragments which have a half-life of about 45 minutes in healthy humans in comparison to 5 minutes for intact PTH (Blind et al., 1988). Fairly large pulses of intact PTH were found with a sensitive cytochemical bioassay (Dixit et al., 1987): concentrations of 1.3 0.6 ng/l at the nadir (mean f SD) and peak concentrations of 1 9 k 4 ng/l. Hypocalcaemia increased both the frequency and the amplitude of hormone secretion. 511
Clinical Endocrinology (1992) 37
512 K. Herfarth e t a / .
Partly conflicting reports have appeared on the pulsatility and rhythm of intact PTH in men (Harms er al., 1989; Kitamura et al., 1990). Kitamura et al. (1990) observed only small pulses with amplitudes of about 1.3 pmol/l and no narrow pulses. However, they found waves (‘pulse’ duration of 42.8 minute) and a circadian rhythm (higher values in the afternoon and during the night compared to morning values). Harms et al. (1989) found both narrow pulses (6 pulses/h with 1.6 pmol/l amplitude) and broad pulses (1 pulse/h). Therefore we studied PTH secretion in healthy adults and patients with primary hyperparathyroidism using frequent blood samples (every 1 minute) looking for narrow pulses of intact PTH secretion. In addition, we investigated ionized calcium, total calcium and phosphate as causes for PTH variation and circadian rhythm during daytime.
Study subjects, materials and methods Study subjects
Twelve healthy men (mean age 26.4 years, range 21-45) volunteered for the daylight variation study of calcium, phosphate and intact PTH. All were in good physical condition and no abnormalities in serum calcium, phosphate, alkaline phosphatase, total protein, urea and creatinine were observed. Six healthy subjects (three males, three females; mean age 27.7 years, range 21-56) volunteered for the pulsatility study. All were in good physical condition and serum calcium, phosphate, alkaline phosphatase, total protein, urea and creatinine were in the normal range. Twenty-one patients (five males, 16females; mean age 58-8 years, range 25-84) with surgically confirmed hyperparathyroidism (pathology: 17 adenomas, one double adenoma and three cases of hyperplasia) were also studied for the pulsatility pattern of intact PTH. The volunteers and the patients were informed about the risks and the aim of the following procedure in agreement with the ethical committee and gave their consent before blood was taken. An indwelling cathetcr was placed in a forearm vein at 0800 h in the 12 volunteers of the circadian rhythm study. Blood samples were taken from 0900 h to noon every 30 minutes and from noon to 1700 h hourly. Each time we sampled about 7 ml EDTA-plasma and closed the catheter with a stylet. The volunteers had been on a low-calcium diet (no milk products) for two days before sampling; on the sampling day they had a low-calcium breakfast at 0730 h and a low-calcium lunch after the 1300 h collection. For the pulsatility study the blood samples were taken in
the later afternoon. An indwelling catheter was placed in the antecubital vein and blood was taken every minute for one hour in the healthy subjects and for 30 minutes in the patients. In order to avoid thrombosis of the indwelling catheter we sampled 4 ml blood slowly into an EDTA-tube so blood stasis lasted only for about 30 seconds. In the circadian rhythm study, the blood samples were centrifuged within 4 hours. In the pulsatility study, blood samples were centrifuged after drawing was finished. The samples were stored frozen at - 30°C until assayed. For the ionized calcium series, an additional 2 ml blood were taken in a syringe, and determination was done immediately. lnta ct P TH
