Background

Clozapine in an effective therapy for psychosis; however, its side effect profile limits its use to refractory presentations. Acute Interstitial Nephritis (AIN) is a rare adverse effect of clozapine, but has significant implications. Eleven cases of clozapine-induced AIN have been described in the literature. To our knowledge, this is the first case reported in Ireland. This case raises our awareness of this rare but significant complication of clozapine therapy. The risk issues in this case warranted deep consideration before clozapine was rechallenged and withdrawn.

Case presentation

This case of clozapine-induced AIN occurred in a 56-year-old married Caucasian lady who had a long-standing diagnosis of schizoaffective disorder. The nature of her illness was such that her symptoms were very florid in onset with violent harm to others and self-harm. She had a history of repeated prolonged inpatient admissions, including a protracted admission to a secure forensic unit. Prior to a trial of clozapine, this patient had multiple trials of typical and atypical antipsychotic medication over the years in oral and depot form at unlicensed high doses with minimal success.

Regarding medical history, this patient has a diagnosis of hypothyroidism, bilateral sensorineural hearing loss and a diagnosis of Chronic Obstructive Pulmonary Disease (COPD). Prior to admission in 2012, she drank three to four units of alcohol at weekends.

Regarding family history, she reportedly has two sisters with a history of mental illness; their exact diagnoses are unclear. She has a cousin who reportedly died by suicide. Her brother died of complications of alcohol dependence. There was no significant familial renal or cardiac history. With regards to early life, she achieved limited literacy and suffered significant childhood adversity. There was a prominent focus on religious ideologies from an early age, which would later be portrayed as a central theme in her relapses. She was recalled to a national forensic unit in 2012 after suffering a severe relapse of psychosis.

First trial of clozapine

October 2013: This patient presented with treatment-resistant psychotic symptoms and the treating team decided to commence clozapine following normal standard baseline investigations. Seven days later, the patient complained of feeling generally unwell with flu-like symptoms, a sore throat and cough. Eleven days after initiation of clozapine, she exhibited tachycardia of 127 beats per minute (BPM) and temperature of 37.4 Celsius (C). Her blood pressure (BP) was 126/74. Electrocardiogram (ECG) revealed a sinus tachycardia of 116 BPM. Her clozapine dose on this date was 125 milligrams (mg) mane and 150 mg nocte. Her full blood count (FBC) was normal. C-reactive protein (CRP) was raised at 26.6 mg/l (normal level 0–5 mg/l). Creatinine was slightly raised at 84 umol/l (normal level 44–80 umol/l) and remainder of her urea and electrolyte profile was normal. Her renal function was normal on subsequent renal tests until re-trial of clozapine 2 years later. There is no record of urinalysis. Clozapine was reduced by 125–150 mg once per day (OD). However, she became more unwell over subsequent days. Fifteen days after initiation of clozapine, vitals revealed a BP of 99/60, temperature of 38.7 C, respiratory rate of 25 and a persistent tachycardia of 110–130 BPM. CRP was raised further to 55.5 mg/l. Her FBC on this date indicated a normal FBC with a rise in eosinophils from 0.1 to 0.2 × 10⁹/l (normal range 0.04 – 0.4 × 10⁹/l). Clozapine was reduced further, and she was transferred to the general hospital where she was admitted for 14 days. Clozapine was discontinued on admission to the general hospital and replaced with olanzapine.

The diagnosis at this point was uncertain, and considered possibly to be a lower respiratory tract infection given her diagnosis of COPD. She was treated with the intravenous antibiotic tazocin for 12 days. However, her symptoms did not resolve as quickly as expected. CT thorax was performed because of the refractory nature of her respiratory symptoms and this showed an inflammatory infiltrate with no consolidation. It is possible that she was suffering clozapine-related parenchymal lung disease (hypersensitivity reaction). The patient expressed a preference not to undergo re-trial of clozapine on return to the forensic unit, and so she was commenced on fluphenazine depot. Her mental state remained quite fluctuant over the next 2 years.

Re-trial clozapine

August–September 2015: The patient suffered another major relapse of psychotic symptoms. Her behaviour became unpredictable, she was observed to be responding to command hallucinations and she described clear verbal urges to harm others. Her symptoms did not abate despite very high doses of typical and antipsychotics, and she required extended seclusion periods for the safety of herself and others. The circumstances around the previous cessation of clozapine were weighed up on balance with the high-risk refractory nature of her symptoms, and a decision was made for a re-trial of clozapine. Her concomitant medications at this time were sodium valproate chrono 1200 mg nocte, lanzoprozole 30 mg OD, levothyroxine 50 mcg OD, salmeterol 50 mcg b.d. (twice daily) and tiotropium 18 mcg OD.

