Editor, Since the beginning of the outbreak of coronavirus disease 2019 (COVID 19) on December 2019, Italy has become one of the most affected countries in the world. At the time of preparing this article, on 11 May 2020, the WHO reported in Italy 219 070 confirmed cases, the highest number of COVID 19 cases in the world after USA, Spain, Russian Federation and the United Kingdom, with 30 560 deaths, the third total number after USA and the UK.1 In particular in Lombardy, the most affected Italian region, the survey of the Italian National Institute of Health reported on 3 April 2020, the maximum number of 1381 patients admitted in ICU and requiring invasive mechanical ventilation due to the associated severe acute respiratory syndrome coronavirus 2 (Sars-Cov-2).2 In addition to the viral pneumonia, COVID 19 can cause additional extrapulmonary manifestations. It has been reported to bind to the angiotensin-converting enzyme 2 (ACE2), a surface molecule localised on many extrapulmonary tissues including heart, kidney, intestine and immune cells, to enter the host cell, leading to a systemic disease.3 Due to the severity of acute respiratory distress syndrome (ARDS) with need of prolonged mechanical ventilation and high frequency of respiratory asynchronies, multimodal sedation with propofol, benzodiazepines and opioids with or without paralysis has been employed. Sevoflurane has been used for early sedation in paralysed ARDS patients with no renal adverse effects.4 Sevoflurane is potentially nephrotoxic during prolonged use for sedation, with polyuria and diabetes insipidus and while some case reports of ICU patients have been described, no data have been reported in patients affected with Sars-Cov-2. We hereby report two cases of Sars-Cov-2 complicated with an acute onset of polyuria and hypernatraemia, compatible with a diagnosis of nephrogenic diabetes insipidus (NDI) associated with sevoflurane delivery in addition to the intravenous sedation. Ethical approval for this study was provided by the Institutional Review board of our hospital (Comitato Etico Interaziendale Milano Area A, Milano, Italy) and written informed consent was obtained according to the Italian regulations. A 48-year-old man and 68-year-old woman with no significant medical history were admitted to our ICU with ARDS requiring mechanical ventilation. After 48 h of deep sedation and neuromuscular paralysis, intravenous sedation with propofol 3 mg kg−1 h−1, midazolam 0.05 mg kg−1 h−1 and remifentanil 0.2 μg kg−1 min−1 was administered due to the requirement of continued ventilatory support and increased agitation. After 2 to 3 days, the persistent agitation associated with respiratory asyncronies led to the introduction of sevoflurane for sedation using ventilators of the operating theatre along with continuous end tidal gas monitoring. The age-corrected minimum alveolar concentration was maintained between 0.3 and 0.9 according to the end tidal concentration of 0.5 to 1.1 vol% for a period of 9 days for the women and of 0.5 to 1.6 vol% for 8 days for the men. After 6 to 8 days of sevoflurane administration, a progressive increase in urine output, plasma sodium and plasma osmolality were observed without glycosuria suggesting the diagnosis of diabetes insipidus. Maximum values of plasma sodium and osmolality were reported after 1 to 3 days from sevoflurane discontinuation (157 mmol l−1 and 336 mOsm kg−1 for the women; 179 mmol l−1 and 386 mOsm kg−1 for the men) (Figs. 1) (Figs. 2).Fig. 1: Evolution of plasma and urine sodium concentration (upper panel), and of urinary output (lower panel) during ICU stay in a 68-year-old woman. Grey and black square areas refer to the period of administration of the indicated drug.Fig. 2: Evolution of plasma and urine sodium concentration (upper panel), and of urinary output (lower panel) during ICU stay in a 48-year-old man. Grey and black square areas refer to the period of administration of the indicated drug.Both patients developed acute renal failure with a maximum decrease in creatinine clearance to 30 ml min−1. Intravenous desmopressin was given (4 μg) without any effect on diuresis and urinary sodium confirming the diagnosis of NDI. After sevoflurane discontinuation, patients continued to have an inappropriate polyuria (with a maximum urine output of 7200 and 12 800 ml day−1, respectively) with hypernatraemia, managed with intravenous infusion of Ringer Lactate to compensate for renal losses together with free water through the nasogastric tube, parenteral nutrition along with intravenous FreAmine 10% to compensate for the extrarenal fluid losses. Subsequently, a combination of thiazide with a potassium sparing diuretic (50/5 mg respectively for a total daily dose of 150/15 mg for 3 days) to increase the urinary sodium excretion, and an inhibitor of prostaglandin synthesis (indomethacin with a maximum daily dose of 150 mg 24h−1 for 5 days) to reduce the urine output were given. During the following days a slow reduction in urine volume with a concomitant increase of urinary sodium and osmolality occurred (Figs. 1) (Figs. 2). In cases of hypernatraemia, hypotonic polyuria with hypovolaemia, the administration of a therapeutic dose of intravenous desmopressin can easily differentiate cranial diabetes insipidus from NDI, characterised by an inability to concentrate the urine due to failure of the distal nephron to response to antidiuretic hormone [arginine vasopressin (AVP)]. After a careful review of nephrotoxic medications and common causes of diabetes insipidus, the temporal sequence between the administration of sevoflurane and the beginning of hypotonic polyuria is a strong element to support the hypothesis of the role of sevoflurane in the NDI development. The mechanisms of potential nephrotoxicity of sevoflurane can be related to its biotransformation by the hepatic cytochrome P450 in inorganic fluoride that seems to produce a resistance to AVP in the ascending limbs of the loop of Henle, to the formation of compound A during low flow sevoflurane anaesthesia in presence of CO2 absorbers, that produces a transient swelling and/or necrosis in the tubular epithelium and to the impairment of the aquaporin-2 function, water channels regulated from AVP, localised in the renal collecting duct.5 Despite its common use in the operating room and more recently in ICUs only few cases of NDI during and after sevoflurane exposure are described. Using AnaConDa system, Cabibel et al.6 reported three cases of ICU patients exposed to sevoflurane for 13 days with sevoflurane end-tidal concentration of 1%. We administered sevoflurane for 9 and 10 days with similar values of end-tidal concentrations. Mechanically ventilated ARDS patients are at higher risk of acute kidney injury (AKI) because of inflammatory, neurohormonal and haemodynamic effects, in particular right heart failure, high intrathoracic pressure or volume overload, that may result in increased interstitial and tubular hydrostatic pressure. AKI is a relevant complication of patients affected with Sars-Cov-2 occurring in 2.9 to 23% of ICU patients.7 Pan et al.8 examined whether the cytokine inflammatory systemic response or a specific coronavirus cytopathic effect could be responsible. They concluded that the co-expression of the receptor ACE2 and transmembrane serine protease genes on podocytes and proximal straight tubule cells, make the kidney an important target organ for Coronavirus. The major limitations of this report are the absence of the measurements of plasma concentration of arginine vasopressin and the urine inorganic fluoride as a marker of sevoflurane related damage. Although the use of sevoflurane in ARDS patients has gradually emerged as an alternative to sedative agents, physicians should be aware that the use of sevoflurane with its nephrogenic effects, together with the ARDS related inflammatory and haemodynamic mechanisms on renal function could add to the renal viral attack making Sars-Cov-2 patients at higher risk for severe tubular dysfunction. In our cases, the use of the combination of thiazide with a potassium sparing diuretic was able to increase the urinary sodium excretion and the use of the inhibitor of prostaglandin synthesis (indomethacin) reduced the urine output. Acknowledgements relating to this article Assistance with the letter: none. Financial support and sponsorship: none. Conflicts of interest: none.