Bleeding

Bleeding during RRP is most commonly encountered during division of the dorsal venous complex of the penis. Before the description of the anatomic approach to radical prostatectomy by Walsh, severe hemorrhage was often encountered during uncontrolled division of the venous complex. On occasion, such bleeding was severe enough to require cutdown onto the prepubic penile shaft and identification of the retracted, uncontrolled venous complex beneath Buck’s fascia for ligation.

Despite the advances in surgical technique, the potential for significant bleeding during division of the dorsal venous complex and the remainder of the procedure still exists. The mean intraoperative blood loss reported in large series of radical prostatectomy has varied tremendously and ranges from 579 mL to >2 L and is likely influenced by operative technique, the presence or absence of nerve sparing, time, and the experience of the surgeon. Several surgeons have suggested various means of perioperative blood management to minimize the need for transfusion. Rates of allogeneic transfusion in contemporary series have varied from 2.4% to 21% independent of autologous blood donation. Reoperation for bleeding is extremely rare, with a reported incidence as low as 0.3% and 0.6% in two large series.

The use of autologous blood donation was previously shown to reduce the risk for allogeneic transfusion. Several investigators challenged the need for blood donation. Koch and Smith reported an overall transfusion rate resulting from intraoperative bleeding of 2.4% in 124 patients undergoing radical prostatectomy without preoperative autologous blood donation. Because it is unlikely that most surgeons will be fortunate enough to experience such a minimal likelihood of intraoperative and postoperative bleeding, it is advisable that some form of preventive management be instituted in patients undergoing radical prostatectomy.

One of the limitations of autologous blood donation is all men are destined to have significant preoperative and postoperative anemia. It has been estimated that the endogenous response to autologous donation produces about one unit of red blood cells. Because of potential risks associated with transfusing autologous blood, the indications for transfusing autologous and allogeneic blood are similar. Therefore the only men who benefit from autologous blood are those who have significant bleeding and require replacement of all three stored units. Therefore the impact of autologous blood donation is a modest reduction in allogeneic blood transfusion at the expense of significant postoperative anemia.

Sultan et al. have reported that return to work and physical activities are dependent on discharge hemoglobin and timing for removal of the urinary catheter. Therefore a blood management strategy should consider both the potential advantages of decreasing allogeneic transfusion rates and minimizing postoperative anemia.

Chun and associates proposed the use of preoperative erythropoietin injections instead of autologous donation. Patients were given one to two injections of erythropoietin, provided their preoperative hematocrit was <48%. On average, an increase in the hematocrit of 3% was noted when 600 U/kg was administered 7–14 days preoperatively. Using this method, the investigators demonstrated an identical risk of allogeneic transfusion in individuals donating autologous blood. Based on safety, efficacy, and dose ranging studies, we currently recommend administering epoetin alfa 300 IU/kg 7 and 14 days preoperatively in men with a baseline hematocrit <48%. The postoperative hemoglobin of the group receiving erythrocyte-stimulating proteins was significantly greater than in the autologous blood donation group. The short- and long-term cardiovascular and oncologic safety of erythrocyte-stimulating proteins has been demonstrated.

Although dorsal venous complex bleeding can occur regardless of the chosen technique of control, several anatomic points can be helpful for avoiding profuse hemorrhage. The dorsal venous complex courses beneath the pubis symphysis as a bundle of veins over the membranous urethra before it fans over the anterior surface of the prostate and bladder neck. On uncontrolled division, variable amounts of bleeding can occur. Initial descriptions of the anatomic retropubic prostatectomy included isolation of the dorsal venous complex with passage of a right-angle clamp between the membranous urethra and the complex, followed by sharp division with a knife blade. We have, in recent years, chosen to ligate the complex proximally and distally before division.

The dorsal venous complex is covered by an overlying reflection of the endopelvic fascia. To control all the veins fanning over the prostate within the complex effectively, the endopelvic fascia must be fully incised along the lateral sulcus of the anterior prostate. Removing the periprostatic adipose tissue overlying the anterior surface of the prostate allows visualization of the structures required to control the dorsal venous complex. This incision is made in the groove between the prostate and levator musculature during gentle retraction of the prostate contralaterally ( Fig. 42.1A ). The incision must be carried to the apex of the prostate and bladder neck to allow effective mobility of the proximally located veins. Small perforating vessels may be controlled with surgical clips.

