Keywords

1 Introduction

Previously, 25 % of the patients with ulcerative colitis (UC) underwent colectomy for medically refractory UC or UC-associated neoplasia. In recent years, however, elective colectomy rates in UC patients have decreased, as the efficacy of antitumor necrosis factor (TNF) antibodies has been confirmed and their use has increased significantly [1]. Nonetheless, patients with long-standing UC continue to be at higher risk of colorectal cancer (CRC) than the general population and therefore often require surgery for UC-associated neoplasia. The early detection of cancers in UC patients means that the tumor can be treated at an earlier stage, which corresponds to a better prognosis [2].

Total proctocolectomy (PC) with ileal pouch-anal anastomosis (IPAA) is the procedure of choice for patients undergoing elective surgery for UC and is the standard procedure for those with UC-associated CRC. Clearly, there are fewer indications for IPAA in the treatment of UC-associated neoplasia than for medically refractory UC, given that in the treatment of CRC, the primary aim is to improve the oncologic prognosis, which takes precedence over the functional prognosis.

Recent technical progress in gastroenterological endoscopy has allowed the more accurate detection of dysplasia, which has led to renewed debate regarding the utility of endoscopic surveillance or resection for dysplastic lesions. The larger proportion of younger patients with UC-associated cancer than sporadic cancer in individuals with a normal mucosa has also led to a reconsideration of surgery for UC-associated cancer with respect to operative indication, extent of resection, and pros and cons of rectal mucosectomy. In this chapter, we discuss the various options for the surgical treatment of UC-associated cancer while taking into account the related controversies.

2 Surgical Indications for UC-Associated Neoplasia

Although the choice of treatment is influenced by the site and stage of the cancer, the detection of CRC by biopsy is an absolute indication for surgery in UC patients. UC-associated cancer is characterized by its broad range of macroscopic appearances, its tendency to spread diffusely and invasively, and its poorly marginated lesions; these features distinguish it from sporadic CRC [3]. Another important difference between sporadic and UC-related colonic neoplasia is that in the latter the entire colonic mucosa is at risk for neoplastic transformation, which can be multifocal [4]. The underlying mechanism is the inflammation → dysplasia → carcinoma sequence in patients with long-standing UC-related inflammation of the colon and rectum [5–7]. In addition, so-called field effects of multiple epigenetic alterations, including methylation, have been shown in both the neoplastic and non-dysplastic mucosa of UC patients [8–10]. Field effects have been attributed to the constant reepithelialization of the ulcerated and chronically inflamed colonic mucosa by abnormal cell clones that arise during healing and subsequently expand [11] and to changes in the local environment, such as oxidative stress and an altered bacterial flora, both of which can give rise to cellular mutations [12].

Dysplasia, the earliest histologic manifestation of neoplastic transformation, is defined as an unequivocal neoplastic change in the colonic epithelium without invasion into the lamina propria. Dysplasia is grouped into five main categories: low-grade dysplasia (LGD), high-grade dysplasia (HGD), dysplasia-associated lesion or mass (DALM), adenoma-like mass (ALM), and adenoma-like DALM. The appearance of any of these is an early clinical alert to the development of carcinogenesis because the probability of a coexistent carcinoma is relatively high [13, 14]. However, some researchers maintain that the presence of dysplasia is associated with a low risk of unsuspected cancer at the time of colectomy [15]. Accordingly, the decision-making process for patients with UC who are considering intensive surveillance vs. surgical intervention after a diagnosis of dysplasia is a controversial one.

2.1 High-Grade Dysplasia

Over 30 % of the patients with HGD will develop cancer during subsequent surveillances; however, there are no reported instances in which HGD was subsequently downgraded to negative [7, 16, 17]. Among patients with HGD who underwent prophylactic colectomy immediately after diagnosis, about 40 % had cancer in the resected sample [7, 18–20]. Moreover, the HGD was detected at a colonic site distant from that of the synchronously detected cancer [19, 21]. Therefore, a diagnosis of HGD is an absolute indication for surgical resection [22–24].

