Jonathan S. Berek
In order to surgically manage gynecological malignancies, it is frequently necessary to perform surgical procedures beyond the genital tract. These include selected operations on the intestinal and urologic tracts and plastic reconstructive operations, including the creation of a neovagina. In addition, central venous access is frequently required for hyperalimentation or chemotherapy.
Responsibilities for these procedures vary from one center to another. In some centers, most or all of the operations are performed by the gynecologic oncologist, while in other centers, the emphasis is on the development of a multidisciplinary team with involvement of colorectal surgeons, urologists, plastic surgeons, and anesthesiologists. The surgical techniques for these nongynecologic procedures are presented in this chapter.
Central venous-access catheters are often necessary in the critically ill gynecologic oncology patient for either central venous pressure monitoring or centrally administered hyperalimentation or chemotherapy (1,2,3,4). The most frequently used veins are the subclavian and the jugular. The brachial veins are used for peripherally inserted central catheters (PICC lines) (5).
Subclavian Venous Catheter
Infraclavicular Technique Although there are many different techniques for the insertion of a central venous catheter into the subclavian vein, the infraclavicular technique remains the one most commonly employed and the simplest. The subclavian vein lies immediately deep to the clavicle within the costoclavicular triangle, where the vein is more commonly approached from the right side (Fig. 20.1A). The costoclavicular-scalene triangle is bounded by the medial end of the clavicle anteriorly, the upper surface of the first rib posteriorly, and the anterior scalene muscle laterally (1). The anterior scalene muscle separates the subclavian vein anteriorly from the subclavian artery posteriorly. Just deep to the subclavian artery are the nerves of the brachial plexus. The subclavian vein is covered by the medial 5 cm of the clavicle. Just deep to the medial head of the clavicle, the right internal jugular vein joins the right subclavian vein to form the innominate vein, which then descends into the chest, where it joins the left innominate vein to form the superior vena cava in the retrosternal space.
Figure 20.1 Central venous catheter insertion sites. The right subclavian and right internal jugular vein insertion sites are illustrated. The insertion sites for the subclavian venous catheter: (A) via the infraclavicular technique; (B) via the supraclavicular technique; (C) the site of insertion for the internal jugular vein. The needle is directed toward the suprasternal notch.
There are several other vital structures in the scalene triangle. The phrenic nerve courses anterior to the anterior scalene muscle and therefore lies immediately deep to the subclavian vein. If the deep wall of the vein is penetrated, the phrenic nerve can be injured. If the subclavian artery is penetrated, the brachial plexus, lying just deep to the vessel, can be injured. The right lymphatic duct and the thoracic duct on the left enter their respective subclavian veins near the junction with the internal jugular veins and therefore may be injured by a misplaced needle. The most common injury is to the pleura, the apex of which is just beneath the subclavian vein at the junction of the internal jugular vein.
The technique for infraclavicular insertion of a catheter into the right subclavian vein is as follows:
If central venous pressure readings are to be determined, the intravenous line is attached to a manometer, and the base of the water column is positioned at the level of the right atrium, which is about 5 cm posterior to the fourth costochondral junction when the patient is in the supine position. The normal central venous pressure should be between 5 and 12 cm of water.
The complication rate for central venous catheter insertion through the subclavian route is about 1% to 2% (1,2,3,4). Most serious complications are related to puncture of the pleura and lung or perforation and laceration of vessels, resulting in a pneumothorax or hemothorax. Catheter-related infection is seen in about 0.5% of patients, and the catheter should be removed if this source of infection is suspected.
Supraclavicular Insertion An alternative route of insertion into the subclavian vein is the supraclavicular route (Fig. 20.1B). Some prefer this to the infraclavicular route, but the morbidity of insertion is comparable with the two methods, and the preference is related to the technique that is most comfortable for the operator.
The technique for insertion is identical to that of the infraclavicular route, except that the needle is inserted above the clavicle, approximately 5 cm lateral to the midsternal notch. The angle of insertion is about 30 degrees from a line drawn between the two shoulders and directed caudally. The needle is aimed at the suprasternal notch.
Jugular Venous Catheterization
Another alternative for central venous access is the use of the jugular veins, either the internal or external vein. Jugular venous catheterization is frequently the method of choice when the catheter is inserted intraoperatively and the catheter is to be used primarily for acute monitoring. The advantage is that there is relatively easy access while the patient is anesthetized and draped for surgery, whereas the disadvantage is that it is more difficult to anchor the catheter because the neck is more mobile than the anterior chest wall. The location for the insertion site is illustrated in Fig. 20.1C.
The technique for insertion is as follows:
While individual studies have reported varying results, a meta-analysis of ultrasonic guidance for central venous cannulation reported a lower technical failure rate, reduction in the complication rate and faster access when using two dimensional ultrasonic guidance (5). The benefit was more significant for internal jugular versus subclavian venous catheterization.
External jugular catheters may also be used in patients who are under general anesthesia. Some patients have relatively prominent external jugular veins, and they are very easily catheterized. The external jugular is not durable, however, and this route is not useful for central hyperalimentation. The complication rate for jugular venous catheterization is essentially the same as that for the subclavian route.
The placement of semipermanent lines is useful in patients who require prolonged access to the central venous system, such as those with a chronic intestinal obstruction or fistula who are to receive hyperalimentation after discharge from the hospital (2).
Broviac, Hickman, and Quinton Catheters
The most common types of lines are catheters made of flexible, synthetic rubber (e.g., Broviac, Hickman, or Quinton catheters). The catheters are available in several sizes, although the adult type is used for most patients; the length is adapted by cutting the catheters as necessary. The catheters are available with either a single or double lumen. The single-lumen catheters usually are sufficient for parenteral nutrition, whereas the double-lumen ones may be necessary for patients requiring frequent bolus medication, such as intravenous pain or antibiotic medications (2,3).
The most common site for insertion of a semipermanent catheter is the right subclavian vein. The method of insertion is initially identical to the technique employed for the insertion of a temporary catheter, but an insertion cannula, called a Cook introducer, can simplify and facilitate insertion of the catheter (Fig. 20.2). It is preferable to insert the catheter under fluoroscopic guidance.
The technique is as follows:
Figure 20.2 Semipermanent catheter insertion. The technique for insertion of the semipermanent (e.g., Hickman) catheter. A: A needle is inserted into the right subclavian vein, a guide wire is inserted through the needle, and the needle is withdrawn. B: The Cook introducer then is inserted over the guide wire. C:After the introducer with its outer sheath is in place in the right subclavian vein, the wire is withdrawn.
Figure 20.2 (Continued) D: The central catheter of the Cook introducer is withdrawn, and the free end of the semipermanent catheter is inserted through the outer sheath. E: The outer sheath of the Cook introducer is peeled away. F: The semipermanent catheter is tunneled in the subcutaneous tissue under the skin of the right side of the chest, and the free end is exteriorized.
Peripherally Inserted Central Catheter Lines
Another type of semipermanent line is inserted through a peripheral access site and is called the peripherally inserted central catheter or PICC line (6). The PICC line is inserted into the brachial vein in the antecubital fossa. The catheter is passed cephalad until it reaches the central subclavian vein. This line is suitable for the infusion of parenteral nutrition as well as chemotherapy. The PICC line may be more desirable than a totally implantable line when short-term use is contemplated, e.g., 3 to 4 months.
This approach is less durable than the centrally inserted catheters and somewhat more cumbersome because of the location of the insertion site. However, its main advantage is that it can be easily inserted at the bedside. Furthermore, it can be placed by a certified nurse or an intravenous technician trained in the insertion technique. Alternately, an implantable port can be inserted in the antecubital fossa by a physician.
Peritoneal catheters are used in gynecologic oncology for the instillation of intraperitoneal chemotherapy. A commonly used catheter is the Tenckhoff peritoneal dialysis catheter.This dialysis catheter is designed to minimize the risk of infection, even though it is left in place many months (7). Alternatively, a Hickman venous access catheter can be used.
The catheter is implanted into the peritoneal cavity lateral to the midline laparotomy incision (Fig. 20.3). The catheter is tunneled in the subcutaneous tissue and brought out through a stab incision lateral to the fascial incision. The tip of the catheter in the peritoneal cavity is directed toward the pelvic cul-de-sac.
