Nephrol Dial Transplant (1996) 11 [Suppl 3]: 22-28
Nephrology
Dialysis
Transplantation
A deficit of calcitriol synthesis may not be the initial factor in the
pathogenesis of secondary hyperparathyroidism
I. Martinez, R. Saracho, J. Montenegro and F. Llach
Servicio Nefrologia, Hospital de Galdakao, Vizcaya, Spain, and Newark Beth Israel Medical Center, Nephrology Division,
Newark, New Jersey, USA
Abstract. Secondary hyperparathyroidism (HPT)
develops early in chronic renal failure (CRF) at a time
when plasma calcitriol levels are normal. At this time,
PTH are higher than normal controls and serum
phosphorous levels are lower. A decrement in total
serum Ca is noted, after an oral phosphate load, only
in patients with ERF. These data suggest that factors,
other than a decrease in calcitriol synthesis, may be
involved in the pathogenesis of HPT. A hypothesis is
forwarded suggesting that an alteration in the newly
cloned calcium sensor receptor may be the earliest
abnormality in the HPT, preceding a decrease in
plasma calcitriol levels.
Key words: calcitrol; calcium sensor receptor; early
renal failure; hyperparathyroidism; vitamin D receptor
Introduction
calcium and phosphorus may not reflect actual concentrations of these ions throughout the day. Although
serum calcium is usually normal in ERF, it may not
be sufficient to inhibit PTH secretion; since it is known
that in uraemia a greater serum calcium is needed to
inhibit PTH [12,13]. Finally, a major role has been
assigned to a deficit of calcitriol as the initial step in
the triggering of HPT. However, few data are available
in this regard.
The present study evaluates in a large number of
patients with CRF the following questions. (1) When
in the course of CRF do abnormalities in calcium,
phosphorus and calcitriol occur and when does HPT
develop? (2) What is the earliest divalent ion abnormality noted in these patients? (3) Since fasting values
may not reflect actual changes in divalent ions, what
is the effect of an acute phosphate load on divalent
ions?
Patients and methods
Secondary hyperparathyroidism (HPT) develops early
in chronic renal failure (CRF). Both hypocalcaemia
and hyperphosphataemia are early factors considered
in the pathogenesis of HPT [1,2]. Most of the published
observations in patients and in experimental animals
have been done in advanced CRF [3-6]. The few
available data in patients with early renal failure
(ERF) have shown normal serum calcium and phosphorus. In fact, normophosphataemia or mild hypophosphataemia have been consistently observed [7-9].
More recently, a deficit of calcitriol synthesis has been
suggested as the major factor in HPT [10]. Such a
deficit may lead to a decrease in parathyroid hormone
(PTH) inhibition, gut absorption of calcium and
worsening of the calcaemic response to PTH. It is
likely that changes in serum calcium, phosphorus and
calcitriol are all important factors in HPT. However,
it is not clear how these factors may interact in ERF.
Thus, the seemingly normal blood concentrations of
This study included 165 patients with various degrees of
CRF. Of these, 67 were females and 98 were males. Their
mean age was 55 years, ranging from 17 to 82 years. Patients
excluded from this study included those taking agents such
as diuretics, phosphate binders, calcium supplements or
vitamin D analogues. Also, patients with a serum albumin
less than 3.5 g were excluded. The primary cause of CRF in
these patients was as follows: vascular aetiology and/or
hypertension was present in 16%; glomerular disease and/or
diabetic nephropathy in 40%; interstitial disease and polycystic kidney disease in 28%. The primary cause of CRF
could not be established in 16% of the patients.
Patients were distributed into 10 groups according to
creatinine clearance (CCr) (Table 1). Thus, group 1 included
patients with CCr between 10 and 19ml/min, group 2
20-29 ml/min and so on. Finally, group 10 included patients
with CCr greater than 100 ml/min.
Twenty-eight patients with CCr> 100 ml/min were noted
to have PTH values in the upper limit of normal; they were
compared with the 14 control subjects with normal renal
function and comparable CCr. These patients had a mean
age of 42, ranging from 21 to 67 years; 18 were males and
Correspondence and offprint requests to: F. Llach, Newark Beth 10 females. Their various aetiologies of the renal disease
Israel Medical Center, Nephrology Division, 201 Lyons Avenue, included primary glomerulonephritis (proven by renal
biopsy) in 16, renovascular hypertension (proven by arterioNewark, NJ 07112, USA.
