Venous thrombosis in a pseudohypoparathyroidism patient with a novel GNAS frameshift mutation and complete resolution of vascular calcifications with acetazolamide treatment

Abstract

Introduction

Pseudohypoparathyroidism type IA (PHP1A) is characterized by end-organ resistance to multiple hormones and Albright’s hereditary osteodystrophy (AHO). PHP1A is caused by inactivating mutations of the GNAS gene encoding the α-subunit of the stimulatory G protein (Gsα). In line with the underlying genetic defect, impaired inhibition of platelet aggregation has been demonstrated in some patients. However, no PHP1A case with thrombotic events has been described. Also, PHP1A cases typically have subcutaneous ossifications, but soft tissue calcifications are another common finding. Treatment options for those and other non-hormonal features of PHP1A are limited.

Case Presentation

A female patient presented with short stature, fatigue, and exercise-induced carpopedal spasms at age 11^7/12 years. Diagnosis of PHP1A was made based on hypocalcemia, hyperphosphatemia, elevated serum PTH, and AHO features, including short stature and brachydactyly. A novel frameshift variant was detected in the last exon of GNAS (c.1065_1068delGCGT, p.R356Tfs*47), showing complete loss of baseline and receptor-stimulated activity in transfected cells. The patient developed venous thrombosis and vascular and subcutaneous calcifications on both forearms after venous puncture on the right and extravasation of calcium gluconate during treatment on the left. The thrombosis and calcifications completely resolved following treatment with low molecular weight heparin and acetazolamide for 5 and 8 months, respectively.

Conclusions

This case represents the first PHP1A patient displaying thrombosis and the first successful use of acetazolamide for PHP1A-associated soft tissue calcifications, thus providing new insights into the treatment of non-endocrinological features in this disease.

Introduction

Pseudohypoparathyroidism (PHP) refers to the biochemical features of hypocalcaemia, hyperphosphatemia, and elevated PTH levels due to end-organ resistance to the actions of parathyroid hormone (PTH). Patients with PHP can present with Albright’s hereditary osteodystrophy (AHO), characterized by obesity, round face, short stature, brachydactyly, subcutaneous ossifications, and intellectual disability [1].

GNAS is a complex, imprinted gene giving rise to multiple products, including transcripts that encode the α-subunit of the stimulatory guanine nucleotide-binding protein (G protein) (Gsα) encoded by exons 1–13 [2, 3]. Genetic and epigenetic defects of GNAS lead to impaired Gsα function and cAMP deficiency and, thereby, underlie the pathogenesis of PHP type 1 (PHP1) [2]. Heterozygous inactivation mutations in the Gsα-coding exons of the maternal GNAS allele cause PHP1A/1C [2, 4], while alterations in the methylation status of GNAS lead to PHP1B [5, 6]. Consistent with Gsα being a ubiquitous signalling protein, PHP1 patients also present with resistance to other hormones, including TSH, growth hormone-releasing hormone, and gonadotropins [1, 2, 4, 7–10]. These additional hormonal defects are commonly found in PHP1A and PHP1C patients. Although the hormone resistance is primarily confined to PTH in PHP1B, some cases can also display mild TSH-resistance [2, 5–8].

PHP1A is characterized by a 50% reduction of Gsα activity in patient-derived erythrocytes [7, 9, 11]. PHP1C patients also carry mutations in coding GNAS exons and exhibit indistinguishable clinical and biochemical features from PHP1A patients. Mutations in PHP1C affect receptor-coupling but not adenylyl cyclase activity and are located near the C-terminal end of the Gsα molecule [7, 10, 12]. Accordingly, when using biochemical assays that evaluate basal rather than receptor-stimulated activity, erythrocyte Gsα activity appears normal in PHP1C patients [13]. Gsα activity is detected as normal in PHP1B patients [2, 7, 8].

