Hand, foot and mouth disease (HFMD) is an acute infectious disease caused by enterovirus infection, with high incidence and high infectivity (1–3). HFMD is more common in children, especially in children under 5 years of age (4). According to reports by the Chinese National Center for Disease Control and Prevention (https://www.chinacdc.cn), the annual incidence and mortality rates of HFMD in China has been among the top three class C infectious diseases from 2014 to 2019. HFMD has become a major threat to public health and a considerable economic burden.

Coxsackievirus A16 (CV-A16) and enterovirus A71 (EV-A71) represent the predominant pathogens responsible for HFMD. However, the prevalence of other pathogens, such as coxsackievirus A6 (CV-A6) and coxsackievirus A10 (CV-A10), is increasingly recognized (5–7). HFMD is mostly mild and self-limiting, presenting with fever, macular papules or blisters on hands, feet and buttocks, and ulcers in the oral mucosa (8). In severe cases, children with HFMD may develop neurological, respiratory, and circulatory systems, precipitating rapid clinical deterioration. Critical complications include neurogenic pulmonary edema (NPE), aseptic meningitis, brainstem encephalitis, myocarditis and cardiopulmonary failure (4). These complications are associated with accelerated disease progression and significantly elevated mortality. NPE is the leading cause of death among children with HFMD, with most fatalities occurring within 12–18 h of symptom onset. Persistent neurological sequelae are frequently observed among survivors (9, 10).

As a major infectious public health problem in China, the prevention and control system of HFMD still faces severe challenges. At present, no specific antiviral treatment plan has been established for enterovirus infections (11, 12). Although the EV-A71 inactivated vaccine has been marketed in China and included in the immunization program, limited by the inherent high variability of RNA viruses and the lack of cross-protective effects of the EV-A71 vaccine against other serotypes of enteroviruses, the existing vaccines are difficult to achieve long-term broad-spectrum immune defense (13). Therefore, developing multivalent combination vaccines to cover the main prevalent strains and designing new antiviral drugs targeting the key links of viral replication have become the core research directions for safeguarding the health of children and maintaining public health security.

At present, the mechanism of enterovirus infection, pathological characteristics, and the development of vaccines and antiviral drugs remain the focuses of research on HFMD. Advances in pediatric medicine and vaccine deployment are altering the morbidity and mortality patterns of HFMD in children (14). Nevertheless, many questions remain unanswered regarding the epidemiology, etiology, and pathogenesis of HFMD in children. To better characterize the clinical features and causes of death among children with severe HFMD, a retrospective analysis was conducted on 31 fatal cases managed in the pediatric intensive care units (PICUs) of Children's Hospital Affiliated to Zhengzhou University.

HFMD remains a serious threat to children's health as a common infectious disease in China. Although often considered a mild and self-limiting disease usually resolving within 7 days, a subset of children with HFMD progress to severe cases with high mortality. HFMD, a viral infectious disease, lacks specific treatment (8). Early identification of the clinical manifestations of fatal cases is a key measure to reduce the mortality of severe HFMD. Our study analyzed the comprehensive clinical features of children who died from HFMD.

The susceptibility and severity of HFMD are age-related (18). Children under 5 years of age constitute the most affected group, accounting for more than 90% of cases (2). Furthermore, children under 2 years of age are at higher risk for serious complications and mortality; and 84.02% of deaths occurred in this group (2). Our findings align with these findings. All cases were children under 5 years old with an average age of 23.13 months. The highest proportion of cases was 58.06% from 12 months to 36 months, and 24 cases (77.42%) were under 2 years old. Increased susceptibility in young children may relate to cluster epidemics and immature immune protection. Epidemiological data show that morbidity and mortality are greater in younger children (19). The median hospital stay was 1.25 (0.35, 6.00) days. The median critical illness time point, mechanical ventilation time and disease duration were 3.00 (2.00, 4.00) days, 1.15 (0.48, 6.00) days and 4.50 (3.50, 8.00) days, respectively. These results suggest that the disease progresses rapidly and that the disease duration is short. Most fatalities occurred within one week. In addition, none of the fatal cases had underlying disease.

Timely and accurate identification of risk signs and symptoms during disease progression is particularly important for the diagnosis and treatment of severe HFMD. Identification and proper treatment can effectively reduce the mortality (4). Therefore, our study provided a more detailed description of the signs and symptoms. All patients had clinical manifestations of fever, rash, and dyspnea. Relevant studies have shown that a fever exceeding 39°C for more than 3 days is closely associated with severe or fatal HFMD (4, 20). In our study, high fever was the main form of fever, and 83.87% of the children had a history of fever for more than 3 days. In addition, disturbance of consciousness, abnormal heart rate, CRT ≥ 3 s, limb tremors, vomiting and hypotonia were the common clinical manifestations. Disturbance of consciousness, limb tremors, vomiting and hypotonia all indicate nervous system involvement and poor prognosis. These clinical manifestations indicate more extensive or severe neurological damage.

EV-A71 and CV-A16 are the most commonly reported causative agents of HFMD. Other enteroviruses, such as CV-A6 and CV-A10, currently account for a significant proportion of children with HFMD, and their epidemiological influence is expanding (21, 22). EV-A71 infection is the leading cause of death and an independent risk factor for death from severe HFMD (6, 23). Consistently, the most common pathogen in our fatal cases was EV-A71, with a positive rate of 87.10%. EV-A71 is one of the most important neurotropic viruses, and the mechanism of its induction of neurological deficits remains unclear. EV-A71 infection of intestinal epithelial cells progresses to viremia, which may invade the central nervous system later by disrupting the blood-brain barrier and retrograde axonal transport (6). Human SCARB2 (24) and CXCL1-CXCR2 signaling pathway are involved in the neurological function impairment of EV-A71, and CXCL1-CXCR2 signaling pathway has been confirmed to be related to the severity of neurological complications (25).

