Background

Total knee arthroplasty (TKA) is an effective treatment that helps to relieve severe pain and handicap induced by knee joint disease. However, TKA may cause significant perioperative blood loss ranging from 800 to 1800 mL, and 10 to 38% of patients need allogeneic blood transfusion [1–5]. Blood loss and subsequent blood transfusion can increase cost, and many complications, such as infection with HIV or other infectious diseases, fluid overload, and graft-versus-host disease can occur [6].

Many methods have been used to decrease blood loss during TKA, including the use of tourniquet, fibrin sealant, and tranexamic acid (TXA) [7–9]. The hemostasis effect of tourniquet in TKA is controversial. Wang et al. [10] revealed that total and intra-operative blood losses were reduced only with long-duration tourniquet use, while Zhang et al. [11] found that tourniquet release at the end of the TKA will activate local fibrinolysis and might increase blood loss. Auguilera et al. [12] reported that TXA can be more effective than fibrin sealant in reducing postoperative bleeding and transfusion requirement and that fibrin sealant use was not superior to routine hemostasis.

The use of TXA in primary TKA is today widely accepted, and several studies and meta-analyses have confirmed the efficacy of TXA at decreasing blood loss without increasing complications and costs [9, 13, 14]. Several clinical trials have identified that intravenous (IV) TXA is effective in reducing perioperative blood loss and the need for subsequent blood transfusions in TKA [15–17]. However, deep venous thrombosis (DVT) after the systemic administration of TXA is still a serious and fatal complication [18]. Therefore, many researchers have focused on the topical application of TXA via drain tube or intra-articular administration, and the topical application of TXA is considered an alternative effective route that entails less risk than IV after TKA [19, 20]. A previous meta-analysis has been published that only included five RCTs and one non-RCT [21]. Nevertheless, the comparative efficacy and safety of topical versus intravenous TXA often differs between studies. Thus, we performed a meta-analysis to compare the effectiveness and safety of topical versus intravenous TXA after TKA.

Methods

Search strategy

The following electronic databases were searched: PubMed, Embase, the Cochrane Library, Web of Science, Chinese Biomedicine Literature (CBM), Wanfang Database and China National Knowledge Infrastructure (CNKI), and Google Scholar were searched by two reviewers. Relevant studies comparing the topical administration of TXA and intravenous TXA for the management of blood loss during TKA were identified by two reviewers; the search was performed in August 2015. The keywords and Medical Subject heading (MeSH) terms used for the search were the following: “TXA,” “tranexamic acid” “total knee arthroplasty,” “total knee replacement”, “TKA”, “TKR”, and “Arthroplasty, Replacement, Knee”[Mesh]; the terms were connected by the Boolean operators AND or OR. The search string used in our research is presented in appendix A. Additionally, the reference lists of all identified full-text articles were reviewed to identify any initially omitted studies. There were no restrictions regarding language. Since this is a meta-analysis, no ethics committee or institutional review board approval was required.

Eligibility criteria and study quality

Study selection was performed according to the following inclusion criteria: (1) studies of patients undergoing primary TKA intervention; (2) with topical administration TXA and intravenous TXA as therapy to control blood loss; (3) studies assessing primary outcomes such as total blood loss, blood loss in drainage, hidden blood loss, need for transfusion, hemoglobin decline, and the rate of complications (DVT); and (4) studies designed as randomized controlled trials (RCTs). All the studies were required to be clinical trials. Trials on cadavers or artificial models were excluded. Letters, comments, editorials, and practice guidelines were also excluded.

According to the Cochrane Handbook for Systematic Reviews of Interventions (version 5.1.0) [22], Review Manager, version 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, 2014) was used to describe the quality assessment.

Data extraction

Two reviewers independently extracted and recorded following data in a spreadsheet: (1) patient demographic data, author’s name, publication date, sample size, location of study, ratio of male and female subjects, the dose and method of TXA application (topical or intravenous), and whether TKA was unilateral or bilateral; (2) the method of anesthesia; and (3) total blood loss, blood loss in drainage, hidden blood loss, need for transfusion, hemoglobin decline and rate of complications (deep venous thrombosis (DVT)). If the data were presented as figures, the software “Getdata Graph Digitizer” was used to extract the data for meta-analysis.

