Chromoendoscopy techniques improve the visualization of mucosal structures. This article reviews and summarizes key studies addressing the impact of chromoendoscopy on colonic neoplasia detection and differentiation of neoplastic from non-neoplastic polyps in average and high-risk populations, including patients with colonic inflammatory bowel disease (IBD). In this context, there are convincing data that chromoendoscopy differentiates neoplastic from non-neoplastic polyps in average-risk populations with high accuracy. Moreover, dye-based chromoendoscopy improves neoplasia detection in colonic IBD surveillance.
Key points
- •
Chromoendoscopy techniques improve the visualization of mucosal structures.
- •
Chromoendoscopy differentiates neoplastic from non-neoplastic polyps in the average-risk population with high accuracy but does not distinguish both reliably in inflammatory bowel disease (IBD).
- •
Dye-based chromoendoscopy improves neoplasia detection in colonic IBD surveillance, with the potential to replace random colonic biopsies as the preferred surveillance option.
Chromoendoscopy
Chromoendoscopy refers to image-enhanced endoscopy through the use of dye spraying or optical techniques. Although initially limited to dye spraying, over the last decade, equipment-based imaging-enhanced optical colonoscopy techniques have been developed that are commonly referred to as dyeless or digital chromoendoscopy . Chromoendoscopy techniques improve the visualization of mucosal structures and, thus, improve recognition of borders and surface topography of pathologic lesions compared with standard white light colonoscopy. This review focuses on the role of these imaging techniques in the detection of colonic neoplasia.
Why Do We Need Chromoendoscopy?
Polypectomy of colonic neoplasms is the backbone of colorectal cancer (CRC) screening and health prevention measures, as it is associated with a 53% reduction of mortality. However, colonoscopy does not fully protect against CRC, with interval CRCs representing between 2% and 6% of all CRCs. There is evidence that most interval cancers arise from missed, rather than new, colorectal neoplastic lesions.
Adenoma miss rates average 24% and are highest for diminutive adenomas (26%) compared with adenomas greater than 10 mm (2%). The reasons for missing colorectal neoplasms during colonoscopy include inadequate bowel preparation; presence of flat polyps, which often resemble normal mucosa at first glance; and technical challenges of colonoscopy limiting mucosal visualization behind folds and in the right colon. Fortunately, flat neoplasms, which have a prevalence of 5% to 10%, appear preferentially in the right colon allowing endoscopists to focus their attention in this region when looking for such polyps. Sessile serrate adenoma/polyps account for an important subtype of mostly flat and right colonic neoplastic lesions, of which 9.5% contain high-grade dysplasia.
Since Kaminski and colleagues and Corley and colleagues found that a low adenoma detection rate (ADR) is an independent predictor for interval CRC, significant attention has been directed at increasing the ADR. Unfortunately, the introduction of high-definition (HD) colonoscopy has resulted in a diagnostic average gain of only 3.8% compared with standard white light colonoscopy. The marginal increase of ADR is mainly limited to diminutive polyps.
Chromoendoscopy has emerged as a method to improve ADR for both average and high-risk CRC screening populations, including those with inflammatory bowel disease (IBD). Chromoendoscopy may also have a role in distinguishing between neoplastic and non-neoplastic colonic lesions allowing for a resect-and-discard strategy for diminutive colonic lesions bypassing formal pathologic assessment.
Application of Dye for Dye-Based Chromoendoscopy
Dye-based chromoendoscopy uses color dyes that are either absorbed by the mucosa (vital dye) or remain on the mucosal surface (nonvital dye). The dye can be applied in a nontargeted fashion to the entire colonic mucosa (pan-chromoendoscopy) or to target certain colonic sections to define borders and predict histology of an area of interest. The two most common dyes used for staining are indigo carmine and methylene blue. Both dyes seem to be equally effective in enhancing dysplasia detection ( Table 1 ).
