Introduction

Idiopathic scoliosis surgery is a major surgery associated with multiple blood loss, causing loss and shortage of coagulation factors, platelet and fibrinogen. This further induces the disturbance of blood coagulation. All above initiates a vicious circle and results in a deteriorating condition, which presents great difficulties for surgical procedure and forces higher morbidity and mortality on patients. However, from another side, this makes it a good candidate to investigate the bleeding-reducing and blood-saving effect of hemostatics and their influences on patients’ coagulation function. Batroxobin and Tranexamic acid (TXA) can significantly diminish blood loss and transfusion requirements in spine surgeries [1–4]. We assumed that batroxobin and TXA are of different work mechanisms, and, theoretically, could be administrated jointly to exert their own superiority to generate superposition or synergistic effects and achieve a superior hemostatic effect.

Materials and methods

After obtaining approval from the institutional review board (IRB) of our institution and written informed consent from each patient, 80 patients were enrolled in our study. The exclusion criteria are subjects with (1) preexisting cardiac, pulmonary, renal and hepatic disorders; (2) intake of NSAIDs within 7 days before surgery; (3) history of coagulation disorders, Deep vein thrombosis (DVT) or pulmonary embolisms; (4) lower preoperative Hb (<100 g/l); (5) abnormal clotting tests, such as prothrombin time (PT) and platelet counts. All patients were randomly allocated to receive 0.9% saline (group A); batroxobin (group B), 0.02 U/kg intravenously 20 min before skin incision with a maximum of 1 U, followed by a superaddition of the same dosage every 2 h until the end of operation; TXA (group C), a bolus of 20 mg/kg at skin incision, followed by a continuous infusion of 10 mg/kg h during operation; and both two drugs (group D) in above manner.

Total intravenous anesthesia (TIVA) was performed on every patient. The anesthesia was induced with midazolam (0.1–0.15 mg/kg), sufentanil (1.0–2.0 μg/kg), propofol (2–2.5 mg/kg) and rocuronium (0.6–1.2 mg/kg) and maintained with remifentanil (0.2–0.25 μg/kg min) and propofol (4–12 mg/kg h). The anesthetic depth was adjusted according to intraoperative vital signs. Volume ventilation was controlled to maintain PaCO2 between 30 and 35 mmHg. Besides standard monitoring, the arterial blood pressure (ABP) of radial artery was routinely measured and intraoperative mean arterial pressure (MAP) was controlled between 55 and 65 mmHg with intermittent administration of isosorbide dinitrate. Nasal temperature was monitored and maintained at 36.0–37.0°C with a heating apparatus during operation. Fluid therapy (6% hydroxyethyl starch and Ringer’s solution) was adjusted with ABP, urinary output (≥1 ml/kg h) and hourly arterial blood gas measurement results, such as pH value, electrolyte and hemoglobin concentration. All fluid was heated to 37°C by an electronic heating device. The autologous blood recovery equipment (Dideco Electa, Sorin Group, Italy) was used routinely and autologous blood was infused immediately after washing and filtering. The uniform transfusion threshold for allogeneic blood transfusion was 80 g/L after infusion of autologous blood. Frozen fresh plasma (FFP) was infused according to British Committee for Standards in Haematology (BCSH) guidelines and the coagulation function check results. FFP was administrated at least one of the following criterions was matched: (a) PT greater than 1.5 times the mid point of the normal range; (b) APTT greater than 1.5 times the top of the normal range [5]. No coagulant was used postoperatively.

Primary outcomes measured included the amounts of intraoperative blood loss, blood transfusion, FFP and overall drainage. The blood loss was the sum of blood volume from surgical field suction, autotransfusion system reservoirs and weighing sponges from the operative field [6]. The results of Hb, hematocrit (Hct) and platelet (PLT) were documented preoperatively, postoperatively and on the first postoperative day. The coagulation indexes of blood samples were analyzed by the conventional method for prothrombin time (PT), thrombin time (TT), activated partial thromboplastin time (APTT) and fibrinogen (FIB) as well as with the Sonoclot Coagulation and Platelet Function Analyzer (SCA, SIENCO, Inc., USA) for activated clotting time (ACT), clotting rate (CR) and platelet function (PF) preoperatively, postoperatively and 1 day after operation.

