Journal of Clinical Virology 59 (2014) 12–17
Contents lists available at ScienceDirect
Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv
Clinical progression and predictors of death in patients with severe fever with thrombocytopenia syndrome in China Ning Cui a,1 , Xiao-Lei Bao b,1 , Zhen-Dong Yang a , Qing-Bin Lu c , Chun-Yan Hu d , Li-Yuan Wang d,e , Bing-Jun Wang a , Hong-Yu Wang d,e , Kun Liu d , Chun Yuan a , Xue-Juan Fan a , Zhen Wang a , Lan Zhang a , Xiao-Ai Zhang d , Liang-Ping Hu b , Wei Liu d,∗ , Wu-Chun Cao d,∗ a
The 154 Hospital, People’s Liberation Army, Xinyang 464000, PR China Consulting Center of Biomedical Statistics, Academy of Military Medical Sciences, Beijing 100850, PR China c School of Public Health, Peking University, Beijing 100191, PR China d State Key Laboratory of Pathogen and Biosecurity, Beijing Institute of Microbiology and Epidemiology, Beijing 100071, PR China e Graduate School of Anhui Medical University, Hefei, PR China b
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 14 October 2013 Accepted 21 October 2013 Keywords: SFTS Clinical progression Predictors of death
a b s t r a c t Background: Severe fever with thrombocytopenia syndrome (SFTS) is an emerging infectious disease of which the clinical progression and factors related to death are still unclear. Objective: To identify the clinical progression of SFTS and explore predictors of fatal outcome throughout the disease progress. Study design: A prospective study was performed in a general hospital located in Xinyang city during 2011–2013. Confirmed SFTS patients were recruited and laboratory parameters that were commonly evaluated in clinical practice were collected. The clinical progression was determined based on analysis of dynamic profiles and Friedman’s test. At each clinical stage, the laboratory features that could be used to predict fatal outcome of SFTS patients were identified by stepwise discriminant analysis. Results: Totally 257 survivors and 54 deceased SFTS patients were recruited and the data of 11 clinical and laboratory parameters along their entire disease course were consecutively collected. Three clinical stages (day 1–5 post onset, day 6–11 post onset and day 12 to hospital discharge) were determined based on distinct clinical parameters evaluations. Multivariate discriminant analysis at each clinical stage disclosed the indicators of the fatal outcome as decreased platelet counts at early stage, older age and increased AST level at middle stage, and decreased lymphocyte percentage and increased LDH level at late stage. Conclusions: The significant indicators at three clinical stages could be used to assist identifying the patients with high risk of death. This knowledge might help to perform supportive treatment and avoid fatality. © 2013 Elsevier B.V. All rights reserved.
1. Background Severe fever with thrombocytopenia syndrome (SFTS) is an emerging infectious disease first identified in central China in 2009. Its etiological cause is a novel bunyavirus classified as a new member of genus Phlebovirus, family Bunyaviridae. The major clinical symptoms include fever, thrombocytopenia, leukocytopenia and gastrointestinal symptoms [1–3]. By July 2013, SFTS cases had
∗ Corresponding authors at: State Key Laboratory of Pathogen and Biosecurity, Beijing Institute of Microbiology and Epidemiology, 20 Dong-Da Street, Fengtai District, Beijing 100071, PR China. Tel.: +86 10 63896082; fax: +86 10 63896082. E-mail addresses: [email protected] (W. Liu), [email protected] (W.-C. Cao). 1 These authors contributed equally to this work. 1386-6532/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jcv.2013.10.024
been reported in at least 12 provinces of mainland China, after the expanded surveillance was initiated [4–7]. Outside China, SFTS like or confirmed SFTS patients had been reported in Dubai, United Arab Emirates [8], Missouri, United States [9] and Korea [10]. In Japan, four fatal cases with SFTSV infection were reported recently and none of them had overseas travel history [11], suggesting the wider distribution of this disease than previously thought. Moreover, recent findings of the potential person-to-person transmission through blood contact made this newly emerged zoonosis a severe threat to public health [12–16]. SFTS patients had an extensively wide clinical spectrum, with some experiencing self-limiting clinical course, while approximately 14% of the cases developing fatal outcome. Life-threatening complications that were found to forecast fatal outcome included neurological manifestations, pulmonary hemorrhage,
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
