Correspondence

Fulminant Myocarditis

N Engl J Med 2000; 343:298-300July 27, 2000

Article

To the Editor:

In their study of fulminant myocarditis as compared with acute (nonfulminant) myocarditis, McCarthy et al. (March 9 issue)1 included all patients in whom the diagnosis was established by endomyocardial biopsy and who met their selection criteria. However, by not incorporating diagnoses established at autopsy, the authors introduced a selection bias in favor of a better outcome of fulminant myocarditis. Excluding cases with a rapidly fatal onset essentially reduces the study to a comparison of patients with acute myocarditis and early survivors of fulminant myocarditis.

In several autopsy series,2-4 fulminant myocarditis was found to be a prominent cause of sudden death. In one study, lymphocytic myocarditis was detected in nearly 6 percent of 2427 cases of sudden death.2 Burlo et al.3 reported a prevalence of 5.1 percent by systematically incorporating standardized methods of myocardial sampling. In a series of 60 patients, lymphocytic myocarditis was detected at autopsy in 6 patients.4 In four additional patients, fulminant myocarditis was diagnosed during life, but all four died after a mean period of 48 months, whereas the mortality rate was 29 percent in the group of patients with acute myocarditis. In a series of five patients with fulminant myocarditis,5 three died within days, and two required transplantation. In addition, numerous cases of fulminant myocarditis in which the initial presentation was either sudden death or circulatory collapse resistant to conventional therapy have been reported.

Another problem with the study by McCarthy et al. is the large proportion of patients who were lost to follow-up, particularly in the group with fulminant myocarditis (47 percent). These patients were assumed to be event-free almost until the termination date of the study. In a sensitivity analysis, the results of this approach could have been contrasted with the results of censoring data for all study subjects with an unknown vital status at a date after but close to the discharge date, when they were last known to be alive, or classifying them as dead at the same date. Furthermore, using the time in the study rather than age as the time variable is arguably suboptimal and fails to address left-censoring selection bias (i.e., does not control for losses before enrollment).6 It is also unclear whether confounders were considered separately or jointly. Moreover, the choice of covariates was based on statistical significance, which is considered by many to be irrelevant in assessing confounding effects on the hazard under study.

These limitations, combined with already unstable estimates because of the small sample, cast serious doubts on the validity of the findings, which prompted Karliner, in his accompanying editorial,7 to suggest counterintuitively that “the sicker the patients are, the more likely they are to survive.” Although we do not dispute the potential for recovery from fulminant myocarditis, there is insufficient evidence to support the claim by McCarthy et al. that patients with this disorder have an “excellent long-term prognosis” that is superior to the prognosis for patients with acute myocarditis.

Paul Khairy, M.D., C.M.
Montreal Heart Institute, Montreal, QC H1T 1C8, Canada

Claire Infante-Rivard, M.D., Ph.D.
McGill University, Montreal, QC H3A 1A3, Canada

7 References

1. 1

   McCarthy RE III, Boehmer JP, Hruban RH, et al. Long-term outcome of fulminant myocarditis as compared with acute (nonfulminant) myocarditis. N Engl J Med 2000;342:690-695
   Full Text | Web of Science | Medline

2. 2

   Wentworth P, Jentz LA, Croal AE. Analysis of sudden unexpected death in southern Ontario, with emphasis on myocarditis. Can Med Assoc J 1979;120:676-80, 706
   Web of Science | Medline

3. 3

   Burlo P, Comino A, Di Gioia V, Passarino G, Mollo F. La miocardite dell'adulto in un ospedale generale: osservazioni su 605 autopsie. Pathologica 1995;87:646-649
   Medline

4. 4

   Sinagra G, Maras P, D'Ambrosio A, et al. Polimorfismo clinico di presentazione e storia naturale della miocardite attiva: esperienza su 60 casi. G Ital Cardiol 1997;27:758-774
   Medline

5. 5

   Reiss N, el-Banayosy A, Posival H, Morshuis M, Minami K, Korfer R. Management of acute fulminant myocarditis using circulatory support systems. Artif Organs 1996;20:964-970
   CrossRef | Web of Science | Medline

6. 6

   Korn EL, Graubard BI, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale. Am J Epidemiol 1997;145:72-80
   Web of Science | Medline

7. 7

   Karliner JS. Fulminant myocarditis. N Engl J Med 2000;342:734-735
   Full Text | Web of Science | Medline

To the Editor:

In the study by McCarthy et al., patients with fulminant myocarditis had a better long-term, transplantation-free survival than those with acute myocarditis, despite a more severe clinical and histologic presentation initially (87 percent of the patients in the fulminant-myocarditis group had active myocarditis, vs. 52 percent of those in the acute-myocarditis group). It would be important to know the results of a subanalysis of outcomes according to the histologic classification (borderline or active myocarditis) in the two groups (patients with fulminant myocarditis and those with acute myocarditis), particularly since the two groups received different treatments. The results of such a subanalysis were not provided in the article, nor was the issue of histologic differences addressed.

The authors hypothesized that patients with fulminant myocarditis (those who are very sick at presentation) may have a better long-term prognosis than patients with acute myocarditis. However, multivariate analysis revealed that long-term, transplantation-free survival was correlated with a higher cardiac output and lower mean pulmonary arterial pressures at base line, which were characteristics of the patients who were less ill at presentation. These data seem to contradict the authors' conclusion.

Finally, the authors do not discuss the possible causes of such strikingly different outcomes in the two groups of patients with myocarditis. It is unclear whether the patients with fulminant myocarditis were simply less genetically predisposed to myocardial damage, had been infected with a less virulent virus or strain, or had less endothelial dysfunction and coronary steal than the patients with acute myocarditis. But what strikes us is that the two groups received different treatments at base line; the group with fulminant myocarditis was treated much more aggressively (with higher doses of pressors and mechanical circulatory support).

