Endocrine Journal
Online ISSN : 1348-4540
Print ISSN : 0918-8959
ISSN-L : 0918-8959
CASE REPORT WITH REVIEW OF LITERATURE
A case of central diabetes insipidus after COVID-19 as a probable diagnosis of lymphocytic infundibulo-neurohypophysitis with positive anti-rabphilin-3A antibodies with review of literature
Yuka NatsukiYuki Nagata Toshiki NagasakiMari MorimotoNorikazu ToiMasafumi KurajohTomoaki MoriokaTetsuo ShojiYasuo ImanishiNaoko IwataHaruki FujisawaAtsushi SuzukiYoshihisa SugimuraMasanori Emoto
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2024 Volume 71 Issue 12 Pages 1165-1173

Details
Abstract

Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2, and various complications have been reported. Furthermore, there have been increasing reports of endocrinopathy related to COVID-19 following the pandemic. We report a 49-year-old healthy woman who developed rapid onset of polydipsia and polyuria three weeks after COVID-19. Laboratory tests indicated low urine osmolarity and increased serum osmolarity, and antidiuretic hormone (ADH) was undetectable. Urine osmolality remained low with water deprivation. Similarly, plasma ADH responses to hypertonic-saline infusion were blunted and urine osmolality increased in response to desmopressin. There was no clear evidence of anterior pituitary dysfunction. T1-weighted magnetic resonance imaging (MRI) showed pituitary stalk thickening and absence of posterior pituitary bright signal spots, suggesting the presence of hypophysitis. Based on these results, we made a probable diagnosis of lymphocytic infundibulo-neurohypophysitis (LINH) which have caused central diabetes insipidus. Positive findings for serum anti-rabphilin-3A antibodies, reported as a potential diagnostic marker for LINH, were also noted. Following oral desmopressin administration, polydipsia and polyuria were quickly improved, though treatment with desmopressin was still required over four months. This is the first report of a patient with a probable diagnosis of LINH after COVID-19 who tested positive for anti-rabphilin-3A antibodies. Positive findings for those antibodies suggest that pituitary dysfunction associated with COVID-19 is hypophysitis involving an abnormal immune mechanism. The presence of anti-rabphilin-3A antibodies may be useful as a non-invasive diagnostic marker of LINH and potentially serve as a valuable diagnostic aid in cases of LINH associated with COVID-19.

 Introduction

The pathogenesis of coronavirus disease 2019 (COVID-19) has been elucidated, while a number of related sequelae have also been reported [1, 2]. It has become widely accepted that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes multiorgan dysfunction, as various tissues and organs can be affected, including the endocrine system [3, 4]. Some of those endocrine disorders appear during the period of SARS-CoV-2 infection, while others develop several months after full recovery [3]. Other studies have shown that SARS-CoV-2 vaccination might also trigger onset of hypophysitis [5, 6]. The mechanisms for autoimmune endocrine disorders reported to follow SARS-CoV-2 infection and vaccination remain under investigation.

Recently, anti-rabphilin-3A antibodies were shown to be a highly sensitive and specific diagnostic marker for lymphocytic infundibulo-neurohypophysitis (LINH) [7, 8]. Since the procedure used is non-invasive, it may be valuable for the differential diagnosis in patients presenting central diabetes insipidus (CDI) [9]. Here, we report a case of Japanese middle-aged woman complicated by continued polydipsia and polyuria after COVID-19. She was diagnosed with CDI based on endocrinological findings, and positive results for anti-rabphilin-3A antibodies strongly supported the possibility of LINH. Additionally, previously reported cases of CDI as a COVID-19-related event are reviewed.

 Materials and Methods

Serum obtained from the present patient was examined with western blotting for detection of anti-rabphilin-3A antibodies using a previously reported method [7, 8]. Briefly, a vector containing the full-length human rabphilin-3A gene was transfected into HEK293FT cells to produce recombinant human rabphilin-3A protein. Expression of recombinant human rabphilin-3A protein was confirmed using an anti-V5 antibody. As a negative control, the same vector without the rabphilin-3A gene was transfected into HEK293FT cells. A protein band presenting a size of 76 kDa, corresponding to the molecular weight of rabphilin-3A, appeared in the lysate of cells transfected with rabphilin-3A protein but not in that of the control cells, which was considered to indicate positive for anti-rabphilin-3A antibodies.

