Chromoendoscopy

INTRODUCTION — Chromoendoscopy involves the topical application of stains or pigments to improve tissue localization, characterization, or diagnosis during endoscopy [1]. Several agents have been described that can broadly be categorized as absorptive (vital) stains, contrast stains, and reactive stains (table 1). Absorptive stains (eg, Lugol's solution and methylene blue) diffuse or are preferentially absorbed across specific epithelial cell membranes. Contrast stains (eg, methylene blue) highlight surface topography and mucosal irregularities by permeating mucosal crevices. Reactive stains (eg, Congo red and phenol red) undergo chemical reactions with specific cellular constituents, resulting in a color change. The stains used for chromoendoscopy are transient, unlike the stains used to tattoo lesions. (See "Tattooing and other methods for localizing gastrointestinal lesions", section on 'Tattooing'.)

Chromoendoscopy has been applied in a variety of clinical settings and throughout the gastrointestinal tract. Interest in chromoendoscopy increased with the development of technologies such as endoscopic mucosal resection and photodynamic therapy that require precise visual tissue characterization. In addition, chromoendoscopy is being used in conjunction with other advances in endoscopic imaging, such as magnification endoscopy, confocal endomicroscopy, and confocal endocytoscopy. (See "Magnification endoscopy" and "Confocal laser endomicroscopy and endocytoscopy".)

Compared with other evolving diagnostic modalities, such as fluorescence spectroscopy, fluorescence endoscopy, and optical coherence tomography, the equipment needed for chromoendoscopy is widely available. Furthermore, the techniques are simple, quick, inexpensive, and safe. However, the interpretation of the findings is not always straightforward, and, like many endoscopic techniques, the impact of chromoendoscopy on clinical outcomes relative to standard endoscopic and histologic methods has not been established in large controlled trials.

This topic will review chromoendoscopy, which generally refers to the application of stains or pigments by spraying through a catheter. Endoscopic tattooing (the injection of dye through a needle to mark a site for future identification), optical coherence tomography, narrow band imaging, and magnification endoscopy are discussed separately. (See "Tattooing and other methods for localizing gastrointestinal lesions", section on 'Tattooing' and "Optical coherence tomography in the gastrointestinal tract" and "Barrett's esophagus: Evaluation with optical chromoscopy" and "Magnification endoscopy".)

INDICATIONS — The role of chromoendoscopy as a surveillance technique for specific conditions (eg, Barrett's esophagus, inflammatory bowel disease) is discussed separately [2]. (See "Barrett's esophagus: Surveillance and management" and "Surveillance and management of dysplasia in patients with inflammatory bowel disease".)

EQUIPMENT — Minimal equipment is required for chromoendoscopy, and the reagents used are generally widely available. The procedure is carried out using standard endoscopic equipment. In addition, a spray catheter is essential, since it delivers a fine mist to the mucosa (movie 1). The catheters are reusable and last several years, even when used frequently. A single use biopsy channel cap is preferable to minimize the amount of stain that leaks out.

Stains that have been used for chromoendoscopy include Lugol's solution, methylene blue, toluidine blue, Congo red, phenol red, and indigo carmine. In order to choose the appropriate stain, the endoscopist needs to be familiar with the characteristics of the individual stains and the staining features of specific tissues (table 1).

GENERAL TECHNIQUE — The technique for staining is simple and easy to learn. Certain staining techniques require specific tissue preparations, which are described below. Once that has been accomplished, application of the stain is generally straightforward. The endoscope and catheter tip should be directed toward the mucosa using a combination of rotational clockwise-counterclockwise movements, while simultaneously withdrawing the endoscope tip.  

SPECIFIC DYES

Lugol's solution — Lugol's solution contains potassium iodine and iodine, which have an affinity for glycogen in non-keratinized squamous epithelium. The stock solution needs to be diluted to 1 to 4 percent (usually 2 to 3 percent strength works best) followed by spraying of 20 to 50 mL through a spray catheter.

