The late Douglas Wong was regarded as an international authority on TRUS and in its interpretation. He previously had been kind enough to provide me a narrative of his technique, and I thought it would be useful to reproduce it here:

The patient is instructed to take a small volume enema 1 hour prior to the examination. The procedure is
carefully explained to the patient and pertinent questions answered. The assistant enters the demographic data in the ultrasound computer as well as the frequency of the ultrasound probe and its focal range. The patient is placed in the left lateral position on the examining table. The study is preceded by a digital rectal examination to evaluate the size, fixation, location, and morphology of the rectal lesion. In most instances the use of a large-bore proctoscope serves several purposes. It allows visual examination of the rectal tumor with exact determination of its location both with respect to circumferential involvement of the rectal wall and the distance from the anal verge. Secondly, it allows suctioning of any residual stool or enema fluid that might interfere with the acoustic pathways of the ultrasound waves that may distort the image. Most important, however, it allows easy passage of the probe above the tumor to ensure that the transducer is advanced above the rectal lesion to allow complete imaging. This is significant since the lower border of a rectal cancer can differ in the depth of invasion than the center or upper portions of the cancer and lymph nodes in the perirectal region are often seen just above the level of the tumor. They will be missed if complete imaging is not obtained. Small distal lesions can be adequately imaged with the ultrasound inserted blindly and advanced above the lesion, but for most midrectal, bulky tumors the use of a proctoscope will facilitate the passage of the transducer.

The probe is prepared by placing a condom over the transducer head followed by a metal ring or rubber band that secures the base of the balloon. The assistant holds the probe with the balloon in the most dependent position and fills the balloon with about 50 mL of water via the connector at the base of the metal shaft. Any air in the system is aspirated through the syringe and expelled. A water-soluble lubricant is liberally applied to the outside of the balloon. The probe is now ready for insertion. The proctoscope is then advanced above the rectal lesion, and water-soluble lubricant is inserted into the lumen of the proctoscope to facilitate passage of the probe. The probe is then gently introduced through the proctoscope and advanced such that the transducer is sited above the rectal cancer. Once the 20 cm mark on the shaft of the probe is at the proximal end of the proctoscope, the proctoscope is then pulled back on the probe as far as possible, thus exposing the transducer for 7 cm beyond the end of the instrument and positioned about the rectal cancer. The balloon is then instilled with 30 to 60 mL of water; this is the volume required to obtain optimal imaging. The transducer is activated by depressing the button on the proximal end of the probe, and the image on the screen is visualized.

When the connector for introducing the water into the balloon is pointing upwards (towards the ceiling), by convention the anterior aspect of the rectum will be on the superior part of the screen, the right lateral rectum will be on the examiner’s left, the left lateral wall will be on the examiner’s right, and the posterior rectum will be in the lower screen. The tip of the ultrasound probe should be maintained in the center
of the rectal lumen in order to achieve optimal imaging of the rectal wall and perirectal structures. Some adjustments may have to be made on the gain of the ultrasound unit in order to provide better imaging. Occasionally, it is possible to perfectly depict all five layers of the rectum circumferentially, but usually only a portion of the rectal wall at a time will be optimally imaged. Minor adjustments will have to be made in the location of the probe relative to the rectal wall at various locations to optimally image all five layers clearly. Once optimal imaging of the rectal wall and surrounding structures has been achieved, the ultrasound probe is gradually withdrawn while carefully observing the screen and the images obtained. Several hard-copy images should be obtained for future reference through the use of a Polaroid image recorder. These images can be obtained by stopping the rotation of the transducer by depressing the activation/deactivation button and holding it longer, thus activating the recorder.

When a critical area of the tumor needs to be visualized at a higher magnification, the window can be activated which gives a large image of the area being examined. The entire length of the rectal tumor is evaluated. It is not uncommon for one to perform several passes along the full length of the tumor in order to acquire all the relevant information.

