Risk Factors for Mortality in Severe Fever with Thrombocytopenia Syndrome Patients with Central Nervous System Complications

Background

Severe fever with thrombocytopenia syndrome (SFTS) is a natural infectious disease caused by a virus transmitted by tick bites. The viral pathogen that causes SFTS was first isolated and characterized as a new bunyavirus, designated Dabie bandavirus or SFTS virus (SFTSV), by China’s CDC in 2010. Belonging to the Phenuiviridae family and Bandavirus genus [1], it has been renamed Dabie bandavirus in 2020 [2]. The main clinical manifestations are fever, digestive tract symptoms, lymph node enlargement, leukopenia, and thrombocytopenia. In addition, some severe cases show several organ disorders and nervous system involvement. The prognosis of most patients is good, but the mortality of critically ill patients is high. The overall mortality rate of SFTS reported in China is 16% (95% CI: 0.13–0.19) [3]. In South Korea, mortality is reported as 46% [4]. In recent years, the incidence of SFTS has continued to increase [5], and SFTS cases have been found in 14 provinces in China [6,7]. Currently, most of the reported studies are from China, Japan, and South Korea [8–13], and some studies have confirmed that SFTSV in South Korea and Japan has homology with the Chinese SFTSV [9]. The United States has also reported similar cases caused by the virus [14]. The WHO has listed severe fever with thrombocytopenia syndrome virus (SFTSV) as a priority pathogen requiring urgent attention.

The onset and severity of SFTS are acute and severe, and the main clinical manifestations are fever, chills, anorexia, nausea, vomiting, abdominal pain, diarrhea, lymphadenopathy, leukopenia, and thrombocytopenia. Some cases may be complicated by multiorgan dysfunction, including heart, liver, kidney, and brain dysfunction, as well as coagulation, and several organ disorders may occur in severe cases [15–17]. There is still a lack of effective antiviral drugs and preventive vaccines for this virus. Choi et al and Gai et al summarized the pathogenesis of SFTS as fever, multiple organ disorder, and recovery stages [8,18], believing that severely affected patients would develop multiple organ failure that would eventually lead to mortality. Neurological complications are among the serious complications of SFTS, most of which occur within 5–7 days. The incidence of neurological involvement in SFTS is 3–77% [4,15,18,19], mainly manifested as disturbance of consciousness, indifference, muscle fibrillation, epilepsy, and coma, and presence of these conditions is closely related to mortality [15,18]. A study on the cerebrospinal fluid of patients with SFTS complicated by neurological damage confirmed that the protein and glucose levels in cerebrospinal fluid were normal, and an increase in leukocytes in the cerebrospinal fluid was not common. Only a few patients tested positive for SFTSV, and the levels of MCP-1 and IL-8 in cerebrospinal fluid were higher than in serum. This suggests that these cytokines are very important in the pathogenesis from virus invasion of the nervous system [19]. About 90% of the non-survivors showed indifference, lethargy, muscle tremor, convulsion, coma, and other central nervous system symptoms before mortality. In comparison, 37% of the survivors showed these symptoms, mainly manifested as indifference and lethargy [18]. Neurological symptoms seriously affect the disease progression and prognosis in SFTS patients, a fact that clinicians should pay great attention to. We studied the clinical manifestations, laboratory test results, imaging examination results, and treatment of SFTS patients with neurological impairment to determine the risk factors for mortality, and to predict the disease progression and prognosis of patients with SFTS combined with neurological impairment, for clinicians.

Material and Methods

SAMPLE SIZE:

From 2013 to 2020, 69 SFTS patients with central nervous system complications treated by Shandong Weifang Yidu Central Hospital were included in our study. If a suspected case met both of the following criteria, it was diagnosed as SFTS: 1. The patient had a history of tick bite 2 weeks before the onset of the disease, or worked, lived or travelled in hills, forests, or mountains during the epidemic season (April to June). 2. The patient had fever (most of the patients had temperatures above 38 degrees Celsius), peripheral blood thrombocytopenia (normal range: 100–400×10⁹/L), and decreased white blood cell count (WBC, normal range: 4–10×10⁹/L). Finally, the patient’s blood samples must be positive for SFTSV as detected by an SFTSV nucleic acid quantitative detection kit and polymerase chain reaction fluorescence probe. In the present study, these positive results served to determine that a suspected case would be classified as a confirmed case of SFTS. The detections were carried out in Weifang Center for Disease Control and Prevention. The patients were split into survival and mortality groups according to the clinical course.