The technique of dorsal venous complex control. A, The endopelvic fascia is sharply incised on either side of the prostate. The incision extends from the base of the prostate to the puboprostatic ligament. B, An Allis clamp is used to bunch the dorsal venous complex centrally. The initial bunching is done over the midprostate, proximal to the prostatic apex. Bunching is performed by inserting the jaw of the clamp into the endopelvic fascial incision on each side and thereby incorporating all veins as they fan over the prostate. C, Proximal and distal bunching sutures are in place in the dorsal venous complex. The surface of the anterior prostate reveals no excluded veins coursing over the gland. The first suture is placed proximally at the level of the bunching clamp. A second suture is then placed distal to the puboprostatic ligaments, around the dorsal venous complex, as far caudad as possible without division of the puboprostatic sutures.

(From Tareen B, Godoy G, Taneja SS, Complications of Radical Retropubic Prostatectomy, in Complications of Urologic Surgery: Prevention and Management, 4th Edition, Taneja SS (ed), Saunders Elsevier, Philadephia, PA, 2010.)

A decussation of the endopelvic fascia, termed the puboprostatic ligament, lies at either lateral edge of the venous complex fusing its anterior surface to the posterior surface of the pubic bone. We do not routinely divide the puboprostatic ligaments unless they fan laterally and limit placement of the distal suture ligature controlling the dorsal venous complex. Incising the puboprostatic ligament may facilitate dropping the venous complex away from the pubis. This is particularly useful when operating on large glands that compress the dorsal venous complex or glands with a large anterior or retropubic component and when the puboprostatic ligaments extend onto the anterior surface of the prostate gland. In these situations, we divide only that portion of the puboprostatic ligament needed for exposure. Preservation of the puboprostatic ligament will hold the vein stump fixed, to some extent, and prevents retraction and uncontrolled bleeding.

Once the endopelvic fascia has been incised on either side of the gland and the superficial dorsal vein has been defined and fulgurated, the deep venous complex is bunched centrally for control (see Fig. 42.1B ). One approach is to use an angled or straight long Allis clamp inserted with one jaw in each facial incision. The venous complex is then bunched centrally at the level of the midprostate, with care taken to include all veins but avoiding prostatic capsular injury. Difficulty with bunching may be corrected with further proximal incision of the endopelvic fascia. Once the clamp is in place, the venous complex is controlled with a suture ligature of 2-0 polyglactin 910 (Vicryl) on a wide curved needle, at the level of the clamp placement. The clamp is removed during ligation of the proximal vein. Next, a suture ligature is placed around the distal venous complex as far caudal as possible (see Fig. 42.1C ). This is achieved by placing the Allis clamp more apically and gently depressing the visible prostatic apex, thus exposing the distal vein as it courses over the membranous urethra. Double passage of each suture ligature will fix it in place and prevent it from slipping off on division of the vein. Alternatively, securing the proximal suture ligature at the midgland can serve the same function as the Alis clamp.

Sharp incision of the dorsal venous complex is then initiated between the proximal and distal sutures ( Fig. 42.2 ). A plane of dissection is developed sharply within the dorsal complex, with great care taken not to injure the anterior prostatic capsule. A sponge stick is used by the assistant to retract the prostate cranially and gently roll it away from the pubis as the venous tissue is divided. The surgeon should always err on the side of cutting out the distal suture rather than injuring the anterior prostate. If the distal suture is cut out, complete division of the dorsal venous complex should be completed, and then the dorsal venous complex stump can be directly suture ligated. Incision of the dorsal venous complex is sequentially carried out until the anterior urethra is exposed. An effort should be made to preserve the anterior thickness of the urethra maximally.

