Diagnostic algorithm for clozapine-induced myocarditis: A systematic review

Abstract

BACKGROUND

Clozapine, the gold standard for resistant schizophrenia, is underused due to risks like clozapine-induced myocarditis (CIM). Non-specific biomarkers and inconsistent imaging, and the significant overlap with clozapine-induced pneumonia (CIP) lead to misdiagnosis and premature discontinuation.

AIM

To develop a diagnostic algorithm for CIM to enhance accuracy, differentiate from CIP, and guide safe clozapine rechallenge.

METHODS

A systematic review of 119 PubMed studies (published between 1990 and April 2025) was conducted in accordance with PRISMA guidelines. The review analyzed CIM diagnosis and rechallenge outcomes, with a focus on biomarkers, imaging, and collaboration with cardiology.

RESULTS

CIM diagnosis relies on troponin and C-reactive protein; electrocardiography and echocardiography are inconsistently applied, and cardiac magnetic resonance imaging (CMR) is underused. Rechallenge was successful in 64.7% to 68.9% of 136 cases, with 2.9% resulting in fatal outcomes. Up to 65% of presumed CIM cases lack confirmation. A proposed protocol integrates chest computed tomography to exclude pneumonia and CMR for CIM confirmation, with echocardiography as an alternative.

CONCLUSION

A protocol involving multidisciplinary collaboration among computed tomography, CMR, and cardiology improves CIM diagnosis. Slow titration prevents CIM; adjust the dose for CIP and discontinue for confirmed CIM.

Key Words: Clozapine; Myocarditis; Pneumonia; Rechallenge; Protocol; Multidisciplinary management

Core Tip: This systematic review proposes a multidisciplinary diagnostic algorithm for clozapine-induced myocarditis to reduce misdiagnosis with clozapine-induced pneumonia. Integrating chest computed tomography to exclude clozapine-induced pneumonia, cardiac magnetic resonance imaging for clozapine-induced myocarditis confirmation, and echocardiography in resource-limited settings, the protocol emphasizes slower clozapine titration and collaboration with cardiology to enhance diagnostic accuracy and ensure the safe use of clozapine in the treatment of treatment-resistant schizophrenia.

INTRODUCTION

Clozapine is unparalleled in reducing hospitalization and suicidality in treatment-resistant schizophrenia[1]. However, its use is limited by rare but potentially fatal adverse effects, including clozapine-induced myocarditis (CIM) and clozapine-induced pneumonia (CIP)[2,3]. These conditions pose diagnostic challenges due to overlapping symptoms (e.g., fever, dyspnea) and elevated biomarkers [e.g., troponin, C-reactive protein (CRP)][4]. Current CIM screening protocols primarily rely on clinical suspicion and nonspecific biomarkers, with inconsistent application of electrocardiography (EKG), echocardiography, or advanced diagnostics, such as cardiac magnetic resonance imaging (CMR) or biopsy. Exclusion of overlapping conditions like CIP and cardiologist consultation are rarely standardized, and guidelines often recommend discontinuing clozapine at the first suspicion of CIM, risking premature abandonment of an effective therapy. The absence of uniform diagnostic criteria for CIM and its overlap with CIP heighten the risk of misdiagnosis. Given CIM’s life-threatening potential, refining diagnostic approaches is critical to balance patient safety with clozapine’s therapeutic benefits.

This systematic review evaluates existing clinical information management screening and diagnostic guidelines, assesses clozapine rechallenge outcomes following presumed myocarditis, examines gaps in differentiating CIM from clozapine-induced psychosis (CIP), and proposes a structured protocol integrating advanced imaging and multidisciplinary care to improve diagnostic accuracy and reduce unnecessary clozapine discontinuation. This review aims to develop a diagnostic algorithm for CIM that minimizes misdiagnosis with CIP and informs safe clozapine rechallenge.

MATERIALS AND METHODS

Search strategy

This work is a systematic review, not a clinical trial. It synthesizes existing literature on CIM diagnosis and management rather than prospectively assigning human subjects to intervention or comparison groups to explore cause-and-effect relationships.

Study selection and eligibility

We conducted a systematic review of literature indexed in PubMed from 1990 to April 2025, using the search string: “clozapine AND (myocarditis OR cardio* OR pericarditis) AND (evaluation OR algorithm OR guidelines OR monitoring)”. This timeframe encompasses early CIM reports, which reflect the evolving diagnostic and management approaches. Studies were eligible if they were clinical investigations, diagnostic guidelines, or high-quality case series addressing the diagnosis of CIM or the outcomes of clozapine rechallenge. Single or low-quality case reports, animal studies, and non-English publications were excluded to prioritize robust, generalizable evidence. Two independent reviewers screened titles and abstracts, followed by full-text evaluation, resolving discrepancies through consensus.

Data extraction and synthesis

The review adhered to the PRISMA framework to ensure a structured and transparent process. The PRISMA checklist guided the formulation of the research question (“How can CIM be accurately diagnosed and differentiated from CIP in clozapine-treated patients?”), the development of the search strategy, study selection, data extraction, and synthesis. PRISMA principles also informed workflow development, including the use of structured questions (e.g., Population: Clozapine-treated patients; intervention: Diagnostic protocols; comparison: Standard vs advanced imaging; outcome: Diagnostic accuracy, rechallenge success) and process validation across reviewers. Figure 1 presents the PRISMA flowchart detailing the study selection process.

Figure 1 PRISMA flowchart showing a selection of 119 studies from 365 initial records, with detailed exclusion criteria. CIM: Clozapine-Induced Myocarditis.

