[Analysis of Clinical Characteristics and Prognostic Risk Factors of HLH Children with Central Nervous System Involvement].

Objective: To investigate the clinical characteristics and prognostic risk factors of HLH children with central nervous system (CNS) involvement so as to provide more reference for further improving the prognosis of HLH children.

Methods: The clinical data of 45 HLH children with CNS involvement treated in our hospital from January 2006 to October 2016 were collected and analyzed retrospectively. The clinical characteristics of HLH children with CNS involvement were recorded, moreover the possible factors influencing the prognosis of HLH children with CNS involvement were analyzed using univariate and multivariate analysis through the establishment of Cox risk ratio model.

Results: Among 45 HLH children with CNS involvement, male was 19 cases and female was 26 cases. The median age of 4.0 years old (1.0-15.1). The detection showed that EBV found in 38 cases (84.44%), CMV infection in 1 case (2.22%), bacterial infection in 3 cases (6.67%), connection tissue disease in 1 case (2.22%) and indefinite etiology infection in 2 cases (4.44%). After lumbar puncture of 27 HLH children with CNS involvement, 10 cases (37.04%) showed cerebrospinal fluid abnormality. In addition, 22 cases showed the craniography abnormality. The follow-up results showed that the OS rate of 1 year was 46.67% (21/45), the OS rate of 3 years was 44.44% (20/45); the median survival time was 5.0 months. The OS analysis indicated that 1 years OS rate of diseased children with cerebrospinal fluid abnormality was significantly lower than that of diseased children with cerebrospinal fluid normality (10/45 vs 17/45) (P＜0.05), and 1 years OS rate of diseased children who not received intrathecal injection was significantly lower that of diseased children who received intrathecal (10/45 vs 17/45) (P＜0.05). The univariate analysis showed that the symptoms of nervous system, abnormal cerebrospinal fluid, absence of intrathecal injection and treatment schedule all were the risk factors affecting the prognosis of HLH children with CNS involvement (P＜0.05). The multivariate analysis by Cox risk model showed that abnormal cerebrospinal fluid and absence of intrathecal injection were independent risk factors for of HLH children with CNS involvement (P＜0.05).

Conclusion: The clinical prognosis of HLH children with CNS involvement is relatively poor, moreover some of HLH children with CNS involvement have neural sequelae. The cerebrospinal fluid abnormality and absence of intrathecal injection are independent risk factors leading to poor prognosis for HLH clildren with CNS involvement.