Central diabetes insipidus with anti-rabphilin-3A antibody positivity causing hypovolemic shock after resection of tumorous lesions in the pelvic cavity

Abstract

A 36-year-old female was pointed out to have liver enzyme elevation by routine health checkup. Subsequent contrast-enhanced CT scan identified gigantic uterine fibroids and retroperitoneal tumor. She was referred to the gynecologist at JA Toride Medical Center and planned to undergo a uterus enucleation and biopsy of the retroperitoneal tumor. The surgery was conducted without any troubles. After the surgery, the patient presented polyuria with urine volume 10–20 L a day and developed hypovolemic shock. Laboratory test revealed hypotonic urine and hypernatremia. Arginine vasopressin (AVP) loading test suggested shortage of endogenous vasopressin. Since the subcutaneous administration of AVP was not sufficient to control the urine volume, continuous intravenous infusion of AVP was initiated. After achieving hemodynamic stability, the treatment was switched to oral desmopressin. MRI finding indicated attenuation of high signal in posterior pituitary in T1 weighted image while neither enlargement of pituitary nor thickening of pituitary stalk was indicated by enhanced MRI. Hypertonic salt solution test indicated no responsive elevation of AVP, confirming the diagnosis of central diabetes insipidus (CDI). Her anterior pituitary function was preserved. Only anti-rabphilin-3A antibody was found positive in the serum of the patient, while other secondary causes for CDI were denied serologically and radiologically. Hence, lymphocytic infundibuloneurohypophysitis　(LINH) was suspected as the final diagnosis. Hormonal replacement therapy by nasal desmopressin was continued and the patient managed to control her urine volume. In cases of CDI considered idiopathic with conventional examinations, anti-rabphilin-3A antibody may be a clue for determining the cause as LINH.

Introduction

Diabetes insipidus (DI) is one of rare diseases, with a reported prevalence of 1:25,000 [1]. Central diabetes insipidus (CDI) is the most common form of DI, occurring in both the sexes equally and at any age. It is caused by impaired synthesis or release of AVP, often secondary to traumatic or surgical injury to the hypothalamus, posterior pituitary gland [2] and neurons originating in the supraoptic and paraventricular nuclei of the hypothalamus [1]. Meanwhile, up to 15% of cases are classified as idiopathic, although many of these are thought to have an autoimmune basis as a secondary cause [3], which clinically present as autoimmune hypophysitis. In lymphocytic infundibuloneurohypophysitis (LINH) or lymphocytic pan-hypophysitis (LPH), neurohypophysis is damaged by chronic inflammation to impair AVP secretion. The definitive diagnosis of lymphocytic hypophysitis requires pituitary biopsy, which is not easily available in clinical practice due to its invasiveness. Recently, rabphilin-3A was identified as a critical regulator of secretory vesicle trafficking including exocytosis in AVP neurons and T cells specific for rabphilin-3A may be involved in the pathogenesis of neurohypophysitis [4, 5]. Anti-rabphilin-3A antibody was shown to be a highly sensitive and specific diagnostic marker for LINH [6]. Herein, we present a female case, in which only positivity of anti-rabphilin-3A antibody was a clue for determining the actual cause of CDI.