Normal squamous epithelium stains black, dark brown, or green-brown after a few minutes (picture 1). An abnormal staining pattern (absence of dye uptake) is associated with conditions that result in depletion of glycogen in squamous cells, such as inflammatory change (eg, reflux esophagitis), dysplasia, or early malignancy. Staining with Lugol's solution is particularly helpful in patients at increased risk for squamous cell carcinoma (such as those with alcohol use disorder) and in those with head and neck cancer [3-9]. It is most commonly used in the esophagus for detection of squamous dysplasia and early squamous cell carcinoma (picture 1) [9]. Because it can help in revealing the extent and delineation of a lesion, it can be used to guide endoscopic mucosal resection of early squamous cell carcinomas. The number and size of Lugol’s voiding lesions can predict those at high risk for having a metachronous or second primary carcinoma in patients with squamous cell esophageal and head and neck carcinoma [10]. (See "Overview of endoscopic resection of gastrointestinal tumors".)

Efficacy of Lugol's staining — For the detection of squamous lesions, staining with Lugol's solution has a sensitivity of 91 to 100 percent and a specificity of 40 to 95 percent [11].

The value of Lugol's staining has been demonstrated in studies looking at the detection and evaluation of esophageal squamous neoplasia:

●One study that included 158 patients at high risk for squamous cell cancer demonstrated that Lugol's staining improved the evaluation of the extent of lesions [12]. Twelve of the patients had cancerous lesions detected by endoscopy before Lugol's staining, with one additional patient with cancer being identified after staining. There was a significant difference in the size of the lesions detected before and after staining. The dye-free surfaces (suggestive of malignancy) after staining averaged 11.6 cm² compared with the endoscopically measured mucosal involvement before staining of 1.4 cm².

●In a series of 225 patients, Lugol's staining increased the identification of both moderate and severe dysplasia, but not of cancer [11]. Of 31 patients with moderate dysplasia, 17 (55 percent) were identified only after staining with Lugol's solution. Of the 35 patients with severe dysplasia, eight (23 percent) were identified only after staining. In addition, 88 percent of the high-grade dysplasia and cancerous lesions were larger or more clearly defined after staining.

●In a study comparing narrow band imaging (NBI) with Lugol's staining in patients with early esophageal squamous cell carcinoma undergoing endoscopic submucosal dissection, there were no significant differences in R0 resection rate, margin status, adverse events, recurrence rate, or survival between the techniques. However, NBI was associated with shorter procedure time and less pain based on visual analog score [13]. In another study, detection of early cancer in patients with pre-existing caustic strictures was better with NBI compared with Lugol's staining [14].

Safety of Lugol's staining — Lugol's staining may lead to a transient retrosternal discomfort. A trial suggested that the application of a sodium thiosulfate solution (20 mL of a 5 percent solution) following staining was associated with a reduction in discomfort 30 minutes after the procedure [15]. An effective but more widely available alternative to sodium thiosulfate for neutralizing residual iodine is N-acetylcysteine [16].

Patients with a history of iodine allergy should not undergo Lugol's staining. Severe allergic reactions have been reported [17], as have cases of chemical esophagitis [18] and gastritis [19].

Methylene blue — Methylene blue is a vital stain taken up by actively absorbing tissues such as small intestinal and colonic epithelium. It does not stain nonabsorptive epithelia such as squamous or gastric mucosa.

Methylene blue staining has been studied in the esophagus, stomach, and small and large intestines. In the small intestine and colon, a lack of staining suggests metaplastic, neoplastic, or inflammatory change, whereas in other areas of the gastrointestinal tract it is used to identify metaplastic absorptive mucosa (eg, Barrett's mucosa).

Methylene blue staining is used to:

●Aid in the detection of Barrett's esophagus and associated dysplasia and/or early cancer. However, dye spraying has largely been replaced by virtual chromoendoscopy using NBI or I-SCAN [20] or by acetic acid mucosal enhancement [21]. (See 'Acetic acid' below.)

●Improve the diagnosis of early gastric cancer, either alone [22] or in combination with Congo red dye [23].

●Identify metaplastic absorptive epithelium, such as intestinal-type metaplasia in the stomach [24]. In addition, because it only stains absorptive epithelium, it can identify metaplastic nonabsorptive epithelium (such as ectopic gastric metaplasia) in areas that have background staining, such as the small bowel.

●Highlight subtle mucosal changes in the small intestine (eg, celiac disease [25]).

●Detect colonic neoplasia (flat adenomas and carcinomas) [26]. Pit patterns, which are highlighted by the methylene blue staining, correlate with neoplastic lesions. (See 'Colonic dysplasia' below.)

●Detect the extent of inflammatory changes and aid in the detection of intraepithelial neoplasia in patients with chronic ulcerative colitis [27].