An alternative approach is the use of a completely implantable port which is attached directly to a fenestrated peritoneal catheter or venous access catheter. The port is inserted into the subcutaneous tissue and positioned in the left or right lower quadrant of the anterior abdomen, or over the lower anterior rib cage for ready access (Fig. 20.4). Both laparotomy and laparoscopy can be utilized for port placement. Postprocedure, the port is entered percutaneously with a 21-gauge needle.
The reported rate of catheter-related complications ranges from 3% to 34% (8,9,10,11,12,13). In the most recent phase III study of IP versus IV chemotherapy for the treatment of ovarian cancer, Walker et al. reported that 40 (19.5%) of 205 patients randomized to the IP arm had complications including infection (n = 21); blockage (n = 10); leakage (n = 3); access problems (n = 5); and vaginal leakage of fluid (n = 1) (8,14). Of the 119 patients discontinuing IP therapy, catheter-related complications accounted for 34% of cases (40 of 119). Both Tenckhoff or implantable ports with attached fenestrated or venous (Hickman) catheters were used.
Minor infections can be treated with antibiotics, and low-grade peritonitis can be treated by the instillation of antibiotics directly via the catheter. For persistent and severe infections, the peritoneal catheter may require removal. In other studies, the most common problem associated with the catheters was blockage (10,11), as there is no effective way to prevent some deposition of fibrin around the catheter. Occasionally, this produces a “ball valve” effect; i.e., fluid will flow in but will not flow out. While some authors report higher fibrin sheath formation and adhesions in association with fenestrated catheters and dacron cuffs (15), a recent retrospective study of fenestrated peritoneal catheters reported that only 9 of 342 patients (3%) required discontinuation secondary to catheter complications (9).
Figure 20.3 Tenckhoff peritoneal catheter. The placement of the Tenckhoff catheter into the peritoneal cavity is illustrated.
Figure 20.4 Port-a-Cath peritoneal catheter. The totally implantable peritoneal access catheter is tunneled through the subcutaneous tissues into the peritoneal cavity.
Particularly important in the operative plan for any patient is the determination of the type of incision to be made. The surgeon should have a general philosophy and modus operandi when planning the surgical procedure. There are certain incisions that are more appropriate in patients who are undergoing surgery for cancer rather than for benign conditions. In addition, special guidelines for the closure of incisions should be followed.
Abdominal incisions used in the gynecologic oncology patient are most commonly vertical. Transverse incisions are also appropriate in certain circumstances. The indications and techniques for these incisions and their modifications are discussed.
Patients with suspected malignancies of the ovary or fallopian tube are best explored through a vertical abdominal incision. With a vertical incision, the patient's disease can be staged properly. Also, this approach permits the removal of any upper abdominal metastases, which cannot always be appreciated preoperatively. The most likely site of resectable upper abdominal disease is the omentum. For an omentectomy, access to the region of the splenic and hepatic flexures is required.
A vertical incision is also necessary in patients being explored for intestinal obstruction or fistulae. The performance of a paraaortic lymphadenectomy is facilitated by a vertical incision. Patients being explored for recurrent malignancies or for possible pelvic exenteration also require a vertical abdominal incision.
The most commonly used vertical incision is in the midline. This incision has the advantage of being easy to perform; it can be accomplished quickly, because the midline is the least vascular area of the abdominal wall, and the smallest depth of tissue must be divided. The principal blood supply to the anterior abdominal wall is from the inferior epigastric vessels, which are located laterally in the rectus sheath posterior to the rectus abdominis muscles, and these vessels are avoided by the mid-line incision.
The principal problem associated with the midline incision is that it has the highest rate of wound dehiscence when compared with all other incisions. The wound disruption rate is about 0.1% to 0.65% (16,17,18,19,20,21), although this rate may be higher in patients with cancer, particularly those with ascites and malnutrition, or those needing postoperative radiation. Dehiscence rates as high as 2% to 3% have been reported in obese, diabetic patients with cancer (18). The majority of wound dehiscences are associated with wound infection or poor closure technique. The occurrence of ventral hernia is associated with wound disruption secondary to infection and is more common in patients with malignancy(21). The use of prosthetic meshes for herniorraphy can significantly reduce the rate of recurrence of hernia.
In patients with a probable benign condition who are undergoing abdominal exploration for the first time, a lower transverse abdominal incision is frequently employed. The advantage of this incision is that it is more cosmetic, is generally less painful, and is associated with fewer incisional hernias. The disadvantage is the relative problem of upper abdominal exposure and the more frequent occurrence of wound hematomas.
If exposure to the upper abdomen is required, the surgeon has several choices. The incision can be modified by division of the rectus abdominis muscles in a transverse direction at the level of the incision (i.e., a Maylard incision), or the rectus abdominis muscles may be detached from the symphysis pubis (i.e., the Cherney incision). After division or mobilization of the rectus muscles, the inferior epigastric vessels are ligated bilaterally and, if necessary, the incision is further extended laterally by incising (with the diathermy) the “strap” muscles of the anterior abdominal wall. The conversion of the incision to a Maylard or a Cherney incision always provides considerably more exposure in the pelvis and low paraaortic area.
If better access to the upper abdomen is required, the incision can be modified further by extending the incision cephalad to form a “J,” a reverse “J,” or a “hockey stick” incision.In general, any of these techniques is preferable to the making of a second incision, i.e., a midline incision coincident with the transverse incision, a so-called “T” incision. The principal difficulty with the latter approach is the weakness of the incision at the point of intersection of the two incisions.
In patients undergoing radical hysterectomy and pelvic lymphadenectomy for early-stage cervical cancer, a lower abdominal transverse incision is acceptable.
Of primary importance is the technique of incisional closure (16,17,18,19,20,21). The closure can be accomplished by closing the peritoneum, fascia, subcutaneous tissue, and skin individually, or a bulk closure can be performed that incorporates the peritoneum and the fascia together. This bulk closure or internal retention suture, the “Smead-Jones” closure, is the strongest closure technique (16). Mass closure with a continuous, single strand of polyglyconate monofilament absorbable suture (Maxon) or polydioxanone (PDS) has been shown to be an effective, safe alternative to the use of interrupted sutures, even in vertical midline incisions (22,23,24,25,26,27).
Internal Retention Suture
The Smead-Jones, or internal retention, technique uses interrupted sutures that are placed as illustrated in Fig. 20.5. The sutures are placed in a far-far, near-near distribution, which is a modified figure-of-eight. The first suture is placed through the anterior fascia, rectus muscle, posterior fascia, and peritoneum and the second through the anterior fascial layer only. The key is to place the sutures at least 1.5 to 2.0 cm from the fascial edge and not more than 1 cm apart (16). The disruption rate for midline incisions closed with this technique should be less than 0.2%.
Figure 20.5 Internal retention abdominal closure. The “Smead-Jones” far-far, near-near closure.
The choice of suture should be dictated by the circumstances (13,14,15,16,17,18). If there is evidence of significant infection, as with an abscess or an intestinal injury, a monofilament, nonabsorbable suture is most appropriate. The most frequently used substances are nylon sutures, such as Prolene.
For vertical incisions, an absorbable, long-lasting synthetic suture offers the best combination of strength, durability, and ease of use. Most suitable is either monofilament polyglyconate suture (Maxon) or monofilament polydioxanone (PDS) (24). Braided, polyglycolic acid (PGA) suture, such as Vicryl or Dexon, is suitable for transverse incisions. A grade 0 or 1 suture is necessary to provide a suitably strong closure. The tissue reactivity to these synthetic materials is less than that of chromic catgut. Nonabsorbable polyfilament materials, such as cotton and silk, are not used for incisional closure because of the higher potential for “stitch abscess” formation (27).
External Retention Suture
Retention sutures that are external can be used to prevent evisceration in patients who are at high risk of this potentially catastrophic occurrence. The routine use of internal retention sutures has reduced the need for the external retention sutures. However, in patients who are morbidly obese, patients who have a major wound infection, and patients whose incisions have eviscerated in the past, the addition of external retention sutures may be indicated. These sutures are placed in a manner similar to internal retention sutures, i.e., far-far, near-near, with the far sutures also placed through the skin so that the retention sutures are knotted externally. The preferable suture material for this closure is nylon. The external retention sutures are inserted through a rubber “bolster” that helps to protect the skin from injury from the suture. Sutures are placed at approximately 2 to 3 cm intervals, and interrupted fascial sutures are placed between them.
