Topically Applied Epsilon-Aminocaproic Acid Reduces Blood Loss and Length of Hospital Stay After Total Knee Arthroplasty

Materials and Methods

This single-center, single-surgeon (M.G.) retrospective analysis compared 3 groups of consecutive patients with primary or secondary osteoarthritis who underwent a unilateral TKA from January 2012 to August 2016. The surgeon was fellowship trained and had more than 20 years of experience performing primary TKA. The first group consisted of patients who were operated on from January 2012 to January 2013. They did not receive ε-ACA during their TKA (control group, N=80). From January 2013 to May 2015, the authors used ε-ACA locally during TKA. Epsilon-aminocaproic acid was applied after tourniquet release. Accordingly, the second group consisted of patients who were operated on from January 2013 to May 2015 and who received ε-ACA locally after tourniquet release during their TKA (ε-ACA-after-tourniquet-release group). From August 2015 to August 2016, the authors continued to use ε-ACA locally during TKA. However, ε-ACA was applied before tourniquet release. Accordingly, the third group consisted of patients who received ε-ACA locally before tourniquet release during their TKA (ε-ACA-before-tourniquet-release group). Patient characteristics are listed in Table 1. This retrospective data analysis was approved by the local Clinical Research Ethics Committee.

Table 1: Preoperative Demographic and Clinical Characteristics of the Patients

The surgery was performed by 1 surgeon using a standardized technique. A midline approach was used, and a medial arthrotomy was performed. Posterior stabilized, cemented TKA was selected using Journey II BCS (Smith & Nephew, Memphis, Tennessee); patellar replacement was performed in all patients. A tourniquet was used for cementing and was inflated after satisfactory alignment of the trial components was achieved. Before release of the tourniquet, the ε-ACA-before-tourniquet-release group received 5 g of ε-ACA diluted in 100 mL of normal saline soaked into the wound and secured with compression dressing, which was left in place for 3 minutes. The dose of 5 g of ε-ACA per 100 mL of normal saline and the exposure time of 3 minutes have been shown to be effective in reducing blood loss in THA.²⁰ The remnant was suctioned after 3 minutes, and then the tourniquet was released. The ε-ACA-after-tourniquet-release group received 5 g of ε-ACA diluted in 100 mL of normal saline after release of the tourniquet. The ε-ACA solution was soaked into the wound and secured with compression dressing, which was left in place for 3 minutes. The remnant was suctioned after 3 minutes. The control group did not receive ε-ACA. In all 3 groups, the capsule and wound were then closed in a standard fashion. No group received a drain or an additional injection.

A multimodal regimen was used for postoperative analgesia in all patients: acetaminophen combined with hydrocodone or oxycodone in the absence of contraindications. A patient-controlled hydromorphone pump was used for rescue analgesia if needed. A standardized postoperative anticoagulation protocol was applied for all patients. Dalteparin (5000 IU/d for 14 days) was used for postoperative thromboembolic prophylaxis. Enteric-coated acetylsalicylic acid (325 mg/d) was used for a further 28 days. The same anticoagulation protocol was used in all 3 groups without exception. Patients with contraindications for dalteparin or acetylsalicylic acid were not included in the analysis. In addition, alternating leg pressure devices were used during surgery and while the patient was in the hospital. All patients used the same postoperative recovery program.

An intraoperative blood salvage device was not used for any patient. Indications for transfusion were a hemoglobin (Hb) value of 7 g/dL in patients free of cardiovascular disease and a Hb value of 8 to 9 g/dL in patients with established cardiovascular disease or cardiovascular risk factors.^17,24,25 A Hb value below 10 g/dL in patients with poor clinical tolerance of lower values was also an indication for transfusion. Symptoms of poor clinical tolerance of lower values were signs of hypoxia such as tachycardia, dyspnea, or syncope.²⁶

Blood loss was calculated using the Hb balance method. The blood volume (BV) was calculated using sex, body mass, and height.²⁷ The preoperative Hb value and the Hb value on postoperative day 2 were used for calculation of blood loss according to the formula: Hb loss=(Hb_pre-Hb_post)×BV+Hb_Transfusion.²

A power analysis based on previous data²⁸ showed that a clinically relevant reduction in total blood loss (280 mL), with 90% power (alpha=0.05), would require at least 67 patients in each group. Data are expressed as mean±SD. The Student's t test was used for parametric data. The chi-square test was used for categorical values. P<.05 was considered significant.