Methylene blue application — Methylene blue staining involves the application of a mucolytic, followed by dye, followed by washing off excess dye. Surface mucus must be removed by applying a mucolytic agent to increase the uptake of the dye into epithelial cells. The original technique for staining involved ingestion of a proteolytic enzyme solution (proteinase or Pronase) followed by methylene blue in a capsule [22,28]. This technique was adapted for use in the United States by substituting a 10 percent solution of N-acetylcysteine for Pronase and a 0.5 percent solution of methylene blue for the capsule [24].

The reagents are sequentially sprayed onto the mucosa using a washing catheter (picture 2). Excess dye is then vigorously washed off with syringes. The endpoint of staining is subjective and is the most difficult part to learn. As a general rule, washing should continue until the staining pattern is stable. The most common mistake is failure to adequately rinse the mucosa following dye application. Positive staining is defined as the presence of blue-stained mucosa that persists despite vigorous water irrigation.

Efficacy of methylene blue staining

Barrett's esophagus — The most extensive experience with methylene blue staining in the United States, Europe, Asia, and South America has been in the evaluation of patients with Barrett's esophagus. Multiple reports that were published in final form or as abstracts that have shown discrepant results [29-45]. The discrepancies appear to be related to the learning curve for this staining technique; investigators with experience and studies with large sample sizes have reported the best results [29-35,46-48]. Perhaps because of these issues, a meta-analysis of nine studies with heterogeneous study designs, methods, and study populations concluded that methylene blue chromoendoscopy did not significantly increase the yield of detection of specialized intestinal metaplasia and dysplasia compared with random biopsies [49].

The potential to improve the detection of intestinal metaplasia was illustrated in a study of 975 patients with areas in the distal esophagus that were macroscopically conspicuous for Barrett's esophagus [48]. All patients underwent conventional biopsies, after which the distal esophagus was sprayed with methylene blue; additional targeted biopsies were obtained based upon the staining pattern. All patients with documented intestinal metaplasia underwent a repeat endoscopy within one year to assess the reproducibility of the method. A total of 3900 conventional biopsy specimens without staining and 130 biopsy specimens with staining were obtained. Conventional biopsies were significantly less likely to show intestinal metaplasia compared with dye-directed biopsies (1.4 versus 88 percent). The conventional biopsies diagnosed Barrett's esophagus in 16 patients (1.6 percent), whereas the dye-directed biopsies led to the diagnosis in 35 patients (3.5 percent), including all patients diagnosed with conventional biopsies. Intestinal metaplasia was confirmed on the follow-up endoscopy in all dye-positive patients.

Colonic dysplasia — Chromoendoscopy has been used in patients undergoing colonoscopy, including those undergoing surveillance for chronic ulcerative colitis. Using pit pattern classifications, chromoendoscopy has a sensitivity of 92 to 98 percent and a specificity of 91 to 95 percent for differentiating neoplastic from non-neoplastic lesions [50]. The Kudo pit pattern classification is most commonly used (figure 1) [51,52]. Lesions with gyrus-like pits or non-structural pits are more likely to be neoplastic than lesions with round pits or stellar pits.

One older study found that when combined with high-magnification endoscopy, 0.25 percent methylene blue can aid in the visualization of aberrant crypt foci, the earliest neoplastic lesions of the colon that are precursors to adenomas and carcinomas [53]. In an examination of 97 biopsy samples, the combination of chromoendoscopy and magnification endoscopy had a sensitivity of 100 percent and a specificity of 97 percent for detecting dysplasia. (See "Magnification endoscopy".)

In a randomized trial, 174 patients with chronic ulcerative colitis were assigned to either undergo colonoscopy with chromoendoscopy using 0.1 percent methylene blue or conventional colonoscopy [50]. Magnification endoscopy was used in the chromoendoscopy arm to classify the lesions detected by chromoendoscopy based upon the pit pattern [51]. The yield of chromoendoscopy for the detection of intraepithelial neoplasia was higher than that for conventional colonoscopy (32 versus 10 lesions). In addition, chromoendoscopy had a 93 percent sensitivity and specificity for differentiating neoplastic from nonneoplastic lesions.

However, for surveillance in patients with inflammatory bowel disease, the use of chromoendoscopy with a standard-definition colonoscope has been largely replaced by electronic or virtual chromoendoscopy using a high-definition colonoscope. Surveillance for dysplasia in patients with inflammatory bowel disease is discussed in more detail separately. (See "Surveillance and management of dysplasia in patients with inflammatory bowel disease".)

