BACKGROUND AND EVOLUTION
Exenteration was first reported by Brunchswig in 1948 as a palliative procedure, and, in his subsequent series of 561 patients collected from 1948 to 1964, the operative mortality was 26% and the 5-year survival rate was 20%.1 Advances in anesthesia, surgical instrumentation, technique, pelvic floor reconstruction, antibiotics, fluid management, blood products, the use of intermediate or intensive care units, and the training of gynecologic oncologists have decreased operative mortality; in addition, improved patient selection preoperatively and intraoperatively have increased survival rates. At the Mayo Clinic from 1950 to 1986, the operative mortality rate decreased from 12% to 6.7%, and the 5-year survival rate increased from 26% to 41%.2 Patient rehabilitation has also remarkably improved. From loss of vaginal function and 1 or 2 permanent stomas, reconstructive techniques have resulted in no stoma bags (due to low rectosigmoid anastomosis and continent urinary reservoir), and restoration of vaginal function by using different type of pedicle flaps. Other factors to consider for all exenteration candidates are the need for family support due to long rehabilitation, usually 6 months or longer depending on the extent of the operation and whether there were or not postoperative complications, quality of life before and after surgery, and the need for appropriate care if an exenteration cannot be performed or cure is not achieved.
INDICATIONS AND CLINICAL APPLICATIONS
Pelvic exenteration remains a salvage operation when other forms of therapy have been exhausted. Most candidates are patients with cervical cancer postirradiation or those with advanced primary disease. Because of dosimetry improvements and the addition of sensitizing concomitant chemotherapy, the number of exenteration candidates has decreased, with only about 8% of patients with advanced cervical cancer being candidates. Other candidates are patients with recurrent endometrial, vaginal, urethral, vulvar, rectal, bladder, and, even occasionally, ovarian malignancies. Most often, the clinical scenario for consideration of exenterative surgery involves a cervical or endometrial cancer patient presenting with a recurrence after having received pelvic irradiation for a postsurgical recurrence. Occasionally, patients with an advanced primary malignancy of the vulva or vagina, who are not candidates for irradiation, are best treated by primary pelvic exenteration.
Prognostic factors impacting survival and the risk of morbidity include the patient’s age, body mass index (BMI), general medical condition, the origin site of the primary lesion, the time interval from primary treatment to recurrence, the volume and location of the recurrent lesion, histological type of tumor, lymphatic invasion, degree of tissue invasion and pelvic viscera involved, parametrial invasion, levator muscle invasion, postresection margins, positive nodes, peritoneal involvement, small bowel involvement, lateral pelvic wall extension or fixation, invasion of other soft tissues or bone, and extrapelvic metastases.2,3,4,5,6 The combination of several adverse prognostic factors decreases survival rates.5
Optimal candidates for pelvic exenteration are young patients in good health with a normal BMI presenting with an isolated central pelvic recurrence from squamous cervical carcinoma postirradiation or chemoirradiation, localized only to the cervix or to the cervix and the upper-most portion of the vagina, with a tumor volume of less than 5 mL, no lymphatic invasion, no parametrial extension, negative nodes, no metastases, and an interval of more than 2 years from completion of irradiation.5
Older age is not a contraindication to exenterative surgery; although it is associated with increased morbidity, age does not adversely impact survival.7 Similarly, obesity is not a contraindication but may affect the likelihood of morbidity. Patients in poor medical condition may simply not survive the operation or its complications. Properly selected patients with endometrial adenocarcinoma recurrence have a similar survival outcome as patients with cervical cancer recurrence.8 Patients with persistent or recurrent tumor within 6 months from primary treatment have poor survival rates even with complete resection.5 Factors such as tumor volume, location, degree of invasion, and extension to parametria and levators are important because they may impact the likelihood of obtaining satisfactory surgical margins. In general, large tumors portend decreased survival as compared with small lesions. Tumor volume is a factor in type I exenterations due to the cephalad extension of the lesions with direct peritoneal or small bowel involvement, but is not a factor in type II or III indicating that in these cases complete tumor resection is more important than the lesion volume.2 Complete resection (negative margins) is associated with increased survival. Incomplete resection (positive margins) has a dismal prognosis,3,5 with a median survival of only 4 months.9 A long interval (> 1 year) from primary treatment to recurrence is a favorable prognostic factor.5 Low rectal anastomosis has been associated with an increased risk of local recurrence.10,11 Although the cause is unknown, a potential factor could be an increased propensity for narrow margins resulting from attempted preservation of the rectum.