disseminated intravascular coagulation (DIC) and multiple organ failure (MOF). According to the clinical data, no consensus was obtained as to when these severe complications might develop. However, the emergency of these complications were accompanied by elevated levels of the laboratory parameters that were commonly tested in clinical practice. For example, aspartate aminotransferase (AST), alanine transaminase (ALT), and lactate dehydrogenase (LDH) were markers of liver damage, decreased platelet and increased blood coagulation times (activated partial thromboplastin time (APTT) and thrombin time (TT)) were indicative of the coagulation disturbances, while early organ failure could be forecasted based on elevated serum levels of the albumin (ALB), creatine kinase (CK) and LDH. The close monitoring of these laboratory parameters could help to recognize the severe complications in early phase to attain an intensive treatment in clinical practice. To our knowledge, there is only one study focusing on the clinical progression of SFTS [17]. Previous cross-sectional research [18–20], as well as our recent study [21] have been performed to identify the risk factors that might be correlated with disease severity. Incongruent conclusions were obtained due to the limited samples size and the lack of adjustment for potential confounding effects by multivariate analysis in certain studies. 2. Objective The current prospective study was performed on 357confirmed SFTS patients to identify the clinical progression that SFTS patients experienced and to explore predictors of fatal outcome, which could assist clinicians to identify patients with high risk of death in a real time manner. 3. Study design 3.1. Study sites and patients inclusion The study was performed in a hospital designated for SFTS treatment (154 Hospital) in Xinyang administrative district of Henan Province in 2011 and 2012. Xinyang is the most severely endemic area of SFTS in China, where 98.75% of the SFTS cases in Henan Province were reported from 2010 and 2012. The detailed information of the hospital and procedure of recruiting patients had been described in the previous study [21]. Briefly, all clinically diagnosed patients admitted into 154 Hospital were defined by an acute fever with thrombocytopenia and/or leukopenia [4]. Serum samples were collected from all clinically diagnosed patients on admission for laboratory detection. Laboratory measurements of SFTSV RNA by real-time RT-PCR and SFTSV-specific IgM and IgG antibodies by enzyme-linked immunosorbent assay (ELISA) were performed as following described [1]. A laboratory-confirmed SFTS patient was defined as meeting one or more of the following criteria: (1) isolation of SFTSV in cell culture, (2) detection of SFTSV RNA by a molecular method and (3) seroconversion or ≥4-fold increase of antibody titers between two serum samples collected at least 2 weeks apart. All laboratory-confirmed SFTS patients were included in this study. 3.2. Source of information and data collection A medical record review was performed to collect the information on demographic characteristics, symptoms and signs, laboratory test results and treatment regimens during the entire hospitalization. Eleven clinical and laboratory parameters that were most frequently tested and biologically related to the complications of fatal SFTS patients were evaluated prospectively, including body temperature, white blood cell (WBC), lymphocyte
13
percentage, platelet, ALT, AST, LDH, CK, ALB, alkaline phosphatase (ALP) and gamma-glutamyl transpeptidase (GGT). The research protocol was approved by the Human Ethics Committee of the hospital, and all participants provided written informed consent. 3.3. Statistical analysis Descriptive statistics were performed with continuous variables estimated as median and range, and categorical variables summarized as frequencies and proportions. To determine the difference between the fatal and non-fatal group, categorical variables were compared with 2 or Fisher exact tests and continuous variables with two-sample Wilcoxon rank-sum test. In order to determine the distinct disease phases that manifested with the most diversified characteristics, the dynamic data of the eleven parameters mentioned above were tracked from disease onset till hospital discharge. Briefly, three clinical phases were randomly categorized and each evaluated parameter was compared among three phases by Friedman’s test for fatal and non-fatal cases, respectively. The grouping modes which produced highest number of significant differences among three clinical phases were determined as the clinical phases. After the clinical phases were determined, the log10 transformed data were tested by univariate discriminant analysis which included all the demographic characteristics and clinical data for analysis at each clinical stage. Significant variables in the univariate analysis (P < 0.10) were entered into a multivariate discriminant model to determine the appropriate parameters that could classify the subjects into the fatal and non-fatal group with the highest discriminating power; stepwise selection method was used to retain all selected variables in the model. Adequacy of the discriminant model was determined by examining resubstitution error rates and leave-one-out cross-validation error rates at each clinical stage [22]. All statistical analyses were performed using SAS 9.3 software (SAS Institute, Cary, NC) and P < 0.05 was considered statistically significant. 