Statistical approach

Statistical analysis was performed with SPSS software, version 11.5 (SPSS Inc, Chicago, IL). All measurement data were presented as \( {\bar{\text{X}}} \pm {\text{SD}} \) or ratio. The Kolmogorov–Smirnov test was performed to test the normality of the distribution of the continuous variables. We used the Chi-square test to evaluate percentages and One-Way ANOVA to test normally distributed variables. A p value <0.05 was considered statistically significant.

Results

The demographic data and surgical profiles are shown in Table 1. The male to female ratio, age and weight were comparable in four groups. For the number of fused levels, the Cobb angle and the surgical duration, no statistical differences were detected among the four groups. All operations were completed by a fixed panel of surgeons.

Table 1 Demographics and surgical characteristics of patients
Full size table

The blood loss of the control group (2045.1 ± 599.3 ml) was similar with those reported in the literature (1971 ± 831 ml [1] and 1554 ± 1106 ml [7]). The intraoperative blood loss of group B (1323.3 ± 373.3 ml) was similar with group C (1169.5 ± 270.5 ml) (p = 0.212), while group D was significantly lower than the rest three groups with decreases of 64.5% (compared with group A, p < 0.001), 45.1% (compared with group B, p < 0.001) and 37.8% (compared with group C, p = 0.001). For the allogeneic blood transfusion, there were no significant differences between group B (360.0 ± 263.7 ml) and group C (235.0 ± 163.1 ml) (p = 0.069). However, the allogeneic blood transfusion of both two groups were significantly lower than group A (p < 0.0001) and higher than the group D [p < 0.0001 (group B) and p = 0.006 (group C)]. As to the autologous blood transfusion, group A (1005.8 ± 267.5 ml) differed significantly with group B (710.7 ± 186.5 ml) and group C (634.7 ± 162.2 ml) (p < 0.0001). Although the autologous blood transfusion of group C was lower than group B, it was comparable between the two groups (p = 0.211). The autologous blood transfusion of group A, B and C was much higher than group D, which decreased by 61.2, 45.1 and 38.5% as compared to the other groups (p < 0.0001). Finally, the overall drainage of every group decreased orderly with p values were all <0.0001. The overall drainage of the group D (301.6 ± 94.8 ml) reduced by 68.0, 58.4 and 41.7% compared with group A, B and C, respectively. FFP was greatest in group A (505.0 ± 139.5 ml), while group B (185.0 ± 168.1 ml), group C (100.0 ± 85.8 ml) and group D (25.0 ± 44.4 ml) diminished conspicuously versus group A (p < 0.0001). Furthermore, as compared to group A, B and C, group D decreased by 95.0% (p < 0.0001), 87.6% (p < 0.0001) and 75% (p = 0.047), respectively. The FFP of group C was also markedly lower than the group B (p = 0.025) (Fig. 1).

Fig. 1
figure 1

Amounts of intraoperative blood loss, allogeneic and autologous blood transfusion and FFP as well as postoperative overall drainage. I Intraoperative blood loss of group A was significantly higher than the other three groups (p < 0.001); intraoperative blood loss of group B and group C were comparable (p = 0.212), while that of group D was significantly less than group B (p < 0.001) and group C (p = 0.001). II compared with group A, a marked reduction of allogeneic blood transfusion was found in group B, C and D (p < 0.001), while results of group B and C were comparable (p = 0.069). Group D displayed much better results in comparison with the other three groups (p < 0.001, compared with group A and B; p = 0.0006, compared with group C). III there is no significant difference in autologous blood transfusion between group B and C (p = 0.211), however, autologous blood transfusion was significantly higher in group A than in group B and C (p < 0.001). Group D received significantly less homologous blood than the rest groups (p < 0.001). IV FFP of four groups was decreased orderly (p < 0.001, group A vs. group B, C and D; p = 0.025, group B vs. group C; p < 0.001, group B vs. group D; p = 0.047, group C vs. group D). V postoperative overall drainage of every group was reduced sequentially (all p values were <0.001)

Full size image

Though there were some transient changes of routine blood test results (Figs. 2, 3, 4) and coagulation parameters (Tables 2, 3) at the end of operation and on the first postoperative day, they were not clinically relevant.