2.2 Low-Grade Dysplasia

Whether patients with flat LGD should undergo surgery continues to be debated. Among patients with LGD with subsequent surveillance colonoscopy, 5–50 % subsequently had the diagnosis downgraded to indefinite or negative [17, 18, 20, 25]. However, in approximately 30 % of these patients, there was eventual progression to HGD, DALM, or cancer [7, 16–18, 20, 22, 25]. Cancer was eventually detected in approximately 10 % of the patients with LGD and subsequent surveillances [16, 18, 20, 22, 25–27]. Zisman et al. [28] showed that patients with three or more biopsies demonstrating LGD at a single colonoscopy had an increased risk of progression to advanced neoplasia [relative risk (RR) = 5.8; 95 % confidence interval (CI): 1.29–26.04].

Analyses of the outcomes of patients who underwent colonoscopic surveillance for 10 years following the detection of LGD did not show a statistically significant difference from the outcomes of control patients. This conclusion is supported by histopathologic reviews, which have demonstrated the unreliability of LGD diagnosis [29]. Thus, the current opinion is that a diagnosis of LGD does not justify prophylactic colectomy.

However, there are opposing views regarding whether only flat lesions should be followed up endoscopically, because dysplastic lesions, which eventually progress to invasive cancer, cannot be consistently and reliably detected through successive surveillances [30]. Bernstein et al. [22] analyzed ten prospective studies (1225 patients), in which the lesions of 16–29 % of the patients progressed from untreated LGD to DALM, HGD, or cancer. In their retrospective study, Kiran et al. [31] showed that the rate of the risk of cancer in postoperative pathologic findings was 3 % even if preoperative biopsies demonstrated LGD. In general, because advanced neoplasia can be found in association with dysplastic changes of any grade, patients with confirmed dysplasia of any grade should undergo colectomy [13, 22, 26, 31, 32].

According to the current American College of Gastroenterology guideline [33], surgery should be promptly considered in patients with flat LGD to prevent progression to a higher grade of neoplasia. In the Medical Position Statement of the American Gastroenterological Association [34, 35], multifocal LGD is a strong indication for colectomy. In addition, although controversial, there is evidence to suggest that patients with flat, unifocal, LGD should also be considered for colectomy [36, 37].

Primary sclerosing cholangitis (PSC) is typically associated with inflammatory bowel disease (IBD), particularly UC. PSC-IBD patients are at an increased risk of colorectal neoplasia [38]. In one-third of PSC-UC patients, LGD will progress to HGD/CRC. Venkatesh et al. [39] evaluated ten PSC-UC patients with LGD who underwent surveillance colonoscopy. In three (30 %) patients, LGD progressed to raised HGD over a mean follow-up of 13 ± 11 months, and HGD occurred more frequently within the first year of the initial detection of LGD (23.5 per 100 patient-years of follow-up). Therefore, PSC-UC patients with LGD should be closely and carefully followed.

2.3 Dysplasia-Associated Lesion or Mass

It was initially suggested that any dysplasia found in association with a DALM, in particular with a polypoid mass, indicates a high likelihood of the presence of synchronous or metachronous neoplasia [40]. Bernstein et al. [22] evaluated 40 patients with DALM from ten prospective studies (1225 patients); 17 (43 %) of the patients already had cancer at immediate colectomy. In another report of patients with HGD in DALM who were followed for over 5 years, none of the patients had carcinoma, either in surveillance biopsies or in resection specimens [41]. Thus, the latter authors concluded that the presence of HGD in DALMs does not warrant colectomy with continued close observation.

In a series of 348 patients from 1984 to 2007, Kiran et al. [31] demonstrated that those with a preoperatively detected DALM had a significantly higher risk of cancer than patients with flat dysplasia (25 % vs. 8 %; P < 0.001). They also found that the risk of cancer was not significantly higher in LGD with DALM than in flat LGD (7 % vs. 2 %; P = 0.3), but the risk of cancer or HGD was threefold higher (29 % vs. 9 %; P = 0.015).

