Effect of Additional Administration of Topical Tranexamic Acid in Patients Undergoing Primary Total hip Arthroplasty Without Drainage

Introduction: This study was designed to compare the clinical results of additional administration of topical tranexamic acid (TXA) with intravenous TXA (IV-TXA) and to determine its effect in patients undergoing primary total hip arthroplasty (THA) without drainage using propensity score matching analysis. Methods: A total of 248 patients (248 hips) underwent primary THA from March 2016 to June 2018. Patients who received topical TXA and IV-TXA were included in the combined group (46 patients), and patients who received IV-TXA were included in the IV only group (202 patients). After propensity score matching, both groups consisted of 44 patients (88 patients in total). We compared the results of total blood loss (TBL), haemoglobin (Hb) level, Hb drop, and the incidence of deep vein thrombosis (DVT) and pulmonary embolism (PE). Results: In the combined group, the TBL was 280.4 ± 64.2 mL, which was signicantly lower than that in the IV only group (335.3 ± 75.3 mL; p < 0.001). From the preoperative period to 1-week postoperatively, Hb levels were not signicantly different between the two groups. There was signicantly more Hb drop in the IV only Group from immediate postoperative to postoperative day 3 (POD 3). In both groups, no patient presented with postoperative symptomatic DVT or PE. Conclusion: In patients undergoing primary THA without drainage, combined use of IV-TXA and topical TXA can signicantly decrease blood loss without inuencing postoperative complications.


Introduction
Total hip arthroplasty (THA) is a common and effective orthopedic surgical procedure for patients with end-stage hip disease. However, it is associated with a signi cant amount of blood loss (approximately 700-2,000 mL); hence, several patients require postoperative blood transfusions, with a reported transfusion rate of 16-37% [1]. Furthermore, although the incidence is low, some serious complications related to transfusions have been reported, including volume overload, immunological reactions, intravascular hemolysis, infections, renal failure, and even death [2][3][4].
Notably, two main factors are known to contribute to substantial blood loss during THA: overt blood loss (OBL) caused by surgical trauma and hidden blood loss (HBL) caused by brinolysis. HBL accounts for approximately 60% of the total blood loss (TBL) [5]. To reduce blood loss, several blood-conserving techniques have been used, including autologous blood transfusion, application of autologous brin tissue adhesive, and use of hypotensive anesthesia. These techniques can reduce the risk of transfusion and infection rates. However, some techniques such as autologous blood transfusion are expensive [6][7][8]. Another method used to control intraoperative blood loss is the administration of anti brinolytic agents, such as aprotinin, tranexamic acid (TXA), and epsilon-aminocaproic acid [9][10][11]. Of these agents, TXA has attracted the most attention due to its low cost and minimal complications.
TXA, a synthesized anti brinolytic agent, competitively inhibits the activation of plasminogen by blocking lysine-binding sites and inhibiting clot breakdown, resulting in reduced blood loss and transfusion requirements [12]. Numerous studies have con rmed that TXA could effectively reduce blood loss and transfusion rates in patients undergoing THA without increasing the risk of complications such as deep vein thrombosis (DVT) and pulmonary embolism (PE). To con rm the e cacy and safety of TXA, these studies have evaluated the use of TXA via single or repeated intravenous (IV) administration, topical application, or IV administration combined with topical application [10,[13][14][15][16][17][18].
In all these studies, topical application was evaluated in patients undergoing surgery with drainage. Thus, the complete effect of topical TXA administration may have been diminished due to drainage. Therefore, to determine the complete effect of topical administration of TXA, we conducted this retrospective study to assess whether the additional administration of topical TXA with IV-TXA is more e cient than administration of single-dose IV-TXA in patients undergoing unilateral primary THA without drainage.