There are a number of both absolute and relative contraindications to exenterative surgery. The presence of extrapelvic metastasis at any site (such as supraclavicular nodes, grossly enlarged aortic nodes, liver, lung, omentum, and peritoneum) and extensive lateral pelvic wall involvement, are contraindications for an exenteration because the disease is not resectable or curable.
Absolute contraindications to exenterative surgery are extrapelvic involvement of any organ, grossly positive aortic nodes, peritoneal metastases, or situations in which completely resection cannot be achieved because the disease is not curable. Other adverse factors, such as microscopic aortic node metastases, lateral pelvic wall involvement, positive pelvic nodes, or involvement of bony pelvic structures are not absolute contraindications (see relative contraindications) but they may result in a lower survival. The combination of several adverse factors has an increasingly unfavorable impact on survival.
Contraindications to an exenteration are discovered in about 50% of patients explored for an exenterative procedure. In an early series, reasons for abandoning the procedure were pelvic sidewall involvement, positive aortic nodes, small bowel or cecal involvement, and liver metastases.12 Patients undergoing an abandoned exenteration attempt have a dismal prognosis. Approximately 75% of such patients die during the first year, 95% have expired by the end of the second year, and only 5% of patients survived for longer than 2 years.12 Adverse surgical factors are evaluated both preoperatively and at surgical exploration. Minimally invasive surgical exploration 48 to 72 hours prior to a planned exenteration can identify patients with surgical contraindications and avoid a laparotomy (see MIS exploration prior to exenteration in Minimally Invasive Surgery Applications section).
The presence of positive pelvic or aortic nodes, involvement of the pelvic peritoneum, sigmoid colon, small bowel, ovaries, tubes, fixation to the pelvic wall fascia, involvement of other soft tissues or bone, even when the lesion is completely resected, have been shown to result in decreased survival rates. Any of these findings reduce the likelihood of cure, and, because there are no other therapeutic alternatives, one may still consider proceeding with exenterative surgery. However, the presence of several adverse prognostic factors may be an indication for not proceeding with the exenteration. In our experience, factors associated with a decreased survival2,6 were observed in 15% of patients with type I, in 27% of patients with type II, and in 17% of patients with type III exenterations.2 When analyzing the entire group of patients with adverse factors, the 2- and 5-year survival rates were 47% and 17%, respectively.6 In a separate series of patients with adverse factors analyzed by the type of exenteration, there were no survivors among patients undergoing a type I or III exenteration, while a 28% 5-year survival rate was obtained for patients with type II.2 Exenterations with decreased survival have been named palliative, however, this term is misnomer since palliative procedures are directed to improve symptomatology and not survival rates.
There are several issues to consider in the presence of positive nodes prior to deciding to proceed with an exenteration. These factors include lymph node location (pelvic, aortic, sigmoid mesentery, groin, or other), whether the disease is microscopic or macroscopic, the number of involved nodes, whether nodal disease is unilateral or bilateral, extracapsular involvement, and surgical resectability. For instance, 1 or 2 unilateral positive pelvic nodes may not influence survival when other unfavorable factors are not present.6 In our experience, the survival of patients with positive nodes was 46% at 2 years and 23% at 5 years.6 In an earlier series of the Mayo Clinic there were no survivors with positive pelvic nodes,13 probably due to additional negative prognostic factors. A single aortic node with microscopic metastasis may not be a reason to abort an exenteration but multiple microscopic or grossly positive aortic nodes, especially if fixed, are an absolute contraindication. Positive groin nodes, when resectable and treatable with adjuvant groin irradiation, are not a contraindication in the presence of negative pelvic nodes.