Gastric neoplasia — Methylene blue staining has been combined with magnification endoscopy in the evaluation of gastric neoplasia [54,55]. In a study of 136 patients, the sensitivity and specificity of magnification chromoendoscopy for intestinal metaplasia were 76 and 87 percent, and for dysplasia were 97 and 81 percent, respectively [54]. In clinical practice, electronic chromoendoscopy such as NBI, combined with high-definition or magnification endoscopy, provides an accurate and efficient alternative to methylene blue chromoendoscopy. However, randomized trials are needed to compare these techniques.

Safety of methylene blue — Methylene blue staining is generally considered to be safe. However, a concern has been raised regarding the potential to induce oxidative damage to DNA in tissues exposed to methylene plus white light (which is used during endoscopy) [56]. Such oxidative damage has the potential to accelerate carcinogenesis. However, this theoretical risk for increasing neoplastic transformation has not been proven by clinical studies showing increased cancers in patients who have undergone staining.

Indigo carmine — Indigo carmine is derived from a blue plant dye (indigo) and a red coloring agent (carmine) [1]. Its availability is limited in many areas due to a shortage of the raw materials used in its production [57]. Unlike the vital stains (which are taken up by tissues), indigo carmine is not absorbed by gastrointestinal epithelium. It pools in crevices between epithelial cells, highlighting small or flat lesions and defining irregularities in mucosal architecture, particularly when used with high-magnification or high-resolution endoscopy.

It is used primarily in the colon for the detection and evaluation of colorectal neoplasia and is the most common form of chromoendoscopy applied in the colon. As with many other stains, indigo carmine is used to evaluate pit patterns (figure 1). These patterns can help discriminate between hyperplastic polyps (which have a typical "pit" pattern) and adenomatous polyps (which have a "groove" or "sulci" pattern) [58]. Pit patterns can also aid in the diagnosis of minute, flat, or depressed colorectal tumors and increase the detection of flat adenomas (picture 3) [59-63]. Indigo carmine can assist in the detection of dysplastic changes in patients with ulcerative colitis undergoing surveillance colonoscopy [64,65], as well as aid in the detection of adenomas in patients with hereditary nonpolyposis colorectal cancer [66].

Indigo carmine has also been used:

●In combination with high-magnification endoscopy and Lugol's staining to diagnose the villiform appearance of Barrett's esophagus [17,67] (see "Magnification endoscopy").

●To diagnose small gastric cancers used alone [22] or in combination with acetic acid staining [68].

●To evaluate villous atrophy in patients suspected of having malabsorption from celiac disease or tropical sprue [69].

●To detect duodenal adenomas in patients with familial adenomatous polyposis [70].

There are several techniques described for indigo carmine staining. The oral route involves ingestion of a capsule or a dye-containing colonic electrolyte lavage solution. The dye (0.1 to 0.8 percent) can also be sprayed directly onto the mucosa. Indigo carmine has also been injected into the celiac artery (intra-arterial dye method) to facilitate endoscopic delineation of the size and extent of gastric cancers [71].

Efficacy for finding colorectal neoplasia — The use of indigo carmine in the colon has been examined in randomized controlled trials with variable results [62,63,72-74]. A meta-analysis found a benefit from chromoendoscopy for the detection of polyps, both neoplastic and non-neoplastic [75]. The analysis found that patients who underwent chromoendoscopy were more likely to have a polyp detected (odds ratio [OR] 2.1), a benefit that persisted when only neoplastic lesions (OR 1.6) or diminutive neoplastic lesions (OR 1.7) were considered. Finally, chromoendoscopy also detected more patients with three or more neoplastic lesions (OR 2.6).

Representative trials (including trials done subsequent to the meta-analysis) have demonstrated the following:

●In one trial, 660 patients were assigned to undergo either high-definition chromoendoscopy with spraying of the entire colon or high-definition white light colonoscopy [73]. Chromoendoscopy showed a nonsignificant trend toward higher mean rates of detection of patients with at least one adenoma (56 versus 48 percent for white light colonoscopy) and number of adenomas per patient (1.3 versus 1.1). While the absolute differences were small, chromoendoscopy detected more flat adenomas per patient (0.6 versus 0.4), more adenomas less than 5 mm in diameter per patient (0.8 versus 0.7), and more non-neoplastic lesions per patient (1.8 versus 1.0). The mean procedure time was longer for chromoendoscopy (30 versus 22 minutes).

