Dysplasia at the Anal Transition Zone after IPAA

Abstract

Ileal pouch-anal anastomosis is a popular way of reconstruction the gastrointestinal tract after total proctocolectomy for ulcerative colitis. The pouch-anal anastomosis is usually stapled, which requires the preservation of a small amount of upper anal canal and lower rectum. This includes the anal transition zone (ATZ), a surprisingly small and irregular ring of tissue at and just above the dentate line. The ATZ and rectal cuff is prone to inflammation and neoplasia, particularly in patients who had a colon cancer or dysplasia at the time their large bowel was removed. This high-risk group needs ATZ/rectal cuff surveillance before and after the surgery. Those without colorectal dysplasia preoperatively are at low risk of developing ATZ/rectal cuff dysplasia postoperatively and follow-up can be more relaxed. Treatment of ATZ dysplasia is difficult and may mean mucosectomy, pouch advancement, pouch removal, or a redo pelvic pouch.

When the ileal pouch-anal anastomosis (IPAA) was first described by Parks and Nicholls in 1978 it was a real game-changer for patients needing a total proctocolectomy. ¹ The operation allowed a permanent ileostomy to be avoided and gave patients the gift of per anal defecation. A key part of the procedure was the IPAA, where the lower rectum and anus were stripped of their mucosa to form a muscular cuff through which the pouch and efferent ileum were pulled, to create a handsewn ileoanal anastomosis at the dentate line. Parks and Nicholls' intent was to completely remove all large intestinal mucosa so that the mucosal disease that they were treating was cured. However, the process of stripping the mucosa from the lower 8 cm of rectum, which is usually inflamed in patients with ulcerative colitis, was burdensome, while preserving the long rectal muscular cuff seemed not to provide any benefit. The operation therefore evolved, with the S-pouch superseded by a J design, and the sutures of a handsewn anastomosis by staples. This evolution represented a compromise, as a transanally inserted stapling cartridge needed to be accommodated by some bowel above the anus. Total removal of the diseased epithelium was sacrificed in the interests of an easier procedure with better function and fewer complications. ² Now, in the most popular version of Parks and Nicholls' operation, the ileal pouch is stapled to the low rectum just above the dentate line, creating a strip of residual large intestinal mucosa that is still potentially diseased. This strip of mucosa has become known as the anal transition zone (ATZ). The main issues caused by the retained ATZ are inflammation and dysplasia, and this article is concerned with the latter.

The discussion that follows is limited to the ATZ in the setting of ulcerative colitis. The other large group of patients undergoing IPAA has familial adenomatous polyposis (FAP) and their clinical course is shaped by their tendency to form adenomas. Instead of flat, barely visible dysplasia, FAP patients form obvious adenomas at a frequency that increases with time and can pose similar problems to those seen in patients with ulcerative colitis. ³ All FAP patients with a pouch need yearly surveillance of the ATZ and pouch, and the dysplasia is handled in a similar fashion to that arising in patients with ulcerative colitis.

ATZ Anatomy and Physiology

The anal transitional zone was defined by Fenger as a strip of anal epithelium between uninterrupted stratified squamous mucosa of the anal canal below and uninterrupted columnar epithelium of the rectum above. ⁴ Our concept of this area of the anal epithelium is shaped by the simplistic drawings of the anus that appear in various textbooks and manuscripts, but a recent study by Thompson-Fawcett et al has shown that the real anatomy is much more complex. Based on computer modeling of anal histology, Thompson-Fawcett and his colleagues found that the ATZ usually begins at or just below the dentate line, an average of 1.05 cm above the lower border of the internal sphincter. ⁵ In 8 patients the ATZ extended 3 to 6 mm below the dentate line, in 100 patients it started at the dentate line and extended upwards, in 4 patients the lowest point of the ATZ started 1 to 7 mm above the dentate line, and 1 patient had no ATZ. Overall, the ATZ ranged from 6 mm distal to 20 mm proximal to the dentate line. ⁵ Perhaps most surprisingly, the median length of the ATZ was only 4.5 mm (compared with 7.3 mm using the alcian blue stain).

Both Fenger and Thompson-Fawcett et al have shown considerable interpersonal variation in the length and position of the ATZ, and in the shape of its borders. The simple diagram we are used to seeing must be replaced by a much smaller length of transitional epithelium with wavy borders that straddles the dentate line but is not absolutely limited by it. ^{5 6} The significance of this new, more realistic appreciation of the limits of the ATZ is in the way that an IPAA is approached and performed, and in the interpretation of all the previous studies that compared handsewn versus stapled anastomoses. In patients with an IPAA 2 cm proximal to the dentate line the residual mucosa is a mix of ATZ and rectal mucosa. In some patients who underwent ATZ “stripping,” unless the stripping began 6 mm distal to the dentate line there is still ATZ present.