The platelets express Gsα-coupled prostanoid receptors, including the IP receptor (PTGIR) [14], and Gsα activity can also be evaluated using patient-derived thrombocytes [15, 16]. The endothelium-derived prostaglandin I₂ (prostacyclin, stable mimetic iloprost), along with the stable prostanoid prostaglandin E₁ (PGE₁), bind to the IP receptors to trigger platelet cAMP production by stimulating adenylyl cyclase (AC)[17]. Elevation in cAMP activates broad spectrum c-AMP dependent protein kinase A (PKA)[18], which initiates key cellular events, including Ca²⁺ fluxes and integrin α_IIbβ₃ activation [19, 20], that are necessary for the thrombotic processes like platelet adhesion, shape change, cytoskeletal changes, secretion, aggregation and procoagulant activity [21, 22]. Therefore, platelet based Gsα activity test is based on the inhibition of platelet aggregation by cAMP, induced by agonists of the Gsα-coupled IP receptor. In PHP1A cases compared to the healthy controls, the agonist concentration required to achieve 50% inhibition of platelet aggregation is at least ten times higher, indicating severe Gsα hypofunction in platelets [15]. Based on these platelet assays, it is plausible that Gsα deficiency in platelets causes a tendency towards thrombosis. Strikingly, a case with PHP1A with severe Gsα deficiency has been described to be in a prothrombotic state based on ex vivo platelet function assays [23]. A recent study also demonstrated significant alterations in collagen-dependent thrombus formation ex vivo in several PHP1A patients [24].

Although biochemical assays have shown defects in the inhibition of platelet aggregation ex vivo in PHP1A patients, no patients with thrombosis have been hitherto reported. Here we describe a PHP1A case carrying a complete null mutation in GNAS. During treatment, the patient manifested a thrombotic event and subcutaneous and vascular calcifications, which were completely resolved with antithrombotic therapy and acetazolamide treatment. Acetazolamide, a carbonic anhydrase inhibitor, has been demonstrated to be effective in the treatment of soft tissue calcifications in tumoral calcinosis due to its metabolic acidosis-producing and phosphaturic effects [25–28]. Here we report its successful use, for the first time, in vascular and soft tissue calcification in PHP.

Case Description

An 11^7/12-year-old girl presented with short stature, fatigue, and effort-induced carpopedal spasms. She was born at term with a birth weight of 2700 g (−1.7 SD score) to non-consanguineous parents. Her past medical history was unremarkable. Her height was 127.6 cm (−3.37 SD score), weight was 30 kg (−1.81 SD score), and body mass index (BMI) was 18.1 kg/m² (−0.03 SD score). On physical examination, a round face, despite normal BMI, and short 3^rd to 5^th metatarsals were remarkable. Carpopedal spasm and positive Chvostek’s sign were noted. The patient had Tanner stage II breast and Tanner stage I pubic hair development and no detectable subcutaneous nodules. Laboratory tests showed hypocalcaemia [5.4 mg/dL (8.8–10.8 mg/dL)], hyperphosphatemia [7.4 mg/dL (3.3–5.4 mg/dL)], markedly elevated PTH levels [PTH: 460 ng/L (15–65 ng/L)], and normal thyroid hormones (Table 1). The growth hormone stimulation test with clonidine showed a peak level of 11.5 ng/mL(Cut-off: 10 ng/mL). Family history revealed short parents [father: 158.5 cm (−2.87 SD score), mother 140 cm (−3.93 SD score)] and similar phenotypic features with the patient, such as round face and brachydactyly, in the mother and maternal grandmother without any known history of hypocalcaemia. Nevertheless, laboratory evaluation of the parents as a part of the patient’s etiological evaluation revealed hypocalcaemia, hyperphosphatemia, and elevated PTH levels in the mother (Table 1). Although the patient had vitamin D deficiency, clinical and laboratory features, i.e., hyperphosphatemia and AHO features, together with a family history, are consistent with PHP1A.

Table 1.

Laboratory values before, during, and after acetazolamide treatment.