Primary Closure Skin closure of vertical incisions in cancer patients generally should be interrupted, generally with metal skin clips. Subcuticular closures are not appropriate in most circumstances for vertical incisions, but they are quite cosmetic and acceptable for small transverse incisions where the risk of wound infection is low.
Secondary Closure A delayed or secondary skin closure is useful in patients whose incisions are infected, that is, after the drainage of an intraabdominal abscess or repair of an intestinal fistula. This is achieved by placement of interrupted mattress sutures in the skin, which are not tied, so that the skin remains unapproximated. Thus the skin can be closed later, usually after 3 to 4 days, when the infection is under control.
Preoperative Intestinal Preparation
If bowel resection is planned or contemplated, a mechanical and antibiotic “bowel preparation” may be undertaken preoperatively. If the intestine is prepared properly, the segment is well vascularized, and there is no sepsis, prior irradiation, or evidence of tumor at the site of anastomosis, colonic reanastomosis can be accomplished without leakage in 98% of the cases (25). More proximal resection of the small intestine can be performed without a bowel preparation, because this portion of the intestine does not contain bacteria.
An effective protocol for bowel preparation is presented in Table 20.1. Recently, the FDA issued a Safety Alert discussing the risk of acute phosphate nephropathy, a rare type of renal failure associated with oral sodium phosphate bowel cleansing which may lead to permanent renal impairment. Increased risk has been associated with multiple factors including advanced age, kidney disease, and use of medications affecting renal perfusion and function. In response, over-the-counter oral sodium phosphate preparations such asFleet® Phospho-soda have been recalled. Bowel preparation regimens should be individualized for patients after assessment of risk factors. Alternative bowel preparation regimens include magnesium citrate and Golytely.
Table 20.1 Intestinal Preparation
There is controversy regarding the use of bowel preparations in general as well as the optimal “bowel prep,” as it is uncertain whether, in addition to the mechanical preparation, the antibiotic preparation is necessary (28). Several meta-analyses of patients undergoing elective colorectal procedures have found either no difference or increased rates of anastomotic leakage and wound infection when prophylactic mechanical bowel preparation is used (29,30,31). However, bowel resections for gynecologic malignancies are performed in a different patient population, often in the presence of ascites, extensive carcinomatosis, and multiple sites of bowel involvement. Lacking studies in the gynecologic oncology population, the use of mechanical bowel preparation and antibiotics are predominantly determined by surgeon preference. Before laparotomy for small intestinal obstruction caused by ovarian cancer, it is useful to insert a nasogastric (NG) tube for 24 to 48 hours preoperatively to avoid the possibility of vomiting and aspiration (32).
Minor Intestinal Operations
The most common intestinal operations are lysis of adhesions, repair of an enterotomy, and creation of an intestinal stoma.
Repair of Enterotomy
Intestinal enterotomy is a common inadvertent occurrence in abdominal surgery, and it can occur in the most experienced hands. Factors that predispose to serosal and mucosal injury include extensive adhesions, intraabdominal carcinomatosis, radiation therapy, chemotherapy, prior abdominal surgery, and peritonitis.
An enterotomy usually does not cause any problems, provided it is identified and repaired. Any defect should be repaired when it occurs or marked with a long stitch so that it will not be overlooked later. At the completion of any intraabdominal exploration necessitating significant lysis of adhesions, the surgeon must “run the bowel,” carefully inspecting it to exclude either a serosal injury or an enterotomy.
Serosal defects through which the intestinal mucosa can be seen must be repaired. Less complete defects must be repaired in all patients who have had radiation treatment to the abdomen. When in doubt, the defect should be repaired to minimize the risk of intestinal breakdown, peritonitis, abscess, and fistula.
When there is an enterotomy, the repair should be made with interrupted 3-0 or 4-0 sutures on a gastrointestinal needle, placed at 2- to 3-mm intervals along the defect. The suture materials most commonly employed for this purpose are silk or PGA (Vicryl or Dexon). The direction of closure should be perpendicular to the lumen of the bowel to minimize the potential for lumenal stricture (Fig. 20.6).
With small defects (i.e., <5-6 mm), the closure can be accomplished with a single layer of sutures passed through both the serosa and the mucosa. However, it is preferable to close more extensive defects in two layers: an inner full-thickness layer covered with an outer seromuscular layer. Care should be taken to approximate the tissues carefully without cutting through the fragile serosa.
Figure 20.6 Closure of an intestinal enterotomy. A: The edges of the enterotomy are trimmed. B: The enterotomy is closed perpendicular to the lumen in two layers.
A gastrostomy may be necessary in patients with chronic intestinal obstruction, usually from terminal ovarian cancer. It is particularly useful in those who require prolonged intestinal intubation and in whom the underlying intestinal blockage cannot be relieved adequately. This procedure may permit the removal of an uncomfortable nasogastric tube that is irritating to the nasopharynx. The two most common procedures are the Witzel and the Stamm gastrostomies (32).
Stamm Gastrostomy The simplest technique is the Stamm gastrostomy, in which a small incision is made in the inferior anterior gastric wall. A Foley catheter with a 30-mL balloon is brought into the peritoneal cavity through a separate stab incision in the left upper outer quadrant of the abdomen. Two or three successive pursestring sutures, with 2-0 absorbable suture material, are used to invert the stomach around the tube. Interrupted 2-0 silk or PGA sutures are placed in the serosa, and the same material is used to suture the serosa to the peritoneum, approximating the gastric wall to the anterior abdominal wall in an effort to prevent leakage.
Witzel Gastrostomy The Witzel technique is similar, but the catheter is tunneled within the gastric wall for several centimeters with Lembert sutures of 2-0 silk or PGA. This technique results in a serosal tunnel that may further reduce the risk of leakage. The most important step in preventing gastrostomy leakage is approximation of the gastric serosa to the anterior abdominal wall.
Percutaneous Gastrostomy Another technique for gastrostomy in patients not otherwise undergoing laparotomy is the percutaneous placement of a catheter into the stomach. This method involves the initial passage of a gastroscope. The site for catheter insertion is illuminated by a fiberoptic light source through the gastroscope, and the catheter is introduced into the stomach percutaneously.
The performance of a cecostomy may be useful in the occasional patient who has an obstruction of the colon and a grossly dilated cecum and in whom a simple palliative measure to relieve the obstruction is indicated. A more definitive procedure for relief of the obstruction may be appropriate when the patient's condition is more stable.
The cecostomy is performed by placement of a Foley catheter into the dilated portion of the cecum. The tube is sutured into place by the technique employed for a Stamm gastrostomy. The tube is exteriorized through a stab incision in the right lower quadrant of the abdomen and attached to gravity drainage.
Colostomies may be temporary or permanent. A temporary colostomy may be indicated for “protection” of a colonic reanastomosis in patients who have had prior radiation therapy or to palliate severe radiation proctitis and bleeding. It is indicated also in patients who have a large bowel fistula (e.g., rectovaginal fistula) to allow the inflammation to subside before definitive repair. A permanent colostomy is indicated in patients who have an irreparable fistula or a colonic obstruction from a pelvic tumor that cannot be resected. A permanent colostomy is also indicated in patients undergoing total pelvic exenteration, unless the distal rectum can be preserved and the colon reanastomosed, and in those who require anoproctectomy because of advanced vulvar cancer.
The site of the colostomy should be selected so that the stomal appliance and bag can be applied to the skin of the anterior abdominal wall without difficulty. The best site is approximately midway between the umbilicus and the anterior iliac crest. The most distal site possible should be employed in the large intestine. After selection of the stomal site, a circular skin incision is made to accommodate two fingers. The subcutaneous tissue is removed, and the fascia of the rectus sheath is incised similarly (Fig. 20.7). The end of the colon is brought through the stoma and sutured to fascia with interrupted 2-0 silk or PGA suture, and the stoma is everted to the skin to form a “rosebud” with the use of interrupted 2-0 or 3-0 absorbable braided suture.
