Abstract
Introduction
Asymptomatic mild primary hyperparathyroidism is increasingly being identified during pregnancy. Recent studies have demonstrated inconsistent findings with regard to pregnancy complications and the need for surgical intervention during pregnancy.
Method
A retrospective audit of outcomes of pregnancies complicated by hypercalcaemia over a 15-year period was performed.
Results
Twenty-nine pregnancies to 26 women with hypercalcaemia were identified, corresponding to 37 cases per 100,000 deliveries. Hypercalcaemia was due to primary hyperparathyroidism in 90% of cases, with mean serum calcium of 2.89 mmol/l and mean ionised calcium 1.43 mmol/l. Four women underwent successful neck exploration during pregnancy. Pregnancy complications were limited to three cases of pre-eclampsia and one case of symptomatic neonatal hypoparathyroidism.
Conclusion
Close observation without surgical intervention would seem reasonable in women with mild hypercalcaemia during pregnancy.
Introduction
The principal cause of hypercalcaemia in pregnancy is primary hyperparathyroidism (PHPT). Up until 1999, only 145 cases of PHPT complicating pregnancy had been described. These earlier studies of women with severe symptomatic hypercalcaemia in pregnancy revealed maternal complications in 67%, fetal complications in 80%, and fetal/neonatal mortality in 30%. 1 The increasing use of multiple biochemical analysis (MBAa) has resulted in the identification of women with asymptomatic mild PHPT. A recent study screening woman aged 20–40 years found the incidence of PHPT to be 0.05%. 2 As many women are asymptomatic the true incidence in pregnancy is unknown. Other major causes of hypercalcaemia in pregnancy include familial hypocalciuric hypercalcaemia (FHH), milk-alkali syndrome, and parathyroid hormone (PTH)-related peptide-mediated hypercalcaemia due to malignancy or the pregnancy itself. 3 Three recent studies of pregnancy outcome in the setting of mild PHPT demonstrated inconsistent results.2,4,5 The need for surgical intervention when mild PHPT is identified during pregnancy is therefore unclear. This retrospective audit reviews the outcome of 29 pregnancies to 26 mothers with hypercalcaemia, demonstrating a relatively low rate of maternal and fetal complications, suggesting careful observation is safe in pregnancies complicated by mild maternal hypercalcaemia.
Methods
A retrospective audit was performed at the Mater Mother’s Hospital Brisbane, a quaternary obstetrics hospital. The hospital pathology database was interrogated to identify pregnancies of women with hypercalcaemia between 1 January 2000 and 31 December 2015. This approach would only identify pregnant women with hypercalcemia identified on routine biochemical testing, but would not identify those with hypercalcemia and an admission diagnosis of miscarriage.
Results
Twenty-nine pregnancies to 26 women with hypercalcaemia were identified during the 15-year time period, corresponding to 37 cases per 100,000 deliveries (Table 1). Mean maternal age was 30 years (range 21–42 years), and 20 were primigravida. Twenty-six pregnancies were complicated by PHPT. In seven pregnancies women were known to have hyperparathyroidism prior to conception but had declined neck exploration. Seven were diagnosed in first trimester as a result of a MBA being done with first visit bloods, nine cases at 26–28 weeks gestation as a result of a MBA being done at the time of screening for gestational diabetes mellitus, and three in third trimester as an incidental finding on MBA. None of the women with PHPT were symptomatic. The mean maternal ionised calcium (iCa) at diagnosis was 1.43 mmol/l (range 1.32–1.8; adult reference interval 1.13–1.3 mmol/l) and mean serum corrected calcium (Cacorr) was 2.89 mmol/l (range 2.62–3.43; non-pregnant adult reference interval 2.1–2.6 mmol/l). Mean PTH level in women with PHPT was 8.85 pmol/l (range 3.4–25; non-pregnant reference interval 1–7 pmol/l). Urine calcium:creatinine ratio was elevated above the non-pregnant reference range in all women with PHPT. iCa was checked in all women on a regular basis (usually monthly) during pregnancy and remained unchanged prior to delivery (or surgery). Four mothers with PHPT underwent successful parathyroid surgery during pregnancy (at 20, 20, 31, and 34 weeks of gestation) without maternal or fetal complications. Medical therapy for hypercalcaemia was not offered during pregnancy. A further seven women underwent surgery post-partum, the remainder either declining intervention or being lost to follow-up.