4. Results 4.1. Patient information During April 2011 to July 2013, 357 laboratory-confirmed SFTSV infected patients were enrolled. The median age was 61 years old (range 7–87), and 202 (56.6%) were female. Fifty-four patients died. The age and gender distribution, as well as the duration from disease onset to admission were significantly different between the fatal and non-fatal patients. Regarding clinical therapy, ribavirin and doxycycline administration composed major therapies, which were highly comparable between two groups (Table 1). 4.2. Laboratory parameters profile and clinical phases defining The dynamic patterns of 11 clinical and laboratory parameters were derived for the fatal and non-fatal group, respectively (Fig. 1). When we divided the clinical stages into 0–5, 6–11 and >11 days, a distinct pattern among stages could be observed. As displayed in Fig. 1, at early stage (0–5 days), most of the laboratory parameters deviated slightly from normal value ranges, while lymphocyte percentage, ALB, ALP and GGT remained at normal levels. At middle stage (6–11 days), all laboratory parameters began to deviate from normal level progressively. The ALT, AST, LDH and CK kept the increasing trend, attaining the peaking level at about day 11, while the platelet sustained the decreasing trend, declining to a nadir level at day 11. The original normal ALB, ALP and GGT levels deviated slightly in fatal cases. In the late stage (>11 days), all of the
14
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
Table 1 Basic information of patients infected with SFTSV. Characteristic Demographic features Gender, female, No. (%) Age, y, median (range) Duration from onset to admission, median (range) Therapy Ribavirin (%) Doxycycline (%) a b c
Total (N = 357)
Non-fatal (n = 303)
Fatal (n = 54)
P value
202 (56.6) 61 (7, 87) 5 (0, 12)
180 (59.4) 60 (7, 87) 5 (0, 12)
22 (40.7) 66 (34, 85) 6 (2, 11)
0.011a 11 days). 4.3. Factors associated with fatal outcome of SFTS at three stages By univariate analysis, temperature, platelet counts, ALT, AST and CK had statistically significant difference between the fatal and non-fatal group at early stage. At middle stage, nine laboratory parameters (temperature, lymphocyte percentage, platelet counts, ALT, AST, LDH, CK, ALP and GGT) differed significantly between two groups. At late stage, all parameters, except for temperature and WBC, were statistically significant between two groups (Table 3). At each clinical stage, discriminant analysis was performed to explore factors that were associated with the fatal outcome after considering the effects from age, gender and delay duration. The multivariate analysis showed that the factor related to the fatal outcome of SFTS was decreased platelet counts at early stage, older age and increased AST level at middle stage, decreased lymphocyte percentage and increased LDH level at late stage (Table 3). Adequacy of the discriminant analysis was tested by resubstitution and leave-one-out cross-validation. The resubstitution error rates, which measured the proportion of misclassified observations using the same data that were used to derive the discriminant function, were 26.1%, 9.5% and 2.4% at the three stages, respectively. The cross-validation error rates, which evaluated the generalization capacity of the achieved classifiers by computing the proportion of misclassified observations under cross validation, were 26.1%, 23.7% and 5.4% at the three stages, respectively (Table 4). 5. Discussion The current study prospectively observed the fever and laboratory parameters that were commonly evaluated in clinical practice on a cohort of 357 SFTSV infected patients. By valid statistical analysis, three stages were defined, from which the laboratory features that could be used to predict fatal outcome of SFTS patients were identified. In early disease within 5 days post disease onset, most parameters were comparable between two groups; only platelet counts could be used to discriminate the patients at higher risk of fatal outcome. Thereafter, the clinical pathological changes of reduced PLT, and elevated LDH, AST, ALT and CK were significantly increased at middle stage (6–11 days post infection), however, only older age and elevated AST had a significantly predictive value for
determining fatal outcome by applying multivariate analysis. At late stage (day 12 to hospital discharge), decreased lymphocyte percentage and increased LDH level could be used to predict sever cases. This knowledge regarding the clinical progression and risk factors of SFTS associated with death might help initiate supportive treatment and avoid fatality. Our study demonstrated that the measurement indicative of pathological lesions mainly involved the hematological system at early stage, with liver damage involved at middle stage and gradually developed to kidney, muscle and lymphoid system hurt at late stage. In SFTSV infected mice model, kidney and liver were shown to be the major target organ [23], and it’s therefore not surprising that the liver and renal function parameters were identified as the critical predictors of fatal outcome. This also