There are now also questions about whether the ATZ is truly prone to dysplasia and inflammation, or are we just seeing residual rectal mucosa affected by these conditions? Fenger has explored the histology of anal transitional epithelium and has shown that it consists of a mixture of columnar, cuboidal, or more flattened cells, some 5 to 9 layers thick. ⁴ Electron microscopy show the details of the ATZ to include normal squamous and columnar cells (a minority), but mostly a cobblestone pattern of row-forming cells with short microvilli. ⁷ In addition, endocrine-like cells and melanocytes can be found. ⁸ With these uncertainties and this new anatomy in mind, the literature concerning ATZ dysplasia can be reexplored.

Techniques

Stapled IPAA : Any complex medical procedure is liable to the variations in technique that develop based on differences in surgical training, experience, skill, and different levels of understanding of what is being attempted. The IPAA is particularly susceptible to this sort of uncertainty which is typified by the height of the IPAAs. Some patients undergo a stapled IPAA 4 to 6 cm above the dentate line and are left with a pouch-rectal anastomosis. Others have a true IPAA, stapled distal to the dentate line. In a few patients, the anastomosis is angled, leaving rectum at the back and taking dentate line at the front. The ideal IPAA is within 1 to 2 cm from the dentate line, and is straight across, parallel to the pelvic floor.

Handsewn IPAA : Although the majority of IPAA currently performed are stapled, handsewn IPAA are still done after pouch redo, and in cases where ATZ dysplasia demands it. Stripping the mucosa from the upper anus and low rectum is done transanally, and in patients with ulcerative colitis whose mucosa is chronically inflamed, this can be technically difficult. Raising the mucosa on saline can facilitate the dissection and performing the procedure in strips rather than attempting a circumferential dissection is also easier. If the dissection begins at the dentate line and continues proximally for 2 cm, the entire ATZ will be removed in 90% of cases. In practice, the need to avoid damaging anal muscle and the difficulty of dissection means that islands of ATZ mucosa may remain after the stripping. The literature shows that ATZ dysplasia and cancer can happen in handsewn anastomosis, although this is significantly less common than after stapled anastomoses. ^{9 10}

Pouchoscopy

Another critical aspect of technique concerning IPAA is the endoscopic examination and biopsy of the ATZ. The diagnosis of dysplasia is made on histology and to obtain histology an adequate biopsy is needed. This can be difficult in patients with an IPAA, where there may be an anal stenosis, where the perianal skin may be excoriated by chronic fecal seepage, and where endoscopic skill and patience are lacking. Just getting the biopsy forceps to impact directly on the anal epithelium is tricky and to do this in each of the four quadrants in such a way to produce a decent amount of tissue requires experience. Use of a flexible scope such as a gastroscope or a pediatric colonoscope may allow retroflexion within the pouch and a more effective biopsy of the ATZ from above. Biopsy is even more difficult in patients with a handsewn anastomosis, where there is often a stenosis, and the seepage makes pouchoscopy painful. There is therefore a tendency for undersampling of the ATZ in pouch patients. The literature is interpreted with this factor in mind.

Dysplasia

Cellular dysplasia is simply defined as disordered cell growth, representing loss of growth regulation at a genetic level. There is a progression in the severity of dysplasia from mild to severe, reflecting progressive loss of growth control due to increasing tumorigenic genetic changes. Low grade (mild) dysplasia is early in the progression to cancer and treatment can be conservative. High grade (severe) dysplasia is much closer to malignancy and treatment should be without delay. Some patients have a report describing changes that are “indefinite for dysplasia,” which reflects the subjectivity of interpretation of the material and a possible effect of active mucosal inflammation that resembles dysplasia. The evolution of “indefinite” for dysplasia to actual dysplasia is possible but uncommon. ¹¹ However, careful follow-up is needed. Patients with chronic ulcerative colitis develop mucosal dysplasia due either to the chronic inflammation accelerating the instability of the epithelium, or to the development of adenomas which tend to be made less obvious by the chronic inflammation. Prevention of cancer in patients with chronic, burnt out colitis requires careful surveillance and ultimately prophylactic colectomy. In such patients the ATZ and residual rectal epithelium are at risk for dysplasia, either developing after the surgery or present at the time of the operation. This is a risk factor for cancer in the ATZ.

