Parathyroid Hormone Resistance and Autoantibodies to the PTH1 Receptor

SUMMARY

We describe two cases of acquired parathyroid hormone (PTH) resistance consequent to the development of serum PTH type 1 receptor (PTH1R) autoantibodies, which block PTH binding and signaling. Both cases were associated with other autoimmune manifestations, and one case was associated with atypical membranous glomerulonephritis. In vitro binding and signaling assays identified the presence of PTH1R-blocking IgG autoantibodies, which were not present in serum samples from patients with other renal or autoimmune disorders. (Funded by the Intramural Research Programs of the National Institute of Diabetes and Digestive and Kidney Diseases and others.)

Impaired parathyroid hormone (PTH) activity — as occurs in patients with PTH deficiency (hypoparathyroidism) — or resistance (pseudohypoparathyroidism) leads to hypocalcemia and hyperphosphatemia. PTH mediates its actions by binding to the PTH type 1 receptor (PTH1R), which is located on the plasma membrane of target cells; this in turn leads to activation of the G protein G_sα and to stimulation of intracellular cyclic adenosine monophosphate (cAMP) production. In the kidney, PTH decreases the reabsorption of phosphate and stimulates the synthesis of the active form of vitamin D (1,25-dihydroxyvitamin D) in the proximal tubule, whereas it increases calcium reabsorption in the distal tubule. In bone, PTH stimulates calcium and phosphate mobilization. PTH resistance in the absence of kidney failure is typically associated with genetic or epigenetic abnormalities (pseudohypoparathyroidism types 1A and 1B, respectively) involving GNAS, the gene encoding G_sα. These abnormalities are congenital disorders in which PTH resistance is established during early childhood.¹ In this report, we describe two cases of acquired severe hypocalcemia resulting from the development of PTH resistance in association with PTH1R-blocking autoantibodies.

CASE REPORTS

PATIENT 1

A 70-year-old Black woman with a history of hypertension and previously normal serum calcium levels, who had been referred to the National Institutes of Health at 60 years of age with new-onset hypocalcemia, presented with muscle cramps in her legs, a tingling sensation in her hands and feet, numbness around her mouth, irritability, and impaired memory. She reported no family history of calcium or parathyroid abnormalities but two sisters had both end-stage kidney disease of unknown cause and autoimmune thyroid disease. The patient did not have any physical features of Albright hereditary osteo-dystrophy, such as short stature, brachydactyly, subcutaneous ossifications, or obesity. The initial biochemical evaluation showed a markedly low level of total serum calcium (5.7 mg per deciliter [1.4 mmol per liter]), an elevated phosphorus level (10 mg per deciliter [3.2 mmol per liter]), a mildly low magnesium level (1.7 mg per deciliter [0.7 mmol per liter]) (lower limit of normal, 1.8 mg per deciliter [0.75 mmol per liter]), a very high PTH level (>1000 pg per milliliter), a normal 25-hydroxyvitamin D level (47 ng per milliliter [117 nmol per liter]), and undetectable levels of 1,25-dihydroxyvitamin D (<8 pg per milliliter [<20 nmol per liter]). Renal ultrasonography was performed and showed no evidence of nephrolithiasis or nephrocalcinosis. Treatment with calcium carbonate and calcitriol was initiated, but the patient had extreme fluctuations in serum calcium and PTH levels, leading to frequent adjustments in treatment (Fig. 1A). Despite aggressive treatment, the PTH levels generally remained extremely high, whereas more aggressive treatment with oral calcium and calcitriol was accompanied by increases in serum creatinine levels (Fig. 1A). Despite the persistently elevated PTH levels, the bone mineral densities, as measured on dual-energy x-ray absorptiometry, remained high, with T scores of 1.7 for the anteroposterior spine, 1.2 for the total hip, and 1.7 for the forearm. Owing to the continuing increase in the serum creatinine levels (Fig. 1A) and the development of nephrotic-range proteinuria (with a urinary protein excretion of 7 g per 24 hours), albuminuria (with a urinary albumin excretion of 6 g per 24 hours), and hypoalbuminemia (with a serum albumin level of 3.1 g per deciliter), a biopsy specimen of the kidney was obtained. The specimen showed an atypical, membranous-like nephropathy and markedly thickened basement membranes, with predominantly intramembranous aggregates of unusual microspherules (Fig. 1B through 1E). The composition of these structures was unknown, but they lacked the appearance or immunofluorescence staining pattern of immune complexes characteristically seen in patients with membranous nephropathy. Immunofluorescence staining for the M-type phospholipase A2 receptor was negative.² In addition, serologic tests for antibodies to both M-type phospholipase A2 receptor and thrombospondin type-1 domain-containing 7A (THSD7A)^2,3 were negative, as were tests for hepatitis B and C, which may be associated with membranous nephropathy. Tests for antinuclear antibody, extractable nuclear antigen, and anticentromere antibodies were positive (for the complete results of serologic testing, see Table S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). The patient was treated with an angiotensin-converting–enzyme inhibitor, which led to gradual abatement of the proteinuria and albuminuria. Other autoimmune manifestations included alopecia totalis, autoimmune thyroiditis (for which thyroid hormone–replacement therapy was required), Jaccoud’s arthropathy in the left hand, a small erosion of the metacarpophalangeal joint of the left thumb, and dry eyes and mouth. Biopsy specimens of the salivary glands showed a Sjögren’s focus score of 4, with absence of discrete lymphoid aggregates, and one gland showed diffuse involvement. On the basis of the rheumatologic signs and symptoms, a diagnosis of undifferentiated connective-tissue disease was made. The patient met the American College of Rheumatology–European League against Rheumatism 2016 criteria for Sjögren’s syndrome and had features of lupus, but she did not meet the Systemic Lupus International Collaborating Clinics’ 2019 criteria for lupus.