Pregnancy and neonate details.
k-gestation at diagnosis (0 preconception); FHH- familial hypocalciuric hypercalcaemia ; MAS- milk alkali syndrome; iCa Dx- maternal ionised calcium at diagnosis; iCa end- maternal ionised calcium at end of pregnancy (or prior surgery); Cacorr- peak maternal corrected calcium ; PTH-maternal parathyroid hormone level; u Ca:Cr- maternal urine calcium:creatinine ratio (non-pregnant reference 0.05-0.45mmol/mmol) ; Neonatal iCa - nadir neonatal ionised calcium; Bthwt- neonatal birthweight; Surg- gestation of neck exploration surgery; PP- postpartum.
One patient with PHPT had Multiple Endocrine Neoplasia type 1 (MEN1) identified during the pregnancy, following the diagnosis of antenatal hypercalcaemia and interrogation of her family history. Two mothers presented with symptomatic milk-alkali syndrome, with moderate hypercalcaemia (mean iCa 1.62 mmol/l; Cacorr 3.36 mmol/l), acute kidney injury, and alkalosis. One patient had genetically proven FHH.
All 29 pregnancies resulted in the delivery of a live neonate. Three pregnancies were complicated by pre-eclampsia at 30, 38, and 38 weeks of gestation. Maternal hypercalcaemia in the pregnancies complicated by pre-eclampsia was mild, with mean iCa 1.38 mmol/l (range 1.35–1.41) and mean Cacorr 2.77 mmol/l (range 2.73–2.81). With one exception all deliveries were at term gestation. One neonate developed symptomatic hypocalcaemia, presenting on the 15th day of life with seizures and tetany, and an iCa of 0.89 mmol/l (term neonatal ref range 1.08–1.4 mmol/l). The mother had PHPT with an iCa of 1.51 mmol/l and Cacorr of 3.17 mmol/l immediately prior to delivery at 39 weeks of gestation. Two neonates had transient asymptomatic hypocalcaemia which did not require treatment. There was one case of omphalocoele presumed to be unrelated to the maternal hypercalcaemia. No other maternal, fetal, or neonatal complications were observed.
Discussion
Three recent studies have described pregnancy outcomes in women with PHPT. Norman et al. 4 described 77 pregnancies in 32 women with PHPT who had been referred to a single practice limited to the treatment of PHPT. The mean serum calcium was 2.85 mmol/l. Fifteen of the women underwent parathyroid surgery between 13 and 23 weeks of gestation. There were no pregnancy losses in these 15 women. In the 62 pregnancies to women who did not have parathyroid surgery, the rate of pregnancy loss was 48%, occurring between 7 and 23 weeks of gestation. Three neonates born to conservatively treated mothers developed hypoparathyroidism requiring intravenous calcium treatment. Neonatal hypoparathyroidism was transient in two neonates and permanent in the other. Surprisingly these cases of neonatal hypocalcaemia manifested in the first three days of life. Neonatal hypoparathyroidism as a result of maternal PHPT usually manifests after the fifth day of life, though has been reported as late as the 60th day of life. Early neonatal hypocalcaemia typically occurs in the setting of prematurity, birth asphyxia, fetal growth restriction, or maternal diabetes mellitus.