Data extraction focused on diagnostic protocols and rechallenge outcomes, which were synthesized and compared across studies (Tables 1 and 2), allowing for a critical assessment of current practices and identifying evidence gaps. To address the specificity of biomarker thresholds for CIM, we extracted troponin and CRP cut-offs from CIM-focused studies (e.g., Ronaldson et al[5]; Tirupati et al[6]) where available. Where CIM-specific thresholds were not reported, we adopted conservative thresholds (troponin > 0.04 ng/mL, CRP > 10 mg/L) from the literature on viral myocarditis, acknowledging potential differences in clozapine-related inflammation. These thresholds were selected to prioritize sensitivity for early detection, with the understanding that imaging [chest computed tomography (CT), CMR, or echocardiography] is required for confirmation and verification. We also assessed studies for dynamic biomarker trends (e.g., rising troponin over 24-48 hours) to enhance diagnostic precision.

Table 1 Summarizes studies and guidelines for screening and diagnosing clozapine-induced myocarditis.

Ref.	Study type	Sample size	Monitoring protocol	Biomarkers used	Imaging techniques	Outcomes reported	NOS scores	Certainty of evidence	Strengths	Limitations
Berk et al[45], 2007	Guidelines	N/A	Baseline: Clinical evaluation, EKG, plasma troponin, CK-MB, TEE-days 7 and 14: Clinical evaluation, EKG, plasma troponin, CK-MB-6 months and annually: TEE	Troponin, CK-MB (routine); WBC, eosinophil count, ESR, CRP (suspected cases)	Echo (routine)	No outcome reported for guidelines	6/9	Low: Expert-driven with no supporting data	First suggested protocol	Based on expert opinion and clinical consensus, with no data supporting the guidelines
Ronaldson et al[5], 2010	Study	75	Baseline: Echo, troponin I or T, CRP-weekly for 4 weeks: Troponin I or T, CRP	Troponin and CRP	Echo (baseline), EKG (suspected cases)	Reduced mortality and early detection	7/9	Moderate: Strong evidence but a moderate sample	Strong evidence, comprehensive	Moderate sample size, limited follow-up, limited imaging role, and lack of guidelines for overlapping conditions
Murch et al[52], 2013	Study	122	Baseline Echo and repeated Echo at 3 and 6 months	CBC and CRP	Echo	73% had the Echo before starting clozapine 65% had one follow-up echo	4/9	Low: Small study with high bias risk	Examine the role and benefit of echo in patients with myocarditis	Limited utility, high cost of using echo for screening, and lack of guidelines for overlapping conditions
			Symptomatic patients had CBC, CRP			3 patients screened positive, 2/3 had findings in the echo suggesting myocarditis
Youssef et al[55], 2016	Retrospective study	129	No protocol. They had criteria for myocarditis based on symptoms and one of the following (elevated troponin, EKG changes, or echo changes)	Troponin	Echo and EKG were inconsistently assessed	3.88% met the diagnosis of myocarditis	6/9	Moderate: Large sample but retrospective	Large sample	Inconsistent use of imaging and lack of exclusion of overlapping conditions
McNutt et al[53], 2021	Study	38	Baseline and weekly for 4 weeks: Troponin I or T, CRP	Troponin, CRP	Cardiology involved 2 cases had echo/CMR 2 other patients only had elevated CRP	50% of patients who screened positive were confirmed to have myocarditis	4/9	Low: Small study, inconsistent imaging	Cardiologist for suspected cases	Small study, inconsistent use of images, and lack of guidelines for overlapping conditions
Anıl Yağcıoğlu et al[46], 2019	Study	38	Baseline and weekly for 4 weeks: Clinical evaluation, troponin, CRP, ESR, eosinophil count	Troponin, CRP, ESR, and eosinophil count	Echo (suspected cases), CMR (1 patient)	11.3% suspected myocarditis, 1.4% confirmed	6/9	Low: Small study	Echo for suspected cases	Small study, no clear management consensus
Nachmani Major et al[48], 2020	Retrospective study	24	Routine: Clinical symptoms, WBC, troponin, CRP, BNP- Suspected cases: Clinical assessment, EKG, echo, MRI	Troponin, CRP, WBC, BNP	EKG, echo, MRI (suspected cases)	8.6% suspected myocarditis, 1.4% confirmed with imaging	5/9	Low: Small retrospective study	Moderate sample size	Retrospective study and lacks guidelines for overlapping conditions
Kanniah et al[51], 2020	Guidelines	N/A	Baseline and weekly for 4 weeks: Vital signs, CRP, ESR, CPK, troponin-after 4 weeks: Eosinophil count monitoring	CRP, ESR, CPK, troponin, eosinophil count	EKG (symptom-driven)	No outcome reported for guidelines	5/9	Low: Screening-focused, expert-driven	Focused on screening	There are no work-up recommendations for suspected cases or guidelines for overlapping conditions
Sandarsh et al[47], 2021	Study	100	Baseline and weekly for 4 weeks: Clinical symptoms, CRP, eosinophil count, CPK	CRP, eosinophil count, CPK, troponin I	EKG, echo (2/6 positive cases)	5.3% suspected myocarditis after initial exposure, 3.5% after restarting	6/9	Moderate: Larger study	Larger study	Screening-focused, no further workup guidance
Segev et al[54], 2021	Retrospective study	228	Undetermined	CRP and troponin	Echo/MRI for suspected cases	Myocarditis was ruled out in 78.7%, confirmed in 9.8%, and undetermined in 11.5%	6/9	Moderate: Large study with cardiology input	A large study involved cardiology for further workup	Retrospective study, studied other antipsychotics other than clozapine, no workup for overlapping conditions
de Leon et al[56], 2022	Guidelines	N/A	Baseline and weekly for 4 weeks: CBC, CRP; troponin optional in resource-rich settings	WBC, CRP, troponin (optional)	None specified	No outcome reported for guidelines	NA	Low: International guidelines, no outcomes	International guidelines	Focused on screening with no workup for presumptive cases or workup to exclude overlapping conditions
NSW Health[50], 2022	Guidelines	N/A	Baseline: Clinical symptoms, CRP, CPK, troponin, EKG, echo-weekly for 6 weeks: Troponin, CRP - every 6 months: EKG - every 12 months: Echo	CRP, CPK, troponin	EKG (routine), echo(routine), MRI (suspected cases)	No outcome reported for guidelines	NA	Low: Expert-driven, no outcomes	Clear workup for suspected cases	No outcome data
Tirupati et al[6], 2024	Retrospective study	327	Baseline and weekly for 6 weeks: Troponin I or T, CRP	Troponin, CRP	EKG, echo (most positive cases)	9.8% incidence of clozapine-induced myocarditis	7/9	Moderate: Large study, comprehensive	Large study, comprehensive follow-up	Retrospective, no exclusion of overlapping conditions
Griffin et al[49], 2021	Expert opinion and case series	NA	Baseline and weekly for 8 weeks: Troponin, CRP, BNP/NP-pro BNP and EKG	CRP, BNP/pro-BNP	Consult cardiology for suspected cases. TEE, CMR, and LVEF should be assessed in 6 months, and based on the results, clozapine can be reinstated	No outcome reported for guidelines	7/9	Low: Expert opinion, no outcomes	Expert opinion. Includes more workup for presumptive cases	The suggested protocol does not have data to support it and does not include a workup for overlapping presentations

