The authors state that they have no conflicts of interest.
Autosomal dominant hypocalcemia (ADH) is an inherited form of hypoparathyroidism caused by activating mutations in the calcium-sensing receptor (CaR). Treatment with PTH(1–34) may be superior to conventional therapy but is contraindicated in children, and long-term effects on the skeleton are unknown. The patient is a 20-yr-old female with ADH treated with PTH continuously since 6 yr and 2 mo of age. A bone biopsy was obtained for histomorphometry and quantitative backscattered electron imaging (qBEI). Her data were compared with one age-, sex-, and length of hypoparathyroidism-matched control not on PTH and two sex-matched ADH controls before and after 1 yr of PTH. The patient's growth was normal. Hypercalciuria and hypermagnesuria persisted despite normal or subnormal serum calcium and magnesium levels. Nephrocalcinosis, without evidence of impaired renal function, developed by 19 yr of age. Cancellous bone volume was dramatically elevated in the patient and in ADH controls after 1 yr of PTH. BMD distribution (BMDD) by qBEI of the patient and ADH controls was strikingly shifted toward lower mineralization compared with the non-ADH control. Moreover, the ADH controls exhibited a further reduction in mineralization after 1 yr of PTH. These findings imply a role for CaR in bone matrix mineralization. There were no fractures or osteosarcoma. In conclusion, long-term PTH replacement in a child with ADH was not unsafe, increased bone mass without negatively impacting mineralization, and improved serum mineral control but did not prevent nephrocalcinosis. Additionally, this may be the first evidence of a role for CaR in human bone.
Hypoparathyroidism is a disease of altered mineral homeostasis. An inherited form, autosomal dominant hypocalcemia (ADH), is caused by activating mutations in the calcium-sensing receptor (CaR). The result is symptomatic hypocalcemia with tetany and seizures, hyperphosphatemia, and hypercalciuria that can lead to nephrocalcinosis and impaired renal function.(1) Whereas standard treatment has consisted of calcium and vitamin D, human PTH replacement may be superior.(2,3) Recombinant human PTH(1–34) (teriparatide) is contraindicated in children because of a potential elevated risk of osteosarcoma and carries a “black box” warning limiting treatment to 2 yr in adult patients.(4) Additionally, the long-term effects of exogenous PTH on bone are unknown. Herein we present the case of a 20-yr-old woman with ADH secondary to sporadic CaR mutation who was treated with PTH replacement continuously since 6 yr and 2 mo of age. To our knowledge, this patient's 13.5 yr of continuous PTH treatment, most of it as a child, represent the longest exposure to exogenous PTH in pediatric or adult populations.
The patient was born in Iceland by Cesarean section to a G5P3 mother. She was a dizygotic twin delivered at 36.5 wk of gestational age for concerns of growth restriction in the patient's twin sister. Apgars were 8 and 9 at 1 and 5 min, respectively, and the initial physical exam was unremarkable. Serum calcium was 2.2 mM (reference range, 2.04–2.4 mM) with an albumin of 31 g/liter (reference range, 37–47 g/liter). The pregnancy was complicated by hyperemesis gravidarum, gastric ulcer, and β-blocker therapy for tachycardia. The mother also reported taking triazolam (pregnancy category X) for sleep. Shortly after birth, the baby was noted to be a poor feeder, hypotonic, and lethargic and was admitted to the intensive care unit (ICU) for three apneic episodes complicated by seizures.
There were recurrent, worsening episodes of seizure, laryngospasm, and respiratory insufficiency until the age of 17 mo when hypocalcemia secondary to hypoparathyroidism was diagnosed. Other symptoms included thirst, anorexia, abdominal pain, muscle cramps, and painful urination. At diagnosis, serum calcium was 1.23 mM and serum phosphorus was 3.16 mM (reference range, 0.81–1.55 mM). There were no disorders of mineral metabolism in any first-degree relatives, including the patient's twin.