Fig. 2
figure 2

HB results of pre- and post-operation as well as the first post-operative day (g/L). I Pre-operative HB results of every group were comparable (p > 0.05). II post-operative HB results of group A, B and C were much less than that of group D (p < 0.001). HB results of group B and C were of no significant difference (p = 0.071) as well as those of group A and group B were also comparable (p = 0.064). HB result of group C was much higher than that of group A (p < 0.001). III HB results of the first post-operative day of group A, B and C were all of significant differences in comparison with group D (p < 0.001), while comparisons between these three groups were comparable (p = 0.201, group A vs. group B; p = 0.299, group A vs. group C; p = 0.308, group B vs. group C). [*, # and & indicate differences vs. group A, B and D were statistically significant, respectively (p < 0.05)]

Full size image
Fig. 3
figure 3

Hct results of pre- and post-operation as well as the first day after operation (%). I Pre-operative Hct results of every group were comparable (p > 0.05). II post-operative Hct results of group A, B and C were significantly lower than that of group D (p < 0.001, p = 0.012, p = 0.013, respectively). Hct of group A was also lower than that of group B and C (p = 0.001), while difference of group B and group C was not significant (p = 0.26). III Hct results of the first post-operative day of group A, B and C were all of significant differences in comparison with group D (p < 0.001), while comparisons between these three groups were comparable (p > 0.05). [*, # and & indicate differences vs. group A, B and D were statistically significant, respectively (p < 0.05)]

Full size image
Fig. 4
figure 4

PLT values of pre- and post-operation as well as the first post-operative day (109/l). I Pre-operative PLT values of every group were comparable (p > 0.05). II post-operative PLT results of group A, B, C and D displayed marked increases orderly (p < 0.001). III PLT results of the first post-operative day of group A, B, C and D were also of significant differences compared with each other (p < 0.001). [*, # and & indicate differences vs. group A, B and D are statistically significant, respectively (p < 0.05)]

Full size image
Table 2 Coagulation parameter results
Full size table
Table 3 Results of Sonoclot coagulation analyzer and platelet function analysis
Full size table

Discussion

The main findings of this study were that batroxobin and TXA could significantly attenuate haemorrhage and blood products, such as allogeneic blood requirements and FFP, while TXA displayed a relatively better performance in ameliorating FFP and the overall drainage. However, the combined administration of the two drugs worked more effectively in these respects than either alone.

Idiopathic scoliosis surgery is one of the major operations associated with enormous blood loss and transfusion requirement. However, there are various disadvantages for hemostasis, such as particular spinal anatomical structure and spongy vertebras. Moreover, the venous plexus wall within the spinal canal is thin, easily broken and of no self-contraction after injury. Additionally, hemorrhage is oozing mainly at incision or wound and hemostasis depends largely on local coagulation. Therefore, conventional hemostatic measures, including electrocoagulation and oppression, are of no ideal effect. Besides these adverse factor, a number of factors may also impact perioperative bleeding. It is reported that the blood loss of scoliosis surgery is associated with the number of fused levels and duration of surgery but independent of Cobb angle, MAP, central venous pressure (CVP), or muscle relaxants and opioids [1]. In a recent study, Maria J et al. [8] reported that the surgical duration was the main variable having an impact on operative bleeding and transfusion requirements in complex spine surgery. As to our study, nevertheless, demographic materials, fused levels and surgical duration were not statistically significantly different.

After the suspension of aprotinin, the arduous problem that whether there are effective measures to decrease massive hemorrhage in spinal surgeries has persecuted surgeons for a long time. However, few studies found effective and safe solutions. Batroxobin, refined and purified from the Brothrops atrox venom, can promote blood fibrin monomer conversion and induce the formation of component FX particles on the surface of phospholipid under the condition that the injured vascular points release platelets factor III. Therefore, its hemostatic effect occurs only at the point of injury, avoiding large-scale intravascular coagulation [9]. However, there were rare articles concerning the efficacy and the side effect of batroxobin in spinal surgeries except for some Chinese papers, which stated that batroxobin performed well in spinal surgeries [10, 11]. TXA, a widely used fibrinolytic inhibitor, can competitively inhibit the plasminogen activator and the adsorption of plasminogen to fibrin, which prevents its activation and fibrin from being degraded by plasmin. It can even suppress plasmin directly in high concentration. Moreover, it has the protective effect on platelet. Therefore, in consideration that these two agents act at different points of coagulation cascade and their mechanisms seem to be complementary, we hypothesized that the combination of TXA and batroxobin may produce better results than either alone. However, there are rare articles concerning the combined administration of hemostatic agents. In 2005, Bulutcu and his colleagues conducted a study about the comparison of the efficacy of TXA, aprotinin and the combination of both two styptics on reducing the postoperative bleeding. They found that the combination of TXA and aprotinin was not more effective than either of the two drugs alone. They also reported that the coagulation parameters were not significantly different between the three groups [12].

