Case report

A 5 month old girl was seen in the department of paediatrics of a municipal hospital because of lethargy and poor feeding. Six days before admission she had a fever, which persisted for five days. Because skin eruptions appeared on her body after the fever subsided, she was diagnosed with exanthem subitum by the attending paediatrician (one day before admission). From the late afternoon of the day before admission, she became lethargic and exhibited poor feeding. Because her condition deteriorated, she was transferred to a municipal hospital. She had previously been healthy, with normal development, and there was no family history of particular note.

On examination, axillary temperature was 36.1°C, pulse was 186 beats/min, and blood pressure was not evaluated. She was unconscious, and her extremities were flaccid. There was no significant lymphadenopathy, hepatosplenomegaly, respiratory tract signs, or murmurs. The skin rash had already disappeared, and the anterior fontanel was flat. Cranial computed tomography gave normal results. The peripheral leucocyte count was 15 200/mm³, with 66% lymphocytes, and the following enzymes were greatly increased: aspartate aminotransferase (228 U/litre; normal range, 10–40), lactate dehydrogenase (2154 U/litre; normal range, 220–430), and creatine kinase (815 U/litre; normal range, 50–210), indicating myocardial damage. Moreover, wide QRS tachycardia was evident on the electrocardiogram. During examinations, she had generalised convulsions and premature ventricular contractions. She was given diazepam and disopyramide phosphate intravenously. Finally, she was diagnosed with acute myocarditis, and transferred to Seirei Hamamatsu General Hospital to receive intensive treatment. However, she experienced cardiopulmonary arrest at the time of arrival. Resuscitation attempts were unsuccessful.

Serological examinations were carried out on a serum sample obtained at the time of necropsy. HHV-6 and human herpesvirus 7 (HHV-7) antibody titres were measured by an immunofluorescent assay, as described previously.^{6, 7} The results of the serological assays were: HHV-6 IgG, < 1 : 8; HHV-6 IgM, 1 : 8; HHV-7 IgG, 1 : 10; and HHV-7 IgM, < 1 : 10.

The postmortem examination disclosed that the main pathological lesion was in the heart. Diffuse infiltration of mononuclear cells and neutrophils, with oedematous change and degeneration and loss of myocardial cells, were seen in the interstitium of the myocardium and pericardial tissue, compatible with typical acute interstitial myocarditis (fig 1). The most characteristic observation was that many of the infiltrating cells in the lesion had apoptotic nuclei, which was confirmed by the TdT mediated dUTP biotin nick end labelling (TUNEL) method using ApoTag (Oncor, Gaithersburg, Maryland, USA), according to the manufacturer's instructions (fig 2). However, no apoptotic changes were seen in myocardial cells. In addition, calcification was occasionally seen in some myocardial cells. No arteriovascular malformation was found. No other abnormal findings were seen in lung, liver, kidneys, spleen, and thymus, except for generalised lymphoadenopathy, with mild to moderate degrees of follicular hyperplasia.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Diffuse infiltration of mononuclear cells and neutrophils is seen in the myocardium with oedematous change and degeneration and loss of myocardial cells (haematoxylin and eosin stained; original magnification, ×200).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Infiltrating mononuclear cells in the myocardium are confirmed to be apoptotic by means of the TUNEL method using an ApoTag peroxidase kit (DAB-peroxidase; original magnification, ×200).

At necropsy, we obtained serum and postmortem tissue specimens including heart, lung, lymph node, thymus, liver, kidney, and spleen, and stored these samples at −80°C. DNA was extracted by phenol/chloroform treatment, followed by ethanol precipitation from the tissue specimens. Total DNA (1 μg, measured spectrophotometrically) extracted from each sample was used for the polymerase chain reaction (PCR). Nested PCR was carried out to detect HHV-6 and HHV-7 DNA in these samples. Details of the procedure for HHV-6 and HHV-7 PCR have been described elsewhere.⁸ Figure 3A shows the PCR results. HHV-6 DNA was detected in heart, liver, kidney, spleen, and lymph node. However, HHV-7 DNA was not detected in these samples (data not shown). Aliquots of Tris/EDTA buffer were processed as negative controls, and appropriate precautions to avoid cross contamination were taken during sample preparation and while performing PCR. Moreover, all specimens were analysed using a primer to the β globin gene to confirm the absence of residual PCR inhibitors in the sample. Reverse transcription PCR was also carried out to detect enteroviral RNA in myocardial tissue, and was found to be negative (data not shown).The HHV-6 variant was identified by the presence or absence of the characteristic HindIII site in the second round PCR product; all products had the site that is indicative of the B variant (fig 3B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

(A) Human herpesvirus 6 (HHV-6) DNA amplified by nested PCR. Positive bands are seen in the lanes indicating heart, liver, kidney, spleen, and lymph node. The β globin gene was amplified in all DNA specimens. (B) Second round PCR products extracted from heart, kidney, liver, spleen, and lymph node were digested in an HHV-6 variant B specific manner by HindIII. Variant B, FG-1 strain isolated from a patient with exanthem subitum in our facility; variant A, GS strain.

