|Home | About | Journals | Submit | Contact Us | Français|
Sudden infant death (SID) is an unresolved problem of high relevance and previous studies have indicated a role of viral heart infections. The diagnosis remains difficult in clinical practice using routine diagnostic tests and must be substantially improved.
A prospective study based on post-mortem samples from SID victims whose heart disease was not clinically recognized was conducted for 4 years in a Tunisian University Hospital. Pediatric cases of unnatural death served as controls. Both SID victims and controls were investigated for possible coxsackievirus-B (CV-B) infection in heart tissue.
During the study period, 39 cases with a male predominance (77%) were reported. There was no positive family history of coronary artery disease among the victims. In 35 cases (90%), low birth weight and/or critical development period were reported. All SID victims had complained of mild fever and insomnia for a few days preceding death, which required infectious laboratory investigations marked with an elevated white blood cell count (WBC) and C-reactive protein (CRP). The cardiac biomarkers were also elevated. The histopathological investigations of the heart tissue samples revealed signs of myocardial and pericardial inflammation. Enterovirus was detected by immunohistochemistry (IHC) and PCR from myocardial samples from 6 cases (15.3%) having myocarditis and 3 cases (7.7%) having perimyocarditis.
The current study is of great interest and is aimed at urging health professionals to adopt systematically long intensive heart care in infants with underlying vulnerability as well as new diagnostic approaches including histopathology complemented with IHC and molecularpathology.
Sudden death is mainly defined as rapid, unexpected and natural death of an individual who appears healthy but dies suddenly within a short time due to a pre-existing disease or a functional disorder. Most often, sudden unexpected death due to natural causes results from previously unknown cardiovascular diseases though extra cardiac causes should not be ruled out [1-3]. New medical research has reported that infants have a higher risk of sudden unexpected death . Ruling out extra cardiac causes of death essentially cerebral, pulmonary and other factors such as prematurity and exogenous stressors, infants died suddenly due to silent cardiovascular infections [5, 6]. Infectious heart diseases include a group of entities involving the heart wall, mainly myocarditis and pericarditis. Myocarditis is clinically defined as an inflammatory myocardial disease . Patients who have suffered from a heart attack may develop pericarditis pathologically defined as an acute inflammation of the pericardium . The coexistence of acute myocarditis and pericarditis is not uncommon since both are commonly caused by cardiotropic viruses. The two terms (perimyocarditis) and (myopericarditis) are used to describe the disease. While perimyocarditis implies predominant myocardial involvement and myopericarditis implies predominant pericardial involvement, both terms are used interchangeably [9, 10]. Recent clinical studies and anecdotal communications reported on different viruses. Among them, CV-B, a small nonenveloped single-stranded, positive-sense RNA virus in the Picornaviridae family and Enterovirus genus, has been implicated in 25% to 40% of acute myocarditis and dilated cardiomyopathy cases in infants and young adolescents .
In this framework, the current study is aimed at investigating into how accurately CV-B heart infection is involved in SID by using a multitude of tests: conventional histopathology for the detection of inflammatory infiltrates and necrosis sites, molecularpathology for the investigation of the enteroviral genomic RNA (conserved sequences and VP1-capsid-protein coding region) and modern IHC to reveal the enteroviral VP1-capsid-protein and immune inflammatory markers (CD3-T and CD19-B-lymphocytes).
The present study reported 39 SID victims (study group) 30 males and 9 females aged 3 to 9 months. The postmortem samples, heart necropsies and -when present- pericardial fluids were obtained from 2010 to 2014. Some clinical information including low birth weight and critical development period were reported by the victims’ families. All SID victims had complained of mild fever and insomnia for a few days prior to death requiring infectious laboratory investigations marked with an elevated WBC count and CRP. The cardiac biomarkers were also elevated. In terms of susceptibility, there was no positive family history of coronary artery disease and the maternal smoking factor was excluded. A total of 17 cases of unnatural death home accidents all males aged 2 to 11 months (control group) were used as controls (Table 1).
