Efficacy and safety of dabigatran for venous thrombosis prophylaxis after knee replacement surgery in Thai patients: a prospective non-randomized controlled trial

This was an investigator blind, single-center, prospective, non-randomized, paralled-group, controlled study. The study protocol was approved by Mahidol University Ethics Committees (certificate of approval no. MU-IRB2012/169.1810), and all eligible patients signed written informed consent forms. The study has been registered at the Thai Clinical Trials Registry (TCTR20150504001).

Between May 2015 and December 2016, there were 152 total knee replacements performed in 150 patients in our center. Inclusion criteria were age between 18 and 90 years and primary elective total knee replacement. Exclusion criteria were any abnormal bleeding history within 1 year, a history of DVT or PE, a history of hemorrhagic stroke or gastrointestinal bleeding, treatment with anticoagulants or antiplatelet within 7 days before surgery, severe liver disease, chronic kidney disease stages 4 and 5, women of child-bearing potentials, and active malignant disease. A total of 150 patients met our inclusion criteria. Twelve patients were excluded from the study: four with chronic kidney disease staged 4 and 5, three with previous use of anticoagulants, three with a history of hemorrhagic stroke or gastrointestinal bleeding, and two with a history of the previous DVT. Of the 138 eligible patients, 74 refused to participate, leaving 64 enrolled patients. We informed all enrolled patients about the study protocol as well as the risks and benefits of dabigatran and allowed the patients to decide independently whether or not to receive the prophylactic agent. Thirty-two patients who decided to receive oral dabigatran were included in the study group, whereas another 32 patients who refused to receive dabigatran were included in the control group. All 64 patients were studied prospectively. Sixty-two patients were diagnosed with primary osteoarthritis of the knee and two diagnosed with rheumatoid knee arthritis. Primary TKA was performed by three orthopedic surgeons. All patients received regional anesthesia with an adductor canal nerve block or femoral nerve block combined with spinal anesthesia. All prostheses were cemented, fixed bearing, and posterior stabilized knees (NexGen; LPS, LPS-flex, Gender Solutions; Zimmer, Warsaw, Indiana). We used the intramedullary guide for the femoral cut and extramedullary guide for the tibial cut. One dose of antibiotic prophylaxis with Cefazolin or Clindamycin was given before the operation. The operated leg was exsanguinated before inflation of the tourniquet with over 150 mmHg systolic blood pressure. The postoperative protocol included cold compression, compressive dressing in the first 24 h, drain removal in 48–72 h after operation, and continuous passive movements initiated on the second day followed by active mobilization and full weight-bearing ambulation.

All patients in the study group received a single dose of oral dabigatran 110 mg in the first postoperative day and then 220 mg once daily for 14 days. The control group received no chemoprophylaxis. The patient demographic data, preoperative functional knee score (Oxford Knee Score), operative data (operative time, intraoperative blood loss), lengths of stay, and complications were recorded.

Between May 2015 and December 2016, 150 patients underwent TKA and were assessed for eligibility. There were 138 patients met all eligibility requirements. Among them, 74 patients refused to participate in this study and 64 patients were enrolled. Of these, 32 patients were included in the study group and 32 patients were included in the control group. All patients in both groups were evaluable for outcome assessment and completed the study (Figure 1). Intergroup differences were not significant for BMI, age, underlying comorbidity, preoperative knee score, and operative time (Table 3). However, the length of hospital stay for the control group was shorter than the study group (7.4 days vs. 8.4 days, P = 0.02)

Figure 1

There has been some controversy over the routine use of pharmacological prophylaxis for DVT after TKA in Asian countries because there is a substantial variation reported in the incidence of DVT in Asian centers [3, 4, 5, 6, 7]. Cohen [14], on behalf of the Asia-Pacific Thrombosis Advisory Board, reviewed the incidence of DVT from local Asian studies including eight countries. The incidence of DVT ranged from 0 to 81.3%; he concluded that current guideline recommendations should be implemented in Asia for the routine use of postoperative thromboprophylaxis. In contrast, in Thailand, the Thai conference on VTE in hip and knee surgery, held in September 2015, recommended to use chemoprophylaxis for VTE in Thai patients undergoing total joint arthroplasty only in the high-risk because of the risk of bleeding complications.