gave clues as to the treatment targeted at clinical practice. Previous studies have identified high viral RNA load in blood at admission, as well as higher serum liver transaminase levels, more pronounced coagulation disturbances (APTT, TT), higher levels of acute phase proteins (phospholipase A, fibrinogen, hepcidin), higher urea nitrogen level, and lower platelet counts as risk factors for death [18–20,24]. The shared features of these observational studies were small sample sizes (9–25 fatal, and 32–67 non-fatal cases) and retrospective nature. In Gai’s study, the risk factors of fatal outcome were elevated serum AST, LDH, CK, and CK fraction, as well as the appearance of central nervous system (CNS) symptoms, hemorrhagic manifestation, DIC, and MOF. This study, although with prospective nature, only compared 11deaths and 48 survivors. The small sample size had severely impaired the reproductivity of the deduced association, especially under the situation of lacking multivariate analysis to consider the complex interaction between the measured variables. The strength of our study is its large sample size. Actually in our univariate analysis, a wide spectrum of laboratory evaluations had produced significant associations with fatal outcome, which were comparable with previous findings. However, the effects from most of the variables were masked after adjustment for confounders and potential interaction effects by the multivariate analysis. Considering the complex correlation among the evaluated parameters, the results derived from multivariate discriminant analysis could be more reliable. According to our recent study performed in the same hospital, older age, elevated levels of LDH and CK that were evaluated on hospitalization were identified to be useful predictors of death [21]. After recruiting more patients and expanding the observation to the whole hospitalization period, different indicators were found that could be used to predict the fatal outcome. These findings further corroborated the necessity of performing prospective study on SFTS patients. Another strength of the current study is its prospective study design. Risk factors were identified in various disease stages, by which characteristics of the high risk patients should be differentiated and get more attention in treatment to prevent fatal outcome. It’s also noted that the clinical phases of the previous study were
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
15
Fig. 1. Dynamic profiles of 11 clinical and laboratory parameters in 357 SFTS patients. Dynamic profiles were delineated using mean and 95% confidence intervals (CIs) of log10 -transformed data of each parameter. Blue lines represent survivors and red lines indicate deceased patients. The dashed lines indicate the normal level of each parameter. The light, median and dark pink represents the early, middle and late stage of clinical progression, respectively. P values were computed by comparing each log10 -transformed parameter between the two groups by GEE models. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
defined in a different way from the current study. According to Gai’s results, the period of 7–13 days after disease onset was considered as a critical stage for SFTS patients. The clinical risk factors for death in this period by univariate logistic regression analysis were identified as CNS symptoms development, hemorrhagic tendency and DIC [17]. Our middle stage, or the deterioration phase, was set at 6–11 days, earlier than the previous study. This was logical since our analysis was mainly based on the laboratory evaluations, which developed earlier than the emergency of clinical syndromes,
therefore the predictors identified in the current study might present more useful as risk factors that could be used to warn fatal outcome. There is concern that treatment strategy could influence the disease outcome. In the studied hospital, highly comparable treatment regimen were administered in two groups, moreover, results from both our study [21] and other research [17] demonstrated no significant effect of ribavirin treatment in determining clinical outcome. Therefore we believe the effect from treatment was minimized.
16
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
Table 2 Evaluation of laboratory parameters among three stages in the fatal and non-fatal group by Friedman’s test. Variable
Non-fatal, median (range)
Temperature (◦ C) WBC (×109 /L) Lymphocyte (%) Platelet (×109 /L) ALT (U/L) AST (U/L) LDH (U/L) CK (U/L) ALB (g/L) ALP (U/L) GGT (U/L)
Fatal, median (range)
Stage I
Stage II
Stage III
P value
Stage I
Stage II
Stage III
Pvalue
37.0 (34.4, 39.5) 2.6 (0.7, 21.5) 25.1 (0.3, 74.4) 82 (11, 260) 37 (9, 380) 56 (14, 697) 270 (40, 1302) 230 (37, 6000) 38 (23, 50) 77 (31, 318) 23 (1, 265)
36.8 (36.2, 39.2) 3.6 (0.7, 12.2) 27.5 (5.6, 63.8) 62 (11, 256) 77 (11, 657) 135 (15, 1134) 510 (98, 1881) 361 (27, 7000) 34 (15, 46) 84 (33, 398) 48 (8, 667)
36.6 (36.2, 37.7) 4.6 (1.8, 20.0) 29.7 (3.3, 52.7) 112 (13, 400) 79 (13, 418) 78 (12, 653) 364 (145, 2145) 128 (21, 3673) 35 (19, 51) 102 (41, 565) 101 (14, 900)