In this prospective study, the 39 SID victims were classified as follows: in 3 cases, myocarditis revealed histologically during autopsy was undoubtedly fatal (non-SIDS). The 36 remaining victims did not present any significant pathological changes and were categorized as SIDS.
An autopsy was performed at the Forensic Medicine Departments of Fattouma Bourguiba University Hospital (Monastir) and Farhat Hached University Hospital (Sousse). Five necropsies (1 cm3) were taken from each of the three standardized heart locations (the right ventricle, the septum and the left ventricle i.e, fifteen necropsies from each victim's heart). The necropsies were divided into two categories: in each case, 9 of the 15 necropsies were fixed in formalin and embedded in paraffin for histopathology and IHC while the remaining 6 necropsies were frozen at −80°C for enteroviral genome amplification. The postmortem samples (necropsies and pericardial fluids) were obtained by a forensic doctor 24 hours after death and were sent to our Department to detect viral infectious agents. The research protocol related to the SID cases was referred to the Ethics and Medical Research Committees which gave their approval. As for the samples, they were taken in compliance with the Tunisian law (Act 91- 22; March 25th 1991) pertaining to human organ removal and transplantation.
Neutral buffered formaldehyde 30% diluted to 1/10 was used as fixative. The first category of necropsies was fixed for 24 hours and embedded in paraffin. The sections (5μm) were cut from the paraffin-embedded tissues with a microtome. All sections were stained with hematoxylin eosin (Invitrogen: Vermont, USA) and the slides were investigated for myocarditis according to the Dallas criteria [6, 9-11].
The viral RNA was extracted from pericardial fluids and frozen myocardial tissues (second category of necropsies) using TRIzol® Plus RNA Purification Kit (Invitrogen) according to the manufacturer's instructions. DNase treatment during RNA purification was adopted using PureLink™ DNase (Invitrogen) in order to obtain DNA-free total RNA. The total extracted RNA was investigated for enteroviral genome (conserved sequences and VP1-capsid-protein coding region).
For the conserved sequences, a fragment of 155 bp of the extracted RNA was amplified by one-step Reverse Transcriptase-PCR (RT-PCR) (Invitrogen SuperScript™ One-Step RT-PCR with Platinum® Taq) using 006 and 007 primers  directed to the conserved sequences in the 5'-UTR of the enterovirus genome. The RT-PCR was performed on a mixture containing 25 μl 2X reaction mix (a buffer containing 0.4 mM of each dNTP, 2.4 mM MgSO4), 0.2 μM each of sense and anti-sense primers, 1 μl enzyme mix (RT/ Platinum® Taq; invitrogen), and RNase free water to 50 μl. The reaction was conducted with an initial reverse transcription step at 42°C for 30 min, followed by PCR activation at 94°C for 5 min, 30 amplification cycles (94°C, 30 sec; 42°C, 1 min; 72°C, 2 min) and a final 10-min extension at 72°C in an Eppendorf Mastercycler Thermal Cycler. The RT-PCR products were run on a 2% agarose gel stained with ethidium bromide and visualized under UV light.
Enterovirus PCR from VP1- capsid protein region was conducted in two stages as follows: (1) first stage-10 μl of previously extracted RNA were reverse-transcribed into cDNA at 42°C for 45 min using 200 units of Super-ScriptIII reverse transcriptase and 2.5 ng/μl of random primers (Invitrogen, Cergy Pontoise, France) in the presence of 10 units of RnaseOUT recombinant RNase inhibitor (Invitrogen, Cergy Pontoise, France) and (2) second stage- 5 μl of cDNA were amplified using 50 pmol of the 292 and 222 primers directed to the enteroviral VP1-capsid-protein coding region (Table 2) and 1.25 units of Platinum Taq DNA polymerase (Invitrogen, Cergy Pontoise, France) in 50 μl of reaction mixture according to the protocol . A band of the expected size of 357 bp was observed after agarose gel electrophoresis.