Recent studies broadly separated the raised (endoscopically visible) dysplastic lesions in IBD into those resembling non-IBD-related sporadic adenomas (adenoma-like) and those that do not resemble adenomas (non-adenoma-like) [42–44]. Biopsy specimens of non-adenoma-like DALMs may contain the surface of an invasive adenocarcinoma, which is regarded as endoscopically unresectable [40]. Thus, patients with UC and an endoscopically unresectable, non-adenoma-like DALM, regardless of the grade of dysplasia detected on biopsy analysis, should undergo colectomy, because of the high association of these lesions with metachronous or synchronous carcinoma [35, 45].

2.4 Stricture

A colonic stricture is regarded as a manifestation of chronic UC, although carcinoma may occur at the site of a stricture [46], and a fibrous stricture as an indication for surgery, owing to the possibility of malignant degeneration [47]. Gumaste et al. [14] investigated 1156 UC patients; in this group, 17 of the 70 strictures (24 %) proved to harbor a malignancy. In addition, they described three features that distinguish a malignant from a benign stricture: (1) appearance late in the course of UC (61 % probability of malignancy in strictures that develop after 20 years of disease vs. 0 % in those occurring before 10 years); (2) location proximal to the splenic flexure (86 % vs. 47 %, 10 %, and 0 % when the stricture is in the sigmoid colon, rectum, and splenic flexure and descending colon, respectively); and (3) symptomatic large bowel obstruction (100 % probability of malignancy vs. 14 % in the absence of obstruction or constipation) [14]. Lashner et al. [48] described 15 patients with UC-related strictures identified by air-contrast barium enema or on colonoscopy; within this group, 11 had dysplasia and two had cancer. Thus, a stricture should be considered as a strong risk factor for cancer, requiring intensive colonoscopic surveillance. If dysplasia is discovered or if the stricture cannot be adequately biopsied, then surgical treatment should be considered [48].

3 Surgical Procedure

The surgical procedure for neoplasia in patients with UC varies, ranging from colonoscopic resection to total PC. Although PC with permanent ileostomy and restorative PC with mucosectomy are the only surgical procedures that will reliably eliminate the cancer risk in UC, the risk of subsequent morbidity and impaired anal function is not small. The choice of surgical treatment is influenced by the site and stage of the neoplasia, the functional state of the rectum, the presence of multifocal lesions, the patient’s age, and the duration of UC [49] (Fig. 6.1).

Fig. 6.1
figure 1

Surgical options for ulcerative colitis patients with dysplasia found on colonoscopy. IPAA ileal pouch-anal anastomosis, IRA ileorectal anastomosis, TAC total abdominal colectomy, TPC total proctocolectomy

Full size image

3.1 Abdominoperineal Total PC

Abdominoperineal total PC is the most definitive treatment for the eradication of undiagnosed synchronous dysplasias and/or carcinomas and the prevention of subsequent metachronous lesions in UC. It allows the resection of synchronous colonic and rectal dysplasia/cancer and avoids the development of metachronous colonic and rectal lesions. It also obviates the need for further colonoscopic surveillance. However, patients must accept a permanent stoma and the postoperative complications are significant, including urinary and sexual dysfunction or a nonhealing perineal wound.

Although this procedure is not an attractive option for patients with comorbidities or for those wanting to preserve anal function, it is indicated for patients with advanced rectal and anal canal cancer or for patients with poor anal sphincter function, such as the older postpartum female. These patients will also require an end ileostomy or a continent Kock ileal pouch.

3.2 Segmental/Partial Resection

There is limited debate regarding segmental colectomy in the treatment of lesions in patients with UC in long-standing remission. The indication for the procedures considers the difficulty of further PC and IPAA after lymphadenectomy or adhesions at the target surgical site.