Intact PTH concentrations were determined in duplicate using a two-site immunoradiometric assay. We used the antihPTH (1-34) antibody ‘Rm’ raised in a rabbit (reported in detail by Bohler er al. 1989) and coated it onto a new polystyrene immunoradiometric assay tube (Sarstedt Co., Niimbrecht, FRG, order no. 73.1022) with enlarged surface (Klemm et al., 1990) using a method described by Wood (1989). We received a monoclonal anti-hPTH (44-68) antibody from Medgenix, Brussels, Belgium, which has previously been described by Bouillon e f al. (1990). This second antibody was labelled with 1251to a specific activity of 15Ci/g. 200 p1 EDTA-plasma were pipetted into a coated tube and incubated at 4°C overnight. The next day the sample was aspirated and the tubes were rinsed three times with 0.9% NaC1. Then 200 p1 of 1251-anti-hPTH(44-68) were added to each tube and incubated again at 4°C. After 4 hours the contents were aspirated and the tubes washed four times with 0.9?4 NaCI. The tubes were centrifuged upside down for 1 minute at 1000 r.p.m. to empty them completely. The tubes were then counted in a gamma-counter for 4 minutes (using the multi-crystal counter LB 2104 from Berthold, Wildbad, FRG). Assay characteristics The sensitivity of this assay was 0.13 pmol/l(2 SD above the zero point based on 14 samples of the zero standard); this is equal to 1.2 pg/l. The precision profile of the assay was determined by measuring each standard 12 times. The coefficient of variation ranged between 3.4% (125 pmol/l) and 7.6% (250 pmol/ 1) of the measured concentration (0, 1,4,8, 16,32,64,125 and 250 pmol/l). The interassay coefficient of variation was determined with control sera at two different concentrations in 14 different assays. A coefficient of 17.2% was found at 2.2 pmol/l (SDkO.38) (low normal range) and of 4.0% at 120-2 pmol/l (SDk4.78).
Clinical Endocrinology (1992) 37
Rhythms of PTH secretion
Table I Linearity of the IRMA for intact PTH
Patient 1
Dilution (factor) 1:l 1:2 1.4 1:s 1.10
Conc. found (pmol/l) 20.1 11.2 5.0 2.4 1.5
Hyperparathyroid subjects Thirty-two patients with surgically confirmed (n = 24) or diagnosed (n = 8) primary hyperparathyroidism showed values ranging from 5.4 to 341.3 pmol/l (mean 32.6, median 12.7).
Patient 2
Percentage of expected value
Conc. found (Pm 01/I )
Percentage of expected value
-
358.7 174.7 80.3 35.3 29.6
97.4 89.6 78.7 82.5
111.4 99.5 95.5 746
Added (pmol/l) -
2
4 16
32
Conc. found (pmolil) 7.4 10.8 11.9 22.6 35.0
Sample 2
Recovery
(YO)
Added (pmol/l)
-
-
114.9 104.4 96.6 88.8
0.5 2 8 32
Conc. found @mol/l) 7.8 7.3 8.4 13.5 34.2
Hypoparathyroid subjects Intact PTH concentration was below the detection limit in three hypoparathyroid patients.
-
Table 2 Recovery of added samples in the IRMA for intact PHT
Sample 1
513
Recovery
Ionized calcium
The ionized calcium and pH values were determined by ion selective electrode (Ionized Calcium Analyzer ICA 2, Radiometer Copenhagen) directly after the blood sample had been drawn. The measured ionized calcium value was automatically adjusted to a pH of 7.4. Clinical chemistry analyses (serum phosphate, alkaline phosphatase, total protein, urea and creatinine) were determined with a Hitachi-737 analyser. Serum calcium was determined by atomic emission spectrophotometry corrected for sodium interferences (CAFM, Eppendorf Geratebau, Hamburg, FRG).
(%) -
88.0 85.7 85.4 85.9
Linearity Two samples from patients with primary hyperparathyroidism were diluted with hypoparathyroid serum. The results are shown in Table 1. Recovery This was tested by adding known amounts of intact PTH to two different serum samples drawn during neck vein catheterization of a patient with primary hyperparathyroidism. The results are shown in Table 2. fntevference Several synthetic hormone fragments did not cross-react in the physiological range. We tested interference with several synthetic hPTH-fragments obtained from Bachem Co., Bubendorf, Switzerland. HPTH (1-34) did not interfere up to 250 pmol/l, hPTH (1-44) up to 125 pmol/l, hPTH (28-48) up to 500 pmol/l and hPTH (44-68), hPTH (39-84) and (hPTH (53-84) did not interfere up to concentrations of 5000 pmol/l. Normal range We measured intact PTH concentration in 38 ostensibly healthy blood donors in the fasting state. Intact PTH was detectable in all samples. The concentrations ranged from 0.8 to 5.3 pmol/l (mean 2.77, median 2.7, SD 1.18).
