Continuous Blood purification on Influenza-Associated Neurological Disease in children: a retrospective cohort study

Background Due to lack of proven therapies, we evaluated the effect of CBP on Influenza-Associated Neurological Disease in children. Methods A single-center, retrospective, cohort study was conducted in Luoyang, Henan province, China from January 2018 to January 2020. Children (<18 years) with influenza-associated neurological disease were enrolled in the study. Children with indications for CBP and parental consent received CBP (Continuous Blood purification), while others received maximal intensive care treatment because of the absence of parental consent. The outcomes of the CBP and non-CBP groups were compared. Categorical variables were presented as percentage and compared by Chi-square test. Continuous variables were expressed as median (interquartile ranges) and compared with non-parametric independent sample test. Statistical analyses were carried out by SPSS (version 26.0) and p < 0.05 (2 tailed) was considered to be statistically significant. Results 30 children with influenza-associated neurological disease were recruited to the study. 18 received CBP and the other 12 received maximal intensive care. There were no differences between CBP and non-CBP children in age, sex, body weight, type of influenza virus, neurological complications, Glasgow score, PIM-2 score and PCIS at admission (p > 0.05). The inflammatory factors (CRP, PCT and IL-6) of 30 cases were tested at admission and after 3 days of admission. In the CBP group, there was a significant decrease in IL-6 levels at 3 days of admission (p = 0.003) and a decrease in CRP and PCT levels, but no significant difference (p > 0.05). In the non-CBP group, there were no significant difference on levels of CRP, PCT and IL-6 at admission and 3-day of admission (p > 0.05). The 28-day mortality was significantly lower in the CBP group compared with the non-CBP group (11.11% vs. 50%, p = 0.034). Conclusions CBP definitely reduces IL-6 levels significantly. We did find that the survival rate of patients in the CBP group was improved. But we don’t know if there is a relationship between the reduction of IL-6 levels and the survival rate. Trial registration: http://www.chictr.org.cn/index.aspx(ChiCTR2000031754).


Background
Seasonal influenza epidemics cause between 3-5 million cases of severe illness and approximately 290,000 to 600,000 deaths every year in the world [1] . Most people are susceptibility to influenza and self-limited in the population, but some patients with high risk factor, especially for infants can develop to severe complications. These complications include pneumonia, neurological complications and even multi-organ dysfunction. Although pneumonia is the most common complication of childhood influenza, neurological disease is one of the most severe complications [2] . The Influenza-associated neurological disease include seizures, meningitis, transverse myelitis, acute disseminated encephalomyelitis, Guillain-Barre syndrome, and encephalopathy/encephalitis， and the incidence rate of children is significantly higher than that of adults [3] . The Influenza-associated neurological disease is associated with severe neurologic sequelae and high mortality [4] . Except supportive treatment, there is no effective treatment for the Influenza-associated neurological disease until now. The systemic inflammatory response syndrome after influenza virus infection have been shown to play an important role for neurological complications [5] . CBP can remove inflammatory factors and reduce the level of inflammatory factors in the serum of sepsis patients [6] .
It is unknown whether the outcome and prognosis would be improved, if CBP is applied for removing inflammatory factors in influenza children with neurological complications. The aim of our study is to determine whether CBP have superior outcome and prognosis compared with non-CBP in influenza children with neurological complications.

Study design
A single center, retrospective cohort study in Henan, China from January 2018 to January 2020 was conducted. A total of 545 pediatric patients (<18 years old) were admitted to PICU of Luoyang Maternal and Child Health Hospital for diagnosis of influenza. Among these children, the cases with neurological complications were enrolled into this study. The inclusion criteria include:①29d-18y；②Diagnosed with influenza [7] ；③Nervous system abnormalities such as consciousness and convulsions quickly appeared； ④Respiratory support by ventilator.
Exclusion standard： ①The children received cardiopulmonary resuscitation before admission， and the pupils were dilated and fixed, or both pupils were not equal, or the EEG voltage was low (<5Hz)；②The children who gave up treatment due to family members within 72 hours of admission；③There are only frequent convulsions, but no progressive aggravation of consciousness disorder during the course of the disease；④There are high risk factors for bacterial purulent meningitis such as rhinorrhea, middle ear deformity or skull base fracture; ⑤The children who has basic brain disorders such as metabolic encephalopathy or epileptic encephalopathy. A 5F double lumen catheter was used for vessels with diameter≤ 0.6cm, 7F double lumen for vessel between 0.6cm and 1.0cm and 11.5F for vessels with diameter over 1cm.

Outcomes
The primary outcome was 28-day mortality at admission. Secondary outcomes included ICU stay, Glasgow score, PIM-2, PCIS, ventilator-time, hospitalization costs.