| Methylene Blue | Indigo Carmine | Crystal Violet | |
|---|---|---|---|
| Dye category | Vital | Nonvital | Vital |
| Staining mechanism | Active absorption into intestine epithelial cells Less or no absorption by inflamed mucosa and neoplasia | Not absorbed by cells but coats intestinal lining | Active absorption by Lieberkühn crypts |
| Staining pattern | Highlights pit pattern Increases contrast between mucosal structures Inflamed mucosa or neoplasia appears brighter than normal mucosa | Highlights pit pattern Increases contrast between mucosal structures | Highlights pit pattern |
| Dye application | Spray catheter or standard water pump | Spray catheter or standard water pump | Spray catheter |
| Concentration | 0.1% | 0.03%–0.5% | 0.05%–0.2% |
| Color | Blue | Blue | Blue |
| Staining time | 1 min to allow staining of normal mucosa | No waiting time following dye application | Typically sprayed after indigo carmine application |
| Duration of staining | Lasts for up to 20 min | Lasts for few minutes, disappears because of dilution throughout colon | n/a |
Indigo Carmine
Indigo carmine is not absorbed by cells (nonvital dye). It coats the mucosa highlighting mucosal pits, grooves, erosions, depressions, and subtle colonic contour irregularities. Its deep-blue color enhances the visualization of mucosal structures and allows better distinguishing of borders, depth, and surface topography of lesions.
Methylene Blue
Methylene blue is actively absorbed by small intestine and colonic epithelium (vital dye). This absorption requires waiting about 60 seconds before adequate staining is achieved. Colonic dysplastic and inflamed tissue absorb less or no dye resulting in different staining characteristics compared with normal mucosa. The different staining characteristics provides better resolution to distinguish borders and surface topography of lesions, similar to the application of indigo carmine.
Crystal Violet
Crystal violet is also a vital dye that stains colonic crypts by being preferentially absorbed by the crypts of Lieberkühn. Similar to methylene blue, crystal violet is absorbed by noninflamed mucosa better than by neoplasia and inflamed tissue. Crystal violet was shown to be useful in characterizing pit patterns, particularly in conjunction with indigo carmine. This dye is not used commonly in practice.
Safety of Dye Application
Generally, mucosal dyes used by gastroenterologists are safe; no significant adverse drug reactions are reported with the exception of methylene blue, which was shown to damage DNA when used in the evaluation of Barrett esophagus. The clinical importance of this finding is not known, and methylene blue is generally considered to be safe. Recent unpublished studies (Alessandro Repici, MD, verbal communication including data submitted to the US Food and Drug Administration, 2014) suggest that there is no clinically relevant DNA damage caused by methylene blue for colonic dye spraying.
Method of Dye Application
Adequate colonic preparation is paramount to achieve good mucosal visualization during chromoendoscopy. The endoscopist should lavage the colon on insertion to remove any remaining material. Once the cecum is reached, the dye is applied to the colonic mucosa. In this context, decompression of the colon allows better mucosal dye coverage.
The concentration of the indigo carmine and methylene blue has varied in studies from 0.03% to 0.5% mixed in water (indigo carmine 0.03%–0.5% ; methylene blue 0.1% ). The dye may be sprayed directly through the accessory channel of the colonoscope using a 60-mL syringe and a spray catheter. The spray catheter does, however, increase the cost of the procedure. Picco and colleagues outlined a more practical approach of dye application. The dye mixed in 1 L of sterile water was administered by a standard water pump device attached to the colonoscope, enabling the endoscopist to spray the dye by pressing the foot pedal with results similar to catheter-based applications (video example http://www.youtube.com/watch?v=6PJ91qYUPcE ).
Most recently, Repici and colleagues reported a coated methylene blue capsule allowing dye delivery limited to the colon by using a specific capsule formulation. However, it is not yet clear whether this method is an acceptable alternative to conventional chromoendoscopy.
For pancolonic chromoendoscopy, colon segments of 20 to 30 cm are sprayed with dye. Immediate evaluation can be performed following indigo carmine application after the excess dye is suctioned. Methylene blue, however, requires 60 seconds after application to achieve adequate mucosal staining. The steps are repeated during colonoscope withdrawal until the entire mucosa is visualized.
