A case–control study of risk factors for HIV-negative children with cryptococcal meningitis in Shi Jiazhuang, China

Cryptococcosis is an emerging disease arising due to the global HIV/AIDS pandemic. Cryptococcus neoformans is the most common cause of cryptococcosis worldwide. Cryptococcosis is responsible for cryptococcal meningitis (CM), the most common fungal infection of the central nervous system and a significant cause of morbidity and mortality among patients with HIV/AIDS [1]. Few studies of pediatric CM have been reported, but CM is believed to be less frequent in children than in adults, regardless of immune status [2]. Recently, higher numbers of CM cases in children have been reported worldwide [3–13]. However, these reports come from small case studies, literature reviews, or studies comprised mostly of adults. In addition, the epidemiologic characteristics, clinical features, laboratory findings, and outcomes for pediatric patients are inconsistent among the different studies. Data on the overall incidence of children with CM are scarce, and understanding which children are at highest risk of CM is lacking.

Here, we analyzed the epidemiologic characteristics, clinical features, laboratory findings, outcomes, and risk factors for CM in pediatric patients without HIV/AIDS. Study subjects were identified by the Acute Meningitis and Encephalitis Syndrome (AMES) surveillance project in Shi Jiazhuang, China.

Here, we analyzed epidemiology, clinical and laboratory characteristics, and risk factors associated with CM due to C. neoformans in HIV-negative children. Similar to the incidence in children of Taiwan with unknown HIV status reported previously [11], we found an average annual incidence of 4.3 cases of CM per million HIV-negative children with a fatality rate of 1.7% in mainland China.

Two species in the genus Cryptococcus, C. neoformans and C. gattii, can infect humans. The former, which is found commonly worldwide, infects mostly immunocompromised patients [14, 15], whereas the latter affects both immunocompetent and immunocompromised individuals [16, 17]. In our study, children without underlying diseases comprised 74% of CM patients caused by C. neoformans, consistent with other reports from mainland China [18]. This finding may be explained by the genetic homogeneity between Chinese C. neoformans strains [18]. However, data demonstrate that C. gattii, which was formerly described as a geographically restricted species [19], is spreading at a great rate around the globe. Cases caused by this species have been reported not only from the tropical and subtropical areas, but also from the Pacific coast of Canada and the United States [20]. Moreover, China, Europe, and Japan confirm that this fungus can also occur in temperate climates [18, 21, 22].

The proportion of immunocompetent patients with CM varies greatly between the HIV-negative patients in different regions and countries [8, 23–28], regardless of the age of patients. For example, in Taiwan, about 50% of HIV-negative children with CM exhibited underlying disease [8, 23] whereas 38% of CM patients in India had no underlying disease [24]; in Vietnam, 81% of the CM patients had no underlying disease [25], but in the United States, about 70% of HIV-negative adults with CM had underlying disease [26]. Subtle defects in immunity [27] or idiopathic CD4 lymphopenia [28] could account for these variations, as could ethnic differences. Further studies are needed to determine the cause of regional variation in the incidence of CM among immunocompetent populations.

However, the epidemiologic characteristics, clinical features, and laboratory findings associated with CM were similar for children with and without underlying diseases. The median age of infected children was about 12 years old, consistent with prior studies of CM in HIV-negative children [8, 9]. Like in other hospital-based reports of CM in children, most of the patients in our study were boys [12, 29]. A population-based surveillance in South Africa also showed that boys are significantly more likely to become infected with cryptococcosis than adults, but it did not evaluate the gender distribution for CM in children [30].

Of the children included in our study, 73.9% of the patients were from rural areas, suggesting that there is increased environmental exposure in rural areas. Similar results were obtained in a population-based study in Australia and New Zealand [4], although few HIV-negative children with CM were included in these studies. Furthermore, about half of the CM cases in our study occurred in summer in Shijiazhuang; the reason for this is unclear. Although time distribution is an important epidemiologic characteristic of diseases, the seasonal distribution of CM in children has not been reported. When compared to children with meningitis caused by other pathogens, a conditional logistic analysis showed that age, gender, and geographic location were not risk factors for CM in pediatric patients. We did not test onset of illness as a risk factor, because it was the matching variable in the case–control study. Thus, further studies are needed to confirm whether or not age, gender, geographic location, and onset of illness are risk factors associated with CM in children.