NOS: Newcastle-Ottawa Scale scores; N/A: Not applicable; NA: Not available; EKG: Electrocardiogram; TEE: Transesophageal echocardiography; CK-MB: Creatinine kinase; WBC: White blood count; ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein; CBC: Complete blood count; CMR: Cardiac magnetic resonance; BNP: B-type natriuretic peptide; MRI: Magnetic resonance imaging; CPK: Creatinine phosphokinase; LVEF: Left ventricular ejection fraction.

Table 2 Summarizes studies on clozapine rechallenges after presumptive diagnosis of myocarditis.

Ref.	Study type	Sample size	Rechallenge success rate	Fatal failure	Diagnostic certainty	Confirmatory test used	Key findings	Limitations
Richardson et al[57], 2021	Systematic review	88 cases from 88 case reports/case series	64.7%	2.9%	Low (4.5% biopsy-confirmed)	Biopsy (4.5%) No CMR	High failure rate, rare fatalities	Based on the case report, we rely on biomarkers for diagnosis, with no exclusion of overlapping conditions
McMahon et al[58], 2024	Systematic review	45 cases	68.9%	0%	Moderate (EKG/TEE in < 33%)	EKG, TEE (< 33%)	No fatalities, 1/3 failed	Based on case reports
						No CMR		Most of the studies diagnosed myocarditis based on biomarkers alone
								With no workup overlapping conditions
Noël et al[59], 2019	Case series	3 cases	33.3%	0%	High (EKG/TEE evidence)	EKG, TEE	Success is only in mild cases	Case series

CMR: Cardiac magnetic resonance; EKG: Electrocardiogram; TEE: Traneosophageal echocardiography.

Certainty assessment

Evidence certainty for CIM incidence, diagnostic accuracy, and rechallenge outcomes was evaluated narratively, considering study design (retrospective vs. prospective), sample size, diagnostic method consistency (e.g., CMR use), and risk of bias. Evidence was rated as high, moderate, low, or very low, reflecting the level of confidence in clinical reliability.

Risk of bias assessment

Risk of bias in included studies was assessed using a modified Newcastle-Ottawa Scale for observational studies and a checklist for guidelines. Newcastle-Ottawa Scale scores for each included study are now included in Table 1 for transparency. To assess reporting bias, we searched conference abstracts from the American Psychiatric Association Annual Meetings and European Society of Cardiology (ESC) Congresses (1990 to April 2025) via available online archives (e.g., American Journal of Psychiatry supplements, European Society of Cardiology 366 ESC 365) and manually reviewed reference lists for unpublished reports.

RESULTS

Of the 365 initial articles, 119 met the inclusion criteria (Figure 1), excluding single case reports. Fourteen key protocols and studies on CIM screening and diagnosis were identified (12 via PubMed and two from reference lists), and three studies met the criteria for clozapine rechallenge.

Bias varied

Retrospective studies carried a moderate risk, case series had a high risk, and guidelines had a low risk, although they were expert-driven (e.g., New South Wales Health guidelines). Rechallenge success (64.7%-68.9%) evidence had low certainty due to small sample sizes, inconsistent diagnostics, and rare fatal failures (2.9%). CIM incidence evidence (0.06%-3.88%) had moderate certainty, supported by large cohorts but weakened by diagnostic variability.

Diagnostic protocols/studies

Troponin and CRP were the most commonly used biomarkers across 14 protocols or studies that utilized various imaging techniques. Suspected CIM ranged from 5.3 to 11.3% (median 8.6%), and confirmed cases ranged from 1.4% to 9.8% (median 3.5%). Moderate certainty is reflected in large cohorts (e.g., n = 327, 228) but is undermined by variable diagnostics and the lack of CIP exclusion (Table 1).