Kiesslich and Neurath proposed several technical guidelines (SURFACE) for the use of chromoendoscopy in ulcerative colitis that included assurance of optimal visualization with excellent bowel prep and targeted chromoendoscopy with indigo carmine or methylene blue. The use of a spasmolytic was also included, but it is typically not necessary. Classification of polyp pit pattern was also recommended but does require significant expertise.
Chromoendoscopy
Chromoendoscopy refers to image-enhanced endoscopy through the use of dye spraying or optical techniques. Although initially limited to dye spraying, over the last decade, equipment-based imaging-enhanced optical colonoscopy techniques have been developed that are commonly referred to as dyeless or digital chromoendoscopy . Chromoendoscopy techniques improve the visualization of mucosal structures and, thus, improve recognition of borders and surface topography of pathologic lesions compared with standard white light colonoscopy. This review focuses on the role of these imaging techniques in the detection of colonic neoplasia.
Why Do We Need Chromoendoscopy?
Polypectomy of colonic neoplasms is the backbone of colorectal cancer (CRC) screening and health prevention measures, as it is associated with a 53% reduction of mortality. However, colonoscopy does not fully protect against CRC, with interval CRCs representing between 2% and 6% of all CRCs. There is evidence that most interval cancers arise from missed, rather than new, colorectal neoplastic lesions.
Adenoma miss rates average 24% and are highest for diminutive adenomas (26%) compared with adenomas greater than 10 mm (2%). The reasons for missing colorectal neoplasms during colonoscopy include inadequate bowel preparation; presence of flat polyps, which often resemble normal mucosa at first glance; and technical challenges of colonoscopy limiting mucosal visualization behind folds and in the right colon. Fortunately, flat neoplasms, which have a prevalence of 5% to 10%, appear preferentially in the right colon allowing endoscopists to focus their attention in this region when looking for such polyps. Sessile serrate adenoma/polyps account for an important subtype of mostly flat and right colonic neoplastic lesions, of which 9.5% contain high-grade dysplasia.
Since Kaminski and colleagues and Corley and colleagues found that a low adenoma detection rate (ADR) is an independent predictor for interval CRC, significant attention has been directed at increasing the ADR. Unfortunately, the introduction of high-definition (HD) colonoscopy has resulted in a diagnostic average gain of only 3.8% compared with standard white light colonoscopy. The marginal increase of ADR is mainly limited to diminutive polyps.
Chromoendoscopy has emerged as a method to improve ADR for both average and high-risk CRC screening populations, including those with inflammatory bowel disease (IBD). Chromoendoscopy may also have a role in distinguishing between neoplastic and non-neoplastic colonic lesions allowing for a resect-and-discard strategy for diminutive colonic lesions bypassing formal pathologic assessment.
Application of Dye for Dye-Based Chromoendoscopy
Dye-based chromoendoscopy uses color dyes that are either absorbed by the mucosa (vital dye) or remain on the mucosal surface (nonvital dye). The dye can be applied in a nontargeted fashion to the entire colonic mucosa (pan-chromoendoscopy) or to target certain colonic sections to define borders and predict histology of an area of interest. The two most common dyes used for staining are indigo carmine and methylene blue. Both dyes seem to be equally effective in enhancing dysplasia detection ( Table 1 ).
