Introduction

Bothersome Lower Urinary Tract Symptoms (LUTS) is a common clinical manifestation of Benign Prostatic Hyperplasia (BPH) for men above 50-years of age [1]. Medical therapy with 5-alpha reductase inhibitor (5-ARI), including finasteride and dutasteride, is commonly recommended as part of conservative management to improved symptoms [2]. Surgical management options are considered when patient failed to achieve adequate symptomatic relief with medical therapy [3]. Despite availability of various urological procedures developed over the last two decades, TURP remains the most prevalent surgical option [4]. Despite its effectiveness, TURP entails significant bleeding-risk in the immediate 2-3 weeks post-operative period in which patient frequently represent acutely with haematuria and/or clot retention requiring further intervention [5].

5-ARI had been established to reduce the risk of BPH-related haematuria [6]. Its mechanism is thought to be secondary to 5-ARI blockade, thereby preventing conversion of testosterone to dihydrotestosterone, thereby reducing prostatic angiogenesis with decreasing Micro-Vessel Density (MVD) in the hyperplastic prostatic tissue [7]. Several studies had demonstrated its use in reducing peri-operative blood loss in patients undergoing TURP. Pre-operative treatment with finasteride for as short as two weeks had been demonstrated to reduce prostate MVD and reduced operative estimate blood loss [8,9]. The effectiveness of pre-treatment with dutasteride to reduce operative bleeding remains inconclusive within current literature [9]. These studies were mostly conducted on non-anticoagulated patients, and there remain a paucity of evidence on effect of 5-ARI in the anticoagulated group.

Patient receiving Anticoagulant (AC) undergoing TURP is known to have increased morbidity related to clinically significant haematuria even with appropriate peri-operative management of the AC [10]. With the aging population, the prevalence of men requiring management of symptomatic BPH is expected to increase [11]. In addition, prevalence of patient with co-morbidities requiring anticoagulation is expected to rise. For instance, it is estimated that patients with atrial fibrillation is projected to triple in the coming decades, and at present, approximately 65% of 46.3 million patient with AF globally receive anticoagulation therapy [12,13]. This epidemiological trajectory highlights the importance of minimizing post-operative haematuria risk in patients undergoing TURP, particularly for the anticoagulated patients.

Previous studies on 5-ARI and TURP that demonstrated reduced peri-operative haematuria had been conducted on non-anticoagulated patients. At present, there is a paucity of evidence in current literature to guide clinician in minimizing haematuria risk for TURP patients, except for withholding anticoagulant in the peri-operative period [10,14]. Our objective is to evaluate the effectiveness of 5-ARI in anticoagulated patients in reducing development of clinically significant haematuria and compare the risk to a non-anticoagulated cohort.

Methods

Study Population

A single centre retrospective cohort analysis was conducted at a high-volume referral centre at Gold Coast University Hospital, Australia. A cohort of patients that underwent elective TURP for symptomatic BPH were collated from 2019 to December 2023. All patients were reviewed in a multidisciplinary fashion pre-operatively. Decision to proceed to TURP would be made with a urologist, and patient would subsequently review by an anaesthetist, and relevant specialist if patient was deemed high surgical risk. A clinical pharmacist would review and document a patient’s up-to-date medication lists. Patient’s use for 5-ARI and/or anticoagulant will be documented at this encounter. For patients receiving anticoagulation, perioperative management of their AC would be decided via a multidisciplinary approach after assessment of their thromboembolic risk. Routine blood test including patient’s full blood count, renal function, liver function, and coagulation profile would be conducted for baseline assessment. Clinical data were retrieved from an integrated electronic medical record system utilized by the state-wide health care system. Over 23 relevant data variables including peri-operative, and operative details were collected. Patient demographics such as age, body mass index, and ethnic origin were recorded. Pre-operative parameter including indication for TURP, urinary catheter dependence, treatment with 5-ARI, pre-operative Prostate Specific Antigen (PSA) level, prostate size, and prostate cancer history were recorded. Anticoagulant details were collected for patients receiving AC therapy. These included indication for AC, pre-operative and post-operative withholding duration, and use of bridging therapy. Several well-established scoring systems were utilized as surrogate measure for patient’s baseline parameters. Patients’ co-morbidities were recorded with American Society of Anaesthesiologist (ASA) physical status classification [15]. Baseline bleeding risk was recorded with HAS-BLED score [16] as a surrogate measure. CHA₂DS₂-VASc score was used to stratified a patient’s thromboembolic risk [17].

Bleeding complications were defined as acute hospital representation with clinically significant haematuria, requiring further investigation and review by urology service, insertion of indwelling catheter for bladder irrigation, clot retention requiring manual washout, or haematuria requiring cytotropic intervention under general anaesthetics. Blood transfusion requirement was also recorded as a surrogate measure for severity of haematuria. All representations would be captured through the state-wide integrated electronic medical record. All patients had routine outpatient follow-up 3 months after TURP.

Statistical Analysis

Statistical analyses were performed using STATA 18 software (Stata Corp., College Station, Texas, USA). The cohort were separated into two groups, the Anticoagulated (AC) group, and the non-anticoagulated (No-AC) group. Comparison was made between patient treated with 5-ARI preoperatively versus patient without 5-ARI pre-treatment amongst the two groups. Two-sided t-test were performed to compare the continuous variable, and Chi-square test was performed to compare categorical variables between the groups. Complex probably-weight adjusted regression analysis was performed as a form of treatment-effect analysis to compared risk of clinically significant haematuria between patients receiving 5-ARI versus no 5-ARI in both anticoagulated, and non-anticoagulated cohort. The variables in the weight-adjustment analysis were identified using univariate logistic regression to look for co-variables, which might have influenced on outcome measured. A cut off threshold of p-value = 0.10 for co-variables to be included in the treatment-effect was used.