	Initial	2nd week in conventional therapy	1st week of acetazolamide therapy	2nd week of acetazolamide therapy	8th month of acetazolamide therapy	6 months after discontinuation of acetazolamide	1 year after discontinuation of acetazolamide	Mother	Father
Calcium (N:8.8–10.8 mg/dL)	5.4	9.2	9.5	9.8	10.1	9.9	9.2	7.6	9.4
Ionized Ca (N:4.6–5.2 mg/dL)	2.2	4.9	5	4.9	5.1	5.1	5
Phosphorus (N:3.3–5.4 mg/dL)	7.4	5	5	6	5.2	4.6	5.1	4.4	3.3
Mg (N:1.8–2.2 mg/dL)	1.8	1.9	2.0	2.2	1.8	2.0	2	2.0	2.2
Alkaline phosphatase (N: 122–400 U/L)	251	172	149	122	136	113	243		79
PTH (15–65 ng/L)	461	145	71	88.9	37.2	19.1	95.4	511	25
25-OH D vitamin (N:20–100 μg/dL)	3.86	-	-	62.5	16.3	25.3	9.36	21.1	6.7
TSH (N:0.34–5.6 mU/L)	2.17	8.54	-	4.93	3.54	4.9	5.02	3.77
sT4 (N:0.61–1.12 ng/dL)	0.67	0.75	-	0.72	0.81	1.09	0.7	0.89
pH (N:7.31–7.41)	7.38	7.39	7.32	7.24	7.28	7.37	7.38
Bicarbonate (N:24–28 mmol/L)	27	28	18.2	20.7	22.6	28.1	26.4
TRP	99.8%	99.5%	-	-	98%	98.9%
TMP/GFR (N:3.9–8.2 mg/100 ml)	9.4	7.35	-	-	7.25	7.28
Treatment Calcitriol (μg/day)	-	1.0	0.5	0.5	0.5	1.0	1.0
Calcium Carbonate (mg/kg/day)	-	53	-	-	36	36	-

Radiological examination displayed symmetrical shortening of all metacarpal bones with variable shortening of phalanges on hand radiographs (Figure 1A) and basal ganglia and subcortical white matter calcification on cranial computerized tomography imaging (Figure 1C). These laboratory and radiological findings were consistent with PHP1A.

Figure 1.

Radiological examination demonstrated shortening of all metacarpal bones (A) and 3^rd, 4^th, and 5^th metatarsal bones (B), and ganglia and subcortical white matter calcification on cranial computerized tomography imaging (arrows) (C). A visible rigid mass at the volar site of the left wrist (D.1) and the right antecubital fossa (circles) (D.2), bilateral radio dense masses at antecubital area (E. Right arm - F. Left arm) with radio-dense vascular tract (arrows) were compatible with physical examination (F).

Hypocalcaemia was initially treated with intravenous calcium gluconate concomitant with oral calcium carbonate, cholecalciferol and calcitriol. During intravenous (IV) calcium therapy, calcium gluconate extravasation was recognized by the nurse at the left distal wrist as hyperaemia on the skin and resolved spontaneously within a few days. The patient was discharged on the 7th day of hospitalization when her serum calcium rose to 8.1 mg/dL. Three weeks later, on the follow-up visit, a palpable rigid mass was detected at the volar side of the left distal wrist where the extravasation occurred (Figure 1.D.1). This mass extended along the vascular path as a rigid palpable string at the left forearm up to the antecubital fossa. Venous thrombosis with vascular calcification was considered to have been the consequence of calcium extravasation during IV treatment. However, a 2×1 cm of solid mass was also palpated at the antecubital fossa of the opposite arm (Figure 1.D.2), where venous punctures were previously performed, although there was no history of an IV line or extravasation of IV fluid at this site. Like the left arm, a rigid string was palpable along the vascular path to the wrist. X-rays demonstrated bilateral radio-dense masses in the antecubital area, which were compatible with physical examination (Figure 1.E and 1.F). Doppler ultrasound (USG) and CT angiography of the upper extremities revealed thromboses and calcifications in both antecubital veins, the distal radial and ulnar veins of the left arm, and the cephalic vein of the left arm (Figure 2.A–2.C). No thrombosis was detected in deep veins. Antibiotics and low molecular weight heparin (LMWH) were initiated for thrombophlebitis. Additionally, due to the extensive subcutaneous and vascular calcifications, acetazolamide treatment was considered, as this treatment has successfully reduced soft tissue calcifications in tumoral calcinosis [25–28]. Haematological evaluation for thrombosis showed normal coagulation parameters, including prothrombin time (PT), partial thromboplastin time (PTT), and the levels of D-dimer, fibrinogen, antithrombin 3, protein C, protein S, and factor VIII. No abnormalities were detected in platelet aggregation or collagen/ADP and collagen/epinephrine tests. Further etiological analysis of thrombosis showed normal lipoprotein (a) and homocysteine levels.

