{"paper_id":"f6619e85-3b2a-4308-a7a0-1fd44e12f321","body_text":"Thrombosis is a very important and fatal complication after surgery, generally\noccurring in the form of deep vein thrombosis (DVT) or pulmonary embolism (PE)\n[ 1 ][ 2 ].\nIn patients undergoing major gynaecologic surgery, in the absence of\nthromboprophylaxis, the prevalence of DVT ranges from 15% to 40% [ 3 ]. Venous thromboembolism (VTE) is one of the\nmain causes of mortality after gynaecologic and obstetric surgeries [ 1 ] In general, the risk of VTE in cancer\npatients is five to six times higher than in non-cancer patients [ 2 ][ 3 ]. VTE\nis an independent prognostic factor for mortality and the second leading cause of\ndeath in cancer patients [ 4 ][ 5 ]. Also, asymptomatic DVT strongly increases\nthe risk of PE [ 6 ].\nSince most deaths associated with PE occur within 30 minutes of the onset, the time\nfor therapeutic intervention is very limited and it is necessary to identify those\nat high risk of VTE and to implement effective thromboprophylaxis to minimize\nmortality in these patients [ 7 ]. Despite the\nadvances made in recent years, venous thromboembolism (VTE) still accounts for a\nhigh percentage of mortality. Also, cancer increases the risk of VTE 4-7 times,\nmaking it the second leading cause of death in these patients [ 8 ].\nTherefore, patients undergoing surgical intervention for gynaecologic cancer are at\nhigh risk of VTE due to both risk factors. One method of preventing thrombosis is\nthe use of anticoagulant drugs such as rivaroxaban, heparin, and enoxaparin [ 9 ]. Heparin in combination with antithrombin III\nprevents clot formation by inactivating factor Xa and inhibiting prothrombin\nconversion [ 9 ][ 10 ][ 11 ]. Enoxaparin is\nalso a low molecular weight heparin that binds to and activates antithrombin III,\nthereby inhibiting factors Xa and IIa [ 10 ].\nIn fact, the main effect of this class of drugs is on factor Xa inhibition, with\nlittle effect on thrombin (IIa) and clotting time [ \n12 ]. On the other hand, rivaroxaban is an oral anticoagulant (NOAC). It\nis the first direct oral factor Xa inhibitor, a small molecule oxazolidinone\nderivative that binds directly and reversibly to factor Xa through S1 and S4\nreceptors, and competitively inhibits factor Xa [ \n13 ][ 14 ].\nUnlike heparin and enoxaparin, rivaroxaban inhibits both free and clot-bound factors\nand inhibits prothrombinase activity, thereby prolonging clotting time [ 15 ].\nGiven the importance of thrombosis in patients undergoing surgery, the present study\nwas conducted to compare the efficacy and safety of rivaroxaban, heparin, and\nenoxaparin in preventing thrombosis in gynaecologic oncology surgeries.\n\nThis study was designed as a single-center pilot randomized controlled trial (RCT)\nconducted at Hospital, affiliated with Semnan University of Medical Sciences, Iran,\nThe trial was registered in the Iranian Registry of Clinical Trials\n(IRCT20151020024625N19; https://www.irct.ir/trial/24625) and approved by the Ethics\nCommittee of Semnan University of Medical Sciences (IR.SEMUMS.REC.1402.223). Written\ninformed consent was obtained from all participants prior to enrollment.\nEligible patients were women scheduled for gynecologic oncology surgeries, including\nstaging hysterectomy or cytoreductive surgery, with histologically confirmed\novarian, endometrial, or uterine sarcoma. Exclusion criteria included\ncontraindications to anticoagulation, severe renal or hepatic dysfunction, or\nrefusal to participate.\nAs a pilot RCT, the target sample size was pragmatically set at 30 patients per group\n(total=90), consistent with recommendations for pilot studies. This number was\nintended to provide preliminary effect estimates for transfusion requirements and\ncomplication rates to guide future definitive trials. During the study, 5 patients\nwere excluded, resulting in 85 patients available for final analysis (25 enoxaparin,\n30 heparin, 30 rivaroxaban).\nPatients were randomly allocated into three groups (enoxaparin, heparin, rivaroxaban)\nusing a computer-generated block randomization sequence (block size=3). Allocation\nconcealment was ensured with sealed opaque envelopes prepared by an independent\nresearcher not involved in patient enrollment. This was a double-blind trial:\npatients and outcome assessors were blinded to treatment allocation, while nurses\nadministering the anticoagulants were not involved in outcome evaluation.\n*Enoxaparin group: received subcutaneous enoxaparin.\n*Heparin group: received subcutaneous unfractionated heparin.\n*Rivaroxaban group: received oral rivaroxaban.\nAll patients received perioperative care according to institutional protocols.\nA structured clinical checklist was used to record demographic data, clinical\nvariables, and outcomes. Content validity of the checklist was confirmed by three\nindependent experts in gynecologic oncology. Reliability was assessed by inter-rater\nagreement in 10 pilot cases (>90% agreement). Cronbach’s alpha was not applied,\nas the checklist was not a multi-item psychometric scale.