Flat adenomas and flat adenocarcinomas of the colorectal mucosa in Japanese and Swedish patients. Comparative histologic study

Two histologic compartments in nonpolypoid conventional colon adenomas

Two intertwined compartments coexisting in nonpolypoid conventional (i.e. tubular or villous) adenomas are highlighted in this review: one built of dysplastic tissue on top and the other portraying crypts with irregular, corrupted shapes, albeit lined with normal epithelium, below. The latter compartment has remained unattended in the literature. Recently, however, the histologic characteristics of the nondysplastic compartment in nonpolypoid conventional adenomas were closely examined, and some of its biological attributes were unveiled. Studies with the proliferation marker ki67 showed that the crypts with irregular, corrupted shapes in the nondysplastic compartment displayed haphazardly distributed proliferating cell-domains. Given that the proliferating cells are generated by stem cells, the relocation of proliferating cell-domains in those crypts seems to be the result of a reorganization of the stem cells within the crypts. The abnormal distribution of proliferating cells, the finding of p53-upregulated cells, and of crypts in asymmetric fission suggest that the crypts in that compartment are histo-biologically altered, probably somatically mutated. This new information might contribute to unravel the riddle of crypto-histogenesis of nonpolypoid conventional adenomas of the colon. More research along these lines is necessary, before the biology of the crypts in the nondysplastic compartment can be fully translated into molecular terms.

The unique pathology of nonpolypoid colorectal neoplasia in IBD

Patients with inflammatory bowel disease may develop dysplasia in the cryptal epithelium, polypoid neoplasias, and nonpolypoid (flat) adenomas, lesions at risk to proceed to colorectal carcinoma. The onset of invasion in nonpolypoid adenomas may occur without changes in the shape or the size of the lesion. In experimental animals, some colonotropic carcinogens induce polypoid and nonpolypoid neoplasias and others induce polypoid neoplasias exclusively. Some of the biologic attributes of nonpolypoid adenomas in humans can be demonstrated in laboratory animals.

Risk factors associated with colorectal flat adenoma detection

OBJECTIVES

Colorectal flat adenomas have been associated with a higher risk of colorectal malignancy. We describe demographic characteristics and endoscopic findings in patients with colorectal flat adenomas.

METHODS

In total, 1934 consecutive patients undergoing colonoscopy were prospectively included. Polyp shape was classified according to the Japanese classification. Chromoendoscopy was applied whenever a flat lesion was suspected. Indications for colonoscopy, demographic data, and characteristics of neoplastic lesions were recorded. Patients were classified as follows: group 1, no adenomas (n=1250); group 2, only protruding adenomas (n=427); group 3, protruding and flat adenomas (n=118); and group 4, only flat adenomas (n=139).

RESULTS

Approximately one in every 10 patients (13.2%) had flat adenomas. Among them, concomitant protruding adenomas were identified in approximately half of the cases. In multivariate analysis, age older than 50 years [odds ratio (OR)=1.62; 95% confidence interval (CI)=1.08-2.43, P=0.02], protruding adenomas (OR=2.17; 95% CI=1.65-2.87, P<0.001), follow-up colonoscopy for polyps or cancer (OR=2.22; 95% CI=1.59-3.10, P<0.001), screening colonoscopy (OR=1.60, 95% CI=1.15-2.22, P=0.005), and specifically trained endoscopist (OR=2.02, 95% CI=1.53-2.68, P<0.001) were associated independently with flat adenoma detection.

CONCLUSION

Flat adenomas have specific demographic factors that might help to improve detection. Particularly, age older than 50 years, colorectal neoplasia surveillance, and the presence of protruding adenomas should alert endoscopists to the possible presence of these lesions. Trained endoscopists may offer a greater chance of detecting these lesions.

Risk for high-grade dysplasia or invasive carcinoma in colorectal flat adenomas in a Spanish population

AIM

to determine the frequency and malignancy risk of colonic flat adenomas among patients with colorectal polyps in a Spanish population.

METHODS

1300 consecutive colonoscopic examinations were reviewed; 640 polyps were detected and removed endoscopically in 298 patients. Chromoendoscopy with 0.2% indigo carmine was applied to clarify the macroscopical appearance of flat lesions. The following data was collected for flat and protruding polyps: size, location (proximal or distal to splenic flexure), histology (neoplastic or non neoplastic), high grade dysplasia (HGD) and submucosal invasive carcinoma (SIC) or beyond.

RESULTS

490 polyps (76.6%) were adenomas and 150 (23.4%) hyperplastic; 114 (23.3%) adenomas were flat (3 flat-depressed) whereas 376 (76.7%) were protruding. The diameter of flat and protruding adenomas was 9.2 +/- 7.9. mm and 7.0 +/- 5.9 mm, respectively (p < 0.001). A proximal location was more frequent for flat (63.1%) than for protruding adenomas (48.7%) (p = 0.003). The rate of HGD or SIC was significantly higher in flat than in protruding adenomas (7.0 vs 2.6%; p < 0.04). Two of the 3 flat-depressed lesions (both <or= 10 mm in diameter) were carcinomas (T1 and T2, respectively). Flat adenomas had an increased risk for HGD or SIC (OR = 2,7; CI, 1,04-7,04; p < 0.05).

CONCLUSIONS

In a Spanish population, flat adenomas represent nearly one quarter of all colorectal neoplastic polyps, their most frequent location being the right colon and they bear a higher risk of malignancy than protruding adenomas, especially for the flat depressed type.

Short- and long-term outcomes of standardized EMR of nonpolypoid (flat and depressed) colorectal lesions > or = 1 cm (with video)

BACKGROUND

Nonpolypoid (flat and depressed) colorectal lesions are increasingly recognized. Their endoscopic removal requires specialized EMR techniques, which are more complex to perform. Outcomes data on EMR of nonpolypoid neoplasms in the United States is needed.

OBJECTIVE

To determine the safety and efficacy of EMR in the resection of nonpolypoid colorectal neoplasms > or = 1 cm.