© 1996 European Dialysis and Transplant Association-European Renal Association
Calcitriol and secondary hyperparathyroidism
23
Table 1. Mean biochemical parameters of patients, grouped according to creatinine clearances
Group
No. of cases
CCr (ml/min)
PTH (pg/ml)
Ctrl (pg/ml)
ICa (mg/dl)
P
(mg/dl)
1
2
3
4
5
6
7
8
9
10
Control .
9
9
11
12
14
30
17
13
14
10-19
10-29
30-39
40-49
50-59
60-69
70-79
80-89
90-99
>100
>100
164 + 142**
119 + 60
115 + 89**
69 + 27**
52 + 20
57 + 22
48 ±26
34 ±20
38+21
42±13*
28 + 15
17±2.6
5.13 + 0.5
4.54 + 0.3
4.67 ±0.3
4.67 ±0.2
4.68 ±0.2
4.69 + 0.3
4.79 + 0.2
4.75+0.3
4.72+0.3
4.54±0.3
4.71 ±0.2
4.23 ±0.7
3.83±0.5
3.65 ±0.7
3.22 ±0.4
2.89 + 0.4
2.91 ±0.5
3.07 ±0.4
3.00 + 0.6
3.11 ±0.5
3.05 ±0.6
3.08 ±0.3
•
28
19
18 + 7.5
21±5
20 + 8
22±9
23±9
22+12
27±13
32±13t
31±9t
24±9
*.P<0.013 compared with controls; **P<0.05 compared with groups 5-10 and controls; ti'<0.05 compared with groups 1-7.
graphy) in three, chronic pyelonephritis in three, polycystic
kidney disease in two, incipient diabetes nephropathy (microalbuminuria) in two and unilateral nephrectomy in two
patients. The control subjects were all members of our
nephrology division; they included eight females and six
males without history of renal disease. Mean age was 39,
ranging from 26 to 55 years. They were normotensive and
had normal physical exam and urine sediment.
An acute phosphate load was performed in five control
subjects and five patients with ERF. The five controls
included four males and one female, mean age 31, ranging
from 27 to 35 years. These subjects were all members of our
nephrology section; they did not have a history of renal
disease or hypertension; biochemical parameters and urinary
sediment were normal. None of the normal subjects was
taking therapeutic agents. The ERF patients included four
males and one female, mean age 55 years, ranging from 41
to 69. Mean CCr was 73+ 12 ml/min, ranging from 70 to
80 ml/min. Primary renal disease in this group was IgA
nephropathy in two, chronic pyelonephritis in two and
polycystic kidney disease in one patient.
Procedure
Both controls and all patients had blood drawn in fasting
conditions for determination of total and ionized calcium,
phosphorus, creatinine, calcitriol and PTH. Twenty-four
hour urine was collected for calcium, phosphorus and creatinine. Glomerular filtration rate was evaluated using standard
CCr corrected for body surface of 1.73 m2. Calcium, phosphorus and creatinine were determined with standard autoanalyser; calcitriol was determined by radioimmunoassay—
normal values in our laboratory range between 18 and
78 pg/ml. PTH was determined using a radioimmunometric
assay (IRMA) which detects the intact PTH molecule.
Normal values in our laboratories are between 15 and
65 pg/ml.
The acute phosphate load was performed as follows: both
patients and controls had a phosphate load consisting of
regular breakfast containing 400 mg phosphorus, 250 mg of
calcium and 350 calories. Immediately after, they had 1 g of
elemental phosphorus administered orally at 08.00 h. Prior
to breakfast, baseline blood and urinary samples were collected. At the beginning of the study, all subjects had a
catheter placed in the cephalic vein for repeated blood
sampling. To avoid clotting problems, 1 ml diluted heparin
was placed on the catheter. At sampling, the first 10 ml of
blood were always discharged to avoid possible contamination with the heparin solution. Prior to the phosphate load
and on hourly basis x 6, blood was sampled for determination of creatinine, total calcium and phosphorus. Calcitriol
and PTH were determined at baseline and at the end of the
6 h study. Urinary collections were made at 2 h intervals
during the study. Urinary determination included creatinine,
phosphorus and calcium.