Outcome measures and statistical analysis

The main parameters studied were total blood loss, blood loss in drainage, hidden blood loss, need for transfusion, hemoglobin decline, and rate of complications (DVT). Hidden blood loss (mL) was calculated as total blood loss minus blood loss collected from drains. The results are presented as the mean differences (MDs) with 95% confidence interval (CI) for continuous outcomes such as total blood loss and blood loss in drainage. Binary outcomes, such as the need for transfusion and the occurrence of DVT were presented as relative risk (RR) values with 95% CI. Statistical significance was set at P < 0.05. To summarize findings across the trials, the software RevMan 5.3 (The Cochrane Collaboration, Oxford, UK) was used to conduct the meta-analysis. Statistical heterogeneity was tested using the chi-squared test and the I 2 statistic. Chi-squared test results with P > 0.1 were considered suggestive of statistical heterogeneity. An I 2 statistic greater than 50% was considered to indicate substantial heterogeneity [23]. A fixed effect model was adopted if the I 2 statistic value was less than 50%. If the I 2 statistic value was greater than 50%, a sensitivity analysis was performed using Stata 12.0 (Stata Corp., College Station, TX, USA) to explore the effect of an individual study.

Results

Search result and quality assessment

The initial search yielded 385 potentially relevant studies; of these, no duplicate was found using Endnote Software. Based on the inclusion criteria, 374 studies were excluded after reading the titles and abstracts. Finally, 12 clinical trials involving 1130 patients were included in the meta-analysis [24–36]. The process for selecting the included studies is presented in Fig. 1. One study included four IV groups [30], and one report included two topical groups that differed in the dose and timing administration of TXA [32]. Differently dosed TXA groups were included in this meta-analysis. The characteristics of the included studies are shown in Table 1, and the doses and methods used to apply TXA are summarized in Table 2. In the included studies, 1130 TKAs were performed, and the numbers of topical TXA and intravenous TXA were 529 and 601, respectively. Eight articles were in English, and four studies were in Chinese; all were published between 2013 and 2015. All participants in the 12 studies were elderly patients aged between 42 and 72.5 years who were planning to undergo TKA. Tourniquet was applied in all of the included studies, and the pressure of the tourniquet ranged from 100 to 350 mmHg. Detailed information regarding the risk of bias in the included studies is shown in Figs. 2 and 3.

Fig. 1
figure 1

Flow diagram for the included studies

Full size image
Table 1 Characteristics of included studies
Full size table
Table 2 Dose and method for topical and intravenous administration of TXA
Full size table
Fig. 2
figure 2

Summary of the risk of bias in each study

Full size image
Fig. 3
figure 3

Graph of the risk of bias for the included studies

Full size image

Meta-analysis results

Blood loss (total blood loss, blood loss in drainage, and hidden blood loss)

Five trials [12, 25, 27, 29, 30] reported data on topical versus intravenous administration for reducing total blood loss after TKA. Statistical heterogeneity was found (χ 2 = 41.56, I 2 = 83%, P < 0.00001, Fig. 4), and a random-effects model was performed. Topical administration of TXA showed no significant difference (MD2.08, 95% CI −68.43 to 72.60, P = 0.95, Fig. 4) in reducing total blood loss compared with intravenous TXA. A total of 10 studies [24–28, 32–34, 36] compared topical versus intravenous TXA in terms of reducing blood loss in drainage. Statistical heterogeneity (χ 2 = 864.26, I 2 = 98%, P < 0.00001, Fig. 4) occurred, and a random-effects model was performed. Similarly, the topical application of TXA did not significantly decreased the blood loss in drainage after TKA (MD18.49, 95% CI −40.01 to 76.98, P = 0.54, Fig. 4) compared with intravenous administration. For hidden blood loss, statistical heterogeneity (χ 2 = 39.60, I 2 = 95%, P < 0.00001, Fig. 4) was found, and a random-effects model was performed. There was no significant difference between the topical and intravenous groups (MD 4.75, 95% CI −337.94 to 347.44, P = 0.99, Fig. 4).