DESIGN

Retrospective analysis.

SETTING

Veterans Affairs Palo Alto Health Care System.

PATIENTS

Over a 5-year period, patients who underwent EMR for nonpolypoid colorectal lesions > or = 1 cm.

INTERVENTION

A standardized approach that included lesion assessment, classification, inject-and-cut EMR technique, reassessment, and treatment of residual tissue.

MAIN OUTCOME MEASUREMENTS

Complete resection, bleeding, perforation, development of advanced cancer, and death.

RESULTS

A total of 100 patients (125 lesions: 117 flat and 8 depressed) met inclusion criteria. Mean size was 16.7 +/- 7 mm (range, 10-50 mm). Histology included 5 submucosal invasive cancers, 5 carcinomas in situ, and 91 adenomas. Thirty-eight patients (48 lesions) did not receive surveillance colonoscopy: 8 had surgery, 16 had hyperplastic pathology, and 14 did not undergo repeat examination. Surveillance colonoscopy was performed on 62 patients (77 lesions). Complete resection was achieved in 100% of these patients after 1 to 3 surveillance colonoscopies. All patients received follow-up (mean [standard deviation] = 4.5 +/- 1.4 years); none developed colorectal cancer or metastasis.

LIMITATIONS

Single endoscopist, retrospective study.

CONCLUSIONS

A standardized EMR (inject-and-cut) technique is a safe and curative treatment option in nonpolypoid colorectal neoplasms (> or = 1 cm) in the United States.

Flat adenomas

This article was presented at the conjoint CSSA and RACS (colorectal section) spring meeting Queensland, Australia, September 2004. The adenoma-carcinoma sequence describes a succession of events from polypoid adenoma to colorectal cancer. However, this model only accounts for up to two-thirds of colorectal cancers. There is growing evidence that flat adenomas are precursor lesions to a flat type of colorectal cancer and certain subtypes of these polyps are at greater risk of malignant transformation. If confirmed, the implications for screening, endoscopic recognition and management will become of increasing importance if we are to decrease the incidence of colorectal cancer.

Olympus Lucera high-resolution vascular ectasia mapping in combination with the type V crypt pattern for the invasive depth estimation and nodal disease estimation in Paris type II colorectal cancers: a comparative prospective analysis to 20 MHz ultrasound

INTRODUCTION

Flat and depressed neoplastic lesions of the colorectum [Paris type (PT) 0-II] localized to the superficial submucosal (sm) layer can be managed using endoscopic mucosal resection. Successful endoluminal management can be enhanced using endoscopic or ultrasound tools that help predict the degree of sm invasion. Previous studies addressing invasive depth estimation using high-magnification chromoscopic colonoscopy showed a low specificity for deep sm layer 3 invasion with miniprobe ultrasound demonstrating better nodal and T stage in vivo prediction. High-resolution vascular mapping of lesions can show microvascular superficial changes that may predict sm invasive disease.

AIMS

Vascular mapping in combination with high-magnification chromoscopic colonoscopy (HMCC) may provide an accurate tool for the invasive depth estimation of PT type II neoplastic lesions as compared with high frequency 20/12.5 MHz miniprobe ultrasound.

METHODS

Sixty-eight patients with a known diagnosis of PT II neoplasia were imaged using 3 "back to back" imaging modalities. Phase 1-vascular ectasia mapping; phase 2-HMCC with crypt analysis according to Nagata criteria; phase 3-12.5/20 MHz miniprobe ultrasound. Lesions predicted as T0/1/N0 were resected using endoscopic mucosal resection with the remaining referred for surgery. Each imaging modality was then compared with the resected histopathologic specimen used as the "gold standard."

RESULTS

N=68 lesions (19 sm1/13 sm2/36 sm3). Overall accuracy of Nagata criteria, Nagata criteria combined with vascular mapping, and ultrasound staging was 65%, 78%, and 94%, respectively (P<0.001) when observing the between phase differences. Fifty-two lesions were resected surgically. The prevalence of node positive disease was 16% (8/52) with the remaining 44/52 (84%) being confirmed pN0 at histopathology. The kappa coefficient of agreement between invasive depth estimation (using histopathology as the gold standard), Nagata stage, Nagata stage plus vascular ectasia mapping and ultrasound stage was 0.47, 0.65, and 0.9, respectively. A significant improvement in between phase differences was observed (P=0.001).

CONCLUSIONS

This is the first study to address the in vivo clinical utility of vascular mapping in combination with HMCC for the T and N staging of PT II neoplasia. Combination imaging may provide an adequate clinical tool for both T and N stage assessment in vivo and help stratify those patients at high risk for T2/N1 disease that may benefit from further high-frequency miniprobe ultrasound (HFUS) assessment and possible primary surgical excision. This is important in the clinical context, given the high overall costs of a second HFUS examination, limitation of HFUS resources, and safe selection of patients undergoing primary endoscopic resection versus surgical resection.

Perceptive errors in CT colonography

Published results to date have indicated a good per patient sensitivity of computed tomographic colonography (CTC) for colorectal cancer and for polyps measuring 10 mm or more together with a very good specificity. Sensitivity and specificity for polyps in the range of 6-10 mm are moderate. These results, however, can be achieved only with meticulous attention to technique including adequate colonic distention, and acquisition of supine and prone thin-section computed tomographic (CT) images. Moreover, there is a significant learning curve involved in the interpretation of CTC studies, with performance statistics improving with operator experience. Radiologists must be comfortable in reporting directly from workstation monitors and have access to and be familiar with software for multiplanar and endoluminal reconstructions. In addition to maximize polyp detection and minimize false positive results, reporting radiologists must have a working knowledge of normal colorectal anatomy and pathology on CTC and be familiar with potential pitfalls in interpretation. Besides the description of several possible causes for perceptive errors, also a literature search of perceptive errors in CTC is included in this paper.

Differential gene expression between flat adenoma and normal mucosa in the colon in a microarray analysis

BACKGROUND

Flat adenomas in the colon are associated with a relatively higher potential for malignancy. Distinct genes may be involved in the development of flat adenoma. The aim of this study was to profile gene expression changes in flat adenomas in the colon.