Contents lists available at ScienceDirect
Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv
Clinical progression and predictors of death in patients with severe fever with thrombocytopenia syndrome in China Ning Cui a,1 , Xiao-Lei Bao b,1 , Zhen-Dong Yang a , Qing-Bin Lu c , Chun-Yan Hu d , Li-Yuan Wang d,e , Bing-Jun Wang a , Hong-Yu Wang d,e , Kun Liu d , Chun Yuan a , Xue-Juan Fan a , Zhen Wang a , Lan Zhang a , Xiao-Ai Zhang d , Liang-Ping Hu b , Wei Liu d,∗ , Wu-Chun Cao d,∗ a
The 154 Hospital, People’s Liberation Army, Xinyang 464000, PR China Consulting Center of Biomedical Statistics, Academy of Military Medical Sciences, Beijing 100850, PR China c School of Public Health, Peking University, Beijing 100191, PR China d State Key Laboratory of Pathogen and Biosecurity, Beijing Institute of Microbiology and Epidemiology, Beijing 100071, PR China e Graduate School of Anhui Medical University, Hefei, PR China b
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 14 October 2013 Accepted 21 October 2013 Keywords: SFTS Clinical progression Predictors of death
a b s t r a c t Background: Severe fever with thrombocytopenia syndrome (SFTS) is an emerging infectious disease of which the clinical progression and factors related to death are still unclear. Objective: To identify the clinical progression of SFTS and explore predictors of fatal outcome throughout the disease progress. Study design: A prospective study was performed in a general hospital located in Xinyang city during 2011–2013. Confirmed SFTS patients were recruited and laboratory parameters that were commonly evaluated in clinical practice were collected. The clinical progression was determined based on analysis of dynamic profiles and Friedman’s test. At each clinical stage, the laboratory features that could be used to predict fatal outcome of SFTS patients were identified by stepwise discriminant analysis. Results: Totally 257 survivors and 54 deceased SFTS patients were recruited and the data of 11 clinical and laboratory parameters along their entire disease course were consecutively collected. Three clinical stages (day 1–5 post onset, day 6–11 post onset and day 12 to hospital discharge) were determined based on distinct clinical parameters evaluations. Multivariate discriminant analysis at each clinical stage disclosed the indicators of the fatal outcome as decreased platelet counts at early stage, older age and increased AST level at middle stage, and decreased lymphocyte percentage and increased LDH level at late stage. Conclusions: The significant indicators at three clinical stages could be used to assist identifying the patients with high risk of death. This knowledge might help to perform supportive treatment and avoid fatality. © 2013 Elsevier B.V. All rights reserved.
1. Background Severe fever with thrombocytopenia syndrome (SFTS) is an emerging infectious disease first identified in central China in 2009. Its etiological cause is a novel bunyavirus classified as a new member of genus Phlebovirus, family Bunyaviridae. The major clinical symptoms include fever, thrombocytopenia, leukocytopenia and gastrointestinal symptoms [1–3]. By July 2013, SFTS cases had
∗ Corresponding authors at: State Key Laboratory of Pathogen and Biosecurity, Beijing Institute of Microbiology and Epidemiology, 20 Dong-Da Street, Fengtai District, Beijing 100071, PR China. Tel.: +86 10 63896082; fax: +86 10 63896082. E-mail addresses: [email protected] (W. Liu), [email protected] (W.-C. Cao). 1 These authors contributed equally to this work. 1386-6532/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jcv.2013.10.024
been reported in at least 12 provinces of mainland China, after the expanded surveillance was initiated [4–7]. Outside China, SFTS like or confirmed SFTS patients had been reported in Dubai, United Arab Emirates [8], Missouri, United States [9] and Korea [10]. In Japan, four fatal cases with SFTSV infection were reported recently and none of them had overseas travel history [11], suggesting the wider distribution of this disease than previously thought. Moreover, recent findings of the potential person-to-person transmission through blood contact made this newly emerged zoonosis a severe threat to public health [12–16]. SFTS patients had an extensively wide clinical spectrum, with some experiencing self-limiting clinical course, while approximately 14% of the cases developing fatal outcome. Life-threatening complications that were found to forecast fatal outcome included neurological manifestations, pulmonary hemorrhage,
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