In September 2016, Henan Province officially offered EV-71 vaccination to children. With the development and application of the EV-A71 vaccine, the morbidity and mortality patterns of HFMD have changed (14). Our hospital is the national children's regional medical center, which undertakes the diagnosis and treatment of most critically ill children in Henan Province and some of the neighboring provinces, which can better reflect the regional disease epidemic situation. Therefore, we analyzed the morbidity and mortality of children with HFMD treated in our hospital from 2014 to 2019. The results revealed that with the continuous improvement of diagnosis and treatment experience and the promotion of vaccines, the acute mortality of children with severe HFMD has gradually decreased. However, the potential for possible short-term or long-term sequelae remains a concern. The long-term sequelae of severe HFMD primarily affect preschoolers under 5 years of age, a critical stage in their growth and development. In recent years, the molecular epidemiology of HFMD pathogens has changed, potentially leading to outbreaks, and the development of pan-enterovirus vaccines is a possible solution (26).

Children with HFMD may have increased WBC counts, elevated of liver or myocardial enzymes, and abnormal blood gas (4). Our laboratory results showed increased WBC counts dominated by neutrophils and mostly elevated PCT. Abnormal liver or myocardial enzymes were common. Thirteen patients had abnormal pulmonary function, with significantly reduced PaO2/FiO2 ratios. Several studies have shown that WBC, BNP, and PCT can be used as early biomarkers of severe or fatal HFMD and are positively correlated with disease severity (27–29). Severe HFMD can lead to rapid cardiopulmonary failure, and BNP has been widely used in the diagnosis of cardiopulmonary failure and the prognosis of cardiovascular diseases (27, 28).

Our study revealed that more than half of the pediatric patients exhibited coinfection with enteroviruses and other pathogens, with adenoviruses being the most common. Relevant studies have shown that both adenovirus and influenza A virus can cause serious damage to the central nervous system (30–32). However, potential interactions between these viruses and enteroviruses, especially EV-A71, in promoting severe central nervous system complications remain unknown. One patient had Pseudomonas aeruginosa coinfection, likely related to long-term mechanical ventilation. He was mechanically ventilated for 62 h and was hospitalized for 125 days. The correlation between multipathogen coinfection and the severity of HFMD needs further study. Coinfection is closely related to the severity and complications of the disease, and the coinfection of different enteroviruses may enhance the pathogenic effect of the disease (33). Indeed, there have been reports describing the synergistic effects among coinfected pathogens that can lead to pathogenic mechanisms (34), and also promote the emergence of atypical clinical manifestations and disease severity (35–38). Therefore, monitoring coinfection is crucial for understanding atypical presentations and disease severity in children with HFMD.

In this study, NPE and brainstem encephalitis were the main causes of death in children with HFMD, consistent with prior studies (4, 9). Enterovirus invasion of the central nervous system can cause brain edema and injury, further inducing a systemic inflammatory response and immune damage, and eventually leading to NPE and other serious complications (9, 10, 39). NPE is one of the most serious central nervous system complications and a leading cause of rapid death in children with severe HFMD, and its development is closely linked to central nervous system inflammation and cytokine storms (6, 14, 40). NPE progresses rapidly, carries high mortality, and survivors face significant neurological sequelae risk (9). Lijun Peng pointed out that drowsiness, vomiting, tachycardia, hypertension, respiratory rhythm change, and limb tremor were risk factors for NPE in children with severe HFMD, and approximately 90% of patients died within 12 h (10). NPE progresses rapidly and has a very high mortality rate. Even those who survive are at great risk of neurological sequelae (9). Previous studies have shown that NSE and S100 are highly expressed in neurons and can be used as sensitive markers of brain injury (41, 42). These markers are significantly upregulated in inflammatory processes such as trauma, infection, heat and stress and play a role in inflammation regulation (43, 44). We found that NSE and S100 were elevated in children with HFMD. It is urgent for pediatricians to identify these potential risk factors early and take countermeasures to reduce mortality in children with severe HFMD.

There are several limitations in this study. First, the sample size of this study is small, and the comparison of children who died from neurogenic pulmonary edema vs. those with brainstem encephalitis. It is still necessary to collect relevant cases for basic comparative analyses. Second, the quality of the historical imaging data (2014–2019) utilized in this study is constrained by the acquisition parameters and storage technologies available at the time. Consequently, these images possess inherent limitations in resolution that could not be retrospectively enhanced. Third, the coinfection in children with HFMD awaits further research. Further studies are needed to clarify the potential synergistic mechanisms among pathogens and their impact on the severity of the disease. Forth, some clinical indexes in the fatal cases may not be detected, which affects further study of the pathogenesis of these fatal cases. Fifth, the serology markers of hepatitis, e.g., HBV surface antigens, were missing, which may be helpful for explaining the elevated AST and ALT levels in our cohort (45–47).

Fatal HFMD predominantly affects children, especially those under 36 months of age. Mortality rate of HFMD has declined in recent years. EV-A71 is the primary pathogen of death from HFMD. Neurological complications were predominant, and the main causes of death were NPE and brainstem encephalitis. Children with brainstem encephalitis had longer hospital stay and mechanical ventilation dependence. In comparison, NPE and pulmonary hemorrhage progressed more rapidly, with shorter disease duration, and these children died within 7 days of symptom onset.