Clozapine-induced nephritis and monitoring implications

To the Editor

Clozapine-induced interstitial nephritis has been increasingly recognised as a potential serious complication of clozapine therapy. A recent report (Kanofsky et al., 2011) has summarised eight cases of clozapine-induced acute interstitial nephritis (AIN). In all except one case onset of AIN occurred within 2 weeks of clozapine initiation. We report a delayed onset of AIN in a 53-year-old woman with chronic schizophrenia and discuss the clinical implications with regards to monitoring renal function.

Ms W had a 19-year history of schizophrenia characterised by treatment-resistant positive symptoms. Clozapine was started in May 2012, during an admission to hospital for an acute on chronic relapse of psychosis. Relevant baseline investigations were done as per protocol (Novartis, 2008) and were unremarkable. Ms W had commenced sodium valproate 3 weeks prior to starting clozapine.

Within a fortnight, clozapine was ceased for 12 days after Ms W experienced fever, vague abdominal pain and diarrhoea. Urinalysis revealed proteinuria and she received trimethoprim for a possible Escherichia coli infection. Eosinophilia was first noted on day 6 since clozapine initiation and persisted for 2 months (mean 0.93 ×10⁹/L). Ms W was recommenced on clozapine by slow titration when her abdominal symptoms subsided. The maximum clozapine dose achievable was 200 mg/day.

After 60 days of recommencing clozapine Ms W was referred to the medical team owing to hypotension (90/70 mmHg), pyrexia (38.6ºC), tachycardia (148 bpm), tremors and an unsteady gait. She was clinically suspected to have pneumonia but the chest X-ray was unremarkable. She was empirically treated with azithromycin 500 mg and ceftriaxone 1 g for 2 days. At the time of admission there was renal dysfunction, with creatinine 149 µmol/L (reference range 50–100 µmol/L) and urea 7.9 mmol/L (reference range 2.7–8 mmol/L). Septic screens were negative and autoimmune markers were inconclusive. The nephrology team suspected an acute kidney injury due to clozapine-induced AIN.

Clozapine and sodium valproate were ceased. Renal function continued to worsen, with elevated creatinine and urea (rising to 252 µmol/L and 15 mmol/L, respectively, 23 days after presentation). Apart from brief episodes of diarrhoea Ms W did not report any major physical symptoms. She underwent a renal biopsy for a definitive diagnosis. Histology confirmed the clinical suspicion of AIN, with interstitial scarring noted to be developing in 50% of the cortical area. She was given prednisolone 25 mg daily with rapid taper to avoid further deterioration. Ms W’s renal parameters slowly improved after the course of steroids. Aripiprazole 15 mg was commenced to prevent an exacerbation of psychosis. She was eventually transferred to a psychiatric unit when her renal parameters were near baseline (creatinine 135 µmol/L, urea 5.9 mmol/L).

Our patient developed AIN 90 days after clozapine was initiated. The diagnosis was confirmed by biopsy and other potential contributory factors were eliminated. The Naranjo scores were clozapine, 5; sodium valproate, 4; and azithromycin and ceftriaxone, 0; thus implicating clozapine as the probable cause for the AIN (Naranjo et al., 1981). Antibiotics administered with the presumption of infection could potentially worsen the course of AIN (Rossert, 2001).

Clozapine-induced AIN raises important issues for clinicians and on the current clozapine monitoring protocol. In our case, and in other reported cases of AIN, the diagnosis was not readily considered owing to a lack of pathognomonic markers to warrant early intervention. Eosinophilia or fever may be sensitive early warning signs of AIN (Kanofsky et al., 2011), but may not be specific markers given that > 50% of patients develop eosinophilia or fever within 6 weeks of clozapine initiation (Roge et al., 2012).

Recent literature suggests clozapine’s immunomodulatory effects play a role in clinically significant adverse effects such as benign hyperthermia, myocarditis, agranulocytosis, and eosinophilia (Roge et al., 2012). AIN possibly falls under this umbrella of immune-mediated effects, but has received less attention in psychiatric literature.

Existing protocols (Novartis, 2008) do not include monitoring for the possibility of AIN. In the absence of specific symptoms or signs that would raise clinical suspicion of AIN, and with the emerging literature that this adverse effect is not uncommon, we propose that renal parameters are incorporated into existing clozapine monitoring protocols for up to 6 months given the possibility of delayed AIN.