We also carried out immunohistochemical analysis to detect HHV-6 antigens in tissue specimens including heart, lung, lymph node, thymus, liver, kidney, and spleen. A variant specific monoclonal antibody reactive with the p101 structural protein of HHV-6 variant B was used for this assay. The antibody, which does not crossreact with the homologous protein of variant A HHV-6, was kindly supplied from Dr PE Pellet (Centers for Disease Control and Prevention, Atlanta, Georgia, USA). Moreover, the antibody was used to detect the virus antigen in postmortem tissues.⁹ Immunohistochemical analysis was performed on paraffin wax embedded sections by an avidin–biotin complex (ABC) peroxidase method. None of these specimens was positive for HHV-6 antigen (data not shown).

Discussion

HHV-6 is now classified into two distinct variants (variants A and B). Exanthema subitum, benign infantile febrile disease, is caused by primary infection with HHV-6 variant B.¹ Although the isolation of HHV-6 was not attempted, our patient demonstrated a typical clinical course of exanthem subitum, and was positive for anti-HHV-6 IgM. Moreover, both infection with HHV-7 (another aetiological agent for exanthem subitum) and enteroviruses was excluded. To identify the HHV-6 variant, we looked for the presence of a variant specific HindIII site in the second round PCR products from several tissues. All second round products would be digested at the appropriate site by the enzyme, indicating HHV-6 variant B. Therefore, we think that the aetiological agent in our patient was HHV-6 variant B.

Prezioso et al reported a 13 month old immunocompetent girl with fatal disseminated HHV-6 infection.¹⁰ According to the postmortem findings, many organs including heart showed parenchymal damage. However, details of the pathological findings of the myocardium in the patient were not clearly described. Moreover, the clinical course of the patient was quite different from our case. In the Prezioso case, it seems that parenchymal damage in the patient's myocardium occurred secondarily to the main pathology caused by disseminated HHV-6 infection, and was not linked directly to the cause of her death. Recently, two children with myocardial damage have been reported, one suffered sudden infant death syndrome¹¹ and the other had encephalitis with subsequent infantile spasm.¹² However, the clinical features of these patients were clearly different from our patient. Our case showed a typical clinical course with pathological findings of acute myocarditis. Myocardial calcification has been seen in acute myocarditis,¹³ and the absence of fibrotic changes excluded chronic myocarditis. The myocardial calcification was therefore considered to be caused by the acute myocarditis. To our knowledge, this is the first report of a fatal case with acute myocarditis, which appears to be caused by primary HHV-6 infection.

Our patient's past history did not suggest the presence of an obvious pre-existing immunodeficiency. Moreover, the patient's thymus and lymphoid tissues appeared to be morphologically adequate. Although productive infection of HHV-6 (HHV-6 antigen positive cells in the myocardial tissue) was not demonstrated, probably because of diminishing viraemia, HHV-6 DNA was detected in several tissues, including the myocardium. We do not know whether these findings are related to unidentified immunosuppression in the host or whether they represent normal occurrences in immunocompetent children. Immunodeficiency caused by direct damage to thymic epithelial cells, which was speculated in the patient reported by Prezioso et al,¹⁰ might be a possible mechanism for the dissemination of viral DNA. Alternatively, the HHV-6 DNA identified in these tissues might be attributable to infiltrating mononuclear cells latently infected with HHV-6. In any event, it is important to elucidate the complete pathophysiology of primary HHV-6 infection. The establishment of an animal model of HHV-6 infection might help to clarify the issue.

Mechanisms by which viral agents cause cardiac damage are still not completely understood; however, direct virus induced myocardial injury and virus induced autoimmune myocardial damage have been postulated. In our patient, it seems that myocarditis occurred just after cessation of HHV-6 viraemia. It is thought that natural killer cell activity and a virus specific immune response were induced at that time. Thus, it is speculated that the host immune response might have an important role in the pathogenesis of myocarditis. Moreover, in our case apoptotic changes were present in infiltrating mononuclear cells and neutrophils. Gebhard et al reported that, in perforin positive mice, which developed severe myocarditis as a result of coxsackie B infection, apoptotic cells were detected within the inflammatory infiltrate.¹⁴ This finding is similar to that of our patient. Although the mechanisms of myocardial damage are not known, the infiltrating apoptotic cells might play an important role in the pathogenesis of the viral myocarditis. To elucidate the precise mechanisms for myocardial damage, further studies are necessary to characterise infiltrating cells, and to detect perforin in myocardial tissues by immunohistochemical analysis. Even though either direct myocardial damage by the viral infection or immune mediated myocardial impairment is involved in the pathogenesis of the myocarditis, it is probable that replication of the virus in the myocardial tissues is important. It is not known why viral antigen was not detected in the patient's myocardial tissues. The investigation of similar cases and pathological analysis using a more sensitive immunohistochemical method to detect the viral antigen is needed to determine whether replication of the virus occurs in the myocardial tissues.