Figure 2.

CT angiographic images of both upper extremities revealed thromboses and wall calcifications in the right antecubital vein (arrow on A and single arrow on B), the left distal radial vein (double arrow on B), and the left antecubital vein (arrow on C). Doppler ultrasound images of the left upper extremity demonstrate wall calcifications (arrow) in the left antecubital vein and thrombosis of the cephalic vein (arrow head) (D).

LMWH treatment (100 units/kg/day) was planned to be given for three weeks; however, thrombosis was still present in the 3^rd week of treatment on Doppler USG, and therefore, the treatment continued for five months according to D-dimer levels. Acetazolamide was initiated at 12.5 mg/kg/day and then increased to 25 mg/kg/day on the 4^th day of treatment. Acidosis (pH: 7.24) was observed by the 3^rd week of treatment when the dose was decreased to 16 mg/kg/day (Table 1). The reduction in the size of calcifications was detected in the first week of acetazolamide treatment. Complete resolution of subcutaneous and vascular calcifications was observed on physical examination and Doppler USG at eight months of treatment (Supplementary figure 1). Thus, we discontinued the acetazolamide treatment. Over the eight-month treatment period, we did not observe any adverse effects. Despite the complete resolution of vascular and subcutaneous calcifications, intracranial calcifications were unchanged. In contrast, several subcutaneous calcific nodules, detected only by cranial CT, appeared to have shrunken upon acetazolamide treatment (Figure 3.F1 and F.2). Additionally, oral calcium and calcitriol doses required adjustment after initiation of acetazolamide treatment. Oral calcium was discontinued on the 4^th day of acetazolamide treatment since the PTH level declined from 145 μg/dl to 29 μg/dl. Subsequently, the calcitriol dose was decreased from 1.0 μg/day to 0.5 μg/day in the first week of treatment. Treatment doses and biochemical parameters are presented in the table. Neither hypercalciuria nor nephrocalcinosis on ultrasound examinations had been detected throughout the follow-up.

Figure 3.

Cranial CT images demonstrated probable ossifications as subcutaneous opacities (arrows) (A, B, F) and intracranial calcifications (D, E) at initial presentation (C.1-F.1) and after acetazolamide treatment (C.2-F.2). Although the same sections were not available due to different positioning during CT scans (C1, C2), no obvious differences were detected in basal ganglia and subcortical white matter calcification before (D1, E1) and after the acetazolamide treatment (D2, E2). After acetazolamide treatment, substantial improvement was observed in the two largest subcutaneous ossification foci (arrows) and in multiple tiny opacities (F2 vs F1).

Genetic Testing and Functional Analysis

Genomic DNA was extracted from peripheral blood lymphocytes. Polymerase chain reaction (PCR) amplification of exons and intron-exon boundaries of the GNAS gene was performed. The polymerase chain reaction (PCR) products were sequenced via ABI PRISM^® 3130xl genetic analyzer (Applied Biosystems, Foster City, CA, USA). The pathogenicity of the variant was assessed using ACMG/AMP criteria [29], including, for our case, the loss-of-function effect, segregation analysis, allele frequencies in control population databases (1000G, gnomAD, ExAC), and the results of our in vitro functional assays.

A panel analysis of thrombophilia gene polymorphisms was performed for factor V Leiden (G1691A and R506Q), factor II/prothrombin (G20210A), MTHFR (A1298C and C677T), PAI-1 (5G>4G), and factor XIII (V14L) by using a Light Cycler instrument according to the guidelines of the manufacturer. Mutation screening was performed by real-time PCR. The increase in the fluorescence of the product obtained through DNA amplification using RT-PCR was monitored in real-time. Gene polymorphism was detected based on a detailed melting curve analysis of PCR products.