Rechallenge outcomes

Three studies reported rechallenge success rates of 64.7%-68.9%, with rare cases of fatal failure. Inconsistent use of cardiologists, imaging, and confirmatory tests raises safety concerns (Table 2).

DISCUSSION

CIM

CIM was first reported in 1980 during post-marketing surveillance, with a suspected overdose case[7]. Subsequent reports linked myocarditis to overdose[8] and rapid titration[9], while an Australian case series (n = 23) solidified the association[2]. Several hypotheses have been proposed to explain the mechanism of CIM. One suggested that a type 1 immunoglobulin E-mediated hypersensitivity reaction is involved, supported by eosinophilia in two-thirds of patients[5,10]. Other reports indicate that clozapine causes an increase in both epinephrine and norepinephrine[11]. This increase in catecholamines has been found to worsen myocarditis in animal studies[12]. Another possible mechanism is associated with oxidative stress, downregulation of antioxidants, and apoptosis[13].

Several risk factors have been identified, including genetic susceptibility, advanced age, and rapid titration of clozapine doses[14]. Additionally, the concurrent use of sodium valproate, a common augmenting strategy for augmentation with clozapine, is suggested to increase the risk of CIM[15]. Myocarditis typically presents within the first two to eight weeks of clozapine initiation, with the highest risk occurring in the first 30 days. However, delayed myocarditis can occur years after initiating clozapine[14,16].

The reported rates of CIM varied significantly, ranging from as low as 0.06% to as high as 3.88%[14]. This variability may stem from challenges in diagnosing CIM, as the clinical symptoms often overlap with those of other conditions, such as pneumonia, pulmonary embolism, and other cardiopulmonary diseases. Additionally, the diagnostic methods for myocarditis lack specificity, which can result in mistakenly diagnosing other conditions as myocarditis or failing to recognize this potentially life-threatening condition[14].

Clinical presentations and diagnosis of CIM

CIM varies from mild to severe, with symptoms overlapping those of CIP and clozapine side effects, including fever, fatigue, dizziness, chest pain, tachycardia, shortness of breath, abnormal heart rhythm, and peripheral edema[14,17,18], necessitating advanced diagnostics beyond clinical suspicion.

Diagnosis of CIM

The current diagnostic criteria for CIM are not well-defined, leading to treatment termination based solely on clinical suspicion[19]. Up to 65% of patients presumed to have CIM do not meet ESC criteria for myocarditis[1]. The ESC working group on myocardial and pericardial disease has proposed diagnostic criteria based on an integrative approach for myocarditis. This approach was developed in response to the challenges of confirming or excluding myocarditis based on a single diagnostic test. Their recommendations were primarily focused on viral myocarditis, the most common cause, but have also been extended to consider other underlying causes, such as hypersensitivity or toxicity from medications, including clozapine[20]. These criteria and the workup are designed to assist in making a diagnosis and ruling out other conditions with similar clinical presentations.

These criteria include the presence of suggestive clinical presentations such as acute chest pain, new onset or worsening of dyspnea, fatigue, subacute/chronic worsening of dyspnea and fatigue, palpitations, unexplained arrhythmia, or unexplained cardiogenic shock. These clinical features should be followed by investigative findings of the following: (1) Biomarkers of myocardial injury; (2) EKG/Holter/Stress test features; (3) Cardiac imaging or CMR changes[21].

The utility of standard tests in diagnosing myocarditis can be significantly limited. EKG changes were seen in 78% of patients with CMR-proven myocarditis; however, these features are primarily nonspecific, and up to 22% of patients with CMR-diagnosed myocarditis showed normal EKG[21]. Cardiac markers such as high-sensitivity cardiac troponin can be a valuable indicator of myocardial tissue damage. Cardiac troponin was reported to have a sensitivity of 34%, specificity of 89%, and positive predictive value of 82%. CRP is another marker used for myocarditis, with a sensitivity of 52% and a specificity of 81%[4,22]. However, these data were recovered from studies for all causes of myocarditis, where viral myocarditis is the most common etiology[4].

Echocardiographic changes can be nonspecific and reported in about 30%-60% of patients with myocarditis[23,24]. Hence, the absence of abnormalities in cardiac echocardiograms does not exclude myocarditis. Endomyocardial biopsy (EMB) is the gold standard for diagnosing myocarditis. It involves using the updated Dallas criteria and polymerase chain reaction, as well as conducting immunohistochemical analysis of inflammatory cells. EMB provides detailed information about the histological composition of inflammatory cell infiltrations, cardiomyocyte damage, necrosis, fibrosis, atrophy, and hypertrophy. These findings can help determine the underlying cause of myocarditis, particularly when combined with polymerase chain reaction or immunohistochemical analysis of inflammatory cells[25]. However, it is essential to note that EMB is an invasive procedure, and the results can vary depending on the location of the biopsy, particularly in cases of patchy myocardial involvement. Sensitivity can increase from 50% for one biopsy to 90% for seven biopsies[26,27]. The updated Dallas criteria, used for interpreting EMB results, have a sensitivity of 60% and a specificity of 80%[28]. EMB is generally recommended for patients with severe or rapidly progressing symptoms, especially when less invasive tests are insufficient for a definitive diagnosis[29]. CMR has become a valuable and non-invasive option for diagnosing myocarditis in recent years due to its ability to characterize inflamed tissues in multiple ways[30]. The CMR findings of myocardial inflammation were outlined in the Lake Louise Criteria 2009. However, the initial criteria were criticized for their moderate diagnostic sensitivity and high subjectivity in qualitative assessment[31]. The Lake Louise Criteria was revised in 2018 to include parametric mapping and several quantitative evaluations, which increased sensitivity to approximately 88% while maintaining high specificity for patients with a strong clinical suspicion of myocarditis[32,33]. The results of CMR can help identify the cause of myocarditis, especially when combined with the patient's clinical symptoms[31]. This capability was used during the coronavirus disease 2019 pandemic to distinguish between myocarditis related to the coronavirus disease 2019 virus and that caused by vaccination[31]. Despite the clear benefits of CMR, challenges such as cost, availability, and accessibility hinder its widespread use for diagnosing myocarditis. In many centers, the approximate wait time for patients is between 2 weeks and 4 weeks[33]. Nonetheless, CMR has emerged as a non-invasive, highly sensitive tool for diagnosing myocarditis. Table 2 provides a summary of the significance of various tests used for assessing myocarditis.