Involvement of the peritoneum overlying the recurrent lesion or direct contact of the lesion with the adnexa, sigmoid, small bowel, cecum, or omentum are not absolute contraindications, but these conditions require removal of all the involved tissues.6 Small bowel resection was the most common additional tissue resected in type I exenterations due to the cephalad growth of the lesion with peritoneal or adjacent bowel involvement. In contrast, there were no bowel resections in patients with a type III exenteration due to the distal location of the lesions.2
Extension to the pelvic wall was considered an absolute contraindication in the past when discovered on pelvic examination. However, complete resection has been achieved in 29% of patients with apparent extension to lateral pelvic wall with a 32% 10-year survival rate.3 In some patients, fixation to the pelvic wall is secondary to radiation fibrosis and the recurrence can be easily resected. Others have lesions invading the pelvic wall muscular fascia of the obturator internus, levators, and coccygeus muscles, and are resectable (see Extended Radical Parametrectomy). Others have unresectable disease involving into or past the large sciatic notch or have large fixed recurrences attached to the external iliac vessels, which may be resectable, but these patients are left with subclinical residual disease and are not curable unless intraoperative irradiation is administered. Partial resections should not be attempted because they will not result in cure.
When preoperative evaluation of patients with postirradiation disease with lateral pelvic wall extension appears resectable, we recommend additional external pelvic irradiation (usually 4000 cGy in about 4 weeks), followed by an exenterative procedure 4 weeks later. An effort to achieve complete resection must be made in these instances, because local control is directly related to the size of the residual tumor. Once the resection is completed, intraoperative electron irradiation (Mobetron, IntraOp Medical, Sunnyvale, California) is delivered through a metal cylinder of an appropriate diameter to cover the selected target area (Figure 8-1). A dose of 1250 cGy is delivered in the presence of microscopic disease and may be increased according to the amount of residual visible tumor. Distant metastases and survival rates are directly related to obtaining a complete tumor resection. Tumor bed and local control rates of 81% and 67%, respectively, have been achieved when a complete resection was accomplished in patients with different types of recurrent gynecologic malignancies at different sites and with different surgical procedures.14 In the absence of intraoperative electron irradiation, immediate postoperative brachytherapy with iridium 192 needles is an alternative, providing a similar local control rate of 68%.15 Hollow plastic tubules are sutured directly over the tumor bed providing a homogenous dose (Figure 8-2) and are exteriorized through the anterior abdominal wall (Figure 8-3). An omental flap is created and sutured over the tubules to prevent direct bowel contact with them. The tubules are pulled out once the brachytherapy is completed. For intraoperative electron and brachytherapy irradiation, the survival rates were 55% and 44%, respectively, and severe complications were observed in 15% compared with 33% of patients, respectively.14,15
Bone involvement is different than periosteal attachment or fixation, which is easier to resect. Factors to consider prior to bone resection are whether complete resection of the entire lesion is possible, the risk of associated morbidity, long-term sequelae, and any other adverse prognostic factors. Bone invasion is associated with decreased survival rate even when all disease is resected.6 We recommend a similar approach to patients with lateral pelvic wall extension: preoperative irradiation, followed by tumor and bone resection, and administration of intraoperative irradiation. When the recurrent lesion is fixed to the periosteum, periosteal, or superficial bone resection with an osteotome will achieve complete clearance. The pubic symphysis can be resected without significant disability, as can the ischial tuberosity and a portion of the ischium ramus if necessary to achieve disease clearance. In time, a callus is formed between both ends of the transected bone providing some degree of stability. Most recurrences invading the sacrum are resectable since, as a rule, they tend to be distally located. Sacral involvement below S2 is considered resectable (Figure 8-4) while more proximal disease is generally unresectable. In our institution, preoperative consultation is obtained with an orthopedic oncologist who performs the bone resection.