Figure 1 (facing page). Mineral Metabolism and Renal Pathology in Patient 1.

Panel A shows the changes in the patient’s serum levels of parathyroid hormone (PTH) (in blue) and creatinine (in gray) and blood levels of ionized calcium (in purple) over 10 years during treatment. Black arrows denote representative time points of undertreatment and red arrows representative time points of more aggressive treatment. Panels B and C show representative glomeruli with atypical, membranous-like nephropathy manifested by diffuse thickening of the glomerular capillary loops (yellow arrows; hematoxylin and eosin staining in Panel B and periodic acid-Schiff staining in Panel C; magnification of 600 in both panels). In Panels D and E, images obtained on electron microscopy show the ultrastructure of a thickened glomerular capillary loop containing atypical, microspherule-shaped, intramembranous and subepithelial aggregates of unknown composition (white arrows; magnifications of 12,000 in Panel D and 20,000 in Panel E).

PATIENT 2

A 39-year-old Black woman with a history of immune thrombocytopenia, α-thalassemia, and lactose intolerance presented with hypocalcemia. She had had a normal calcium level at 33 years of age and new-onset hypocalcemia at 37 years of age, at which time she presented with muscle cramps, tingling in the arms and legs, perioral numbness, lethargy, disorientation, and weakness. There was no family history of autoimmune disorders or abnormalities in mineral metabolism. At the time of her recent presentation, an initial biochemical evaluation showed markedly low levels of total serum calcium (5.5 mg per deciliter [1.4 mmol per liter]), an elevated phosphorus level (5.8 mg per deciliter [1.9 mmol per liter]), a low magnesium level (2.9 mg per deciliter [1.2 mmol per liter]), a very high PTH level (1362 pg per milliliter), and a low 25-hydroxyvitamin D level (14 ng per milliliter [34.9 nmol per liter]). The serum levels of urea nitrogen, creatinine, and thyrotropin were normal, and the albumin level was low (3.2 mg per deciliter). The T scores on tests of bone mineral density in the femoral neck, anteroposterior spine, and forearm were 3.0, 1.5, and 1.3, respectively. After initial treatment with intravenous calcium, which was followed by oral calcium, calcitriol, ergocalciferol, and magnesium, the patient remained mildly symptomatic, with a serum calcium level of 8.8 mg per deciliter (2.2 mmol per liter), a PTH level of 685 pg per milliliter, and a creatinine level of 0.87 mg per deciliter (76.9 μmol per liter). She had had development of immune thrombocytopenia at 25 years of age and had received intermittent treatment with glucocorticoids, intravenous immunoglobulins, and romiplostim as well as blood transfusions. She had had a splenectomy at 33 years of age. At the time of the most recent presentation, anti–thyroid peroxidase antibodies were undetectable.

METHODS

This study was approved by the institutional review board of the National Institute of Diabetes and Digestive and Kidney Diseases, and the patients provided written, informed consent. For Patient 1, whole-exome sequencing was performed by GeneDx, and monogenic autoimmunity and autoinflammatory panels (Fig. S1) were screened by Invitae. Patient 2 was screened with the use of a clinical exome sequencing panel prepared by Fulgent Genetics. GNAS methylation analysis was performed by means of Southern analysis⁴ for Patient 1 and by CGC Genetics for Patient 2. Luciferase immunoprecipitation systems (LIPS) technology was used for the detection of autoantibodies to PTH and PTH1R, as described in the Methods section of the Supplementary Appendix. A cell-based assay measuring cAMP responses to PTH 1-34 (Forteo, Eli Lilly) in PTH1R-expressing cells was used to confirm the presence of PTH1R-blocking IgG antibodies, as described in detail in the Methods section of the Supplementary Appendix. Lymphocyte immunophenotyping was performed (also as described in the Methods section of the Supplementary Appendix).