METHODS

A genomewide expression analysis was carried out by using flat adenoma and adjacent normal mucosa in the colon to detect differences in gene expression. Because the right and left colon have different embryonic origins, each sample was classified according to its location, and the gene expression levels between flat adenoma and adjacent normal mucosa were also compared among samples derived from the right or left colon.

RESULTS

A total of 180 genes were differentially expressed between flat adenoma and normal mucosa in the colon, including matrix metalloproteinase 7 (MMP7), cadherin 3 (CDH3), S100P, and dual oxidase 2 (DUOX2). In addition, a total of 89 and 49 genes were differentially expressed between flat adenoma and normal mucosa among the samples from the right and left colon, respectively. Subsequent quantitative real-time reverse transcriptase-polymerase chain reaction supported the reliability of the expression analysis. Immunohistochemical analysis confirmed differential CDH3 and MMP7 protein expression.

CONCLUSIONS

This is the first report characterizing the genes differentially expressed in flat adenomas using a microarray analysis. Considerable differences in the gene expression profiles of flat adenomas also exist between the right and left colon. These data should lead to new insights into the pathogenesis of flat adenomas in the colon as well as to new therapeutic strategies.

Flat and depressed neoplasms of the colon in Western populations

Flat and depressed neoplasms of the colon are defined endoscopically as visible non-exophytic, flat and/or depressed mucosal lesions with a height less than half the diameter of the lesion. These neoplasms are typically smaller than their polypoid counterparts, and might be associated with a more aggressive biological behavior. While these lesions have been described in cohorts of Japanese patients for over two decades, their existence in Western populations has been less well described. This review focuses on the epidemiology and biological behavior of flat and depressed neoplasms in Western populations as well as the strategies for their identification, endoscopic staging, and therapy.

Microsatellite instability caused by hMLH1 promoter methylation increases with tumor progression in right-sided sporadic colorectal cancer

OBJECTIVE

A subset of sporadic colorectal cancers (SCRCs) exhibits microsatellite instability (MSI). Most MSI in SCRCs is caused by hMLH1 inactivation due to promoter methylation. However, the role of MSI in the progression of SCRCs remains unclear.

METHODS

Thirty-two intramucosal cancers and 63 cancers with submucosal invasion were assigned to group 1 (early-stage cancer), and 30 Dukes' B and 26 Dukes' C cancers to group 2 (advanced-stage cancer). hMLH1 promoter methylation status was determined by methylation-specific PCR. MSI was determined using five markers. hMLH1 expression was determined immunohistochemically.

RESULTS

MSI was found in 1 of 95 (1.1%) tumors in group 1, compared with 4 of 56 (7.1%) tumors in group 2. In right-sided tumors, the overall frequency of hMLH1-methylation-positive tumors in group 1 was not significantly different from that in group 2 (17 of 43, 39.5%, vs. 9 of 23, 39.1%). In right-sided tumors with hMLH1 promoter methylation, the frequency of MSI-positive tumors in group 1 was significantly lower than that in group 2 (1 of 17, 5.9%, vs. 4 of 9, 44.4%, p=0.0081).

CONCLUSION

The frequency of MSI caused by hMLH1 promoter methylation increases with tumor progression in right-sided SCRCs.

On the role of endothelial progenitor cells in tumor neovascularization

The exact role that bone marrow (BM)-derived endothelial progenitor cells (EPCs) play in tumor neovascularization is heavily debated. We develop a quantitative three-compartment model with predictive power regarding the dynamics of tumorigenesis. There are two distinct processes by which tumor neovasculature can be built: angiogenesis is the formation of new blood vessels from preexisting vessels; vasculogenesis is the formation of new vessels by recruiting circulating EPCs. We show that vasculogenesis-driven and angiogenesis-driven tumors grow in different ways. (i) If angiogenesis is the prevailing process, then the tumor mass (and volume) will grow as a cubic power of time, and BM-derived EPCs will stay at a constant level. (ii) If vasculogenesis is the dominant process, then the tumor mass will be characterized by a linear growth in time, and the number of circulating EPCs (after possibly increasing to a maximum) will decrease to low levels. With this information, one can identify the "signature" of each of the processes in the observations of tumor growth and the dynamics of the relevant characteristics, such as the level of BM-derived EPCs. We show how our results can help explain some apparently contradictory experimental data. We also propose ways to couple this study with directed experiments to identify the exact role of vasculogenesis in tumor progression.

Flat-type early colorectal cancer preferentially develops in right-sided colon in older patients

BACKGROUND

Flat-type colorectal cancer is frequently reported in Japan and Europe, but its clinical features remain obscure. Thus, we investigated the clinical features of flat-type early colorectal cancer with respect to tumor location and patient age and compared them with those of polypoid-type early and advanced cancer.

METHODS

Between January 1999 and June 2001, total colonoscopy was performed in 6,178 patients (mean age, 61 years; 4,290 males and 1,888 females). Of these patients, 402 patients with 429 colorectal cancers were found: 202 at advanced stage (invading beyond muscularis propria) and 227 at early stage (carcinoma in situ or invading within submucosa). Early-stage cancer was classified into two macroscopic subgroups: flat-type and polypoid-type.

RESULTS

Out of 227 early cancers, 44 were flat type and 183 were polypoid. Flat-type early cancer was more frequently located in the right colon (57 percent, 25/44) than polypoid-type cancer (19 percent, 35/183; P < 0.001). Adenomatous component in flat-type early cancer was less frequent than in polypoid-type cancer (23 percent vs. 92 percent, P < 0.001). The proportion of right-sided colon in flat-type early cancer increased with age (33 percent in patients < or = 59 years, 50 percent in patients between 60 and 69 years, and 72 percent in patients > or = 70 years), whereas polypoid-type early cancer showed minimal change (16 percent, 18 percent, and 25 percent, respectively). An increase in the proportion of right-sided colon with age was also found in advanced cancer (20 percent, 38 percent, and 52 percent, respectively).