Results

The total blood loss was reduced from 1478.8±367.1 mL in the control group and 1424.0±249.3 mL in the ε-ACA-after-tourniquet-release group to 1052.3±419.1 mL in the ε-ACA-before-tourniquet-release group (P<.05) (Figure). In the ε-ACA-before-tourniquet-release group, 5 of 80 patients received transfusions, as compared with 3 of 80 patients in the ε-ACA-after-tourniquet-release group and 5 patients in the control group (P>.05). A total of 7 units of packed red blood cells were transfused in the ε-ACA-before-tourniquet-release group, 5 units of packed red blood cells in the ε-ACA-after tourniquet release group, and 9 units of packed red blood cells in the control group.

Figure: Comparison of blood loss among the control group not receiving epsilon-aminocaproic acid (ε-ACA) (1), the ε-ACA-after-tourniquet-release group (2), and the ε-ACA-before-tourniquet-release group (3). Descriptive statistics include the quantile plot, standard deviation, and standard error in each column.

The mean length of hospital stay was 2.6±1.1 days in the ε-ACA-before-tourniquet-release group, 3.2±0.9 days in the ε-ACA-after-tourniquet-release group (P<.05), and 3.3±1.3 days in the control group (P<.05) (Table 2). The proportion of patients being discharged to a skilled nursing facility remained the same within the 3 groups (average, 16%).

Table 2: Surgical Outcomes, Blood Loss, and Packed Red Blood Cell Transfusion

In this analysis, all patients in the ε-ACA groups received 5 g of ε-ACA. At the authors' hospital, the cost for 5 g of ε-ACA is $1.03. The average cost for a day of hospital stay in the United States is $2212.30.²⁹ With an average reduction of length of hospital stay of 0.7 days (difference of length of stay between the control group and the ε-ACA-before-tourniquet-release group), the cost savings per patient using ε-ACA locally before tourniquet release is $1547.37 (ie, 0.7×$2212-$1.03) (Table 3).

Table 3: Hospital Cost Summary

At follow-up after 6 weeks, 1 patient in the control group and 1 patient in the ε-ACA-before-tourniquet-release group had developed a distal venous thromboembolism. This was treated with low-molecular-weight heparin followed by warfarin for 6 months.

Discussion

Epsilon-aminocaproic acid is a safe and effective drug for reducing blood loss after TKA. This single-surgeon, retrospective, case-control investigation into the use of topical ε-ACA in TKA did show a significant decrease in perioperative blood loss when ε-ACA was applied before tourniquet release. The authors chose to use calculated blood loss with the Hb balance method, rather than surgical estimated blood loss, to account for third space losses and postoperative bleeding that would not be measured using estimated blood loss alone. The use of ε-ACA saved $154,737 per 100 patients based on hospital stay costs alone. In the context of bundled payments for elective primary TKA, it is the responsibility of the health care delivery system to maintain lucrative and sustainable patient care.³⁰ The number of units of blood administered per patient did not differ between the ε-ACA groups and the control group. Two thromboembolic events occurred—1 in a patient from the ε-ACA-before-tourniquet-release group and 1 in a patient from the control group. The incidence of 0.8% is similar to the reported incidence of venous thromboembolism after TKA.³¹ To the best of the authors' knowledge, this is the first analysis to investigate the use of topical ε-ACA in TKA.