October 2015: Two days after clozapine rechallenge, the patient suffered an episode of haemoptysis, tachycardia and chest pain. CTPA revealed no evidence of pulmonary embolism and showed minor ground glass in the right lower lobe, considered to possibly represent early infection. Urinalysis revealed a trace of blood, and routine bloods and ECG were normal. The haemoptysis did not persist.

November 2015: As the clozapine dose was being titrated upwards, the patient displayed signs of over-sedation and sialorrhea. Three weeks after recommencing clozapine on a dose of 50 mg mane and 125 mg nocte, unusually high concentrations of clozapine (1300 ng/ml) were detected in the serum and the dose was tapered down. A typically desired range for serum clozapine is 250–450 ng/ml. Routine blood samples were drawn due to a persistent tachycardia (100–120 BPM) showing deterioration in renal function. Amoxicillin/clavulanic acid 625 mg b.d. was commenced to treat possible respiratory infection given her COPD history. Four days later, the patient was subsequently transferred to the general hospital for further investigation and management. The prescribed dose of clozapine that had been tapered steadily was only 25 mg b.d. at the time of transfer. A standard dose of clozapine is typically 300–500 mg/day.

Investigations

Admission bloods revealed a creatinine of 193 umol/l (44–80 umol/l). Urea and electrolytes were normal. Estimated GFR was 24/ml/min/1.73 m². eGFR was biochemically consistent with CKD stage 4. Albumin was low at 23 g/l (35–40g/l) and liver function testes were normal. CRP was raised at 108 mg/l. FBC was normal apart from peripheral eosinophilia (0.9 × 10⁹/l), which raised suspicion of AIN. Renal biopsy confirmed AIN - interstitium revealed diffuse and marked inflammatory infiltrate, comprising mainly plasma cells (90%) with a few lymphocytes and rare eosinophils (see Fig. 1). There was no evidence of granulomas and blood vessels were normal. Congo Red staining for amyloid was negative. Antibody and complement screen were negative. Serum Protein Electrophoresis and Urine Protein Electrophoresis did not show evidence of a monoclonal band. Renal ultrasound demonstrated that both kidneys measured 13.3 cm in bipolar length, with mild prominence of both renal pelvises. Chest radiograph revealed borderline cardiomegaly with bibasal atelectasis.

Fig. 1. Renal biopsy image showing changes consistent with acute interstitial nephritis.

Differential diagnosis

Clozapine-induced AIN was the primary diagnosis, given the clear temporal association of clinical findings after clozapine initiation. AIN secondary to infection was less likely, given the lack of clear infective source, normal white cell count and neutrophils throughout admission. There was no evidence of rash, oral ulceration, photosensitivity or arthralgia; vasculitic screen was negative.

A differential included amoxicillin/clavulanic acid-induced AIN. However, this was less likely, given the previous administration of amoxicillin/clavulanic acid with no reaction and the clinical presentation began prior to addition of this antibiotic. Protein pump inhibitors and sodium valproate have also been implicated in AIN (Rossert, Reference Rossert2001), but both were long-standing medications prior to this presentation.

Treatment and outcome

The patient spent 4 days in the general hospital under the care of the renal team with input from the forensic and liaison psychiatric teams. Clozapine, lansoprazole and amoxicillin/clavulanic acid were discontinued. Ranitidine was added in place of lanzoprazole. Prednisolone 60 mg once a day was commenced with caution, and given her vulnerable mental state was tapered quickly. Renal function improved quickly after clozapine withdrawal, with no requirement for dialysis (see Fig. 2). Respiratory opinion was sought in relation to right lower lobe infiltrate on radiological investigations and single episode of haemoptysis, and formed the view that the inflammatory infiltrate was likely to be also secondary to clozapine therapy and advised against performing a bronchoscopy. Weekly monitoring of renal function was advised post discharge. She remained on high dose olanzapine of 20 mg b.d. and amisulpride 400 mg b.d. with stabilisation of psychotic symptoms. The patient has not regained normal renal function. Her eGFR remains stable at 50/ml/min/1.73 m². Her mental state has remained stable on a combination of high dose mood stabilisers, typical and atypical antipsychotics and other psychosocial interventions. She has been discharged from the National Forensic inpatient unit, remaining stable in the community with forensic rehabilitation and nephrology outpatient follow-up.

Fig. 2. Trend indicating renal function and CRP pre- and post-clozapine withdrawal.