Division of the dorsal venous complex begins by incising the centrally bunched tissue between the proximally and distally placed sutures. A sponge stick is used to rotate the gland dorsally and cranially away from the pubic symphysis as the dorsal complex is divided. In doing so the dorsal complex is rotated into the field as it is sequentially divided. Antegrade dissection is carried out within the dorsal venous complex until the membranous urethra is rotated into the field of view. Caution must be exerted to avoid anterior prostate injury.

(From Tareen B, Godoy G, Taneja SS, Complications of Radical Retropubic Prostatectomy, in Complications of Urologic Surgery: Prevention and Management, 4th Edition, Taneja SS (ed), Saunders Elsevier, Philadephia, PA, 2010.)

An alternative approach and advantage of the open approach is the ability to palpate the prostatourethral junction between the thumb and forefinger. A McDougal clamp is then passed at this precise location in order to guide division of the dorsal venous complex so that the entire complex is divided while maximally preserving the rhabdosphincter.

Bleeding can occur throughout the procedure. This bleeding often originates from the bed of the neurovascular bundle if nerve sparing is performed, from the seminal vesicular arteries, or from the bladder neck. The surgeon must balance securing hemostasis while minimizing trauma to the neurovascular bundles. Using a lap pad to pack the pelvis is an effective technique to minimize pelvic bleeding during the dissection and minimizes blood loss through venous compression. The prostatic pedicles are usually directly ligated before division and for this reason should not be a major source of bleeding. Bleeding from the neurovascular bundles can occasionally be controlled by clipping prostate perforating branches before division, but this is often a fruitless endeavor. In a nonnerve-sparing procedure, early ligation of the neurovascular bundle can save a great deal of trouble when one tries to suture ligate the stump once it is retracted. When steady bleeding is encountered, it is usually best to move ahead with the procedure because operative time likely predicts the cumulative blood loss. Fulguration or clipping of the neurovascular bundles may inadvertently injure the nerve and therefore should be avoided.

Some investigators have advocated an antegrade approach to radical prostatectomy to avoid early bleeding, but in general, if the procedure is to be performed in a timely fashion with early control of the dorsal venous complex, this should not be necessary.

In a personal experience of almost 5000 cases, intraoperative control of bleeding was not achievable in a single case and the patient was closed after tightly packing the pelvis with multiple lap pads. The lap pads were removed the following day with preservation of anastomotic integrity.

Significant and life-threatening hemorrhage typically occurs within a few hours after completion of the case. The surgeon is confronted with the decision to return immediately to the operating room or attempt conservative management with the hope bleeding will spontaneously subside. Significant blood in the pelvis is associated with prolonged hospitalization and catheter drainage with increased likelihood of strictures. In cases with significant hemodynamic instability due to bleeding, we recommend immediate return to the operating room to control bleeding, evacuate the pelvic hematoma, and re-do the anastomosis if necessary. This approach results in an expedited recovery and limits postoperative strictures and incontinence.

Rectal Injury

Injury of the rectum occurs in <1–3.6% of patients undergoing radical prostatectomy. Risk factors for rectal injury include small prostates, previous radiation, and periprostatic inflammation resulting from previous transurethral resection or excessive bleeding at the time of biopsy. Salvage prostatectomy carries the highest rates of intraoperative rectal injury, with reports ranging from 6% to 15%. Immediate repair of the injury results in excellent outcome without long-term sequelae in the majority of patients.

Following division of the posterior urethra, the surgeon encounters the rectourethralis muscle along with underlying Denonvilliers’ fascia. It is desirable to divide this layer in a single incision and thus enter the pararectal space and allow an adequate posterior margin of resection. If the surgeon is too timid in division, aggressive retraction of the prostatic apex can result in shearing of Denonvilliers’ fascia from the posterior apex and splitting of the gland. If the surgeon is too aggressive in achieving this plane, sharp injury of the rectum can occur. In general, rectal injury results from an overzealous attempt to develop the posterior plane bluntly with the tip of the finger. If the rectourethralis is not completely divided, or if the plane does not develop easily with blunt dissection, the attached rectal wall can be torn. Sharp dissection is preferable if the apex of the gland does not lift away from the rectum easily. It is also advisable to first incise the visceral layer of the endopelvic fascia and develop the plane between the prostate and rectum beginning laterally. An obscured operative field from excessive bleeding hinders the ability to see the tissue planes clearly. Therefore adequate hemostasis is preferable at the time of the apical dissection.

