Levels of peripheral blood routine, biochemical and coagulation parameters in patients with hemorrhagic fever with renal syndrome and their relationship with prognosis: an observational cohort study

Hantavirus can cause two diseases in humans: Hemorrhagic fever with renal syndrome (HFRS) and Hantavirus Cardiopulmonary Syndrome (HCPS) [1]. HCPS is mainly caused by Sin Nombre virus (SNV), which is prevalent in the Americas, with a high mortality rate of more than 30%-50% [1]; HFRS is mainly caused by Hantaan virus (HTNV), Seoul virus (SEOV), Puumala virus (PUUV) and other viral serotypes of the Hantavirus [2]. HFRS caused by PUUV infection usually has mild manifestation and is mainly prevalent in Europe, also known as epidemic nephropathy (NE), with a fatality rate of about 0.1% [3]. HFRS in China is mainly caused by HTNV and SEOV infection, and the patients infected by HTNV are more serious and the fatality rate can reach 5%-10% [4].

The pathogenesis of HFRS is complex and has not been fully elucidated. Currently, it has been confirmed that hantavirus infection can cause increased vascular endothelial permeability, inflammatory cytokine storm, severe thrombocytopenia and coagulation disorder [4, 5]. The refractory shock, hemorrhage related complications, disseminated intravascular coagulation (DIC) and multiple organ dysfunction syndrome (MODS) have been considered as major causes of death in the severe HFRS patients [4, 6].

There is still no specific treatment for HFRS. Early detection, early diagnosis and early symptomatic support treatment are still the main treatment principles. The pathophysiology of HFRS in humans is complex and the determinants associated with mortality, especially the coagulation and fibrinolysis disorders, are still not been fully elucidated. Based upon the background, we aimed to observe the levels of peripheral blood routine, biochemical and coagulation parameters during the early stage, so as to find independent risk factors closely related to the prognosis. Furthermore, the dynamic changes of coagulation and fibrinolysis parameters were observed and analyzed, which will also be benefit to providing theoretical basis for targeted evaluation and intervention.

In this study, we explored the predictors for prognosis based on the routine blood chemistry and hematological parameters from patients infected with HTNV on admission and the dynamic changes of the coagulation and fibrinolysis parameters within 1 week after admission.

The kidney is the major organ involved in HFRS. This study showed that there was no significant difference in the level of SCr between the survival group and the death group. Some studies even showed that the level of SCr in the dead patients was lower than that of the survival patients [8]. The reason may be that in the AKI patients without other serious complications, some will regularly receive blood purification therapy, which may be helpful to clear urea nitrogen and SCr effectively [8, 9]. Furthermore, this study also showed different degree of electrolyte disturbance on admission, such as hypokalemia and hyponatremia, while their levels were not correlated with prognosis.

Previous studies have demonstrated that the liver function injury can be appeared in the patients with HFRS [10‐12]. This study also showed that most HFRS patients manifested abnormal levels of ALT and AST on admission. High level of AST and ALT were correlated with prognosis, and the level of AST was also higher than that of ALT, which may be related to the location of AST. In addition to the release by injury liver cell, multiple organ injuries such as myocardium, muscle, kidney and pancreas can also cause the release of AST. Among them, the myocardial cell injury has a greater impact on the elevation of AST, and the viral myocarditis and heart failure caused by Hantavirus can reflect the severity of the patient's disease and may be directly related to the prognosis [13]. This study also showed that hyperbilirubinemia was more common in the death group, suggesting that severe HFRS patients may have obvious decrease in the liver metabolic function. Furthermore, most of the coagulation factors can be synthesized in the liver, and their half-life is short and degradation is fast, so the liver injury can further aggravate coagulation and fibrinolysis dysfunction.

