Abdominoperineal resection was initially described by Miles and was considered the standard of care for cancers of the mid and distal rectum for several decades. However, total mesorectal excision with sphincter preservation has supplanted abdominoperineal resection as the standard in these patients because of changes in our understanding of cancer behavior, management of localized disease, and techniques for tumor excision and anastomosis construction.

Regardless, abdominoperineal resection is still employed for advanced tumors of the distal rectum, mainly in the presence of direct elevator or anal sphincter involvement. Even without sphincter involvement, if the patient has sphincter dysfunction or has significant co morbidities, abdominal perineal resection is a viable alternative. Initial studies reported worse oncological outcomes associated with abdominoperineal resection compared to sphincter preserving procedures. Specifically, abdominoperineal resection was associated with worse local recurrence and overall survival rates when compared to low anterior resection. Differences in local recurrence rates were quite significant, ranging from 15-33% for abdominoperineal resection and 1-13% for low anterior resection. In addition, 5-year overall survival rates after abdominoperineal resection were within 38-60% and 57-76% for low anterior resection.[1], [2], [3], [4] Moreover, these differences persisted despite the introduction and standardization of total mesorectal excision.

Two large prospective European trials confirmed that outcomes were worse in patients undergoing abdominoperineal resection.[5],[6] Both trials also reported that circumferential margin positivity and intraoperative tumor perforation were independent predictors of worse oncological outcome. The increased positive circumferential resection margins after abdominoperineal resection (41% versus 12%) generated considerable discussion regarding conduct of the procedure. Accordingly, many centers focused on their operative technique and adopted a modified approach (extralevator or "cylindrical" abdominoperineal resection) whereby care is taken to halt the abdominal portion of the operation before the anorectal ring is reached and the levator muscles are transected further laterally, leaving a cuff of levator muscle attached to the specimen (Figure 10.1). This better ensures proper circumferential margins and avoids "waisting" of the surgical specimen in the area where the tumor is often located. Levator excision may be tailored as indicated by tumor involvement (partial versus total levator excision).

Figure 10.1

The abdominal part of the operation is performed through an open or minimally invasive approach while adhering to the principles of total mesorectal excision. Specifically, the dissection is conducted outside the mesorectal fascia. However, the dissection is halted upon reaching the levator muscles, leaving the levator muscles attached to the lowest portion of the mesorectum. This avoids "waisting" of the specimen that usually occurs at the level of the anorectal ring, which is usually near the tumor location (Figure 10.2). The lower limits of the dissection are clearly defined: upper third of the coccyx in the posterior part of the dissection; the seminal vesicles or uterine cervix in the anterior dissection, and; at the level of the hypogastric plexuses in the lateral portion. Once these landmarks are reached, the transabdominal dissection is halted, and a surgical laparotomy pad or gauze can be placed in the mobilized retrorectal space to guide the perineal dissection. A silicone drain may be left in the pelvis and exteriorized through the lower abdominal wall. The sigmoid colon is transected, an end colostomy is constructed (Figure 10.3).

Figure 10.2

A common technique is to complete the perineal portion in the prone position.[7] If so, then the abdomen is closed, and the patient is placed in a prone jack-knife position with the hips slightly abducted. If this technique is employed, prior to abdominal closure, a plan needs to be made for closure of the perineal defect. While it is frequently the final part of the operation, planning starts early. Since the bony confines of the pelvis prevent tissue collapse and lead to significant dead space, pelvic infection requiring opening of the perineum with resultant prolonged wound healing and a chronic perineal wound is not uncommon. Rotating a pedicle of well-vascularized omentum or a formal vascularized myocutaenous flap into the pelvis should be considered to reduce dead space and facilitate perineal healing, especially in patients who have received neoadjuvant pelvic radiation or have other risk factors for poor wound healing. If a primary closure is to be employed, then mobilizing an omental flap can help to reduce the dead space. An adequate omental pedicle flap can be easily fashioned by dividing the transverse colon and lesser sac attachments of the omentum, detaching the omentum from right or left side, ligating the gastroepiploic pedicle and short gastric vessels. Care is taken to avoid injury to the contralateral gastroepiploic vessels because this provides the blood supply to the flap. In cases of anterior exenteration, extensive perineal skin loss, need for vaginal reconstruction, or sacrectomy, a myocutaneous (e.g., rectus abdominis, gluteal, gracilis) flap is preferred (Figure 10.4).