RESULTS

ABSENCE OF EVIDENCE FOR KNOWN CAUSES OF PTH RESISTANCE

No mutations in genes associated with hypocalcemia were identified, including PTH, PTH1R, CASR, GATA3, GNA11, and GNAS. Furthermore, no methylation defect associated with pseudohypoparathyroidism type 1B⁴ was identified in either patient on methylation analysis.

IDENTIFICATION OF PTH1R-BLOCKING AUTOANTIBODIES

The presence of multiple autoimmune manifestations raised suspicion that an autoimmune disorder might underlie the acquired PTH resistance. We hypothesized that PTH resistance developed in these patients because of impaired PTH binding to its receptor, as a result of the presence of either autoantibodies to PTH or blocking autoantibodies to PTH1R, and we assessed these possibilities with the use of LIPS technology.⁵ Plasmids encoding light-emitting Gaussia luciferase fusion proteins for both PTH and PTH1R were constructed and transfected into Cos1 cells. The crude cell extracts were then used for LIPS assays for the measurement of autoantibodies. No increase in immunoreactivity against PTH was detected in either patient as compared with healthy controls (Fig. 2A). In contrast, both patients had very high serum levels of autoantibodies against PTH1R as compared with heathy controls and patients with autoimmune polyendocrinopathy–candidiasis–ectodermal dystrophy, membranous nephropathy, or other kidney disorders associated with nephrotic syndrome (Fig. 2B). Three patients in the latter group had uremia. Additional mapping experiments performed with PTH1R truncation mutants encompassing the extracellular domain (PTH1R-ECD), the region involved in PTH binding, or the transmembrane domain (PTH1R-TMD) showed that PTH1R autoantibodies were directed largely to the ECD in Patient 1 and directed exclusively to the ECD in Patient 2 (Fig. 2C).

Figure 2. Identification of PTH1R-Blocking Autoantibodies.

Panel A shows the results of a luciferase immunoprecipitation systems (LIPS) autoantibody analysis for human PTH in six healthy volunteers and in Patients 1 and 2, in which there is no evidence of autoantibodies to PTH. Panel B shows the results of LIPS autoantibody profiling for human, near full-length PTH type 1 receptors (PTH1Rs). Levels of PTH1R autoantibodies in the serum samples from both Patient 1 and Patient 2 were high as compared with the levels in samples from 14 healthy volunteers, 19 patients with various renal diseases (including membranoproliferative glomerulonephritis, C3 glomerulopathy, minimal change disease, thrombotic microangiopathy, vasculitis, light-chain deposition disease, and class III or IV lupus nephritis), 20 patients with autoimmune polyendocrinopathy candidiasis with ectodermal dystrophy (APECED), and 24 patients with membranous nephropathy (including 3 patients with uremia). Panel C shows LIPS autoantibody profiling against the PTH1R extracellular domain (PTH1R-ECD) and PTH1R transmembrane domain (PTH1R-TMD) in 15 healthy volunteers and in Patients 1 and 2, indicating that PTH1R autoantibodies bind predominantly to the extracellular domain in Patient 1 and exclusively to the extracellular domain in Patient 2. In Panels A, B, and C, raw light units reflect autoantibody levels determined by means of LIPS. The two data points for Patient 1 in Panels B and C represent repeated measurements of the same sample. Panel D shows stimulation of cyclic adenosine monophosphate (cAMP) in response to PTH 1-34 (182 pg per milliliter) in the presence of antibodies isolated from 4 healthy volunteers and from Patients 1 and 2, with all individual responses normalized to the mean response in healthy volunteers. In healthy volunteers, levels of PTH 1-34–stimulated light units were approximately five times as high as those of unstimulated light units. In Patient 1, results similar to those shown in Panel D were also observed at a higher PTH 1-34 dose (22,750 pg per milliliter), although twice the amount of IgG was included in that experiment (data not shown). Means and standard deviations are shown for each group.

To confirm that the PTH1R autoantibodies identified in both patients could block PTH signaling through PTH1R, we purified IgG free of endogenous PTH from the serum samples of patients and controls and used a cell-based assay to examine the capacity of the purified IgG to block the intracellular generation of cAMP in response to PTH 1-34.⁶ As compared with samples from healthy volunteers, purified IgG from both patients blocked the PTH-dependent increase in cAMP activation by more than 90% (Fig. 2D). These results support the hypothesis that the patients’ anti-PTH1R autoantibodies blocked PTH action and thereby mediated PTH resistance.