 Case Report

A 49-year-old Japanese woman was presented with a fever and dry cough, and subsequently diagnosed with COVID-19 by antigen test. Approximately one and a half years prior to the infection, she had undergone a third COVID-19 mRNA vaccine inoculation (BNT162b2/Pfizer-BioNTech). While the fever and respiratory symptoms gradually improved with nirmatrelvir-ritonavir administration for 5 days, she developed dysgeusia one week after the SARS-CoV-2 infection. Administration with zinc supplements was started and the dysgeusia was resolved, though polydipsia and polyuria developed approximately three weeks after onset of the infection. The symptoms gradually became exacerbated and she visited her family physician. There was no loss of body weight after COVID-19. Biochemical evaluation results showed that serum sodium, calcium, creatinine, and plasma glucose levels were within normal ranges. Paired serum and urine osmolalities were 292 and 124 mOsm/kg, respectively. There was no evidence of active salivary gland inflammation in echography findings nor any pre-existing comorbidity that required regular medication. The physician suspected CDI and referred her to our hospital.

Approximately 10 weeks after the COVID-19 diagnosis, she was admitted to our hospital. Upon admission, she was conscious and hemodynamically stable, though tachycardia (pulse rate 109 bpm) and dry tongue were noted. She did not have a headache or visual impairment, and urine volume reached 11,950 mL/day. Laboratory results obtained at the time of admission are shown in Table 1. Those indicated elevated serum osmolality with a low level of antidiuretic hormone (ADH) which was measured by a radioimmunoassay kit (YAMASA Shoyu Corporation, Choshi, Japan), while the prolactin (PRL) level was approximately three times the upper limit of the normal range. There was no evidence of infection, tumor, or vasculitis. Urinary findings were also normal other than low urinary osmolality (74 mOsm/kg).

Table 1 Laboratory data at time of admission

Investigation Results Normal range Investigation Results Normal range
Hematology Endocrinology
WBC (/μL) 8,700 3,300–8,600 PRA (ng/mL/hr) 1.4 0.2–2.3
RBC (×104/μL) 458 386–492 PAC (pg/mL) 33.0 4.0–82.1
Hb (g/dL) 13.3 11.6–14.8 ACTH (pg/mL) 29.3 7.2–63.3
Ht (%) 40.7 35.1–44.4 Cortisol (μg/dL) 9.6 3.7–19.4
Plt (×104/μL) 30.4 15.8–34.8 DHEA-S (μg/dL) 294 19–231
Biochemistry TSH (μIU/mL) 1.25 0.50–5.00
TP (g/dL) 6.9 6.6–8.1 F-T4 (ng/dL) 1.08 0.90–1.70
Alb (g/dL) 4.0 4.1–5.1 GH (ng/mL) 0.21 0.13–9.88
AST (U/L) 46 13–30 IGF-1 (ng/mL) 138 81–218
ALT (U/L) 66 7–23 FSH (mIU/mL) 43.0 <157.8
γ-GTP (U/L) 27 9–32 LH (mIU/mL) 15.2 5.7–64.3
T-bil (mg/dL) 0.6 0.4–1.5 PRL (ng/mL) 42.1 3.1–15.4
ALP (U/L) 79 38–113 ADH (pg/mL) <0.4 *
Amy (U/L) 85 44–132 hCG (mIU/mL) <2.3 <5.0
CK (U/L) 68 41–153 ACE (U/L) 10.8 7.7–29.4
BUN (mg/dL) 6 8–20 Immunochemistry
Cr (mg/dL) 0.44 0.46–0.79 CRP (mg/dL) 0.06 0–0.14
eGFR (mL/min/1.73 m2) 115.2 ≥60 sIL-2R (U/mL) 242 204–587
UA (mg/dL) 3.8 2.6–5.5 PR3-ANCA (IU/mL) 0.6 <2.0
Na (mEq/L) 145 138–145 MPO-ANCA (IU/mL) 0.2 <3.5
K (mEq/L) 3.8 3.6–4.8 anti-nuclear antibody (–) (–)
Cl (mEq/L) 111 101–108 anti-SS-A antibody (U/mL) <1.0 <10.0
Ca (mg/dL) 9.0 8.8–10.1 anti-SS-B antibody (U/mL) 2.2 <10.0
Pi (mg/dL) 3.4 2.7–4.6 IgG4 (mg/dL) 42.1 11–121
FPG (mg/dL) 95 73–109 T-SPOT.TB test (–) (–)
HbA1c (%) 5.3 4.6–6.2 β-D glucan (pg/mL) <5.0 ≤20.0
Osmolality (mOsm/kg) 293 276–292
CEA (ng/mL) <1.7 ≤5.0
AFP (ng/mL) 4.6 ≤20.0