| Methylene Blue | Indigo Carmine | Crystal Violet | |
|---|---|---|---|
| Dye category | Vital | Nonvital | Vital |
| Staining mechanism | Active absorption into intestine epithelial cells Less or no absorption by inflamed mucosa and neoplasia | Not absorbed by cells but coats intestinal lining | Active absorption by Lieberkühn crypts |
| Staining pattern | Highlights pit pattern Increases contrast between mucosal structures Inflamed mucosa or neoplasia appears brighter than normal mucosa | Highlights pit pattern Increases contrast between mucosal structures | Highlights pit pattern |
| Dye application | Spray catheter or standard water pump | Spray catheter or standard water pump | Spray catheter |
| Concentration | 0.1% | 0.03%–0.5% | 0.05%–0.2% |
| Color | Blue | Blue | Blue |
| Staining time | 1 min to allow staining of normal mucosa | No waiting time following dye application | Typically sprayed after indigo carmine application |
| Duration of staining | Lasts for up to 20 min | Lasts for few minutes, disappears because of dilution throughout colon | n/a |
Indigo Carmine
Indigo carmine is not absorbed by cells (nonvital dye). It coats the mucosa highlighting mucosal pits, grooves, erosions, depressions, and subtle colonic contour irregularities. Its deep-blue color enhances the visualization of mucosal structures and allows better distinguishing of borders, depth, and surface topography of lesions.
Methylene Blue
Methylene blue is actively absorbed by small intestine and colonic epithelium (vital dye). This absorption requires waiting about 60 seconds before adequate staining is achieved. Colonic dysplastic and inflamed tissue absorb less or no dye resulting in different staining characteristics compared with normal mucosa. The different staining characteristics provides better resolution to distinguish borders and surface topography of lesions, similar to the application of indigo carmine.
Crystal Violet
Crystal violet is also a vital dye that stains colonic crypts by being preferentially absorbed by the crypts of Lieberkühn. Similar to methylene blue, crystal violet is absorbed by noninflamed mucosa better than by neoplasia and inflamed tissue. Crystal violet was shown to be useful in characterizing pit patterns, particularly in conjunction with indigo carmine. This dye is not used commonly in practice.
Safety of Dye Application
Generally, mucosal dyes used by gastroenterologists are safe; no significant adverse drug reactions are reported with the exception of methylene blue, which was shown to damage DNA when used in the evaluation of Barrett esophagus. The clinical importance of this finding is not known, and methylene blue is generally considered to be safe. Recent unpublished studies (Alessandro Repici, MD, verbal communication including data submitted to the US Food and Drug Administration, 2014) suggest that there is no clinically relevant DNA damage caused by methylene blue for colonic dye spraying.
Method of Dye Application
Adequate colonic preparation is paramount to achieve good mucosal visualization during chromoendoscopy. The endoscopist should lavage the colon on insertion to remove any remaining material. Once the cecum is reached, the dye is applied to the colonic mucosa. In this context, decompression of the colon allows better mucosal dye coverage.
The concentration of the indigo carmine and methylene blue has varied in studies from 0.03% to 0.5% mixed in water (indigo carmine 0.03%–0.5% ; methylene blue 0.1% ). The dye may be sprayed directly through the accessory channel of the colonoscope using a 60-mL syringe and a spray catheter. The spray catheter does, however, increase the cost of the procedure. Picco and colleagues outlined a more practical approach of dye application. The dye mixed in 1 L of sterile water was administered by a standard water pump device attached to the colonoscope, enabling the endoscopist to spray the dye by pressing the foot pedal with results similar to catheter-based applications (video example http://www.youtube.com/watch?v=6PJ91qYUPcE ).
Most recently, Repici and colleagues reported a coated methylene blue capsule allowing dye delivery limited to the colon by using a specific capsule formulation. However, it is not yet clear whether this method is an acceptable alternative to conventional chromoendoscopy.
For pancolonic chromoendoscopy, colon segments of 20 to 30 cm are sprayed with dye. Immediate evaluation can be performed following indigo carmine application after the excess dye is suctioned. Methylene blue, however, requires 60 seconds after application to achieve adequate mucosal staining. The steps are repeated during colonoscope withdrawal until the entire mucosa is visualized.
Kiesslich and Neurath proposed several technical guidelines (SURFACE) for the use of chromoendoscopy in ulcerative colitis that included assurance of optimal visualization with excellent bowel prep and targeted chromoendoscopy with indigo carmine or methylene blue. The use of a spasmolytic was also included, but it is typically not necessary. Classification of polyp pit pattern was also recommended but does require significant expertise.