IMMUNOPHENOTYPING

Patient 1 had elevated levels of several autoantibodies, C-reactive protein, and IgG as well as an elevated erythrocyte sedimentation rate (Table S1 in the Supplementary Appendix). The results of genetic screening for monogenic autoimmunity and autoinflammatory genes were negative (Fig. S1). Immunophenotyping of peripheral-blood lymphocyte subsets with the use of flow cytometry showed a normal frequency of CD3+ T lymphocytes, a normal CD4-to-CD8 ratio, and normal representation of CD4+ and CD8+ T-cell subpopulations. The patient also had normal frequencies of natural killer (NK) cells and NK T cells, total B cells, immature and transitional B cells, and plasmablasts but reduced frequencies of switched and nonswitched memory B-cell subsets.

In Patient 2, the frequency of CD3+ T lymphocytes and CD8+ T cells was normal, but the CD4-to-CD8 ratio was low, as was the frequency of CD4+ T cells. The frequency of CD8+ T cells was normal. Regarding T-lymphocyte subpopulations, Patient 1 had decreased frequencies of central memory CD4+ and CD8+ T cells and naive CD4+ cells and increased frequencies of effector memory CD4+ T cells. Patient 2 also had normal frequencies of NK and NK T cells, increased frequencies of total B cells and switched and nonswitched memory B-cell subsets and a decreased frequency of early transitional naive B cells, whereas frequencies of immature and transitional B cells were normal.

Patient 2 had a markedly increased frequency of autoreactive CD21-low–CD38-negative B cells, which were also found at the upper limit of the normal frequency in Patient 1 (Table S2). CD21-low–CD38-negative B cells are enriched in patients with certain monogenic autoimmune diseases (e.g., autoimmune polyendocrinopathy candidiasis with ectodermal dystrophy⁷), in patients with common variable immunodeficiency-associated autoimmune cytopenia⁸ and cytotoxic T lymphocyte–associated protein 4 haploinsufficiency–associated immune dysregulation,⁹ and in some patients with polygenic autoimmune diseases (systemic lupus erythematosus or rheumatoid arthritis) associated with autoantibody production.¹⁰

DISCUSSION

These two cases of hypocalcemia were associated with PTH resistance owing to the presence of blocking autoantibodies to PTH1R. As compared with most cases of primary PTH resistance, which are congenital syndromes involving genetic or epigenetic defects in GNAS, the gene encoding the G protein that mediates PTH1R action, these cases had several unusual features: late-onset, markedly elevated PTH levels that were difficult to correct with standard therapy, and an association with other autoimmune manifestations. Treatment with glucocorticoids, plasmapheresis, or other immunosuppressive therapies may prove useful in addressing PTH resistance and other manifestations of autoimmune disease. It is interesting that both patients had high bone mass despite having a clinically significant history of severely elevated PTH levels, a finding that is consistent with global resistance to PTH, as would be expected given the presence of PTH1R-blocking autoantibodies. It is unlikely that intermittent changes in PTH levels during the treatment period contributed to the high bone density in Patient 1, since there was no substantial change in bone density during the treatment period.

One challenge involved in the treatment of Patient 1 was worsening renal function. A renal biopsy specimen showed an unusual form of membranous nephropathy, with global and diffuse spherular deposits in glomeruli and tubules. Membranous nephropathy with spherules is a rare variant of membranous nephropathy that is poorly understood.^11,12 It is unknown whether these distinct ultrastructural morphologic features reflect a distinct subset of membranous nephropathy with a new antigen–antibody complex or a different form of pathogenesis. Given the distribution of deposits within the tubules where PTH1R is expressed,¹³ it is worth considering a potential role for circulating PTH1R autoantibodies in the development of the renal manifestations. Although a previous study in which a competitive binding assay was used suggested that most patients with uremia have blocking autoantibodies to PTH1R,¹⁴ our findings, based on the use of a more direct binding assay, showed that PTH1R autoantibodies were absent in all 41 patients with kidney disease, including 3 patients with uremia. It therefore does not appear that PTH1R autoantibodies are a common feature of uremia. It is of interest that our patients had increased frequencies of the CD21-low–CD38-negative B-cell subset in peripheral blood, an enrichment of which has also been reported in patients with autoimmunity associated with autoantibody production.^7–10

In conclusion, we identified two patients with acquired autoimmune PTH resistance caused by PTH1R-blocking autoantibodies. These autoantibodies, which prevent PTH from binding to its receptor, were primarily directed at the extracellular domain, a region required for PTH binding.