Necrotic tissue secondary to irradiation ischemia may appear as necrotic tumor recurrence and may not provide viable tissue for a pathologist to confirm or rule out recurrent malignancy. Infrequently, an exenterative procedure is performed in a patient with necrotic, malodorous tissues, with an annoying vaginal discharge and severe pain only to find no evidence of malignancy on final pathological examination. In a Mayo Clinic series, 4 (3.4%) patients were exenterated with a preoperative clinical suspicion of cancer but without histological confirmation of recurrence.13 Two (1.2%) patients had actual recurrence on the specimen but no tumor was found in the other 2. These procedures result in a marked improvement of patient’s quality of life, especially if they have intractable pain, fistulous tracts, or both.
The triad of ipsilateral hydronephrosis, leg swelling, and sciatica has usually been considered a contraindication for an exenteration. Most of these lesions, especially when large, are unresectable. However, the authors have operated in patients with such symptomatology where the symptoms were not directly related to recurrent disease, the lesion was resectable, and the exenteration was curative. In many instances, the final decision of unresectability will be made at the time of surgical exploration.
The classification of exenterations into 3 main categories—anterior, posterior, and total—reflects only the specific viscera removed. This classification fails to address the caudal level of resection of the viscera and the excision of other soft tissues, bone, or both. The Mayo classification of exenterations16 provides a further understanding of the extent of the operation by taking into consideration the preservation or removal of additional tissues in addition to the pelvic viscera, including:
Level of resection of the pelvic viscera (above or below levator muscle),
Vulva or vulva-perineum-anus (vulvectomy or vulvo-anusectomy),
Additional soft tissues, such as small bowel, sigmoid, groin nodes, psoas muscle (Figure 8-5), pelvic wall muscles (Figure 8-6), or bone, such as pubic symphysis, pubic ramus, ischial tuberosity, and sacrum (Figure 8-4), or other structures (extended exenteration).
Extended exenteration. Resection of leiomyosarcoma recurrence involving the left external iliac vessels and psoas muscle. The recurrence has been dissected off the external iliac vessels and a segment of the psoas muscle has been resected. The left femoral nerve can be seen lateral to the external iliac vessels exiting the pelvis below the left inguinal ligament. Narrowing of the middle portion of the external iliac vessels secondary to tumor compression is also seen.
Using the above criteria, there are 3 main categories of exenterations, ie, anterior, posterior, and total, depending on the specific pelvic viscera removed. Each category is divided into 3 types, ie, types I (supralevator), II (infralevator), and III (infralevator with vulvectomy or vulvoanusectomy; Figure 8-7). The tissues removed in each type are summarized in Table 8-1. Type I involves resection of the pelvic viscera above the levator muscle (with preservation of levators, distal vagina, perineal membrane, urethra, vulva, and anus; Figure 8-8). Type II includes resection of the pelvic viscera below the levator muscles as well as the levator muscles (Figure 8-9A). There are 2 levels of visceral resection in type II:
Proximal to the urethral meatus, vaginal introitus, and anus, with preservation of the perineal membrane and perineum (Figure 8-9B), or
|Pelvic viscera||Above levator||Below levator||Below levator|
|Levator muscles||None||Partial or extensive||Extensive|
|Perineal membrane||None||(A) None or (B) minimal||Extensive or complete|
|Urethral meatus-vaginal introitus-anus||None||(A) None or (B) circumscribed resection||Extensive|
|Other soft tissues and/or bone with any typea|
(A) Total pelvic exenteration type II. Level of resection of pelvic organs. (B) Total pelvic exenteration type IIA. Specimen of large recurrent cervical carcinoma with preservation of urethral meatus, vaginal introitus, perineum, and anus. (C) Total pelvic exenteration type IIB. Specimen of recurrent cervical carcinoma with resection of urethral meatus, vaginal introitus, perineum, and anus.(D) Total exenteration type IIB showing the limited resection of urethral meatus, vaginal introitus, perineum, and anus as in the specimen of Figure 8-9C.