To avoid false-positivity from contaminations, negative controls were performed in all experiments. CV-B3 prototype (Nancy) infected cells were used as positive controls.
The enteroviral VP1-PCR-products were sequenced in three stages as follows: the amplicons were firstly purified using the ExoSAP-IT –PCR Clean-Up Reagent (USB® Products from Affymetrix, Inc) which constitutes a one-step enzymatic cleanup of PCR products. Then, they were sequenced in forward and reverse directions with the respective PCR primers. Finally, the chromatogram sequencing files were inspected with FinchTV (version 1.4.0). The obtained sequences were compared to the corresponding ones available in the GenBank using the Basic Local Alignment Search Tool (BLAST) in order to identify the CV-B serotype [14, 15].
For the detection of CD3-T and CD19-B-lymphocytes (DAKO: Vermont, USA) and the enteroviral VP1-capsid-protein Mab 5-D8/1 (DAKO: Vermont, USA), IHC was performed on all the necropsies previously fixed in formaldehyde and embedded in paraffin. Tris-buffered NaCl solution with tween 20, the target retrieval solution, serum-free protein block, antibody diluent, mayer's hematoxylin, EnVision+ system-HRP (AEC) and glycergel® mounting medium aqueous were all purchased from DAKO (Vermont, USA). The Immunohistochemical procedures included antigen exposure, blocking, incubation with primary antibody, incubation with the secondary antibody in the En-Vision detection system, and appropriate wash between steps using Tris-buffered saline with tween 20. All incubations were carried out at room temperature [16-19]. Briefly, paraffin-embedded tissue sections (5 μm) were dewaxed with xylene and rehydrated with graded ethanol. Antigen exposure was achieved by heat in a water bath (95-99°C) mediated by target retrieval. The endogenous peroxidase activity was blocked with a peroxidase-blocking reagent for 15 minutes. The tissue sections were blocked with the protein block for 10 minutes and then incubated with primary antibody (appropriately diluted to 1:100-1:500 in antibody diluent) for 30 minutes and washed. The secondary antibody in the En-Vision detection system is the goat anti-mouse Ig conjugated with dextran polymer, on which many peroxidase molecules were labeled. The sections were incubated with this reagent for 30 minutes, washed, and then reacted with substrate chromogen for 5-10 minutes. The slides were immersed in aqueous hematoxylin for counterstaining. The color reaction was stopped by a wash in distilled water. Finally, the mounted sections were examined and confirmed under a Nikon Eclipse 50i microscope. In all tests, a negative control was performed and the primary antibody was replaced with an antibody diluent. CD3-T-lymphocytes and CD19-B- lymphocytes were counted in 20 randomized high-power fields (HPF; magnification x400), and the mean value was calculated.
With routine autopsy, all internal organs were investigated histologically. In the study group, pericardial effusion and ascites were seen in only 3 cases suggesting infection as a cause of death. In these cases, the myocardium exams showed dispersed foci of disturbed architecture of muscle fibers. No relevant cardiac pathological findings were seen in the other cases. The controls were completely negative with no relevant cardiac pathological findings.
In both groups, the autopsy showed no abnormalities in the kidneys, brain, liver and lungs, which excluded extra-cardiac causes of death.
The histopathological diagnosis of myocarditis according to Dallas criteria was applied in all heart necropsy samples taken from both the study group and controls. Severe myocardial necrosis, myocytes degenerations and large inflammatory infiltrates were revealed in the three cases suggesting heart infection at autopsy. These inflammation markers, which confirm active myocarditis, were seen in all myocardial samples taken from different locations (the right ventricle, the septum and the left ventricle) (Figure 1A,B,C,D,E and F). The slides from the paraffin tissue blocks from the controls showed no significant pathological findings (Figure 1G and H).