For patients who require temporary diversion, a transverse or sigmoid colostomy is usually created. The most distal portion of the colon should be used to allow the most formed stool possible. A loop colostomy is usually created: A loop of the colon is brought out through an appropriately placed separate incision in the abdominal wall. The loop is maintained by suturing it to the fascia beneath it. It can be reinforced with a rod of glass or plastic passed through a hole in the mesentery. The stoma can be opened immediately by means of an incision along the taenia coli in the longitudinal direction. Alternatively, the loop may be “matured” 1 to 2 days later to minimize the risk of sepsis if the bowel is unprepared.
The colon can be brought out as an end colostomy, which requires transection of the colon. This can be readily accomplished by means of a gastrointestinal anastomosis (GIA) stapler, which closes and transects the colon simultaneously. The distal end is sutured to the fascia, and the proximal end is brought out as the colostomy. If the distal colon must also be diverted (because of distal obstruction), a double-barrel colostomy can be created.
Figure 20.7 The formation of a colostomy. A: The end of the colon is brought through the abdominal wall. B: It is sutured to the fascia and skin. C: The “rosebud” stoma is formed.
A permanent colostomy is an end or terminal colostomy, performed as far distally as possible to allow the maximum amount of fluid reabsorption. The distal loop of the transected colon may be oversewn to create a Hartman's pouch if there is no distal obstruction. In patients in whom there is complete distal obstruction, a mucous fistula should be created.
If the colon is surgically inaccessible because of extensive carcinomatosis or radiation-induced adhesions, it may become necessary to palliate the bowel obstruction by the creation of a small intestinal stoma. Because the small-bowel contents are loose and irritating compared with colonic contents, an ileostomy or a jejunostomy should be undertaken only when absolutely necessary.
Major Intestinal Operations
Intestinal Resection and Reanastomosis
After a segment of bowel, along with its wedge-shaped section of mesentery, has been resected, a reanastomosis may be performed (32,33,34,35,36,37). The most commonly used technique for reanastomosis is the end-to-end anastomosis, which is performed as either an open two-layered closure or a closed one-layered anastomosis. An end-to-side anastomosis may be used to create a J-pouch, i.e., a segment of bowel created to improve low colonic continence (38,39,40,41,42,43,44,45,46,47,48,49). A side-to-side anastomosismay be useful to increase the size of the lumen at the site of anastomosis. Increasingly, the use of surgical stapling devices has permitted more rapid performance of the reanastomosis, which is particularly useful when more than one resection is being carried out or when the duration of the procedure is of major concern.
When the reanastomosis is to be hand sewn, the proximal and distal ends are clamped with Bainbridge clamps (Fig. 20.8A), and the posterior interrupted, seromuscular Lembert stitches are placed with 3-0 silk or PGA sutures (Vicryl or Dexon) (Fig. 20.8B). The clamps are removed, the devitalized ends are trimmed, and an inner continuous full-thickness layer of 3-0 silk or PGA suture is placed to complete the posterior portion of the anastomosis. After the corner is reached, the needle is brought through the wall to the outside, and the continuous layer is completed anteriorly with a Connell stitch (outside-in, inside-out) to complete the inner layer (Fig. 20.8C). The anterior seromuscular layer is then placed with interrupted 3-0 silk or PGA sutures (Fig. 20.8D). The defect in the intestinal mesentery is repaired. A single-layered closed technique is occasionally used for colonic reanastomosis in obstructed, unprepared bowel in an effort to minimize peritoneal contamination. In these circumstances, however, the use of the surgical staplers is now recommended (36).
The side-to-side anastomosis is particularly useful in patients who are undergoing intestinal bypass rather than resection to palliate bowel obstruction, e.g., in patients with unresectable or recurrent tumor. The loops of intestine are aligned side to side, and linen-shod clamps are applied to prevent spillage of intestinal contents. A posterior row of 3-0 silk or PGA sutures is placed with interrupted Lembert sutures, and the lumina are created. An inner layer of continuous, full-thickness 3-0 PGA sutures is placed and continued anteriorly to complete the layer with a Connell stitch. The anastomosis is completed by placement of an anterior seromuscular layer with the use of interrupted 3-0 silk or PGA sutures.
The principal advantage of the gastrointestinal staplers is the speed with which they can be employed. There is no increase in the complication rate with the use of staplers as compared with hand-sewn anastomoses (32,33,34,35,36,37). The staplers are especially useful in facilitating reanastomosis after low resection of the rectosigmoid colon, because a hand-sewn anastomosis is technically difficult when performed deep in the pelvis. A disadvantage of the staplers is their increased cost, and staplers are difficult to use when the intestinal tissues are very edematous.
Types of Stapling Devices
The staplers are available in either reusable metal devices or in single-use disposable devices (Fig. 20.9A-C).
Thoracoabdominal Stapler The thoracoabdominal (TA) stapler comes in several sizes, the TA-30, TA-55, TA-60, and TA-90, corresponding to the length, in millimeters, of the row of staples. Individual staples are either 3.5 or 4.8 mm long. The TA closes the lumen in an everting fashion. A TA device is available with a flexible, rotating end, called a Roticulator-55 that can be adjusted for placement into narrow areas (e.g., the deep pelvis).
Figure 20.8 Hand-sewn end-to-end enteroenterostomy. A: The tumor and bowel are resected along with the mesentery. B: The posterior seromuscular layer is sutured. C: The Connell stitch is placed. D: The anterior seromuscular layer is placed. E: The completed anastomosis.
Gastrointestinal Anastomosis Stapler The gastrointestinal anastomosis (GIA) device places two double rows of staples and then cuts the tissue between the two rows.
End-to-End Anastomosis Stapler The end-to-end anastomosis (EEA) stapler is used primarily to approximate two ends of the colon, especially to facilitate the reanastomosis of the lower colon after pelvic exenteration or resection of pelvic disease in patients with ovarian cancer. The stapler places a double row of staples, approximates the two ends of the intestine, and cuts the devitalized tissue inside the staple line. It is available in diameters of 21, 25, 28, 31, and 35 mm, and a metal sizing device is used to measure the diameter of the intestinal lumen (37).
Intraluminal Stapler The intraluminal stapler (ILS) is a disposable EEA stapler that has a detachable anvil. This removable feature can facilitate the placement of the anvil into a portion of one intestine that is difficult to mobilize. The anvil can be reattached to the rod of the ILS device after it has been placed in the anastomosis.
Figure 20.9 Stapling devices for intestinal anastomosis. The single-use staplers: (A) thoracoabdominal (TA); (B) gastrointestinal anastomosis (GIA); (C) end-to-end anastomosis (EEA).
Functional End-to-End Enteroenterostomy Anastomosis
This operation is illustrated in Fig. 20.10. The GIA stapler is used to staple and divide each end of the bowel segment to be resected. The antimesenteric borders of the bowel loops are approximated, and the corners are resected. A fork of the GIA device is inserted into each bowel lumen, and after alignment, the stapler is fired. The defect where the stapler was introduced then is closed with a TA stapler.
Figure 20.10 Functional end-to-end anastomosis using the stapling technique. A: The gastrointestinal anastomosis (GIA) stapler is used to resect the intestine. B: The segments of the transected intestine are placed side to side, and each antimesenteric corner is incised to create two holes into which the two forks of a second GIA stapler are placed. The GIA stapler is fired to create the new intestinal lumen. C: The thoracoabdominal (TA) stapler is placed over the end and “fired” to close the remaining defect. Note the cross section at a-a'.
Side-to-Side Enteroenterostomy Anastomosis
When a bypass enteroenterostomy is performed, the two loops of bowel to be anastomosed side to side are aligned, an enterotomy is created in each loop, and a fork of the GIA stapler is slid into each lumen, fired, and removed. This creates the lumen between the two bowel segments, and the enterotomy that is left when the instrument is withdrawn is then approximated with a TA stapler.
Figure 20.11 Low colonic end-to-end anastomosis using the end-to-end anastomosis (EEA) stapler. A: After resection of the rectosigmoid colon, the distal end of the descending colon is mobilized and a pursestring suture is placed by hand or with a special instrument (illustrated). A pursestring suture is also placed around the rectal stump. The open end of the EEA stapler is inserted through the anus, and the rectal pursestring is tied around the instrument. The end of the descending colon is placed over the end of the EEA, and the second pursestring is tied. B: The EEA device is closed and “fired.”