CIP

The Food and Drug Administration has issued a warning about second-generation antipsychotics due to the increased mortality among older adults with behavioral disturbances. Pneumonia was reported to be the most common cause of death in this population[34]. Patients taking clozapine are significantly more susceptible to pneumonia, with an odds ratio of 4.07 compared to the general population[3]. The risk of pneumonia depends on the dosage, with the frequency among clozapine patients ranging from 19% to 34%[35].

Several factors are believed to increase the risk of pneumonia in patients taking clozapine. These factors include sedation, excessive salivation, difficulty swallowing, and alterations to the immune system[35]. Managing excessive salivation and monitoring sedation levels is recommended to lower the risk of inhaling saliva or food particles and developing pneumonia[36]. Patients with pneumonia may experience increased clozapine levels due to the release of cytokines that can inhibit the activity of the cytochrome P450 enzyme, the primary enzyme responsible for clozapine metabolism[37]. Therefore, many medical professionals recommend reducing the clozapine dose by half during an upper respiratory infection accompanied by fever. The microbiology of clozapine-associated pneumonia has not been extensively studied; however, aspiration is a suspected mechanism of infection. Several bacteria are common during aspiration, including Streptococcus pneumoniae, Staphylococcus aureus, Haemophilus influenzae, and Enterobacteriaceae[38].

The symptoms of CIP can overlap with the symptoms of myocarditis. They include fever, cough, chest pain, shortness of breath, fatigue, rapid heartbeat, and dizziness[39-41]. While chest radiographs are considered the gold standard for diagnosing pneumonia, they have a lower positive finding among culture-positive patients than chest CT scans (51.5% vs 90.7%)[42], all of which should mandate the need for careful interpretation of a normal chest X-ray (CXR) and a follow-up with a chest CT scan might occasionally be required. Table 3 highlights the overlap between symptoms of clozapine-associated pneumonia and myocarditis.

Table 3 Highlights the overlap between symptoms of clozapine-associated pneumonia and myocarditis.

Symptom	CIM frequency (%)	CIP frequency (%)
Fever	40-70[14]	71-75[39]
Difficulty breathing	27-70[14]	67-75[39]
Dizziness with postural changes	50[14]	35[39]
Rapid heartbeat	46-50[14]	53[40]
Chest pain	32-40[14]	39-49[39]
Fatigue	20[14]	90[39]
Lower limbs swelling	20[14]	Uncommon

CIM: Clozapine-induced myocarditis; CIP: Clozapine-induced pneumonia.

Diagnostic challenges in CIM and CIP differentiation

The current approaches to diagnosing CIM are hindered by an overreliance on nonspecific biomarkers and inconsistent imaging practices, which are further complicated by inadequate differentiation from CIP. Troponin and CRP are cornerstone screening tools, reflecting myocardial inflammation with sensitivities of 34% and 52%, and specificities of 89% and 81%, respectively[4,22]. However, these markers lack specificity for CIM, as they are elevated in up to 52% and 75% of pneumonia cases, respectively[43,44], a condition that is frequently observed among clozapine users (odds ratio = 4.07)[3]. The proposed troponin (> 0.04 ng/mL) and CRP (> 10 mg/L) thresholds are based on viral myocarditis studies, as there is limited CIM-specific data available. However, CIM may present with higher troponin elevations, as seen in cohorts by Ronaldson et al[5] and Tirupati et al[6], suggesting a need for recalibration. For instance, Ronaldson et al[5] reported troponin levels greater than 0.1 ng/mL in confirmed cases of CIM, indicating a potential for higher specificity with adjusted cut-offs. Dynamic monitoring of biomarkers over 24-48 hours, rather than single-point measurements, may further enhance diagnostic accuracy, particularly when distinguishing between CIM and CIP, as CRP elevations are common, occurring in up to 75% of cases. Additional biomarkers, such as erythrocyte sedimentation rate, eosinophil count, and creatine phosphokinase (CPK), are variably employed across protocols without a consensus on their optimal use (Table 1), further clouding diagnostic precision. The overlap in clinical presentation - fever, dyspnea, chest pain, and tachycardia - which is common to both CIM and CIP (Table 3) - exacerbates this uncertainty, risking misdiagnosis in up to 65% of presumed CIM cases that fail to meet the ESC criteria.