Fig. 4
figure 4

Forest plots comparing topical versus intravenous TXA administration regarding the need for transfusion. A Mantel-Haenszel fixed-effects model was used. Mean differences are shown with 95% CI

Full size image

Need for transfusion

Fifteen studies [24–27, 29–34, 36] mentioned a need for transfusion in 1206 enrolled knees. No significant heterogeneity (χ 2 = 13.20, I 2 = 2%, P = 0.43) was found. This paper uses a fixed-effects model, and no significant difference was found between topical TXA groups and intravenous TXA groups regarding the need for transfusion (RR = 0.92, 95% CI 0.67~1.25, P = 0.58) (Fig. 5).

Fig. 5
figure 5

Forest plots comparing topical versus intravenous TXA for the need for total blood loss, blood loss in drainage, and hidden blood loss. An inverse-variance random-effects model was used. Mean differences are shown with 95% CI

Full size image

Hemoglobin decline

No significant difference was found between the IV TXA with topical TXA groups (MD-0.42, 95% CI −0.89 to 0.05, P = 0.08, Fig. 6). Statistical heterogeneity (χ 2 = 107.72, I 2 = 94%, P < 0.00001, Fig. 6) exists, and a random-effects model was performed. To eliminate heterogeneity, a subgroup analysis was carried out for hemoglobin decline (Table 3).

Fig. 6
figure 6

Forest plots comparing topical and intravenous TXA administration in terms of Hb decline. An inverse-variance random-effects model was used. Mean differences are shown with 95% CI

Full size image
Table 3 Subgroup analysis for the hemoglobin drop
Full size table

Thromboembolic complications

Fourteen studies provided data about the rate of thromboembolic complications; the results of the meta-analysis indicate that the rate of thromboembolic complications for the intravenous group is 1.93% and that for the topical group is 1.58%; no significant difference was found between the two groups (RR = 1.17, 95% CI 0.55~2.50, P = 0.68, Fig. 7). No statistically heterogeneity (χ 2 = 5.72, I 2 = 0%, P = 0.68, Fig. 7) was found; therefore, a fixed-effects model was performed.

Fig. 7
figure 7

Forest plots comparing topical and intravenous TXA administration in terms of the occurrence of DVT. A Mantel-Haenszel fixed-effects model was used. Relative risk values are shown with 95% CI

Full size image

Other outcomes

Since the sample of other identified outcomes was not large enough for meta-analysis, we performed a descriptive analysis. For instance, Seo [33] found no significant difference between the range of motion between the two groups; the average ranges of motion (ROMs) at 2 months were 2.6°–123.3° and 2.5°–120.4° in the intravenous and intra-articular groups, and no significant difference was found in the intergroup analysis. The ranges of motion 1 month postoperatively were 104° ± 10° (95% CI, 101° to 108°) in the topical group and 105° ± 11 (95% CI, 101° to 109°) in the intravenous group (P = 0.612) [27]. There were 0 and 1 cases of atrial fibrillation in the topical TXA and IV TXA groups, respectively. Saezaeem [32] reported just one postoperative complication in the injected TXA group; this involved a patient with skin necrosis and more joint swelling than that seen for the intravenous group. One case of acute kidney injury was found in each of the topical and IV TXA groups [31].

Sensitivity analysis

A sensitivity analysis of blood loss in drainage was conducted, and the results indicated that none of the studies affected the stability of the final results (Fig. 8).

Fig. 8
figure 8

Sensitivity analysis of each end-point for total blood loss (a), blood loss in drainage (b) and hidden blood loss (c) after TKA

Full size image

Discussion

The results of the current meta-analysis indicate that there were no significant differences in total blood loss, blood loss in drainage, hidden blood loss, need for transfusion, and minimal hemoglobin decline between topical TXA and IV TXA groups in primary TKA. Moreover, no significant difference was found in the rate of DVT between the two groups. Moreover, no adverse effects were reported in the included studies.

