Dengue virus infection is widely distributed in the tropical and subtropical regions of the globe affecting up to 100 million people per year; 2.5 billion people are at risk [1]. The infection is caused by 1 of 4 antigenically distinct but related single stranded RNA viruses in the family Flaviviridae and is transmitted by mosquito vectors, primarily Aedes aegypti [1]. Dengue virus infections cause a spectrum of illnesses ranging from self-limiting fever to severe dengue haemorrhagic fever(DHF) where increased vascular permeability is the main pathology leading to shock [1–3]. In some cases, uncommon complications such as acute hepatic failure, acute renal failure, dengue encephalitis and myocarditis have been recognized [4–7]. In general, many viruses cause myocarditis and their pathogenesis has been described in the literature. However, the dengue virus as a cause of viral myocarditis is not emphasized in most of these reports [8–14].

In Sri Lanka, dengue fever emerged in the early 1980s and all dengue viral serotypes, DEN-1, DEN-2, DEN-3, and DEN-4 have been found in circulation causing DHF [2, 3] The worst recorded outbreak of dengue in the island occurred in 2009 with 24629 cases and 245 deaths reported [15]. During this epidemic, the General Hospital, Peradeniya in the hilly Central Province of the island received a substantial number of dengue cases. Of them, a proportion of patients died with predominant cardiac dysfunction, despite meticulous fluid management in the Intensive Care Unit. Cardiac complications in dengue fever were reported from the same unit in 2005, where 75 patients had ECG changes with cardiac dysfunctions, all patients recovering [4]. However, lack of histological diagnosis from myocardial biopsies was the main concern in this study to confirm dengue virus as a cause of myocarditis [4]. Therefore, in 2009, we performed a limited number of autopsies on patients who died due to dengue complications and the tissue material was subjected to histopathological examination. Material from five cases showed histopathological changes in the myocardium. In this study we aim to describe the histopathological features of the myocardium in these 5 cases and also to describe other organ system involvement and the main clinical manifestations they showed during the illness.

The presence of interstitial oedema, infiltration of inflammatory cells and necrosis of myocardial fibers in the myocardium suggested severe myocarditis in all five cases and these features tally with the clinical evidence of cardiac dysfunction they developed during the illness causing profound hypotension and death. In addition to the heart, cellular changes occurred in organs such as lungs and liver together with evidence of bleeding and effusions in to serosal cavities. This particular outbreak carried a high death rate and only in a few cases was DEN-1 shown to be the causative serotype. By performing autopsies and histology, we attempted to confirm the occurrence of myocarditis in dengue infection as a true manifestation. Cardiac dysfunction could also be explained as the result of an indirect mechanism affecting the myocardium. However, without demonstrating viral RNA in the myocardium, the indirect myocardial damage due to cellular immune responses and/or cytokines cannot be excluded. This question was correctly addressed in a recent study where immunohistochemistry was used to demonstrate direct dengue viral infection in the myocardial tissues causing myocarditis [16]. Therefore, the gross histopathological features detected in the current case series (Figure 2) could be due to direct dengue viral infection. The cardiotropism property of the dengue virus and its contribution to morbidity and mortality needs further evaluation. Due to lack of technology in Sri Lanka, the current autopsy study did not proceed to detect dengue viral RNA or antigen in tissues or staining for T-cells (CD3) or macrophages (CD68) to strengthen the histopathological diagnosis of myocarditis.

The characteristics of dengue infection in Sri Lanka have been changing recently in relation to the severity of infection contributing to high morbidity and mortality [2, 4]. In 2005, an epidemic showed high incidence of myocarditis, in which the diagnosis of myocarditis was made solely on clinical, ECG and enzyme evidence [4]. Published evidence of cardiac dysfunction in dengue based on clinical manifestations, ECG changes, demonstration of reduced cardiac output by echocardiography and using specific cardiac markers are available in the literature [17–19]. The current study has gone one step further to provide histopathological evidence in support of myocarditis in dengue infection.

There are many viruses recognized as aetiological agents of myocarditis. The pathogenesis of myocartitis caused by these viruses and the related complications are well described in the literature [8–14]. In future, the dengue virus should also qualify as a causative agent of viral myocarditis taking the current evidence into consideration. So far in the literature there are no reports documenting the incidence of myocarditis in dengue infections. The reasons could be due to the rarity of dengue myocarditis or its temporal nature of occurrence or due to under-diagnosis. The experience in Sri Lanka supports its temporal nature of occurrence along with an emergence of new causative dengue serotype. In an epidemic in 2005, DEN-3 was identified as the causative dengue serotype of myocarditis [4], whilst in 2009, DEN-1 was found to be the main serotype responsible for both the island wide outbreak and the regional outbreak at Peradeniya [15]. It is interesting to note that in 2009, DEN-1 emerged as the serotype of importance and caused severe secondary infection among the regional population who had been exposed to DEN-3 in 2005. Thus, viraemia would have been more pronounced in the epidemic in 2009 causing multiple organ dysfunction in addition to excessive fluid leak.

It is well known that the dengue virus has the predilection to infect peripheral blood and vasculature more than vital organs, particularly the myocardium [1, 3]. An interesting report published in 2004, used immunohistochemistry and in situ hybridization to localize dengue virus in blood and different human tissues obtained from biopsies and autopsies [20]. They demonstrated the presence of dengue viral antigen in the liver, spleen, lung, kidney and peripheral blood leucocytes, but not in thymus, lymph nodes, thyroid, pancreas, heart, adrenal gland, skeletal muscles, intestine and brain [20]. Even though clinical data are lacking in this paper to understand organ specific clinical manifestations in their patients, it has proved the ability of dengue virus to infect organs. A subsequent study demonstrated direct viral invasion of the myocardium and derangement of calcium storage in the infected cells contributing to damage of myocardial cells [16]. In the current case series, involvement of other organs apart from the heart such as lungs and liver was found. Thus, multiple organ dysfunction and excessive fluid leak as suggested by effusions would have contributed to the final fatal outcome.

In conclusion, primary cardiac failure due to myocarditis in dengue infection is very often overlooked. The current study presents histopathological evidence to support myocarditis in dengue infection. We emphasise the need to consider the possible existence of myocarditis in severely ill patients with dengue infection in clinical settings. Management guidelines for dengue should stress myocarditis as an important issue.