Low Colonic End-to-End Anastomosis
A low colonic resection is performed by isolating and removing the portion of the rectosigmoid colon involved with disease. The EEA stapler is inserted through the anus and advanced to the site of the anastomosis. The instrument is opened to allow the anvil to accommodate the proximal colon, which is mobilized and tied over the distal end of the EEA. The distal colon is likewise tied over the EEA with a pursestring suture (Fig. 20.11). The EEA is then closed, approximating the two ends of the colon, and the instrument is fired and removed. A reinforcing layer of interrupted 3-0 silk or Vicryl Lembert sutures is placed anteriorly. The anastomosis is palpated to confirm that it is intact. Also, the pelvis can be filled with saline solution, and air can be insufflated through the rectum to search for bubbles, which would indicate a defect in the anastomosis (32).
Low Colonic End-to-Side Anastomosis: J-Pouch
An alternative end-to-side (functional end-to-end) low colonic anastomosis can be performed with the use of one of the newer disposable stapling devices, which has a removable distal onepiece anvil: the intraluminal stapler. In this manner, a J-pouch can be created, which has the potential to improve the continence of patients (Fig. 20.12A-D). Studies comparing the colonic J-pouch with the direct end-to-end anastomosis have suggested that there is a lower leak rate, better continence rate, fewer stools per day, and better control of urgency and flatus (34,48). The problem with this approach is that some patients have more difficulty emptying the pouch, a problem that can be minimized by limiting the size of the pouch to about 5 centimeters in length (40).
Figure 20.12 Low colonic end-to-side anastomosis (EEA) to create a J-pouch. A: The end-to-side anastomosis allows the mesocolon to be preserved and to cover the sacral hollow. B: The terminal end of the colon has the J-pouch created by stapling a loop side-to-side using a gastrointestinal anastomosis (GIA) stapler. C: The EEA device is then used to anastomose the rectal stump end-to-side with the pouch. D: The end of the pouch is stapled closed with a GIA or thoracoabdominal (TA) stapler.
The J-pouch is created by first folding the distal colon onto itself and stapling side to side with a GIA stapler (Fig. 20.12A). The pouch is then anastomosed to the rectal stump using an end-to-side technique with an EEA stapler (Fig. 20.12B) and by detaching the anvil, which is inserted in the proximal colon segment. The center rod of the open EEA instrument without the anvil is inserted through an opening in the bowel or through the anus (Fig. 20.12C). Then the rod is inserted through or near the staple line. In the other segment of bowel, a pursestring suture is placed, and the free anvil is inserted within the lumen of the bowel within the pursestring suture. The anvil is then screwed onto the rod, the device is closed, and the anastomosis is created (Fig. 20.12D).
Low Colonic Side-to-Side Anastomosis
An alternative side-to-side technique (functional end-to-end) anastomosis of the rectosigmoid colon can be used when the portion of removed bowel is proximal enough to permit this operation (i.e., 10 to 15 cm of preserved rectum). The GIA instrument is used to perform the colorectal anastomosis. After the segment of colon to be resected is mobilized, the proximal colon to be reanastomosed is closed with either the GIA or the TA-55 instrument. A stab wound is made in the antimesenteric border of the colon about 5 cm proximal to the staple line closure. A corresponding stab wound is made in the left anterolateral wall of the rectum at the proximal point of the planned site of anastomosis. The proximal colon is placed into the retrorectal space, side to side along the rectum; the GIA device is placed into the proximal and distal segments; and the instrument is closed and fired. The remaining single defect is closed with either a handsewn, double layer of 3-0 sutures or the rotating TA-55 device (Roticulator-55).
Low Colonic Coloplasty
A newer form of colorectal anastomosis is the coloplasty (see Chapter 22), which may have advantages over the colonic J-pouch for patients undergoing pelvic exenteration. The principal advantage is that this technique may be easier to perform in a narrow pelvis, such as in those patients undergoing a pelvic exenteration and simultaneous creation of a neovagina. In a randomized study of the three techniques, patients undergoing the coloplasty and colonic J-pouch had significantly more favorable compliance, reservoir volume, and fewer bowel movements per day than those having a straight anastomosis (49).
The technique of coloplasty (Fig. 20.13) is illustrated. The lower part of the proximal colon is incised longitudinally and then reapproximated in the transverse direction (Fig. 20.13A). In this manner, a small, simple reservoir is created. The EEA stapler is used and then closed to create the anastomosis just as with the other techniques outlined above (Fig. 20.13B).
Historically, after resection of the small bowel, a nasogastric tube was usually placed for about 24 to 48 hours to reduce the volume of intestinal secretions that must pass through the site of anastomosis. However, numerous studies including a recent meta-analysis of 33 studies with 5,240 patients undergoing open abdominal procedures (including 14 studies of gastroduodenal or colorectal procedures) have demonstrated that prophylactic nasogastric decompression is associated with a slower return to bowel function and increased pulmonary complications, with no difference in anastomotic leaks (50,51). While prophylactic nasogastric decompression is often not indicated, it remains integral to the management of small bowel obstruction. Postoperative use is individualized based on preoperative diagnosis and intra-operative assessment. In patients who have received pelvic or abdominal irradiation, the upper intestinal tract may remain intubated until bowel function has returned, as signified by the passage of flatus or stool. Oral feeding can begin as patients develop an appetite, and early feeding has been shown to be safe following both small and large bowel resection (52,53,54). In patients who have undergone colonic resection and reanastomosis, enemas and cathartics should be avoided (49).
Intravenous fluids must be continued while the patient is receiving nil by mouth. In patients whose recovery is likely to be prolonged beyond 7 days, such as those who have previously received whole-abdominal irradiation, consideration should be given to the use of parenteral nutrition, as discussed in Chapter 19. In such patients, a gastrostomy tube may be useful to avoid prolonged nasogastric intubation.
Figure 20.13 Low colonic coloplasty. A: The distal colon is incised in the longitudinal direction and resutured in the transverse direction to create a widening of the portion of the colon to be anastomosed to the rectum. B: The rectum is anastomosed to the distal colon using the end-to-end anastomosis (EEA) stapler.
Urinary Tract Operations
The preoperative evaluation of the urinary tract is important in patients with gynecologic malignancies because of the frequent involvement of the urinary organs, especially the bladder and the distal ureters (55,56,57,58). Renal function and ureteric patency must be assessed preoperatively.
Cystoscopy should be performed as part of the staging for cervical and vaginal cancers unless the disease has been diagnosed early (55). Cystoscopy is also indicated in patients with a lower urinary tract fistula or unexplained hematuria. Cystoscopic examination may demonstrate external compression of the bladder by a tumor, bullous edema produced by the blockage of lymphatic vessels from adjacent tumor growth, or mucosal involvement with tumor. When a mucosal lesion is seen, a biopsy can confirm the diagnosis.
Technique Cystoscopy is performed with the patient in the dorsal lithotomy position. After preparation and draping of the area, the cystoscopic obturator and sheath are inserted into the urethra and carefully advanced into the bladder, after which the obturator is removed. The cystoscope is inserted into the sheath. About 250 to 400 mL of normal saline solution is instilled into the bladder to permit a thorough inspection of the entire mucosa.
A suprapubic cystostomy catheter is useful in patients who require prolonged bladder drainage. This catheter may be useful in patients undergoing radical hysterectomy for cervical cancer or extensive resection of pelvic tumor because of the temporary disruption of bladder innervation that occurs with these dissections. The suprapubic catheter can be easier for the patient to manage than a transurethral Foley catheter, and the rate of bladder infection is lower (55). The other convenient aspect of this catheter is that it can facilitate trials of voiding. The patient can clamp the catheter for a specified interval, void, and then unclamp to check for residual urine. When the residual urine is less than 75 to 100 mL, the catheter can be removed. However, many patients are also managed successfully after radical hysterectomy with a period of continuous trnsurethral catheterization followed by intermittent self-catheterization (59,60).
Technique The catheter used is an 18F Silastic Foley catheter with a 5- to 10-mL balloon. This catheter is well tolerated by patients, produces minimal local tissue irritation, and is of sufficient caliber that blockage of the catheter lumen is not a major problem.