Overexpression of inflammatory cytokines is usually involved in the pathophysiological process of the HFRS patients [2, 4]. The increased expression and imbalance of cytokines and chemokines further aggravate the vascular endothelial injury, which in turn leads to increased permeability [4]. Our results showed that the level of WBC increased during the acute stage. The increasing of WBC was correlated with the prognosis and the level of WBC in the dead cases can reach to more than 30 × 10⁹/L, suggesting that the early severe inflammatory cytokines storm might play an important role in the pathogenesis of HFRS.

Coagulation disorders are one of the main pathophysiological changes in HFRS, with hemorrhage as common clinical manifestation. The mechanism of bleeding in HFRS has not been fully elucidated. Studies have shown that coagulation and inflammation are mutually causal, and together as part of the innate immune response [14], which can be activated during viral infection. Inflammation itself can also promote thrombosis through various mechanisms, finally enabling the unlimited formation of microvascular thrombus, and ultimately cause DIC and multiple organ failure [15]. A study of PUUV infection showed that 28% of the HFRS patients developed DIC during the early stage of the disease [16]. Another study on HCPS patients showed that DIC was also an important clinical manifestation [17]. In this study, we used ISTH DIC unified scoring system to analyze and found that 27.30% of the HFRS patients enrolled manifested DIC on admission, which was correlated with the prognosis (death), and was similar with previous study [18]. This result indicates that extensive coagulation activation can occur early in the course of HFRS, demonstrating rapid disease progress and is closely related to the prognosis, which requiring early intervention.

Thrombocytopenia is also a major factor in the disorder of coagulation mechanism of HFRS. Current studies suggest that decreased PLT is one of the factors involved in the vascular endothelial injury and increased permeability [19]. A reduction in PLT on admission had been shown to help predict disease severity in HCPS patients [20]. This study showed that most HFRS patients have significant decrease of PLT early in the course of the disease, and severe PLT reduction was more common in the dead cases. Until now, the mechanism of PLT reduction in HFRS is still unclear. Some studies suggested that the reduction of PLT was related to its continuous consumption [21]; Others indicated that the reduction of PLT was mainly related to adhesion to the surface of endothelial cells infected with hantavirus [22].

This study showed that PT and APTT can reach the maximum on the first day after admission. Prolongation of PT and APTT reflects the activation of coagulation factors, and prolonged PT can be considered as a predictor of death, suggesting that coagulation activation was more prominent in the dead patients, which was consistent with the study in HCPS [23]. The continuous activation and consumption of coagulation factors and PLT may indicate an increase in the body's production of thrombin. Thrombin is a "double-edged sword", an appropriate amount can promote blood coagulation, and an excessive amount may aggravate the activation of the inflammatory response, thereby increasing the permeability of blood vessels [24]. PLT activation and blood coagulation processes corelated each other [25]. Previous studies have found that increased thrombin generation in the patients with PUUV infection and SNV infection can predict the severity of the disease, and the increase in thrombin was more obviously in critically ill patients [26, 27].

As an acute phase protein mainly produced by liver, Fib is a bridge for platelets aggregation and a target of thrombin, playing an important role throughout blood coagulation. This study showed that the reduction of Fib was closely related to the disease severity, which was consistent with the results of some previous studies [28], while there were also studies with different result [16]. In a clinical study of the patients with PUUV infection, the median level of Fib was 4.2 g/L [16], which was much higher than the result in our study. The reason may be related to the different viral serotypes. The majority of the cases enrolled in this study were HTNV-infected patients, whose viral pathogenicity was significantly higher than that of the patients infected by PUUV, and the degree of coagulation and fibrinolysis activation caused by HTNV infection may be stronger. In addition to persistent consumption of Fib, the decreased capacity of liver synthesis will also cause further reduction of Fib. Fib was strongly associated with poor prognosis in HFRS patients. Therefore, clinical attention should be paid to the dynamic monitoring and evaluation of Fib.