Statistics
Student’s t-test for paired samples was employed for statistical analysis of the daylong concentration shifts. The Cluster analysis program of Veldhuis and Johnson (1986) was used to search for all significant pulses in the pulsatility study. The occurrence of a peak is defined as a significant increase followed by a significant decrease and a nadir is defined as a decrease followed by an increase. Employing a 2 x 2 cluster configuration for test peak and test nadir we took a t-statistic of 4.1 for detection of significant increases and decreases, so the false positive rate was reduced to less than 1 % on signal-free noise, generated by assaying 70 replicates of a single serum pool. Results
The mean concentration of intact PTH of all 507 observations in the healthy subjects was 2.75f 1.12 pmol/l (mean fSD) (median 2.3 pmol/l). We observed a significant circadian rhythm of intact PTH in healthy men during the daytime. There were low concentrations in the late morning at 0930 h (2.0k0.94 pmol/l, range 1.15-4.15) with a continuous increase of intact PTH during the next hours. The highest values were found in the afternoon at 1400 h (3.6f 1.33 pmol/l, range 1.55-5.1). This increase of 1.6 f0.99 pmol/l (range 0.4-3.15) was statistically highly significant ( P i 0.0005). Thereafter, only a slight but rather insignificant decrease of intact PTH concentrations occurred until 1700 h.
K . Herfarfh e t a / .
514
Clinical Endocrinology (1992) 37
The ionized and protein-adjusted total calcium decreased in the afternoon; these changes were small but statistically significant. The mean total calcium concentration (proteinadjusted) was 2.46 0.06 mmol/l at 0930 h and 2.42 & 0.06 mmol/l at 1400 h (decline: 0.038+0.027 mmol/l, range (- 0.071)-( + 0.02); P < 0.0005). Ionized calcium concentrations (adjusted to pH 7.4) also showed a small but significant decline of 0.014 0.018 mmol/l from 0930 h to 1500 h (range (-0.04)-( + O.O2)mmol/l; P < 0.02). Ionized and total calcium revealed an inverse relationship to intact PTH (Fig. l). Phosphate showed no changes in the morning and a significant increase in concentration of 0.28 +_ 0.23 mmol/l (range (-0. I3)-( + 0.57) mmol/l; P < 0.002) in the afternoon (1.008~0.178mmol/l at 1000 h a n d 1.288kO.169 mmol/l at 1500 h). Serum phosphate concentration closely paralleled that of intact PTH (Fig. 1). All four measured parameters correlated well. Table 3 shows the correlation coefficients of all mean data (Table 3).
*
*
-
09y
oaoo
1 . 5 0900 1000 1100 1200 1300 1400 1500 1600 1700 Time ( h I
.,
Fig. 1 0 , Mean ionized calcium concentration (adjusted to pH
mean phosphate concentration; A, mean total calcium concentration (protein-adjusted); 0, mean iPTH concentration during the sampling time. 7.4);
Table 3 Correlation coefficients of the mean of PTH, total calcium (protein adjusted), ionized calcium (adjusted to pH 7.4) and phosphate on the different timepoints
12
PTH
PTH tCa (adj.) iCa (adj.) Phosphate
1 -0.872*++ -0617*
It=
042***
tCa (adj.)
iCa (adj.)