The patient was started on calcium and 1-α vitamin D3, and symptoms improved markedly. The child presented to the NIH at 6 yr and 2 mo of age for enrollment in a clinical trial of PTH replacement. On exam, she was noted to be 118.4 cm (70th percentile) and 20.2 kg (52nd percentile), with no dysmorphic features suggestive of DiGeorge syndrome. Enamel hypoplasia of the front teeth was present, but there was no evidence of candidiasis, cataracts, or other endocrinopathies. She was Tanner 1 with normal prepubertal genitalia and no focal neurologic findings.
Serum calcium was 2.07 mM, and serum phosphorus was elevated at 2.20 mM. PTH and 1,25(OH)2 vitamin D3 were undetectable. Hypomagnesemia was noted (serum Mg, 0.54 mM; reference range, 0.75–1.00 mM). Twenty-four-hour urine collections showed elevated calcium (0.35 mmol/kg/24 h; reference range, <0.1 mmol/kg/24 h) and magnesium (0.25 mmol/kg/24 h; reference range, <0.075 mmol/kg/24 h) excretion, and the creatinine clearance was markedly reduced (56.9 ml/min/1.73 m2). Loss of renal function because of prolonged hypercalciuria is common in patients with hypoparathyroidism and even more common in patients whose hypoparathyroidism is on the basis of CaR mutations, because these patients have a greater degree of calciuria caused by loss of the calcium-resorbing effect of the CaR in the distal tubule. CT scans showed calcifications of the basal ganglia but no evidence of nephrocalcinosis, however. Genomic analysis performed at a later date showed a previously described(5) heterozygous F806S mutation in the CaR. The mutation was not present in either parent.
The patient had continuous access to PTH for the entire 13.5 yr. Other medications included calcitriol, magnesium, and the intermittent use of thiazides. Each study was approved by the Institutional Review Board of the National Institute of Child Health and Human Development. Informed consent was obtained.
Preparation of PTH
From 1994 to 1998, synthetic PTH(1–34) was given as previously described.(2) From 1998 to 2003, synthetic PTH(1–34) was prepared by Swedish Orphan International (Stockholm, Sweden). After October 2003, the patient used commercially available PTH(1–34) as teriparatide (Eli Lilly).
From 1994 to 1998, and 2005 to the present, laboratory determinations were performed at the NIH Clinical Center, with the exception of osteocalcin. Laboratory determinations done before initiating care at NIH and from 1998 to 2005 were obtained from her medical record. Osteocalcin was measured at Corning-Nichols Institute Diagnostics (San Juan Capistrano, CA, USA).
BMD of the spine (L1–L4), total hip, whole body, and forearm (one-third radius) was performed by DXA on QDR2000 and QDR4500 instruments (Hologic). BMD measurements were compared with the Hologic reference database for children(6) and to the Bone Mineral Density Childhood Study reference database.(7) From measurement of BMD of an anthropomorphic spine phantom, the CV for determination with the QDR2000 instrument was <0.5% and for the QDR4500 instrument was <0.4% over 6-mo periods. Comparison of a single spine phantom scanned on both instruments and duplicate scans performed on 38 patients showed that the two instruments gave BMD determinations that were within 0.5%. A nuclear medicine physician (J.C.R.) reviewed all scans.
Control subjects and bone biopsies
After standard tetracycline labeling, transcortical iliac crest bone biopsies were obtained from the patient and three control subjects. In an effort to control for sex, age, and length of hypoparathyroidism, we included control 1, who was a 22-yr-old woman diagnosed with idiopathic (non-ADH) hypoparathyroidism at age 7 and treated with conventional therapy alone (calcium, calcitriol). To control for sex and PTH treatment, we included controls 2 and 3, who were women with ADH whose biopsies were taken before starting PTH (at ages 14 and 12, respectively) and after 1 yr on PTH replacement (at ages 15 and 13, respectively).