Gsα deficient HEK293 cells expressing the luciferase-based GloSensor cAMP reporter have been previously described [30, 31]. Cells were transfected with empty vector (pcDNA3.1), HA-tagged wild-type Gsα (WT), HA-tagged Gsα-R356Tfs*47 (R356fs), or HA-tagged Gsα-Y391* using lipofectamine 3000 (L3000008, L3000008). The full-length HA epitope tag (YPYDVPDYA) replaced the residues 76–81 (DPQAAR) in the exon 3-encoded portion of Gsα. Transfected cells in 96-well plates were treated with luciferin and two mM 3-isobutyl-1-methylxanthine for 30 min before challenge with the vehicle, isoproterenol, or prostaglandin E2. The luminescence indicating the binding of cAMP to the reporter was determined as previously described [32]. For Western blots, whole-cell lysates were prepared in RIPA buffer 72 h after transfection, and proteins were separated by SDSPAGE using a gradient gel (4–20%). After blotting onto a PVDF membrane, the wild-type and mutant Gsα proteins were identified using an anti-HA antibody (sc-805, Santa Cruz Biotechnology, 1:500 dilution). Anti-tubulin (#5346, Cell Signaling Technology, 1:2000 dilution) was used as a loading control.

Statistics

Data from transfected cells were analysed by two-way ANOVA considering the expressed protein (wild-type, Gsα-R356Tfs*47, or Gsα-Y391*) and agonist concentration. Tukey’s posthoc test was used to conduct multiple comparisons.

Results

Sanger sequencing of the GNAS gene revealed a heterozygous frameshift variant (NM_000516.5: c.1065_1068delGCGT, p.R356Tfs*47), replacing the last 38 residues of the Gsα protein with a novel sequence of 46 amino acids (Supplementary figure 2). This variant was not observed in the Genome Aggregation Database (gnomAD) and not registered in ClinVar, the Human Genome Mutation Database (HGMD), or the Leiden Open Variation Database. We classified this novel variant as “pathogenic” according to ACMG/AMP guidelines.

The patient carried the normal genotypes for factor V Leiden-G1691A and R506Q, factor II/prothrombin-G20210A, MTHFR-A1298C, PAI-1 5G>4G, and factor XIII-V14L polymorphisms but was heterozygous for MTHFR-C677T polymorphism.

To determine whether the mutant functions like native Gsα, we transfected Gsα-deficient HEK293 cells with cDNA plasmids encoding wild-type Gsα or Gsα-R356Tfs*47. As expected, cells transiently expressing wild-type Gsα demonstrated increased basal cAMP accumulation and raised the cAMP levels upon treatment with the β-adrenergic receptor agonist isoproterenol or prostaglandin E2 (Figure 4). In contrast, cells expressing Gsα-R356Tfs*47 behaved indistinguishably from cells transfected with empty vector, displaying no basal or receptor-stimulated cAMP accumulation. We also compared these findings to a previously described nonsense Gsα mutant, Y391*, identified in a patient with PHP1C [10]. As expected, this mutant showed basal, but not receptorstimulated, activity. Western blot analysis confirmed that Gsα-R356Tfs*47 was expressed in transfected cells, although its level appeared modestly lower than wild-type Gsα and Y391X. Based on these results, the frameshift mutant is completely inactive, consistent with the absence of 39 functionally critical C-terminal residues (Figure 4). Within this missing portion, at least eight missense variants have been classified as pathogenic or likely pathogenic according to ClinVar and/or the Leiden Open Variation Database.