In addition to coagulation activation, fibrinolytic activity is also activated in the HFRS patients. The fibrinolytic process is a series of protease-catalyzed reactions, and the final product is plasmin. Plasmin can degrade microthrombi generated by coagulation activation and prevent intravascular fibrin deposition, while uncontrolled plasmin production can also cause systemic hemorrhage within minutes [29, 30]. Fibrinogen degradation products and fibrin degradation products are as all referred to as FDP, and D-dimer is the specific degradation products of the cross-linked fibrin. Previous studies have shown that HFRS patients with HTNV infection had low expression of plasminogen and high expression of FDP [8, 10]. This study also showed that the levels of D-dimer and FDP were increased on admission, while there was no significant correlation with the prognosis. This result was consistent with previous studies on other hantavirus infections (PUUV and SNV) [31, 32].

Different hantavirus serotypes may cause different degrees of fibrinolytic activation in the body. Studies have shown that the median D-dimer in the patients with PUUV infection is 4.8 mg/L [16], and 1.82 mg/L in the HCPS patients [31]. The possible reason is that if a large number of thrombus forms in the body but does not dissolve, the increase in D-dimer may be mild. It can be seen that low fibrinolysis may be more common in the patients with HCPS, which was associated with less severe bleeding complications in the HCPS patients, while hypercoagulability and severe organ dysfunction were the more common manifestations in HCPS [33]. In contrast to HCPS, hyperfibrinolysis was obvious in the patients with PUUV infection. Some studies have shown that increased fibrinolysis activation in the patients with PUUV infection during the acute stage, and excessive fibrinolysis in PUUV patients can reduce the formation of microthrombi, and have a preventive effect on organ dysfunction and may be benefit to clinical recovery [16], Another study has shown that increased fibrinolysis was associated with the increased bleeding risk in the patients with PUUV infection [34]. Other studies also shown that patients with high fibrinolytic activation are prone to have bleeding complications, but the incidence of organ dysfunction is low, and patients with low fibrinolytic activation are more prone to have organ dysfunction [35]. These results may largely explain the enhanced bleeding risk observed in some patients with HFRS.

Compared with PUUV infection, patients with HTNV infection have more serious bleeding complications. This study also showed that HTNV-infected HFRS patients have hyperfibrinolysis during the early stage. First, on the 7th day after admission, the levels of PLT, PT, APTT and Fib gradually became normal, while the levels of D-dimer and FDP remained high. In addition, the proportion of D-dimer and FDP abnormalities are higher than PT, APTT, Fib, and more dramatic range of increase. This result may at least partially indicate that fibrinolysis activation was greater than coagulation activation in the patients with HTNV infection during the course of disease progression, tilting the balance towards fibrinolysis. However, in this study, hyperfibrinolysis is not associate with disease prognosis, its relationship with bleeding complications needs further study. Finally, we ventured to speculate that HTNV could directly activate fibrinolysis independently of the coagulation system and it could be a common manifestation of HTNV infection. The problem which need to be emphasized was that this study was retrospective, and some conclusions might be still speculative. The mechanism of hyperfibrinolysis caused by HTNV infection has not been elucidated, and further exploration and research are needed in the future.

This study has some limitations. First, the number of cases enrolled is still relatively small. In recent years, with the continuous updating and development of intensive care and treatment technology, the incidence of HFRS in Xi'an has decreased gradually. Second, some HFRS patients enrolled in this study may have received different treatment prior to admission because of different timing for seeking medical assistance, some laboratory parameters may change rapidly during the disease progression, so comparing their expression at different timing may lead to based error. Third, some parameters (APTT, TT) are limited by their upper limit of detection and may underestimate the differences between the two groups. In addition, this study is a retrospective design and has no validation analysis. Therefore, it is necessary to conduct multicenter, prospective cohort studies with larger and follow-up samples in the future.

Detection of PT, Fib and TBIL on admission is necessary, which might be benefit to early predicting prognosis. It is also important to pay attention to the dynamic coagulation disorders and hyperfibrinolysis during the early stage in the severe HFRS patients.