●In a second large trial, 400 patients were assigned to undergo colonoscopy with chromoendoscopy performed by an endoscopist experienced in the technique (group A, n = 200), chromoendoscopy performed by an endoscopist who was not experienced in the technique (group B2, n = 100), or colonoscopy without chromoendoscopy, but with a minimum withdrawal time of 10 minutes (group B1, n = 100) [76]. After adjusting for confounders, the odds of finding a polyp were lower in group B1 compared with group A (OR 0.44). There was no significant difference in polyp detection between group A and group B2. While there was not a statistically significant difference in the rate of adenoma or cancer detection between group A and group B1, flat adenomas were detected less often in patients in group B1 compared with group A (OR 0.24).

●In another trial, 259 patients were assigned to undergo either colonoscopy with spraying of the entire colon or routine colonoscopy [62]. The study found no difference in the proportion of patients with at least one adenoma or the total number of adenomas detected between the two groups, though there were more adenomas less than 5 mm detected proximal to the sigmoid colon and more patients with three or more identified adenomas in the chromoendoscopy arm. The withdrawal time was longer in the chromoendoscopy arm compared with the control arm (median 9 versus 5 minutes).

●A fourth trial compared the use of pancolonic chromoendoscopy and targeted chromoendoscopy [72]. A total of 260 patients were assigned to one of the two arms. In an attempt to control for withdrawal time, a minimum withdrawal time of 8 minutes was established, and there was no difference in median withdrawal times (17 minutes in the chromoendoscopy group and 15 minutes in the control group). More adenomas were detected in the pan-chromoendoscopy group compared with the control group (57 versus 34), including more diminutive adenomas (<4 mm) and more patients with greater than three adenomas detected.

●A fifth trial included 1008 patients who were assigned to either pancolonic chromoendoscopy or standard colonoscopy [74]. The proportion of patients with at least one adenoma was higher in the chromoendoscopy group compared with the control group (46 versus 36 percent). In addition, chromoendoscopy increased the per patient detection rate for adenomas overall (09.5 versus 0.66), flat adenomas (0.56 versus 0.28), and serrated lesions (1.19 versus 0.49). Mean withdrawal times were longer in the chromoendoscopy group (11.6 versus 10.1 minutes).

Randomized trials using tandem colonoscopy studies have also been performed to evaluate adenoma detection rates for chromoendoscopy. In a tandem colonoscopy study, a standard colonoscopy is performed, and any polyps detected are removed. A second colonoscopy is then immediately carried out with patients assigned to have the exam either with chromoendoscopy or without (in an attempt to control for the fact that some lesions that were missed on the first examination might be detected on a second examination, even without dye spraying).

Tandem colonoscopy trials have also had variable results:

●In a trial of 50 patients with a history of either colon cancer or colon adenomas, chromoendoscopy found additional adenomas in 44 percent of patients, compared with 17 percent for those who underwent a second colonoscopy with intensive inspection (inspection time of more than 20 minutes) [77]. The adenomas detected by chromoendoscopy were smaller than those detected with intensive inspection (mean 2.7 versus 3.2 mm). Chromoendoscopy took longer than intensive inspection (mean procedure time 37 versus 27 minutes). After controlling for inspection time, chromoendoscopy still detected more adenomas than intensive inspection.

●A second trial examined 54 patients with Lynch syndrome using the same protocol described above [78]. After controlling for age, number of prior colonoscopies, procedure time, and history of prior colonic resection, chromoendoscopy detected more polyps overall, but there was no difference in the adenoma detection rate. (See "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Clinical manifestations and diagnosis".)

●There was no difference in polyp detection rate in a tandem colonoscopy trial that included 292 patients [79]. The patients in the chromoendoscopy group were also examined with structure enhancement (an electronic function that is supposed to enhance medically relevant structures). While more hyperplastic polyps were detected in the chromoendoscopy group, there was no difference between the groups with regard to the number of patients found to have adenomas or the total number of adenomas detected. The median procedure times were significantly longer in the chromoendoscopy group compared with the control group (27 versus 19 minutes).

Given mixed results with regard to the effectiveness of chromoendoscopy for adenoma detection and the increased time required to perform the procedure, the use of chromoendoscopy for the routine detection of colorectal neoplasia needs to be individualized.