Bone specimens were embedded in methyl methacrylate and prepared following a standard protocol for histomorphometric analysis.(8) Quantitative analysis was performed by using a digitizing image-analysis system OsteoMeasure (OsteoMetrics, Atlanta, GA, USA). Cancellous and cortical bone structure was evaluated and expressed by the variables of cancellous bone volume (BV/TV), trabecular width (Tb.Wi), trabecular number (Tb.No), and trabecular separation (Tb.Sp). Cortical and dynamic variables are not shown in Table 2 because the patient's prescribed tetracycline label was inadvertently substituted with doxycycline, a poor fluorochrome. Bone remodeling activity was evaluated on the cancellous (Cn.) bone surface and expressed by the variables of osteoid perimeter (O.Pm) and eroded perimeter (E.Pm). All variables are expressed according to the recommendations of the ASBMR nomenclature committee.(9) Z-scores were derived by comparing samples to age-matched control samples from the literature.(10)
Quantitative backscattered electron imaging
Quantitative backscattered electron imaging (qBEI) was used to measure the bone mineralization density distribution (BMDD) of cancellous bone in the patient's biopsy and in biopsies from three control subjects. Full details have been published elsewhere.(11,12) The embedded undecalcified bone samples were trimmed, polished, and coated with carbon before being analyzed in a digital scanning electron microscope (DSM 962; Zeiss, Oberkochen, Germany). Four parameters characterizing the BMDD were evaluated: CaMean, the weighted mean Ca concentration of the bone area; CaPeak, the peak position of the histogram that indicates the most frequent calcium concentration of the bone area; CaWidth, the full width at half-maximum of the distribution, describing the variation in mineralization density; CaLow, the percentage of bone with low mineralization density (<17.68 weight percent calcium), which is normally the amount of bone area undergoing primary mineralization.
Skeletal growth and development
Growth did not differ significantly from her twin sister either before or after initiating treatment with PTH (data not shown), suggesting that neither hypoparathyroidism nor PTH treatment affected growth. Her current height, 167 cm, exceeds the expected midparental adult height (164.6 cm) and her twin sister's height (165 cm). Bone age was slightly delayed until 8 yr and 8 mo of age, but became concordant with chronological age (data not shown). Tooth eruption occurred normally, although there was severe enamel hypoplasia, consistent with hypoparathyroidism, requiring restoration and caps.
Dose of PTH
The patient was initially enrolled in a trial comparing an equal total PTH dose (0.7 μg/kg/d) given once daily or in two divided doses. Twice daily dosing resulted in less fluctuation in serum calcium and more normal urine calcium excretion (data not shown), consistent with a prior study.(13) The patient was rerandomized to PTH for a second trial lasting 3 yr and was able to access PTH from an outside source in Sweden thereafter. She continued twice daily dosing until the present, except for a 6-mo period in 2005 (age 17) when she took PTH three times per day as part of a research protocol. To avoid elevation in urine calcium levels, the PTH dose was titrated to maintain serum calcium at just below the normal range at which the patient was asymptomatic. Total daily PTH doses were typically 0.6 μg/kg/d (range, 0.4–1.7 μg/kg/d). Doses needed to be increased acutely for illness (e.g., upper respiratory tract infection, gastroenteritis). Because doses (corrected for body weight) did not change appreciably over time, there is no evidence of PTH resistance. Magnesium supplementation was required two to three times per day.
Serum mineral levels
There were almost no occurrences of hypercalcemia after starting PTH (Fig. 1A). Hyperphosphatemia improved with administration of PTH (Fig. 1C) but hypomagnesemia persisted (Fig. 1E). By self-report, symptoms of hypocalcemia improved on PTH.
Urinary mineral excretion
Twenty-four-hour urine calcium and magnesium excretion were chronically elevated (Figs. 1B and 1F), and urine phosphorus excretion remained mostly in the normal range (Fig. 1D), despite hyperphosphatemia (Fig. 1C). There was a trend toward lower calcium, phosphorus, and magnesium excretion as a function of time on PTH replacement. Renal function, as reflected by creatinine clearance, improved significantly over time (Fig. 1G), despite the finding of nephrocalcinosis on renal CT by age 19. Glomerular filtration rate (GFR) estimated by the Schwartz formula mirrors the trend seen in creatinine clearance.
Markers of bone formation
The patient's serum osteocalcin level was generally markedly elevated compared with an age- and sex-matched pediatric reference population (data not shown).(14) Alkaline phosphatase remained within the age-corrected range of normal (data not shown). No other bone formation or resorption markers were available.
