Ex-vivo comparison of the haemostatic properties of standard transurethral resection and transurethral vaporization resection of the prostate




To compare the haemostatic properties of standard transurethral resection of the prostate (TURP) and transurethral vaporization resection of the prostate (TUVRP), as perioperative bleeding is still regarded as the major complication of prostate resection.


Isolated blood-perfused porcine kidneys were used to determine the haemostatic efficacy of TURP and TUVRP (using two different electrodes). Bleeding was quantified precisely in relation to tissue ablation for the two techniques, and specimens were evaluated histologically.


Both TUVRP groups had significantly less bleeding (P = 0.005) than the TURP group for a standardized ablation volume of perfused kidney tissue (18.9, 19.5 and 24.1 mL/min, respectively). The different TUVRP electrodes had no significant haemostatic differences. The histology showed significantly (P = 0.03) larger coagulation zones for the TUVRP groups than for standard TURP.


TUVRP ex-vivo was associated with significantly better haemostasis than TURP. The haemostatic properties of different active electrodes for TUVRP seem to be equivalent.


Various alternative treatment devices, e.g. electrovaporization, needle ablation, laser, ultrasound or microwave therapy, have recently become available for treating BPH [1–6]. However, for efficacy, TURP is still regarded as the reference standard by most urologists. TURP produces immediate and enduring success rates. In an attempt to combine the efficacy of conventional TURP with the favourable safety profile of transurethral vaporization of the prostate (TUVP), the so-called vaporization resection of the prostate (TUVRP) was introduced [7–9]. This procedure uses modified, broader resection loops and higher electrosurgical energies. The technique is claimed to integrate the excellent tissue debulking properties of TURP and the haemostatic tissue ablation by vaporization, desiccation and coagulation, as seen in TUVP. Contrary to TUVP, TUVRP allows high-quality specimens to be recovered for histopathological examination [7,10].

Using an elaborate ex-vivo animal model we have quantified precisely the haemostatic properties of TUVRP and compared it with standard thin-loop TURP. The need for this basic experimental evaluation of TUVRP is supported by the divergent findings of the few published clinical trials prospectively comparing haemostasis of TUVRP and TURP [11–14].


We modified an ex-vivo model of isolated blood-perfused porcine kidney (IBPK), introduced by Koehrmann et al.[15] and Michel et al.[16], to quantify bleeding during different high-frequency current ablation procedures (Fig. 1). In all, 17 kidneys were taken from pigs within 6–10 min after slaughtering. The organs were immediately perfused through the renal artery with sodium chloride solution (0.9%). Organs were preserved at 4°C when the effluent from the renal vein was clear. The experiments started no later than 3 h after death. For perfusion, autologous blood was harvested, heparinized with 8000 IU/L and stored at 4°C.

Figure 1.

The experimental system.

A roller pump carried the blood through an 8 F catheter intubating the renal artery, at a perfusion rate set to 60 mL/min with a perfusion pressure of 100–115 cmH2O. Experiments were carried out strictly with the corresponding autologous blood of each pair of kidneys. The organs were placed in a glass basin, irrigated at 20°C with standard mannitol/sorbitol solution and earthed with a standard neutral electrode.

The high-frequency current for the resections was generated by the ICC 350 device (ERBE GmbH, Germany). Resections were carried out in the HighCut (automatic electric arc control) mode and a power output of 150 W, ‘effect 3’ for standard thin-loop TURP. For ‘thick-loop’ TUVRP, as in the clinical setting, the HighCut mode and a power output of 200 W, ‘effect 3’ was used.

As the active electrode for standard thin-loop TURP, a conventional 24 F loop (Olympus GmbH, Germany) was used. For ‘thick-loop’ TUVRP the so-called Band electrode (Olympus) and the gold-plated Wing electrode (Richard Wolf, Germany) were assessed (Fig. 2).

Figure 2.

Different active electrodes: the Band electrode (left), Wing electrode (middle) and conventional TURP wire loop (right).

The renal tissue was resected in a standard way to ablate a total volume of ≈ 16 mL (4 × 4 × 1 cm deep). The electrodes were attached to a standard 26 F resectoscope. As described previously, a constant drag speed and pressure applied were provided by clamping the resectoscope into a stepper motor, which was controlled by custom-made software [17]; the drag speed was set to 10 mm/s for all experiments.

After the resection was complete the irrigation fluid was immediately drained from the basin to monitor the induced bleeding. As reported by Michel et al.[16], the blood loss was quantified by weighing a swab before and after it was held on the bleeding surface for 30 s. Finally, the surface and the surrounding tissue were dissected and fixed in 10% formalin, and processed for histological examination using haematoxylin-eosin staining. The depths of the coagulation zones were determined with a measuring calliper under the microscope. Depths were recorded eight times along the resection surface within each group and then averaged.

Unpaired data were analysed using a nonparametric Mann-Whitney U-test between TURP and TUVRP, and between the different TUVRP resection electrodes.