Attention must also be focused on the perirectal tissues in order to search for potentially involved lymph nodes. In general, normal lymph nodes are not visualized with the ultrasound, and, therefore, any hypoechoic structure in the surrounding perirectal tissue should be suspected of harboring metastatic disease. Lymph nodes often exhibit hypoechoic echogenicity comparable to that of the primary tumor, are more often round than oval, and are frequently irregular in appearance. Lymph nodes can be distinguished from blood vessels that are also circular hypoechoic areas, but when followed distally and proximally, they seem to extend further and may be seen to elongate and to branch. Once the study is completed, the balloon is deflated, and the probe is removed. Figure 24-5 illustrates the five layers of the rectal wall seen schematically and in the TRUS image of the normal rectum. The inner white
line represents the interface of the balloon with the mucosal surface of the rectal wall. The inner black line represents the mucosa and muscularis mucosa, whereas the middle white line corresponds to the submucosa. It is this middle white line that is the most crucial layer to visualize in order to determine whether the tumor is invasive.²⁹² The outer black line corresponds to the muscularis propria, whereas the outer white line represents the interface between the muscularis propria and the perirectal fat. Once it has been ascertained that the middle white line is broken, then the presence of an invasive tumor has been confirmed. It is then a matter of determining the depth of invasion. Figure 24-6 demonstrates an ultrasound of a minimally invasive cancer with the corresponding resected, microscopic area. Figure 24-7 shows transmural involvement by tumor.

The TNM (tumor, node, metastasis) classification is used with a u modifier to describe depth of invasion and the presence or absence of metastatic lymph nodes as described by Beynon and coworkers^51,53:

In the foregoing, uN₀ represents evidence of lymph nodes, uN₁ represents lymph nodes positive for tumor, MX indicates metastatic tumor status unknown, M0 represents no evidence of metastatic tumor, and M1 means that metastatic tumor is present.

In addition to the cited indications, this technique offers the opportunity for clear visualization of the full thickness of an anastomotic area in those patients who have undergone restoration of rectal continuity, especially with respect to the possibility of early detection of recurrent cancer and for assessing the effects of preoperative radiotherapy.^{52,102,296,400,454,519} Further information can be obtained by means of ultrasonographically guided biopsy and histologic determination of lymph node status.^562,664

The Creighton University group in Omaha, Nebraska, performed preoperative ultrasound staging on 107 patients with rectal cancer.⁴ TRUS identified 18 of 19 patients with uT₃ tumors and 44 of 51 individuals with less invasion than uT₃ (86.3%). Garcia-Aguilar and colleagues reported the University of Minnesota group’s experience of 1,184 patients with rectal adenocarcinoma or villous adenoma who underwent TRUS, comparing their assessment with pathologic specimens obtained by either resection or transanal excision without neoadjuvant treatment.¹⁹⁸ Somewhat disappointingly, overall accuracy in assessing the degree of rectal wall invasion was only 69%, with 18% overstaged and 13% understaged. Overall accuracy in assessing nodal involvement in those who underwent radical surgery (238 patients) was 64%, with 25% overstaged and 11% understaged. The primary benefits appear to have been in determining which tumors were benign and those that either went completely through the rectal wall and those that did not.¹⁹⁸

I prefer this instrument to the traditional Yankauer suction because of its length and also because I frequently use high-pressure water irrigation to open up planes, especially in redo surgery. Once an incision is made in the correct plane, water irrigation can develop the plane with ease. Furthermore, the long sucker does not have a bend in it. I find that an angled sucker can interfere with the line of vision deep in the pelvis, whereas straight suction permits a clear line of view.

A long diathermy is essential because sharp dissection is used throughout the procedure. I do not use hands or fingers for mobilizing.

For the past 3 years, I have increasingly adapted the use of different energy sources for open as well as for laparoscopic surgery. I have found the long Harmonic scalpel and the Wave Harmonic scalpel particularly useful; they have complemented, and in some steps of the operation, they have replaced the use of diathermy.