Type III is for distal or vulvovaginal lesions and involves type II b resection with excision of the perineal membrane plus a vulvectomy or vulvoanusectomy (Figures 8-10A to D). The term extended exenteration is added to any one of the described types when additional tissues or bone are removed.
(A) Total exenteration type III. The vulva, including the labia majora, the perineum and perianal tissues are excised in addition to the pelvic viscera, levator muscles, and perineal membrane. (B) Total exenteration type III. Surgical specimen of type III total pelvic exenteration showing large ulcerated vulvar carcinoma (on the right) and bladder, vagina (opened), and rectum on the left. (C) Surgical specimen of total exenteration type III for recurrent vaginal carcinoma postirradiation involving the distal left lateral vaginal wall, vaginal introitus, urethra, and posterior vaginal wall (ulcerated lesion). (D) Higher magnification of the left vaginal ulcerated lesion seen in Figure 8-10C.
Box 8-1 KEY INSTRUMENTATION
Long curved clamps, such as Shallcross clamp, long curved scissors, and Russian forceps
Electrocautery with long tip; vessel-sealing device
Stapling instruments for intestinal and vascular applications
Clip applier with large and small size clips
Preoperative evaluation is directed to assess the medical condition and the extent of the disease (pelvic and extrapelvic) of potential exenteration candidates. Patients in poor medical condition may simply not survive the operation because of the fluid and metabolic changes experienced during and after surgery and are not appropriate candidates. A medical examination should include palpation of nodal areas, especially the supraclavicular nodes. A detailed pelvic examination is performed to determine the lateral and caudal margins of the lesion, which in the presence of radiation fibrosis may be difficult or impossible to ascertain. In patients with severe pain or a narrow, tender, or irradiated vagina, it is preferable to perform an examination under anesthesia, at which time biopsies can be obtained, and cystoscopy and proctoscopy examinations can also be performed. Smooth parametrial induration may be fibrosis, not tumor, and be resectable. Positron emission tomography/computed tomography scan is the preferred imaging test to evaluate for metastatic disease. The presence of extrapelvic metastatic disease is an absolute contraindication. In the absence of metastases, the magnitude of involvement of the viscera, soft tissues, and bones of the pelvis must be carefully evaluated.
Patient counseling about the proposed operation and the resulting permanent life style changes related to urinary, intestinal, and vaginal functions require time and compassion. A strong family support is necessary for the long postoperative recovery, usually 6 months or longer when facing complications, which, unfortunately, are common. Consultation with a stoma nurse is necessary to assess for potential stomal sites and patient and family education of stoma care and potential complications.
On the day prior to the exenteration, the stoma site is marked in the standing position, the patient should remain on clear liquids after lunch and until midnight, have a bowel preparation at home, and take an evening shower with a chlorhexidine brush for the abdominal and vulvoperineal skin, which is repeated in the morning of the surgery day. All patients are typed and screened, and depending on the planned type of exenteration, with 2 or more units of blood crossmatched. Stockings and intermittent pneumatic devices are used in all patients and placed prior to the operation. Prophylactic antibiotics are administered 1 hour before the operation and repeated at 3 hours. It is advisable to have an anesthesia team familiar with the intraoperative management of the serious potential complication that can accompany exenterative surgery.