Methods
We retrospectively reviewed the collected registry data of 248 patients who underwent primary THA from March 2016 to June 2018. The patients who received topical TXA with IV-TXA were included in the combined group (46 patients), whereas patients who received IV-TXA only were included in the IV only group (202 patients). Patients who met the following inclusion criteria were included: (1) age ≥ 18 years, (2) diagnosis of osteoarthritis (OA) or osteonecrosis of the femoral head (ONFH), for which unilateral, 2incision, minimally invasive THA without drainage was performed, and (3) use of IV-TXA only or a combination of topical TXA and IV-TXA. The preoperative exclusion criteria were as follows: (1) known allergy to TXA, (2) cardiovascular disease (history of angina, myocardial infarction, or atrial brillation), (3) cerebrovascular pathology (history of stroke), (4) thromboembolic disorders (history of DVT or PE), and (5) clotting disorders. Patients who underwent THA before September 2017 were not administered topical TXA during surgery, whereas patients treated during and after September 2017 were administered topical TXA to effectively reduce blood loss and postoperative complications. Since the additional administration of topical TXA depended only on the date of surgery, no patient selection bias was present. Four patients, two in each group, were lost to follow-up postoperatively.
Before propensity score matching (PSM) analysis, the combined group included 44 patients and the IV only group included 200 patients. To minimize the effect of possible confounding factors, both groups underwent PSM prior to analysis. The match tolerance and maximum difference between the propensity scores of any matched pair were set at 0.051. One-to-one PSM was performed to minimize the selection bias using variables including age, sex, body mass index (BMI), weight, and preoperative haemoglobin (Hb), hematocrit (HCT), prothrombin time (PT), and activated partial thromboplastin time (APTT) values to identify comparable patients. The included variables were not signi cantly different between the two groups before PSM. Patients were successfully selected such that no signi cant differences would be observed after matching (Table 1). Finally, 44 patients from the combined group were matched to 44 patients from the IV only group (Fig. 1). This study was approved by the Institutional Review Board of our hospital.
All operations were performed by a single senior surgeon using cementless acetabular and femoral components, and all procedures were performed via 2-incision MIS under general anesthesia. In the combined group, 1 g IV-TXA was administered 10 min before skin incision, and 1 g TXA in 50 mL physiological saline was injected intra-articularly after capsule and fascia closure. In the IV only group, patients received a single IV dose of 1 g TXA, 10 min before skin incision. In all patients, drainage was not performed. Additionally, for preventing venous thromboembolism (VTE) events, an intermittent pneumatic compression device was used postoperatively as a routine practice. Patients were examined daily for DVT symptoms (postoperative swelling in the affected leg, red or discolored skin on the leg, feeling of warmth in the affected leg); if any DVT symptoms were observed, venous Doppler ultrasound was performed.
The postoperative outcomes included TBL, OBL, and HBL. TBL was calculated using the Gross and Nadler formula [5,19], OBL was de ned as intraoperative blood loss, and HBL was de ned as TBL excluding OBL. As drainage was not performed in all patients, a small amount of blood remained in the joint cavity postoperatively. In this study, this residual amount of blood was considered under HBL. We assessed Hb levels preoperatively; immediately post-operation; on postoperative day (POD) 1, POD 3, and POD 5; and 1-week postoperatively. Moreover, we calculated the Hb drop from immediately post-operation to 1-week post-operation. Regarding hip function, we assessed the Harris hip score (HHS) [20] preoperatively and at 3 months post-operation. Blood transfusion was deemed essential in patients presenting with symptomatic anemia (de ned as lightheadedness, presyncope, fatigue precluding participation in physiotherapy, and palpitations or shortness of breath not attributed to other causes), characterized by Hb levels between 7 and 10 g/dL or Hb levels below 7 g/dL postoperatively. Independent t-tests were applied for analyzing continuous data, which are expressed as means ± standard deviations. The Pearson chi-square test and Fisher's exact test were used to compare binary data, expressed as percentages. For all analyses, a p value of < 0.05 was considered to indicate statistical signi cance. IBM SPSS, version 24.0 (IBM Corp., Armonk, NY) was used for all statistical analyses in this study.
No difference was observed in the HHS between the two groups preoperatively (combined group: 60 ± 9.2; IV only group: 58.7 ± 9.3; p = 0.575) or 3 months postoperatively (combined group: 92.2 ± 2; IV only group: 91.2 ± 2.2; p = 0.327). Furthermore, there was no difference in operation time between the two groups (combined group: 78.6 ± 11.6; IV only group: 77.5 ± 12.3; p = 0.658) ( Table 2). In the IV only group, two patients needed transfusion postoperatively. No symptomatic DVT or PE events were observed in the two groups; no patients presented with hematomas postoperatively (Table 3).