Results

Cohort Summary

There are 629 patients in the retrospective cohort, 113 (17.9%) received anticoagulation and 516 (82%) were not anticoagulated (Table 1a). In the overall cohort, the number of patients receiving preoperative 5-ARI, and no 5-ARI were almost even, 325 (51.7%) and 304 (48.3%) patients, respectively (Table 1b). Of all the patients receiving 5-ARI, majority of patients were treated with dutasteride, and only 28 (4.5%) received finasteride. This pattern remained similar in both the anticoagulated and the non-anticoagulated cohort, with patients 5.9% and 4.3% of patients receiving finasteride, respectively. Almost even number of patients received 5-ARI (50.9%) in the non-anticoagulated cohort (Figure 1a). An increased proportion of patients received pre-operative 5-ARI in the anticoagulated cohort, 60.8% (Figure 1b) and this finding was statistically significant (P = 0.02).

Table 1a: Overview of anticoagulation status in 629 patients and management in the perioperative period.

Table 1b: Overview of 629 patients receiving 5-ARI.

Figure 1: a) 5-ARI treatment breakdown of the non-anticoagulated cohort and b) 5-ARI treatment breakdown for the anticoagulated cohort.

Overall Cohort Outcome

Cohort characteristics were compared between the AC vs. no-AC cohort, and 5-ARI vs. No 5-ARI cohorts (Table 2). The independent variables and outcomes between the AC versus no-AC cohorts were important. Any statistically significant differences between the two cohorts will need to be accounted for final adjusted analysis. It was noted that patients in the anticoagulated cohort were significantly older, higher in BMI, higher incidence of pre-operative urinary tract infection, and more likely to received 5-ARI. It was also important to note that patients in the anticoagulated had higher baseline bleeding risk, as this is reflected in higher portions of patient categorized to have high to very-high risk HAS-BLED score. It was also worth noting that anticoagulated patients had significantly higher co-morbidities reflected in higher proportion of higher ASA score. The anticoagulated cohort patient had significantly higher risk of developing clinically significant haematuria post TURP (Table 2). AC patients were more likely to represent with haematuria requiring further assessments and interventions compare to the no-AC group, 27% vs. 9.9% (P < 0.01). AC patients were more likely to developed transfusion requirement (AC: 4.4% vs. No-AC: 1%, P <0.01). reflecting more severe bleeding. We have observed that more patients from the AC group require bladder catheterization for manual and continuous irrigation for haematuria management. However, this finding was not statistically significant. We observed that there were more patients in the no-AC cohort requiring cystoscopy washout under general anaesthesia. This finding was statistically insignificant and given this was a rare event (less than 1% of cases), it would require a much larger study cohort to be able to observe demonstrable statistical difference if existent. It was worth noting that more patient in the AC cohort was more likely to report prolonged haematuria of greater than 14 days (AC: 19% vs. No-AC 1.6%, P < 0.01) with a longer mean duration of haematuria of 42 days. There were differences in mortality rate between the two groups.

Cohort characteristics were compared between the 5-ARI and no 5-ARI group. This comparison was important in the subsequent adjustment for the treatment-effect analysis as statistically significant covariables between the two cohort would need to be accounted for. We have observed that patients who received 5-ARI were statistically significant older in age, had higher pre-operative PSA, and had higher prostate volume. On our unadjusted analysis, patients receiving 5-ARI were more likely to developed clinically significant haematuria prompting acute representations (5-ARI: 16.3% vs. No 5-ARI: 9.4%, P = 0.01). Other than this, there were no significant differences in complications between the two groups.

5-ARI Treatment-Effect Analysis (Non-Anticoagulated Vs. Anticoagulated)

Weight-adjusted analysis was performed from the above results to further assess the treatment effect of pre-operative 5-ARI on reducing clinically significant haematuria. Covariable with statistically significant difference in the above analyses (Table 2) between the AC vs. no-AC group, and 5-ARI vs. no 5-ARI group were carried forward for adjusting final analysis. Separate univariate linear regression analyses were performed on remainder covariables to potential variables that may impact outcome for adjustment. The weight-adjusted result from our treatment-effect model were demonstrated in Table 3. In the overall cohort, patients who received pre-operative 5-ARI were at 0.56 times reduced risk of developing clinically significant haematuria compared to the no 5-ARI group (5-ARI: 5.8%, P = 0004; No 5-ARI: 10.2%, P < 0.001). Similar risk reduction ratio was observed in the anticoagulated patients as well. In the AC group, patients who received pre-operative 5-ARI were at 0.61 times reduced risk of developing clinically significant haematuria compared to the no 5-ARI group (5-ARI: 11.6%, P = 0.012; No 5-ARI: 18.9%, P = 0.032). It was also worth noting that patients in the anticoagulated group had overall higher risk of developing haematuria related complications regardless of whether pre-operative 5-ARI was administered. However, it is worth noting that the risk of clinically significant haematuria in anticoagulated with 5-ARI treatment was similar to patients without receiving 5-ARI in the non-anticoagulated group.

Table 2: Cohort comparison of patient characteristics and outcome - with or without oral anticoagulation, and with or without 5-ARI treatment.