CONCLUSION

The incidence of flat-type early cancer in right-sided colon increased with age, similar to the pattern of advanced cancer. This suggests that flat-type early cancer may be a precursor of advanced cancer in the right colon, especially in older people.

Chromosomal instability in flat adenomas and carcinomas of the colon

Flat adenomas are flat or slightly elevated dysplastic lesions of the colorectal mucosa, mostly with a tubular architecture. Compared with polypoid adenomas of similar size, flat adenomas show a higher frequency of high-grade dysplasia and rapid submucosal invasion. The aim of this study was to survey whether flat colorectal lesions differ in their pattern of chromosomal aberrations from their polypoid counterparts. Six flat adenomas and 12 flat carcinomas were analysed by comparative genomic hybridization (CGH) and the pattern of chromosomal aberrations was compared with a previously published series of 112 polypoid adenomas and 82 polypoid carcinomas. In addition, multiplex ligation-dependent probe amplification (MLPA) for identifying DNA copy number changes of 25 individual genes on chromosome 20 was performed on 14 flat and 15 polypoid tumours. With CGH, flat adenomas showed on average 1.8 gains (range 1-4) and 3.2 losses (range 0-4), and the flat carcinomas 4.5 gains (range 0-8) and 3.5 losses (range 1-6). In both adenomas and carcinomas, high frequencies of 20q gain (83% and 92%, respectively) and 18q loss (83% and 92%, respectively) were found. This correlation between 20q gain and 18q loss had previously been observed in a subgroup of polypoid colorectal tumours. Both flat and polypoid colorectal tumours with 20q gains by CGH showed similar patterns of copy number ratios for the individual genes tested. TOP1, BCL2L1, and E2F1 had median copy number ratios of 2 or higher, while ZNF217 had a ratio around 3. In conclusion, flat adenomas and carcinomas of the large intestine show a similar pattern of chromosomal aberrations to that observed in a specific subgroup of polypoid lesions. The transcription factor ZNF217 is an important candidate for driving the 20q gain.

Colorectal cancer screening and surveillance with CT colonography: current controversies and obstacles

Computed tomographic (CT) colonography has been advocated as an alternative colorectal screening method because studies in populations with a high prevalence of polyps have demonstrated that sensitivity for patients with large (> or =10 mm) polyps is generally high (approximately 90%). In three recent studies in low-prevalence populations, however, these values vary from 55% to 94%. Many questions have been raised as to the cause of this remarkable variability, which hampers the implementation of CT colonography in colorectal cancer screening and surveillance. We provide an overview of some potential causes and discuss the available, often indirect, evidence. In addition, several other obstacles that may influence implementation are discussed. Many differences between the study with high sensitivity (94%) and the two studies with low sensitivity (55% and 64%) exist: the primary method to review the data (two or three dimensional), bowel preparation (with or without oral contrast agents), study design (verification method and analysis of adenomas only), reader's experience, and scanning technique (single vs. multislice, thin vs. thick sections). Additional obstacles for implementation in prevention of colorectal cancer may be controversial results concerning patient acceptance, the large-scale use of ionizing radiation, difficulties in detecting flat adenomas, and extracolonic findings. Use of primary three-dimensional review methods, addition of oral contrast agents to bowel preparation, and endoscopic verification of false-positive results on CT colonography are speculated to have a positive influence on sensitivity. Future investigations should demonstrate the influence of these potential factors on sensitivity of CT colonography. Despite a growing body of evidence, it remains uncertain to what extent patient acceptance, radiation issues, flat lesions, and extracolonic findings will be a stumbling block to using CT colonography for colorectal cancer screening.

Cyclooxygenase-2 overexpression is related to polypoid growth and K-ras gene mutation in T1 colorectal carcinomas

PURPOSE

Cyclooxygenase-2 is thought to play a role in the development of intestinal tumors and levels are elevated in approximately 80 to 90 percent of human colorectal carcinomas. To clarify the role that cyclooxygenase-2 plays in the development of colorectal carcinoma, we studied the relationship between cyclooxygenase-2 expression and tumor morphology and that between cyclooxygenase-2 expression and K-ras mutation.

METHODS

We classified 48 T1 colorectal carcinomas as polypoid or nonpolypoid and analyzed the clinicopathologic features. The expression of cyclooxygenase-2 was determined immunohistochemically, and nested polymerase chain reaction-restriction fragment length polymorphism detected a K-ras codon 12 mutation.

RESULTS

Cyclooxygenase-2 expression was higher in polypoid carcinomas than in nonpolypoid carcinomas (P < 0.001). The K-ras codon 12 mutation was associated with higher levels of cyclooxygenase-2 expression compared with carcinomas without this mutation (P = 0.028).

CONCLUSIONS

Polypoid growth and K-ras gene mutation are both associated with increased levels of cyclooxygenase-2 expression in T1 tumors.

Endoscopic morphological anticipation of submucosal invasion in flat and depressed colorectal lesions: clinical implications and subtype analysis of the kudo type V pit pattern using high-magnification-chromoscopic colonoscopy

OBJECTIVE

Focal submucosal invasive colorectal cancers (submucosa-sm1) can be managed by endoscopic mucosal resection (EMR) as local lymph node metastasis (LNM) are rare. Lesions are usually flat, depressed or mixed. In deeper vertical submucosal invasion (sm2-3) LNM rates exceed 10-15%. EMR within this group can be complicated by perforation, noncurative resection and may leave LNM untreated. It is therefore essential to differentiate accurately focal sm1 disease from submucosal sm2/3 disease. The aim of this study was to evaluate the relationship between the invasive type V pit pattern using high-magnification-chromoscopic-colonoscopy (HMCC) and submucosal invasive depth for flat and depressed colorectal lesions.