Segmental colectomy is a short operative procedure and maintains continence, but most patients will later require not only further medication but also excision and colectomy and ileostomy. Moreover, a right-sided colo-anal anastomosis is unsuitable for the treatment of left-sided UC [50]. Thus, in general, partial resection of the colon should be avoided because of the high frequency of occult carcinomas and multifocal carcinogenesis. Schwarz et al. [51] reported the case of a patient who underwent a left hemicolectomy and a mucosal proctectomy but then had a macroscopic recurrence of the colitis within 2 months postoperatively and eventually required excision of the remaining colon and an end ileostomy. This suggests that the risk of UC relapse in the residual colon must be taken into account, even if the right side of the colon seems to be in remission.

Patients with UC-associated CRC have a twofold higher mortality than patients with sporadic CRC [52]. Whether segmental or partial resection is the optimal procedure for UC patients with sporadic cancer remains questionable.

3.3 Total Abdominal Colectomy with Ileorectal Anastomosis

The cumulative probability of total abdominal colectomy with ileorectal anastomosis (IRA) after 10 years of UC is about 50 %. Total excision was performed following the detection of dysplasia in 5.9 % of these patients [53] and is no longer an acceptable procedure for patients with UC. Although UC is generally considered to always involve the rectum and in some patients also the more proximal portions of the colon, albeit in a diffuse and non-segmental fashion, rectal sparing has been documented [53–57]. If the colitis is totally quiescent or shows rectal sparing, total abdominal colectomy with IRA may be an option for UC patients with a single cancer in the colon. However, since most colectomy specimens with an absence of macroscopic activity show histologic features of chronicity or activity [58], these patients should be monitored for a relapse of proctitis.

The indication of low anterior resection for patients with quiescent UC and rectal cancer or dysplasia should be considered very carefully, because further PC and IPAA would be difficult after this procedure. In addition, patients undergoing total abdominal colectomy with IRA and treated postoperatively with immunomodulators or biologics have a risk of a relapse of inflammation in the residual rectum. Thus, the prognosis, and especially the risk of cancer in the residual rectum, after IRA in patients with UC-associated cancer is a concern. Among patients who received an IRA, regardless of the indication, the estimated cumulative cancer risk after a disease duration of 20 years is 2.1–20.0 % [59–63]. The high long-term risk of cancer after total abdominal colectomy with IRA suggests that this procedure is an interim solution in younger patients. However, IRA with a close follow-up still plays a major role in treating UC patients because it is an easier surgical procedure than IPAA, has excellent functional results, shorter hospitalization, and, importantly, fewer severe complications, unlike in IRA [61].

A particular indication of IRA for UC-associated cancer is a non-obstructing tumor located above the pelvic floor with remote metastases in remission. In the case of an advanced tumor causing obstruction, then a primary colectomy and IRA with a covering ileostomy is advisable.

3.4 Subtotal Colectomy

Subtotal colectomy (STC) with end ileostomy and rectal stump pouch are less ideal options because of the retained rectum, which poses a continued cancer risk. However, STC was shown to be a safe treatment option for patients satisfied with an ileostomy or for those with comorbidities that make them ineligible for other procedures or who do not choose to undergo later pelvic pouch surgery. Nonetheless, the potential for the proliferation of residual dysplastic cells or malignant change within the rectal stump in patients who have undergone STC with rectal stump preservation for UC-associated CRC is of serious concern. In addition, whether the potential for malignancy in the rectal stump of STC patients with UC-associated CRC outweighs the morbidity associated with complete proctectomy is difficult to determine. The rate of cancer occurrence later on was shown to be low (1.4 %) in one study [62] but eightfold higher in another [64]. PSC and disease duration until STC were shown to be significant risk factors for rectal stump cancer in a closed rectal stump after STC [65]. Thus, considering the risk of rectal cancer, the low success rate of long-term rectal preservation, and the safety of surgery, a more aggressive approach to early complete proctectomy is recommended in this situation. If this is not possible, patients treated with STC should be followed with close endoscopic surveillance of the closed rectal stump.