Box 8-2 MASTER SURGEON’S PRINCIPLES
Adequate exposure with an adequate size incision and a self-retaining retractor are a must for every exenteration
Sharp dissection of pelvic spaces is usually the rule for previously irradiated patients
Be prepared for unexpected bleeding at any time
Division of the ureters, rectum, or both is not performed until there is assurance of no contraindications
The omentum must be preserved for layering the denuded pelvic floor, an important part of the operation
Inferior epigastric arteries must be preserved when vaginal reconstruction involves the rectus muscle
Because prolonged anesthesia time is associated with increased morbidity and mortality, the surgeon must be prepared to proceed expeditiously and have an extensive knowledge of the pelvic anatomy and surgical techniques for the management of ureters, bladder, vagina, rectum, small bowel, pelvic and aortic nodes, and vulva, as well as being versed in reconstructive techniques for the urinary, vaginal, and intestinal tracts. Equally important is preparation for the management of severe hemorrhage, which can occur even with the best technique. The addition of other surgical teams in the operation room may not be in the best interest of the patient due to difficulty with coordination of schedules, different approaches, difference in level of experience of the surgeons, and the need for different operating room personnel, resulting in a “cluttered” operating room. We prefer a single surgeon for the exploration, exenterative, and reconstructive abdominal phases and a second surgeon for the reconstructive vulvovaginal phase simultaneously with the reconstructive abdominal phase.
The surgeon must exercise creativity and surgical judgment, and possess physical endurance and mental strength to determine and carry out the type of resection best suited for each patient. When the resection is tailored to the patient’s disease and anatomy, each operation is different. For instance, a fixed, right parametrial recurrence may require complete resection of the right hypogastric vessel system while on the contralateral side these vessels can be preserved. A lesion extending into the right lower vaginal third near the introitus, will require ipsilateral levator, perineal membrane, and vulvar resection, while the symmetrical contralateral tissues may be spared. Radiation fibrosis, which increases with the passage of years, may obliterate the usual pelvic spaces the surgeon is familiar with, requiring a more extensive resection than one may have planned originally. Alternatively, a small central recurrence in a patient with preservation of the spaces and planes of dissection may result in a radical or extended radical hysterectomy, instead of an exenteration. Obliteration of the vesicovaginal or rectovaginal spaces may require a total exenteration instead of the planned anterior or posterior exenteration, which would have been adequate to obtain clear margins. We have observed that patient survival is not affected by the type of exenteration, assuming complete tumor resection is accomplished.2
The type of instrumentation used for the division of vascular pedicles and soft tissues is a personal preference. Our preference is to use electrocautery for transection of soft tissues and a vessel-sealing device for vascular pedicles, unless it appears unsafe in which case the standard clamp-cut-suture technique is used. Soft rubber rugs are placed on the floor of both sides of the operating table for the surgeon and assistants to stand on.
The patient is placed in a low semilithotomy position using Allen stirrups with heel but not calf pressure. Leg position is intermittently checked during the operation by a circulating nurse. The full abdomen, vulva, perineum, anus, buttocks, and upper thighs are prepped and draped to provide immediate access when needed. A small degree of Trendelenburg facilitates bowel retraction in the upper abdomen, improves venous flow, and allows direct lighting to the deep aspects of the pelvis. One operating room light is positioned perpendicular to the surgical field and focused on the upper pelvis and lower abdomen while a second light is positioned over the patient’s head at an angle and directed to the deepest part of the pelvis. The small bowel and sigmoid are placed in the upper abdomen and packed with a long gauze pack or blue towels depending on the weight of the patient, and held in place with malleable self-retracting blades attached to the retractor’s ring.
Exenterative surgery follows a series of segments or phases:
Exploratory phase: surgical exploration to determine resectability and absence of contraindications
Exenterative phase: removal of pelvic viscera and other tissues
Vulvo-vaginal-anal phase: completion of the exenterative phase with removal of additional vulvo-vaginal-anal tissues
Hemostasis phase: meticulous hemostasis after the exenterative phase is completed
Reconstructive phase: there is an abdominal reconstructive phase and a vulvo-vaginal-anal reconstructive phase. They consist of the reconstruction of the urinary and intestinal tracts and restoring vaginal function. In our opinion, this is when 2 surgical teams may be working concurrently to shorten the operating time
Patients with no obvious preoperative contraindications are submitted to an examination under anesthesia, cystoscopy, proctoscopy, and surgical exploration prior to proceeding with an exenteration because almost half of the patients are found to have contraindications or unresectable disease during this evaluation.12 Both examinations can be performed by laparoscopy or robotics 48 to 72 hours prior to the planned exenteration to avoid an unnecessary laparotomy in case of adverse findings (see MIS exploration prior to exenteration in Minimally Invasive Surgery Applications section).
