Bleeding is considered one of the most common and threatening complications during percutaneous nephrolithotomy (PCNL).¹ The overall blood transfusion rate for PCNLs has been reported as 5.7%.¹ However, when patients are stratified according to their stone complexity, the greater the stone complexity, the higher the blood transfusion rate.² The risk of receiving blood transfusion due severe bleeding, according to the Guy’s Stone Score (GSS), is 0%, 2.2%, 4.8% and 22.2%, respectively to GSS I, GSS II, GSS III and GSS IV.^2–4

Given the potential occurrence of bleeding during PCNL, several strategies have been proposed to minimize the risk: 1) knowledge of anatomical landmarks of the kidney arteriovenous network and collecting system;^5–7 2) use of a nephrostomy tube;⁸ 3) avoiding causing intercostal neurovascular bundle injury;⁸ 4) avoiding excessive torque;⁹ 5) use of fibrin sealant in the nephrostomy tract;¹⁰ and 6) miniaturization of PCNL. However, these strategies have not sufficiently mitigated the risk of bleeding during PCNL.¹¹

In the last decade, interest in the use of tranexamic acid for decreasing perioperative bleeding has grown.¹² Tranexamic acid is a synthetic analogue of the amino acid lysine that reversibly binds to lysine receptor sites on plasminogen, thus postponing fibrin degradation. Since the 1960s, tranexamic acid has been known to block the active-site of urokinase, the plasminogen activator present in the urine, reducing clot breakdown in the urinary tract.¹³ Despite promising results from the use of tranexamic acid, few studies have investigated the role of intravenous administration of tranexamic acid in PCNL and conflicting results have been reported.^14–16

Given the limited direct and high-quality evidence on the effect of intravenous tranexamic acid administration to patients who undergo PCNL, we conducted the TrAc Trial (Tranexamic Acid Trial). The TrAc Trial was a prospective, randomized, double-blind, placebo-controlled trial conducted to investigate the effect of a single dose of tranexamic acid in patients with complex kidney stones undergoing PCNL. The trial protocol was approved by the University of São Paulo Research Ethics Committee (1.076.701) and registered on ClinicalTrials.gov (NCT02966236). The study followed the Good Clinical Practice (GCP) and Consolidated Standards of Reporting Trials (CONSORT) guidelines.^17,18

The study population included patients aged ≥18 years with complex kidney stones (GSS III and IV) and candidates for PCNL. The GSS was adopted to stratify patients according to stone complexity as the GSS has proven to be reproducible, easy to use in the clinical setting, and able to facilitate objective comparison in clinical trials.^3,19 Moreover, the GSS allowed identification of the cohort at higher risk of receiving blood transfusion during PCNL (ie GSS III and IV patients). Individuals who met 1 or more of the following criteria were excluded: known allergy to tranexamic acid; use of antiplatelet and anticoagulation therapies; history of venous thrombosis, pulmonary embolism, and coronary artery disease treated with a drug-eluting stent; known coagulopathies; baseline hemoglobin level <10 mg/dl and severe chronic renal failure (Modified of Diet in Renal Disease [MDRD] <30 ml/min/1.73 m²).

Eligible patients were randomly assigned in a 1:1 ratio to either tranexamic acid group or placebo group. At the beginning of anesthetic induction, a loading dose of 20 ml of the IMP (Investigational Medicinal Product, ie 20 ml [1 g] of tranexamic acid or 20 ml of 0.9% sodium chloride), was diluted in 250 ml of saline and intravenously infused over 15 minutes before skin incision (fig. 1).²⁰ A 1 g dose of tranexamic acid was demonstrated to be sufficient to obtain an anti-fibrinolytic effect.²¹ Trial participants, surgeons, anesthesiologists and those assessing outcomes were blinded to the intervention assignment.

Figure 1. IMP: tranexamic acid 20 ml (1 g) or sodium chloride 0.9% 20 ml. The tranexamic acid and placebo ampules were identical, and packaging was indistinguishable. Apart from randomization number, all labels were identical for tranexamic acid and placebo.

