MYTHS – MYocarditis THerapy with Steroids
San Diego, California
Enrico Ammirati, MD, PhD
Acute myocarditis (AM) is a common condition characterized by histological evidence of inflammatory infiltrates associated with myocyte necrosis of non-ischemic origin. Clinical presentation spans from indolent form to cardiogenic shock also called fulminant myocarditis (FM). Patients can be stratified on the basis of their clinical presentation: patients with left ventricular (LV) ejection fraction (EF)<50% at first echocardiogram, and those with sustained ventricular arrhythmias, called complicated AM, have a worse prognosis compared with uncomplicated cases with preserved left ventricular ejection fraction (LVEF) and without arrhythmias. Among complicated AM, FM patients are those ones at the highest risk, presenting with severely impaired LVEF (generally <40%), and with need for inotropes and/or temporary mechanical circulatory supports (t-MCS).The pathogenesis of AM is felt to be due to an immune-mediated response against the myocardium.
As such, the overall objective is to evaluate the efficacy of pulsed IV corticosteroids therapy for the treatment of AM. It is proposed to test the efficacy of pulsed IV methylprednisolone in a single blind randomized controlled trial versus standard therapy on top of maximal support. The rationale for using pulsed corticosteroid therapy in the acute setting (within 3 weeks from cardiac symptoms’ onset) to reduce myocardial inflammatory infiltrates favoring recovery appears strong. Nevertheless, no trial has tested this hypothesis in the very acute phase of AM, despite the high mortality rate of this condition and the fact that AM mainly affects young patients.
Currently, no specific medications in the acute phase of lymphocytic AM are recommended beyond supportive therapy with inotropes and t-MCS. One Cochrane review on corticosteroids showed that almost all studies focused on inflammatory cardiomyopathies with 6 months of symptoms of heart failure (HF), and despite an improvement of cardiac function observed in low quality and small size studies, there was no improvement in the survival. In the past, only one study assessed the efficacy of immunosuppression in AM, the Myocarditis Treatment Trial (MTT) that reported no benefit from immunosuppression. Neutral results in the MTT could be ascribed to a delay in the initiation of this potentially effective treatment. Thus, 55% of patients started immunosuppressive therapy after 1 month from the onset of myocarditis, when the left ventricle (LV) was already dilated, as highlighted by a mean LV end-diastolic diameter (EDD) of 64 mm. It is expected that patients with FM have normal LV dimension during the acute phase despite severe LV systolic dysfunction. Based on a study from PI group, it was observed that FM patients recover most of the LVEF in the first 2 weeks after admission, with a median absolute increase of 30%. This finding further suggests that an immunosuppressive treatment should be started as soon as possible to demonstrate effectiveness. As little has changed in the medical treatment of this condition in the last 30 years, identification of effective drugs is needed.
Patients admitted to hospital for suspected AM complicated by acute HF/cardiogenic shock and LV systolic dysfunction will be screened for randomization.
Patients will be randomized in the two arms in a 1:1 ratio (Pulsed methylprednisolone therapy vs Placebo). Randomization will be performed with stratification by country.
The primary objective is to demonstrate a reduction in the rate of the primary composite endpoint on patients treated with pulsed methylprednisolone therapy vs. standard therapy and maximal supportive care.
Endpoints will be analyzed according to the following principles:
- Intention-to-treat (ITT) population
- Per Protocol (PP) population:
- “Safety population”
- A sensitivity analysis will also be performed on the previously defined populations after excluding patients (1) with histological diagnosis of giant cell myocarditis (GCM) or (2) who did not reach the final diagnosis of acute myocarditis based on CMRI or histology.
Sample size calculation: we plan to recruit a total of 360 patients, and we expect that about 20% of these patients or local physicians will refuse randomization. This would leave a total of 288 randomized patients (144 per arm).
Considering as relevant a reduction in the probability to reach the primary endpoint at 6 months from 25% in the standard therapy on top of maximal supportive care arm to 12% in the pulsed corticosteroid therapy arm (absolute risk reduction of 13% in absolute corresponding to a hazard ratio (HR) pulsed corticosteroid therapy vs. standard therapy of 0.44), the planned sample size will allow achieving a power of 0.80 with a one-sided log-rank test and an overall type I error of 0.025. The 25% figure considered for the standard therapy derives from a retrospective analysis of the patient’s cohort spanning over 20 years. The calculation includes an interim analysis planned at 50% recruitment (O'Brien-Fleming method). This interim analysis is accounted for in the sample size calculation with an alpha level of 0.001525 (final analysis 0.023475 alpha level) and is planned on the primary endpoint to assess a possible early treatment effect. No specific stopping rules are planned, given the multiplicity of aspects involved, but the report on safety will be reviewed by the Data and Safety Monitoring Committee (DSMC) will advise on possible aspects of the trial that need reconsideration.
Sample size adaptation: We will consider, based on the DSMC advise an adaptive approach to sample size in two regards:
- At the interim analysis, if the baseline incidence is lower than the expected 25%, the sample size calculation may be re-evaluated keeping the same HR of 0.44. For instance, if the observed incidence is 20%, maintaining the same HR of 0.44 (corresponding to an incidence of 9% in the pulsed corticosteroid therapy group, i.e. 11% in absolute risk reduction) the effective sample size needed to achieve 80% power should be increased to 360 patients. If the baseline incidence is higher than 25%, the planned actual sample size will achieve a power greater than 80% to detect a HR of 0.44 and no action will be taken.
- Based on the conditional power method, and on the DSMC advise, we may reconsider if a less promising result than HR=0.4444 is worth pursuing, given the current observed estimate. This case would require an increase in sample size that will be discussed in terms of relevance and feasibility. For example, if at the planned interim analysis, the estimated absolute risk reduction is at least 10% (HR=0.56) with 25% baseline, and the conditional power of meeting this 10% target (instead of the planned 13%) would be at least 60%, the sample size may be increased to reach the 80% desired power. In this case, the final effective sample size should be increased to 254 patients per arm to preserve an 80% power of demonstrating the less marked difference. The flexibility allowed in the sample size estimate will be considered based on the evaluation of the interim report by the DSMC and no data will be disclosed on interim treatment estimates to the study coordinator and steering committee.
The overall duration of the study from first patient first visit to last patient last visit will be 39 months. The follow-up will be up to 6 months and with additional 3 months to lock the database. Enrollment will last 30 months.
In parallel, there will be a prospective registry of patients that are eligible for the trial, but they are not randomized.
A second registry, called MYOCARDITIS REGISTRY will prospectively recruit all patients with acute myocarditis demonstrated by CMRI or EMB who are not eligible for randomization (not all centers will take part in this registry).
The study is supported by a grant from Italian Ministry of Health (GR-2019-12368506) and Lombardy Region.
Exemption from the investigational new drug (IND) regulations by FDA on August 2nd, 2021 (PIND: 15727)
EudraCT identifier: 2021-000938-34