Case report

A 36-year-old female was pointed out to have an elevation of liver enzymes at a routine health checkup. Contrast-enhanced computed tomography was performed by the first contact physician whereby liver nodular lesions, retroperitoneal tumor, gigantic uterine fibroids and ovarian cysts were identified (Fig. 1). Liver lesions were suspected of hemangioma on the basis of the contrast enhancement image. She was then referred to a gynecologist at JA Toride Medical Center, where magnetic resonance image (MRI) was done, which suggested suspected malignant neoplasm in the retroperitoneum cavity (Fig. 1). The patient was planned to undergo an enucleation of the gigantic uterine fibroids, resection of the ovarian cysts and biopsy to the retroperitoneal tumor. On the previous day before the surgery, the patient was asymptomatic and the physical examination did not indicate any notable findings except for abdominal distention due to the gigantic uterus fibroids. She had never been pregnant. Besides, she had no notable family history of specific illnesses. Her temperature was 37.0 ℃, heart rate was 89 beats/min and blood pressure was 121/71 mmHg. Laboratory test was normal except for relatively low value of blood urea nitrogen and creatinine (Table 1). At that time, the attending gynecologist was not aware of the patient’s history of polydipsia and polyuria, although the patient had presented them since almost 2 years before. After the surgery, the urine volume of the patient increased over 400 mL/hr. On the first postoperative day (POD 1), she developed hypovolemic shock. She was then referred to the nephrology department. Her blood pressure was 102/68 mmHg, and heart rate was 152 beats/min. Laboratory test revealed elevated serum sodium level of 157.3 mEq/L and urinary test showed hypotonic urine with its osmolality of 96 mOsm/kgH₂O. Five-percent glucose intravenous infusion was initiated immediately with its speed set to surpass urine volume to maintain patient’s volume status. Nevertheless, urine volume continuously increased, leading to hemodynamic instability. AVP loading test was subsequently performed, showing a positive response with increasing urine osmolality and decreasing urine volume after the subcutaneous injection of 5 unit of AVP (Table 2), indicating shortage of endogenous AVP. After the treatment with sufficient intravenous fluid infusion and subcutaneously administered AVP, the patient gradually recovered from shock. Since subcutaneous AVP injection was insufficient to control her urine volume, we switched the treatment to continuous intravenous injection on the POD 3. Over the next day, her urine volume gradually decreased, but she started presenting nausea and headache with a decreased serum sodium level of 124.7 mEq/L on the POD 5. Suspecting water intoxication, the intravenous administration of AVP was tapered off and the patient was added oral desmopressin instead. Again, her urine volume gradually started decreasing with amelioration of water intoxication (Fig. 3). On POD 6, pituitary MRI indicated attenuation of high signal in the posterior pituitary gland in T1 weighted image, compatible with central diabetes insipidus (Fig. 2). Hypertonic salt solution test with 5% sodium chloride solution showed minimal secretion of AVP in response to the elevation of serum sodium level (Table 3), confirming the diagnosis of CDI. The anterior pituitary function was preserved except for low TSH level and high ACTH/cortisol level (Table 4). This finding could be explained by postoperative stress response and reexamination showed normal values of each hormone.

Fig. 1.

Contrast-enhanced CT showed a gigantic uterine fibroids, b retroperitoneal tumor (arrow head) and ovarian cysts (arrow). Gadolinium-enhanced MRI showed irregular central enhancement effect in retroperitoneal tumor (arrow head)

Table 1.

Laboratory data on admission

Urianalysis			Biochemistry
specific gravity	1.003		Total protein	7.5	g/dL
Protein	( −)		Albumin	4.6	g/dL
Sugar	( −)		Urea nitrogen	5	mg/dL
Occult blood	( ±)		Creatinine	0.45	mg/dL
Urea nitrogen	77.8	mg/dL	Uric acid	8.6	mg/dL
Creatinine	16.31	mg/dL	Na	144.9	mEq/L
Na	21	mEq/L	K	4.06	mEq/L
K	14	mEq/L	Cl	111	mEq/L
Osmolality	96	mOsm/kgH2O	Ca	9.6	mg/dL
Blood cell count			Alkaline phosphatase	80	U/L
White blood cell	70.6	× 102 /μL	Asparate aminotransferase	40	U/L
Red blood cell	516	× 104 /μL	Alanine aminotransferase	62	U/L
Hemoglobin	14.7	g/dL	Lactate dehydrogenase	222	U/L
platelet	33.2	× 104 /μL	γ-Glutamyl transpeptidase	80	U/L
Coagulation			Amylase	63	U/L
Prothrombin time	12.7	sec	Total bilirubin	1	mg/dL
Prothrombin time (%)	109	%	blood sugar	106	mg/dL
Activated partial thromboplastin time	31.1	sec	C reactive protein	0.67	mg/dL
Fibrinogen	379	mg/dl
D-dimer	1.29	mg/dl

Table 2.

The result of AVP loading test

Time after AVP injection (min)	30	60	90	120
Urine volume (mL)	90	52	30	45
Urine osmolality (mOsm/kgH2O)	70	211	367	425

Fig. 3.

Detection of anti-rabphilin-3A antibodies by Western blotting. Recombinant full-length human rabphilin-3A expressed in HEK293FT cells (RPH3A + HEK293FT, left lanes) or negative control (HEK293FT, right lanes) were probed with serum from the present patient (Patient), or from a patient who was diagnosed with LINH previously (Positive control patient). The arrowhead indicates the presence of anti-rabphilin-3A antibodies. The dashed arrowhead indicates the absence of anti-rabphilin-3A antibodies

Fig. 2.

Sagittal (a) and Coronal (b) view of pituitary MRI. Findings of gadolinium-enhanced MRI (c, d). Pituitary gland is indicated by an arrow

Table 3.

The result of hypertonic salt solution test

Time after injection(min)	0	30	60	90	120
serum Na (mEq/L)	141	149	151	152	151
arginine vasopressin (pg/mL)	< 0.4	0.4	< 0.4	< 0.4	< 0.4

Table 4.