Neurological symptoms caused by SFTSV were not defined in the 2010 version of the guidelines for SFTS. We defined and classified these symptoms based on published literature reports and the symptoms observed in clinical diagnosis and treatment [15,18–21]. As a result, SFTSV-associated neurological complications refer specifically to the occurrence of 1 or more symptoms as follows: (1) muscle tremor and limb tremor: in particular, involuntary muscle tremor of the tongue, jaw, or limbs; (2) cognitive dysfunction: decreased orientation, decreased memory and calculation ability, slow reaction time, inability to accurately answer questions, inability to complete simple instructions, inability to do fine movements and aphasia, speech communication disorders, including difficulty in speech and word finding, and/or naming disorders; (3) consciousness disorder: drowsiness, coma, delirium; (4) convulsions or seizures. The patients with these neurological complications in confirmed cases were split into survival and mortality groups. According to the condition of the patients at discharge, if all the indicators were improved or entered the recovery period, the patients were classified into the survival group. The patients who did not survive until discharge were classified into the mortality group. Mortality was determined by the following criteria: cardiac and respiratory arrest, electrocardiogram showing no cardiac electrical activity, dilated pupils, and absence of light reflex.

TREATMENT DATA:

Demographic factors, date of onset/admission/mortality, disease outcome, clinical manifestations, physical check, and laboratory parameters were collected retrospectively. These data were imported into EpiData (the designated database), and exported as an Excel file.

STATISTICAL VALIDATION:

Data were analyzed by SPSS 26.0 (IBM, New York, NY, USA). Categorical data were compared using the Chi-square test or Fisher’s exact test, and the results were expressed as n (%). Quantitative data were compared with the t-test, expressed in the form of χ̄±s. In univariate analysis, variables with statistically significant differences were included in multivariate analysis for binary logistic regression analysis. The test level for statistical significance was α=0.05.

Results

COMPARISON OF BASIC DATA BETWEEN THE MORTALITY GROUP AND THE SURVIVAL GROUP:

Neurological complications occurred in 59 patients in the survival group and 10 patients in the mortality group (Table 1). There was no obvious gender difference between the 2 groups; the male: female patients’ ratio in the survival and mortality group was 25: 34 and 5: 5, respectively. The mean age of the survival group was 64.02±10.8, and that in the mortality group was 71.7±14.91. There was no obvious age difference between the 2 groups. The mean duration of occurrence of encephalitis was 7.47±1.91 in the survival group and 7.2±3.16 in the mortality group; there was no significant difference in encephalitis occurrence between the 2 groups. The mean duration of hospitalization was 12.36±4.04 days in the survival group and 4.25±2.11 days in the mortality group, while the mean course of the disease was 17.98±4.50 days in the survival group and 10.00±2.68 days in the mortality group. These differences between the groups were statistically significant (P<0.01).

COMPARISON OF LABORATORY TEST RESULTS:

The patients in the mortality group had a disease duration of up to 15 days. Therefore, we compared the laboratory measures during the first 2 weeks of disease duration. There was no significant difference in WBC, neutrophils (N), lymphocyte percentage (LY%), lymphocytes (LY), red blood cells (RBC), hemoglobin (HGB), creatinine kinase (CK), C-reactive protein (CRP), chloride (Cl), prothrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), alanine aminotransferase (ALT), total bilirubin (TBIL), direct bilirubin (DBIL), globulin (GLO), gamma glutamyl transferase (GGT), alkaline phosphatase (ALP), and brain natriuretic peptide (BNP) between the 2 groups. In contrast, the differences in platelet count (PLT), AST, LDH, CK-MB, procalcitonin (PCT), potassium (K) and sodium (Na) were statistically significant (P<0.05). The mean values of PLT, K, and Na in the mortality group were significantly lower than those in the survival group, while the mean values of AST, LDH, CK-MB, and PCT were significantly higher than those in survival group (Table 2).