Discussion
Patients undergoing THA demonstrate a relatively high risk of requiring allogeneic blood transfusion. In the context of primary THA, several investigations regarding IV and topical TXA administration have reported the e cacy and safety of TXA in reducing blood loss and transfusions [13][14][15][16]. The combination of IV-TXA with topical TXA has been introduced and is used by surgeons in patients undergoing total knee arthroplasty (TKA) and THA with satisfactory results [17,18,21,22]. Following preoperative IV-TXA administration, TXA is widely distributed across extracellular and intracellular compartments, rapidly reaching the maximum plasma concentration in 5-15 min. Furthermore, IV-TXA inhibits local brinolysis as soon as surgery is initiated, with maintenance of the plasma TXA concentration above the minimum therapeutic level for approximately 3 hours [23,24]. However, concerns regarding the risk of DVT and PE due to systemic administration of high-dose TXA persist, hindering the widespread application of IV-TXA [25].
Compared with IV-TXA, topical TXA application has advantages such as ease of administration, inhibition of clot breakdown with maximum concentration at the bleeding site, and reduction in joint swelling, which leads to improved wound healing, with minimal systemic absorption [26,27]. During surgery, most of the bleeding occurs during soft tissue release and acetabular and femoral canal preparation. Topical TXA administration directly targets the bleeding site in a surgical wound, maintaining maximum local TXA levels to induce partial microvascular hemostasis by preventing breakdown of the brin clot [28]. Considering the biological half-life of TXA in the bloodstream or joint space [29], topical TXA administration during wound closure can extend the effective time of TXA at the surgical site.
In the present study, the administration of 1 g topical TXA in combination with IV-TXA was more effective in reducing TBL, OBL, and HBL than the administration of IV-TXA alone. The post hoc power analysis revealed that the current study had > 80% power for comparing TBL, OBL, and HBL between the two groups. From immediately post-operation to 1-week post-operation, there was no signi cant change in Hb levels in the combined group compared with that in the IV only group. We speculate that this may be due to the differences in preoperative Hb levels, because the average Hb level in the IV only group was lower than that in the combined group (0.5 g/dL) preoperatively (combined group 12.9 ± 1.3 g/dL vs. IV only group 13.4 ± 1.4 g/dL; p = 0.086). However, the Hb drop was signi cant from immediately post-operation to POD 3, and the post hoc power analysis showed > 80% power, demonstrating no signi cant difference in Hb drop between POD 5 and 1-week post-operation. Therefore, TXA is effective in controlling the Hb drop from immediately post-operation to POD 3. In a randomized double-blind controlled trial, Yue et al. [16] compared the application of 3 g topical TXA with the application of a placebo, which revealed that the Hb drop was signi cantly lower in the topical group at POD 1 and POD 3. These ndings are consistent with our results. Interestingly, in our study, although the Hb drop at POD 5 was not signi cantly different, it demonstrated a strong tendency toward signi cance (p = 0.051).
Notably, the TXA dose has not been standardized and remains controversial. TXA is frequently administered intravenously with a loading dose of 10 or 15 mg/kg, followed by continuous infusion or repeated bolus doses [30,31]. Husted et al. randomized patients to receive TXA as a bolus IV injection of 10 mg/kg (maximum 1 g) for 10 min, approximately 15 min before incision. This resulted in reduced blood loss and a reduced need for blood transfusion. Furthermore, no patient reported prolonged drainage, infection, clinical DVT, or PE during hospitalization or at the last follow-up. For topical administration, the TXA dose ranged from 0.5 g TXA/100 mL normal saline to 3 g TXA/100 mL normal saline, as reported by Zhao et al. [32] in a meta-analysis of six randomized controlled trials. As mentioned earlier, Yue et al. [16], in a randomized double-blind controlled trial, administered 3 g TXA in 150 mL saline at three points during THA and reported that the topical application of 3 g of TXA signi cantly reduced bleeding and transfusions in patients undergoing primary THA without increasing the risk of DVT and PE. Several investigators have reported another strategy for the combined administration of TXA in the setting of THA; in most of these studies, the maximum dose of TXA was < 3 g [17,18]. Xie et al. [17] performed a prospective randomized controlled trial and compared the results according to administration route: IV, topical, and a combination of the two. In all three groups, the TXA dose was < 3 g.
