Venous thromboembolism in radical prostatectomy: is heparinoid prophylaxis warranted?


Mark S. Soloway, Department of Urology, PO Box 016960 (M814), Miami, Florida 33101, USA.



To review the incidence of venous thromboembolism (VTE) after radical retropubic prostatectomy (RRP) and evaluate the need for heparinoid prophylaxis as opposed to mechanical compression devices after RRP.


RRP is classified as a category 1 (high risk) procedure for VTE by the American College of Chest Physicians and several international guidelines recommend subcutaneous heparinoids as the preferred prophylaxis. However, this regimen may be associated with a greater risk of bleeding. We have not used heparinoid prophylaxis but place a mechanical compression device for prophylaxis of VTE, and report our clinical experience over a 12-year period. Between 1992 and 2004, all RRPs carried out by one surgeon (M.S.S.) at our centre were retrospectively reviewed after obtaining institutional review board approval. The protocol for prophylaxis of VTE consisted of compression stockings and a sequential compression device from the time of entry into the operating room until complete ambulation (we encourage early ambulation). Patients were evaluated for VTE if they developed any clinical signs or symptoms. Patients were followed at 7 days, 6 weeks and 3 months after RRP in the first year and 6-monthly thereafter. All relevant clinical data and complications were entered in a database.


In all there were 1364 RRPs; the mean (sd) age of the patients was 61 (7) years and the mean follow-up 44 (38) months. All patients had a mechanical compression device and ambulated on the first day after surgery. None received heparinoid prophylaxis. Three VTE events were identified (0.21%); two patients had a lower limb VTE and one an upper limb VTE. All were successfully treated with anticoagulation. No patient had a documented pulmonary embolus and none died from VTE. There was one death after RRP, from myocardial infarction.


The incidence of VTE after RRP is low, possibly related to the use of a mechanical compression device and early aggressive mobilization. Despite the recommendations by some, we feel that routine heparinoid prophylaxis is questionable.


radical prostatectomy


intermittent compression device


thromboembolic deterrent stockings


sequential compression device


venous thromboembolism


deep vein thrombosis


pulmonary embolism.


Radical prostatectomy (RP) is one of the more common oncological procedures. Venous thromboembolism (VTE) is an important cause of death in patients with pelvic malignancy, and the prevention of VTE is a concern in abdominal and pelvic surgery for cancer. It is estimated that 1–5% of contemporary patients undergoing major urological surgery have a symptomatic VTE [1]. The risk of an associated fatal pulmonary embolism (PE) is estimated to be <0.25%[1]. Although all urological surgeons agree that early ambulation is critical to preventing VTE, various other methods are used as prophylaxis. This includes low molecular weight heparin or other heparinoid prophylaxis, and sequential mechanical compression devices. In Europe, anticoagulation is standard as VTE prophylaxis, while mechanical devices are preferred in North America.

The primary attraction of either graduated compression stockings or the use of an pneumatic intermittent compression device (ICD) is their safety. While mechanical methods of prophylaxis have been shown to reduce the risk of deep vein thrombosis (DVT) in several groups of patients, they have been studied much less intensively than anticoagulant-based options and are generally thought to be less effective than the latter for preventing DVT [1].

There are no large prospective randomized trials to evaluate different methods of VTE prophylaxis in urological procedures [1]. At our institution, we routinely use a mechanical device for thromboprophylaxis in RP. We present our experience with prophylaxis against VTE during or after retropubic RP (RRP) for cancer of the prostate.


After obtaining Institutional review board approval, all RRPs performed by one surgeon (M.S.S) between January 1992 and September 2004 were retrospectively reviewed. Demographic, clinical, and pathological data were entered into a computerized prospective database; all complications were also prospectively entered into the database. Patients with clinically localized prostate cancer were admitted on the day of surgery. Mechanical thromboprophylaxis in the form of thromboembolic deterrent stockings (TEDs) were placed before RRP and a sequential compression device (SCD) placed on the lower extremities before inducing anaesthesia. Most patients had spinal anaesthesia supplemented by sedation. A standard RRP through a lower midline incision was used until 2002; from 2002, a modified Pfannenstiel incision was used [2]. A bilateral modified pelvic node dissection was used in 67% of patients, and intraoperative cell salvage where blood loss required replacement. No patient received an allogenic blood transfusion. A pelvic drain was not routinely inserted from 2002 [3].

Patients were aggressively encouraged to ambulate on the first day after RRP, with TEDs and SCDs continued until full ambulation. Patients were discharged 1 or 2 days after RRP, with the Foley catheter removed at 7–10 days. Patients were followed at 6 weeks, 3 and 6 months, and 6-monthly thereafter, at each visit being routinely assessed for clinical evidence of DVT; if there was clinical suspicion of DVT, radiological investigations were ordered.


Between January 1992 and September 2004 there were 1373 RRPs; the patients’ characteristics, clinical and pathological data and complications are listed in Table 1. There were three instances of DVT; two were in a lower limb and one in an upper limb. No patient had a clinical PE. There was one death after RRP from a myocardial infarction, confirmed on autopsy. Patients with DVT were treated with heparin initially, followed by warfarin; all three cases resolved with anticoagulation treatment, with no sequelae.