Working flow and basic information of influenza children in PICU
The study collected 545 children with influenza admitted to PICU from January 2018 to December 2019 for retrospective analysis. 39 of them had Influenza-associated neurological disease. 9 cases did not meet the inclusion criteria and were excluded。Among these 9 cases, 2 children received cardiopulmonary resuscitation before admission，after admission, the pupils were dilated and fixed or pupils were not large, and the brain voltage was lower than 5 Hz. One child was given up within 72 hours of treatment; 6 children gradually recovered their consciousness after admission and gradually recovered without receiving advanced life support such as ventilator-assisted ventilation.30 children were enrolled in this study. 18 received in CBP group with parents' consent, and other 12 were in non-CBP group due to failure of parents' consent (see Figure 1).
The demographic characters and clinical information at admission were showed as follow.
There were no differences between CBP and non-CBP children in age, sex, body weight, type of influenza virus, neurological complications, Glasgow score, PIM-2 score and critical ill score and the level of Inflammatory factors at admission (p > 0.05). (see Table 1)

Inflammatory factors after CBP
The level of CRP, PCT and IL-6 of serum in CPB group were lower than in non-CBP group at 3-day of admission, there was no significant difference (p > 0.05). In CBP group, the level of IL-6 was significant lower at 3-day of admission than it at admission (p = 0.003), the level of CRP and PCT also decreased at 3-day of admission, but there was no significant difference 6 (p > 0.05). In the non-CBP group, there were no significant difference on level of CRP, PCT and IL-6 between at admission and 3-day of admission (p > 0.05) (see Figure 2).

Outcomes
The 28-day mortality in CBP group was significantly lower than in non-CBP group (11.11% vs 50.00%, p = 0.034). Glasgow score, PIM-2 score and PCIS also significantly improved in CBP group. At the same time, the length of PICU stay was significantly prolonged, and the hospitalization costs were significantly increased (see Table 2). per million, respectively. The median duration of fatal cases from the onset of influenzarelated encephalopathy to death was 1 day [3] .In China ， the Beijing Children's Hospital of flu-related encephalopathy [10] .
Besides serious respiratory complications, some children can quickly progress to coma, and influenza related neurological complications are more severe in children and the main causes of influenza death [3,11,12] . Until now, the pathogenesis of influenza-related neurological complications is still unclear. Autopsy of these patients showed necrosis and epithelial hemorrhage of thalamus and posterior cerebral cap of pons, pale myelin sheath in white matter of brain and cerebellum, and no clear obstruction of vascular endothelium and peripheral vascular edema [13] . Due to the acute onset of influenza-related neurological diseases, severe brain dysfunction can occur quickly. The treatment time window is tight.
Before the symptoms of nervous system appear, the children's respiratory system disorder is not prominent or even no respiratory system performance, and lack of early warning indicators. Although immunization can reduce its incidence to a certain extent [9] , but most children are not actively vaccinated. Once the patient developed into an Influenza-associated neurological disease, treatment is extremely difficult, and some children develop cardiopulmonary failure within a short period of time. Many children have undergone cardiopulmonary resuscitation before admission, and showed pupil light loss or even brain herniation after admission, the brain failure will soon appear even if with active treatment.
Even if they survived, some children will still have serious neurological complications. The mortality of acute necrotizing encephalopathy is 30%, only 10% of patients can fully recover and more patients survive in the form of neurological sequelae. The neurological sequelae recover very slowly and some patients received unacceptable long-term rehabilitation [14] .
Therefore, effective treatment of influenza-associated neurological disease, especially acute necrotizing encephalopathy is a problem that we urgently need to solve. In this study, we found that the mortality of children in the CBP group was lower than that of the non-CBP group, and the neurological coma score (Glasgow score), PCIS score and PIM-2 score were improved compared with the non-CBP group. It is suggested that CBP has association with decreased mortality.
The pathogenesis of influenza-associated neurological disease is unclear. Current opinion was that systemic inflammatory response syndrome triggered by influenza virus infection especially the elevated of cytokines (interleukin-6, -8, -10) were involved. Previous studies found that inflammatory factors significantly increased in the cerebrospinal fluid of patients, especially IL-6 and TNF-α [14] .So, the elevated cytokines (interleukin-6, -8, -10), caused by systemic inflammatory response syndrome triggered by influenza virus infection, have been hypothesized to play an important role in the pathogenesis of neurological complications [11,15] .
However, there is no effective therapy for the "system cytokine storm" caused by influenza virus. Corticosteroid is the most commonly used and the effect is still controversial. The current treatment for influenza-associated neurological disease is symptomatic treatment and immune regulation. Kawashima [16] reported three children with influenza encephalopathy recovered from influenza after hormone and plasma exchange. The CBP model in this subject includes PE and CVVHDF, which can remove inflammatory factors between 0.5k-60 kDa from serum. In this study, the level of inflammatory factors, including CRP, PCT and IL-6, reduced in CBP group after blood purification. IL-6 molecular weight 21-30kD. IL-6 levels were significantly lower after blood purification than at admission.
However, due to the effective number of samples in this study, the mechanism and efficacy of CBP therapy in the treatment of Influenza-associated neurological disease require further research and discussion.

Conclusions
CBP could Reduce inflammatory factors and may reduce 28-day mortality and improve neurologic function of influenza children. Due to the limited samples of this study, the correlation between the improvement of neurological function, mortality and blood purification still needs further well-designed clinical trials.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors have no conflicts of interest or funding to disclose.

Funding
The efficacy and mechanism of continuous blood purification in the treatment of children's