\nBaseline variables: age, body mass index (BMI), employment status, cancer type, type\nof surgery, history of venous thromboembolism (VTE).\nPrimary outcome: intraoperative transfusion requirement (≥1 unit of packed red blood\ncells administered intraoperatively).\nSecondary outcomes: dyspnea, chest pain, peripheral edema, lower limb pain, bleeding\n(WHO criteria), infection, hematoma, recovery, and mortality, assessed during\nhospitalization and at 1-week, 2-week, and 1-month follow-ups.\nData were analyzed using SPSS v.22 (IBM Corp., Armonk, NY, USA).\nNormality of continuous variables was assessed with the Shapiro-Wilk test. Continuous\nvariables were presented as mean ± standard deviation (SD) and compared using\none-way ANOVA or Kruskal-Wallis test, as appropriate. Categorical variables were\nexpressed as frequencies and percentages and compared using chi-square or Fisher’s\nexact test. Relative risks (RR) with 95% confidence intervals (CI) were calculated\nfor key outcomes. A two-sided P<0.05 was considered statistically significant.\n\nValues are mean ± standard deviation (SD). Statistical test: ANOVA.\nAbbreviation:  BMI =Body Mass Index.\nValues are n (%). Statistical test: Chi-square test.\nAbbreviation:  VTE =Venous Thromboembolism\nValues are n (%). Statistical test: Chi-square test.\nAbbreviations: RR =Relative Risk;  CI =Confidence\nInterval\nValues are n (%). Statistical test: Chi-square or Fisher’s exact test\nValues are n (%). Statistical test: Fisher’s exact test\nValues are n (%). Statistical test: Chi-square test.\nThis study was conducted on 85 patients undergoing gynaecologic oncology surgery.\nBaseline demographic and clinical characteristics of the patients are summarized in\nTable- 1 a and Table- 1 b. Table- 1 a shows\ncontinuous variables (age and BMI), while Table- \n1 b presents categorical variables including employment status, type of\nsurgery, cancer type, and history of vascular events, The mean age of the patients\nstudied, enoxaparin, heparin, and rivaroxaban groups was 58.43±9.92, 58.12±12.28,\n59.60±9.51, and 57.53±8.21 years, respectively. No significant difference was\nobserved (Table- 2 ) in terms of mean age\nbetween the groups (P=0.532). Also, the mean BMI of the patients studied was\ncalculated to be 28.34±5.16 kg/m2, and no significant difference was observed in\nterms of BMI between the groups (P=0.056). The types of surgery performed included\ntotal hysterectomy and cytoreductive in 74 (87%), and 11 patients (13%),\nrespectively. Also, the types of cancer in the patients studied included\nendometrial, cervical, ovarian, and sarcoma in 51 (60%), 8 (9.4%), 20 (23.5%), and 6\npatients (7.1%), respectively. There was no significant difference between the\ntreatment groups in terms of fre. Based on the results, 2 patients (3.4%) had a\nhistory of vascular events. In the enoxaparin group, none of the patients had a\nhistory of vascular events, and in the heparin and rivaroxaban groups, one patient\nhad a history of vascular events. No significant difference was observed between the\ntreatment groups in terms of history of vascular events (P=0.653). In terms of\nintraoperative complications, 14 patients (16.5%) required blood transfusion, of\nwhich 2 (8%), 2 (6.7%), and 10 patients (33.3%) in the enoxaparin, heparin, and the\nrivaroxaban groups required blood transfusion, respectively. The need for blood\ntransfusion in the rivaroxaban group was significantly higher than in the other two\ngroups (P≥0.05, Table- 3 ). In terms of\npostoperative complications, dyspnea, chest pain, lower limb pain, and peripheral\nedema were reported in 1, 1, 4, and 2 patients in the heparin group, respectively.\nAlso, lower limb pain was reported in 2 patients in the rivaroxaban group. No\nstatistically significant difference was observed between the groups regarding\npostoperative complications (P≤0.05, Table- 4 ).\nOne week after the surgery, dyspnea was reported in 1 patient in the heparin group.\nLower limb pain was observed in 1 and 3 patients in the heparin and rivaroxaban\ngroups, respectively. Bleeding was observed in 3 patients, 1 in each (Table- 5 A, Table- \n5 B) treatment group, and there was no statistically significant\ndifference between the different groups in terms of the complications (P≤0.05,\nTable- 6 ). This is while peripheral edema was\nobserved in 1 and 5 patients in the heparin and rivaroxaban groups, respectively\n(P≥0.05).\nTwo weeks after discharge, dyspnea, chest pain, peripheral edema, and lower limb pain\nwere observed in 1, 1, 2, and 1 patients in the heparin group, respectively. Also,\nchest pain, peripheral edema, lower limb pain, and bleeding were observed in 1, 5,\n2, and 2 patients in the rivaroxaban group, respectively. In the enoxaparin group,\nno complications were reported, and no statistically significant difference was\nobserved between the groups in terms of complications two weeks after discharge\n(P≤0.05). Also, one month after discharge, dyspnea, chest pain, and bleeding were\nnot observed in any of the patients, but peripheral edema was seen in 5 patients (1\nin the heparin and 4 in the rivaroxaban group). Also, lower limb pain was observed\nin 3 patients (1 in the heparin