The placement of a suprapubic catheter involves the following steps:
Ureteral obstruction is the most common urinary complication in patients with gynecologic malignancies. This problem is seen particularly in patients with cervical or vaginal cancer, either at the time of diagnosis or with recurrent disease. It may result from direct tumor extension into the bladder or distal ureters or from compression by lymph node metastases. In patients with intraabdominal carcinomatosis, most often from ovarian cancer, extensive pelvic tumor may cause significant progressive ureteral obstruction. The most frequent site of lower urinary tract obstruction in gynecologic patients is the ureterovesical junction (56).
Postoperative obstruction is usually incomplete and results from edema, possible infection, and partial devascularization of the distal ureter. However, the obstruction may be complete, and when it is, it most often results from inadvertent suture ligature of the distal ureter when the surgeon is attempting to ligate the blood vessels of the cardinal ligament (55). Chronic obstruction can result from stenosis after pelvic irradiation, particularly if pelvic surgery is also performed.
In patients who have a partial ureteral obstruction, the passage of a retrograde stent at the time of cystoscopy might bypass the site of blockage. The retrograde stent used is a 7F to 9F flexible, double “J” retrograde ureteral stent; it is inserted with the aid of a stent-placement apparatus that has an elevator attachment to the cystoscope. Great care must be taken, as this procedure has the risk of ureteral perforation. When the stent does not pass readily, the performance of a percutaneous nephrostomy is preferable.
In patients in whom complete ureteral obstruction is suspected (i.e., because of a rising serum creatinine level or the development of an acute unilateral hyelonephrosis), a computed tomography urogram should be performed if the serum creatinine value is less than 2.0 mg/dL; an ultrasonogram should be obtained if the level is higher. In patients with complete ureteral obstruction, the problem must be corrected immediately, either by temporary urinary diversion by means of a percutaneous nephrostomy or by reexploration and repair of the ureter. Repair may be by either reanastomosis or reimplantation.
Mild degrees of hydroureter are managed by bladder drainage alone in most patients, as these problems are usually temporary and resolve gradually as the edema subsides. Infection should be treated with appropriate antibiotics.
In patients undergoing radical hysterectomy and bilateral pelvic lymphadenectomy, postoperative ureteral damage from devascularization can be decreased by minimizing the collection of fluid in the retroperitoneal space by leaving the pelvic peritoneum open.
If an excretory urogram cannot be performed (e.g., because of dye sensitivity) or if the study is inconclusive, retrograde pyelography may be necessary. This procedure is potentially morbid and should be performed only if the information to be gained is critical to the decision regarding diversion of the affected kidney (55,56,57). Contrast injected beyond a high-grade obstruction can produce pyelonephritis and sepsis and may require urgent drainage through a percutaneous nephrostomy. The attempted passage of a retrograde ureteral catheter or stent may be useful for diagnosis, and it will stent the ureter if the obstruction has not resulted from a misplaced suture ligature.
In patients with an obstructed ureter that cannot be decompressed by means of a retrograde ureteral stent, a percutaneous nephrostomy tube can be placed under fluoroscopic guidance (56). This procedure is relatively easy to perform, and the tube can be changed or replaced as necessary. In addition, an antegrade ureteral catheter or stent can occasionally be passed through a nephrostomy to allow removal of the percutaneous stent in patients in whom a retrograde catheter cannot be passed.
Ureteral Reanastomosis (Ureteroureterostomy)
When the ureter has been transected or damaged beyond repair, it will have to be revised and reanastomosed or reimplanted into the urinary bladder. If the ureteral injury is above the level of the pelvic brim, a simple reanastomosis is the procedure of choice. The two ends of the ureters are trimmed at a 45-degree angle. A double “J” ureteral stent is passed into the distal ureter with one “memory” end inserted into the bladder. The proximal end of the ureter is placed over the stent and sutured to its distal counterpart (58). Interrupted 4-0 absorbable PGA sutures are placed at close intervals in a circumferential fashion (Fig. 20.14). After several weeks, the absence of leakage can be established by means of intravenous pyelography, and the stent can be removed through a cystoscope.
The reimplantation of the distal ureter into the bladder is known as the Leadbetter procedure, or ureteroneocystostomy (44). This operation is preferred for the ureter that has been disrupted distal to the pelvic brim, as long as the bladder can be sufficiently mobilized on the side of reimplantation. Integral to successful ureteral reimplantation is the creation of a submucosal tunnel (Fig. 20.15). The tunnel minimizes the risk of vesicoureteral reflux and chronic, recurrent pyelonephritis (44).
The technique is as follows:
Figure 20.14 Ureteroureterostomy. A: The two ends of the ureters are cut diagonally. B: A ureteral stent is inserted into the proximal and distal ureter, and interrupted full-thickness sutures are placed. C: The completed anastomosis.
Figure 20.15 Ureteroneocystostomy. A: A submucosal tunnel is created. B: The ureter is brought into the bladder. C: The ureter is passed through the tunnel and sutured to the bladder serosa and mucosa. The serosa of the bladder is sutured to the psoas muscle to stabilize the anastomosis.
In cases where there is inadequate length of viable ureter, additional techniques can be utilized. The Boari flap utilizes a segment of bladder to bridge the defect to the remaining ureter (62). A U-shaped flap is created, turned superiorly and tabularized to create the additional length required. It should not be performed in patients with a history of prior radiation because of the increased risk of flap ischemia and failure. Ileal interposition can also be used to replace a portion of the ureter (63,64). First the required length of ileum is isolated and the ureter is anastomosed to the proximal end of the ileal segment in an end-to-side fashion with 4-0 delayed absorbable sutures. The ileum is then anastomosed in a tension-free fashion to the dome of the mobilized bladder.
In patients with a contracted, fibrotic bladder or when a portion of bladder requires resection for disease, the ileum can also be used for an augmentation ileocystoplasty. In this procedure, the isolated ileum is detubularized by opening it along its antimesenteric border and folding it into a “U” shape. The medial sides are sewn together using a running delayed absorbable suture, creating an ileal segment which can be anastomosed to the bladder (63).
Another procedure that can be useful in the carefully selected patient is the transureteroureterostomy (TUU) (65). When the distal ureter must be resected on one side, and the proximal ureter is too short to permit ureteroneocystostomy, it is possible to anastomose the distal end of the resected ureter into the contralateral side (66). The distal end of the partially resected ureter is tunneled under the mesentery of the sigmoid colon and approximated, end to side, to the recipient ureter. A ureteral stent is used to protect the anastomosis and is left in place for at least 7 to 10 days.
Permanent Urinary Diversion
Permanent urinary diversion must be performed after cystectomy or in patients who have an irreparable fistula of the lower urinary tract. Lower urinary tract fistulae can result from progressive tumor growth or from radical pelvic surgery and/or pelvic irradiation. The most common fistula is ureterovaginal.
The most frequently employed techniques for urinary diversion are the creation of an ileal conduit (the “Bricker procedure”) (67), the creation of a transverse colon conduit (68), and the creation of a “continent” urinary conduit (e.g., the Koch, Miami, Indiana, Mainz, and Rome, pouch) (69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84). The ileal conduit has been the most widely used means of permanent urinary diversion, and it is suitable for most patients. A segment of transverse colon can be used if the ileum has been extensively injured (e.g., by radiation therapy). The transverse colon is usually away from the irradiated field, and thus its vascularity is not compromised.
The continent urinary conduit may be helpful for gynecologic oncology patients who require exenterative surgery. The first such conduit was the continent ileal conduit (or “Koch pouch”). It requires a longer portion of the ileum (up to 100 cm), a longer operative time (4 to 6 hours), and greater technical skills for the creation of the continence mechanism. It is now rarely performed (69).
The continent colon conduit utilizes the intestine from the terminal ileum to the mid-portion of the transverse colon and has been popularized in Indiana, Miami, and Mainz. These pouches are technically somewhat easier to perform than the Koch pouch. The type of continent conduit created is generally determined by the training and preference of the surgeon (70,71,72,73,74,75,76,77,78,82,83).