Imaging modalities provide limited support in resolving these challenges. EKG reveals non-specific changes in 78% of CMR-confirmed myocarditis cases, yet 22% of affected patients show normal results[21]. Echocardiography, used routinely in some protocols (e.g., Berk et al[45] in 2007) or reserved for suspected cases (e.g., Anıl Yağcıoğlu et al[46] in 2019), detects abnormalities in 30%-60% of CIM patients but lacks specificity and often appears normal in CIP[23]. CXRs, the standard for CIP diagnosis, miss the condition in nearly half of the culture-positive cases (51.5% sensitivity vs 90.7% for chest CT)[42], undermining their utility in ruling out this confounder.

Endomyocardial biopsy, the gold standard for CIM, has a sensitivity of 60% and specificity of 80%[28], but is invasive and impractical for routine use. In comparison, CMR offers high sensitivity (88%) and specificity (91%) non-invasively[32] but remains underutilized due to cost and access barriers[33]. Current protocols rarely standardize advanced imaging or CIP exclusion, thereby perpetuating diagnostic variability. Suspected CIM rates range from 5.3% to 11.3%, with confirmed cases spanning 1.4% to 9.8% (Table 1), highlighting the need for a more robust and integrated approach. Table 4 summarizes the significance of various tests for both CIM and CIP.

Table 4 Summary of the significance of various tests for both clozapine-induced myocarditis and clozapine-induced pneumonia.

Test	CIM sensitivity/specificity	CIP sensitivity/specificity	Dynamic monitoring	Comments
EKG changes	Non-specific and seen in 78% of patients[21]	Non-specific[40]	NA	Feasible in all settings
				Non-specific changes (e.g., ST changes, T-wave inversion) in CIM; normal in 22% of cases
				Limited use in CIP
cTnl	34%/89%	Elevated by 52%[43]	Cornerstone for CIM screening; elevated in CIP	Feasible in all settings
				Repeating testing is critical for trends (> 20% rise suggests escalation)
CRP	52%/81%[22]	Elevated in 75%[44]	Yes (24-48 hours) in CIM Repeat testing (e.g., > 50 mg/L) informs interim management	Feasible in all settings
Echo changes	Nonspecific and seen in 30-60%[23]	Likely normal	Yes (48 hours-5 days)	Feasible in most settings
				Detects wall motion defects in CIM, normal in CIP. Serial testing is key in resource-limited settings (echo-only pathway)
Chest X-ray	NA	51% sensitivity	NA	Feasible in most settings
				Misses CIP in approximately 50% of cases; used in resource-limited settings. Normal X-ray requires clinical correlation
Chest CT	NA	90.7%	NA	Gold standard for CIP; critical for excluding CIP in CIM workup
				Limited to well-resourced settings due to cost (300 dollars-1000 dollars)
Endomyocardial biopsy	60%/80%[28]	NA	NA	Gold standard for CIM; invasive, rarely used. Feasible only in specialized centers with severe cases
Cardiac MRI	88%/91%[32]	NA	NA	Non-invasive, highly sensitive for CIM
				Limited by cost (1000 dollars-3000 dollars) and access (2-4-week wait)
				Preferred in well-resourced settings

Dynamic monitoring: Indicates whether serial testing (e.g., 24-48 hours for biomarkers, 48 hours to 5 days for echocardiography) enhances diagnostic precision, as per the interim management strategy in Figure 2. Resource availability: Chest X-ray, electrocardiogram, echocardiography, and biomarkers are feasible in resource-limited settings; however, chest computed tomography and cardiac magnetic resonance imaging are typically limited to well-resourced settings. Differentiation: Clozapine-induced pneumonia exclusion via chest computed tomography (or X-ray in resource-limited settings) is critical before clozapine-induced myocarditis (CIM) confirmation. Echocardiography and cardiac magnetic resonance imaging are key for CIM, with serial testing improving reliability in resource-limited settings. Sensitivity and specificity data for CIM are primarily derived from viral myocarditis studies[4,30]. CIM-specific data are limited and require validation. cTnl: Cardiac tropinin; CRP: C-Reactive protien; CT: Computed tomography; MRI: Magnetic resonance imaging; CIM: Clozapine-induced myocarditis; CIP: Clozapine-induced pneumonia; NA: Not available.

Strengths and limitations of current protocols for screening and diagnosing CIM

The serious risk of inducing myocarditis necessitates effective monitoring protocols to ensure early detection and management. Hence, several studies and protocols have been developed for screening and diagnosing CIM (Table 1). Variability in diagnostic thresholds across protocols has led to significant disparities in reported CIM rates. Suspected cases range from 5.3%[47] to 11.3%[46], while confirmed cases span 1.4%[48,46] to 9.8%[49]. These differences likely arise from variations in monitoring intensity, patient demographics, and diagnostic criteria, as detailed in Table 1. For instance, protocols like Ronaldson et al[5] emphasize weekly troponin and CRP monitoring for four weeks, while others extend this to eight weeks[49] or incorporate annual echocardiography[50]. Such inconsistencies reflect a lack of consensus on optimal timing and tools.

A key observation of most protocols is their reliance on troponin and CRP as cornerstone biomarkers, widely recognized for their ability to detect myocardial inflammation early[5,6]. Additional markers - ESR, eosinophil count, creatinine phosphokinase, B-type natriuretic peptide, and white blood count - are variably employed, offering flexibility but no standardized panel (e.g., Kanniah et al[51]; Nachmani Major et al[48]. Some protocols integrate routine imaging, like echocardiography[45,50], or reserve it for suspected cases[46,48], while advanced tools like CMR are occasionally used for confirmation[50,52]. Large-scale studies (e.g., Tirupati et al[6], n = 327) and clear workup pathways outlined in the New South Wales Health guidelines[50] further bolster the evidence base, with progression from expert-driven guidelines, such as those by Berk et al[45], to data-driven approaches marking a significant advancement.