In general the blood loss induced by the tissue resection was recorded precisely using the IBPK model; the results were highly reproducible, generating low sds. Generally, during and after tissue ablation, blood loss induced by ‘thick-loop’ TUVRP was visibly lower than with ‘thin-loop’ TURP. The mean (range) blood losses with TURP (seven replicates), and TUVRP with the band (six) and Wing (six) electrodes were 24.1 (20.1–28.5), 18.9 (17.2–21.2) and 19.5 (17.6–22.1) mL/min, the last two being significantly lower than TURP (P = 0.005). The differences in blood loss between the band and Wing electrodes were not significant.

Because of the experimental system the coagulation zones were very homogeneous and reproducible. For the power settings noted and a constant drag speed of 10 mm/s there were no apparent carbonization effects. The mean (range) coagulation zone for TURP, and TUVRP by band and Wing were 0.6  (0.2–0.8), 1.0 (0.6–1.1) and 0.9 (0.5–1.2) mm, respectively, with the difference between TURP and the TUVRP band being significant (P = 0.03), but not between TURP and TUVRP wing (P = 0.08).


Over recent years many alternative methods have been introduced for treating BOO caused by BPH [1–7,16], the rationale for this continuous search to improve surgical techniques being the persistent complication rate of conventional TURP [18–20]. Perforation of intra- and periprostatic arteries and veins during resection, causing bleeding and absorption of irrigation fluid, remains the main problem of TURP. With a rapidly ageing population, leading to more high-risk patients requiring treatment, the new techniques claim to be less invasive than TURP. However, the major drawback of most minimally invasive procedures for treating BPH is the failure to attain the immediate success rates of conventional TURP. TUVRP is a recent modification of TURP, using several types of thicker and broader resection loops than the standard wire loop. With additional electrosurgical energy, this modification claims to combine the effective tissue ablation of pure TURP with the superior haemostatic properties of electrovaporization.

Despite the expanding clinical use of TUVRP little basic research information has yet been published for this treatment [7,17,21]. Basic work emphasising haemostatic properties has been reported mostly for TURP and TUVP [17,22–24] but these authors unanimously rely on in vitro results for their studies. By measuring the depths of coagulation zones the haemostatic performance of the different treatments is inferred, but the extent of the coagulation zone or the ‘heat-affected’ zone merely serves as an indicator of in vivo haemostasis. The present ex-vivo model enables bleeding to be monitored and measured, thus allowing a direct comparison of the haemostatic properties of the different techniques.

In this model, with TUVRP (with both electrodes) there was significantly less haemorrhage than with TURP; the bleeding was slightly less for the band than the gold-plated Wing electrode but the difference was not significant. The better haemostasis with TUVRP than TURP arises from two main effects: (i) With the geometry of the electrode, TUVRP provides four times the duration of contact between active electrode and tissue. Given a constant drag speed this increases the amount of energy penetrating the tissue; (ii) coupled with a higher baseline power output in TUVRP, this energy, converted into thermal energy, creates expansive and thorough coagulation and desiccation of the tissue. These changes can reduce haemorrhage, as seen in this ex-vivo model.

There are only four prospective, randomized clinical trials comparing TUVRP and standard TURP published to date. Only two could confirm the better haemostasis with TUVRP [12,13]; the others reported no clinically relevant advantages for TUVRP over TURP [11,14]. An explanation for these contrary findings might be that, clinically, haemorrhage must be viewed as a result of several factors, some of which might not occur in an ex-vivo model. Intrinsic coagulopathy, e.g. as in patients taking aspirin (41% of the TUVRP cohort vs 27% in the TURP cohort in [14]) will obviously not be assessed by a laboratory model. Moreover, the vascular pattern of the porcine renal tissue used only represents intraprostatic vascularization. Any division of periprostatic or sinus vessels, with significantly larger lumina, will have very different haemostatic properties. This could partly explain the apparent lack of better haemostasis in TUVRP in humans. If the prostatic tissue is resected radically, not respecting the prostatic capsule, massive periprostatic or sinus bleeding may not be reversed by TUVRP.

Coagulation zones in the present TUVRP samples were> 150% larger than those generated by TURP; these findings correlate well with results published recently [17]. In a similar laboratory in-vitro model, coagulation depths averaged 156% for TUVRP and 233% for TUVP above standard TURP. This correlation of coagulation depths in vitro and haemorrhage ex vivo supports the assumption that the extent of coagulation zones in vitro can serve as a reasonable indicator of the haemostatic properties of a novel ablation technique. However, Ishikawa et al.[21] reported no significant differences in heat-affected zones in vitro in porcine muscle among four different vaporization-resection electrodes and a standard loop. Using these different electrodes in a canine in-vivo model created significant differences in coagulation zones among the different electrodes. The coagulation volume depends on several separate variables, e.g. coagulation increases with a decrease in drag speed, an increase in energy and an increase in pressure applied [17,21,22]. In the present ex-vivo model we could adjust each variable precisely and thus eliminate any interference. In clinical or in-vivo settings these variables obviously cannot be kept constant, necessarily leading to different coagulation zones [10,21].

Thus with the present ex-vivo method, we provide substantial evidence that TUVRP could cause significantly less perioperative bleeding than TURP.


transurethral vaporization of the prostate


transurethral vaporization-resection of the prostate


isolated blood-perfused porcine kidney.