METHODS

Total colonoscopy was performed by a highly selected single endoscopist using the Olympus C240Z on 850 patients between January 2001 and July 2003. Kudo type V pits were identified using 0.05% crystal violet (CV) applied directly to the lesion using a steel tipped catheter. Type V pits were graded into class V(n)A-C as described by Nagata. Morphology was documented using the Japanese Research Society classification (JRSC). Histological sections, with reference to mucosal invasive characteristics, acquired using EMR or surgical excision were then compared with the pit pattern.

RESULTS

Fifty-one lesions showed a type V pit pattern. The kappa coefficient of agreement between pit the type V pit pattern and histologically confirmed submucosal invasion was 0.51 (95% CI). Following resection, 97% of lesions were correctly anticipated to have sm2 + invasion using pit type Vn(B) and Vn(C) as clinical indicators of invasive disease. Specificity was low at 50% with an accuracy of 78%.

CONCLUSIONS

The type V pit pattern is useful for the in vivo staging of submucosal invasive depth in flat and depressed colorectal lesions and is as sensitive as conventional 7.5 MHz EUS. There was a tendency to over-stage lesions and hence the technique is limited by its low overall specificity.

A prospective clinicopathological and endoscopic evaluation of flat and depressed colorectal lesions in the United Kingdom

OBJECTIVES

Flat and depressed colorectal lesions are now reported in Western populations. The malignant potential, anatomical distribution, and other clinicopathological features have not been established in this group. This study aimed to assess prospectively the prevalence, clinicopathological, and endoscopic features of flat and depressed colorectal lesions in the United Kingdom.

METHODS

A single endoscopist performed colonoscopy on 850 consecutive patients presenting for routine colonoscopy. All endoscopies were performed using a high magnification colonoscope with chromoscopy to facilitate detection of flat and depressed colorectal lesions.

RESULTS

A total of 458 flat lesions were identified. Of these, 173 (38%) were hyperplastic and 285 (62%) adenomatous or beyond. Of the 173 hyperplastic flat lesions, 162 (94%) were located in the recto-sigmoid region. Of the 267 adenomas, 66 (25%) had areas of high grade dysplasia (HGD), with 54/66 (82%) being present in the right colon. Flat lesions <8 mm in diameter was more likely to contain HGD than those <8 mm (p<0.001). Nine of the 10 (90%) flat invasive adenoacarcinomas were in the right colon and all had a depressed morphological component. In contrast, HGD was observed in 58/466 (12%) of protuberant (sessile/pedunculated) adenomas of which 95% (55/58) were located in the left colon. In addition, HGD was present in 17% of all sessile adenomas versus 44.6% of flat lesions >8 mm in diameter (p=0.001). Of the 14 protuberant carcinomas, 13/14 (93%) were in the left colon. Synchronous lesions were found in 96/816 (12%) of cases. Of the 816 patients with two or more left-sided protuberant adenomas <8 mm (with or without HGD), 89 (11%) had one or more flat lesions in the right colon with HGD.

CONCLUSIONS

Flat adenomas and carcinomas have a high malignant potential compared to protuberant lesions and have a propensity for developing in the right hemi-colon. Total colonoscopy is required to detect such lesions, as only 18% of flat lesions would be in reach of the flexible sigmoidoscope.

Colorectal flat neoplasia

BACKGROUND AND AIMS

An attempt has been made to evaluate the clinicopathological characteristics of flat colorectal neoplastic lesions, and analyse the factors associated with the malignancy.

PATIENTS AND METHODS

A total of 115 flat neoplastic lesions, > or = 5 mm in size, diagnosed in 87 patients by colonoscopy, were investigated.

RESULTS

The rectum was the most common location. Almost half (49.6%) of the flat neoplasms were small (5-10 mm), 27.8% were 11-20 mm and the remainder (22.6%) larger than 20 mm. The surface was smooth in 55.7%, granular in 20.0% and nodular in 24.3%. Histologically, the flat lesions were tubular, tubulovillous and villous adenomas in 69.6%, 20.9% and 5.2%, respectively. Five lesions (4.3%) were composed of carcinomas without adenoma. High-grade dysplasia, intramucosal carcinoma and invasive carcinoma were diagnosed in 9.6%, 7.8% and 6.1% of all flat neoplasms, respectively. Univariate analysis demonstrated that the location, size, surface pattern and histologic type of the flat lesions were factors associated with malignancy. However, in multivariate analysis, the size of the flat lesions was the only significant risk factor for malignant transformation.

CONCLUSIONS

Flat neoplastic lesions of the colorectum have a relatively high rate of malignancy, and size is the most important factor associated with malignancy.

Flat and depressed colorectal tumours in a southern Swedish population: a prospective chromoendoscopic and histopathological study

BACKGROUND

Flat and depressed colorectal tumours are common in Japan but are very rare or non-existent in Western countries.

AIMS

To study the occurrence of flat colorectal tumours in a southern Swedish population.

METHODS

In this prospective study, 371 consecutive European patients were examined by high resolution video colonoscopy combined with chromoendoscopy. The nature of the lesions was determined by histopathological examination.

RESULTS

A total of 973 tumours were found; 907 (93.2%) were protruding and 66 (6.8%) were flat or depressed. Of the flat/depressed tumours, five (7.7%) were early adenocarcinomas infiltrating the submucosa. Eleven carcinomas (1.2%) were found among protruding tumours. High grade dysplasia was observed in 18% (n=11) of flat/depressed adenomas in contrast with 7.3% (n=65) of protruding adenomas, and occurred in smaller flat/depressed tumours compared with protruding ones (mean diameter 8 mm v 23 mm, respectively). Furthermore, high grade dysplasia was significantly more common in flat elevated tumours with central depression or in depressed adenomas (35.7%; 5/14) than in flat elevated adenomas (12.8%; 6/47).

CONCLUSION

Flat and depressed tumours exist in a Western population. Future studies should address whether or not chromoendoscopy with video colonoscopy is necessary in the search for flat colorectal neoplasms.

Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population

BACKGROUND & AIMS

To assess the prevalence of flat and depressed (F&D) colorectal adenomas in the United States, we performed a prospective study of 211 American patients.