Pituitary function and screening test for CDI. Reexamined values are described in brackets

Pituitary function
Arginine vasopressin	< 0.4	pg/mL
Adrenocorticotropic hormone	82.4 [24.9]	pg/mL
Growth hormone	0.39	ng/mL
Prolactine	12.2	ng/mL
Luteinizing hormone	1.12	mU/mL
Follicle stimulating hormone	5.95	mU/mL
Thyroid stimulating hormone	0.251 [2.027]	μU/mL
Cortisol	23.6 [6.33]	pg/mL
Free T3 (triiodothyronine)	2.81	ng/dL
Free T4 (thyroxine)	1.24	ng/dL
Screening test for CDI
Angiotensin converting enzyme	8.3	U/L
lysozyme	2.6	μg/mL
IgG4	10.6	mg/dL
T-SPOT○r	( −)
Anti-nuclear antibody	< 40
Human chorionic gonadotropin	< 1.00	nU/mL
α-fetoprotein	4.6	ng/mL

Screening laboratory testing for secondary causes of CDI showed no notable findings as shown in Table 4. In addition, gadolinium-enhanced MRI showed no pituitary enlargement or intra/supra sellar mass and there was no significant enhancement of pituitary or thickening of pituitary stalk (Fig. 2). Since cases of LINH in chronic phase possibly lack any notable MRI findings, we further examined the anti-rabphilin-3A antibody in the patient’s serum by western blotting, which was positive, supporting the diagnosis of LINH (Fig. 3). Since immunosuppressive therapy has not been established in cases of LINH presenting CDI, only hormonal replacement therapy with desmopressin was continued. Desmopressin can be administered through both of oral and nasal routes. Considering its convenience and stability of blood concentration, oral administration was initially applied. Subsequently, a nasal route was added to gain a more rapid increase in serum AVP concentration. After urine volume stabilized, the oral administration of desmopressin was tapered on the POD 14. Then, liver injury probably caused by oral desmopressin developed on the POD 16 (Fig. 4). Hence, we discontinued oral route. Her liver function recovered by this alternate administration. Since her urine volume and electrolyte balance were well controlled with 10 µg of nasal desmopressin twice a day, the patient was discharged on the POD 28. The retroperitoneal tumor was diagnosed as shwannoma by the pathological examination and the patient has been followed up by scheduled checkups.

Fig. 4.

Clinical course. *AVP arginine vasopressin

Measurement of anti-rabphilin-3A antibody by western blotting

A vector containing the full-length human rabphilin 3A gene was transfected into HEK293FT cells to produce a recombinant human rabphilin-3A protein. As a control, the same vector but without the rabphilin 3A gene was transfected into HEK293FT cells. Anti-rabphilin-3A antibody in the serum were detected by Western blotting using the recombinant human rabphilin-3A protein lysate as the antigen and the serum as the primary antibody. A protein band presenting a size of 76 kDa appeared in the lysate of cells transfected with rabphilin-3A protein but not in that of control cells, which was considered to be positive for anti-rabphilin-3A antibody, as reported previously [3, 6].

Discussion

The present case presented polyuria after the surgery, leading to the diagnosis of CDI.

Screening tests for secondary causes of CDI were all negative except for anti-rabphilin-3A antibody, which ultimately diagnosed the patient as LINH.

CDI is caused by variable causes which include tumorous or invasive lesions of germinoma, craniopharyngiomas, IgG4-related disease, Langerhans cell histiocytosis, along with mechanical derangement by trauma and surgical procedures. Genetic impairment of AVP synthesis also causes CDI [7, 8]. However, up to 15% of CDI causes remain idiopathic [9–11]. In the present case, neither pituitary enlargement nor intra/supra sellar mass was identified by MRI, denying any of tumors related to CDI. In addition, the serological screening also denied IgG4-related disease, sarcoidosis, systemic lupus erythematosus, tuberculosis. With all possible secondary causes for CDI denied by the examinations, the present case was apparently considered as an idiopathic case. A large proportion of such cases may be explained by autoimmune hypophysitis, which is a chronic form of inflammation of the pituitary gland with varying degrees of pituitary dysfunction [7]. Depending on the infiltrated cell types, autoimmune hypophysitis is histologically classified into lymphocytic, granulomatous, xanthomatous, IgG4-plasmacytic, or necrotizing forms, among which more than 70% are lymphocytic [12]. Lymphocytic hypophysitis is also divided into 3 forms according to the localization of the hematopoietic infiltrate: lymphocytic adeno-hypophysitis (LAH), lymphocytic infundibuloneurohypophysitis (LINH), and lymphocytic pan-hypophysitis (LPH) [13]. LINH, in which lymphocytes infiltrate the neurohypophysis, causes CDI and often presents an enlargement of posterior pituitary or thickening of pituitary stalk, which can be detected by contrast-enhanced MRI [6]. We summarized the characteristics of the reported cases of LINH with CDI in Table 5 [3, 14–18]. It is notable that the present case was absent from any of the known features of MRI while all cases in Table 5 had some abnormal findings.