Dabigatran, a direct thrombin inhibitor, has been approved as an orally administrated anticoagulant in many countries including Thailand. Food and Drug Administration (FDA) of Thailand has approved dabigatran since 2009 for the prevention of stroke and VTE in atrial fibrillation (AF), orthopedic surgery, and treatment of DVT and PE. Recently, dabigatran has been widely accepted as chemoprophylaxis after orthopedic operations because it is convenient to administer, does not need coagulation monitoring and has comparable efficacy for venous thromboprophylaxis to enoxaparin [8]. However, the uncertainty in the safety profile may be a major limitation of DOACs in clinical practice [15].

For the efficacy outcome, we found no thromboembolic complication after TKA in Thai patients in both the control and the dabigatran groups. This finding supports our previous study that the incidence of thromboembolism after primary TKA in Thai patients is very low with or without chemical prophylaxis [16]. We used Doppler ultrasonography as the diagnostic tool for detecting DVT in all patients in this study because it was noninvasive and reliable; however, it was an operator-dependent method and had low sensitivity for the detection of asymptomatic DVT in the calf [17], which was a limitation in this study, however, we believe that asymptomatic distal DVT has low clinical significance and needs no treatment. Vaitkus et al. [18] studied the mortality rate of asymptomatic DVT in medical patients. He concluded that asymptomatic proximal DVT associated with increased mortality rate; however, no significant difference was observed in the mortality rate between the asymptomatic distal DVT and no DVT groups. We also used the mean DVT and PE clinical score as a secondary efficacy outcome measurement; however, no significant difference was observed between the two groups. We developed these scores in our recent study [16] to assess clinical signs of DVT and PE. The DVT clinical score had no correlation with the diagnosis of DVT, and there was no patient had clinical signs suspected for PE.

For safety outcome measurement, our study shows no significant difference in the composite of major and clinically relevant nonmajor bleeding events as well as minor bleeding events between the control group and the dabigatran group. Friedman et al. [8] reported the incidence of bleeding complications from a pooled analysis of three major trials (RE-MOBILIZE, RE-MODEL, and RE-NOVATE). Major bleeding incidence of dabigatran 220 mg was 1.4% which was lower than in our study (3.1%). The composite of major and clinically relevant bleeding event was 5.6% compared with our finding (6.2%) and the minor bleeding event of dabigatran 220 mg was 6.6% which was lower than our finding (12.5%). The difference in incidence between pooled data and our finding in minor bleeding events is because the criteria for minor bleeding events are different in details. The mean length of stay is longer in the dabigatran group than in the control group, 8.4 days vs. 7.4 days (P < 0.05) because one patient with Grade III surgical site hematoma in the dabigatran group had delayed rehabilitation program, and the length of stay was 17 days.

The findings in this study suggest that dabigatran has a questionable benefit–risk profile for the prevention of VTE after TKA Thai patients in general. However, the small number of the study population, in this study, could not show any statistically significant differences in all outcome measurements, which is a limitation in this study. We found that the calculated sample size based on the incidence of venous thromboembolic events from the past literature was higher than in our study because most studies used venography as the diagnostic tool. However, we used ultrasonography because the major advantages in terms of its noninvasive nature and the absent risk of thrombosis, skin irritation, and necrosis secondary to contrast injection required in venography. Grady-Benson et al. [19] demonstrated 100% sensitivity, specificity, and accuracy of duplex ultrasonography compared with venography in detecting proximal DVT and 88% sensitivity, 98% specificity, and 98% accuracy in detecting distal DVT. Further studies with larger sample size are needed to confirm our finding.

We concluded that dabigatran might have no clear benefit for the prevention of VTE after TKA in Thai patients. We do not recommend the routine use of dabigatran as a chemical thromboprophylaxis after TKA in general Thai patients. High-risk patients for VTE should be considered for using chemical prophylaxis, depending on each patient's characteristics. Further studies, including a larger study population, are needed to confirm our findings.