METHODS

Dye-assisted colonoscopy was performed in the presence of both an American and a Japanese investigator.

RESULTS

F&D lesions were found in 22.7% of patients, and these were more likely to be adenomatous than polypoid lesions (82% vs. 67%; P = 0.03) and contained more invasive cancer (4.5% vs. 0%; P = 0.04), which also appeared to be at a disproportionately advanced stage. The average size of all F&D advanced lesions (high-grade dysplasia and cancer) was significantly smaller than comparable polypoid lesions (10.75 +/- 2.7 mm vs. 20 +/- 2.9 mm; P < 0.05). F&D adenomas showed significantly stronger fragile histidine triad (FHIT) expression and lower p53 reactivity than similarly sized polypoid adenomas, whereas proliferative and apoptotic indices were similar in both groups.

CONCLUSIONS

We conclude that there is a significant prevalence of colonic F&D colorectal adenomas in this country and that these lesions have significantly different biologic features than polypoid lesions. The clinical and epidemiologic implications of these findings for American patients need to be addressed in further studies.

Predictive factors for detecting colorectal carcinomas in surveillance colonoscopy after colorectal cancer surgery

PURPOSE

The aim of this study was to identify the high-risk groups for metachronous colorectal carcinoma among patients who undergo colorectal cancer surgery.

METHODS

Three hundred forty-one patients undergoing colorectal cancer surgery who had undergone surveillance colonoscopy at least twice during a period of more than three years were analyzed. A metachronous colorectal carcinoma was defined as a new colorectal carcinoma detected by surveillance colonoscopy after surgery.

RESULTS

Surveillance colonoscopy was performed 4.6 times per patient during an average of 6.2 years. Twenty-two metachronous colorectal carcinomas in 19 patients were detected, and 14 (64 percent) of 22 were detected within five years of surgery. The cumulative incidence of developing colorectal carcinomas during a five-year period was 5.3 percent. Seventeen (77 percent) of 22 carcinomas were 10 mm or less in size. Ten (71 percent) of the 14 carcinomas in early stages showed a flat appearance. Univariate analysis showed that extracolonic malignancy, coexistence of adenoma, and synchronous multiple colorectal carcinoma were significant predictive factors for detecting colorectal carcinomas in surveillance colonoscopy and that family history of colorectal carcinoma was a possible predictive factor. Multivariate analysis performed with Cox proportional hazards regression model showed that extracolonic malignancy and the coexistence of adenoma were significant predictive factors.

CONCLUSION

We recommend that patients with the above predictive factors receive surveillance colonoscopy meticulously and regularly.

Apoptosis and its regulation in flat-type early colorectal carcinoma: comparison with that in polypoid-type early colorectal carcinoma

PURPOSE

The aim of this study was to investigate the relationship among apoptotic cell death, proliferative activity, and the expression of apoptosis-regulating proteins (p53, p21 (WAF1/CIP1), and bax) in flat-type early colorectal carcinoma and to compare these factors with those in polypoid-type early colorectal carcinoma.

METHODS

Formalin-fixed, paraffin-embedded tissues of 11 flat-type early colorectal carcinomas and 17 polypoid-type early carcinomas were studied. The histologic diagnosis was either well-differentiated adenocarcinoma or carcinoma in adenoma, and the depth of invasion was limited to mucosa or submucosa. Apoptotic cells were detected by terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling method, and proliferative activity was determined by Ki-67 immunohistochemistry using monoclonal antibody MIB-1. Apoptosis-regulating proteins were determined by immunohistochemistry using antibody DO-7 (p53), Cip1 (p21 (WAF1/CIP1)), and Bax (bax).

RESULTS

There was no significant difference in terminal deoxynucleotide transferase-mediated deoxy-uridine triphosphate-biotin nick end-labeling index between flat-type early colorectal carcinoma and polypoid-type early carcinoma, at 1.9 vs. 1.1, respectively. In flat-type carcinoma terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling index in the p53 protein overexpression group was significantly smaller than that in the p53 protein-negative group (P < 0.05). The Ki-67 labeling index/terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling index ratio in the p53 protein overexpression group was significantly higher than that in the p53 protein-negative group (P < 0.05). In polypoid-type carcinoma, the terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling index and Ki67/terminal deoxynucleotide transferase-mediated deoxyuridine triphosphate-biotin nick end-labeling index ratio showed no significant difference between the p53 protein overexpression group and p53 protein-negative group.

CONCLUSION

p53-dependent apoptosis may contribute to the development of flat-type early colorectal carcinoma. Apoptosis and its regulation in flat-type early colorectal carcinoma may differ from those in polypoid-type carcinoma.

Flat adenoma of the large bowel: re-evaluation with special reference to central depression

PURPOSE

The aim of this study was to re-evaluate the clinicopathologic features of flat adenomas with special reference to the role of central depression found in flat adenomas.

METHODS

Clinicopathologic features, such as grade of atypia by size, site, central depression, coexisting lesions, gender, and family history of cancer, were evaluated in 236 flat adenomas from 183 patients selected retrospectively and prospectively.

RESULTS

Of the 236 flat adenomas, 175 had mild, 33 had moderate, and 28 had severe (i.e., intramucosal carcinoma) atypia. The frequency of severe atypia correlated positively with size. Severe atypia were significantly more frequent in females (21 percent in females and 10 percent in males; P < 0.05). Sixteen percent of flat adenomas in patients with coexisting cancer showed severe atypia, which was significantly more than those without coexisting cancer (P < 0.05). The frequency of multiple flat adenomas in patients with two or more family members with cancer was 53 percent, which was significantly higher than in patients with one or no family members with cancer (P < 0.05). Central depression was observed in 19 percent of flat adenomas. The rate of severe atypia (22 percent) of flat adenomas with central depression was significantly higher than that (9 percent) of flat adenoma without central depression (P < 0.05). Patients who had flat adenomas with central depression had a higher incidence of coexisting cancer and cancer in the family (P < 0.05 and P < 0.01).