Table 1. 
Patient, operative and pathological characteristics
Number of patients1373
Mean (sd):
 Age, years  60.8 (7)
 Initial PSA level, ng/mL   9.1 (12.7)
Clinical stage, n (%)
 T1 851 (62)
 T2 495 (36)
 T3  26 (2)
Neoadjuvant treatment, n (%) 259 (19)
Adjuvant treatment, n (%)  56 (4)
Mean (sd, range) follow up, months  44.7 (38.2, 0.13–149.6)
Mean (sd) estimated blood loss, mL 572 (330)
Gleason score:
 Mean (sd)   6.6 (1)
 Distribution, n (%)
 2–6 592 (43)
 7 593 (43)
 8–10 188 (14)
n (%):
Surgical margins 484 (35)
Lymph nodes  21 (1.5)
Seminal vesicle involvement 147 (11)
Extraprostatic extension 261 (19)


VTE is considered to be one of the most critical and potentially fatal complications after major urological procedures [1]. Most of the epidemiological data related to VTE in this population were derived 10–30 years ago. Subsequent changes in surgical care, early mobilization, and possibly greater use of prophylaxis have been associated with declining rates of VTE [1].

Patients undergoing RRP often have many risk factors for VTE, including advanced age, malignancy, and pelvic surgery with or without lymph-node dissection [4,5]. Additional factors for DVT may include the duration of surgery and hypotension [4,5]. RRP has been classed as a high-risk procedure for VTE by various international guidelines for preventing VTE, including the American College of Chest Physicians and the Scottish Intercollegiate Guidelines Network [1,6–8]. It is recommended that heparinoid prophylaxis with low-dose unfractionated heparin or low molecular weight heparin be used for VTE prophylaxis for this procedure. The use of mechanical devices like TEDs and SCDs should be considered if heparinoids cannot be used. There are no randomized studies to compare the use of heparinoids and mechanical devices in RRP, or indeed any urological procedure. Thus, the optimum approach to thromboprophylaxis in these patients is unknown. However, in many parts of the world, notably in Europe, anticoagulation is not only common, but considered mandatory.

Anaesthetic techniques have improved and there is generally a more aggressive approach to mobilizing patients after surgery. RRP currently causes less blood loss and is faster; at our centre there has been no example of a PE in >1300 consecutive procedures. The low incidence of VTE events could possibly be attributed to early mobilization in addition to the use of TEDs and SCDs. Another factor could be the maintenance of normotension during surgery aided by the use of intraoperative cell salvage where necessary. The use of a Pfannenstiel incision is not a factor, as this is a recent change. A bilateral modified pelvic-node dissection, which may be considered an additional risk factor, was used in more than two-thirds of the patients, but the VTE rate was low despite this.

A recent non-randomized prospective study of patients undergoing RP, in which early ambulation, ICDs and postoperative warfarin were use, reported an incidence of VTE (detected by duplex ultrasonography) of 2.8% on screening at 3–4 weeks after RP, and 0.6% had a symptomatic VTE event [5]. In a series of 1300 consecutive RP, Cisek and Walsh [9] reported clinically manifest DVT and PE in 0.3% and 0.9% of 784 patients. The patients received neither ICD nor pharmacological prophylaxis for VTE. The authors compared this result with that in 516 patients who received ICD prophylaxis alone, where the rates of clinically manifest DVT (0.6%) and PE (1.7%) were slightly higher. There was a statistically significant delay (20 days, sd 12) in the onset of thromboembolic symptoms in patients who received ICD prophylaxis. In a retrospective analysis of 1000 patients undergoing RP, Lepor et al.[10] reported clinically manifest DVT/PE in 0.5% but the method of VTE prophylaxis was not specified. Heinzer et al.[11] retrospectively studied 508 patients who had RP, with low-dose heparin prophylaxis; for the first 320 patients the rate of PE (1.6%) and DVT (4.7%) was higher than that of a subsequent 188 patients, at PE (0.5%) and DVT (1.6%). The latter 188 patients all had an ultrasonographic evaluation of the pelvis 1 day after RP and every 2 days thereafter until discharge. The authors concluded that a pelvic lymphocele or haematoma were important cofactors in the pathophysiology of thromboembolic complications.

Regardless of the method of prophylaxis for VTE, patients are encouraged to ambulate on the day after surgery and in our centre are discharged from the hospital 1 or 2 days after RRP. Most DVTs are silent; we do not screen for a DVT because of the low incidence of clinical DVT/PE.

We do not provide prolonged prophylaxis to prevent a delayed VTE. The use of prolonged prophylaxis of VTE was examined by Bergqvist et al.[12], who concluded that the administration of enoxaparin prophylaxis for 4 weeks after general surgery in patients undergoing abdominal or pelvic surgery for cancer was both safe and effective in reducing venographically detectable thrombosis. The rates of DVT determined by bilateral venography 3 months after surgery were 5.5% for the patients who received 4 weeks of enoxaparin and 13.5% in those who received enoxaparin for 1 week after surgery. The study was not sufficiently powered to detect a significant decrease in PE or mortality. In contrast, Lausen et al.[13] were unable to show that prolonged prophylaxis with tinzaparin had a significant effect on the incidence of DVT occurring late after general surgery. The value of prolonged prophylaxis of VTE specific to patients having urological surgery is unknown.

We acknowledge that the present study has limitations; it was not randomized and was retrospective. We did not routinely use ultrasonography of the lower extremities or pelvis, and the clinical examination at follow-up was the only method used to detect VTE. However, the very low incidence of clinically significant VTE is noteworthy and raises questions about the need for heparinoid prophylaxis, which is a standard practice in many centres and remains as a guideline by some authorities.

In conclusion, in a retrospective review of a RRP database which captures complications, the incidence of DVT and clinical PE was 0.21% and none. The routine use of SCDs and TEDs, as well as early mobilization, may be an important aspect of prophylaxis for VTE. Because of the excellent outcomes with mechanical devices and early mobilization, and the potential risk and expense of heparinoid prophylaxis, we feel that the routine use of heparinoid prophylaxis is questionable.


None declared.