Study Type – Therapy (RCT)
Level of Evidence 1b
Study Type – Therapy (RCT)
To compare bipolar with the conventional monopolar transurethral resection of the prostate (TURP) for blood loss and speed of resection.
PATIENTS AND METHODS
In all, 202 consecutive patients from the hospital waiting list were randomized to undergo TURP using either a bipolar system (Surgmaster TURis, Olympus, Tokyo, Japan) or a monopolar system (24 F, Storz, Tübingen, Germany). The blood loss during and after surgery was measured using a photometer. Other variables compared included indices of resection speed and transfusion rate.
There were no statistically significant differences in operative duration, resection weight, resection speed or radicality of resection. However, the median blood loss was 235 mL for the bipolar and 350 mL for monopolar TURP (P < 0.001). The decrease in blood haemoglobin concentration during the day of surgery was smaller in the bipolar group (5.5% vs 9.6%P < 0.001). Fewer patients were transfused with erythrocytes (4% vs 11%, P < 0.01), which can be explained by the much lower 75th percentile for blood loss in the bipolar group (at 472 vs 855 mL, respectively).
Bipolar TURP using the TURis system was performed with the same speed as monopolar TURP but caused 34% less bleeding, the difference being greatest (81%) for the largest blood losses. Bipolar TURP also required fewer erythrocyte transfusions than the conventional monopolar technique.
transurethral incision of the prostate
prothrombin time/International Normalized Ratio
TURP was developed in the USA during the 1930s for treating BPH; despite the more recent introduction of medical therapies, the high efficacy of TURP still makes it one of the most common surgical procedures worldwide. In Sweden, ≈6500 TURPs are performed every year. TURP is rarely conducted as day surgery, as the morbidity of bleeding events during and after surgery, as well as the TUR syndrome, can require special attention. Erythrocytes are transfused in 3–23% of the patients who undergo monopolar TURP [1–3]. In previous studies comparing bipolar and monopolar TURP, transfusion rates are reported as 0–11%.
In the past few years bipolar TURP has become readily available worldwide [5,6]. This method has theoretical advantages that have not yet been fully evaluated. In the present study we compared bipolar and conventional monopolar TURP for the speed of resection and the blood loss during surgery. The haemorrhage was also measured during the first few hours after TURP, as the bleeding cannot be stopped with sutures. Blood loss is measured more accurately in transurethral surgery than in other operations, because all lost blood is directly dispersed in the irrigating fluid [7,8]. With the Low HemoCue system (HemoCue, Ängelholm, Sweden), a precise measurement (±5%) can be obtained at the bedside [8,9].
PATIENTS AND METHODS
Between December 2005 and August 2008, 202 patients were randomized to undergo TURP using either the bipolar Olympus TURis (in saline) system or the conventional monopolar technique. Inclusion criteria were symptomatic BPH that required surgery (due to failed medical therapy or urinary retention) and a TRUS-estimated prostatic volume of 30–100 mL. Patients with prostate volumes of <30 mL were treated with transurethral incision of the prostate (TUIP) and those with prostates of >100 mL had open surgery. Exclusion criteria were evident prostate cancer and a core biopsy of the prostate within 3 months before scheduled surgery. The local Ethics Committee in Stockholm approved the study, and written informed consent was obtained from all patients before surgery.
Factors such as preoperative medication, indwelling catheter, surgical treatment and the surgeon’s experience, that could affect the results, were assessed in subgroup analyses.
Conventional monopolar TURP was performed with a 24 F resectoscope, using an ERBE ICC 350 generator (Storz, Tübingen, Germany) set at 130 W (cutting mode) and 50 W (coagulation mode). All resections were made using standard loops and factory-made irrigating fluid that contained mannitol 3% and ethanol 1% (Baxter, Kista, Sweden) . Bipolar TURP was performed with the TURis system using the UES 40 generator (Olympus, Tokyo, Japan) set at 280 W for cutting and 100 W for coagulation. All resections were performed with the 0.2 mm 12° medium loop. Normal saline solution (0.9% NaCl; Baxter) was used as the irrigant. Ethanol was added to reach the same concentration as in the electrolyte-free irrigating fluid . This offered the same possibility for early detection and quantification of fluid absorption by a hand-held alcolmeter in both groups. Surgery was terminated when there was evidence of uptake of irrigation fluid, i.e. when there was ethanol uptake registered in the alcolmeter.