Efficacy in chronic ulcerative colitis — Chromoendoscopy using indigo carmine has been studied in patients with chronic ulcerative colitis (CUC). The evidence supporting the use of chromoendoscopy along with guidelines regarding endoscopic surveillance techniques in patients with CUC are discussed elsewhere. (See "Surveillance and management of dysplasia in patients with inflammatory bowel disease", section on 'Chromoendoscopy'.)

Toluidine blue — Toluidine blue (also called tolonium chloride) is a basic dye that stains cellular nuclei. These properties make it useful for identifying malignant tissues, which have increased DNA synthesis and a high nuclear to cytoplasmic ratio [80].

Staining is accomplished by spraying 1 percent acetic acid (which acts as a mucolytic) before and after spraying a 1 percent aqueous solution of toluidine blue. The second application of acetic acid washes off excess dye. After staining, abnormal tissue appears blue, but false-positive results may occur if inflammatory or fibrotic lesions are present.

Toluidine blue staining has been applied in several clinical settings:

●Screening for early squamous esophageal cancers [81] and for use in patients with head and neck cancers [82,83].

●Helping to distinguish benign from malignant ulcers in the stomach [84].

●Detecting Barrett's esophagus. In one report, the sensitivity and specificity were 98 and 80 percent, respectively [85]. A limitation is that it cannot discriminate between gastric and intestinal metaplasia.

No adverse effects from toluidine blue staining have been reported.

Cresyl violet — Cresyl violet has been used for staining uterine cervical lesions and enhancing the endoscopic diagnosis of early malignancies. It acts by staining cell nuclei and is applied as a 0.05 to 0.2 percent solution. It has been used in the colon to highlight pit patterns [86] and has been combined with magnification endoscopy to diagnose a characteristic staining pattern of early gastric carcinomas [87].

Crystal violet — Crystal violet stains cell nuclei and can be used in the esophagus for the detection of Barrett's esophagus and Barrett's associated dysplasia [88]. It has also been used to highlight pit patterns in the colon [89]. It is applied as a 0.05 to 0.1 percent solution.

A double-staining technique has been used in the esophagus and colon to aid with lesion delineation. In the esophagus, methylene blue staining is performed to delineate the lesion, followed by crystal violet staining to enhance mucosal patterns. A similar technique has been used in the colon, substituting indigo carmine for methylene blue [90]. In one study of 1250 patients, high-magnification chromoendoscopy, using a combination of indigo carmine and crystal violet staining in the colon, predicted incomplete endoscopic mucosal resection of flat, sessile colonic lesions [91]. (See 'Methylene blue' above and 'Indigo carmine' above.)

Acetic acid — Acetic acid has been used to stain abnormal tissues during examination of the cervix, where it whitens immature and dysplastic cervical squamous epithelium. (See "Cervical cancer screening tests: Visual inspection methods".)

Acetic acid enhancement has also been applied in the gastrointestinal tract, and it is an inexpensive, simple, and accurate alternative to digital techniques that use optical or electronic enhancement. Studies have suggested that the application of acetic acid may help identify intestinal metaplasia in the esophagus and gastric cardia (picture 4 and picture 5) and help detect dysplasia or early cancer in patients with Barrett's esophagus [92,93]. It has been used to improve visualization of areas of duodenal mucosal atrophy in patients suspected of having celiac disease [94]. Acetic acid has also been mixed with indigo carmine to improve the visualization of early gastric cancer [68] and as a mucolytic agent to enhance the visualization of the mucosal pattern with high-resolution endoscopy [95]. (See "Magnification endoscopy".)

Acetic acid provides contrast enhancement of the surface epithelium. It is typically used in combination with magnification endoscopy [93,96], in which case it is referred to as "enhanced magnification endoscopy." It also provides contrast enhancement when used together with unmagnified high resolution white light endoscopy. Approximately 10 mL of 1.5 to 3 percent acetic acid is sprayed onto the esophageal wall. Initially, both the esophageal mucosa and the gastric mucosa turn white, but after two to three minutes, the normal esophagus remains white ("acetowhitening"), whereas Barrett's mucosa and gastric columnar mucosa will turn red. The effect is transient, lasting only two to three minutes, so repeated applications of acetic acid may be required.