3.5 Total Proctocolectomy with IPAA

IPAA by ileal J pouch, first described in 1980, is now the gold standard surgical procedure for UC refractory to medical treatment [66]. The long-term quality of life of these patients after this procedure is excellent and the level of fecal continence is satisfactory [67–69]. However, this procedure also has several technical difficulties such as mesenteric lengthening of the pouch [70, 71] and mucosal proctectomy [72]. For surgeons, extensive experience is required to obtain acceptable results [73]. Patients with refractory UC who suffer complications after PC have a poor quality of life [74]. Thus, a double-stapled anastomosis without mucosal proctectomy is the preferred procedure, as there are fewer anastomotic complications and superior rectal continence is achieved; however, a cuff of rectal mucosa is retained, which is the main concern as well as the main argument of opponents of the double-stapling technique. Whether with or without mucosal proctectomy, IPAA is indicated for any colonic or rectal lesion in the surgically fit patient who has unifocal or multifocal dysplasia and refuses a stoma. Relative contraindications of IPAA for UC-associated neoplasias are preoperative incontinence/poor anal sphincter tonus, severe backwash ileitis suggesting Crohn’s disease, and very low rectal or anal dysplasia that threatens the sphincters.

However, the use of a stapled anastomosis without mucosal proctectomy in patients with UC-related dysplasia or cancer remains controversial because of the risk of developing synchronous or metachronous neoplasias in the retained anal transitional zone (ATZ) mucosa. Although an association has yet to be reported between dysplasia and any the following: age, sex, preoperative length of disease, use of a double- vs. single-staple technique, or anastomotic distance from the dentate line [75], the risk of cancer can be reduced by ensuring that the minimal length of rectal columnar mucosa is retained. It is therefore recommended that, in carrying out a stapled IPAA, the anastomosis is performed at the anorectal junction, about 1–1.5 cm above the dentate line, because of the deterioration in anorectal function [76]. Additionally, this procedure is indicated for patients with UC and right-sided colon cancers who require lymph node dissection along the superior mesenteric vein and excision of the marginal arcade of the ileocolic artery, because an insufficient extension of the ileal pouch to the anus precludes a hand-sewn IPAA with mucosectomy.

Stapled IPAA has also been advocated in patients with UC associated with coexisting neoplasia [77, 78]. In these cases, long-term surveillance to monitor dysplasia is recommended; if repeat biopsy confirms persistent dysplasia, ATZ excision with a neoileal pouch-anal anastomosis should be performed [78]. However, restorative PC with mucosectomy does not necessarily eliminate the risks, as after this procedure cancer can occur in a residual ATZ [79, 80]. Thus, in patients with long-standing ileal pouches even after mucosectomy of ATZ, and especially in cases in which dysplasia or cancer is detected in the PC specimen, routine long-term endoscopic surveillance is recommended.

There are many reported cases in the indexed medical literature of carcinoma arising after stapled IPAA for UC [79]. In some studies, the incidence of dysplasia in the ATZ at the time of total colectomy was 2.5–5 %, and duration of UC and patient age at colectomy were significant risk factors [81, 82]. In these cases, mucosal proctectomy is the definitive procedure for patients with preoperatively detected dysplasia in the ATZ.

The incidence of dysplasia after stapled IPAA is 3.0–4.5 % [75, 76, 83]. The development of cancer in the ATZ after stapled IPAA without mucosectomy has been reported [76, 84, 85] and was shown to be significantly associated with a preoperative pathologic diagnosis of UC with concurrent dysplasia or cancer [75]. Based on these data, mucosal proctectomy and hand-sewn IPAA are strongly recommended for patients with neoplasia, especially those with cancer or HGD outside the ATZ [75, 86].

For the reasons stated above, in young patients with UC-associated cancer, mucosal proctectomy with IPAA is recommended, whereas for older patients, particularly those with lower rectal cancer who will accept a permanent stoma, total PC may be proposed. Patients older than 50 years have a significantly higher rate of concurrent dysplasia and malignant degeneration than younger patients, probably because of a longer duration of disease [87]. In these cases, restorative PC with mucosal proctectomy may reduce this risk by eliminating all of the colorectal mucosa.