Technique The technique for the creation of an ileal conduit involves the isolation of a segment of ileum at a site where the intestine appears healthy and nonirradiated. This is typically about 30 to 40 cm proximal to the ileocecal junction. The conduit requires a segment of ileum measuring approximately 20 cm and its associated mesentery (67). After isolation of the segment, the ileum is reanastomosed (Fig. 20.16). The ureters are implanted into the closed proximal end of the ileal segment, and double “J” ureteric stents are placed into both ureters. A no. 8 pediatric feeding tube made of soft flexible plastic can also be employed for the ureteric stent, as it is relatively atraumatic. The “butt” end of the conduit is sutured to the area of the sacral promontory.
The distal end of the conduit is brought through the anterior abdominal wall of the right lower quadrant, approximately midway between the umbilicus and the anterior superior iliac crest. The ureteral stents should be left in place for about 10 days.
When a transverse-colon conduit is selected, the technique is essentially the same. Care must be taken in both techniques to ensure that the vascularity of the intestinal mesentery is not interrupted. The mesentery of the reanastomosed bowel must be reapproximated to prevent herniation of intestinal loops through the defect.
Figure 20.16 Ileal urinary conduit. A: A segment of nonirradiated ileum is used for the conduit. B: The ileum is reanastomosed, and the ureters are sewn into the “butt” end of the conduit. Note that ureters are stented individually.
The technique for creation of a continent Miami or Indiana pouch (70,71,72,73,74,75,76) involves resection of the intestine from the last 10 to 15 cm of ileum to the midportion of the transverse colon. The colon is opened along the antimesenteric border through the teniae coli (Fig. 20.17A). The ileum is used to create the continence mechanism (Fig. 20.17B). The ileal-cecal valve serves as the principal portion of the mechanism; the terminal ileum is narrowed, and several pursestring sutures are placed near the valve to reinforce the continence portion of the conduit (Fig. 20.17C). The ascending colon is sutured or stapled to the transverse colon to create a pouch. An ileotransverse anastomosis is performed to reconstitute the intestine (Fig. 20.17D).
Other modifications of the ileocolonic continent reservoir have been described. The Rome pouch utilizes multiple transverse teniamyotomies of the cecum instead of detubularization to create a low pressure reservoir (82,83). Bochner et al. describe the use of a modified ureteroileocecal reservoir, which utilizes the appendix to create the cutaneous stoma (85).
In rare instances, a terminally ill patient undergoing exploratory surgery will have a bladder fistula. In such circumstances, one ureter can be ligated, and a skin ureterostomy can be created with the other ureter. The ureter is mobilized from its attachments and brought laterally through the retroperitoneal space to the lateral and anterior abdominal wall. The ureter is tunneled through the fascia and brought out through a stab incision in the skin, where it is affixed to create a small stoma (55).
Figure 20.17 Colon continent urinary conduit: the “Miami” pouch. A: The segment of distal ileum and ascending and transverse colon is isolated, and the segment is opened on its antimesenteric border along the teniae coli. B: The ureters are reimplanted into the mesenteric side of the ascending colon, a continence mechanism is created with pursestring sutures near the ileal-cecal junction, and a noncutting double staple line is performed with a gastrointestinal anastomosis (GIA) stapler. (Continued)
Figure 20.17 C: The conduit is closed with running sutures, and the ileal stoma is created. D: The intestines are reconstituted with an ileal-transverse colon anastomosis.
Reconstructive operations, particularly pelvic floor reconstruction and creation of a neovagina, are important in patients who are undergoing extensive extirpative procedures, such as pelvic exenteration (see Chapter 22). Vaginal reconstruction helps to provide support to the pelvic floor, thereby reducing the prospect of perineal herniation. By helping to fill the pelvis, vaginal reconstruction also decreases the incidence of enteroperineal fistulae. Pelvic floor reconstruction should be performed in all patients undergoing a pelvic exenteration, and vaginal reconstruction should be performed simultaneously in most patients. The surgeon must be well acquainted with the types of graft that can be employed in the performance of these reconstructive operations and the techniques necessary to accomplish them (86).
Grafts used for reconstructive operations in the pelvis are either skin grafts, which can be full or partial (split) thickness (86,87,88,89,90), or myocutaneous grafts, which are composed of the full thickness of the skin, its contiguous subcutaneous tissues, and a portion of a closely associated muscle (91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110). The most frequently used myocutaneous pedicle grafts contain muscle segments from the rectus abdominis muscle of the anterior abdominal wall, gracilis muscle of the inner thigh, the bulbocavernosus muscle of the vulva, the tensor fascia lata muscle of the lateral thigh, and the gluteus maximus muscle.
Skin grafts must be harvested under sterile conditions (61). The donor site most frequently used to obtain a split-thickness skin graft is either the anterior and medial thigh or the buttock. Although the thigh may be more readily accessible to the surgeon, the buttock donor site has cosmetic advantages; however, this latter site may be more uncomfortable in the postoperative recovery period. The selection of the donor site should be made preoperatively after discussion with the patient.
A dermatome is used to harvest the skin graft. Several different types of dermatome are available, including the Brown air-powered, electrically driven dermatome and the Padgett handdriven dermatome. The surgeon should select the instrument with which he or she has the greatest facility, as an equally good graft can be harvested with either one.
The technique for obtaining the skin graft is as follows:
The purpose of the pedicle graft is to provide a substantial amount of tissue along with its blood supply either to repair an anatomic defect or to create a new structure, such as a neovagina (86,87,88,89,90,91,92). The pedicle graft can be either a full-thickness skin and subcutaneous tissue graft, as is used frequently for closure of a vulvar defect, e.g., a “Z-plasty” (a “rhomboid flap”), or a myocutaneous graft, e.g., a rectus abdominis or gracilis.
Before harvesting a pedicle graft, the surgeon should carefully outline the incisions on the skin with a marker pen. During the mobilization of the myocutaneous pedicle, the surgeon must carefully isolate and preserve the neurovascular bundle that supplies the muscle.
Vaginal reconstruction in the gynecologic oncology patient is performed either to revise or replace a vagina that has stenosed as a result of prior vaginal surgery and/or radiation or to create a neovagina when the vagina has been removed (87).
When the vagina is fibrotic after irradiation, the scarred vaginal tissue first must be resected before placement of the split-thickness skin graft (87,88,89,90). The skin graft is placed over a vaginal stent that is then inserted into the space created by resection of the old, scarred vagina (Fig. 20.18A). The Heyer-Schulte stent is the vaginal stent preferred for this purpose, because it is inflatable, can be easily removed and replaced by the patient, and has its own drainage tube (Fig. 20.18B).
Split-thickness skin grafts can also be used in patients undergoing exenteration, but this approach is less satisfactory than the use of myocutaneous pedicle grafts, as discussed below. When an anterior exenteration is performed, or when a portion of the rectosigmoid colon is resected but primarily reanastomosed, a neovagina can be created with the use of skin grafts. The omentum is mobilized by ligating and dividing the short gastric vessels along the greater curvature of the stomach, preserving the left gastroepiploic pedicle (Fig. 20.19A). The omentum is then placed into the pelvis and sutured to the rectosigmoid posteriorly and laterally to create a pocket for the neovagina. Split-thickness skin graft(s) are harvested, sewn over a vaginal stent, and inserted into the newly created pelvic space (Fig. 20.19B).
Figure 20.18 Creation of neovagina after radiation. A: The vaginal scar is resected in preparation for vaginal reconstruction with split-thickness skin grafts. B: A Heyer-Schulte vaginal stent has the skin graft placed around it, and this is inserted into the pelvic space to create a neovagina.
Figure 20.19 Mobilization of the omentum. A: This is accomplished by ligating and dividing the right gastroepiploic artery and the short gastric arteries along the greater curvature of the stomach. B: The omentum is used to create a “pocket” for the placement of a split-thickness skin graft.