Abbreviations: WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; Ht, hematocrit; Plt, platelet count; TP, total protein; Alb, albumin; AST, aspartate aminotransferase activity; ALT, alanine aminotransferase activity; γ-GTP, gamma-glutamyl transferase activity; T-bil, total bilirubin; ALP, alkaline phosphatase activity; Amy, amylase; CK, creatine kinase; BUN, blood urea nitrogen; Cr, creatinine; eGFR, estimated glomerular filtration rate; UA, uric acid; Na, sodium; K, potassium; Cl, chloride; Ca, calcium; Pi, phosphorus; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; CEA, carcinoembryonic antigen; AFP, alpha-fetoprotein; PRA, plasma renin activity; PAC, plasma aldosterone concentration; ACTH, adrenocorticotropic hormone; DHEA-S, dehydroepiandrosterone sulfate; TSH, thyroid-stimulating hormone; F-T4, free thyroxine; GH, growth hormone; FSH, follicle stimulating hormone; LH, luteinizing hormone; PRL, prolactin; ADH, antidiuretic hormone; hCG, human chorionic gonadotropin; ACE, angiotensin-converting enzyme; CRP, C-reactive protein; sIL-2R, soluble interleukin-2 receptor; PR3-ANCA, proteinase 3 anti-neutrophil cytoplasmic antibody; MPO-ANCA, myeloperoxidase anti-neutrophil cytoplasmic antibody.

*The reference range for plasma ADH level is dependent on plasma osmolality.

A water deprivation test showed that body weight was reduced by more than 3% after 210 minutes and the test was finished at that time. Urine osmolality was below 300 mOsm/kg (40–130 mOsm/kg) during the test, and plasma ADH level was undetectable (Table 2A). Similarly, plasma ADH levels remained low throughout the hypertonic saline test (Table 2B). Intranasal desmopressin acetate (DDAVP) at 10 μg was administered, resulting in markedly decreased urine output, while urine osmolality was increased to 519 mOsm/kg (Table 2C). A three-hormone anterior-pituitary test was performed with no clear evidence of anterior pituitary dysfunction. The results of growth hormone-releasing peptide-2 (GHRP-2) test (GH peak 20.9 ng/mL at 15 minutes) and the insulin tolerance test (cortisol peak 19.6 ng/mL at 60 minutes) were also normal. Magnetic resonance imaging (MRI) revealed lack of posterior pituitary bright spots on T1-weighted images and thickening of the pituitary stalk, suggesting lymphocytic hypophysitis (Fig. 1A-C). Based on these results, a probable diagnosis of LINH causing central diabetes insipidus was made. Later, she was found positive for anti-rabphilin-3A antibodies (Fig. 2 and Supplementary Fig. 1), which strongly supported the possibility of LINH. Following administration of oral desmopressin (60–90 μg/day), the symptoms quickly improved, urinary output was maintained at <3 L/day, and she was discharged. Although T1-weighted MRI findings obtained one month later revealed that pituitary stalk thickening was reduced (Fig. 1D-F), she still required treatment with desmopressin for more than four months until the end of the observation period.

Table 2A Water deprivation test

0 min 60 min 120 min 180 min 210 min
Weight (kg) 49.50 48.75 48.25 48.05 47.95
Urine volume (mL/hr) 310 590 300 260
Plasma osmolality (mOsm/kg) 283 289 292 293
Urine osmolality (mOsm/kg) 55 65 98 128 120
Serum Na (mEq/L) 140 144 145 146
ADH (pg/mL) <0.4 0.6 <0.4 0.4
Table 2B Hypertonic saline test

0 min 30 min 60 min 90 min 120 min
Plasma osmolality (mOsm/kg) 293 301 304 309 313
Urine osmolality (mOsm/kg) 63 90 175 259 310
Serum Na (mEq/L) 145 149 152 154 156
ADH (pg/mL) <0.4 0.6 <0.4 0.5 <0.4
Table 2C Response to desmopressin acetate (DDAVP)

0 min 30 min 60 min 90 min 120 min
Urine volume (mL/30 min) 330 480 10 20 20
Urine Na (mEq/L) 13 13 55 55 59
Urine osmolality (mOsm/kg) 62 68 519 496 491
Fig. 1  T1-weighted magnetic resonance imaging (MRI) showed lack of posterior pituitary bright spots (A), and gadolinium-enhanced MRI revealed thickening of the pituitary stalk in the sagittal (B) and coronal (C) sections. T1-weighted MRI images obtained one month later revealed reduced pituitary stalk thickening (D–F).
Fig. 2  Detection of anti-rabphilin-3A antibodies by western blotting. Recombinant full-length human rabphilin-3A expressed in HEK293FT cells (RPH3A, left lane in each case) or a negative control (-, right lane in each case) were probed with serum from the present case (patient) or from a patient who was diagnosed with LINH by biopsy previously (positive control patient). Recombinant full-length human rabphilin-3A expressed in HEK293FT cells was also probed with an anti-V5 antibody as positive control (Anti-V5 antibody) in the first lane from the left. A protein band of 76 kDa that appeared in RPH3A (arrowheads) but not in that of negative control (dashed arrowheads) was considered to be positive for anti-rabphilin-3A antibodies.