Wright et al. 2 compared pregnancy outcomes in 124 pregnancies to 74 women with PHPT, to 431 pregnancies in 175 normocalcaemic pregnant women. Mean Cacorr in the group with PHPT was 2.67 mmol/l, lower than in the study by Norman et al. Six women with PHPT underwent parathyroid surgery during pregnancy. Overall there were no differences in outcome between the women with PHPT and the control group, and no difference in outcome between those who underwent surgery and those managed conservatively.
Abood and Vestergaard 5 performed a register-based cohort study of women with PHPT with age and gender-matched controls. A diagnosis of PHPT did not increase the rate of miscarriage, nor affect birthweight or Apgar score though there was a higher rate of deliveries by caesarean section.
A Swedish registry-based study (years 1973–1997) found that 13% of women treated surgically for PHPT more than two years prior to delivery developed pre-eclampsia, an adjusted odds ratio of 6.9 with a matched comparison group. 6 The association of pre-eclampsia and parathyroid adenoma surgically treated more than five years before delivery was also statistically significant. Thus, it is uncertain as to whether parathyroidectomy performed during pregnancy will reduce the risk of subsequently developing pre-eclampsia.
Investigation of PHPT in pregnancy
During pregnancy, there is a reduction in total serum calcium, approximately 8%, due to volume expansion and hypoalbuminaemia.7–9 Ionised calcium is unchanged and should therefore be measured. Increased calcium absorption results in physiological hypercalciuria, with a 46% increase in urinary calcium excretion in third trimester which may affect biochemical differentiation of FHH and mild PHPT. 10 Differentiating PHPT from FHH is critical to avoid unnecessary neck exploration and associated maternal and fetal surgical risks. Physiological hypercalciuria associated with pregnancy may make measurement of the calcium excretion fraction unreliable for differentiating between the two conditions. More than 100 mutations in the calcium-sensing receptor have been identified and only two-thirds of cases of FHH had a mutation detected. 11 Testing serum and urine calcium in relatives may be a more rapid and reliable way of differentiating between FHH and PHPT in pregnancy. 12
PHPT may occur as a part of MEN syndromes 1 and 2a, familial hyperparathyroidism–jaw tumour syndrome, or isolated familial hyperparathyroidism. The prevalence of MEN1 and MEN2 in the general population has been estimated to be 2–3 per 100,000 and 1 per 30,000, respectively. 13 Interrogation of the patient's family and past medical history is the most important factor in workup of these familial conditions. Screening for MEN1 in adults less than 40 years of age with PHPT in the absence of a family or personal history of MEN phenotype is not indicated. 14
Imaging to localise a parathyroid adenoma is only necessary when considering surgical intervention. In pregnancy, ultrasound (US) is the first-line investigation and has a sensitivity and specificity of 69 and 94%, respectively. 15 Computerised axial tomography (CT) and magnetic resonance imaging (MRI) are relatively insensitive to be used as solitary imaging modalities. Nuclear medicine techniques are usually avoided during pregnancy but may be a consideration in the setting of moderate–severe maternal hypercalcaemia persisting following unsuccessful neck exploration. Radiation dose to the fetus with 99mTc-sestamibi ranges from 11 mGy in early pregnancy to 4 mGy at term. 16 A prospective registry study of 122 women exposed to 99mTc scintigraphy in first trimester at doses less than 5 mGy showed no evidence of teratogenesis or adverse pregnancy outcomes. 17 Three subsequent case reports described the use of 99mTc-sestamibi during pregnancy to successfully locate mediastinal and thymic parathyroid adenomas after negative four gland surgical exploration.18,19 Radiation doses of less than 50 mGy have not been associated with fetal malformation or fetal loss, though the excess childhood cancer risk is 1.1–3.0 patients per 1000. 20