Figure 4:

The R356Tfs*47 mutant display no basal or receptor-stimulated activity. Panel A: Vehicle or isoproterenol-induced cAMP levels in Gsα-deficient HEK293 cells transfected with empty vector (pcDNA3.1), HA-tagged wild-type Gsα (WT) cDNA, HA-tagged Gsα -R356Tfs*47 (R356fs) cDNA, or HA-tagged Gsa-Y391* cDNA, measured by the GloSensor cAMP reporter assay. RLU, relative luciferase units. Veh, vehicle. ****, p<0.0001 by two-way ANOVA and Tukey’s posthoc test. Panel B: Representative images of the GloSensor assay showing the time-course of vehicle, isoproterenol-induced cAMP accumulation. Panel C: Time-course of cAMP accumulation in response to vehicle or PGE2 treatment in transfected Gsα -deficient HEK293 cells. Panel D: Western blot of cells transfected transiently with empty vector (pcDNA3.1), HA-tagged wild-type Gsα (WT), HA-tagged Gsa-R356Tfs*47 (R356fs), or HA-tagged Gsa-Y391*. Cells were lysed 48 h after transfection. Proteins were separated by SDS-PAGE, blotted onto PVDF membrane, and immunoreacted with anti-HA antibody. Anti-tubulin was used as a loading control.

Discussion

PHP1A is characterized by Gsα hypofunction, which can be demonstrated ex vivo in patient-derived erythrocytes. Gsα hypofunction has also been shown in platelets as impaired Gsα-mediated inhibition of the platelet aggregation [15, 24]. Thus, patients with PHP1A can be considered prone to thrombo-embolic disorders, yet no thrombosis case has hitherto been described. We now report a PHP1A case who manifested thrombotic events, as well as soft tissue and vascular calcifications, that were invoked by mild vascular trauma due to venous puncture. Our findings strongly suggest that the patient did not have any other known prothrombotic conditions, such as a deficiency of antithrombin-3, protein C, or protein S, or elevated levels of lipoprotein (a) or homocysteine. We also did not detect any evidence of malignancy. She also did not carry the factor V Leiden variant, nor did we detect any prothrombin variants that cause thrombosis frequently. A prothrombotic state, based on impaired Gsα-dependent inhibition of platelet aggregation, was described in a PHP1A case that was unique due to the presence of severe Gsα deficiency related to biallelic GNAS mutations [23]. In our patient, platelet aggregation tests were normal. Although we could not assess platelet aggregation inhibition or Gsα activity in this context, we could identify a frameshift variant in the last coding exon of GNAS. This variant (Gsα-R356Tfs*47) showed a complete lack of activity, stimulating cAMP production neither basally nor in response to receptor agonists, including prostaglandin E2. Similarly, the three GNAS mutations associated with a prothrombotic state ex vivo [24] also showed severe loss of Gsα activity in independent studies [4, 33]. Moreover, the PHP1A case described to have a prothrombotic state based on ex vivo platelet analyses carried compound heterozygous Gsα mutations causing a drastic cAMP deficiency [23]. We thus speculate that complete loss of Gsα activity, rather than loss of only the receptor-stimulated activity, may be required for thrombotic tendency. Furthermore, a study of patients with hypoparathyroidism or PHP found no cases of intravenous or oral calcium treatment induced thrombosis [34], but deep vein thrombosis after intravenous calcium injection was reported in a patient with gouty nephropathy who was under haemodialysis [35]. Calcium ion is the clotting factor IV, important in extrinsic, intrinsic, and common coagulation pathways [36]. Thus, it is conceivable that the rapidly increased local calcium concentration may have contributed to the formation of thrombosis in our case. However, thromboembolic events have rarely been reported concurrently with hypercalcemia related to primary hyperparathyroidism, and the reported cases appear to have an underlying predisposing condition, such as an MTFR mutation or vasculitis [37, 38]. Additionally, it is also considered that hyperphosphatemia promotes coagulation through the increased level of platelet-derived extracellular polyphosphate and induces secretion of pro-coagulant microvesicles derived from endothelial cells and platelets, although these mechanisms have been studied mainly in the context of chronic kidney disease [39–42]. Therefore, a prothrombotic state, together with loss of endothelial integrity, hyperphosphatemia, and intravenous calcium infusion, is the likely cause of thrombosis and vascular calcification in our patient.