Statistical analysis
For the purpose of comparing the various groups of patients,
the ANOVA test was used. To evaluate significant differences
between pair groups, the Duncan test was used. Whenever
repeated determinations were made in the same patient, the
ANOVA test was used [31]. Both bivariant and multivariant
analysis were used whenever it was necessary to assess the
effect of one variable or to avoid the effect of confounding
variables. The paired test was used to test significant differences between paired measurements of the same patients; the
P value, when used for multiple comparisons, was adjusted
using the Bonferroni method. Linear regression analysis was
used to test quantitative associations between two variables.
A value of P< 0.05 was considered significant. All the results
are expressed as the mean +SD, except when otherwise
indicated.
Results
The mean values of biological parameters of patients
distributed in 10 groups according to CCr are displayed
in Table 1. Individual values of calcium and phosphorus, as well as calcitriol and PTH, are shown in
Figure 1. As can be seen from Table 1, serum ionized
calcium was within normal range, while a decrease in
serum phosphorus (P<0.05) was noted in patients
with CCr between 50 and 59 ml/min; also, an increment in serum phosphorus was observed in groups
1-3 (P<0.05). A progressive decrease in calcitriol
values was observed as renal function deteriorated
(Figure 1C). When groups 9 and 10 (CCr>90 ml/min)
were compared with groups 1-7, a decrement in serum
calcitriol was observed (P<0.05); of interest is that
calcitriol levels from groups 8-10 (CCr > 80 ml/m) were
higher (P<0.05) than those from controls.
I. Martinez et al.
24
Controls
Patients
Controls
Patients
70
60
. D p <o.oi
*
ng/ml
50
40
i
30
I
o.
•*•
•
—
20
10
Controls
0
Patients
Controls
T
Patients
Fig. 1. Biological parameters of ERF patients. Individual (A) calcium and (B) phosphorus values of all patients are displayed. Mean
(C) calcitriol and (D) PTH values from groups according to creatinine clearances are shown; the mid-point of creatinine clearance for each
group is displayed in the horizontal axis.
B
A
Intact PTH, pg/ml
Calcitriol pg/mL
70-
p<0.001
60-
p<0.0001
50403020-
>- . • • . ^ H r r . v Y . . i
ii
5
i
:
i
i
i
i_
25 45 65 85 105 125 145 165
Creatinine Clearance, ml/min
. " ' • • '
1000
20
40 60 80 100 120 140 160 180
Creatinine clearance, ml/min
Fig. 2. Individual values of (A) PTH and (B) calcitriol from all patients with CRF, correlated with creatinine clearances.
Calcitriol and secondary hyperparathyroidism
A steady and progressive increment in PTH values
was noted as renal function worsened (Figure ID).
Values from groups 1-3 were greater than those from
groups 4-10 (P<0.05). Control subjects had lower
PTH values than groups 8-10. Overall, in the 165
patients there was a lineal and inverse correlation
between PTH and calcitriol values (P < 0.01). As shown
in Figure 2A, there was a hyperbolic correlation
between PTH and CCr (/»<0.001), and a positive
lineal correlation between calcitriol values and CCr
(P< 0.001) (Figure 2B). Similar correlations were
observed if the above mentioned parameters were
analysed in patients when grouped according to their
primary renal disease.
When 28 patients with CCr > 100 ml/min, with seemingly normal PTH, were compared with normal control
subjects, significant differences were noted in biochemical parameters (Figure 3). As it can be noted in
Figure 3D, these patients have a greater PTH
(41±14pg/ml) than control subjects (28±15pg/ml;
P<0.01); this occurred despite the presence of similar
calcitriol values in both groups (Figure 3B). Also,
mean serum phosphorus was less in these 28 patients
(2.9±0.5mg/dl) compared with normal subjects
(3.4 + 0.4 mg/dl; P < 0.008).
