• acute renal failure
• nephrology

Sodium–glucose cotransporter-2 (SGLT-2) inhibitors are effective antihyperglycemic agents with proven cardiovascular and renal benefits.1 2 The risk for acute kidney injury (AKI) in patients on SGLT-2 inhibitors is considered to be relatively low. However, there have been postmarketing reports of episodes of AKI, some requiring hospitalisation and even dialysis in patients receiving SGLT-2 inhibitors.3 This phenomenon may be related to volume contraction and haemodynamic changes, particularly in patients with other risk factors (such as patients on renin–angiotensin–aldosterone system blocking agents or non-steroidal anti-inflammatory drugs, patients with chronic kidney disease).4 Here, we present a case of acute interstitial nephritis (AIN) and acute tubular necrosis likely related to a SGLT-2 inhibitor.

A 67-year-old woman was admitted with weakness, dizziness and abdominal pain. Her medical history was remarkable for hypertension, diabetes mellitus, ischaemic heart disease and peripheral vascular disease. Her medications included bisoprolol, losartan, amlodipine, sitagliptin and aspirin. Two months prior the current presentation, empagliflozin was prescribed by her primary physician for her diabetes mellitus. However, several days after initiation of empagliflozin she felt weak and dizzy and empagliflozin was discontinued. A week before the current presentation, empagliflozin was reinitiated by her cardiologist.

The patient reported no recent illness, fevers, rash, arthralgias, respiratory symptoms or bone pain. She denied exposure to other new medications (including NSAIDS and antibiotics).

On admission, blood pressure was 165/76 mm Hg. The rest of the physical examination was unremarkable. Laboratory studies revealed AKI (creatine 3.19 mg/dL, blood urea nitrogen 28 mg/dL, baseline creatine 0.9 mg/dL). Complete blood count demonstrated no eosinophilia or thrombocytopenia.

Urinalysis showed a few leukocytes but no red cells or casts. The urine protein to creatine ratio was 5160 mg of protein per gram creatine. Abdominal ultrasonography showed normal-size kidneys and no hydronephrosis.

Immunologic and infectious serologies were unremarkable.

Over several days, the patient became oligoanuric, creatine peaked to 9.22 mg/dL and haemodialysis was initiated. Empiric treatment with prednisone was given and a renal biopsy was performed. Four glomeruli were seen on light microscopy. The glomeruli were normocellular. An interstitial infiltrate of lymphocytes and small numbers of eosinophils were seen. Thinning of the renal tubular brush border with intratubular necrotic content was also seen. Immunofluorescence was negative. These findings were compatible with AIN and acute tubular necrosis (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Kidney biopsy. Interstitial oedema with lymphocytic and eosinophilic infiltrate. Dilated tubules with thinning of the brush border containing necrotic content were also demonstrated.

The patient was discharged with a regimen of intermittent haemodialysis and steroid therapy. Three months later, urine output and kidney function improved and the patient was weaned off dialysis.

In this case report, we describe previously unreported histologically proven AIN with evidence to suggest a relationship to empagliflozin exposure. Although SGLT-2 inhibitors have been reported to be associated with AKI, it has been supposed to be related to volume contraction and haemodynamic changes.4 In our case, in addition to interstitial infiltrate, signs of acute tubular injury were seen on kidney biopsy. As our patient received an angiotensin receptor blocker together with an SGLT-2 inhibitor, this might have had a synergistic effect on renal perfusion; it appears that an element of drug-related perfusion also contributed to AKI in our case.

Furthermore, as only a small number of glomeruli were seen on light microscopy, we cannot completely rule out additional glomerular pathology, particularly as our patient had proteinuria within the nephrotic range.

As virtually any drug can cause AIN,5 further evidence of potential drug causation in our case was an onset of renal dysfunction with reinstitution of empagliflozin, biopsy-proven AIN and absence of other drugs, infection or systemic disease that might cause AIN. With the dramatic increase in use of SGLT-2 inhibitors worldwide, treating physicians should be aware of AIN as a possible cause of AKI in this context.