Case Presentation: We present a case of a 50-year-old male with a history of developmental disability and hypothyroidism who presented from a nursing home with fatigue and cough for three days. The patient was found to have a left upper lobe consolidation on chest x-ray consistent with pneumonia. Labs revealed neutrophilic leukocytosis. He was started on vancomycin and piperacillin-tazobactam for healthcare-associated pneumonia (HCAP). His course was complicated by acute hypernatremia to 166 mmol/L and a creatinine of 1.97 mg/dL, reaching a peak of 2.41, from a baseline of 0.6. Further work-up revealed a serum osmolality of 335 mosm/kg and urine osmolality (Uosm) of 174 mosm/kg, raising suspicion for diabetes insipidus (DI). He also had a fractional excretion of sodium (FENa) of 2.6%, eosinophiluria, peripheral eosinophilia, and polyuria (> 3 liters/day). Piperacillin-tazobactam was discontinued due to concern for acute interstitial nephritis (AIN). His serum Na remained elevated despite aggressive parenteral rehydration, calculating and replacing his free water deficit daily. A water deprivation test was performed, during which time Uosm remained below 300 mosm/kg. He was started on D5W and desmopressin 2 mcg daily with partial improvement in sodium and urine osmolality, consistent with partial nephrogenic diabetes insipidus (NDI). Serum creatinine improved to 1.29 after discontinuation of piperacillin-tazobactam.
Discussion: Piperacillin-tazobactam has been well-established as a cause of drug-induced AIN. The diagnosis can be made definitively with kidney biopsy, although most cases are diagnosed clinically. Timing of onset as related to initiation of the suspected offending agent is important in diagnosis. In our patient, peripheral eosinophilia, eosinophiluria, acute renal failure, and initiation of piperacillin-tazobactam with subsequent improvement after discontinuation point to a diagnosis of AIN. NDI is a condition in which polyuria occurs due to lack of response to ADH at the level of the kidney. NDI can be primary (genetic or congenital) or acquired, most commonly due to lithium use. Other acquired causes include hypercalcemia, hypercalciuria, and obstructive uropathy, none of which were present in our patient. NDI can be completeor partial (as in this case). Given the time course of symptoms and laboratory derangements, we propose that piperacillin-tazobactam was the offending agent causing AIN, which subsequently led to NDI. AIN associated with NDI has not been previously reported.
Conclusions: Broad-spectrum antibiotic use for suspected HCAP is being called into question. Antibiotic stewardship programs are being deployed to reduce overutilization. This unique, rare case of piperacillin-tazobactam causing AIN and also NDI introduces another reason to de-escalate antibiotics and limit use as feasible. Hypernatremia without another apparent cause should raise suspicion of NDI in a patient on piperacillin-tazobactam.