Biochemical parameters of normal subjects and five
A
25
patients with early renal failure (mean CCr 73 ml/min)
before and after an acute phosphate load are displayed
in Figure 4. A progressive linear decrease in serum
calcium after the phosphate load was noted in
patients (Figure 4A); it decreased from 9.24±0.18
to 8.89±0.28 mg/dl (P<0.001). There was a similar
linear increment in serum phosphorus in both normals
and patients (Figure 4B). Calcitriol values increased
significantly from 22±4 to 38 + 7 pg/ml (P<0.01) only
in patients (Figure 4C). At this stage of CRF, basal
values of calcitriol were less in patients (22+4 pg/ml)
than in normals (33 + 5 pg/ml; P<0.02), while basal
PTH was greater in patients (40 +12 pg/ml) than in
controls (19 + 8 pg/ml; P<0.01). The increment in
PTH after phosphate load was parallel and of similar
magnitude in patients (46 + 13 pg/ml) and nonnal subjects (29 + 7 pg/ml) (Figure 4D). Multivariant analysis
for repeated values showed lower PTH values in
normal compared with patients (P<0.01).
Discussion
The present study demonstrates that patients with
CRF have normal serum calcium and phosphorus at
various stages of renal insufficiency. Once CCr
B
Ionized Calcium mg/dL
Phosphorus mg/dL
6-
6-
5-
5-
4-
4-
3-
3-
2-
2-
1-
1-
00
0
20
40
60
80
20
100 120 140 160 180
40
60
80
100 120 140 160 180
Creatinine Clearance, ml/min
Creatinine Clearance, ml/min
D
Calcitriol pg/mL
p<.05
ntact PTH pg/mL
200 -]
£
j
P
—
1 ANOVA, Duncan
I
150100-
I
n
15 25 35 45 55 65 75 85 95>100
Creatinine Clearance, ml/min
15
25 35 45
55
65
75
85
Creatinine Clearance, ml/min
Fig. 3. Individual PTH and calcitriol values of 28 ERF patients compared with nonnal subjects.
95>100
26
I. Martinez el al.
-
10.0
E
5
Controls
I
"o
a 3
9.0
g
E
8.5
Patients
in
9.5
9
(0
6
0 12
3 4 5 6
2
0 12
(0
Time, hours
Time, hours
99
i
55
45
o>
;itri«
a
35
25
u
3 4 5 6
s
s.
s /
/
7
0
*
6
Time, hours
45
a
X
25
&
15
5
0
6
Time, hours
Fig. 4. Mean values of biological parameters of ERF patients and normal subjects during the acute phosphate load. Normal subjects are
represented by an interrupted line; ERF patients by a continuous line. *P<0.05, ANOVA test for repeated measurement of each variable.
decreases to less than 20 ml/min a change in serum
calcium may be detected. Likewise, serum phosphorus
was within normal limits and did not increase until
CCr was less than 20 ml/min. These observations are
in agreement with previous observations [7-9]. Of
interest is that patients with CCr between 55 and
65 ml/min had a relative hypophosphataemia. This
also is in agreement with previous observations [7-9].
Furthermore, the seemingly normal serum phosphorus
observed in patients with GFR > 100 ml/min was in
fact lower when compared with control subjects
(P< 0.008). The reason for the seemingly normal or
low serum phosphorus may reflect the HPT already
present in ERF, which in turn leads to a decrease in
TmP leading to phosphaturia. The serum calcitriol was
noted to decrease at a CCr < 80 ml/min, when compared with that from patients with CCr > 90 ml/min.
However, when the calcitriol values of the former
groups were compared with controls, there was no
significant difference. The apparent normal calcitriol
in patients with CCr > 30 ml/min may be the result of
the effect of PTH stimulating calcitriol synthesis; clinically, these observations emphasize the difficulties of
interpreting isolated serum calcitriol values in patients
with CRF. Surprisingly, calcitriol concentrations in
patients with a CCr > 90 ml/min were similar or greater
than those in control subjects. The reason for these
differences (as we will discuss later) may be the HPT.
Thus, mean calcitriol is seemingly normal in ERF
patients and it is only later as renal mass decreases
that it declines to less than normal.