Tranverse (TRAM) and Vertical (VRAM) Rectus Abdominis Myocutaneous Pedicle Grafts
A single rectus abdominis pedicle graft can be used to create a neovagina (91,92,93,94,95,96,97,111,112,113,114,115), or to repair a pelvic or perineal defect (95,96). This is our preferred technique for creation of a neovagina performed simultaneously with a pelvic exenteration (Fig. 20.20). The technique is relatively straightforward and has the advantage of a single pedicle harvested from the same site of the abdominal incision used to perform the exploratory surgery. This approach avoids the use of separate incisions on the inner aspects of the thigh as needed for the gracilis myocutaneous pedicle graft. The disadvantage is that the amount of tissue mobilized from the anterior abdominal is limited, and thus, the ability to adjust the size of the neovagina is somewhat limited. If too large a pedicle is created, there will be too much tension for the abdominal closure, and this can also create distortion of the anterior abdominal wall skin.
The pedicle location is shown in Fig. 20.20A. The oval-shaped pedicle should measure approximately 6 to 8 by 10 centimeters. The skin of the pedicle is incised, and the cephalad portion of the rectus abdominis muscle and attached myofascial tissues are transected (Fig. 20.20B). The tubular neovagina is created by suturing together the sides of the pedicle. One end is left open, and this becomes the distal neovagina. The pedicle is harvested, mobilized, and brought into the pelvis (Fig. 20.20C). The pedicle graft is then sutured to the preserved vaginal introitus to complete the procedure (Fig. 20.20D).
Figure 20.20 The transpelvic rectus abdominis myocutaneous (TRAM flap) pedicle graft. A: The location of the myocutaneous pedicle flap of the rectus abdominis muscle. B: The pedicle is harvested, and the tubular neovagina is created by suturing the full-thickness of the muscle, subcutaneous tissue, and skin of the anterior abdominal wall. C: The pedicle graft is brought down into the pelvis, and the leading edge is sutured to the preserved vaginal introitus. (Continued)
Figure 20.20 D: The final result is a neovagina that also helps to protect the pelvic floor from intestinal adhesions.
Gracilis Myocutaneous Pedicle Grafts
Bilateral gracilis myocutaneous pedicle grafts can be used to construct a neovagina (86,87,102,103). In addition, the grafts provide excellent support for the pelvic viscera. The gracilis myocutaneous graft is harvested (Fig. 20.21A) from the inner aspect of the thigh.
A line is drawn from the pubic tubercle to the medial epicondyle, and this delineates the anterior margin of the graft. The graft should be about 5 cm wide and about 10 cm long. A skin bridge is preserved between the vulva and the pedicle. The myocutaneous pedicle graft is mobilized by transecting the gracilis muscle distally in continuity with the skin and subcutaneous tissue (Fig. 20.21B). The vascular pedicle is proximal, and it must be carefully identified and preserved.
The pedicle is “harvested,” brought under the skin bridge of the vulva, and exteriorized through the introitus (Fig. 20.21C). The two grafts are sutured together to create a hollow neovagina (Fig. 20.21D, E). The entire neovagina is placed into the pelvis by posterior and upward rotation and sutured to the introitus (Fig. 20.21F). The apex is sutured to the symphysis pubis and/or the anterior sacrum. At the completion of the procedure, an omental pedicle is brought down over the graft to reconstruct the pelvic floor (Fig. 20.21G).
Bulbocavernosus Pedicle Grafts
The bulbocavernosus myocutaneous pedicle graft has been used for repair of radiation-induced rectovaginal fistulae (Martius procedure), but the procedure has been adopted for the creation of a neovagina (103,104,116). The procedure is performed by making an incision over the labium majus, isolating the bulbocavernosus muscle superiorly and anteriorly, and mobilizing it on a posterior vulvar pedicle. The graft is tunneled under a skin bridge at the posterior introitus and sutured to the pedicle of the other side.
Some authors have preferred to use a segment of colon to create a neovagina (86,105). This technique has had mixed success in the past, but an approach using a portion of the ascending colon may be an improvement over earlier procedures.
Figure 20.21 The gracilis myocutaneous pedicle graft. A: The pedicle graft is outlined on the inner thigh overlying the gracilis muscle. B: The myocutaneous pedicle graft is mobilized. C: The pedicle is brought under the skin bridge of the vulva. D,E: The two grafts are sutured together. (Continued)
Figure 20.21 F: The neovagina is placed into the pelvis and sutured to the introitus. G: An omental pedicle is used to cover the graft. (Reproduced from Berek JS, Hacker NF, Lagasse LD. Vaginal reconstruction performed simultaneously with pelvic exenteration. Obstet Gynecol 1984;63:318, with permission from the American College of Obstetricians and Gynecologists.)
Vulvar and Perineal Reconstruction
Whenever feasible, the vulva should be closed primarily after radical vulvectomy (106,107,108,109,110,117). With radical local excision or a separate incision approach for the groin dissection, primary closure of the vulvar skin can be accomplished in almost all patients.
Rhomboid Pedicle Graft
If there is any tension on the skin edges, the skin can be mobilized by means of a Z-plasty using the adjacent skin and subcutaneous tissue. This is called a rhomboid flap (106). The technique (Fig. 20.22) involves the repositioning of a rhomboid flap of full-thickness skin and subcutaneous tissue. Use of these pedicle grafts will usually allow for the primary closure of vulvar defects after radical vulvar surgery, but if necessary, a split-thickness skin graft can be used. Myocutaneous pedicle grafts, such as a unilateral gracilis graft, can also be used to cover a large vulvar defect.
Tensor Fascia Lata Pedicle Graft
The tensor fascia lata pedicle graft, harvested from the lateral aspect of the thigh, can be useful in covering large defects of the lower abdomen, groin, and anterior vulva (107). The flap is particularly useful in patients who require extensive resection of large groin recurrences or large, fixed groin nodes.
Figure 20.22 A: The “rhomboid flap” is used to close a posterior vulvar defect. B: The pedicle grafts are bilateral “Z-plasties” that are sutured together in the midline.
The graft is obtained by harvesting a myocutaneous pedicle from its proximal origin at the anterosuperior aspect of the iliac bone to its distal insertion on the lateral condyle of the tibia. The length of the proposed flaps is determined by measuring the distance from the muscle's vascular supply, located 6 to 8 cm distal to the anterior superior iliac spine, to the most inferior or distal point of the recipient site (e.g., the posterior vulva). The blood supply is from the lateral circumflex femoral artery located deep to the fascia lata between the rectus femoris and the vastus lateralis. The posterior border of the graft is defined as a line from the greater trochanter of the hip down to the knee, and the distal border is located about 5 cm proximal to the knee. The width of the flap is determined by the width of the defect to be covered, but typically it is 6 to 8 cm with a length of up to 40 cm.
The pedicle graft is harvested after the defect has been created in order to permit a more accurate measurement of the flap. The flap is first incised distally, and care is taken to avoid injury to the proximal blood supply. Once the flaps are elevated, they are rotated into place and sutured from their most distal point to the proximal. The donor site is closed primarily.
Gluteus Maximus Pedicle Grafts
The gluteus maximus muscles, or a portion thereof, can be used to reconstruct the pelvic floor and the perineum (109,110). This approach might be particularly useful for very large defects, such as for those patients who have undergone a total infralevator pelvic exenteration (see Chapter 22).
Although the preferred methods for vulvar and vaginal reconstruction are outlined above, there is occasionally a need to perform a vulvovaginoplasty, the so-called William's procedure. This procedure (117) involves the incision of a horseshoe-shaped flap on the vulva to create a marsupialized pouch that can be used as a neovagina. This operation has the advantage of being relatively simple to perform, and it does not require pelvic dissection. It has the disadvantage of being less anatomically suitable for vaginal intercourse, but its direction can improve with regular use. It may be helpful in a patient who has undergone a pelvic exenteration without vaginal reconstruction.
Pelvic Floor Reconstruction
At the completion of a pelvic exenteration, the pelvic floor must be reconstructed. Probably the most effective procedure is to perform an omental pedicle graft (provided there is sufficient omentum) and to use myocutaneous pedicle grafts whenever possible to reconstruct the vagina. In patients in whom this is not possible, alternatives include the use of a variety of graft materials, either natural or synthetic (118). A natural material that has been used is dura mater, but this is often unavailable (119,120). All areas that can be directly peritonealized should be carefully covered with peritoneal pedicle grafts.
Synthetic grafts using Marlex have been associated with a high incidence (>20%) of infectious morbidity and are therefore much less desirable. However, if a pedicle graft is not feasible, the synthetic material Gore-Tex may be the best alternative (121,122).