Recently, 18 Ffluorocholine and 11 Cmethionine positron emission tomography (PET) have been described as having high sensitivity and specificity in localising parathyroid adenomas undetected by other imaging modalities.21,22 PET-MRI is associated with a fetal radiation dose of less than 5 mGy and PET-CT has an estimated fetal dose of 2.8–11.8 mGy.23,24 To date, 19 PET-CT/MRI examinations have been performed on 18 women during pregnancy, almost all for evaluation of malignancy. In one case PET-CT detected a supraclavicular lymph node metastasis in a woman with a history of parathyroid carcinoma where US and MRI had failed to detect recurrent disease in pregnancy. 25 The effective radiation dose with four-dimensional CT is 10.4 mGy; however, the dose to the maternal thyroid is about 57 times higher than with sestamibi scans. 26
Management of PHPT in pregnancy
Where intervention is required, the primary management option for PHPT is surgery. It is recommended that neck exploration be performed during the second trimester, although successful surgery has been reported as late as 34 weeks of gestation.5,27 Where a parathyroid adenoma can be localised preoperatively, minimally invasive parathyroidectomy may be performed using a cervical plexus block, avoiding the need for general anaesthesia. 25 Options for medical therapy are limited in efficacy, with little experience in pregnancy. Bisphosphonates may be used to lower calcium in PHPT, though the hypocalcaemic effect is often limited to 10–14 days. Bisphosphonates cross the placenta and may persist in the bone matrix for years after administration, thus the fetus may be exposed to bisphosphonates released from the skeleton years after maternal exposure. Animal studies found that gestational exposure to bisphosphonates may be associated with reduced fetal bone growth, decreased fetal survival, prolonged labour, and decreased birth weight, albeit at far higher maternal doses than those used in humans. Other potential concerns with fetal transplacental exposure to bisphosphonate include the risk of hypocalcaemia. A review of 65 mother–infant pairs citing bisphosphonate exposure prior to conception or during pregnancy found no evidence for serious fetal or neonatal adverse effects. 28 Cinacalcet has been used to treat hypercalcaemia in six pregnancies with variable efficacy though no adverse fetal or neonatal outcomes were seen.29,30 Similarly no adverse effects were observed in animal studies where cinacalcet was used in pregnancy. 31 Cinacalcet carries a theoretical risk of contributing to neonatal hypocalcaemia due to suppression of fetal PTH secretion, though whether the suppression from cinacalcet would exceed that from maternal hypercalcaemia is not known. Cinacalcet also inhibits active placental calcium transport. Cinacalcet may increase the risk of maternal renal calculi due to an increase in urinary calcium excretion. Calcitonin does not cross the placenta and is safe for the mother and fetus though efficacy is limited and tachyphylaxis may develop after 48 h of use. High doses of calcitonin in animals may result in low neonatal birthweight. Historically intravenous magnesium and oral phosphate have been used to treat hypercalcaemia due to hyperparathyroidism in pregnancy.32,33 The use of oral phosphate is associated with an increased risk of soft tissue calcification and diarrhoea. The use of denosumab in pregnancy is contraindicated as it is associated with increased neonatal mortality, skeletal effects, and decreased growth and development in animal studies.34,35
Limitations
Limitations of this study are associated with its retrospective nature and small numbers of patients. There was no standardised protocol for maternal or fetal calcium monitoring, or maternal treatment. The means of case identification meant that we were unable to capture pregnancies which resulted in miscarriage.
Conclusion
In this small cohort maternal PHPT with mean Cacorr of 2.89 mmol/l was associated with a relatively low risk of maternal and fetal complications. The possibility of an increased risk of pre-eclampsia should be explored in larger studies. Overall our findings concur with those of Hirsch et al., that it is reasonable to carefully observe pregnancies complicated by mild maternal hypercalcaemia.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical approval
This study was approved by the Human Research and Ethics Committee, Mater Health, Brisbane. The requirement for patient consent was waived by the Human Research Ethics Committee.
Guarantor
AM
Contributorship
JG performed a literature review and co-wrote the manuscript, AM conceived the study, performed the audit of medical records, performed a literature review, and co-wrote the manuscript.