COMPARISON OF CLINICAL SYMPTOMS AND SIGNS:

There were no significant differences between the 2 groups in headache, chills, muscle pain, arthralgia, nausea and vomiting, loss of appetite, abdominal pain, diarrhea, cough, expectoration, melena, hemorrhagic spots, lymphadenopathy, pulmonary rales, hemoptysis, gastrointestinal bleeding, sinus tachycardia, sinus bradycardia, premature beat, conduction block, atrial fibrillation, and t-wave abnormality (Table 3).

MULTIVARIATE LOGISTIC REGRESSION ANALYSIS:

Variables with statistically significant differences in univariate analysis were included in multivariate analysis. With mortality (mortality =1, survival =0) as the dependent variable and age, PLT, AST, LDH, CK-MB, PCT, K, and Na as the independent variables, a binary logistic regression model was established and the step-by-step regression method was adopted. The results are listed in Table 4.

Table 4 indicates that the regression coefficient of the variable PLT was P=0.038 (P<0.05), OR=0.758 (OR<1), indicating that low PLT is an independent risk factor for patient mortality. The probability of mortality decreased by 24.2% for each unit increase in the variable PLT.

The CK-MB regression coefficient was P=0.048 (P<0.05), OR=1.186 (OR>1), indicating that high CK-MB is an independent risk factor for patient mortality. Each unit increase of CK-MB increased the probability of mortality by 1.186 times.

Therefore, low PLT and high CK-MB were found to be independent influencing factors associated with patient mortality, while age, AST, LDH, PCT, K, and Na were not found to be independent influencing factors of patient mortality (Table 4).

Discussion

The pathogen that causes SFTS is Dabie bandavirus, or SFTSV, which is in the family Phenuiviridae and genus Bandavirus. SFTS is a tick-borne infectious disease with a high mortality rate. Dabie bandavirus was first discovered in the Henan and Anhui provinces of China, and was isolated from blood samples of patients in 2010. The virus can cause serious damage to various systems of the human body through various mechanisms of action [1]. In a study of pathogenesis, cytokine storm was found to be an important cause of morbidity. Induction of inflammatory cytokines occurs in the acute stage of infection [22]. Other studies have shown that cerebrospinal fluid in a few patients tested positive for SFTSV, and the levels of MCP-1 and IL-8 in cerebrospinal fluid were obviously greater than those in serum, suggesting that these cytokines are important in the pathogenesis of virus invasion into the nervous system [19]. In addition, excessive production of pro-inflammatory cytokines in SFTS patients may lead to systemic inflammatory response syndrome (SIRS), which may lead to excessive activation of T cells and macrophages, leading to hemophagocytic lymphohistiocytosis (HLH) and other clinical diseases [23]. Due to the limitations of laboratory conditions, we have not been able to detect cytokines in the serum and cerebrospinal fluid of patients with nervous system involvement. This will become our next research direction.

The nervous system symptoms seen in SFTS differ from those seen in other types of viral encephalitis. They include meningitis, muscle tremors, and limb jitter. In serious cases, gradual disturbances of consciousness are seen, such as delirium, sleepiness, and even convulsions and coma. Some of the patients in the mortality group showed early onset of severe disturbance of consciousness before multiple organ failure. At present, many studies have found that neurological symptoms are closely associated with mortality [1,12,17,18,24], indicating that neurological symptoms of SFTS are of great significance in clinical diagnosis and differential diagnosis. The main peripheral nervous system symptoms include muscle tremors and limb jitter, with the muscles of the patient’s lips and hands involuntarily quivering. The main central nervous system symptoms include irritability, delirium, sleepiness, nonsense speech, disorientation, memory loss, abnormal fidgetiness, and insomnia. Speech disorders can also occur, manifested as tongue-tied speech, inarticulate speech, or motor aphasia. Patients may experience heavy lethargy, convulsions, limb spasm symptoms, epilepsy, loss of consciousness, squinting or double vision, occasionally foaming at the mouth, urinary incontinence, a significant rise in heart rate, and blood oxygen saturation decline. A tic, which can be relieved by the application of sedative drugs, can manifest in 2 days; this was seen in 1 case of patient mortality. Another 3 patients in the mortality group had convulsions on the fourth day after onset, with frequent convulsions. Sedative drug treatment was not effective, and these patients fell into coma after frequent convulsions. All 3 of these patients died on the admission day or the next day. Among the surviving patients, there were 3 patients with convulsions. The neurological symptoms of these 3 patients appeared on the seventh, seventh, and eighth day, respectively. The initial neurological symptoms all appeared as muscle fibrillation. Convulsions occurred less frequently, and when they did occur, it was on the second or third day of hospitalization, and they could be relieved by sedative drug treatment after the seizures. Compared with the surviving patients, the convulsions of the non-surviving patients were more frequent in the early onset of the disease, and were difficult to control with drugs. The patients fell into coma after frequent convulsions, while the convulsions of the surviving patients occurred later and less frequently. After the control of convulsions, consciousness could be cleared, or the patient could be aroused from drowsiness. The duration of hospitalization and the course of disease in the mortality group were significantly shorter than those in the survival group, and most patients died within 2 weeks. It was because these severe neurological symptoms occurred in the early stage of the disease that the survival time of the patients was significantly shortened. Fatal neurological complications can significantly affect patient survival.

