Type 2 diabetes mellitus concurrent with CDI was less frequent in clinical practice. It always occurs in craniopharyngioma, pituitary metastases, and Wolfram syndrome[6], whereas, which were excluded for this patient. The etiology of CDI in this patient was mysterious. She experienced a series of pathophysiological disturbances, such as hemodynamic instability, pregnancy, and a tremendous increase in serum osmolality. During pregnancy, the appearance of placental vasopressinase may dramatically affect vasopressin physiology, with an increase in hormone clearance rates[7]. Gestational diabetes insipidus (GDI) may occur in cases of decreased AVP preservation or partial renal responsiveness to AVP. Symptoms usually resolve when vasopressinase activity declines and is typically undetectable at 4–6 weeks after delivery[8]. Nevertheless, for our patient, GDI could not be considered as her CDI had been sustained more than a half year since the abortion. Additionally, pregnancy can also trigger autoimmune disorders leading to lymphocytic hypophysitis specifically involved in the new onset of CDI[9]. For our patient, there were no findings of adenohypophysis insufficiency or enlargement of the anterior pituitary gland and stalk on MRI imaging. Moreover, there was still no evidence of other autoimmune disorders. Thus, lymphocytic hypophysitis might not afford the etiology of CDI in this patient. Lastly, this patient had undergone hypovolemic shock resulting from fluid loss, and further exacerbated hypotension secondary to a systemic inflammatory response of severe acute pancreatitis. Rarely, CDI has been reported in a patient suffering from severe hemorrhage [10]. We postulated that hypovolemic shock due to severe postpartum bleeding might cause a decreased cranial perfusion, contributing to impairment of any step of the entire process, from synthesis to transport to the storage of AVP in the hypothalamus, stalk, and posterior pituitary. Also, a partial CDI has been reported in an older diabetic patient with HHS in 1996[1]. Our patient presented with a persistent hypernatremia although severe HHS was successfully corrected by substantial fluid replacement. In brief, although the cause of CDI for our patient was still unclear, an impairment of AVP release contributed to the onset of DI, which should be kept in mind with prolonged hypernatremia in patients with HHS.
CDI is characterized by hypotonic polyuria due to impaired physiology of vasopressin, while also presenting hypernatremia and hyperosmolality if fluid loss was not adequately replaced[11]. Discriminating diabetes insipidus is sometimes challenging from primary polydipsia. A water deprivation test still be used for differential diagnosis of polyuria. For our patient, polyuria was not coupled with hypotonic urine with a maximum specific urine gravity of 1.018 when Dapagliflozin continued. Two days following cessation of dapagliflozin, hypotonic polyuria emerged and a water deprivation test was performed with a growing specific urine gravity from 1.005 to 1.010 and an increasing urine osmolality from 19mOsm/kg to 86mOsm/kg. In addition, although the patient was diagnosed with CDI and well-treated, there were still a few intriguing features. The development of CDI began with hypernatremia without a sense of thirst and then presented significant polydipsia of 10L/day. Apart from osmotic diuresis of poorly controlled diabetes, polydipsia and polyuria might be deteriorated by SGLT2 inhibitors, which can reduce glucose and sodium reabsorption in the proximal tubule of the kidneys. Similarly, a recent case report has described a man occurred adipsic diabetes insipidus following a traumatic brain injury[12]. He had a history of T2DM treated with glipizide and empagliflozin. The first desmopressin test was conducted on the day of empagliflozin cessation. Before desmopressin administration, a relatively high urine osmolality of 529mOsm/kg indicated that low urine osmolality might be obscured by empagliflozin. A partial response to desmopressin showed a sharp increase in urine osmolality but no corresponding decrease in urine output. Five days later, a second desmopressin test exhibited a significant reduction in urine volume. Moreover, Refardtet.al demonstrated empagliflozin increased urinary excretion in healthy volunteers with artificially induced Syndrome of Inappropriate Antidiuretic Hormone (SIADH) by administration of desmopressin[13]. Taken together, SGLT2 inhibitors might have an impact on urine osmolality and a slight resistance to the antidiuretic effect of desmopressin. Hence, SGLT2 inhibitors should be ceased when urine osmolality was evaluated to differentiate polyuria.
The inhibitor of SGLT2 simultaneously reduces sodium and glucose reabsorption in the proximal tubule. Either glycosuria or natriuresis can affect urine osmolality and lead to osmotic diuresis. Vasopressin acts on the V2 receptor (V2R) in the collecting duct and is mediated via water transport across aquaporin 2 (AQP2) channels[14]. Dysregulation of the vasopressin-V2R-AQP2 water channel axis has been implicated in either cranial diabetes insipidus or nephrogenic diabetes insipidus. Downregulation of AQP2 was observed in diabetic rat kidneys treated with a 12-week empagliflozin[15]. Under CDI conditions, the deficiency of vasopressin release was sufficient to produce the urine concentration defect in the kidney, which might be further exacerbated by SGLT2 inhibitors through the downregulation of AQP2. For our patient, the polydipsia markedly developing after dapagliflozin treatment further confirmed this postulation.
In summary, CDI should be considered when hypernatremia was sustained following the significant improvement of HHS in T2DM patients. It should be cautioned that SGLT2 inhibitors might affect the diagnosis and management of CDI in patients with T2DM. Whereas, a randomized control clinical trial should be further needed to survey how SGLT2 inhibitors affect patients with T2DM and CDI.