and 2 in the rivaroxaban group), but no significant\ndifference was observed between two groups (P≤0.05). Infection and hematoma in the\nenoxaparin group, respectively. Also, pelvic hematoma and infection was observed in\n(Table- 7 ) 1 patient in the heparin group, and\ninfection and hematoma was observed in 2 patients in the rivaroxaban group. This is\nwhile no statistically significant difference was observed between the treatment\ngroups (P≤0.05). In total, the complications were observed in 7 patients (3 in the\nenoxaparin, 2 in the heparin, and 2 in the rivaroxaban group), and there was no\nsignificant difference between the groups (P≤0.05, Table- 8 ). The mortality rate in the enoxaparin and heparin groups was\n2 (8%) and 1 (3.3%), respectively, and in the rivaroxaban group, all patients had\npartial recovery. There was no significant difference between the groups in terms of\nthe mortality rate (P≤0.05, Table- 9 ). In\ngeneral, all three drugs studied were similar in terms of efficacy and safety, and\nno preference was observed in terms of thromboprophylaxis events.\n\nIn this pilot randomized controlled trial, we compared the effectiveness and safety\nof rivaroxaban, enoxaparin, and heparin for thromboprophylaxis in gynecologic\noncology surgeries. The main findings were: (1) intraoperative transfusion\nrequirements were significantly higher in the rivaroxaban group, with relative risk\nestimates 4-fold higher than enoxaparin and heparin; (2) peripheral edema was more\ncommon with rivaroxaban at one-week follow-up; (3) other short-term postoperative\ncomplications, including dyspnea, chest pain, lower limb pain, and bleeding, did not\ndiffer significantly between groups; and (4) long-term outcomes such as infection,\nhematoma, recovery, and mortality showed no statistically significant differences\namong groups.\nOur results suggest that although rivaroxaban is widely used in other surgical and\nmedical contexts, its application in gynecologic oncology surgeries may be\nassociated with increased intraoperative bleeding risk, reflected by higher\ntransfusion rates. This aligns with prior studies reporting variable bleeding\nprofiles for direct oral anticoagulants compared to heparin-based regimens. However,\nthe absence of significant differences in most postoperative complications and final\noutcomes suggests that rivaroxaban may still be a feasible alternative if bleeding\nrisk is carefully managed.\nEnoxaparin and heparin performed similarly across most outcomes. Both agents\ndemonstrated lower transfusion rates and comparable safety profiles. The modest\nincidence of peripheral edema in the rivaroxaban group may reflect drug-specific\npharmacodynamics, although this observation requires confirmation in larger cohorts.\nThe mortality rate, though low, occurred only in the heparin and enoxaparin groups,\nwhile no deaths were observed in the rivaroxaban arm. Given the small sample size,\nthis finding should be interpreted with caution and not generalized. Importantly,\nthe overall rate of partial recovery was high across all groups, indicating that all\nregimens were broadly effective for postoperative thromboprophylaxis.\nThe strengths of this study include its randomized controlled design, double\nblinding, and prospective data collection on both intraoperative and postoperative\noutcomes. However, several limitations must be acknowledged. First, as a pilot\nstudy, the sample size was not powered to detect small differences between groups,\nlimiting the generalizability of results. Second, unequal group sizes due to\ndropouts may have introduced imbalance despite randomization. Third, some outcomes\nwere rare, reducing the ability to conduct robust statistical comparisons.\nOur findings highlight the need for caution in the use of rivaroxaban in gynecologic\noncology surgeries, particularly regarding intraoperative bleeding risk. Larger,\nadequately powered multicenter RCTs are needed to confirm these results, refine risk\nstratification, and evaluate patient-centered outcomes such as quality of life and\nlong-term thromboembolic events. Until such data are available, enoxaparin and\nheparin remain well-established options for perioperative thromboprophylaxis in this\npatient population.\n\nThe results of our study indicated a greater need for blood transfusion in the\nrivaroxaban group than in the other two groups. However, no significant difference\nwas observed between the groups in terms of discharge time, postoperative\ncomplications, and follow-up on the one week, two weeks, and one month after\ndischarge. These results indicated the importance of thromboprophylaxis in\ngynaecologic oncology surgeries. Though initial research indicates that heparin,\nenoxaparin, and rivaroxaban might be equally safe and effective for\nthromboprophylaxis in gynaecologic oncology surgeries, these findings need to be\nverified. Additional large and multi-center randomized clinical trials are necessary\nto validate these findings and inform clinical practice.\n\nThe authors declare that they have no conflicts of interest.","source_license":"CC-BY-4.0","license_restricted":false}