In our study, fever was the most frequent sign of CM, followed by headache, nausea, altered mental status, and vomiting. Hearing and visual damage were less common during the course of the disease. These frequent characteristics were secondary to increased intracranial pressure (ICP) and consistent with the laboratory findings of ICP. Pleocytosis, elevated CSF proteins, and decreased CSF glucose levels were also common. These clinical and laboratory findings are somewhat different from those previously reported in HIV-negative or HIV-positive children [8, 12, 29]. In HIV-negative children with CM, one study reported a meningeal irritation sign to be the most common manifestation, followed by headache, vomiting, and pleocytosis in 9 HIV-negative children with CM [8]. Yet another study described hearing and visual damage to be common symptoms in 11 HIV-negative children with CM [12]. These different clinical and laboratory findings may be due to different patient populations enrolled in the various studies. The main reason for the difference between HIV-negative or HIV-positive children may be the immune status of the patients. Overall, the characteristics of CM did not differ from those of other forms of meningitis, thus we could confirm the diagnosis of CM only through identification of Cryptococcus in the CSF. The difficulty in diagnosis is likely why patients were misdiagnosed with other forms of meningitis prior to CSF culture.

Although the characteristics of CM are indistinguishable from other forms of meningitis, the treatment may differ. Thus, early identification and prevention are critical for management of CM in the pediatric population. Infection with HIV/AIDS is the most common risk factor for CM in adults, followed by other comorbid conditions, such as cancer, organ transplantation, renal disease, and connective tissue disorders. Furthermore, a population-based multistate study revealed smoking and outdoor occupations are associated with an increased risk of cryptococcosis in adults infected with HIV [3]. However, the risk factors associated with CM in children have not been defined. Our study shows that contact with birds/bird droppings or saprophytes and a longer interval between onset of illness and hospital admission are independent risk factors for CM, but not for other types of meningitis in children. In contrast to children with meningitis due to other pathogens, underlying disease in children with CM is not a significant risk factor. Thus, our findings suggest that contact with birds/bird droppings or saprophytes and a longer interval between onset of illness and hospital admission may be used to distinguish CM from other types of meningitis in pediatric patients.

Although our observations provide valuable insight into pediatric CM, they are not without limitations. The incidence of HIV-negative children with CM was likely underestimated because only fungal culture was used to identify C. neoformans. The traditional diagnostic techniques used to identify this infection include culture analysis, India ink staining of CSF, and cryptococcal polysaccharide antigen detection in the serum and CSF. Lateral flow immunoassay is a new diagnostic method for cryptococcosis, and it shows comparable results to the widely approved enzyme immunoassay for cryptococcal antigen [31]. Although every one of these methods can be used to detect cryptococcosis, a combination of methods may detect more cases with incidences closer to those reported for HIV-negative children with CM in Shi Jiazhuang. Additionally, the sample size used in this study may have limited the power to detect risk factors although our results are statistically significant and support our conclusion that exposure to birds/bird droppings or saprophytes and the interval from illness onset to hospital admission are independent risk factors for CM. However, a larger sample size may allow detection of additional risk factors. Continuous surveillance or a multicenter study would need to be conducted to detect more cases that could be used to assess for additional risk factors. Furthermore, although half of the CM cases emerged in summer, this case–control study lacks a matched variable that would have allowed us to determine if there is a seasonal risk of developing pediatric CM.

Overall, our findings suggest that we should reconsider strategies for managing and preventing CM in children. Instead of simply administering drug prophylaxis, a primary, secondary, and tertiary strategy for prevention and control of CM could be tailored to individuals at high risk for CM. For instance, our results suggest that children should avoid contact with birds/bird droppings and saprophytes, sick children should receive medical care as soon as possible, and clinicians need to consider CM as a potential cause for meningitis especially in boys and in children, who had contact with birds/bird droppings or saprophytes and in children who did not receive prompt medical attention. Furthermore, public health studies are needed to identify and/or develop effective antifungal drugs, vaccines [32], and screening strategies [33].

In summary, we found that although CM is not common in HIV-negative children, the mortality rate associated with CM is high in HIV-negative children. The epidemiology and clinical and laboratory characteristics of CM do not differ between HIV-negative children with underlying diseases and those without underlying diseases. Boys and children from rural areas represented most cases of pediatric CM. Children exposed to birds/bird droppings or saprophytes and children who did not receive prompt medical attention were at high risk for CM, whereas these factors were not associated with risk for other types of meningitis. These findings suggest that avoidance of environmental risk factors and prompt medical attention are important for the prevention and management of CM among children. Clinicians should consider CM as a potential cause for pediatric meningitis in children with factors mentioned above. Further studies are needed to identify the most effective measures to confront CM.