Immune checkpoint inhibitor rechallenge after grade 3 myocarditis: is it ever safe?

58M with metastatic melanoma (BRAF wild-type) who had an excellent partial response to pembrolizumab (8 cycles, tumor burden reduced by 72%). At cycle 9, he developed grade 3 immune-related myocarditis: troponin peak 8.4 ng/mL, LVEF dropped from 62% to 35%, cardiac MRI showed diffuse myocardial edema and late gadolinium enhancement.

Management: Pembrolizumab was permanently discontinued. Treated with methylprednisolone 1g/day x 3 days, then tapered over 8 weeks. He also received mycophenolate for 4 weeks. Complete recovery: troponin normalized, LVEF returned to 58% at 3-month echo, cardiac MRI showed resolution of edema.

6 months later: Surveillance imaging shows progressive disease (new lung metastases, growing hepatic lesions). No targetable mutations on comprehensive genomic profiling.

The question:

All guidelines say "permanently discontinue" after grade 3-4 myocarditis
But this patient has limited treatment options (BRAF WT melanoma, no targetable mutations)
Ipilimumab monotherapy (different checkpoint) has response rates of only 10-15%
Is there ANY scenario where rechallenge with anti-PD1 is considered?

This is one of the most emotionally and ethically difficult scenarios in immuno-oncology. At AIIMS, we have encountered this dilemma 4 times in the past 2 years. Here is my framework:

The standard answer is: No. ICI-related myocarditis has a 25-50% mortality rate, and rechallenge after grade 3-4 cardiac irAE carries an estimated recurrence risk of 30-60% (per the Dolladille 2020 meta-analysis). The NCCN, ESMO, and ASCO guidelines all state "permanent discontinuation."

However, reality is more nuanced. Consider this decision matrix:

Factors favoring rechallenge consideration (ALL must be present):

Complete resolution of myocarditis (normal troponin, LVEF, cardiac MRI) -- THIS PATIENT MEETS THIS
No alternative therapies with meaningful efficacy -- THIS PATIENT MEETS THIS
Life-threatening cancer with documented response to ICI -- THIS PATIENT MEETS THIS
Patient fully informed of 30%+ risk of recurrent myocarditis (potentially fatal)
Cardio-oncology team agrees to intensive monitoring protocol

If rechallenge is attempted, our protocol:

Use ipilimumab (anti-CTLA4) instead of anti-PD1. Mechanism is different and myocarditis is less common with CTLA4 inhibitors (though it can still occur)
If anti-PD1 rechallenge specifically: start at reduced dose (pembrolizumab 100mg instead of 200mg, i.e., approximately 1 mg/kg)
Prophylactic abatacept has been proposed (CTLA4-Ig may modulate the T-cell response causing myocarditis)
Weekly troponin for the first 6 cycles, then biweekly
ECG before every cycle
Echocardiogram monthly for 6 months, then every 3 months
Low threshold for cardiac MRI at any troponin rise

Published rechallenge data: The Dolladille registry identified 29 patients rechallenged after cardiac irAEs. 9/29 (31%) had recurrent cardiac events. 2/29 died of recurrent myocarditis. However, 14/29 (48%) had objective tumor response.

My personal recommendation for THIS patient: I would try ipilimumab monotherapy first (different mechanism, 10-15% response rate). If that fails, discuss anti-PD1 rechallenge with full informed consent, cardio-oncology co-management, and the understanding that this is an off-guideline decision with real mortality risk.

Sharing the Royal Marsden approach and our institutional data:

We have a formal ICI rechallenge committee (medical oncology, cardio-oncology, immunology, ethics) that reviews every case of severe irAE where rechallenge is considered. In 3 years, we have reviewed 41 cases; 8 involved cardiac irAEs.

Our committee approved rechallenge in 3/8 cardiac cases. Outcomes:

Patient 1: Anti-PD1 rechallenge after grade 3 myocarditis. Recurrent grade 2 myocarditis at cycle 3, ICI permanently discontinued. Survived with supportive care.
Patient 2: Switch to ipilimumab after grade 3 myocarditis on pembrolizumab. No cardiac recurrence. Partial response maintained for 14 months.
Patient 3: Anti-PD1 rechallenge with concurrent abatacept. No cardiac recurrence at 8 months. Ongoing PR.

Key innovation we are piloting: Pre-rechallenge endomyocardial biopsy to assess residual subclinical inflammation. Even with normal troponin and echo, some patients have persistent lymphocytic infiltrates that predict recurrence. If EMB shows active inflammation, rechallenge is absolutely contraindicated regardless of other factors.

For this specific patient: I would choose ipilimumab + nivolumab at reduced dose (ipilimumab 1mg/kg + nivolumab 1mg/kg q6 weeks, as per CheckMate 511 dosing) rather than anti-PD1 monotherapy. The combination has synergistic efficacy in melanoma, and the lower nivolumab dose may reduce cardiac risk.

This decision requires documented shared decision-making, and I would strongly recommend an ethics consultation.

2 Answers