The Literature

Incidence

The question of whether routine surveillance of the pouch and/or ATZ is needed in patients originally with ulcerative colitis was answered by Lightner et al, who reported on 9,398 pouchoscopies in 3,672 patients. ¹² Most were diagnostic. There was low grade dysplasia in the ATZ in 7 patients (0.07%), no patient had high grade dysplasia, and 6 had carcinoma (4 in the ATZ and 2 in the pouch). Routine surveillance was not recommended. Another study by Block et al followed 56 out of 629 patients who had some form of neoplasia in the colectomy specimen. ¹³ At a median follow-up of 19 years (range 1–29 years) one patient showed low grade dysplasia in their pouch and 19/20 patients (2 pathologists) had biopsies that were indefinite for dysplasia. Only 2/4 patients had readings of indefinite for dysplasia in the ATZ. There were no cases of high grade dysplasia or cancer. ¹²

Risk Factors

Multiple studies have addressed the incidence of and risk factors for ATZ/rectal mucosal dysplasia after an IPAA. The largest of these was by Kariv et al from the Cleveland Clinic, who described the relationship between preoperative colorectal neoplasia and pouch neoplasia in 3,202 patients followed for 10 years. ⁹ In this study, and others, “neoplasia” is synonymous with “dysplasia” although “dysplasia” is more a histologic, microscopic term applicable to cells while “neoplasia” applies to lesions. The paper also combines “pouch” and “ATZ” neoplasia under the term “pouch neoplasia” which is unfortunate. The authors found neoplasia in 38/3,202 patients (1.19%). Eleven patients had cancer, with 10 of these in the ATZ and 1 in the pouch. Three patients had squamous cell cancer of the ATZ, and 23 had ATZ dysplasia. A preoperative diagnosis of ulcerative colitis-associated cancer or dysplasia was the chief risk factor for ATZ dysplasia and cancer, with adjusted hazard ratios of 13.43 (95% confidence interval [CI] 3.96–45.53; p  < 0.001) and 3.62 (95% CI 1.59–8.23; p  = 0.002), respectively.

Horio et al reported on 1,970 patients who underwent IPAA for ulcerative colitis, and concentrated on ATZ cancers. There were 220 patients who had colorectal cancer before or at the time of IPAA and 14 of these patients developed ATZ cancers. The presence of multiple cancers and rectal cancers were significant risk factors for ATZ cancer (odds ratios 8.79 and 6.48, respectively). ¹⁰

Prevention

Patients with a colorectal cancer or dysplasia at the time of proctocolectomy are at high risk of ATZ dysplasia during the course of their follow-up. Before their surgery, thorough inspection and biopsy of the rectum including the ATZ will establish a baseline. Patients with rectal and ATZ dysplasia preoperatively should routinely have a mucosectomy and handsewn S-pouch at the time of surgery. Patients whose rectum is clear of dysplasia may have either a handsewn or a stapled anastomosis, although the length of retained ATZ should be no greater than 2 cm.

Treatment and Follow-Up

Follow-up of ATZ dysplasia has been reported by only a few studies. In 1994 an early study by Ziv et al showed that low grade dysplasia was found in 8/254 patients at an average of 2.3 years after surgery. On repeat biopsy only 2 of the 8 still had dysplasia and underwent mucosectomy. ¹⁴ Perhaps the most reliable of the surveillance studies had a minimum follow-up of 10 years and found ATZ dysplasia in 8 patients, from 4 to 123 months after surgery. ¹⁵ The risk of dysplasia was significantly associated with cancer or dysplasia in the colectomy specimen. Two patients had high grade dysplasia and 6 patients low grade dysplasia. One patient with high grade dysplasia was rendered negative by a partial mucosectomy while the other was negative at repeat biopsy. Two patients with low grade dysplasia on repeat biopsies had mucosectomy and pouch advancement. The other 4 were dysplasia-free for a median of 119 months.

Mucosectomy and Pouch Advancement

Mucosectomy is a difficult operation that requires an effective removal of the mucosa from the dentate line to the anastomosis, mobilizing the lower part of the pouch, and then creating a pouch-anal anastomosis to the dentate line. Sometimes pouch mobilization is easy but if there has been an anastomotic leak or a perianal fistula the adhesions are difficult to deal with without damaging the pouch. In addition, the “blunt” end of the J-pouch may not reach down to the dentate line through the narrow anal canal. ¹⁶ The longer the distance from the ATZ to the dentate line, the more difficult the transanal approach, and a 4-cm gap is better revised by transabdominal IPAA redo. Mucosectomy and pouch advancement is attempted on a “case-by-case” basis and should be reserved for those experienced with transanal pouch surgery. Because most patients with ATZ dysplasia have cancer or dysplasia at the time of proctocolectomy, they should have a primary mucosectomy and handsewn anastomosis (and an S-pouch), ¹⁶ or if a J-pouch is performed with a stapled anastomosis, at least leave no more than 2 cm and preferably 1 cm of ATZ.

Summary

Recently, a panel of experts in the histology, medical, and surgical management of IPAA published guidelines on the management of pouch neoplasia. ¹⁷ This author was part of that group and contributed to the guidelines. A personal plan for surveying and managing ATZ dysplasia is presented in Fig. 1 as an algorithm.

Fig. 1.

An algorithm for the screening, surveillance, and management of anal transition zone (ATZ) dysplasia.