Branco et al. [88] reported a case in which adenocarcinoma arose in an ileal pouch after IPAA with mucosal proctectomy performed using a cavitron ultrasonic surgical aspirator (Excel, Covidien, Boulder, CO) for UC. This method was introduced to simplify and optimize IPAA with mucosectomy and has been shown to shorten the operative time and reduce blood loss [89]. Its use, however, may increase the number of pathology specimens made uninterpretable on account of tissue ablation. Another ultrasonically activated scalpel (Harmonic; Ethicon Endo Surgery, Johnson & Johnson Medical SPA, Somerville, NJ) also shortened the operative time, decreased blood loss, and was shown to be useful for restorative PC [72]. There has been no report of adenocarcinoma arising in an ileal pouch after mucosectomy performed using this device.

3.6 Endoscopic Resection

The ALMs seen in UC patients are similar to those observed in non-UC patients that have been treated by standard polypectomy. This method is associated with little risk of subsequent malignancy on follow-up [42, 44, 90, 91].

An accurate pathologic diagnosis is very important for distinguishing among the different pathologic entities, given the different therapeutic consequences, such as endoscopic polypectomy for ALM and potential PC for DALM. New and emerging endoscopic imaging techniques, such as chromoendoscopy, magnification endoscopy, and confocal laser endomicroscopy, provide a more accurate diagnosis. Endoscopic resection of an ALM allows confirmation of the biopsy-based adenoma diagnosis and the exclusion of a DALM [91]. However, the endoscopic resectability of a lesion is more important than whether it is an ALM or a DALM [92]. The basic rules for the detection of neoplasia [93] (Table 6.1) should be taken into account and applied in accordance with international guidelines [95–97].

Table 6.1 Basic rules for detecting neoplasia in patients with UC
Full size table

Only a few studies have examined the clinical outcomes of DALMs resembling ALMs that are removed with endoscopic polypectomy, but the safety and efficacy of endoscopic resection have been evaluated [93, 94, 98]. Since DALMs, in particular those with a polypoid mass, are an indicator of a high likelihood of the presence of synchronous or metachronous neoplasia, endoscopic resection is not appropriate [22, 40].

3.7 Perianal Resection Line for Rectal and Anal Canal Neoplasia

In an IPAA performed in a patient with UC without colorectal neoplasia (Fig. 6.2a.), the mesorectum is resected on the inside, close to the rectal wall, to preserve autonomic nerve function. In patients with UC-associated low rectal neoplasia, total mesorectal excision is required in performing an IPAA (Fig. 6.2b). The choice of operative procedure depends on the depth of the tumor and its distance from the anal verge (Table 6.2). Regarding IPAA with intersphincteric resection, there are no data on postoperative anal function from a large number of cases. Thus, consensus on this procedure is lacking. Depending on curability, however, it should at most be confined to a partial intersphincteric resection (Fig. 6.2b).

Fig. 6.2
figure 2

Perianal resection line. Ileal pouch-anal anastomosis (IPAA) for ulcerative colitis (UC) (a), UC-associated low rectal neoplasia (b). TME total mesorectal excision, LAR low anterior resection, IAA ileoanal anastomosis, ISR intersphincteric resection, APC abdominoperineal total proctocolectomy

Full size image
Table 6.2 Indication and procedure for mesorectal excision in UC patients with rectal and anal canal cancer
Full size table

4 Immunomodulators

Medical therapy for UC has advanced dramatically in the last decade, which has led to discussions of the pros and cons of immunomodulators or biologics for UC patients with malignant disease. Previous studies and guidelines showed that patients administered immunomodulators or biologics do not have a higher risk of new cancer development [99]. Anti-TNF antibodies have been linked to a risk of cancer recurrence in rheumatoid patients, thiopurine to a risk of cancer recurrence in transplant patients, and calcineurin inhibitor to a risk of hepatocellular carcinoma recurrence in liver transplant patients [100–102]. However, a meta-analysis of 74 random controlled trials found that anti-TNF therapies are not related to the short-term clinical emergence of cancer [103]. Nonetheless, a relapse residual lesion is not a rare occurrence after segmental/partial resection; thus, in these patients with advanced cancer, the restricted use of immunomodulators or biologics should be considered.