Table 5.

Summary of the cases of LINH with CDI

No	Ref. No	Age/Sex	Anti-rab3A antibody	Biopsy	MRI findings	Treatment	Outcome
1	3	27/F	+	−	Enlargement of the pituitary gland and stalk thickening	Desmopressin	N.A
2	3	69/F	+	−	Stalk thickening	Desmopressin	N.A
3	3	47/F	+	−	Stalk thickening	Desmopressin	N.A
4	3	35/M	+	−	Stalk thickening	Desmopressin	N.A
5	16	10/M	+	−	Enlargement of the pituitary gland and stalk thickening	Prednisolone Desmopressin	Improvement on the stalk swelling DI: not improved
6	17	68/M	−	−	Stalk thickening	Desmopressin	Not improved, progressed to PH
7	18	65/F	−	−	Nodular thickening of the neurohypophysis	Prednisolone Desmopressin	Cured 25 weeks after the onset
8	4	27/M	−	+	Nodular thickening of the stalk	Desmopressin	Not improved
9	4	38/F	−	+	Enlargement of the pituitary gland and stalk thickening	Desmopressin	Not improved
10	5	47/F	−	+	Enlargement of the pituitary gland and stalk thickening	Desmopressin	Not improved
11	5	55/M	−	+	Enlargement of the pituitary gland	Desmopressin	Not improved

Ref No reference number, rab rabphilin, M male, F female, N.A. not available, DI diabetes insipidus, PH pan-hypophysitis

Imura et al. [18] reported that 9 of 17 patients with idiopathic CDI showed stalk abnormalities with thickening and enlargement in MRI scans, which were observed only in patients who had CDI for less than 2 years. It is speculated that LINH in chronic phase might have reduced its inflammation to the extent no radiological abnormality could be detected. The present case was also considered to be in the chronic phase, which might explain the absence of the radiological findings.

Rabphilin-3A, which was identified as an effector protein of Rab3A, is expressed mainly in the brain [19]. Rab3A belongs to the small G protein superfamily and is thought to be a critical regulator of secretory vesicle trafficking including exocytosis [4]. Although the involvement of rabphilin-3A in the pathogenesis of LINH has not been fully elucidated, the immunization of mice with rabphilin-3A led to neurohypophysitis with lymphocytic infiltration in neurohypophysis. Additionally, there was a linear association between the number of T cells specific for rabphilin-3A in peripheral blood mononuclear cells and the number of CD3⁺ T cells infiltrating the neurohypophysis [5]. Hence, rabphilin-3A was suggested to be a pathogenic antigen for LINH.

Anti-rabphilin-3A antibodies were detected in the sera of patients with LINH (100% in patients with biopsy-proven LINH and 72% in patients with clinically diagnosed LINH). Iwama et al. reported a specificity of 100% in distinguishing sellar/suprasellar masses (34 patients including 18 CDI patients) that were difficult to differentiate from LINH in clinical practice [6]. However, one case with germinoma and one case with sarcoidosis were reported to be false-positive for anti-rabphilin-3A antibodies [3]. Even with these cases, anti-rabphilin-3A antibody was suggested to be useful as a highly sensitive and specific marker for diagnosing LINH.

In lymphocytic hypophysitis, immunosuppressive therapy, especially glucocorticoid was utilized for patients with rapid onset of neurological symptoms due to mass effects of pituitary enlargement, which was effective in reducing pituitary swelling and restoring pituitary function [20]. However, CDI itself was resistant to treatments as shown in Table 5 and also a negative prognostic factor in the response of lymphocytic hypophysitis to glucocorticoid therapy [21, 22]. Meanwhile, lymphocytic hypophysitis may be self-limiting and spontaneous remissions may occur in some cases [20]. Additionally, cases of LINH should be monitored for disease progression to pan-hypophysitis [23].

Finally, there were no case reports of CDI accompanied by retroperitoneal shwannoma, uterus fibroid or ovarian cyst. Meanwhile, the gynecological surgery presumably exacerbated the symptoms of CDI by elevating glucocorticoid level in the serum as shown in Table 4, leading to additional inhibition of AVP. Moreover, the loss of blood volume by the surgery without compensatory AVP secretion precipitated circulatory shock, together with increased urine volume.

Declarations

Conflict of interest

The authors have declared that no conflict of interest exists.

Human and animal rights

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from the patient described in the study.