phosphate
1
0.604* -0467***
* Pi0.05;** P
Circadian rhythm and pulsatility of parathyroid hormone secretion in man Klaus Herfarth', Heinrich Schmidt-Gaykt, Stefanie GraV and Andreas Maiert *Department of Surgery, Div.2.I, University of Heidelberg and tEndocrine Laboratory, Im Breitspiel 15, 6900 Heidelberg, Germany (Received 8 April 1992; returned for revision 21 May 1992; finally revised 19 June 7992; accepted 30 July 1992)
Summary OBJECTIVE We wished to investigate the circadian rhythm and pulsatility of parathyroid hormone (PTH) secretion in man, as conflicting results have been published. DESIGN and PATIENTS To investigate the circadian rhythm during daytime, we sampled (a) peripheral blood at hourly intervals in 12 healthy young men from 0900 h until 1700 h. For observation of pulsatility, we sampled (b) peripheral blood at 1-minute intervals for 1 hour in three healthy men and three healthy women (mean 27.7 years, range 21-56 years) and (c) at 1-minute intervals for 30 minutes in 21 patients with surgically confirmed primary hyperparathyroidism (pHPT). MEASUREMENTS The serum levels of intact PTH were measured by two-site immunoradiometric assay and special care was taken to reduce intra-assay variability, especially at the normal PTH concentration. In series (a), ionized calcium, total calcium and phosphate were also determined. RESULTS A circadian rhythm during daytime was found for intact PTH in healthy men and women with a nadir at 0930 h and a peak in the afternoon. Ionized calcium and total calcium (protein-adjusted)decreased and phosphate increased in the afternoon. These changes were all statistically significant (P < 0.02). Pulsetility of PTH: Statistical cluster analysis of the data showed no pulsatility either in healthy persons or in patients with primary hyperparathyroidism. In two healthy women and one healthy man slight changes of longer duration were discovered, but no complete pulses. In five patients with primary hyperparathyroidism, larger differences between the highest and lowest concentrations of intact PTH were found, but no complete pulses. CONCLUSIONS Our data show a signlflcant clrcadian Correspondence: Professor Dr med. H.Schmidt-Gayk, Im Breitspiel 15, D-6900 Heidelberg, Germany.
rhythm during daytime of intact PTH and only minor changes from minute to minute. The alterations in PTHlevels occurred at longer time intervals in healthy persons. In some patients with primary hyperparathyroidism, decreases of PTH-levels were found. The circadian rhythm of PTH may be due to slight changes in calcium or phosphate concentration.
Circadian rhythm is known in many hormones and pulsatility of secretion is very well established for ACTH and gonadotrophic hormones (Gallagher et al., 1973; Perks et al., 1988; Veldhuis &Johnson, 1988; Veldhuis et al., 1987). The change in normal concentration seems to be important for physiological and pathophysiological effects. Disease-specific patterns of rhythm or pulsatility have been described in various diseases (Kazer et a/., 1987; Linkowski et al., 1987; Liu et al., 1987; Spratt et a/., 1987). The hormonal effect might be changed by different pulsatile pattern of secretion (Gross et al., 1987; Paolisso et al., 1988) and this has also been considered for parathyroid hormone (PTH) secretion (Hesch et al., 1988). For parathyroid hormone several authors have described a circadian variation (Jubiz et al., 1972; LoCascio et al., 1982; Logue et al., 1989, 1990; Markowitz et a/., 1988;Packer et a/., 1990; Radjaipour et al., 1986) or a relation to sleep stage cycles (Kripke e t a / . , 1978) of PTH in serum. In most of these reports the circadian changes of parathyroid hormone during the daytime were not observed as the assays employed were of insufficient sensitivity. Robinson et al. (1982) found no pulsatility in healthy young men and patients with primary hyperparathyroidism, while Fox et al. (1981) had detected pulsatile secretion pattern of PTH in dogs. However, these concentrations were determined with assays for parathyroid hormone fragments. The assays employed were often not sensitive enough to determine PTH correctly in the normal range; in addition, pulsatility may be overlooked due to accumulation of inactive fragments which have a half-life of about 45 minutes in healthy humans in comparison to 5 minutes for intact PTH (Blind et al., 1988). Fairly large pulses of intact PTH were found with a sensitive cytochemical bioassay (Dixit et al., 1987): concentrations of 1.3 0.6 ng/l at the nadir (mean f SD) and peak concentrations of 1 9 k 4 ng/l. Hypocalcaemia increased both the frequency and the amplitude of hormone secretion. 511