CONCLUSION

At colonoscopic examination consideration should be given for the increased potential to become malignant in flat adenomas, especially those with central depression and in patients with a history of malignancy or with a family history of malignancy.

Non-polypoid adenomas of the colon are associated with subjacent lymphoid nodules. An experimental study in rats

BACKGROUND

Subjacent lymphoid nodules (SLNs) have been found in 38% of non-polypoid colonic adenomas in humans. In the present work the presence of SLNs in experimentally induced colonic adenomas was investigated in rats.

METHODS

1,2-Dimethylhydrazine was injected subcutaneously in 290 Sprague-Dawley rats for 27 weeks.

RESULTS

An SLN was present in 28.6% of the 84 adenomas, in 8.4% of the 119 adenocarcinomas, and in 9.7% of the 31 small carcinomas without remnant adenomatous tissue. An SLN was found in 35.6% of the 59 non-polypoid neoplasias but only in 9.1% of the 175 polypoid (that is, exophytic) neoplasias. When only adenomas were considered, SLNs were present in 50.0% of the 34 non-polypoid adenomas but only in 14.0% of the 50 polypoid adenomas.

CONCLUSIONS

Non-polypoid colonic adenomas evolve preferentially from the minimal fraction of the colonic mucosa that overlays the few existing lymphoid nodules in rats.

Significance of microsatellite instability in different types of early-stage nonfamilial colorectal carcinomas

PURPOSE

The aim of this study was to investigate the genetic alterations of early-stage nonfamilial colorectal carcinomas regarding microsatellite instability, with special reference to the shape of the tumors and the site of the lesions.

METHODS

Formalin-fixed, paraffin-embedded specimens of 44 early-stage nonfamilial colorectal carcinomas were examined for microsatellite instability with use of polymerase chain reaction.

RESULTS

The 44 carcinomas consisted of 16 flat carcinomas and 28 polypoid carcinomas. Nineteen carcinomas were located in the proximal colon (9 flat type and 10 polypoid type), whereas 25 were in the distal colon and rectum (7 flat type and 18 polypoid type). Ten (22.7 percent) of the 44 carcinomas had at least one positive locus, whereas five (11.4 percent) of them had two or more positive loci. In the proximal colon the percentage of flat carcinomas with at least one positive locus was significantly greater than that of the polypoid carcinomas (4/9 (44 percent) vs. 0/10; P = 0.04). Six patients had synchronous or metachronous colorectal carcinomas or both. They harbored microsatellite instability more frequently than patients with single colorectal carcinomas, and the differences were statistically significant (P < 0.02).

CONCLUSIONS

These data suggest that in nonfamilial carcinomas in the proximal colon, the genetic pathway in flat carcinomas may be different from that in polypoid carcinomas.

Histologic classification of endoscopically removed flat colorectal polyps: a multicentric study

A total of 594 flat colorectal polyps, removed at endoscopy, were histologically classified into non-neoplastic (n = 49) and neoplastic (n = 545) polyps. Non-neoplastic polyps were subdivided into metaplastic (n = 45) and hyperplastic (n = 4), whereas neoplastic polyps were subdivided into adenomas (n = 481), intramucosal carcinomas (n = 28) and invasive adenocarcinomas (n = 36). Several adenoma phenotypes were discerned: tubular (n = 375), serrated (n = 59), villous (n = 39), mixed (n = 7) and fenestrated (n = 1). Intramucosal carcinomas were subdivided into tubular (n = 26) and serrated (n = 2), and invasive adenocarcinomas into tubular (n = 32), serrated (n = 3) and fenestrated (n = 1). The microscopic characteristics of each histologic phenotype described in this communication are defined and illustrated.

Leuchtenberger bodies in flat adenomas of the colorectal mucosa: a comparison between Japanese and Swedish patients

The dysplastic epithelium of flat colorectal adenomas was found to contain DNA inclusion granules, known as Leuchtenberger bodies (LB). Hematoxylin and eosin-stained sections of 263 colorectal flat neoplastic lesions (i.e., flat adenomas, and flat adenocarcinomas) collected in Japan and Sweden were scrutinized for the presence of such intraepithelial inclusions. LB were recorded in 87.1% of the 263 flat colorectal neoplastic lesions. The frequency of flat colorectal lesions having LB was similar in Japan (160 of 178 or 89.8%) to that in Sweden (69 of 85 flat lesions or 81.2%), suggesting that the occurrence of LB may not be influenced by race or environment. Direct immunoperoxidase detection of nuclear DNA fragmentation and transmission electron microscopy studies indicated that the inclusions contained apoptotic material. The semiquantitative assessment of histochemically labeled apoptotic granules indicated that the number of lesions having moderate to large numbers of apoptotic granules (++/ ) was higher in flat adenomas with high-grade dysplasia, suggesting that the occurrence of these granules may be related to the degree of severity of the dysplastic process. A better knowledge of LB may improve our understanding of the possible relationship between apoptosis, host immune response and carcinogenesis in flat colorectal adenomas.

Flat serrated adenomas of the colorectal mucosa

A total of 47 flat serrated neoplasias of the colorectal mucosa are presented: 44 were flat serrated adenomas and the remaining 3 flat serrated adenocarcinomas arising in flat serrated adenomas. These lesions were found among 600 flat mucosal lesions removed at colonoscopy during a 3-year period (1992 and 1994) at the Karolinska Hospital. Thirty-five of the 47 patients (74%) were males and the remaining 12 (26%), females. Depending upon the degree of cellular dysplasia within the epithelium, serrated adenomas were divided into those with low-grade dysplasia (LGD), when the dysplastic nuclei were present in the deeper half of the epithelium, and those with high-grade dysplasia (HGD), when the dysplastic nuclei were found even in the upper half of the epithelium. LGD was present in 37 (84.1%) of the 44 serrated adenomas and HGD in the remaining 7 (15.9%). Depending upon the topographic distribution of the dysplastic epithelium within the crypts, flat serrated adenomas were divided into type I, when the dysplastic epithelium was limited to the lower half of the serrated crypts, and type II, when the dysplastic epithelium was even present in the superficial half of the serrated crypts. Of the 44 serrated adenomas, 38 (86.1%) were type I and the remaining 6 (13.9%) type II. The dysplastic epithelium seemed to originate at the base of the crypts and to progress upwards, replacing the scalloped, serrated epithelium of the sides of the crypts. Invasive adenocarcinomas (i.e., with submucosal extension) were seen to arise from flat serrated adenomas with LGD type I (n = 2) or with HGD type II (n = 1). This preliminary survey suggests that flat serrated adenomas of the colorectal mucosa may be lesions with a propensity to evolve into invasive adenocarcinoma, irrespective of the degree of the epithelial dysplasia or of their extension along the crypts.