Patients undergoing TKA may suffer from large blood loss due to the large area of osteotomy and enhanced local fibrinolysis after tourniquet release [37]. Total blood loss can be 300 mL higher than that occurring with no tourniquet use. Thus, the topical administration of TXA can reduce postoperative blood loss when combined with the use of a drainage clamp. Furthermore, Jawhar et al. [38] reported that tourniquet use can induce ischemia during TKA and can result in higher proteolytic activity within vastus medialis cells, which may increase hidden blood loss. Compared with intravenous TXA administration, intra-articular TXA administration can be rapidly absorbed, maintaining a biological half-life of approximately 3 h within the joint fluid; this may be the reason why patients benefit from topical TXA administration [39]. TXA is an antifibrinolytic agent that activates plasminogen and stops bleeding without increasing plasma fibrin levels [40]. Several studies and meta-analyses have identified TXA as an efficacious and safe way to reduce blood loss in patients who undergo TKA [41–43], and numerous studies have focused on the efficacy of topical TXA administration in reaching a maximum hemostatic effect [44–46]. However, the most effective TXA regimen in primary TKA remains uncertain. A meta-analysis on outcomes was previously conducted to explore the difference between the two groups [21]. However, in their study, five RCTs and one prospective cohort study were included; moreover, they did not compare hidden blood loss between the two groups. The current meta-analysis included twelve studies and did compare hidden blood loss, which comprises a large proportion of total blood loss.

With regard to the need for transfusion, no significant difference was found between the two groups (RR = 0.92, 95% CI 0.67~1.25, P = 0.58); the dose of topical administration ranged from 0.5 to 3 g, and the dose of intravenous TXA administration ranged from 10 to 15 mg/kg. Three studies reported hidden blood loss for the two groups, and Aguilera [24] favors the use of topical administration to reduce hidden blood loss after TKA, whereas Hou and colleagues [47] found no difference between topical and intravenous administration. In this meta-analysis, no significant difference between topical and IV TXA administration was found. Our sensitivity analysis showed that the study of Sarzaeem [32] significantly affected the heterogeneity results. After carefully reviewing the study, the main difference between this study and the other studies is that TXA was administered by irrigating with 3 g of TXA and injecting with 1.5 g TXA. In addition, doses of topically administered TXA ranging from 0.5 g [48] to 3 g [27, 30, 34, 44, 49] have all proven effective and safe. Regarding blood loss in drainage, a large degree of heterogeneity was found between the two groups, and the sensitivity analysis showed that the inclusion of Sarzaeem’s study influences the final conclusion. After carefully reading the article, the topical administration of TXA was applied by irrigation.

DVT and subsequent PE are fatal complications after TKA, and it has been reported that 0.1–2% of patients will suffer from pulmonary embolism following TKA [50]. Theoretically, intravenous TXA administration may increase the risk of thrombosis since it is a synthetic antifibrinolytic agent that can prevent plasminogen activation, and fibrinolysis is delayed [51]. However, it has been ascertained that TXA does not inhibit fibrinolytic activity in the vein wall, and its effect on disrupted endothelium remains unknown [52]. The results of the current meta-analysis indicate that both topical and intravenous administration TXA are safe when used to reduce blood loss during TKA. The frequencies of DVT after intravenous and topical TXA administration are 11/571 and 9/568, respectively (RR = 1.17, 95%CI 0.55~2.50, P = 0.68). The reasons for this may be as follows: (1) the doses used during intravenous and topical administrations are appropriate and do not exceed the optimal dose; therefore, the fibrinolytic system will not be suppressed. (2) Most studies only evaluated symptomatic DVT and PE, and this will have affected the final conclusion.

This meta-analysis has the following limitations: (1) Only twelve studies were included, and the sample sizes used in each study were insufficient, which will affect the conclusions. (2) The studies administered TXA at different doses and using different methods, causing large heterogeneity. (3) The duration of follow-up in some studies was short; therefore, long-term follow-up is needed to measure the occurrence of DVT and PE.

Conclusions

In conclusion, this meta-analysis indicated that topical TXA administration exhibits comparable efficacy and safety to those of intravenous TXA, and both methods can decrease the need for transfusion, total blood loss, blood loss in drainage, and hidden blood loss without increasing the occurrence of DVT. Due to the poor quality of the included studies, more high-quality RCTs are needed to identify the optimal method and dose of TXA after TKA.