Equipment-based imaging-enhanced colonoscopy (digital or dyeless chromoendoscopy)
Since the introduction of HD colonoscopy, several equipment-based imaging-enhanced colonoscopy techniques have become available. The techniques have improved mucosal visualization of surface and vascular structures through the use of optical filters or utilization of software-based technologies. These technologies are reviewed in detail elsewhere in this issue.
Narrow-Band Imaging
Narrow-band imaging (NBI) is an optic filter-based method that allows blue and green wavelengths from the white light spectrum to pass through but blocks red wavelengths. Green and especially blue light wavelengths fall into the peak absorption of hemoglobin. As a consequence, blue and green light are absorbed by superficial and deep mucosal vessels, respectively, but are reflected by the remaining mucosa. This reflection improves visualization of mucosal vessels, which are frequently altered in form, density, and size in neoplastic colorectal lesions.
Fujinon Intelligent Color Enhancement
Fujinon Intelligent Color Enhancement (FICE) (Fujinon Inc, Japan) emits and captures the entire white light spectrum without the use of any optical filters. Following light capture, digital software-based computer algorithms modify the captured images. Thus, certain combinations of wavelengths are selectively enhanced, which results in improved visualization of subtle mucosal surface changes, especially of mucosal vessels and pit patterns.
iScan
Similar to FICE, iScan (Pentax, Japan) is a postprocessing, software-based technology functioning as a digital filter. The digital filter alters certain wavelengths of reflected white light, which results in enhancement of different elements of the mucosa. Three different filter algorithms allow for surface enhancement, contrast enhancement, and tone enhancement.
Chromoendoscopy for average-risk patients
Adenoma Detection Rate in Dye-Based Chromoendoscopy
Early randomized controlled trials failed to demonstrate a significant increase in ADR with dye-based chromoendoscopy compared with standard white light colonoscopy. In these studies, ADR with white light and chromoendoscopy ranged from 25% to 31% and 33% to 39%, respectively. The largest benefit of chromoendoscopy was an increase of the total number of detected diminutive adenomas from 37 in 135 patients to 89 in 124 patients. Two additional studies found significantly more patients with 3 or more adenomas in the chromoendoscopy group compared with the standard white light colonoscopy group (12% [15 of 124] vs 2% [3 of 135] and 10% [13 of 128] vs 3% [4 of 132], respectively). Also, more proximal colonic adenomas were detected by chromoendoscopy, although the results were only significant for diminutive adenomas (80% [71 of 89] vs 73% [27 of 37]). The prevalence of flat colorectal neoplasia in chromoendoscopy was up to 3 times greater than in regular screening colonoscopies. This finding was confirmed by a Cochrane meta-analysis of 5 randomized controlled studies that found that chromoendoscopy detected significantly more patients with at least one lesion neoplastic (odds ratio [OR] 1.67) and significantly more patients with 3 or more neoplastic lesions (OR 2.55).
Results were mixed comparing HD colonoscopy with chromoendoscopy. One study demonstrated similar ADRs (48.4% and 55.5%, no significant difference), whereas another showed a significant increase in ADR from 36.3% to 46.2%. Once again, the differences were mostly limited to a higher detection rate of diminutive lesions (30.4% vs 23.2%), serrated lesions (46.2% vs 29.5%), and flat adenomas. However, no significant difference was found for adenomas larger than 1 cm (12.9% vs 9.4%).
These data indicate that chromoendoscopy achieves marginally higher ADR than standard or HD colonoscopy, with an advantage limited to the detection of diminutive, flat neoplastic and serrated lesions. A disadvantage of chromoendoscopy is a significantly longer procedure duration. Studies addressing the rate of interval colon cancer following screening colonoscopy with chromoendoscopy do not exist but are of interest as chromoendoscopy seems to target lesions that are frequently missed in standard colonoscopy ( Table 2 ).