Clinical Endocrinology (1992) 37
512 K. Herfarth e t a / .
Partly conflicting reports have appeared on the pulsatility and rhythm of intact PTH in men (Harms er al., 1989; Kitamura et al., 1990). Kitamura et al. (1990) observed only small pulses with amplitudes of about 1.3 pmol/l and no narrow pulses. However, they found waves (‘pulse’ duration of 42.8 minute) and a circadian rhythm (higher values in the afternoon and during the night compared to morning values). Harms et al. (1989) found both narrow pulses (6 pulses/h with 1.6 pmol/l amplitude) and broad pulses (1 pulse/h). Therefore we studied PTH secretion in healthy adults and patients with primary hyperparathyroidism using frequent blood samples (every 1 minute) looking for narrow pulses of intact PTH secretion. In addition, we investigated ionized calcium, total calcium and phosphate as causes for PTH variation and circadian rhythm during daytime.
Study subjects, materials and methods Study subjects
Twelve healthy men (mean age 26.4 years, range 21-45) volunteered for the daylight variation study of calcium, phosphate and intact PTH. All were in good physical condition and no abnormalities in serum calcium, phosphate, alkaline phosphatase, total protein, urea and creatinine were observed. Six healthy subjects (three males, three females; mean age 27.7 years, range 21-56) volunteered for the pulsatility study. All were in good physical condition and serum calcium, phosphate, alkaline phosphatase, total protein, urea and creatinine were in the normal range. Twenty-one patients (five males, 16females; mean age 58-8 years, range 25-84) with surgically confirmed hyperparathyroidism (pathology: 17 adenomas, one double adenoma and three cases of hyperplasia) were also studied for the pulsatility pattern of intact PTH. The volunteers and the patients were informed about the risks and the aim of the following procedure in agreement with the ethical committee and gave their consent before blood was taken. An indwelling cathetcr was placed in a forearm vein at 0800 h in the 12 volunteers of the circadian rhythm study. Blood samples were taken from 0900 h to noon every 30 minutes and from noon to 1700 h hourly. Each time we sampled about 7 ml EDTA-plasma and closed the catheter with a stylet. The volunteers had been on a low-calcium diet (no milk products) for two days before sampling; on the sampling day they had a low-calcium breakfast at 0730 h and a low-calcium lunch after the 1300 h collection. For the pulsatility study the blood samples were taken in
the later afternoon. An indwelling catheter was placed in the antecubital vein and blood was taken every minute for one hour in the healthy subjects and for 30 minutes in the patients. In order to avoid thrombosis of the indwelling catheter we sampled 4 ml blood slowly into an EDTA-tube so blood stasis lasted only for about 30 seconds. In the circadian rhythm study, the blood samples were centrifuged within 4 hours. In the pulsatility study, blood samples were centrifuged after drawing was finished. The samples were stored frozen at - 30°C until assayed. For the ionized calcium series, an additional 2 ml blood were taken in a syringe, and determination was done immediately. lnta ct P TH