This is essential for dissection in the pelvis (Figure 24-11).

This retractor has the same shape as a St. Mark’s retractor but is much narrower (Figure 24-12). It is very useful in patients with a narrow pelvis or in those with a bulky tumor in whom it may be difficult to accommodate the width of a St. Mark’s retractor.

If the tumor is in the upper third of the rectum, the perirectal fat is divided 3 to 5 cm distal to the lower end of the tumor. This is usually straightforward in thin patients. However, in obese individuals, this can be a difficult step, and I have found the following technique useful and reproducible.

The visceral peritoneum is divided by diathermy on either side of the rectum at the required level. The rectum is held up between the left index finger and thumb to palpate the antimesenteric border of the rectum. While keeping the index finger in place, the thumb is replaced by a long pair of straight scissors and pushed through from left to right. This opens up the plane between the posterior longitudinal muscle wall of the rectum and the perirectal fat. The perirectal fat is taken in two portions, the larger of which contains the superior rectal artery. Straight Kocher clamps are used for that purpose, and ligation is performed using a heavy suture ligature (Figure 24-27). The back of the rectum can then be inspected to ensure that there is an intact layer of longitudinal muscle, and the rectum is stapled and divided between two double rows of staples with a stapling device. I prefer to use the Contour stapler because its curved shape fits well into the pelvis and also because the rectum can be divided between two rows of staples on either side, thus avoiding any spillage (Figure 24-28). Alternatively, the rectum can be divided between a crushing clamp and a Furness clamp, which can then be used to apply a purse string to the rectum (Figure 24-29). I no longer use the Furness clamp because in a narrow pelvis,
there is the risk of a straight needle injuring structures on the lateral walls as it enters or exits the clamp. Alternatively, a hand-sewn purse string can be used (Figure 24-30).

If a hand-sewn anastomosis is performed, then a rightangled soft bowel clamp is applied to the rectum, and the specimen is divided between that clamp and a Kocher’s clamp. An assistant then irrigates the rectum until the return is clear. The rectal stump is surrounded with packs, and the soft bowel clamp is removed. Four stay sutures, two posteriorly and two anteriorly on either side of the midline, are applied to the rectal stump in preparation for a hand-sewn anastomosis as described later (Figure 24-31).

When the tumor is in the midportion or lower portion of the rectum, the dissection has to proceed in the retrorectal space to the pelvic floor, and the anterior plane of dissection should be as low as possible. The tubular lower end of the rectum is thus fully mobilized. There is little or no perirectal fat at this level, and the rectum can be transected using the Contour stapling device (Figure 24-32).

In obese patients and males with narrow pelvis, it is sometimes useful to ask an assistant to apply pressure to the perineum. This can often push the lower rectum into the pelvis and facilitate stapling and transection (Figure 24-33).

It is essential that the proximal colon be completely mobilized as described earlier. There must be no tension at all on the anastomosis when the colon is pulled through. As a rule, if the proximal colon can reach the patient’s left knee, it will be easy to pull it through without tension. The colon is prepared by stapling it with a linear device. A small incision is made in the middle of the staple line, and a Foley catheter is inserted into the colon; the catheter balloon is inflated. The catheter will be used to pull the colon down through the anal canal.

When the rectum is fully mobilized to the pelvic floor, it is transected by diathermy at least 1 cm below the tumor. A Lone Star retractor is then used for exposure, and eight sutures (at regular round-the-clock intervals) are inserted into it using 2-0 Vicryl on a 26 needle. The sutures are placed from outside in. The needles are not cut off, and the sutures are pulled out and clipped to the rim of the retractor (Figure 24-50). The proximal colon is then pulled through using the Foley catheter as a guide and for traction (Figure 24-51). The staple line is excised, and the colon is sutured to the anal canal by each of the individual sutures that had been preserved. After this has been accomplished, each suture is tied and divided. The suture line disappears out of sight into the anal canal.