The current study had several limitations. The analysis was limited by its retrospective nature. Complications that did not occur in the hospital setting, were not reported at the follow-up visits, or did not require readmission were not recorded. In addition, the authors did not specifically assess postoperative functional recovery to investigate the relationship between blood loss and the outcome of rehabilitation. The assignment of each patient to 1 of the 3 study groups was based purely on the date of surgery. The chronological method of patient inclusion could have influenced the surgeon's proficiency and thereby biased the study results. However, the surgeon had been performing primary TKA for more than 20 years, and a significant change in proficiency was unlikely during the study period. Finally, this analysis was designed to show the superiority of topical application of ε-ACA compared with no treatment. As this was the first analysis evaluating the effect of topically applied ε-ACA in TKA, the authors refrained from comparing topically applied ε-ACA with topically applied TXA or intravenously applied ε-ACA, which would have changed the statistical analysis from a superiority study to a non-inferiority study.

The authors' finding of reduced perioperative blood loss of 427 mL is comparable with the findings of previous meta-analyses of individual studies of intravenous and local TXA administration for patients undergoing TKA.^14,21,32 The mean reduction of total blood loss using intravenous ε-ACA pooled for TKA and total hip arthroplasty is reported to be 331 mL.³³ There are several reasons why ε-ACA topically applied in TKA before tourniquet release might significantly reduce blood loss. When ε-ACA is administered intravenously, it is widely distributed throughout the extracellular and intracellular compartments and has a short half-life of approximately 2 hours.³⁴ Epsilon-aminocaproic acid is eliminated unchanged by glomerular filtration, which typically continues for up to 36 hours.³⁴ The advantage of topical application of ε-ACA is minimal systemic absorption and therefore reduced systemic side effects such as renal dysfunction.³⁵

The advantage of topical application of ε-ACA into the surgical field is the direct targeting of the site of bleeding just before wound closure but after surgical hemostasis has been achieved. Such inhibited local fibrinolytic activity will help to prevent fibrin clot dissolution, maintaining volume and strength at the raw surgical surfaces and thus enhancing microvascular hemostasis.

Before the tourniquet is released, the lower limb is exposed to anaerobic conditions, prompting the vascular endothelial tissue to release tissue fibrin. In response, fibrinolytic enzymes are activated that promote blood clot dissolution and increase postoperative blood loss.³⁶ The cascade of fibrinolysis is best inhibited in the initial stages; therefore, the application of antifibrinolytic agents such as TXA or ε-ACA is most efficient before tourniquet release. As the entire limb is under anaerobic conditions before tourniquet release, fibrinolytic enzymes are activated within the entire operated on limb. Promoting clot formation in the surgical wound before tourniquet release may prevent dispersion of activated fibrinolytic enzymes in the wound. This finding is in accordance with several other studies, which used TXA locally before tourniquet release.^37–39 In the current study, the tourniquet was inflated prior to cementing the components. The time and duration of tourniquet application—whether prior to incision or prior to cementing—does not influence perioperative blood loss when no antifibrinolytic agent is used.⁴⁰ In addition, the timing of tourniquet release (eg, after cementing the components or after wound closure) has no influence on perioperative blood loss when TXA is applied perioperatively.⁴¹ No comparative study using local ε-ACA is available that compares blood loss when the tourniquet is inflated either prior to incision or prior to cementing. However, blood loss using local TXA and tourniquet inflation prior to incision or prior to cementing is comparable.^42,43 Hence, it can be expected that blood loss using local ε-ACA and tourniquet inflation prior to incision is comparable to the current authors' results.

Patients treated with ε-ACA before tourniquet release had a shorter length of hospital stay of 0.7 days. This effect was most likely due to decreased hematoma formation and decreased hidden blood loss leading to less pain, faster rehabilitation, and earlier discharge.³

The TXA preparation costs approximately $80.²⁶ At the authors' institution, the ε-ACA preparation costs $1.03. Given that TXA and ε-ACA have similar safety profiles and efficacies, the application of ε-ACA can save up to approximately $79 per patient. In an institution with 1000 total hip arthroplasties annually, that would lead to savings of approximately $79,000 by using ε-ACA instead of TXA.

Conclusion

Local application of ε-ACA in the setting of TKA is efficient in reducing blood loss, safe with no increased risk of thromboembolic events, and economical. It should be applied before tourniquet release to gain the benefit of its antithrombolytic capacity.