(Hb concentration, g/L × irrigant volume (L))/preoperative blood Hb concentration (g/L)
The Hb concentration was measured 1–7 days before TURP and 6–18 h afterward. The photometer can be conveniently carried into the operating theatre. The atoxic reagent is already present in small disposable vials, and achieves haemolysis within 60 s. The Hb concentration was obtained from the irrigant collected during and after TURP. Bleeding after TURP was measured from end of surgery until the irrigation was switched off. The reliability of the HemoCue system for measuring blood loss during TURP was evaluated previously [1,9]. To prevent coagulation, 15 000 IU of heparin was added to every 10 L container of returned irrigant at the operation, and 5000 IU to each 3 L urinary bag after TURP.
The decision to transfuse erythrocytes was made when the postoperative blood Hb concentration had decreased to 75–100 g/L. The specific ‘transfusion trigger’ for each patient was determined by the patient’s health status.
The radicality of the TURP was expressed as the weight of the resected tissue divided by the preoperative prostate volume, as measured by TRUS. ‘Bleeding velocity’ was obtained by the perioperative bleeding divided by the operating time.
Patients on treatment with anticoagulants (warfarin or salicylate) were told to discontinue treatment 3 and 7 days before surgery, respectively. The prothrombin time/International Normalized Ratio (PT/INR) was measured and a level of ≤1.6 was accepted.
The study was designed to detect a difference in total blood loss of 30% between the groups. The Mann–Whitney U-test was used to evaluate numerical variables with a skewed distribution. Categorical variables were analysed using the chi-square test. Factors of importance to blood loss were identified by stepwise multiple regression analysis. Numerical data showing a skewed distribution were square-root-transformed before being tested by linear regression. In all tests, P < 0.05 was considered to indicate significance.
Of 202 randomized patients, 17 were excluded from the evaluation; four did not fulfil the inclusion criteria, and other reasons for exclusion were changeover to TUIP (four), withdrawal (three), and the detection of bladder cancer (three) or prostate cancer (three) while awaiting surgery. Of the 185 remaining patients, there was no statistically significant difference between the bipolar and monopolar TURP groups before surgery for prostate volume, micturition time, IPSS and bother scores, PT/INR, and Hb levels, the use of 5α-reductase inhibitors (5ARIs) and the prevalence of an indwelling catheter (Table 1). However, the patients in the monopolar group were slightly older than the others (73 vs 70 years, P = 0.007). 5-ARIs were used by 48 patients, of whom 44 had been on treatment daily for >3 months. A third of the patients had an indwelling catheter at the time of TURP, and the others had a timed micturition showing low-flow values as a sign of significant obstruction.
|Mean (sd), n|
|Age, years||69.5 (7.2), 98||72.7 (8.4), 87||0.007|
|Prostate volume, mL||55.6 (18.2), 98||58.2 (17.6), 86||ns|
|Hb before TURP, g/L||141.0 (13.1), 98||141.8 (14.0), 87||ns|
|PT/INR||1.1 (0.1), 98||1.1 (0.1), 87||ns|
|IPSS pre-op, points||21.7 (6.9), 31||20.4 (7.6), 32||ns|
|Bother score pre-op, points||3.9 (0.9), 31||3.7 (1.1), 32||ns|
|Voiding time, s/1st dL||24.5 (14.4), 21||27.7 (17.9), 20||ns|
|Resection weight, g||27.3 (15.1)||26.3 (13.2)||ns|
|Operative duration, min||62 (23)||66 (23)||ns|
|Resection speed, g/min||0.46 (0.21)||0.40 (0.15)||<0.05|
|Resection radicality, %||48.8 (21.0)||45.3 (15.9)||ns|
|Median (25th−75th percentile):|
|Blood loss, mL|
|During TURP||235 (127–415)||350 (175–660)||<0.001|
|After TURP||8.9 (0–34.0)||13.5 (2.0–54.4)||ns|
|Total||262 (150–472)||399 (186–855)||<0.001|
|Hb decrease, %||−5.54 (−9.91, −2.43)||−9.59 (−16.88, −5.51)||<0.001|
|Blood loss/resected tissue, mL/g||11.43 (6.52–18.11)||14.48 (9.83–26.98)||<0.002|