| White Light Colonoscopy (%) | Methylene Blue Chromoendoscopy (%) | Indigo Carmine Chromoendoscopy (%) | NBI/FICE/iScan (%) | |
|---|---|---|---|---|
| Average-risk screening colonoscopy | ||||
| ADR | 25.0–48.4 | Not reported | 33.0–55.5 | 28–64 |
| Differentiating neoplastic from non-neoplastic lesions | ||||
| Sensitivity | 33.0–95.5 | Not reported | 82–98 | 78–98 |
| Specificity | 74–97 | Not reported | 72.7–95.0 | 61.2–93.0 |
| PPV | 87.9–93.4 | Not reported | 98.1 | 89–93 |
| NPV | 63–100 | Not reported | 88 | 76–100 |
| Accuracy | 61.0–93.3 | Not reported | 75–94 | 79.0–98.6 |
| IBD surveillance colonoscopy | ||||
| Dysplasia detection rate per patient | 2.0–8.8 | 13.8–16.7 | 7.0–19.7 | 9–19 |
| Differentiating neoplastic from non-neoplastic lesions | ||||
| Sensitivity | 38–88 | 72–97 | 89–100 | 75–76 |
| Specificity | 65.6–93.8 | 92–93 | 91.0–96.8 | 66–81 |
| PPV | 93.4 | Not reported | 16.5–62.5 | Not reported |
| NPV | 63.3 | Not reported | 99.6–100 | Not reported |
| Accuracy | 72–84 | Not reported | Not reported | 67–80 |
Adenoma detection rate in digital chromoendoscopy
NBI is the most thoroughly studied digital chromoendoscopy in average-risk patients undergoing screening colonoscopy. Multiple studies, including tandem colonoscopy studies and randomized controlled trials, failed to detect a significantly different ADR when comparing standard white light colonoscopy with NBI. A subsequent meta-analysis of 2936 patients showed no significant difference in the overall ADR (36% vs 34%) with NBI. The average number of adenomas per patient also did not differ significantly (0.645 vs 0.59). The only significant advantage of NBI over standard white light colonoscopy was the higher number of detected diminutive adenomas, flat adenomas, and hyperplastic polyps, especially in the distal colon. Moreover, no difference for right- versus left-sided colonic polyps was found. Analogously, a meta-analysis comparing HD colonoscopy with or without NBI also found no significant difference in ADR or detection of flat adenomas and adenomas less than 10 mm of size.
In summary, NBI technology did not show any significant advantage over HD white light colonoscopy in terms of ADR. Therefore, it is not recommended as an ADR-improving strategy.
The data for iScan are less consistent. Hoffman and colleagues demonstrated a higher ADR and rate of flat adenoma detection in screening colonoscopies with iScan in relation to standard colonoscopy (38% vs 13% and 58% vs 23%). These differences were caused by improved detection of diminutive and flat adenomas with mean polyp sizes of 5.6 mm and 6.7 mm, respectively. It is important to note that the ADR in the control group was low, which may have exaggerated the effect of iScan. The same investigators also reported a lower adenoma miss in a tandem study with iScan compared with standard colonoscopy (30.0% vs 62.5%) but again most significantly in nonpolypoid lesions less than 5 mm of size. Others have reported no significant differences in ADR and adenoma miss rate comparing HD colonoscopy and iScan in tandem fashion (31.9% vs 36.5% and 22.9% vs 19.3%). Lastly, following an iScan, a tandem dye-based chromoendoscopy increased the number of detected polyps less than 5 mm in size by 92% in an evaluation limited to the distal colon. Most of the polyps were, however, non-neoplastic polyps.
The least amount of evidence is available for FICE, which did not demonstrate a significant advantage over standard white light colonoscopy in terms of ADR (64% vs 55%). A randomized trial showed similar results with no significant difference in ADR when compared with HD colonoscopy: 28% versus 28%. The only prospective randomized trial comparing FICE with chromoendoscopy for screening colonoscopy also demonstrated no significant difference in ADR (64% vs 68%).
In summary, NBI and iScan primarily improved the detection of diminutive and hyperplastic polyps; however, the overall ADR was not improved. FICE was also unable to improve ADR. Therefore, the routine use of digital chromoendoscopy is not recommended as a tool to improve ADR.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