Enteroviral RNA was amplified with RT-PCR using primers located in the conserved region of enteroviral genome in 9 (23 %) cases of the frozen myocardial tissues obtained from 39 SID victims (figure 2). In 3 of the 39 cases, where pericardial effusion and ascites were seen at autopsy, the genomic RNA was amplified from pericardial fluid samples additionally to the heart necropsies. However, of the heart necropsies samples obtained from 19 SID victims with extra-cardiac cause of death (control group), none was RT-PCR positive.
From the 9 enterovirus-RT-PCR-positive cases, the extracted RNA was reinvestigated with primers directed to the enterovirus coding region (VP1-capsid-protein). Figure 3 shows an example of detection by agarose gel electrophoresis of the VP1-PCR amplification products. The sequence analysis of VP1-PCR amplicons confirmed the presence of CVB-3 serotype (figure 4) by comparison with CVB-3 prototype (Nancy) sequence.
In investigated paraffin heart tissue blocks, CD3-T-lympcytes and CD19-B-lymphocytes were detected for all the nine RT-PCR positive cases. Three of these positive cases presented more than 10 lymphocytes per HPF (>10 lymphocytes/HPF) and six cases 5 to 9 lymphocytes per HPF. They (T and B lymphocytes) were found in diffuse distribution in the 3 fatal cases with pericardial effusion and ascites seen at autopsy and in microfocal accumulations in the six other SID positive cases (Figure 5A and B).
Within the study group, the 9 positive cases were associated with a viral protein synthesis activity suggesting a confluent invasion of enteroviral VP1-capsid-protein (Figure 5C and D). Neither CD3-T-lymphocytes, CD19-B-lymphocytes nor enteroviral VP1-capsid-protein were found in myocardial slides of the control group (figure 5E and F). In the nine RT-PCR enterovirus positive cases confirmed with immunohistochemical inflammatory markers, only three SID cases exhibited severe myocardial necrosis, myocytes degeneration and large inflammatory infiltrates with histopathological diagnosis of myocarditis according to Dallas criteria.
The present study, conducted for the first time in a cohort of Tunisian sudden infant death victims, gives evidence of virus-induced heart infections. The diagnosis of CV-B3 myocardial affections as a causative agent of death represents a high incidence (23%) in the coastal district of Tunisia. The current study is intended to illustrate that a combination of methods provides a powerful tool to detect the viral etiology of heart infections [20-22]. It underscores the importance of histological investigations at autopsy and complementary techniques, especially histopathology, IHC and molecularpathology.
Over a 4-year period, 39 consecutive SID victims aged 1 to 9 months were studied and the findings could have a marked impact on the quality of future investigations in such cases.
More importantly, the present results and the investigation methods suggested herein would enable pathologists to better understand and elucidate the cause of death.
As viral infections are frequently found in children who have died from other unrelated causes, the detection of a virus in a sudden death infant does not necessarily imply that this virus is the cause of death. The link between viral infection and death must be evidenced by the presence of severe pathological changes .
Regarding the time-dependent course of viral myocarditis, as studied in the mouse model, early virus-induced myocardial damage already takes place before the histopathological signs of myocarditis defined by the Dallas criteria can be observed [24, 25]. This is confirmed in our study where a viral heart infection was suggested as a causative agent of death only in the 3 fatal cases according to routine histological investigations at autopsy. However, when supplemented with H&E staining, IHC analysis and molecularpathology assays, death due to CV-B3 heart infections was confirmed in 6 other cases.
Some pertinent data were obtained from studies conducted on sudden deaths in order to isolate viral agents. The most comprehensive study which was directed specifically toward possible viral agent detected enterovirus in 12 of 48 necropsies (25%). Other studies identified viral involvement only in 0.4%. There was no apparent explanation for this discrepancy. It is possible that the high incidence of viral agents isolated in a series may simply reflect the prevalence of these viruses in the community at the time of the study and may not be directly related to the death of these infants. Reporting an incidence of 23%, our findings concur with various other studies, but lack agreement with others [26-34].