|Bleeding velocity, mL/min||3.79 (2.41–6.13)||5.71 (3.33–8.57)||<0.003|
|Bleeding velocity corrected for resection weight, mL/g/min||0.18 (0.12–0.32)||0.24 (0.17–0.39)||0.003|
|Subgroup analyses (P, Mann–Whitney U-test)|
|Median (25th−75th percentile), n bleeding, g/min|
|All patients||0.14 (0.10–0.28), 98||0.23 (0.15–0.33), 87||<0.001|
|5-ARI||0.13 (0.09–0.31), 23||0.23 (0.17–0.34), 25||0.007|
|Anticoagulants||0.11 (0.09–0.14), 4||0.26 (0.19–0.39), 9||0.014|
|Pre-op indwelling catheter||0.13 (0.08–0.20), 33||0.21 (0.15–0.30), 34||0.02|
|Reuter’s trocar||0.12 (0.08–0.14), 19||0.19 (0.14–0.30), 25||0.03|
|Residents||0.27 (0.09–0.32), 9||0.12 (0.11–0.14), 5||0.606|
Two residents performed 14 of the evaluable operations. All the others were performed by 10 specialists in urology with ≥5 years of urological experience. The TURP was terminated due to uptake of irrigation fluid in two patients in the bipolar group and in four in the monopolar group. These resections were all nearly completed when absorption was recorded and thus were terminated late in the surgical procedure.
Intermittent irrigation was used in 79 patients in the bipolar (81%) and in 62 (71%) in the monopolar group. Continuous irrigation with Reuter’s trocar was used in the others. The type of irrigation was chosen according to the operating surgeon’s preference.
There was no difference in resection weight or resection radicality between bipolar and monopolar TURP. The resection speed was slightly higher in the bipolar group. The operative duration tended to be longer in the monopolar group, but the difference was not statistically significant (Table 1).
The blood losses during TURP and in total were significantly smaller in bipolar than in monopolar TURP (P < 0.001; Fig. 1), but the blood loss after TURP was similar (Table 1). Similarly, the bipolar technique showed lower values for bleeding when corrected for operating time and weight of the resected tissue (Table 1). The decrease in Hb level during the day of surgery, both as a percentage and in absolute values, was smaller in the bipolar group (P < 0.001). Four patients in the bipolar group (4%) and 10 patients in the monopolar group (11%) were transfused with erythrocytes (P < 0.01). This difference can be understood from the much lower 75th percentile for blood loss in the bipolar group (472 vs 855 mL, respectively; Fig. 1).
The differences in blood loss corrected for resection weight and operative duration remained in the subgroups containing only patients treated with 5ARIs, or with an indwelling catheter and continuous irrigation (Table 1). However, although few TURP were performed by non-specialists, they seemed to be associated with a larger blood loss when using bipolar than monopolar TURP.
Stepwise multiple regressions based on all patients were used to identify which factors served as statistically significant independent predictors of blood loss. Three factors were found; the operative duration, the weight of the resected tissue and the choice of mono/bipolar surgery (all P < 0.001). The use of suprapubic trocar was not significant. There was a poor, but still statistically significant, linear relationship between the measured blood loss and the relative and absolute change in blood Hb from before TURP to the morning after the surgery (r2 = 12–13%, Fig. 2).
Bipolar TURP gave similar results to monopolar TURP for commonly used measures of surgical efficiency, e.g. the amount of resected tissue per unit of time. However, bipolar TURP resulted in 34% less bleeding during surgery. Moreover, the 75th percentile of blood loss differed as much as 81% between the groups, which means that the bipolar technique had greatest effect on the largest haemorrhages. Therefore, bipolar TURP was followed by a significantly lower need for erythrocyte transfusions. We believe this reduction is an important advantage, as every step taken to limit the haemorrhage benefits both the patient and the surgeon.