RPH3A, rabphilin-3A; TFs, transfections of full-length human rabphilin-3A gene

 Discussion

Although reports of COVID-19 cases have accumulated over the years, only a few have noted suspicion of new onset of lymphocytic hypophysitis following SARS-CoV-2 infection. A notable aspect of the present case is the approach used for clinical diagnosis of LINH based on confirmation of hypothalamic-pituitary function with various hormone stimulation tests and positive results for the presence of anti-rabphilin-3A antibodies.

Rabphilin-3A expressed in neurohypophysis and hypothalamic vasopressin neurons excluding adenohypophysis, proved to be the most useful autoantigen for diagnosis of LINH [7]. That same report presented cases with biopsy-proven diagnosis, with anti-rabphilin-3A antibodies detected in all four patients with histologically confirmed LINH (sensitivity 100%) and in 22 of 29 with LINH (sensitivity 76%) including those clinically diagnosed. In addition, anti-rabphilin-3A was found to have a specificity of 100% for distinguishing sellar/suprasellar masses (34 patients, including 18 with CDI) that were difficult to differentiate from LINH in clinical practice. Notably, it is important to first exclude other possible inflammatory, neoplastic, and infectious diseases, including IgG4-related diseases or sarcoidosis, before determining lymphocytic hypophysitis as the diagnosis. In the present case, those diseases were excluded based on blood test and follow-up pituitary MRI results. As noted in our case, a pituitary biopsy is rarely performed because of high invasiveness, thus the probable diagnosis of LINH was based on criteria of presented by the Japan Endocrine Society [10]. Additionally, the presence of anti-rabphilin-3A antibodies supported the possibility of LINH.

Although the presence of anti-rabphilin-3A antibodies is a useful diagnostic marker for lymphocytic hypophysitis, the correlation of antibody titer with actual disease activity or stage remains poorly understood. Similarly, the correlation between intensity of anti-rabphilin-3A antibody reaction and stage or pathogenesis of CDI associated with COVID-19 infection is unclear, primarily due to lack of comparative data. Findings obtained in our previous foundational research using a mouse model suggest that autoreactive T cells targeting rabphilin-3A contribute to the pathology of LINH. However, other experimental findings have demonstrated that anti-rabphilin-3A antibody administration does not induce LINH, implying that these antibodies may not be directly involved in pathogenesis of the disease [11]. Consequently, while detection of anti-rabphilin-3A antibodies serves as a useful diagnostic tool, their presence does not necessarily reflect the active disease process of LINH. This uncertainty underscores the urgent need for further investigation to explore these relationships more deeply, potentially through use of longitudinal studies or enhanced comparative data collection.

The association between CDI and SARS-CoV-2 infection or vaccination has been reported. To the best of our knowledge, there were 14 cases of CDI after COVID-19 related events including the present case [5, 6, 12-22]. Eight were post-infection and six were post-vaccination cases. Clinical characteristics of each of those patients are shown in Table 3. Median age was 46 years (range 16–74 years), and there were five males and nine females. In general, there is no obvious difference between sex for morbidity rate associated with CDI [23], though the incidence of CDI in those 14 cases suspected to be related to COVID-19 was higher in females. The median duration to onset of CDI was 25 days (range 2–56 days). As for MRI findings, enlargement of the pituitary stalk and lack of high intensity in the posterior lobe, similar to LINH not related to COVID-19 [23], were noted. It is interesting that at least four of the 14 cases had no obvious abnormal MRI findings. In addition, there were few reports of suspected lymphocytic panhypophysitis related to COVID-19 (Cases 3 and 8). In the present case, hyperprolactinemia was noted, which might have been due to decreased supply of dopamine to the anterior pituitary gland as a result of physical compression caused by morphological changes of the pituitary stalk [23]. Only two cases received treatment with steroids; one with suspected central adrenal insufficiency (Case 3) and another complicated with bilateral optic neuritis (Case 13). Nearly all of the patients required continuous desmopressin therapy, except one (Case 11). None of the 14 patients was performed pituitary biopsy, thus there was no case with a histological diagnosis of LINH. The present case is notable, as findings indicating positive for anti-rabphilin-3A antibodies strongly supported the possibility of LINH diagnosis.

