A Case of Partial Nephrogenic Diabetes Insipidus Associated with Anidulafungin Treatment

Abstract

Nephrogenic Diabetes Insipidus (NDI) is classically associated with lithium, however it can be caused by a number of drugs including the anti-fungal agent amphotericin B. Here we describe a case of a 52-year-old male presenting with NDI following treatment with anidulafungin. The patient was admitted with a strangulated ventral hernia complicated by post-operative sepsis for which anidulafungin was initiated. Ten days after initiation of anidulafungin, his plasma sodium level rose to 167 mmol/L with serum and urine osmolality of 357 and 436 mmol/kg, respectively. Anidulafungin was discontinued and 5% dextrose in water (D5W) and vasopressin infusions were initiated. There was a partial response to the vasopressin infusion, suggestive of NDI. Sodium levels eventually returned to normal with continued vasopressin, and later intravenous desmopressin and D5W infusions. The patient received no other medications associated with NDI. Changes in sodium, and serum and urine osmolality, as well as delayed resolution of NDI fit with the fifty-hour half-life of anidulafungin. This case suggests that anidulafungin may be associated with NDI.

Keywords: Diabetes insipidus; Anidulafungin; Nephrogenic diabetes insipidus; Echinocandin; Antifungal

Abbreviations

ADH: Anti-Diuretic Hormone; NDI: Nephrogenic Diabetes Insipidus; CT: Computed Tomography; IV: Intravenous; PEA: Pulseless Electrical Activity; ICU: Intensive Care Unit; D5W: Dextrose 5% in Water; FDA: Food and Drug Administration

Introduction

The osmoregulatory system maintains plasma sodium within a narrow range (135-142 mmol/L) by controlling free water intake and loss [1]. The differential diagnosis for hypernatremia includes (i) unreplaced free water loss, (ii) severe metabolic stress (such as severe exercise or seizures) leading to water entry into cells, or (iii) sodium overload through administration of hypertonic solutions or salt poisoning [1]. Unreplaced free water loss can be caused by insensible or gastrointestinal losses, diabetes insipidus, or osmotic diuresis. Lack of access to free water or an impaired thirst mechanism can lead to hypernatremia. The physiologic response to hypernatremia with free water loss is to maximally concentrate urine. Diabetes insipidus is defined by an impaired ability to concentrate urine, resulting in polyuria with hypotonic urine and serum hypernatremia [2]. This occurs due to central diabetes insipidus, from an inability to produce ADH; NDI, due to inability of nephrons to respond to circulating ADH; or gestational diabetes insipidus, which is due to rapid metabolism of circulating ADH [3]. Response to exogenous ADH (desmopressin) is necessary to differentiate central from nephrogenic types [2].

Medications are the most common cause of acquired NDI in adult patients. Lithium impairs the ability to concentrate urine starting at eight weeks after initiation [4,5]. Other drugs associated with NDI include the anti-viral agents foscarnet [6] and cidofovir [7] and the anti-fungal agent amphotericin B [8]. Demeclocycline, an antibiotic in the tetracycline class, has been associated with NDI in several case reports [9].

Metabolic causes of NDI include hypercalcemia and hypokalemia. Hypercalcemia can cause calcium deposits in the distal tubules and collecting ducts, leading to decreased expression of aquaporin-2 channels [10]. Causes of chronic renal failure associated with NDI include sickle cell anemia, polycystic and medullary kidney disease [11-13], renal amyloidosis, and Sjogren’s syndrome [14]. Postobstructive and osmotic diuresis associated with NDI occurs due to down-regulation of aquaporin-2 channels [15].

Here we present a case of a patient who experienced partial NDI following treatment with the anti-fungal agent anidulafungin.

Case Presentation

A 52-year-old man presented to the emergency department with a three-day history of abdominal pain, obstipation, and fever. His medical history included obesity, recurrent leg ulcers, and a small bowel resection. His home medication was ibuprofen as needed for pain. Abdominal radiographs showed multiple air fluid levels, and subsequent CT scan with Intravenous (IV) contrast confirmed small bowel obstruction and a perforated ventral hernia. The patient was started on IV piperacillin-tazobactam, cefazolin, and lactated ringer’s IV fluid for hypotension. He underwent laparotomy and small bowel resection with hernia repair. He experienced new onset atrial fibrillation during surgery, resulting in a PEA arrest and needed resuscitation with epinephrine and norepinephrine. Atrial fibrillation was managed with metoprolol, amiodarone, and digoxin. A noncontrast CT head for decreased level of consciousness showed no intracranial abnormalities. He was taken to the ICU post-operatively and underwent a second look laparotomy with bowel anastomosis.

The patient had post-operative fevers and blood cultures grew Klebsiella oxytoca and abdominal fluid culture grew Escherichia coli. Piperacillin-tazobactam was switched to meropenem. Due to hemodynamic instability on day 6 of admission, empiric anidulafungin was added. He continued to receive his cardiac medications along with propofol, fentanyl, hydromorphone, midazolam, ondansetron, subcutaneous heparin, and chlorhexidine spray. The patient’s plasma sodium began to rise on day 7 of admission. The sodium peaked at 167mmo1/L on day 16 with concurrent serum and urine osmolality of 357 and 436mmo1/kg, respectively. The serum glucose, potassium, and corrected calcium at the time were normal at 6.4mmo1/L, 3.8mmo1/L, and 2.5mmo1/L, respectively. The renal function had not changed significantly with creatinine at 228umo1/L on the day of admission and 233umo1/L on day 16. Urea was 14.9mmo1/L on day 2 and 15.4mmol/L on day 26. Given the degree of hypernatremia and serum hyper osmolality, normal response to ADH was expected to increase urine osmolality above 500mmol/kg. Inability to fully concentrate urine in the setting of hypernatremia led to the diagnosis of diabetes insipidus.

Anidulafungin was stopped on day 16 due to possible contribution to diabetes insipidus. In addition, D5W and vasopressin infusions were initiated to target a decline in sodium of 10mmol/L per 24 hours. Despite increasing vasopressin to a maximum dose of 2.5 units/hr and administering D5W at 250 mL/hour on day 19 of admission, sodium correction was minimal to 161 mmol/L and urine osmolality only rose to 505 mmol/kg with serum osmolality of 360 mmol/kg. With hyperosmolality, hypernatremia, and vasopressin administration, urine osmolality was expected to be maximally concentrated above 800mmol/kg if ADH response was normal. Therefore, a diagnosis of partial NDI was made.

With ongoing vasopressin and D5W infusions and discontinuation of anidulafungin, his sodium levels returned to 144 mmol/L) by day 22 of admission. The patient was switched to IV desmopressin at 1 mcg twice daily on day 23, which was continued until day 26 to prevent recurrence of hypernatremia. The patient eventually made a full recovery from NDI with sodium level at 142mmol/L on day 27 without desmopressin (Figure 1). This case suggests that anidulafungin may be associated with partial NDI.

    

    

    Figure 1: Plasma sodium, urea, and creatinine levels before, during, and
after anidulafungin. Anidulafungin was initiated on day 6 and discontinued
on day 16 of admission. Vasopressin infusion was started on day 17 of
admission and discontinued on day 22. Desmopressin was administered
starting day 23 and discontinued on day 26 of admission.