Regarding other clinical manifestations in patients with neurological complications, in addition to influenza-like symptoms such as fever, chills and muscle soreness, lymph node enlargement, digestive system symptoms and respiratory system symptoms were more frequent, but there was no statistically significant difference between patients who survived and those who did not survive. According to the autopsy reports of the 2 patients who died of SFTS, there was gastrointestinal, abdominal, and pulmonary bleeding, liver necrosis, and pathological changes in the lymph nodes closest to the tick bite area [25–27]. Pulmonary edema, as well as hemorrhage, infarction, and diffuse alveolar injury, also occurred in the lungs of patients who died, suggesting acute lung injury [25]. Pathological results suggested that the location of organ damage was consistent with the clinical symptoms observed. However, there was no mention of brain biopsies in the autopsy reports.

In addition, SFTS can also cause damage to the bone marrow hematopoietic system, and leukocytes and platelets in patients can be significantly reduced. In our study, leukocytes were as low as 0.52×109/L and platelets were as low as 9×109/L. Platelet counts in the mortality group were obviously lower than those in the survival group, and the decrease in platelets is an independent risk factor related to mortality. We found that the decrease in WBC was mainly in lymphocytes and monocytes, unlike the increase caused by most viral infections. Some studies found that bone marrow specimens from patients showed HLH secondary to virus infection [28,29]. Phagocytosis of platelets and lymphocytes by macrophages leads to thrombocytopenia and lymphocytopenia [30], and the degree of HLH may be associated with the disease severity.

In terms of biochemical indexes, AST, LDH, CK, CK-MB, and ALP increased significantly in the 2 groups; the values were 5–10 times higher than the normal range, and AST, LDH, and CK-MB levels in the mortality group were obviously greater than those in the survival group. The increase in these enzymes often suggests serious myocardial injury. It is also a risk factor associated with mortality [8,12,31], and is the main enzyme that causes elevated SFTS. Our study also found that high CK-MB during the first 2 weeks of the disease course is an independent risk factor associated with mortality, suggesting that the heart is another important organ damaged by SFTSV. In SFTS patients, at the same time, we found that the possibility of electrolyte disorder is higher, especially hypokalemia or hyponatremia in patients with digestive tract symptoms severe enough to lead to lack of potassium and sodium intake. This might be a reason for a result confirmed in studies: that hyponatremia can have a significant influence on mortality [15], and hyponatremia can cause hypotonic encephalopathy, which will aggravate the disturbance of consciousness. Severe hypokalemia can cause arrhythmias, which may be the cause of mortality.

The main limitation of this study was that the sample size of the mortality group was not large enough. Moreover, this study was a single-center study, which can only reflect the characteristics of the disease in a single area, so further multi-center studies are needed. In addition, cerebrospinal fluid and MRI results were not available because of the poor cooperation of patients with neurological symptoms, so the diagnosis of neurological complications depended on clinical diagnosis.

Conclusions

In conclusion, we believe that the clinical manifestations of patients with SFTS complicated by central nervous system complications do not differ in surviving and non-surviving patients, but reduced PLT and increased CK-MB during the first 2 weeks of the disease course were independent risk factors related to mortality. The above indicators should be monitored in clinical diagnosis and treatment, which can help clinicians predict a high risk of mortality in advance, to facilitate reasonable and effective treatment.