Examination under anesthesia, cystoscopy, and proctoscopy are usually performed prior to surgical exploration with particular attention to the caudal and lateral extension of the lesion. Should there be any obvious contraindications, the surgical exploration is abandoned. In the absence of an obvious contraindication, surgical exploration is warranted, as pelvic examination alone can be inaccurate for predicting surgical resectability. In one series, central lesions on pelvic examination were found to be unresectable after exploration in 35% of patients, while 28% of lateral pelvic wall lesions were able to be completely resected.17
There are 2 main objectives and 4 major areas of assessment during the surgical exploration. The 2 objectives are:
Evaluation of resectability: Is it possible to obtain a complete resection?
Evaluation of contraindications: Are there any contraindications?
The 4 major areas of assessment are:
Peritoneal metastases, direct peritoneal involvement, and involvement of attached bowel, adnexa, or other structures to the lesion
Aortic nodal metastases, including the infrarenal nodes
Fixation/extension of the recurrent lesion to the lateral pelvic wall
Pelvic node metastases
The order of performance varies according to the findings on preoperative evaluation and not all elements may be necessary if a contraindication is quickly noted (eg, peritoneal metastases). During the exploration phase the surgeon must avoid cutting any structures (ureters, bowel) to allow for an easy retreat in the case of adverse findings. In the absence of contraindications and after complete resectability has been assured, the procedure can be performed at the same surgical setting or 2 or 3 days later once the negative findings at laparoscopy have been discussed with the patient and her family.
A midline laparotomy incision is made. Intestinal and omental adhesions are sharply removed and efforts are made to preserve the omentum for ensuing pelvic floor covering. Pelvic peritoneal cytology is obtained. A thorough visual exploration of the peritoneal surfaces of abdomen and pelvis and of the omentum is carried out. Palpation is an added benefit of laparotomy. Aortic lymphadenectomy, including the infrarenal nodes, is then performed. Intraoperative frozen section is requested of the nodes and of any peritoneal lesion, suspicious or not. In the pelvis, most patients will have variable degrees of retroperitoneal fibrosis from previous irradiation and/or surgery, and palpation of the pelvic walls without retroperitoneal dissection will not provide valuable information as to resectability. The degree of fibrosis is directly related to the amount of time that has elapsed since completion of irradiation, and sharp dissection of the pelvic spaces will be necessary. The dissection is performed first on the side with the more lateral extension of the disease. The peritoneum over the psoas muscle is incised from the round ligament, which is transected, to above the pelvic brim. The external iliac vessels are exposed up to the common iliac artery bifurcation. Any enlarged nodes are removed for frozen section. The dissection is continued over the internal iliac artery until its anterior division. The superior vesical artery is followed caudally to the lateral aspect of the bladder. Sometimes, it may be easier to identify the superior vesicle artery by pulling it with a clamp or Russian forceps lateral to the bladder and tracing it in a retrograde fashion to its origin at the internal iliac artery. Mobilizing the superior vesical artery and the bladder medially while holding the external iliac vessels laterally, the paravesical space is developed. This dissection can be more or less tedious depending on the degree of fibrosis. Medial to the internal iliac artery and lateral to the ureter and in the direction of the coccyx, the pararectal space is developed. Significant effort must be made sometimes to open these spaces compared with a nonirradiated pelvis. Between both spaces one can palpate the parametrial web extending inferiorly to the levator muscle. Biopsies of the parametrium must be performed if there is a question of involvement. Pelvic lymphadenectomy is performed if there is any remaining nodal tissue.