Two major findings emerge from this population study which examined the prevalence of SID due to CV-B heart infection. First, we found that the postmortem histological examination at autopsy performed by pathologists remains crucial although insensitive and limited. It allows the diagnosis of advanced cardiomyopathies, whether congestive, hypertrophic, restrictive, valvular or congenital. They also facilitate the diagnosis of inflammatory myocarditis and help necropsy sites particularly in difficult cases of myocarditis or inflammatory cardiomyopathies. Second, we found a decrease in unascertained sudden death rate when histological limitations were overcome using a comprehensive combination of methods.
Up to now, the underlying physiopathogeny and immunopathogeny of myocardial damage in humans has remained partially unsolved. However, several studies reported a cardiomyocyte cytopathic effect of enteroviruses in vitro and in animal models . It was demonstrated in vitro that enteroviral protease 2A cleaves and therefore functionally impairs dystrophin, a cytoskeletal protein of cardiomyocytes. During infection with CV-B3, the dystrophin-glycoprotein complex becomes therefore disrupted and the sarcolemmal integrity is lost . Thus, the detection of the enteroviral genomic RNA within the myocardium is itself considered as a pathological finding.
In many recent studies, the common Coxsackie Adenoviral Receptor (CAR) and Decay Accelerating Factor (DAF) have been discovered . Surprisingly, it was established in vitro (cell culture) and using flow cytometry that these receptors were essential for the expression of enteroviral infections . Both CAR and DAF receptors were found to be expressed in cases of chronic inflammatory cardiomyopathies and pancreatitis in adults [39, 40]. Therefore, this observation offers a feasible molecular basis for the explanation of the observed predominance of the above mentioned viruses in the heart.
However, for the CV-B heart infections investigated herein, the physiopathogenesis still remain elusive and is a matter of great scientific interest. However, in clinical practice, it is recognized that non cardiotropic viruses are also incriminated in such heart infections [41-43]. Viral heart infection incidence has been diversely described in the literature. The reported significant differences stem from several factors including the studied samples, the population study (age and gender), seasonal variation, the methods used to detect the infectious agent, result confirmation, correlation with clinical and histopathological findings, and the time of illness when the sample was collected. Furthermore, recent studies have demonstrated a correlation between outcome of a coxsackievirus infection and immune status of the host. SID victims are often premature infants and die mainly between 8 and 16 weeks of age, as the level of maternal antibodies declines .
The absence of abnormalities seen at autopsies in the spleen, kidneys, brain, liver and lungs, indicate an isolated affection of the myocardium in SID victims. This is substantiated by the fact that all heart necropsies from the control group were totally negative for CV-B. Other previous studies with histological investigation of myocardial lesions in cases of SID based on conventional staining revealed only unspecific findings , and the diagnosis is often unreliable because of considerable interobserver variability .
In conclusion, the present study, based on sudden unexpected death in infancy conducted on the coastal region of Tunisia, suggests that the cardiotropic CV-B3 may contribute significantly to sudden death due to myocardial affection. The combined investigations, starting with histology conducted at autopsy by a pathologist, then histopathology based on the Dallas criteria, molecularpathology for the genomic RNA detection and IHC for viral infectious agent and inflammatory marker detection, have improved our ability to diagnose viral heart infections in a rapid and specific way. These findings may be of major interest for the development of further therapies or preventive strategies in the prevention of the development of cardiovascular infections and inflammations. Undertaking a deep clinical assessment of the members of the family and a genetic postmortem analysis of the deceased subject would be very useful in sparing the family members the same affliction.
Further studies are required to determine other virus serotypes involved in these heart infections. Prospective murine experimental myocarditis may be desirable and may have a marked impact on the quality of future investigations to better understand and elucidate the physio and immuno-pathogenesis of such infectious heart diseases.
Histological investigations revealed perimyocarditis as a cause of death in3 cases.
But, using combined investigations, death was confirmed in 6 other myocarditis cases.
Cardiotropic CV-B3 may contribute significantly to sudden death.
The present study was supported by the National Institute of Health grant (HL108371).
The authors are grateful to Mr. Adel Rdissi for proof-reading the paper.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.