5 Prognosis

5.1 Neoplasia in an Ileal Pouch

Previously, cancer of the ileal mucosa was reported in patients who underwent a Brooke ileostomy [104–107] and in those with a Kock pouch [108], but the natural history and prognosis of pouch dysplasia or cancer are poorly understood. Although inflammation, villous atrophy, and colonic metaplasia have been observed within the mucosa of ileal pouches after IPAA, dysplasia may also develop, but the incidence is <0.02 % 20 years after IPAA [109].

In their study of pouch-related adenocarcinoma, Selvaggi et al. [64] showed a pooled cumulative incidence of 0.33 % 50 years after the diagnosis and 0.35 % 20 years after IPAA in a systematic review of the meta-analyses of the literature of pouch-related adenocarcinoma in patients with an IPAA for UC. In that study, one-third of the adenocarcinomas arose from the pouch as a whole and the remainder from the anorectal mucosa [64].

Derikx et al. [110] used the National Registry from 1991 to 2012 to identify 1200 patients with IBD and IPAA; 25 (1.83 %) developed pouch neoplasia, including 16 adenocarcinomas. The cumulative incidence of pouch neoplasia at 5, 10, 15, and 20 years was 1.0 %, 2.0 %, 3.7 %, and 6.9 % for pouch neoplasia and 0.6 %, 1.4 %, 2.1 %, and 3.3 % for pouch carcinoma [110] (Fig. 6.3). A history of colorectal neoplasia was the only risk factor associated with pouch neoplasia. Hazard ratios were 3.76 (95 % CI: 1.39–10.19) for prior dysplasia and 24.69 (95 % CI: 9.61–63.42) for prior carcinoma [110]. Another systematic review similarly concluded that neoplasia in the colectomy specimen was the strongest risk factor (odds ratio = 8.8; 95 % CI: 4.61–16.80) [64].

Fig. 6.3
figure 3

Cumulative incidences of pouch neoplasia (both carcinoma and dysplasia), pouch carcinoma, and dysplasia (From [110])

Full size image

Malignant transformation of the ileal pouch mucosa may occur even in the absence of backwash ileitis or a previous history of cancer [111, 112], as determined in biopsies from the ileal pouch mucosa obtained at least 1 year after the newly formed pouch that was influenced by fecal flow [113]. Chronic inflammation of the ileal mucosa such as occurs with preoperative backwash ileitis and postoperative pouchitis in UC has also been linked to the sequence of malignant transformation [114–117]. An abnormal lesion of the ileal pouch mucosa was shown to have a high risk of adenocarcinoma 20 years or later after the initial IPAA [118].

PSC-IBD patients are at increased risk of colorectal neoplasia [38], but the development of pouch neoplasia in PSC-UC patients following IPAA is unclear. Imam et al. [119] conducted a retrospective chart review of 65 patients with PSC and IBD who underwent colectomy with IPAA followed by pouch surveillance between 1995 and 2012. The cumulative 5-year incidence of pouch neoplasia was 5.6 % (95 % CI, 1.8–16.1 %). Based on this short-term follow-up, they concluded that a frequent surveillance of the pouch was an unnecessary practice in PSC-IBD patients. However, it is recommended that patients with these risk factors be followed by endoscopy and random biopsies for the rest of their lives. If a pouch-related adenocarcinoma is detected during these examinations, abdominoperineal excision is recommended.

5.2 Outcome of Colorectal Cancer in UC

There have been a few reports based on small series that examined the outcome of patients with UC and CRC [12, 120–123]. From a functional aspect, among cancer patients who received an IPAA, no significant differences could be found between those with UC-associated CRC and those with UC without CRC [124]. For UC-associated CRCs, as for non-colitic cancers, histologic stage, site, and mucin content of the tumor are the most important variables determining postoperative survival [49].