The changes in PTH observed through the course
of CRF followed a different pattern to calcitriol. Thus,
in early CRF (CCr> 90 ml/min), at a time when calcitriol is seemingly normal, there was already a significant increase of PTH, which was unchanged until GFR
was < 40 ml/min. In fact, the mean PTH was not
different from groups 5 to 10. It was only when GFR
was less than 40 ml/min that a statistical increase
in PTH was noted. The fact that patients with
CCr > 100 ml/min have a greater PTH than our normal
controls, at a time when calcitriol concentrations were
not different from control subjects (Figure 3), suggests
that HPT may develop prior to the decrease in calcitriol. This observation militates against a deficit of
calcitriol as the initial factor in triggering HPT. Overall,
it appears that early HPT may stimulate calcitriol
synthesis resulting in seemingly normal calcitriol concentration which, in turn, may temporarily ameliorate
the severity of HPT. Once renal mass decrease enough
Calcitriol and secondary hyperparathyroidism
(GFR<30ml/min), calcitriol values decline below
normal, and a more severe HPT is noted.
It is interesting to speculate on what is the initiating
event in stimulating PTH secretions. One possibility
may be the presence of an abnormality of the parathyroid cell receptors, either in the vitamin D receptor
(VDR) or the recently cloned calcium sensor receptor
(CSR) [14]. Earlier, Korkor noted a decrease in VDR
density in patients with CRF [15]. In addition,
VDR density may be regulated not only by calcitriol
but also by dietary calcium intake. Vitamin D-deficient
chicks ingesting a high dietary calcium display an
upregulation of VDR density, while ingestion of a low
calcium diet has the opposite effect [16]. The normal
serum calcium and calcitriol seems to militate against
an early VDR abnormality in ERF. Another possibility
may be the presence of an abnormality of the CSR in
the parathyroids. As we know, the maintenance of a
stable internal environment within complex organisms
requires specialized cells that sense changes in extracellular concentrations of specific ions such as calcium.
Parathyroid cells possess a cell sensing mechanism that
is essential in the regulation of PTH secretion. Though
the molecular nature of this sensor has been known,
recently an extracellular calcium sensor receptor,
BoPCaRI, from bovine parathyroids, has been cloned
and characterized [14]. The bovine calcium receptor
has a deduced amino acid sequence with a topology
similar to the G-protein-coupled receptor superfamily
and a large extracellular region with nine glycoxylation
sites [14]. Thus, an early abnormality of this receptor
may be present in ERF and the seemingly normal
serum calcium may not be sufficient to suppress PTH
synthesis. Once CCR decreases to less than 80 ml/min,
calcitriol decreases and it becomes an important factor
in aggravating the HPT.
Finally, the acute phosphate load resulted in a
significant decrease in serum calcium only in patients
with ERF; this occurred despite a similar increment in
serum phosphorus during the acute phosphate load in
both groups. Though calcitriol and PTH increased
during the acute phosphate load, this was significant
only in patients with ERF. Thus, the mild hypocalcaemia noted in ERF patients occurred despite increased
calcitriol and PTH; this is consistent with the presence
of an intrinsic abnormality in calcium homeostasis
which may be independent of the prevailing levels of
calcitriol.
In summary, hyperparathyroidism may develop
early in CRF before a detectable decrease in CCr. At
this point, the PTH values are greater than those noted
in normal subjects with similar CCr. Since calcitriol
is within normal limits, the mechanism responsible
for HPT may be other than a calcitriol deficit.
Hypothetically, a uraemic generalized defect of receptors could contribute to HPT. Thus, the recently
described BoPCaRI may be involved. This hypothesis
is supported by observations of resistance to other
hormones, such as growth hormone and insulin, in
ERF [17-19]. Once CCr decreases to less than
27
80 ml/min, a decrease in calcitriol synthesis becomes
an additional pathogenic factor in the HPT. PTH is
maintained in the upper limit of normal as long as the
prevailing concentrations of calcitriol are sufficient to
modulate PTH secretion. As CCr decreases to less
than 40 ml/min and renal mass decreases, calcitriol
decreases leading to worsening of HPT. Furthermore,
the ingestion of a high phosphorus diet may unmask
a pre-existing abnormality in calcium homeostasis
resulting in hypocalcaemia and PTH stimulation, despite prevailing normal calcitriol values.
Acknowledgements. The authors thank Miss Sandra Caldwell for
providing secretarial assistance in the preparation of this manuscript.