Soft tissue calcification is not a classical finding in PHP1A. Detectable calcified dermal or subcutaneous nodules are true heterotopic intramembranous ossifications, usually limited to the subcutaneous tissues and highly heterogeneous in number and extent [2, 7, 8, 43]. The ectopic ossifications are a manifestation of G_sα deficiency in mesenchymal stem cells. Gsα reciprocally regulates adipogenesis and osteogenesis, with high cAMP levels being linked to adipogenesis and low levels to osteogenesis [44–46]. G_sα deficient adipose-derived progenitor mesenchymal stem cells differentiate into osteoblast progenitors, leading to de novo formation of ectopic bone islands in the dermis and the subcutaneous fat [44, 47]. These ectopic ossifications are unrelated to serum levels of calcium and phosphorus and are observed regardless of PTH resistance, i.e., both patients with maternal or paternal origin of the genetic mutation can develop ectopic ossifications. Nevertheless, calcifications at the basal ganglia (BGC) or, rarely, at other parts of the central nervous system can be seen in PHP [2, 7, 8, 48–50] and is also observed in hypoparathyroidism but not in patients with pseudo-PHP, who lack PTH resistance [2, 7, 8, 48–51]. Hence, BCG occurs regardless of Gsα deficiency but is associated with hypocalcaemia and hyperphosphatemia. A high calcium × phosphate product (CPP) (≥55 mg²/dl²), calculated by multiplying serum calcium by serum phosphate, has been associated with BGC or soft tissue calcification [8, 48, 51, 52]. However, intracranial and vascular calcifications developed in our patient despite apparently normal CPP. Therefore, CPP may not be the sole determinant of soft tissue calcification, at least in PHP and probably in hypoparathyroidism due to the common presence of severe hypocalcaemia, a multiplier in the formula. However, it could be the primary factor driving calcifications in the setting of chronic renal disease and tumoral calcinosis, where moderate to severe hypocalcaemia is not a typical part of the clinical picture [25–28, 52–54]. However, BCG is not a classical finding of tumoral calcinosis, and only a small calcification was detected in one patient in spite of huge soft tissue calcifications [55]. It appears that hyperphosphatemia can trigger ectopic calcifications, and not only phosphate levels but also an imbalance between inorganic phosphate and pyrophosphate (PPi) levels could be the leading mechanism [56]. Additionally, the local phosphate-to-PPi ratio can be different and present the main pathogenic mechanism governing BGC, as in the case of SLC20A2 mutations, in which altered local phosphate transport leads to idiopathic BGC [57]. The balance between phosphate and PPi is tightly regulated, and vascular and soft tissue calcifications are characteristics of extracellular PPi deficiency [58]. Thus, the changing balance between phosphate and PPi, either due to hyperphosphatemia or PPi deficiency, can be responsible for ectopic calcifications. However, soft tissue calcification demonstrates variation among different disease conditions concerning location and severity. These differences are likely to be associated with the presence of hypocalcaemia, the persistence and duration of calcium/phosphate abnormalities and local tissue texture.