They observed that patients undergoing primary unilateral THA in the combined group demonstrated an effective decrease in TBL, resulting in higher postoperative Hb levels without the risk of higher complication rates, compared to those in the other two groups. Yi et al. [18] performed a prospective randomized controlled trial and compared the results of placebo, IV-TXA, and the combination of IV-TXA and topical TXA. The TXA dose was < 3 g both in the IV-TXA and combined groups. TXA has been administered at different doses and using different routes in patients undergoing primary THA, with most investigators routinely administering TXA at a dose of no more than 3 g. They considered that a dose of < 3 g is safe and e cacious in these patients. In our study, in the combined group, 1 g of IV-TXA was combined with 1 g of topical TXA, and the total TXA dose in the two groups was < 3 g. In both groups, we observed no cases of DVT or PE during hospitalization and until the last follow-up.
The performance of drainage after THA remains controversial. The most important reasons for drainage include preventing the accumulation of hematomas and decreasing the risk of infection [33]. However, some studies have presented differing conclusions, stating that drainage increases blood loss, which may increase transfusion rates and provide an entry point for skin microorganisms, resulting in infections [34,35]. Zhou et al. [36] performed a meta-analysis of 20 randomized controlled studies evaluating the use of closed-suction drains in patients undergoing THA. They reported a signi cant increase in the homologous transfusion rate in patients undergoing drainage, with no signi cant differences in the incidence of infections, hematomas, or thrombosis. Furthermore, they suggested that in patients undergoing elective THA, the routine use of closed-suction drains may result in more damage than bene ts. Walmsley et al. [37] reported that the postoperative transfusion rate was signi cantly higher in the drainage group than in the no drainage group (33% vs. 26.4%, p = 0.042), concluding that drainage provides no clear advantage in the context of THA. Furthermore, a study by Valle et al. [38] showed that drainage confers no bene t in patients undergoing primary uncomplicated THA.
Most studies regarding TXA administration involved the use of drainage. We considered that a small amount of TXA may be lost owing to drainage, reducing the TXA concentration in the joint. A few studies have investigated primary THA without drainage [15]. However, no investigations have compared IV-TXA alone and IV-TXA combined with topical TXA in this population. In our study, to completely retain TXA in the joint, we did not perform any drainage; this approach was considered better for evaluating the effect of topical TXA. In the present study, only two patients needed transfusions, and in both groups, no patient presented with symptomatic hematoma, deep infection, or wound complications postoperatively, until discharge.
This study has several limitations. First, the study population was relatively small. We believe that a larger sample size is needed to effectively detect the difference in DVT and PE incidence between the two groups. Second, this study did not include a placebo group. IV-TXA may be signi cantly superior to a placebo but not as effective as the combination of IV-TXA and topical TXA. We considered that using combined administration as a control rather than a pure placebo was clinically a more useful and appropriate methodology.

Conclusion
The most important nding of this study was in unilateral primary THA, administration of topical TXA with IV-TXA without drainage can effectively decrease the TBL and elicit higher postoperative Hb levels without increasing the risk of complications compared with IV-TXA alone.    Figure 1 Flow chart of patient selection and propensity score matching analysis.

Figure 2
Histograms of total blood loss (TBL), overt blood loss (OBL), and hidden blood loss (HBL) From gure 2, we could see the TBL, OBL and HBL in the combined group were less than in the IV only group postoperatively (The asterisks indicate values that were signi cantly different between the groups). IV = intravenous.

Figure 3
Line diagram of haemoglobin drop Haemoglobin levels dropped postoperatively. Since no change was observed in preoperative haemoglobin levels, we de ned the preoperative value as zero. IP indicates the time at which patients exited the recovery room (as the rst postoperative blood test was performed in the recovery room in our institution). The asterisks indicate values that were signi cantly different between the groups. IP = immediately postoperative; POD 1 = postoperative day 1.