References
1. Reiss E, Canterbury JM, Bilinsky RT. Measurement of serum
parathyroid hormone in renal insufficiency. Trans Assoc Am
Phys 1968; 81: 104-114
2. Malluche HH, Ritz E, Lange HP. Bone histology in incipient
and advanced renal failure. Kidney Int 1976; 9: 355-362
3. Bricker NS. On the pathogenesis of the uremic state. An
exposition of the "trade off" hypothesis. N Engl J Med 1972;
286: 1093
4. Slatopolsky E, Caglar S, Gradowska L, et al. On the prevention
of secondary hyperparathyroidism in experimental chronic renal
disease using 'proportional reduction' of dietary phosphorus
intake. Kidney Int 1972; 2: 147-151
5. Llach F, Massry SG, Singer FR et al. Skeletal resistance to
endogenous parathyroid hormone in patients with early renal
failure. A possible cause for secondary for hyperparathyroidism.
J Clin Endocrinol Metab 1975; 41: 339-345
6. Llach F, Massry SG. On the mechanism of secondary hyperparathyroidism in moderate renal insufficiency. J Clin Endocrinol
Metab 1985; 61: 601-606
7. Wilson I, Felsenfeld A, Drezner MK, Llach F. Altered divalent
ion metabolism in early renal failure: role of 1,25(OH)2D3.
Kidney Int 1985; 27: 565-573
8. Portale AA, Booth BE, Tsai HC et al. Reduced plasma concentration of 1,25 dihydroxyvitamin D in children with moderate
renal insufficiency. Kidney Int 1982; 21:. 627-632
9. Reichel H, Deibert B, Schmidt-Gayk et al. Calcium metabolism
in early chronic renal failure: implications for the pathogenesis
of hyperparathyroidism. Nephrol Dial Transplant 1991; 6:
162-169
10. Chertow BS, Baylink DJ, Wergrdal JE et al. Decrease in serum
immunoreactive parathyroid hormone in rats and in parathyroid
hormone secretion in vitro by 1-25 dihydroxycholecalciferol.
J Clin Invest 1975; 56: 667-678
11. Henderson RG, Russell RGG, Ledingham JGG et al. Effects of
1,25-dihydroxicholecalciferol on calcium absorption, muscle
weakness and bone disease in chronic renal failure. Lancet 1974;
1:379
12. Portale AA, Halloran BP, Morris RC Jr. Dietary intake of
phosphorus modulates the circadian rhythm in serum concentration of phosphorus: Implications for the renal production of
1-25 dihydroxy vitamin D. J Clin Invest 1987; 80: 1147-1154
13. Brown EM, Wilson RE, Eatsman RC et al. Abnormal regulation
of parathyroid hormone release by calcium in secondary hyperparathyroidism due to chronic renal failure. J Clin Endocrinol
Metab 1982; 54: 172-179
14. Brown EM, Gamba G, Riccardi D. Cloning and characterization
of an extracellular Ca-sensing receptor from bovine parathyroid.
Nature 1993; 366: 575
15. Korkor AB. Reduced binding of 3(H)l,25-dihydroxyvitamin
D3 in the parathyroid glands of patients with renal failure. N
Engl J Med 1987; 316: 1573-1577
28
I. Martinez et al.
like growth factor binding protein in uraemia. Pediatr Nephrol
16. Russel J, Bar A, Sherwood LM, Hurwitz S. Interaction between
1991; 5: 539-544
calcium and l,J5-dihydroxyvitamin D3 in the regulation of
preproparathyroid hormone and vitamin D receptor messenger 18. Chan W, Valerie KC, Chan JCM. Expression of insulin like
growth factor-1 in uremic rats; Growth hormone resistance and
ribinucleic acid in avian parathyroids. Endocrinology 1993; 132:
nutritional intake. Kidney Int 1993; 43: 790-795
2639-2644' .
17. Blum WF, Ranke MB, Kietzmann K. Growth hormone resist- 19. DeFronzo RA, Alvestrand A, Smith D, Hendler R. Insulin
ance and inhibition of somatomedin activity by excess insulinresistance in uremia. J Gin Invest 1981; 67: 563-568