Mutation of p53 tumor suppressor gene in flat neoplastic lesions of the colorectal mucosa

PURPOSE

In a recent comparative histologic survey of flat colorectal neoplasias, we found more lesions with high-grade dysplasia (HGD) and carcinoma in Japanese than in Swedish patients. The purpose of this work was to assess the p53 protein overexpression in flat colorectal neoplasias in Swedish patients and to compare results with those reported in Japan.

METHOD

A total of 57 neoplastic lesions of the colorectal mucosa were investigated: 29 had been regarded both at endoscopy and at histology as flat and the remaining 28 as exophytic. Deparaffinized, rehydrated sections were treated immunohistochemically to detect the p53 protein. Lesions having a moderate (++) or high ( ) staining were considered as overexpressing the p53 protein.

RESULTS

Results indicated that 16.7 percent (1/6) of the exophytic adenomas with low-grade dysplasia (LGD) had distinct p53 overexpression as well as 57.1 percent (8/14) of those with HGD and 87.5 percent (7/8) with invasive growth. In flat neoplastic lesions, 7.7 percent (1/13) of the tubular adenomas with LGD, 25 percent (3/12) of tubular adenomas with HGD, and 75 percent (3/4) of adenocarcinomas arising in flat adenomas had p53 overexpression.

CONCLUSIONS

In Swedish patients, the proportion of flat and exophytic colorectal neoplasias showing p53 immunoreactivity increased with increasing degree of dysplasia, the highest percent being recorded in lesions with invasive growth. Because a similar stepwise increase was reported for exophytic and flat colorectal neoplasias in Japan, it seems that the comparison of results in both countries is justifiable. One possible conclusion from this comparison is that the higher proportion of flat neoplastic colorectal lesions with HGD and carcinoma in the Japanese (compared with the Swedish) takes place for reasons extraneous to the overexpression of the p53 protein.

Paneth cell-rich flat adenoma of the rectum: report of a case

A patient having familiar adenomatosis polyposis and an ileo-rectal anastomosis developed a flat mucosal lesion in the rectum. A punch biopsy revealed a villous adenoma with high-grade dysplasia. The subsequent surgical specimen indicated that the flat villous adenoma was rich in Paneth cells. Special stains included lysozyme muramidase (to visualize Paneth cells), MIB1 proliferation monoclonal antibody and single and multilabel immunohistochemistry for Paneth cells. Other methods included transmission electron microscopy and quantification with an image quantifier (Program Optilab 2.1) of lysozyme-stained Paneth cells. The subjective evaluation of hematoxylin-eosin-stained preparations demonstrated that the Paneth cells were mainly located in the lower half of the villi. Sections labeled with a specific stain (lysozyme muramidase) revealed more Paneth cells in the villi and electron microscopy showed even more in lysozyme-negative areas. Obviously some migrating dysplastic Paneth cells had retained their characteristic granules on their way towards the tip of the villi. Quantitative studies indicated that the lysozyme muramidase-positive material accounted for 41% of the adenomatous tissue. MIB1 revealed intense cell proliferation at the base of the adenoma and in the entire slopes of the villi. Despite the wide distribution of Paneth cells in intestinal metaplasia of the stomach, in the normal small intestine and in the large bowel with chronic inflammatory diseases, it is surprising that tumors arising in Paneth cells are extremely rare. The causes of the apparent natural resistance of Paneth cells to tumor development deserve to be investigated. This is the first case of Paneth cell-rich flat adenoma of the rectum in the literature.

p53 overexpression in flat serrated adenomas and flat tubular adenomas of the colorectal mucosa

The expression of the p53 protein was investigated in flat serrated neoplasias as well as in other histological phenotypes of flat or exophytic hyperplasias or neoplasias of the colorectal mucosa. A total of 104 such lesions were analyzed: 24 were flat serrated neoplasias (22 flat serrated adenomas and 2 flat serrated adenocarcinomas), 26 flat tubular adenomas, 17 flat hyperplastic polyps, 29 exophytic tubular and/or villous neoplasias (23 adenomas and 6 exophytic adenocarcinomas) and the remaining 8, exophytic hyperplastic polyps. Deparaffinized, rehydrated sections were treated immunohistochemically to detect those overexpressing the p53 protein. Lesions having slight (+), moderate (++) or intense ( ) staining were considered immunoreactive. The results showed that 50% of the flat serrated adenomas with low-grade dysplasia (LGD) and 66.7% of those with high-grade dysplasia (HGD) had p53 immunoreactivity. None of the flat tubular or of the exophytic adenomas with LGD expressed p53, but immunoreactivity was present in 61.5% of the flat tubular adenomas with HGD and in 52.3% of the exophytic adenomas with HGD. All adenocarcinomas had an intense p53 reaction. Weak p53 expression was demonstrated by 11.7% of the flat hyperplastic polyps but none of the exophytic polyps reacted. The occurrence of p53 expression in flat serrated adenomas with LGD suggested that, despite its low histological profile, one-half of those lesions could be biologically already committed to independent growth. The occurrence of p53 expression in nearly 12% of the flat hyperplastic polyps was totally unexpected and deserves further investigation. Flat serrated adenoma emerges as a novel, independent histological entity among the various phenotypes of flat neoplasias of the colorectal mucosa.