BMD and BMC were neither abnormally elevated nor decreased throughout childhood, with the exception of slightly elevated BMD at the lumbar spine before starting PTH (Fig. 2; Table 1). BMD at the lumbar spine and total hip increased immediately after starting PTH. After 1 yr of PTH, BMD at all sites was flat or trended down until age 12, after which a sustained increase resulted in elevated BMD by age 18 at all sites except the distal radius.
Table Table 1.. Bone Densitometry
A transcortical iliac crest bone biopsy was performed at 18 yr and 11 mo of age, after >13.5 yr of PTH treatment. Cancellous bone volume was markedly elevated (Table 2). This reflects an increase in trabecular number (but not trabecular width) with a reciprocal decrease in trabecular separation. The same trends were seen in both ADH controls after 1 yr of PTH, relative to their baseline biopsies. The cancellous osteoid and eroded surfaces were reduced in all subjects. Cortical and dynamic measurements were unavailable. There was evidence of intratrabecular tunneling (Fig. 3).
The mineral distribution curve was markedly left-shifted toward lower mineralization density at baseline in both ADH controls (Fig. 4) compared with the non-ADH control, whose BMDD did not differ from a population of normal adults. In ADH controls, the amount of bone with low mineralization density (Calow), presumed to represent primary mineralization, was also increased at baseline. After 1 yr of PTH, the curve was further left-shifted with an increased Calow. The patient's bone mineralization density, after 13.5 yr of PTH, was similar to the two ADH controls at baseline. The Calow area under the curve was less than the two ADH controls but greater than the non-ADH control and the normal population.
A very subtle calcification was visible in the left basal ganglia at age 12 (6 yr on PTH) and progressed to include the right basal ganglia and bifrontal subcortical U-fibers (data not shown). The skull itself was diffusely thickened with patchy sclerotic and lytic lesions (Fig. 5). Nephrocalcinosis was not evident on initial evaluation (6 yr and 2 mo of age) but was extensive by age 19, with a small renal calculus but no hydronephrosis (data not shown).
Activating mutations in the CaR cause a left-shift in the PTH dose-response curve (increased ligand sensitivity) to both Ca2+ and Mg2+ and result in hypoparathyroidism, with renal calcium and magnesium wasting, and resultant hypocalcemia, hypomagnesemia, and hyperphosphatemia.(15) Conventional therapy of calcium and vitamin D3 aims to reduce symptoms of hypocalcemia but worsens the already elevated urine calcium excretion and may promote nephrocalcinosis. This case is the first to report the effects of long-term PTH replacement in a child with hypoparathyroidism secondary to activating CaR mutation.
PTH adequately controlled serum Ca2+ and reduced hypercalcemic episodes compared with conventional therapy of calcium and calcitriol. Whereas a prior study comparing PTH to calcium and calcitriol found no difference in serum Ca2+ between treatment groups,(3) previous work has also shown that patients with CaR mutations require higher doses of PTH to achieve a similar calcemic response compared with patients with postsurgical and/or idiopathic hypoparathyroidism.(13) This argues that the disease pathophysiology is not merely PTH deficiency, as has been suggested,(16) but that the mutated receptor itself also plays a role.
PTH treatment also improved urine calcium excretion. However, even on PTH replacement, urine Ca2+ and Mg2+ excretion remained elevated and phosphorus excretion failed to increase as expected. This suggests that the CaR is downstream of PTH and that PTH cannot normalize mineral excretion in patients with CaR mutations. This may explain why urine Ca2+ excretion is higher in ADH than in postsurgical hypoparathyroidism. Although the mechanism remains under study, it is known that PTH stimulates renal tubular epithelium to absorb Ca2+ by upregulating cAMP. Activating CaR mutations mimic elevated basolateral Ca2+, inhibiting PTH-stimulated cAMP increases in medullary and cortical thick ascending limb.(17) This suggests that CaR mutations can partially “undo” PTH action in the renal tubule. Moreover, elevated basolateral Ca2+ inhibits NaCl transport in the thick ascending limb, reducing countercurrent multiplication and urine concentrating ability and decreasing the positive potential in the lumen, the driving force for paracellular Ca2+ and Mg2+ reuptake.(18)
Hypomagnesemia and hypermagnesuria are common in ADH(15) and were observed here. The affinity of Mg2+ for CaR is reported to be too low to detect the physiologic variation in Mg2+ concentration(19); therefore, increased Mg2+ excretion may be secondary to faulty urine Ca2+ reabsorption by the above mechanism. This also implies that efforts to reduce urinary Ca2+ excretion may reduce Mg2+ excretion as well, but further study is needed.
