Images of Early Rectal Cancer

Introduction

Early rectal cancer is enigmatic. The rectum is a hotspot for colonic neoplasia, and in Leeds, United Kingdom, the prevalence of rectal neoplasia is second only to that of neoplasia in the sigmoid colon (Table [1]). Despite this, the rectum is a site at which many early cancers are missed. What is the explanation for this?

Table 1 The prevalence of neoplasia in various sites in the colon (Leeds data) Neoplastic lesions (n) Average width of lesions (mm) Rectum 119 15.6 Sigmoid colon 186 12.0 Descending colon 82 8.3 Splenic flexure (left colic flexure) 33 10.3 Transverse colon 115 7.3 Hepatic flexure (right colic flexure) 25 10.8 Ascending colon 76 10.3 Cecum 101 13.8 Total 733 11.5

The rectum is the first site visualized during colonoscopy, and it is also the last part of the colon that is seen before the endoscope is withdrawn. In addition, neoplastic lesions in the rectum are on average broader than those found elsewhere in the colon (Table [1]). In the rectum, most adenomas are either flat or sessile, rather than polypoid. Of the 119 neoplastic lesions in the rectum listed in Table [1], 80 % were sessile or flat and only 20 % were polypoid.

Another possible reason for the rectum being a hotspot for missed cancers might be that lesions arising in this part of the large bowel are more aggressive than cancers found elsewhere. Again in data from Leeds, the mean size of 37 advanced colonic cancers (T2 or worse) was 43 mm, compared with 34 mm in 16 advanced rectal cancers. Similarly, early cancers (high-grade dysplasia or T1 carcinomas) arising in the colon may be larger than the corresponding early cancers in the rectum (17 mm vs. 13 mm) (Figure [1]). However, these differences are not yet statistically significant. The average risk of a colonic adenoma harbouring high-grade dysplasia or cancer was 4.5 % (25 of 550), compared with a 17 % risk (15 of 89) in rectal adenomas (chi-squared test P = 0.0003; Leeds data).

Figure 1 The range of sizes (mm) of advanced cancers (stage T2 or higher) compared with the sizes of early cancers (high-grade dysplasia or stage T1) in the colon and rectum (Leeds data).

Why, then, is rectal neoplasia more likely to be flat or sessile rather than polypoid? Is it simply that the fecal stream exerts a pulling force on lesions, leading to the formation of a peduncle, while flat or sessile lesions develop in areas with nonflowing feces? If the growth pattern were simply the result of mechanical forces, one would expect the same mutations to be present in flat and polypoid lesions. However, most series have reported a lower prevalence of K-ras mutations in flat neoplastic lesions than in polypoid ones.

There is evidence that adenomas smaller than 5 mm are not yet ”committed” to a particular growth pattern. In a recent study, Morita et al. [1] followed 62 flat or depressed lesions smaller than 5 mm over a 2-year period and reported that 40 % of these lesions became polypoid as they grew larger. Lesions larger than 5 mm in diameter are more likely to have become committed to a flat or polypoid growth pattern. The 8 mm lesion shown in Figure [2 a] was observed, but not sampled, at colonoscopy in 1997. Three years later, the lesion was photographed again (Figure [2 b]) before a mucosectomy was carried out.

Figure 2 This lesion was observed, but not sampled, at colonoscopy in 1997. b The same lesion 3 years later, before a mucosectomy was carried out.

This picture gallery illustrates some types of early malignant neoplasia that may be encountered in the rectum. The small, smooth, plaque-like type of adenoma seen in Case 3 is much more difficult to detect. In spite of their small size, these lesions almost always harbor areas of high-grade dysplasia or early carcinoma. As seen in Case 1, a central depression is a useful endoscopic indicator that the lesion may be an early cancer. However, this sign is unreliable in lesions smaller than 5 mm. Case 2 probably represents a more advanced stage of a precursor lesion with a central depression. In this case, the lesion is spreading laterally and penetrating deep into the submucosa, infiltrating vessels in its path.

One is sometimes requested to remove suspicious rectal lesions that have yielded inconclusive results after surface biopsies. In these cases, it is usually possible to guess the diagnosis even without histology. Small but advanced rectal carcinomas are usually ”chunky” lesions with a fleshy or nodular rim and will feel firm on palpation (Cases 6 and 7). In spite of this, the conclusive diagnosis provided by a partial resection may be necessary before an elderly frail patient is subjected to an anterior resection. As demonstrated in Case 5, even partial removal can be hazardous, as the underlying desmoplastic reaction can make elevation with submucosal saline impossible. To achieve this, it can be helpful to apply an Endoloop around the base of pedunculated lesions that are suspected of harboring invasive cancer (Case 7). Even if surface biopsies have failed to confirm invasive cancer, this technique allows sufficient tissue to be removed, with a minimum risk of perforation.
B. J. Rembacken

Reference

• 1 Morita T, Tomita N, Ohue M. et al . Molecular analysis of diminutive, flat, depressed colorectal lesions: are they precursors of polypoid adenoma or early stage carcinoma?.  Gastrointest Endosc. 2002;  56 663-671