When starting the perineal portion, the anus is closed with a purse-string suture, the perineum is prepared with a bactericidal solution, and the operative site is draped in a sterile manner. An elliptical-shaped incision is performed immediately outside the external sphincters with the posterior extent of the incision at the level of the coccyx and the anterior extent situated over the middle of the perineal body. The lateral aspects of the incision overlie the ischial spines. The dissection continues until the levator muscles are encountered from below, and their insertions/attachments to the lateral wall of the pelvis are identified. The pelvic cavity is entered through continued dissection in a cephalad direction into the posterior pelvis through the presacral fascia with or without disarticulation of the coccyx (Figure 10.5). Next, the laparotomy pad or gauze previously placed in the retrorectal space is identified and removed to afford more space for retraction. The dissection continues to work towards the anterior quadrant by transection of the levator muscles close to their lateral bony insertions/attachments. After muscle transection has been completed, the perineal and transabdominal pelvic dissections are joined. The neurovascular bundles are exposed in the lateral walls of the pelvis and posterior aspect of the prostate or vagina. The transperineal dissection can continue onto the anterior plane or facilitated in selected patients by careful eversion of the specimen from the pelvic cavity through the posterior aspect of the perineal wound. This eversion improves the surgeon’s visualization of the anterior rectal wall and seminal vesicles/prostate or vagina, and resection of the vagina or Denonvilliers’ fascia is typically enabled by this view. In male patients, anterior resection can be facilitated by palpation of the foley catheter to avoid urethral injury. Last, the dissection is concluded with division of the anterior pelvic diaphragm muscles.

In a multicentric study of 300 abdominoperineal resections, the extralevator approach was compared to standard abdominoperineal resection.[8] The extralevator abdominoperineal resection was significantly associated with less circumferential margin positivity (20% versus 49%) and intraoperative perforation rates were less frequently encountered (8% versus 28%). More recently, a literature review that included 5,244 patients found significant differences in circumferential resection margin positivity and intraoperative perforation rates between the extralevator abdominoperineal resection and standard abdominoperineal resection (9.6% versus 15.4% and 4.1% versus 10.4%, respectively).[9] Moreover, local recurrence rates were significantly lower for extralevator abdominoperineal resection (6.6% versus 11.9%) after a median follow-up of 68 months. This review also suggested performance of the procedure in a modified lithotomy position was associated with a higher risk of positive circumferential resection margin and intraoperative perforation, compared to prone position.

In a retrospective study from a single institution comparing standard abdominoperineal resection in the lithotomy versus prone jack-knife position, both approaches resulted in similar pathological findings and postoperative morbidity rates. The circumferential resection margin positivity tended to be lower in the group of patients operated in prone position (2.3% versus 8.5%).[7] However, the study was weakened by the fact that patients operated in the lithotomy position were more frequently managed with neoadjuvant chemoradiotherapy. Lastly, visualization of the perineum and surgeon’s ergonomics are potential advantages of an extralevator abdominoperineal resection performed in the prone jack-knife position. During a standard abdominoperineal resection approach, the surgeon usually operates in a limited space between the patient’s legs and the assistant is retracting from a more remote position. Under these circumstances, any blood accumulated within the pelvis from the prior abdominal procedure compromises the posterior and lateral dissections while impaired optics complicate the anterior dissection. On the other hand, the prone jack-knife position provides several members of the surgical team adequate visualization and any blood pools away for the dissection. Despite the additional time required for repositioning the patient, the benefits in surgical field visualization ease the perineal phase of the operation. Regardless, surgeon preference, and, if applicable, the approach to perineal reconstruction employed by the plastic surgeon, may dictate whether the procedure is performed with combined lithotomy and transperineal approach or prone approach as long as the principles of wide resection are maintained.