We are the first to report the skeletal effects of prolonged exposure to exogenous PTH in a child. During treatment, there was a dramatic and prolonged increase in serum osteocalcin levels compared with an age- and sex-matched normal population. Osteocalcin, an abundant noncollagenous protein in bone matrix thought of as a marker for bone formation, is usually low in hypoparathyroidism, returning to normal with vitamin D3 treatment.(20) In both osteoporotic(21) and hypoparathyroid(3) patients, PTH treatment increases biomarkers of bone turnover, including osteocalcin. In osteoporotic patients, however, bone marker levels tend to return to baseline after ∼15 mo, despite ongoing PTH treatment, in contrast to the sustained elevation seen here. The significance of this observation is unknown. Interestingly, the serum alkaline phosphatase was not elevated above the age-adjusted range of normal (data not shown), a feature we have also observed in some, but not all, of our pediatric ADH patients treated with PTH.
The patient's BMC was normal throughout development. BMD, which is BMC divided by bone area, increased in the first 6 mo after starting PTH at the spine and hip (Fig. 2; Table 1), possibly representing PTH-stimulated deposition of new bone. After the robust increase, BMD was flat or trended down at all sites until age 12, when a sustained increase was evident at the lumbar spine and hip (Fig. 2; Table 1), sites rich in trabecular bone, but not at the distal radius, which contains relatively less trabecular bone.(22) This is consistent with the known disparate effect of PTH on cancellous and cortical bone.(23) Of note, elevated BMD is also consistent with the high bone mass seen in hypoparathyroidism.(24) Direct analysis by histomorphometry and qBEI was done to better understand the effect on bone.
The most striking histomorphometric finding was the dramatically increased cancellous bone volume (BV/TV) after 13.5 yr of PTH replacement (Table 2). A similar finding was seen in both ADH controls after only 1 yr of PTH. It is known that once daily exposure to PTH, at least in postmenopausal osteoporosis, results in new bone formation on cancellous, endocortical, and periosteal surfaces.(25) The robust response seen here far exceeds what is seen in osteoporotic adults, however, even after 36 mo of PTH.(26)
Table Table 2.. Bone Histomorphometry
The new bone formed in this patient is explained by a large increase in trabecular number (with a corresponding loss of trabecular separation) and not increased thickness of existing trabeculae (Tb.Wi), consistent with what is seen in osteoporotic patients treated with PTH.(27) It has been hypothesized that the mechanism involves intratrabecular tunneling, where osteoclasts tunnel through and split trabeculae followed by new bone formation.(28) This phenomenon has been observed in women treated with PTH(1–84) for 18 mo(27) and was seen in our patient (Fig. 3). Despite the increased bone evident on histomorphometry, and in contrast to other men(29) and women(30,31) treated with intermittent PTH, BMD fell or was flat after an initial increase (Fig. 2; Table 1). Given that DXA integrates both bone mass and mineral content and there was no loss of bone mass by histomorphometry (rather the opposite), lower BMD most likely represents relatively less mineralization.
Using qBEI before exposure to PTH, we found low average bone mineralization at baseline in both ADH control patients and an elevated amount of less mineralized bone that was not present in a non-ADH hypoparathyroid control (Fig. 4). After 1 yr of PTH treatment, there was a further left-shift (toward lower mineralization) seen in the ADH controls, presumably because of the formation of a significant amount of new bone of lower mineralization density. A shift toward lower mineralization with greater heterogeneity is seen in osteoporotic patients treated with intermittent PTH for 18–36 mo,(32) although to a lesser degree. Randomized studies in ovariectomized monkeys have shown that intermittent administration of PTH increased BMC by DXA at the spine, hip, and long bones(33) but that bone had lower mineralization and reduced mineral crystallization by Fourier transform infrared imaging.(34) After 13.5 yr of treatment with PTH, BMDD in the patient was not further left-shifted but resembled the ADH controls at baseline. In agreement, BMD by DXA also increased around the time of skeletal maturity. Notably, even after long-term PTH replacement, all BMDD parameters in the patient remained abnormal compared with control 1, a non-ADH age-, sex-, and length of hypoparathyroidism-matched control. Taken together, these data suggest that the underlying CaR mutation may contribute to abnormal mineralization and that the PTH-induced left shift may be overcome with time as more complete secondary mineralization ensues. These data may reflect a direct evidence of a role for CaR in human bone.