Table 3 Summary of reported cases with CDI suspected to be related to SARS-CoV-2

Case no. Age (yrs) Sex Infection or vaccination (dose) Days until symptoms appeared MRI findings Other pituitary hormones impaired Steroid therapy Continuous desmopressin therapy Reference
1 68 M Infection 33 days Normal N/A + [12]
2 28 M Infection 5 weeks Normal N/A + [13]
3 44 F Infection 24 days Normal ACTH↓ + + [14]
4 60 F Infection 8 weeks Enlargement of pituitary stalk, no high intensity in posterior lobe No + [15]
5 54 F Infection 6 weeks Normal No + [16]
6 54 F Vaccination (1st) 3 days Enlargement of pituitary stalk No + [17]
7 16 M Vaccination (1st) A few days Enlargement of pituitary stalk, no high intensity in posterior lobe No + [5]
8 48 F Vaccination (1st) 2 days Enlargement of pituitary stalk LH↓, FSH↓ + [18]
9 17 M Infection 3 weeks No high intensity in posterior lobe No + [19]
10 37 F Vaccination (2nd) 1 week No high intensity in posterior lobe No + [20]
11 35 M Infection 2 weeks N/A No Only
4 weeks
[21]
12 59 F Vaccination (1st) 8 weeks Enlargement of pituitary stalk, no high intensity in posterior lobe No + [22]
13 74 F Vaccination (4th) 1 month Swelling of pituitary gland, enlargement of pituitary stalk, no high intensity in posterior lobe No + + [6]
14 49 F Infection 3 weeks Enlargement of pituitary stalk, no high intensity in posterior lobe PRL↑ + Present case

M, male; F, female; N/A, not available

SARS-CoV-2 can alter the hypothalamic-pituitary axis as well as the central nervous system through direct or immune-mediated mechanisms [24]. Leow et al. were the first to report a possible etiologic role of SARS-associated coronavirus for causing hypophysitis or a direct hypothalamic effect [25]. That report also noted edema and neuronal degeneration along with the SARS-CoV genome identified in the hypothalamus in autopsy studies. In addition, expression of angiotensin-converting enzyme 2 (ACE2) receptors and transmembrane protease serine 2 (TMPRSS2) in several endocrine cells seems to have an important role in the direct pathogenetic mechanism by which the virus infects these organs [26]. A French group examined autopsy findings and showed that the hypothalamus is a highly probable target of SARS-CoV-2 based on its rich expression of ACE2 and TMPRSS2, especially in the paraventricular nucleus [27]. In another report, both hypothalamic and pituitary tissues express ACE2 in case of COVID-19, and could also be potential viral targets [28]. Melo et al. proposed that interference with ADH release into the posterior pituitary via effects on the subfornical organ and paraventricular hypothalamic nucleus lead to CDI [29]. On the other hand, some reports have noted CDI as a delayed complication [15, 16, 22], thus another hypothesis is that SARS-CoV-2 induces a delayed immune response in the central nervous system following the acute clinical phase of COVID-19 [30]. The presence of anti-rabphilin-3A antibodies in the present case suggests that the involved mechanism may have been related with such immune-mediated hypophysitis. Further studies of accumulated cases will be necessary to more fully elucidate the mechanisms of CDI development after SARS-CoV-2 infection.

 Conclusions

This is the first case report of a patient with a probable diagnosis of LINH after COVID-19 who tested positive for anti-rabphilin-3A antibodies. The same as in cases of LINH without SARS-CoV-2 infection, results indicating the presence of anti-rabphilin-3A antibodies could potentially serve as a valuable aid for diagnosing LINH associated with COVID-19.

 Informed Consent

Following our explanation, consent was obtained from the patient for serum collection, anti-rabphilin-3A antibody testing to aid in clinical diagnosis, and publication of this case report.

 Funding Source

None of the authors received any specific grant related to this report from funding agencies in the public, commercial, or not-for-profit-sectors.

 Disclosure

None of authors have conflicts of interest to declare. Y.I. and A.S. are members of Endocrine Journal’s Editorial Board.

Supplementary Fig. 1  Original western blot images used for Fig. 2
References
 
© The Japan Endocrine Society

This article is licensed under a Creative Commons [Attribution-NonCommercial-NoDerivatives 4.0 International] license.
https://creativecommons.org/licenses/by-nc-nd/4.0/
feedback
Top