disseminated intravascular coagulation (DIC) and multiple organ failure (MOF). According to the clinical data, no consensus was obtained as to when these severe complications might develop. However, the emergency of these complications were accompanied by elevated levels of the laboratory parameters that were commonly tested in clinical practice. For example, aspartate aminotransferase (AST), alanine transaminase (ALT), and lactate dehydrogenase (LDH) were markers of liver damage, decreased platelet and increased blood coagulation times (activated partial thromboplastin time (APTT) and thrombin time (TT)) were indicative of the coagulation disturbances, while early organ failure could be forecasted based on elevated serum levels of the albumin (ALB), creatine kinase (CK) and LDH. The close monitoring of these laboratory parameters could help to recognize the severe complications in early phase to attain an intensive treatment in clinical practice. To our knowledge, there is only one study focusing on the clinical progression of SFTS [17]. Previous cross-sectional research [18–20], as well as our recent study [21] have been performed to identify the risk factors that might be correlated with disease severity. Incongruent conclusions were obtained due to the limited samples size and the lack of adjustment for potential confounding effects by multivariate analysis in certain studies. 2. Objective The current prospective study was performed on 357confirmed SFTS patients to identify the clinical progression that SFTS patients experienced and to explore predictors of fatal outcome, which could assist clinicians to identify patients with high risk of death in a real time manner. 3. Study design 3.1. Study sites and patients inclusion The study was performed in a hospital designated for SFTS treatment (154 Hospital) in Xinyang administrative district of Henan Province in 2011 and 2012. Xinyang is the most severely endemic area of SFTS in China, where 98.75% of the SFTS cases in Henan Province were reported from 2010 and 2012. The detailed information of the hospital and procedure of recruiting patients had been described in the previous study [21]. Briefly, all clinically diagnosed patients admitted into 154 Hospital were defined by an acute fever with thrombocytopenia and/or leukopenia [4]. Serum samples were collected from all clinically diagnosed patients on admission for laboratory detection. Laboratory measurements of SFTSV RNA by real-time RT-PCR and SFTSV-specific IgM and IgG antibodies by enzyme-linked immunosorbent assay (ELISA) were performed as following described [1]. A laboratory-confirmed SFTS patient was defined as meeting one or more of the following criteria: (1) isolation of SFTSV in cell culture, (2) detection of SFTSV RNA by a molecular method and (3) seroconversion or ≥4-fold increase of antibody titers between two serum samples collected at least 2 weeks apart. All laboratory-confirmed SFTS patients were included in this study. 3.2. Source of information and data collection A medical record review was performed to collect the information on demographic characteristics, symptoms and signs, laboratory test results and treatment regimens during the entire hospitalization. Eleven clinical and laboratory parameters that were most frequently tested and biologically related to the complications of fatal SFTS patients were evaluated prospectively, including body temperature, white blood cell (WBC), lymphocyte
13
percentage, platelet, ALT, AST, LDH, CK, ALB, alkaline phosphatase (ALP) and gamma-glutamyl transpeptidase (GGT). The research protocol was approved by the Human Ethics Committee of the hospital, and all participants provided written informed consent. 3.3. Statistical analysis Descriptive statistics were performed with continuous variables estimated as median and range, and categorical variables summarized as frequencies and proportions. To determine the difference between the fatal and non-fatal group, categorical variables were compared with 2 or Fisher exact tests and continuous variables with two-sample Wilcoxon rank-sum test. In order to determine the distinct disease phases that manifested with the most diversified characteristics, the dynamic data of the eleven parameters mentioned above were tracked from disease onset till hospital discharge. Briefly, three clinical phases were randomly categorized and each evaluated parameter was compared among three phases by Friedman’s test for fatal and non-fatal cases, respectively. The grouping modes which produced highest number of significant differences among three clinical phases were determined as the clinical phases. After the clinical phases were determined, the log10 transformed data were tested by univariate discriminant analysis which included all the demographic characteristics and clinical data for analysis at each clinical stage. Significant variables in the univariate analysis (P < 0.10) were entered into a multivariate discriminant model to determine the appropriate parameters that could classify the subjects into the fatal and non-fatal group with the highest discriminating power; stepwise selection method was used to retain all selected variables in the model. Adequacy of the discriminant model was determined by examining resubstitution error rates and leave-one-out cross-validation error rates at each clinical stage [22]. All statistical analyses were performed using SAS 9.3 software (SAS Institute, Cary, NC) and P < 0.05 was considered statistically significant. 