The primary outcome was occurrence of blood transfusion during the perioperative period. Secondary outcomes included blood loss (calculated by the Ward formula²²), operative time, stone-free rate (SFR) and complications. To address the potential bias of hemoglobin drop due to haemodilution, we calculated the haemodiluted hemoglobin using Ross’s formula.²³ Complications and adverse effects were investigated. Stone clearance was evaluated by non-contrast computerized tomography (NCCT) performed on postoperative day 1. Complete SFR was defined as no residual fragment (RF) on axial NCCT image. The success rate (SR) was defined as no RFs of >4 mm. The hemoglobin kinetics after PCNL was tracked at scheduled visits. Time to 90% and 95% hemoglobin recovery were estimated through a nonlinear regression growth model. All analyses were performed in the intention-to-treat population.

From May 30, 2016 to May 21, 2019, 217 consecutive patients classified into GSS III and GSS IV underwent PCNL at our institution and were screened for eligibility (fig. 2). A total of 192 patients were randomized and followed up to receive tranexamic acid (95) or placebo (97). Overall, both groups were balanced for baseline characteristics and kidney stone features. Treatment groups were also comparable in terms of operative parameters.

Figure 2. CONSORT diagram. CAD, coronary artery disease; CKD, chronic kidney disease.

Figure 3. Overall and intraoperative blood transfusion rates in tranexamic acid and placebo groups.

Figure 4. Mean hemoglobin for each group vs time since PCNL and hemoglobin kinetics recovery since PCNL. Time “Poi” is moment immediately after PCNL. Compared to placebo, tranexamic acid group had a shortened time to 95% hemoglobin, at a mean (95% CI) of 21.3 (11.5–31.2) time vs 46.8 (35.1–58.4; p=0.001). X-axis was constructed in 2 scales (from baseline to 24 hours, and from 24 hours to 90 days) to clearly demonstrate hemoglobin decrease period. h, hours. d, days.

The overall risk of needing blood transfusion was fivefold reduced in the tranexamic acid group (No. [%]: 2 [2.2%] vs 10 [10.4%]; RR 0.21, 95% CI 0.03–0.76; p=0.033; fig. 3). The number needed to treat was 12. The use of tranexamic acid significantly reduced the risk of intraoperative blood transfusion (p=0.029). The mean number of units of packed red blood cells administered per patient was statistically higher in the placebo group compared with that in the intervention group (p=0.006). The mean intraoperative blood loss was significantly lower in the tranexamic acid group than in the placebo group (p=0.006). We found no statistical difference in the haemodiluted hemoglobin and concluded that haemodilution was equally distributed in both tranexamic acid and placebo groups.

The immediate and 3-month complete SFR and SR were significantly higher in the tranexamic acid group than control group. We hypothesized that the reduction in intraoperative bleeding during PCNL facilitates clear intraoperative visualization of the collecting system and allows easier identification of RFs, without the need to abbreviate the surgery due to active bleeding. The operative time did not differ between groups. The post-hoc hemoglobin analysis showed that patients receiving tranexamic acid recovered their baseline hemoglobin levels faster and could be ready for a second PCNL earlier that the control group (fig. 4).

The antifibrinolitic effect of tranexamic acid was demonstrated within our study population. Patients who received tranexamic acid had lower absolute levels of D-dimer than controls, without changes in the prothrombin time and platelets. These results corroborate previous literature which emphasizes that tranexamic acid is not thrombogenic, but prolongs the degradation of existing clots. There was no statistical difference in complication rate. We underline the importance of stratifying patients according to their thromboembolic risk when considering the use of tranexamic acid. Patients at high risk for thromboembolic events should not use tranexamic acid.

In summary, the TrAc Trial presents strong evidence that administration of tranexamic acid to patients with complex kidney stones undergoing PCNL is safe and reduces the need for blood transfusion by 5 times. Moreover, tranexamic acid may contribute to better stone clearance rate and faster hemoglobin recovery without increasing complications. A single dose of tranexamic acid at the time of anaesthetic induction could be considered standard clinical practice for patients with complex kidney stones undergoing PCNL.