However, limitations persist. Small sample sizes in early studies (e.g., Murch et al[52], n = 122; McNutt et al[53], n = 38) and retrospective designs (e.g., Nachmani Major et al[48]; Segev et al[54]) limit the generalizability of the findings[48]. Inconsistent imaging use - ranging from routine to symptom-driven - undermines diagnostic reliability, as does the rare application of CMR, despite its high sensitivity[32,55]. Many protocols fail to address overlapping conditions, such as CIP, a frequent confounder that shares symptoms and biomarkers with other conditions (Table 3). For example, Kanniah et al[51] and de Leon et al[56] focus narrowly on screening without guidance for suspected cases or CIP exclusion, which can lead to misdiagnosis in up to 65% of presumed CIM cases[1].

Clozapine rechallenges following presumptive diagnosis of CIM

Recommendations for clozapine rechallenge after presumed CIM rest on limited evidence from small case series and reviews of isolated case reports. Diagnoses in these studies often rely on non-specific biomarkers (e.g., troponin, CRP) or suggestive clinical features, lacking rigorous confirmation. Richardson et al[57] analyzed 88 cases, with only 4.5% confirmed by biopsy and no use of transesophageal echocardiography (TEE) or CMR. Despite this, 35.3% of rechallenges failed, with 2.9% resulting in death. McMahon et al[58] studied 45 patients, employing EKG and TEE in fewer than one-third of the cases and involving cardiologists in nearly half; however, they did not include CMR or biopsy. The rechallenge failed in approximately one-third of cases, although no fatalities occurred. In a series of three patients, Noël et al[59] reported success only in one, who had asymptomatic, mild elevations of CRP and troponin; the other two, with flu-like symptoms and EKG/TEE evidence of myocarditis, experienced recurrence.

Current screening protocols further complicate the decision-making process for rechallenge. Anıl Yağcıoğlu et al[46] found that up to 10% of patients screen positive for CIM, yet existing methods rarely confirm the diagnosis. This diagnostic uncertainty, combined with the risk of fatal outcomes[57,58], argues against rechallenging patients with presumed CIM unless definitive confirmation excludes alternative causes.

Suggested recommendations to improve the diagnosis of CIM

This review highlights critical gaps in current diagnostic approaches, including an overreliance on non-specific biomarkers, inconsistent imaging practices, and insufficient differentiation between CIP and CIM. Given the high incidence of CIP, its significant overlap with CIM in clinical presentation and biomarker profiles, and the potential for misdiagnosis, incorporating a workup for CIP is essential when evaluating patients for suspected CIM. CXRs, commonly used to diagnose CIP, frequently fail to detect it, rendering chest CT scans a more reliable tool for this population. Similarly, while EKG and echocardiography are routinely employed for suspected myocarditis, their findings often lack specificity. The diagnostic gold standard - endomyocardial biopsy—is invasive and carries significant risks. In contrast, CMR offers a non-invasive, highly sensitive alternative that confirms CIM, elucidates etiology, and informs treatment strategies. Expanding access to and utilization of CMR is therefore strongly recommended for evaluating CIM[33].

To address these challenges, we propose an algorithm that integrates chest CT to exclude CIP, leverages cardiology collaboration with CMR to confirm CIM, or utilizes CXR and echocardiography when resources are limited. If troponin > 0.04 ng/mL or CRP > 10 mg/L, proceed to chest CT; if negative, perform CMR or perform CXR followed by echocardiography if CMR is inaccessible.

Key recommendations

Monitoring protocols: While current protocols recommend monitoring for myocarditis during the first 4-8 weeks, their utility remains questionable. Elevated troponin and CRP levels alone should not prompt discontinuation of clozapine, as these markers are nonspecific and more likely to indicate pneumonia than myocarditis.

Diagnostic approaches: Prioritize CT scans for diagnosing pneumonia and advocate for the use of CMR in suspected myocarditis cases in well-resourced settings. In limited-resource settings, utilize CXR and serial echocardiography due to their non-invasive nature and reasonable diagnostic value.

Interim management during diagnostic workup: Given CMR wait times of 2-4 weeks in many centers, interim management of clozapine is crucial to balance the risk of CIM with therapeutic continuity. For patients with suspected cardiac injury syndrome (troponin > 0.04 ng/mL or CRP > 10 mg/L), perform immediate echocardiography and repeat biomarker testing within 24-48 hours. The following decision tree is proposed: (1) Mild presentation: (e.g., isolated biomarker elevation, stable vitals, no dyspnea or chest pain): Continue clozapine at current dose, monitor biomarkers daily, and repeat echocardiography in 3-5 days if biomarkers rise (e.g., troponin increase > 20% or CRP > 50 mg/L). Expedite chest CT to exclude CIP; (2) Moderate presentation: (e.g., fever, tachycardia, mild dyspnea, biomarker elevation): Reduce clozapine dose by 25%-50%, monitor biomarkers and vitals twice daily, and repeat echocardiography in 48 hours. If a chest CT confirms CIP, manage the patient according to infectious disease protocols and consider a temporary dose reduction rather than discontinuation; and (3) Severe presentation: (e.g., chest pain, significant dyspnea, arrhythmias, or troponin > 0.1 ng/mL): Hold clozapine, admit for cardiology consultation, and pursue urgent CMR or endomyocardial biopsy if echocardiography is inconclusive. Restart clozapine only after CIM is excluded.