The phosphate species in the extracellular matrix are composed of mono-, di-, and trivalent phosphate, and the negative trivalent phosphate ion is precipitated mostly by positively charged divalent calcium. Only a small fraction of the total plasma phosphate concentration exists as trivalent phosphate, which is significantly pHsensitive [59, 60]. Previous studies have shown that acidosis can prevent and reduce soft tissue calcification by increasing the solubility of calcium phosphate deposits and/or suppressing the expression of various osteogenic transcription factors [59, 61]. Additionally, mild acidosis decreases trivalent phosphate, which slows the rate of precipitation with calcium [60, 62]. Acetazolamide inhibits the enzyme carbonic anhydrase reversibly [63]. This inhibition results in a reduction of hydrogen ion excretion and an increase in renal excretion of sodium, potassium, bicarbonate, and water. Furthermore, it decreases production of cerebrospinal fluid and aqueous humour and resulting in metabolic acidosis [63, 64].As a consequence of these mechanisms, acetazolamide is indicated for the treatment of high-altitude cerebral oedema and acute angle-closure glaucoma and is also used off-label in idiopathic intracranial hypertension and metabolic alkalosis [62–64]. In addition, acetazolamide induces phosphaturia, likely through the inhibitory effect of an altered tubular luminal fluid pH on Na-linked phosphate co-transporters [65]. Therefore, acetazolamide has been used for urinary phosphate excretion in tumoral calcinosis and shown to be successful in decreasing or dissolution of soft tissue calcifications [25–28, 53–55]. However, we did not find any change in phosphate excretion during acetazolamide treatment (Table 1). Similarly, dissolution of calcific mass without any change in serum phosphate levels has been reported in a case with tumoral calcinosis [27]. Therefore, acidosis is likely to underpin the resolution of calcifications in our patient and probably other cases with tumoral calcinosis. In our case, complete resolution of vascular calcification was achieved after eight months of acetazolamide treatment. On the other hand, no detectable alterations in BGC and intracranial calcification by CT were observed. Moreover, BGC and intracranial calcifications are the classical findings of carbonic anhydrase type II deficiency [66]. Therefore, carbonic anhydrase inhibitors may be entirely ineffective for PHP1A-associated intracranial calcifications and should be used with caution in this respect. Acetazolamide treatment may cause minor side effect such as paraesthesia, polyuria, nausea, myopia, and very rarely, aplastic anaemia as an idiosyncratic reaction [63, 64, 67]. Nevertheless, no complication or side effect related to acetazolamide treatment was detected in our patient. However, we had to discontinue oral calcium and decrease the calcitriol doses due to normalization/suppression of PTH levels immediately after starting acetazolamide treatment. Although we did not detect any significant change in the measured ionized calcium levels, we speculate that the rapid normalization of PTH levels might reflect elevated ionized calcium levels resulting from acetazolamide-induced acidosis.

In addition, the size of the subcutaneous calcific nodules, most probably real ossifications, which were only detectable by CT, decreased in size with acetazolamide. Ossifications are different from calcification and no effective treatment exists for the treatment or prevention of ossification in PHP [8, 43]. Currently, surgical removal of ectopic ossifications can be performed if the lesion is disabling and well delimited. Several drugs have been tried with variable responses, including bisphosphonates, steroids and nonsteroidal anti-inflammatory drugs, thiazolidinediones, and retinoids [8, 43, 68]. We cannot speculate that acetazolamide decreases ectopic ossification based on our study, but it is possible that the mineralization of the ossified tissue decreases with treatment [61]. Additional studies are required to figure out the direct and indirect impact of acetazolamide on ectopic ossification.

In summary, this is the first case of thrombosis and the successful use of acetazolamide for soft tissue and vascular calcifications in PHP1A. Here, several other factors, including IV calcium therapy and local trauma, could have contributed to the soft tissue/vascular calcification and thrombosis. Their effects may be in addition to or independently from the Gsα deficiency and biochemical alterations related to PHP1A, such as hyperphosphatemia. Nonetheless, it is essential to consider that individuals with PHP1A may be prone to a prothrombotic state, leading to an elevated risk of thrombosis, especially in the presence of additional risk factors, such as smoking, oestrogen administration, and prolonged immobilization. This case expands the clinical spectrum and provides novel approaches to the treatment of non-hormonal features in this disease.

Statements of Ethics

This study was conducted in compliance with the terms of the World Association Declaration of Helsinki, and written informed consent was obtained from the parent of the patient for publication of the details of their medical care and any accompanying images. Ethical approval was not required for this study in accordance with local/national guidelines.

Established Facts and Novel Insights.

Established Facts

Pseudohypoparathyroidism type IA (PHP1A) is caused by inactivating mutations of the GNAS gene encoding the α-subunit of the stimulatory G protein (Gsα), resulting in end-organ resistance to hormone actions and laboratory evidence for impaired inhibition of platelet aggregation.

In addition to subcutaneous ossifications, patients with PHP1A commonly exhibit soft tissue calcifications.

Acetazolamide has shown successful outcomes in treating soft tissue calcification in patients with tumoral calcinosis.

Novel Insights

The first PHP1A patient with thrombosis and the successful use of acetazolamide for PHP1A-associated soft tissue calcifications have been described, providing new insights into non-endocrinological complications of this disease and their treatment.

Data Availability Statement

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.