To date, the strongest argument for using bipolar TURP is that the 0.9% saline used as the irrigant is believed to prevent the potentially life-threatening TUR syndrome . The present study shows that the bipolar technique markedly reduced surgical haemorrhage, which might be just as important. The reason for the lower haemorrhage is unknown but might be related to the effect of different types of current on opened blood vessels. The difference pertained only to the surgery, when cutting was actually used. After TURP the blood losses recorded for the two methods were similar, although it amounted only to 3.4% of the total blood loss.
Bleeding in conventional monopolar TURP has always been a problem, from different points of view, either expressed as the number of severe haemorrhages requiring erythrocyte transfusion or leading to shock-related complications, such as acute myocardial infarction. Massive perioperative bleeding also impairs visibility in the surgical area and makes it difficult to identify anatomical landmarks. The reduced blood loss in bipolar TURP would alleviate these problems and reduce the stress for the surgeon caused by profuse bleeding, especially when teaching. The smaller haemorrhage should make education in TURP easier and more comfortable for doctors in training.
Bleeding during this type of surgery is an easily overlooked issue, as adequate measurement methods are rarely applied. The ‘Low Haemoglobin’ photometer method is an objective way of measuring the Hb concentration of the irrigating fluid, which is easily transformed into an estimate of the amount of blood lost . Alternative methods include visual estimation, which might be grossly inaccurate [7,12]. The potassium concentration after haemolysation might also be used . A commonly used approach is to take the reduction of blood Hb as an index of the blood loss . The present study, using the Low Haemoglobin photometer method as a reference, showed poor precision for assessing blood loss from changes in Hb concentration. The number of erythrocyte transfusions given, which was fewer among those who were treated with bipolar TURP, is also a less precise measure of blood loss, as the ‘transfusion trigger’ Hb level involves a subjective evaluation of health status.
Previous assessments of the blood loss in bipolar vs monopolar TURP have yielded inconclusive results. Either they were based on decreases in Hb level [15,16] or different estimates of haemorrhage, such as visual analogue scales  or indicator-dilution method [13,17]. However, all of these methods provide only an indirect gross estimate of the haemorrhage and do not quantify the actual amount of blood lost. One study that used the Low Haemoglobin photometer method showed a statistically significant difference in blood loss between bipolar and monopolar TURP, but without reporting the amount of resected tissue . Bleeding after TURP has not been evaluated in any previous study. We found the blood loss after TURP to be small in both study groups, but this variable is still of value when evaluating TURP-associated haemorrhage.
Limitations of the present study include that the differences in medical and surgical treatments created subgroups of patients. However, our analysis of those subgroups suggests that they cannot explain our key finding with respect to blood loss. Hence, a suprapubic trocar modifies fluid absorption but hardly affects blood loss , a result supported by our subgroup analysis (Table 1). Despite early reports of lower surgical haemorrhage in patients treated with a 5ARI, larger randomized studies have not supported such an effect [1,9]; nor was such an effect detected in the present study. Finally, the few TURPs done by residents-in-training tended to weaken rather than explain the overall difference in blood loss between bipolar and monopolar TURP, as they were associated with a larger haemorrhage.
TURP is now challenged by a range of minimally invasive techniques, such as different laser methods and microwave treatments [20,21]. There is a need for further studies in this field, comparing the bipolar technique with laser ablation in terms of bleeding, safety and long-term outcomes. Surgical resection remains favourable when there is a need for a histological specimen, providing a correct pathological diagnosis.
Equipment for bipolar TURP is manufactured by several medical companies and is marketed internationally, despite the lack of high-quality evaluations. However, it is still possible that bipolar TURP is followed by less severe consequences of fluid absorption, bleeding events are less serious, the surgeon’s view during surgery is clearer, the risk of damage to the urethral sphincter is lower, and the time for healing is shorter. Therefore, the possibility of TURP with less bleeding, fewer postoperative bleeding events, and a reduction of the number of erythrocyte transfusions is of value.
In conclusion, bipolar TURP reduced the overall perioperative and total surgical bleeding by 34%. The greatest difference, 81%, was apparent for the largest haemorrhages. The need for transfusion due to TURP was significantly reduced when using bipolar TURP.
CONFLICT OF INTEREST