This approach ensures timely CIP exclusion via chest CT, leverages echocardiography for interim monitoring, and minimizes the need for unnecessary discontinuation of clozapine. This approach strikes a balance between diagnostic precision and clinical feasibility. For CIP, the protocol supports dose reduction during acute illness and gradual re-escalation post-recovery to preserve clozapine’s therapeutic benefits. In contrast, permanent discontinuation is advised for confirmed or strongly suspected CIM due to the risk of recurrence. By harnessing the precision of advanced imaging, this framework addresses overlapping conditions, optimizes the role of clozapine in treatment-resistant schizophrenia, and provides a practical strategy to refine clinical practice. Ultimately, this protocol has the potential to enhance patient outcomes, minimize diagnostic errors, and improve the safe use of clozapine in psychiatric care.

Rechallenge: The current evidence supporting the rechallenge of clozapine following a diagnosis of myocarditis is limited. The limited number of confirmed cases and a high failure rate make it unsafe. Based on this evidence, the authors strongly discourage rechallenge with clozapine following the diagnosis of CIM.

Multidisciplinary care: Manage suspected myocarditis cases with a multidisciplinary team, including a cardiologist familiar with the risks and benefits of clozapine.

Clinical awareness: Educate clinicians about the challenges of diagnosing myocarditis in patients taking clozapine and the importance of considering alternative diagnoses, such as pneumonia, to ensure accurate diagnosis and treatment.

Figure 2 presents our recommendations for improving the recognition of CIM. Our proposed protocol addresses critical gaps in diagnosing CIM and pneumonia by integrating advanced imaging techniques, such as CT and CMR, with multidisciplinary collaboration. This approach not only enhances diagnostic accuracy but also ensures that patients continue to benefit from clozapine's unparalleled efficacy in treatment-resistant schizophrenia.

Figure 2 Diagram for suggested recommendations to improve the recognition of clozapine-induced myocarditis. CIM: Clozapine-induced myocarditis; CIP: Clozapine-induced Pneumonia; CRP: C-reactive protein; CT: Computed tomography; CXR: Chest X-ray.

Strengths and weaknesses: Implementing the proposed CIM diagnostic algorithm requires addressing cost, access, and workflow integration. Chest CT costs range from 300 dollars to 1000 United States dollars, and CMR costs 1000 dollars to 3000 dollars, varying by region and facility[33]. These expenses may deter adoption in resource-limited settings, though echocardiography (100-500 dollars) offers a cost-effective alternative. In the United States, Medicare and private insurers often reimburse CT and CMR for myocarditis evaluation (current procedural terminology codes 75561, 71260). Still, coverage for clozapine-specific indications may require justification of medical necessity. Workflow integration involves coordinating services across psychiatry, cardiology, and radiology, potentially through multidisciplinary clinics or teleconsultation for rural facilities. Training psychiatrists to interpret biomarker trends and recognize the overlap between CIP and CIM can streamline referrals and reduce diagnostic delays.

A preliminary cost–benefit analysis suggests that accurate CIM diagnosis could reduce unnecessary clozapine discontinuations, which incur significant costs from schizophrenia relapses (e.g., hospitalizations averaging 15000 dollars-20000 dollars per episode). By preventing misdiagnosis in up to 65% of presumed CIM cases, the algorithm may yield net savings, though formal economic modeling is needed. Barriers include CMR wait times (2-4 weeks) and limited radiology capacity in low-resource settings, necessitating reliance on echocardiography and clinical assessment.

Our findings are limited by the heterogeneity of the included studies and the lack of standardized diagnostic criteria across protocols. We propose a prospective validation study to transition from narrative recommendations to evidence-based practice. A multicenter cohort of 200-300 clozapine-treated patients with suspected CIM would assess the algorithm’s sensitivity, specificity, positive predictive value, and time to diagnosis. Patients would undergo chest CT, CMR (or echocardiography), and biomarker monitoring per the protocol, with outcomes compared to endomyocardial biopsy or clinical follow-up as reference standards. Secondary outcomes include clozapine continuation rates and the cost-effectiveness of the treatment. This study, ideally conducted across high- and low-resource settings, would refine biomarker thresholds, suggest and validate the clinical CIM Risk assessment, and inform global implementation strategies.

Future studies should focus on validating this protocol in diverse clinical settings and exploring cost-effective strategies to improve access to advanced imaging in resource-limited environments. In addition, studies should prioritize establishing specific biomarker thresholds and validating their predictive value in prospective cohorts of patients treated with clozapine.

CONCLUSION

This systematic review underscores the diagnostic challenges of CIM and its frequent misdiagnosis with CIP, driven by non-specific biomarkers and inconsistent imaging practices. Our analysis of 119 studies reveals that up to 65% of presumed CIM cases lack confirmatory evidence, risking premature clozapine discontinuation, while rechallenge succeeds in 64.7%-68.9% of cases, albeit with a 2.9% fatality risk. To address these gaps, we propose a multidisciplinary diagnostic algorithm that integrates chest CT to exclude CIP, CMR for CIM confirmation in well-resourced settings, and echocardiography as a practical alternative in other settings, such as CXR and serial echocardiography. Emphasizing slower titration and cardiology collaboration, this protocol balances diagnostic precision with clozapine’s unparalleled benefits for treatment-resistant schizophrenia. Our approach aims to enhance patient safety and therapeutic outcomes by reducing misdiagnosis and optimizing management. Future research should validate this algorithm across diverse clinical settings, explore cost-effective imaging solutions for resource-limited environments, and refine the criteria for safe clozapine rechallenge to further support its role in psychiatric care.