A degree of caution is suggested in the interpretation of these data, in that the reference data for qBEI to which the patient and controls were compared are derived from adults. A reference range for normal children does not yet exist. BMDD in normal children seems to be lower, although not nearly to the degree seen here(35) (P. Roschger, personal communication). Our results also suggest that caution is needed when interpreting bone densitometry results, because they may conflate robust changes in bone mass and mineral content. For example, both ADH controls saw only modest, if any, change in BMD after 1 or 2 yr of PTH (data not shown). Similarly, the patient's normal-appearing bone densitometry belies substantial changes in bone mass (increased by histomorphometry) and mineral content (decreased by qBEI), which tend to “cancel” when quantified by DXA. This implies that effects on bone volume and mineralization have to be considered independently, especially when DXA data are interpreted.(36) This casts doubt on densitometry findings from patients with related diseases. It has been argued that patients with familial hypocalciuric hypercalcemia (FHH), which is caused by inactivating mutations of the CaR who experience exposure to supraphysiologic PTH, show no deleterious effects on bone mass as measured by DXA.(37) We posit that it may not be appropriate to conclude that patients with FHH have “normal” bone without independent assessment of bone volume and mineral content. Further work is needed to determine whether mutations in the CaR impart intrinsic defects in bone mineralization, independent of the extrinsic extracellular mineral milieu.
We cannot rule out that altered serum mineral availability (hypocalcemia) negatively affects bone mineralization. Dvorak et al.(38) have shown that osteoblasts are sensitive to minute changes in extracellular Ca2+ concentration, independent of calcitropic hormones. However, adults with hypoparathyroidism develop high bone mass despite hypocalcemia.(23) Notably, the patient has never sustained a fracture.
Last, our data are also the first to suggest that long-term exposure to PTH in a “high-risk” individual may not be harmful. The carcinogenic potential of PTH was raised when osteosarcoma developed in rodents exposed to either teriparatide or full-length PTH(1–84) in a duration- and dose-dependent fashion.(39,40) PTH is contraindicated in children because open epiphyses and the pubertal growth spurt are thought to confer increased risk of osteosarcoma, although the real risk is unknown. After 13.5 yr of continuous PTH replacement, the patient has not developed osteosarcoma. The goal of therapy was to maintain eucalcemia and minimize hypercalciuria. Her current daily dose of PTH (30 μg in divided doses) is higher than the dose approved for osteoporosis treatment (20 μg daily), although the latter is 3- to 60-fold lower than the oncogenic doses in rats.(4) Moreover, given important differences in skeletal growth between rats and humans,(41–44) the rodent data may not apply.
In conclusion, we showed that long-term PTH(1–34) replacement in the treatment of inherited hypoparathyroidism secondary to activating CaR mutation was efficacious, and to date no adverse effects have been shown in this patient. PTH effectively reduced urinary calcium excretion but did not prevent nephrocalcinosis. Whereas there seems to be an inherent mineralization defect in patients with activating CaR mutations, our data suggest that long-term PTH replacement dramatically increases bone mass with little long-term impact on mineralization. Finally, although our analysis was limited to a small number of patients, these findings suggest a direct role for CaR in human bone.
We are indebted to the patient and her family for participation and to Marilyn Kelly for coordinating. This research was supported in part by the DIR, NIDCR and the DIR, NICHD, of the IRP, NIH, DHHS. T.A.T. was supported by the Clinical Research Training Program, a public-private partnership supported jointly by the NIH and Pfizer (through a grant to the Foundation for NIH from Pfizer). P.R. and K.K. were supported by AUVA (Austrian Social Insurance for Occupational Risk), WGKK (Social Health Insurance Vienna), and FWF (Austrian Science Fund) Project P19009-N20.