Intact PTH concentrations were determined in duplicate using a two-site immunoradiometric assay. We used the antihPTH (1-34) antibody ‘Rm’ raised in a rabbit (reported in detail by Bohler er al. 1989) and coated it onto a new polystyrene immunoradiometric assay tube (Sarstedt Co., Niimbrecht, FRG, order no. 73.1022) with enlarged surface (Klemm et al., 1990) using a method described by Wood (1989). We received a monoclonal anti-hPTH (44-68) antibody from Medgenix, Brussels, Belgium, which has previously been described by Bouillon e f al. (1990). This second antibody was labelled with 1251to a specific activity of 15Ci/g. 200 p1 EDTA-plasma were pipetted into a coated tube and incubated at 4°C overnight. The next day the sample was aspirated and the tubes were rinsed three times with 0.9% NaC1. Then 200 p1 of 1251-anti-hPTH(44-68) were added to each tube and incubated again at 4°C. After 4 hours the contents were aspirated and the tubes washed four times with 0.9?4 NaCI. The tubes were centrifuged upside down for 1 minute at 1000 r.p.m. to empty them completely. The tubes were then counted in a gamma-counter for 4 minutes (using the multi-crystal counter LB 2104 from Berthold, Wildbad, FRG). Assay characteristics The sensitivity of this assay was 0.13 pmol/l(2 SD above the zero point based on 14 samples of the zero standard); this is equal to 1.2 pg/l. The precision profile of the assay was determined by measuring each standard 12 times. The coefficient of variation ranged between 3.4% (125 pmol/l) and 7.6% (250 pmol/ 1) of the measured concentration (0, 1,4,8, 16,32,64,125 and 250 pmol/l). The interassay coefficient of variation was determined with control sera at two different concentrations in 14 different assays. A coefficient of 17.2% was found at 2.2 pmol/l (SDkO.38) (low normal range) and of 4.0% at 120-2 pmol/l (SDk4.78).
Clinical Endocrinology (1992) 37
Rhythms of PTH secretion
Table I Linearity of the IRMA for intact PTH
Patient 1
Dilution (factor) 1:l 1:2 1.4 1:s 1.10
Conc. found (pmol/l) 20.1 11.2 5.0 2.4 1.5
Hyperparathyroid subjects Thirty-two patients with surgically confirmed (n = 24) or diagnosed (n = 8) primary hyperparathyroidism showed values ranging from 5.4 to 341.3 pmol/l (mean 32.6, median 12.7).
Patient 2
Percentage of expected value
Conc. found (Pm 01/I )
Percentage of expected value
-
358.7 174.7 80.3 35.3 29.6
97.4 89.6 78.7 82.5
111.4 99.5 95.5 746
Added (pmol/l) -
2
4 16
32
Conc. found (pmolil) 7.4 10.8 11.9 22.6 35.0
Sample 2
Recovery
(YO)
Added (pmol/l)
-
-
114.9 104.4 96.6 88.8
0.5 2 8 32
Conc. found @mol/l) 7.8 7.3 8.4 13.5 34.2
Hypoparathyroid subjects Intact PTH concentration was below the detection limit in three hypoparathyroid patients.
-
Table 2 Recovery of added samples in the IRMA for intact PHT
Sample 1
513
Recovery
Ionized calcium
The ionized calcium and pH values were determined by ion selective electrode (Ionized Calcium Analyzer ICA 2, Radiometer Copenhagen) directly after the blood sample had been drawn. The measured ionized calcium value was automatically adjusted to a pH of 7.4. Clinical chemistry analyses (serum phosphate, alkaline phosphatase, total protein, urea and creatinine) were determined with a Hitachi-737 analyser. Serum calcium was determined by atomic emission spectrophotometry corrected for sodium interferences (CAFM, Eppendorf Geratebau, Hamburg, FRG).