The mobilization of the rectum is the same as has been described earlier for a low anterior resection. The rectum is mobilized along anatomical planes all the way down to the pelvic floor. I aim to perform as much of the dissection as possible during the abdominal dissection before proceeding with the perineal operation. The inferior mesenteric artery is ligated and divided as previously described, and the sigmoid mesentery is divided at an appropriately selected segment (see Figure 24-34). I always test the marginal artery before ligation and division to ensure a good blood supply. Once the mesentery is divided, the colon is transected with a linear stapling device.

General Comments. Various techniques have been described for the perineal dissection. In the first instance, the “traditional” method of dissection will be described. Other techniques that involve a wider and more radical approach will be dealt with later in this text.

I consider the perineal dissection as important, if not more important, than the abdominal procedure. It can be technically demanding, and I strongly recommended that it should be performed either by the principal surgeon or by another experienced colorectal surgeon or by senior-level resident or a fellow under direct supervision. It should not be left to a junior or inexperienced resident. The perineum is prepared and draped. I prefer to have instruments, including diathermy and the Wave Harmonic scalpel, placed on a small perineal table. There should be no leads dangling off the patient’s leg drapes; they get in the way and obstruct the view. The table is tilted head down and the stirrups are elevated. A purse-string suture is applied to the anus to prevent fecal contamination. An elliptical incision is made with the diathermy (Figure 24-57), and a Lone Star retractor is used to retract the skin. In the past, the rest of the dissection was performed using diathermy; however, over the past 3 years, I have used the Wave Harmonic scalpel. This instrument is useful in preventing bleeding during the procedure and an almost bloodless dissection can be achieved (Figure 24-58). The Wave Harmonic scalpel is used to deepen the incision through the ischiorectal adipose tissue. The index finger of the left hand is then used as a guide to locate the tip of the coccyx, and the anococcygeal ligament is divided (Figure 24-59). The finger is then inserted anterior to the coccyx. The abdominal operator lifts the rectum in his left hand and inserts his right hand deep into the retrorectal space until both operators can feel each other’s fingers as separated by Waldeyer’s fascia (Figure 24-60). The perineal surgeon then uses a pair of Abel scissors to break through this layer into the retrorectal space (see also Figure 24-60). Usually, a small amount of blood that has collected at the lower end of that space needs to be sucked out. Attention is then directed to dividing the iliococcygeus part of the levator muscles on either side. The left index finger is inserted into the opening, which has been made in the presacral space, and the finger is hooked under the left levator muscle. The Wave Harmonic scalpel is used to
divide the muscle (Figure 24-61). The left index finger is similarly used to identify the right levator muscle, and the wave instrument is used to divide the muscle on the right side. Attention is then directed to the pubococcygeus and the puborectalis parts of the levator muscle. The muscle is divided with the Wave Harmonic scalpel from lateral to medial on either side. Branches of the inferior hemorrhoidal vessels are in close proximity to the puborectalis and can cause troublesome bleeding. Simply dividing the muscle with diathermy usually does not achieve adequate hemostasis. However, I have found that this can be achieved with the Wave Harmonic scalpel (Figure 24-62). Dissection between the rectum and prostate or vagina can now begin. The abdominal surgeon’s fingers can be used to guide the perineal surgeon. The perineal surgeon’s left hand is used to retract the rectum down, and the diathermy is used to dissect in the plane between the rectum and the prostate in males. The rectourethralis muscle on either side is divided (Figure 24-63); the dissection continues until the abdominal surgeon’s fingers are met, posterior to Denonvilliers’ fascia; and the dissection breaks into the pelvic cavity. By remaining in this plane, injury to the urethra is avoided. Another method of performing the anterior part of the dissection is to deliver the specimen through the perineum; once the posterolateral part of the dissection is performed, the specimen can be passed to the perineal surgeon who then retracts the specimen downward to better identify the anterior plane of dissection (Figure 24-64). In females, the anterior plane of dissection is between the rectum and the posterior wall of the vagina. If the tumor has invaded the posterior wall of the vagina, then the Wave Harmonic scalpel can be used to excise en bloc a portion of the vaginal wall. If the Harmonic scalpel is not available, then the vaginal wall has to be oversewn with an interlocking suture to prevent hemorrhage (Figure 24-65). The wound is closed in layers and drained through the vagina (Figure 24-66).