4. Results 4.1. Patient information During April 2011 to July 2013, 357 laboratory-confirmed SFTSV infected patients were enrolled. The median age was 61 years old (range 7–87), and 202 (56.6%) were female. Fifty-four patients died. The age and gender distribution, as well as the duration from disease onset to admission were significantly different between the fatal and non-fatal patients. Regarding clinical therapy, ribavirin and doxycycline administration composed major therapies, which were highly comparable between two groups (Table 1). 4.2. Laboratory parameters profile and clinical phases defining The dynamic patterns of 11 clinical and laboratory parameters were derived for the fatal and non-fatal group, respectively (Fig. 1). When we divided the clinical stages into 0–5, 6–11 and >11 days, a distinct pattern among stages could be observed. As displayed in Fig. 1, at early stage (0–5 days), most of the laboratory parameters deviated slightly from normal value ranges, while lymphocyte percentage, ALB, ALP and GGT remained at normal levels. At middle stage (6–11 days), all laboratory parameters began to deviate from normal level progressively. The ALT, AST, LDH and CK kept the increasing trend, attaining the peaking level at about day 11, while the platelet sustained the decreasing trend, declining to a nadir level at day 11. The original normal ALB, ALP and GGT levels deviated slightly in fatal cases. In the late stage (>11 days), all of the
14
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
Table 1 Basic information of patients infected with SFTSV. Characteristic Demographic features Gender, female, No. (%) Age, y, median (range) Duration from onset to admission, median (range) Therapy Ribavirin (%) Doxycycline (%) a b c
Total (N = 357)
Non-fatal (n = 303)
Fatal (n = 54)
P value
202 (56.6) 61 (7, 87) 5 (0, 12)
180 (59.4) 60 (7, 87) 5 (0, 12)
22 (40.7) 66 (34, 85) 6 (2, 11)
0.011a 11 days). 4.3. Factors associated with fatal outcome of SFTS at three stages By univariate analysis, temperature, platelet counts, ALT, AST and CK had statistically significant difference between the fatal and non-fatal group at early stage. At middle stage, nine laboratory parameters (temperature, lymphocyte percentage, platelet counts, ALT, AST, LDH, CK, ALP and GGT) differed significantly between two groups. At late stage, all parameters, except for temperature and WBC, were statistically significant between two groups (Table 3). At each clinical stage, discriminant analysis was performed to explore factors that were associated with the fatal outcome after considering the effects from age, gender and delay duration. The multivariate analysis showed that the factor related to the fatal outcome of SFTS was decreased platelet counts at early stage, older age and increased AST level at middle stage, decreased lymphocyte percentage and increased LDH level at late stage (Table 3). Adequacy of the discriminant analysis was tested by resubstitution and leave-one-out cross-validation. The resubstitution error rates, which measured the proportion of misclassified observations using the same data that were used to derive the discriminant function, were 26.1%, 9.5% and 2.4% at the three stages, respectively. The cross-validation error rates, which evaluated the generalization capacity of the achieved classifiers by computing the proportion of misclassified observations under cross validation, were 26.1%, 23.7% and 5.4% at the three stages, respectively (Table 4). 5. Discussion The current study prospectively observed the fever and laboratory parameters that were commonly evaluated in clinical practice on a cohort of 357 SFTSV infected patients. By valid statistical analysis, three stages were defined, from which the laboratory features that could be used to predict fatal outcome of SFTS patients were identified. In early disease within 5 days post disease onset, most parameters were comparable between two groups; only platelet counts could be used to discriminate the patients at higher risk of fatal outcome. Thereafter, the clinical pathological changes of reduced PLT, and elevated LDH, AST, ALT and CK were significantly increased at middle stage (6–11 days post infection), however, only older age and elevated AST had a significantly predictive value for
determining fatal outcome by applying multivariate analysis. At late stage (day 12 to hospital discharge), decreased lymphocyte percentage and increased LDH level could be used to predict sever cases. This knowledge regarding the clinical progression and risk factors of SFTS associated with death might help initiate supportive treatment and avoid fatality. Our study demonstrated that the measurement indicative of pathological lesions mainly involved the hematological system at early stage, with liver damage involved at middle stage and gradually developed to kidney, muscle and lymphoid system hurt at late stage. In SFTSV infected mice model, kidney and liver were shown to be the major target organ [23], and it’s therefore not surprising that the liver and renal function parameters were identified as the critical predictors of fatal outcome. This also gave clues as to the treatment