As reported in some meta-analysis, TXA was effective on reducing blood loss and transfusion requirements in patients managed with spine surgery and its side-effect profiles have not been shown to cause any substantial morbidity or to increase the rate of thromboembolic events [13]. In a meta-analysis, Zufferey and colleagues found that there was a dose-effect relation with TXA [14]. Some experts announced that TXA could be considered as a possible agent to help reduce major hemorrhage in adult spine surgery on the basis of the present available efficacy and safety data [15]. Schindler et al. [16] regarded TXA as an alternative to aprotinin as antifibrinolytic therapy in pediatric congenital heart surgery. However, there were mixed dosages of these two agents reported in the literature. Therefore, the authors selected a compromise that batroxobin was given 0.02 U/kg for a loading dose and the maximum single dose was 1 U [10, 11], followed by super additions with the same dosage every 2 h till the end of the operation. The pre-charge dosage of TXA reported was 10–100 mg/kg or 2 g [2–4, 17], and the maintenance dose was 1–10 mg/kg h [17–19]. Therefore, in consideration of the mixed administration dosages of TXA in idiopathic scoliosis surgery, the pre-charge dose we adopted was 20 mg/kg and the maintenance dosage was 10 mg/kg h for a compromise solution.

In our study, intraoperative administration of batroxobin minimized the blood loss to 1323.3 ± 599.3 ml under the given conditions of this study, decreasing by 36.3% compared to the control group. Our results suggested that the blood loss of the TXA group (1169.5 ± 270.5 ml) decreased by 42.8% than the control group. As reported in a meta-analysis, TXA could mitigate haemorrhage from 25.5 to 49.2% compared to the placebo group [2–4, 9]. From above data, the reduction of blood loss of batroxobin group approximated to that of TXA group, which may be indicated that the hemostatic efficacy of these two styptics was alike. However, the combination of these two drugs was much more effective than either alone. The blood loss of the combination group was 45.1 and 37.8% reduction compared to batroxobin and TXA group. This phenomenon may demonstrate that combined medication in this manner could complementally excavate their potentials to a greater extent and produce superimposed or combined synergies. As to the transfusion requirements, some papers stated that the blood sparing effect of batroxobin and TXA were outstanding. In a meta-analysis, the decrement percentage of TXA ranged from 29.8 to 82.4% compared with placebo group [20]. Batroxobin and TXA, as shown in our trial, mitigated blood transfusion similarly (360.0 ± 263.7 ml vs. 235.0 ± 163.1 ml) with decreases of 57.6 and 72.4%, respectively, compared with the control group (850 ± 305.2 ml). The combination group performed much better than the other three groups with decreases of 94.7, 87.5 and 80.9%, respectively. The autologous blood transfusion changed consistently with the blood loss, including the statistical results. However, TXA seemed to work more efficiently than batroxobin in reducing FFP requirements and the overall drainage with reduction of 45.9 and 28.7%, respectively. Moreover, the combination exhibited much better results than either agent alone with decreases of 96, 86.5 and 75% compared with the other three groups.

Although the safety of combination of hemostatics with different mechanisms in this certain patient population has not been thoroughly established, no thrombotic complications or other adverse events were detected in this clinical trial.

Conclusions

In this study, batroxobin and TXA can reduce blood loss and allogeneic blood transfusion markedly. The combined administration of these two agents can reach much better results than either alone and no DVT or side effects were detected. However, it seems a little weak to draw the conclusion that batroxobin and TXA could be efficaciously employed conjointly because our results need further validation on a larger number of patients, even in a multicenter study considering the limited number of patients in this study.