The specimen is then delivered through the perineum, although sometimes if the specimen is bulky and the pelvis is narrow, then the specimen can be delivered through the abdomen. Once the specimen has been delivered, the perineal wound is carefully examined to ensure hemostasis, especially from the posterior aspect of the prostate or vagina. Because I have started to use the Wave Harmonic scalpel to perform the perineal dissection, there has been a noticeable reduction in blood loss. Once hemostasis is secured, the wound is closed in layers around a drain (Figure 24-67). Occasionally, if there is a bulky tumor with a significant anterior component, the perineal dissection can be performed with the patient in the jackknife position. In order to do so, the colostomy must be exteriorized and the abdomen closed
before the patient is turned over for the perineal dissection. Although repositioning the patient is cumbersome, this position can achieve better exposure and access to the perineal anatomy. It is also more comfortable for both surgeon and assistant. Although I have occasionally used this technique, it is not routine in my practice.

Incidental appendectomy is unnecessary and not indicated at the time of rectal resection, unless the appendix is involved in the tumor mass.

The urinary bladder is not uncommonly directly invaded by rectal cancer. If the dome of the bladder is invaded, an en bloc resection is recommended. It is difficult to know exactly whether the involvement is neoplastic invasion or inflammatory. It is better to assume that it is neoplastic and proceed with en bloc resection of the dome. However, if the trigone is involved, then a specialist urologist should be invited to attend and assist. Under these circumstances, the ureters must be protected from injury, and complex ureteric reconstruction may be required. Depending on the site of invasion, a total cystectomy may be necessary.

Carne and colleagues analyzed 53 patients who underwent en bloc bladder resection for colorectal cancer in New Zealand.⁸⁷ Forty-five had a partial cystectomy. All who did not have en bloc resection developed local recurrence. The authors noted that the decision whether to perform partial or total cystectomy depends on the site of the bladder invasion.

Total pelvic exenteration is defined as the removal of the distal colon and rectum, along with the lower ureters, bladder, internal reproductive organs, perineum, draining lymph nodes, and pelvic peritoneum.³⁷⁰ In a highly selected series from the Ellis Fischel State Cancer Center, Columbia, Missouri involving 24 patients over a 30-year period, Lopez and colleagues reported an operative mortality of about 20% (9% during the last decade).³⁷⁰ The overall survival rate was a remarkable 42%. A review by Williams and colleagues summarizing the results of several series concluded that the procedure can be carried out with a mortality rate of less than 10%.⁶⁵⁴

Sugarbaker has advocated en bloc excision of rectal cancer with sacrectomy for lesions that are fixed posteriorly.⁶⁰⁰ In reporting his experience with six patients, he noted that four survived more than 3 years. Pearlman and colleagues carried out 12 pelvic and 7 sacropelvic exenterations.⁴⁹³ Of the 15 patients without extrapelvic disease, there was 1 operative death, 2 died free of disease, 3 died of cancer, 1 was alive with recurrence, and the remainder were free of disease (none had achieved 5 years). Shirouzu and coworkers performed total pelvic exenteration on 26 patients for locally advanced colorectal cancer.⁵⁶⁷ The operative mortality was 8%. In those with stage II primary disease, the recurrence rate after curative surgery was three of seven, but the mean survival time was 58 months, with a 5-year survival of 71%. Those with stage IV disease had a mean survival time of only 5 months. Others have demonstrated that pelvic exenteration (APR and cystectomy) in selected patients may produce a 25% to 50% 5-year survival.^117,281,316