Discussion and review of similar cases

Clozapine is an effective treatment for psychotic symptoms. However, its side effect profile, particularly hematological sequalae, limits its use to treatment resistant disorders (Walker et al. Reference Walker, Lanza, Arellano and Rothman1997). Renal reactions are rare and potentially fatal, but very treatable if caught in time.

AIN is characterised by a renal lesion caused by an inflammatory infiltrate in the kidney interstitium. It is most commonly caused by drugs (70–75%). Other causes of AIN include infections and systemic autoimmune disease (Raghavan & Eknoyan, Reference Raghavan and Eknoyan2014). Drug-induced AIN develops on average 3 weeks after exposure to the offending drug in 80% of cases and is not dose dependent (Rossert, Reference Rossert2001).

Our literature search revealed 11 cases of clozapine causing AIN (Elias et al. Reference Elias, Bannister, Clarkson, Faull and Faull1999; Fraser & Jibani, Reference Fraser and Jibani2000; Southall & Fernando, Reference Southall and Fernando2000; AU et al, Reference AU, Luthra and Stern2004; Siddiqui et al. Reference Siddiqui, Asim, Shamim, Pillai and Rajan2007; Hunter et al. Reference Hunter, Gaughan, Queirazza, McMillan and Shankie2009; Kanofsky et al. Reference Kanofsky, Woesner, Harris, Kelleher, Gittens and Jerschow2011; Mohan et al. Reference Mohan, Chua, Kartika, Bastiampillai and Dhillon2012; An et al, Reference An, Lee and Noh2013; Parekh et al. Reference Parekh, Fattah, Sahota and Colaco2014; Chan et al. Reference Chan, Cheung, Chan and Chau2015). Demographic details of patients varied. The age range of those affected ranged from 24 to 69 years. The mean age was 35.9 years. Six were male and five were female. With regards to time of onset after initiation of clozapine, the majority of reactions occurred within 2 weeks. One case reported a delay of 2 months. The doses of clozapine at which renal failure occurred varied from 25 mg to 700 mg. In terms of presentation, fever was reported in most cases. Gastrointestinal symptoms were common and no cases of rash were reported. Raised CRP, eosinophilia and proteinuria occurred in the majority of cases. Renal biopsy confirmed AIN in five of these cases. Biopsy, while definitive, is not essential in making the diagnosis of AIN (Perazella & Markowitz, Reference Perazella and Markowitz2010). In terms of clozapine re-challenge, one case was rechallenged with clozapine and renal failure recurred. In six cases, antibiotics were commenced to cover for presumed infection, despite absence of a clear infective source. Regarding treatment, three patients required haemodialysis. Steroid treatment was reported in five cases. In all cases, renal function recovered or improved after cessation of clozapine.

When this patient first commenced clozapine, she developed pyrexia and flu-like symptoms. Her chest radiograph showed an inflammatory infiltrate and she was treated for a presumed respiratory infection; however, her symptoms did not improve with antibiotics and her CT thorax was not characteristic of infection. It is possible that the diagnosis was clozapine-related parenchymal lung disease (hypersensitivity reaction). Bugge et al. (Reference Bugge, Nissen and Wynn2016) reported a case of probable clozapine-induced parenchymal lung disease, where the patient presented with general flu-like symptoms, non-productive cough, fever, raised inflammatory markers and diffuse parenchymal infiltration on chest radiograph with improvement after withdrawal of clozapine. This is very similar to the initial trial of clozapine in our case. It is unclear whether the patient in our case may have had a transient subclinical AIN episode during the initial introduction of clozapine. Of note, there is a higher likelihood of developing AIN during a re-challenge of a patient who has previously demonstrated drug hypersensitivity. The rapid recurrence of disease upon rechallenge in drug-related AIN is a clear manifestation of an immunological memory response and illustrates the immune-mediated nature of the reaction. This has been well described in the literature around drug-induced AIN generally (Baker & Pusey, Reference Baker and Pusey2004) and clozapine specifically (Hunter et al. Reference Hunter, Gaughan, Queirazza, McMillan and Shankie2009).