When rectal injury does occur, it is generally acceptable to repair the defect primarily in the absence of gross fecal soiling. Patients should have been given a preoperative enema specifically for this reason. Débridement of devitalized tissue at the site of laceration should be followed by a two-layer closure of mucosal and muscular layers of the rectal wall. A small opening can be made in the peritoneum to pull down and interpose an omental flap between the rectal repair and the vesicourethral anastomosis. Although this maneuver is not routinely necessary, it can provide an extra safety measure if the surgeon is concerned about the viability of the rectal closure. A pedicalized flap of peritoneum or pararectal fat can serve a similar purpose. In the event of a large injury, a significantly devitalized or devascularized rectal wall, or the presence of gross soiling, serious consideration should be given to a temporary diverting colostomy along with simultaneous rectal wall repair. This procedure is certainly advisable in the setting of a previously irradiated pelvis.

The risk of rectourethral fistula following repair of rectal injury is lowered. Use of prolonged bowel rest, anal dilation at the time of the surgical procedure, and the administration of broad-spectrum antibiotics have all been suggested as potential aids in lowering fistula rates. Perhaps the greatest risk of fistula formation and risk of morbidity stem from unrecognized rectal injury. If a rectourinary fistula does develop and the patient is otherwise well, an initial attempt at conservative management can be attempted using bladder drainage and a low-residue diet. Complicated cases or patients with a history of radiation will likely require use of a temporary diverting colostomy.

Ureteral Injury

Ureteral injury during radical prostatectomy is rare, occurring in 0.05–1.6% of cases. When it does occur, it is usually during the bladder neck division, although with many surgeons advocating extended lymphadenectomy, the ureter may be injured above the iliac bifurcation during node dissection. In the setting of a large median lobe, care must be exercised to avoid the ureteral orifices by directly visualizing them. We prefer to use a partial bladder neck–sparing approach, and when doing so, caution must be exerted to avoid straying into the trigonal musculature.

The administration of intravenous indigo carmine on division of the bladder neck should be routinely performed to visualize ureteral efflux on both sides. This is particularly true for the surgeon inexperienced in prostatectomy, but is a good general habit for all surgeons. Occasionally, in a patient who has undergone aggressive transurethral resection of the prostate, the ureteral orifices can be cut during division of the bladder neck. Our simple solution is to perform a ureterotomy for a length of approximately 1 cm, insert an indwelling stent, and then taper the bladder neck posteriorly to relocate the orifices proximally.

On rare occasions, injury to the ureter can occur during posterior dissection of the seminal vesicles. The ureters generally enter the posterior bladder wall anterolateral to the tips of the seminal vesicles and cross toward the midline above the trigone. For this reason, blind placement of sutures should be avoided in the seminal vesical bed. If one suspects an injury, passage of a 3Fr feeding tube can allay concerns. In rare circumstances, an on-table retrograde pyelogram can be helpful.

If ureteral injury is identified, immediate ureteroneocystostomy is indicated. The cut end of the ureter is pulled into the bladder lumen at a convenient site, spatulated, and sutured to the bladder mucosa. A long submucosal tunnel is not necessary. The vesicourethral anastomosis should be completed first to avoid tension on the ureteral anastomosis. Similarly, if tension is encountered, a Boari flap is preferable to a psoas hitch to avoid undue tension on the vesicourethral anastomosis. Ultimately, this injury remains very rare for experienced surgeons. Heightened operative vigilance is needed when normal anatomic boundaries and tissues have been altered by severe inflammation and fibrosis. For example, in a series of 1500 consecutive open radical prostatectomies, the only case of a ureteral injury occurred in a patient with unrecognized schistosomiasis who had extensive pelvic fibrosis.

Nerve Injury

Injury to the obturator nerve and femoral nerve can occur during pelvic mobilization, and these injuries are discussed in detail in Chapter 40 .