Advanced Endoscopic Imaging in the Lower GI Tract

, Anja Landowski¹ and Mahesh Jayanna¹

(1)

Lyell McEwin Hospital, Haydown Road, Elizabeth Vale, 5112, SA, Australia

(2)

University of Adelaide, Adelaide, SA, Australia

Keywords

Advanced endoscopic imagingEnhanced imagingLower gastrointestinal tractKudo’s pit patternEarly colorectal cancerColorectal lesion assessment

Introduction

Gastrointestinal cancers are often preceded by a curable, non-invasive, premalignant stage that may progress asymptomatically. The transformation process begins with the formation of atypical cells in the epithelium, directly above the basal membrane. Early screening may enable detection of anomalies at very early stages before neoplasia permeates into the deeper submucosal layer and beyond. Morphological characteristics of the mucosa can then be interrogated to obtain further information, which may aid the endoscopist in deciding whether to proceed with endoscopic biopsies or resection.

Detection

Most newer electronic image enhancing modalities perform no better than white light endoscopy in the detection of colorectal neoplasia (Table 6.1). There is simply no substitute for good bowel preparation and a meticulous withdrawal technique, ensuring clear views are maintained to enable careful interrogation of mucosal folds.

Table 6.1

Image enhanced endoscopy in detecting colorectal polyp

Author	Year	Technology	# of pts	Design	Result
Rex [1]	2007	NBI	434	RCT	−
Inoue [2]	2008	NBI	243	RCT	−
Adler [3]	2008	NBI	401	RCT	−
Matsuda [4]	2008	AFI	167	Random, Tandem	+
Pohl [5]	2008	FICE	764	RCT	−
Kaltenbach [6]	2008	NBI	276	Random, Tandem	−
Van Den Broek [7]	2009	AFI	100	Random, Tandem	−
Adler [8]	2009	NBI	1,256	RCT	−

Characterization

Lesions which are detected should be interrogated further to gain additional information should be studied further to gain valuable information. Ideally, this is performed in a methodical manner, initially with a “wide field” overview of the lesion where the Paris classification and granularity are assessed, followed by a closer or “micro” or “magnified” view where the vascular patterns are visualised, with some of the electronic chromoendoscopy techniques. If further information is needed and where possible, the Kudo’s pit pattern is assessed. In vivo classification of polyps using advanced imaging is important for two reasons: (1) the significant cost of histological examination of all polyps, particularly small low-risk lesions, has prompted consideration of a “diagnose, resect, and discard” strategy which could be cost effective, and (2) accurately differentiating invasive from non-invasive cancers may enable clinical decisions to be made in real time, thereby immediately guiding therapy.

Lesion Assessment with Wide Field View

Paris Japanese Classification

The Paris Japanese classification system is especially important not only for standardization but also because it allows prediction of submucosal invasion risk [9].

Polyps can be divided into:

Protruding lesions (Fig. 6.1)

Fig. 6.1

Paris Japanese classification for protruding lesion

(a)

Ip (peduncalated)

(b)

Is (sessile): >2.5 mm* from base of polyp (surrounding mucosa)

Flat lesions (Fig. 6.2)

Fig. 6.2

Paris Japanese classification for flat lesions

(a)

IIa: Slightly elevated (<2.5 mm*)

(b)

IIb: True flat lesion

(c)

IIc: Mildly depressed lesion

*The 2.5 mm limit is used to differentiate sessile (Is) from flat (0-IIa) lesions and approximates the diameter of a closed biopsy forceps.

Flat colorectal lesions account for up to half of all colorectal polyps, while depressed lesions occur less frequently and account for about 1–3 % of all polyps (Table 6.2). The prevalence of high-grade dysplasia (HGD) or invasive cancer, however, increases as lesions become depressed (Paris IIc). Up to 59 % of all Paris type IIc lesions harbor HGD or submucosal invasion (SMI) (Table 6.3).

Table 6.2

Percentage of flat-depressed colorectal lesions

Author	No. of adenomas	% of flat lesions
Jaramillo [10]	261	42
Rembacken [11]	321	36
Saitoh [12]	136	40
Rex [1]	785	56
Okuno [13]	66,670	1.9
Togashi [14]	5,408	2.8
Soetikno [15]	1,535	1.2
Tsuda [15]	973	1.4

Table 6.3

Prevalence of HGD and SMI in colorectal polyps according to Paris Japanese classification

Author	HGD			SMI
Author	Ip	IIa/b	IIc	Ip	IIa	IIc
Rembacken [11]	7.4 %	12.8 %	25 %	0.9 %	1.7 %	50 %
Soetikno [15]	0.5 %	3.5 %	225	0.5 %	1.0 %	11 %
Tsuda [16]	7.3 %	12.8 %	35.7 %	–	–	–
Hurlstone [17]	12 %	15.4 %	59 %	–	–	–

Large colorectal lesions (measuring >20 mm in size) are relatively infrequent and may account for up to 4 % of all polyps (Table 6.4). The size of the lesion does not appear to matter when lesions are assessed for SMI, though. Moss and colleagues prospectively assessed colorectal LSTs in a large cohort of Australian patients; SMI was detected in 33/514 (6.4 %) polyps. The mean size of these polyps was 35.3 mm, in comparison to 35.6 mm when no SMI was detected (p = 0.87) [18].

Table 6.4

Comparison of Paris Japanese classification and size

	Total	≤5 mm	6–10 mm	11–19 mm	≥20 mm
Polypoid (Is, Ip)	14,814	47.6 %	37.7 %	12.6 %	2.1 %
Flat (IIa, IIb)	10,363	73.1 %	13.9 %	9.0 %	4.0 %
Flat-depressed (IIc)	585	45.0 %	29.4 %	21.7 %	3.9 %
Total	24,862	14,892	7,190	2,919	761