targeted at clinical practice. Previous studies have identified high viral RNA load in blood at admission, as well as higher serum liver transaminase levels, more pronounced coagulation disturbances (APTT, TT), higher levels of acute phase proteins (phospholipase A, fibrinogen, hepcidin), higher urea nitrogen level, and lower platelet counts as risk factors for death [18–20,24]. The shared features of these observational studies were small sample sizes (9–25 fatal, and 32–67 non-fatal cases) and retrospective nature. In Gai’s study, the risk factors of fatal outcome were elevated serum AST, LDH, CK, and CK fraction, as well as the appearance of central nervous system (CNS) symptoms, hemorrhagic manifestation, DIC, and MOF. This study, although with prospective nature, only compared 11deaths and 48 survivors. The small sample size had severely impaired the reproductivity of the deduced association, especially under the situation of lacking multivariate analysis to consider the complex interaction between the measured variables. The strength of our study is its large sample size. Actually in our univariate analysis, a wide spectrum of laboratory evaluations had produced significant associations with fatal outcome, which were comparable with previous findings. However, the effects from most of the variables were masked after adjustment for confounders and potential interaction effects by the multivariate analysis. Considering the complex correlation among the evaluated parameters, the results derived from multivariate discriminant analysis could be more reliable. According to our recent study performed in the same hospital, older age, elevated levels of LDH and CK that were evaluated on hospitalization were identified to be useful predictors of death [21]. After recruiting more patients and expanding the observation to the whole hospitalization period, different indicators were found that could be used to predict the fatal outcome. These findings further corroborated the necessity of performing prospective study on SFTS patients. Another strength of the current study is its prospective study design. Risk factors were identified in various disease stages, by which characteristics of the high risk patients should be differentiated and get more attention in treatment to prevent fatal outcome. It’s also noted that the clinical phases of the previous study were
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
15
Fig. 1. Dynamic profiles of 11 clinical and laboratory parameters in 357 SFTS patients. Dynamic profiles were delineated using mean and 95% confidence intervals (CIs) of log10 -transformed data of each parameter. Blue lines represent survivors and red lines indicate deceased patients. The dashed lines indicate the normal level of each parameter. The light, median and dark pink represents the early, middle and late stage of clinical progression, respectively. P values were computed by comparing each log10 -transformed parameter between the two groups by GEE models. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
defined in a different way from the current study. According to Gai’s results, the period of 7–13 days after disease onset was considered as a critical stage for SFTS patients. The clinical risk factors for death in this period by univariate logistic regression analysis were identified as CNS symptoms development, hemorrhagic tendency and DIC [17]. Our middle stage, or the deterioration phase, was set at 6–11 days, earlier than the previous study. This was logical since our analysis was mainly based on the laboratory evaluations, which developed earlier than the emergency of clinical syndromes,
therefore the predictors identified in the current study might present more useful as risk factors that could be used to warn fatal outcome. There is concern that treatment strategy could influence the disease outcome. In the studied hospital, highly comparable treatment regimen were administered in two groups, moreover, results from both our study [21] and other research [17] demonstrated no significant effect of ribavirin treatment in determining clinical outcome. Therefore we believe the effect from treatment was minimized.
16
N. Cui et al. / Journal of Clinical Virology 59 (2014) 12–17
Table 2 Evaluation of laboratory parameters among three stages in the fatal and non-fatal group by Friedman’s test. Variable
Non-fatal, median (range)
Temperature (◦ C) WBC (×109 /L) Lymphocyte (%) Platelet (×109 /L) ALT (U/L) AST (U/L) LDH (U/L) CK (U/L) ALB (g/L) ALP (U/L) GGT (U/L)
Fatal, median (range)
Stage I
Stage II
Stage III
P value
Stage I
Stage II
Stage III
Pvalue
37.0 (34.4, 39.5) 2.6 (0.7, 21.5) 25.1 (0.3, 74.4) 82 (11, 260) 37 (9, 380) 56 (14, 697) 270 (40, 1302) 230 (37, 6000) 38 (23, 50) 77 (31, 318) 23 (1, 265)
36.8 (36.2, 39.2) 3.6 (0.7, 12.2) 27.5 (5.6, 63.8) 62 (11, 256) 77 (11, 657) 135 (15, 1134) 510 (98, 1881) 361 (27, 7000) 34 (15, 46) 84 (33, 398) 48 (8, 667)
36.6 (36.2, 37.7) 4.6 (1.8, 20.0) 29.7 (3.3, 52.7) 112 (13, 400) 79 (13, 418) 78 (12, 653) 364 (145, 2145) 128 (21, 3673) 35 (19, 51) 102 (41, 565) 101 (14, 900)