A 20-year follow-up study of IBD-related CRC from the Mayo Clinic compared patients with sporadic CRC with age- and sex-matched patients with IBD-related cancers. In the latter group, the tumors were more proximally located, with only 55 % distal to the splenic flexure, compared with 78 % among patients with sporadic CRC [125]. However, compared with sporadic tumors, IBD-related CRC was more often in an advanced stage and more likely to have a mucinous component [125]. Yet, no differences were found in the overall survival of patients with sporadic CRC and those with IBD-related CRC [125].

Heimann et al. [126] showed that the 5-year survival rate was significantly worse for patients with non-diploid tumors (76 % vs. 32 %). When stratified by stage, only patients with Dukes’ C lesions had a significant difference in survival for diploid vs. non-diploid tumors. Multivariate analysis showed that Dukes’ classification was the best prognostic indicator, followed by tumor differentiation and DNA ploidy. Tumor location, colloid content, number of cancers, duration of disease, and patient age and sex did not correlate with the prognosis [126].

A retrospective review of 1642 UC patients by Kavanagh et al. [127] showed that patients who undergo surgery for UC-associated CRC (n = 22) have less favorable short-term outcomes but present at a less advanced stage and have a more favorable long-term prognosis than similar patients with CRC and Crohn’s disease. The overall 5-year survival was significantly better in the UC group than in the group with Crohn’s disease (41 % vs. 29 %; P = 0.04).

Watanabe et al. [128] showed that in a group of 108,536 CRC patients, the 169 with UC-associated CRC had a poorer survival than patients with sporadic CRC (43.3 % vs. 57.4 %; P = 0.0320) for stage III disease but not for early-stage disease. The authors concluded that the detection of UC-associated CRC at an early stage results in similar postoperative outcomes as those of patients with sporadic CRC. A Danish population-based study also compared patients with UC-associated CRC (n = 279) and those with sporadic CRC (n = 71,259). Cancer stage and rates of lymph node and distant metastasis were similar between the two groups, but the overall mortality rates at 1 and 5 years after cancer diagnosis were higher in UC-associated CRC than in sporadic CRC (OR = 1.24; 95 % CI: 1.02–1.51 and OR = 1.17; 95 % CI: 1.01–1.36, respectively) [129]. Other population-based studies showed that patients diagnosed with UC-associated CRC at age <60 years had a worse outcome [130, 131], which, according to Shu et al. [131], was more pronounced in males.

5.3 Radiation/Chemotherapy

Although locally advanced rectal cancer requiring multimodality therapy is uncommon in patients with UC, the functional outcome of patients with UC-associated CRC who received adjuvant chemotherapy was shown to be very good if the appropriate surgical technique and chemotherapy protocol were selected [86].

Preoperative chemoradiotherapy (CRT) and total mesorectal excision with or without intersphincteric excision are the current treatment choices for patients with lower rectal cancer. This approach was shown to optimize oncologic outcome and to maintain anorectal function [132]. By contrast, pelvic radiation administered prior to IPAA is associated with poor pouch outcomes for UC patients [86, 133, 134]. In fact, external beam radiation to treat cancer is problematic in UC patients, especially because the small bowel has a lower tolerance than the large bowel [135]. Thus, whether adjuvant CRT increases postoperative complications remains controversial [133, 136].

In patients with cancer located in the ATZ and close to the internal sphincter, restorative PC and partial intersphincteric resection may be indicated [136], whereas preoperative CRT has a negative impact on sphincter function [136–138]. A recent report identified preoperative CRT as a risk factor for impaired anal function after intersphincteric resection [139]. CRT followed by IPAA and partial intersphincteric resection may be even more destructive in terms of postoperative anal function, with several studies showing better outcomes than colonic J pouch reconstruction for lower rectal cancer [140–142]. Previous reports demonstrated a high tolerance of preoperative CRT and pouch surgery with minimum intersphincteric resection [142, 143]. Overall, because prognosis seems to be related to cancer stage, the oncologic benefits and pouch functional outcomes should be carefully balanced before pelvic radiation prior to IPAA is considered [134].