(%) -
88.0 85.7 85.4 85.9
Linearity Two samples from patients with primary hyperparathyroidism were diluted with hypoparathyroid serum. The results are shown in Table 1. Recovery This was tested by adding known amounts of intact PTH to two different serum samples drawn during neck vein catheterization of a patient with primary hyperparathyroidism. The results are shown in Table 2. fntevference Several synthetic hormone fragments did not cross-react in the physiological range. We tested interference with several synthetic hPTH-fragments obtained from Bachem Co., Bubendorf, Switzerland. HPTH (1-34) did not interfere up to 250 pmol/l, hPTH (1-44) up to 125 pmol/l, hPTH (28-48) up to 500 pmol/l and hPTH (44-68), hPTH (39-84) and (hPTH (53-84) did not interfere up to concentrations of 5000 pmol/l. Normal range We measured intact PTH concentration in 38 ostensibly healthy blood donors in the fasting state. Intact PTH was detectable in all samples. The concentrations ranged from 0.8 to 5.3 pmol/l (mean 2.77, median 2.7, SD 1.18).
Statistics
Student’s t-test for paired samples was employed for statistical analysis of the daylong concentration shifts. The Cluster analysis program of Veldhuis and Johnson (1986) was used to search for all significant pulses in the pulsatility study. The occurrence of a peak is defined as a significant increase followed by a significant decrease and a nadir is defined as a decrease followed by an increase. Employing a 2 x 2 cluster configuration for test peak and test nadir we took a t-statistic of 4.1 for detection of significant increases and decreases, so the false positive rate was reduced to less than 1 % on signal-free noise, generated by assaying 70 replicates of a single serum pool. Results
The mean concentration of intact PTH of all 507 observations in the healthy subjects was 2.75f 1.12 pmol/l (mean fSD) (median 2.3 pmol/l). We observed a significant circadian rhythm of intact PTH in healthy men during the daytime. There were low concentrations in the late morning at 0930 h (2.0k0.94 pmol/l, range 1.15-4.15) with a continuous increase of intact PTH during the next hours. The highest values were found in the afternoon at 1400 h (3.6f 1.33 pmol/l, range 1.55-5.1). This increase of 1.6 f0.99 pmol/l (range 0.4-3.15) was statistically highly significant ( P i 0.0005). Thereafter, only a slight but rather insignificant decrease of intact PTH concentrations occurred until 1700 h.
K . Herfarfh e t a / .
514
Clinical Endocrinology (1992) 37
The ionized and protein-adjusted total calcium decreased in the afternoon; these changes were small but statistically significant. The mean total calcium concentration (proteinadjusted) was 2.46 0.06 mmol/l at 0930 h and 2.42 & 0.06 mmol/l at 1400 h (decline: 0.038+0.027 mmol/l, range (- 0.071)-( + 0.02); P < 0.0005). Ionized calcium concentrations (adjusted to pH 7.4) also showed a small but significant decline of 0.014 0.018 mmol/l from 0930 h to 1500 h (range (-0.04)-( + O.O2)mmol/l; P < 0.02). Ionized and total calcium revealed an inverse relationship to intact PTH (Fig. l). Phosphate showed no changes in the morning and a significant increase in concentration of 0.28 +_ 0.23 mmol/l (range (-0. I3)-( + 0.57) mmol/l; P < 0.002) in the afternoon (1.008~0.178mmol/l at 1000 h a n d 1.288kO.169 mmol/l at 1500 h). Serum phosphate concentration closely paralleled that of intact PTH (Fig. 1). All four measured parameters correlated well. Table 3 shows the correlation coefficients of all mean data (Table 3).
*
*
-
09y
oaoo
1 . 5 0900 1000 1100 1200 1300 1400 1500 1600 1700 Time ( h I
.,
Fig. 1 0 , Mean ionized calcium concentration (adjusted to pH
mean phosphate concentration; A, mean total calcium concentration (protein-adjusted); 0, mean iPTH concentration during the sampling time. 7.4);
Table 3 Correlation coefficients of the mean of PTH, total calcium (protein adjusted), ionized calcium (adjusted to pH 7.4) and phosphate on the different timepoints
12
PTH
PTH tCa (adj.) iCa (adj.) Phosphate
1 -0.872*++ -0617*
It=
042***
tCa (adj.)
iCa (adj.)
phosphate
1
0.604* -0467***
* Pi0.05;** P