The elevated plasma clozapine concentration of 1300 ng/mL at the relatively low dose (175 mg) of clozapine is interesting. Plasma clozapine concentrations were not reported in our review of the literature of clozapine-induced AIN. Regarding other inflammatory organ reactions with clozapine, plasma levels were reported in one pancreatitis case and two hepatitis cases, with levels well above the therapeutic range, ranging from 0.831 to 0.910 mg/l (Lally et al. Reference Lally, al Kalbani, Krivoy, Murphy, Gaughran and MacCabe2018). These elevated plasma levels all occurred at standard clozapine doses of around 300 mg. It has been noted previously that in addition to overdose and drug reactions, infection/inflammatory reactions can precipitate clozapine toxicity (Lee et al. Reference Lee, White, Barr, Honer and Procyshyn2016; Clark et al. Reference Clark, Warren, Kim, Jankowiak, Schubert, Kisely, Forrester, Baune and Siskind2018). The relationship between clozapine levels, infection and/or inflammation is multifactorial and complex and could possibly be related to cytokine-mediated down-regulation of the cytochrome P450 system (Haack et al. Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul2003; Leung et al. Reference Leung, Nelson, Takala and Gören2014; Clark et al. Reference Clark, Warren, Kim, Jankowiak, Schubert, Kisely, Forrester, Baune and Siskind2018). Some have proposed that increases in clozapine concentration during infection/inflammation occur as a result of a rise in concentration of the acute phase protein α1-acid glycoprotein, to which approximately 95% of clozapine is bound; so, the rises occur in parallel (Espnes et al. Reference Espnes, Heimdal and Spigset2012; Lee et al. Reference Lee, White, Barr, Honer and Procyshyn2016).

It is notable in our case that the patient presented with over-sedation and hypersalivation (signs of clozapine toxicity) early in her treatment, at lower doses that would be expected for such side effects. Prior to diagnosis of AIN, she complained of chest pain, an isolated episode of haemoptysis and displayed mental state changes. CRP has been found to be elevated in patient with acute psychosis (Jacomb et al. Reference Jacomb, Stanton, Vasudevan, Powell, O’Donnell, Lenroot, Bruggemann, Balzan, Galletly, Liu, Weickert and Weickert2018). These findings all raise the possibility of an underlying inflammatory process mediating a presentation of clozapine toxicity. However, it is difficult to ascertain if the high serum clozapine level was the cause or the consequence of an inflammatory process in our case.

On re-trialling clozapine 2 years later, the patient presented with general flu-like symptoms and fever, tachycardia, eosinophilia and raised CRP. This patient had vague symptoms and her abnormal vital signs prompted a set of routine blood tests. If her renal function had not been checked, she is likely to have had a far worse outcome. The re-trial of clozapine was reasonable, given the unstable nature of her illness. The timing of onset of symptoms in the second trial was 4 weeks. It is possible that the addition of the antibiotic precipitated renal failure, a phenomenon observed previously (Kanofsky et al. Reference Kanofsky, Woesner, Harris, Kelleher, Gittens and Jerschow2011). AIN was confirmed in this case by the gold standard renal biopsy. Overall, her presentation correlates well with previously described cases.

This case and the extant literature raise the question of introducing the routine checking of renal function after commencing clozapine. Clozapine-induced myocarditis is another hypersensitivity reaction and some institutions implement a monitoring protocol for the first four weeks of starting clozapine, which involves checking CRP, FBC, Troponin and ECG weekly for 28 days (Ronaldson et al. Reference Ronaldson, Fitzgerald, Taylor, Topliss and McNeil2011). No such protocol has been implemented for clozapine-induced renal failure, given the low volume of cases reported to date. It is worth noting that tachycardia and eosinophilia are common after clozapine induction and usually benign. Eosinophilia has been well reported and studies report an incidence of transient eosinophilia to vary from 0.2% to 61.7% patients (Grover et al. Reference Grover, Mahajan, Sharma, Aneja and Chakrabarti2015). Clozapine-induced fever is also a common side effect, with a reported incidence of 0.5–55% (Martin & Williams, Reference Martin and Williams2013).

This patient never regained normal renal function. Given the severity of the reaction and level of initial impairment, it is likely that the patient has been left with some residual interstitial fibrosis and tubular atrophy resulting in the persistent reduction of the eGFR. Prognostic factors relating to drug-induced AIN vary, and the most valuable prognostic factor may be renal function a few weeks post diagnosis and/or the duration of acute renal failure (Rossert, Reference Rossert2001).

Take-home points

• This case and the associated literature raise our vigilance around the potential adverse reaction of AIN when prescribing clozapine, particularly in conjunction with other nephrotoxic agents.

• Do not presume persistent fever and eosinophilia are benign or due to infection. It is important to give consideration to a hypersensitivity reaction which may mimic infection, and it may be useful to check serum clozapine level in this context.

• Prescribing unwarranted antibiotics could precipitate AIN and lead to renal failure.

• If renal function deteriorates after clozapine initiation, prompt recognition and treatment in the appropriate centre is likely to preserve renal function and improve outcomes.