Arsad Karcic, M.D.
Arnold R. Conrad, M.D.
Nassau County Medical Center, East Meadow, NY 11554

To the Editor:

McCarthy and colleagues state that a clinical picture suggestive of fulminant myocarditis would prompt them to start aggressive therapy. This statement raises an important issue that is not further addressed in their article. What is the appropriate approach to clinically severe myocarditis, besides supportive treatment?

The first prerequisite for treating a disease successfully is to identify the underlying inciting and perpetuating mechanisms. Since myocarditis is an inflammatory disease in which infectious agents and immunologic disorders go hand in hand, this is an extremely difficult task. Unless diagnostic tools are refined and a causative microbiologic agent (e.g., Toxoplasma gondii, cytomegalovirus, hepatitis C virus, borrelia, or influenzavirus) can be readily identified, treatment will be aimed at an unknown intruder and will have discouraging results.

On the other hand, apart from the findings of small studies in children,1 there is no convincing evidence that immunosuppressive drugs are effective in patients with myocarditis. The conclusions drawn from the results of the Myocarditis Treatment Trial2 were limited by the heterogeneity of the immunosuppressive regimens used and diagnostic inaccuracy. Therefore, this study did not close the discussion of whether immunosuppressive therapy may be valuable in patients with autoreactive forms of myocarditis. Prospective, randomized treatment trials, such as the European Study of Epidemiology and Treatment of Cardiac Inflammatory Disease,3 should be performed.

Viviane Conraads, M.D.
University Hospital Antwerp, B-2650 Edegem, Belgium

3 References

1. 1

   Kleinert S, Weintraub RG, Wilkinson JL, Chow CW. Myocarditis in children with dilated cardiomyopathy: incidence and outcome after dual therapy immunosuppression. J Heart Lung Transplant 1997;16:1248-1254
   Web of Science | Medline

2. 2

   Mason JW, O'Connell JB, Herskowitz A, et al. A clinical trial of immunosuppressive therapy for myocarditis. N Engl J Med 1995;333:269-275
   Full Text | Web of Science | Medline

3. 3

   Maisch B, Hufnagel G, Schonian U, Hengstenberg C. The European Study of Epidemiology and Treatment of Cardiac Inflammatory Disease (ESETCID). Eur Heart J 1995;16:Suppl O:173-175
   Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: Khairy and Infante-Rivard raise the issue of selection bias in our study, which showed that patients with fulminant myocarditis had a higher rate of long-term survival than those with acute myocarditis. We did not include cases of myocarditis discovered at autopsy, since we performed endomyocardial biopsy in all patients with fulminant congestive heart failure, even if they were severely ill. Our hospital autopsy records contain data on only two patients with myocarditis that was not diagnosed while they were alive. Both patients had serious coexisting conditions (e.g., sepsis) and did not present with fulminant congestive heart failure. The autopsy studies cited by Khairy and Infante-Rivard reflect the natural history of untreated myocarditis and obscure the implications for managing the disorder when it is diagnosed during life.

As our data indicate, fulminant myocarditis is rare, yet our clinicopathological scheme has important implications for survival and transplantation. Because the disorder is rare, we evaluated survival in all patients and inferred vital status from a search of the National Death Index for fewer than half the patients.

Karcic and Conrad raise the question of whether the severity of myocarditis contributes to the outcome. In two analyses of the effect of histologic severity — an analysis of survival according to whether the myocarditis was classified as acute or borderline and an analysis in which the severity of inflammation was included in our Cox model — we found no prognostic influence, raising questions about the focal nature of myocardial inflammation and the importance of paracrine or autocrine factors. Our Cox proportional-hazards model also included age and hemodynamic indexes of cardiovascular status. The fact that our survival analysis remained robust even when the independent contributions of cardiac output and pulmonary pressures were accounted for provides further support for our conclusion that patients with fulminant myocarditis have a better long-term survival than those with acute myocarditis. The results of other recent studies are consistent with our findings.1,2

We agree with Conraads that pathophysiologic mechanisms represent a critical area for investigation and that genetic variation, specific viruses, or both are plausible candidate mechanisms. Left-censoring selection bias can be considered in future epidemiologic studies only if warranted by new pathophysiologic insights. Although our study did not address the issue of immunosuppression in different forms of myocarditis, we hope that investigators will incorporate a clinicopathological classification of myocarditis, which has obvious prognostic value, in future studies. Failure to do so, particularly given the rarity of these conditions, may hinder the identification of effective treatments for myocarditis, an effort we support.3

Joshua M. Hare, M.D.
Robert E. McCarthy, III, M.D.
Kenneth L. Baughman, M.D.
Johns Hopkins Hospital, Baltimore, MD 21287

3 References

1. 1

   Marelli D, Laks H, Amsel B, et al. Temporary mechanical support with the BVS 5000 assist device during treatment of acute myocarditis. J Card Surg 1997;12:55-59
   CrossRef | Web of Science | Medline

2. 2

   Chen JM, Spanier TB, Gonzalez JJ, et al. Improved survival in patients with acute myocarditis using external pulsatile mechanical ventricular assistance. J Heart Lung Transplant 1999;18:351-357
   CrossRef | Web of Science | Medline

3. 3

   Hrobon P, Kuntz KM, Hare JM. Should endomyocardial biopsy be performed for detection of myocarditis? A decision analytic approach. J Heart Lung Transplant 1998;17:479-486
   Web of Science | Medline