Barrett's esophagus — Studies have demonstrated high sensitivity and moderate specificity of dye-based chromoendoscopy for detection of intestinal metaplasia and neoplasia when used with magnification endoscopy [92-94,96-104]. In a systematic review and meta-analysis including 12 trials, the overall sensitivity and specificity of dye-based (methylene blue or acetic acid) chromoendoscopy was 91.9 percent (95% CI 89.4-93.8 percent) and 89.9 percent (95% CI 80.1-95.2 percent), respectively [105]. The performance characteristics for acetic acid chromoendoscopy were sensitivity 96.6 percent (95% CI 95.2-97.7 percent) and specificity 84.6 percent (95% CI 68.5-93.2 percent). Hence, acetic acid chromoendoscopy seems to be superior to methylene blue, with greater sensitivity, wider availability, simpler technical application, and no risk for effects on cellular DNA.

Virtual chromoendoscopy using NBI or other digital technology may be more cost effective than dye-based chromoendoscopy. However, dye-based chromoendoscopy using acetic acid may be a good alternative if virtual chromoendoscopy is not available. Surveillance and management of Barrett's esophagus are discussed separately. (See "Barrett's esophagus: Surveillance and management".)

Dysplasia detection in Barrett's esophagus — Acetic acid staining also appears to increase the detection of dysplasia and cancer in patients with Barrett's esophagus. A meta-analysis of nine studies that included 1379 patients undergoing Barrett's surveillance with acetic acid enhancement and targeted biopsies demonstrated high accuracy for prediction of high-grade dysplasia (HGD) and early cancer [104]. The pooled sensitivity and specificity for the diagnosis of HGD or early cancer were 92 and 96 percent, respectively. Furthermore, one prospective study demonstrated the time to the disappearance of acetowhitening could predict high-risk neoplasia in Barrett's esophagus [97]. In a retrospective study of 982 patients with Barrett's esophagus undergoing surveillance, staining with acetic acid led to detection of dysplasia or superficial cancer in 41 of 327 patients (13 percent), compared with 13 of 655 patients (2 percent) in the random biopsy group [103]. In addition, the number of biopsies needed to detect one patient with dysplasia or early cancer was lower with acetic acid (40 versus 604).

A classification system for distinguishing non-neoplastic from neoplastic Barrett's esophagus using acetic acid enhancement has been developed by an international panel [21]. The criteria for recognizing neoplasia include focal loss of acetowhitening and surface patterns of Barrett's mucosa. Application of the acetic acid classification system (known as PREDICT) resulted in improved diagnostic accuracy for detecting Barrett's neoplasia.

Data have suggested that there was no significant difference in dysplasia detection between dye-based chromoendoscopy and virtual chromoendoscopy (narrow band imaging) [105].

LIMITATIONS — While easy to perform and readily available, there are multiple limitations to chromoendoscopy.

●Chromoendoscopy, especially in the colon, can be time-consuming.

●Many endoscopists lack training in chromoendoscopy, which can result in poor staining techniques and misinterpretation of findings (eg, failure to adequately wash following staining with methylene blue, resulting in false-positive results). Interpretation of the staining patterns requires familiarity and is not always straightforward. It is also prone to interobserver variability.

•In a study of 51 patients undergoing magnification chromoendoscopy for detecting Barrett's esophagus, there was a high level of interobserver variability (kappa <0.4) [98].

•In a study of 163 small colon polyps from 104 patients, interobserver variability for interpreting pit patterns (figure 1) was low for experienced endoscopists (kappa 0.85), but high for trainees (kappa 0.40).

•In a study of various techniques for detecting Barrett's esophagus, the interobserver variability for chromoendoscopy using indigo carmine for various findings was low, with kappa values ranging from 0.32 to 0.46 [106]. Chromoendoscopy with acetic acid was not significantly better (kappa values ranging from 0.42 to 0.48). In addition, the values did not change significantly when only experts were evaluated.

●While some classification systems for chromoendoscopic findings exist (eg, the Kudo pit classification system), overall there is a lack of standardization.

●Studies have shown poor reproducibility with regard to the efficacy of chromoendoscopy (eg, as noted above, some studies have found that chromoendoscopy increases adenoma detection rates in the colon, while others have not).

●Studies looking at the impact of chromoendoscopy on patient outcomes are limited.

●Studies comparing chromoendoscopy with other enhanced imaging technologies (such as narrow band imaging) are limited, and primarily focus on chromoendoscopy in the colon [107-110]:

•One study randomized 114 patients to undergo magnifying colonoscopy with either computed virtual chromoendoscopy using Fujinon Intelligent Color Enhancement (FICE), which is similar to narrow band imaging (NBI), or targeted indigo carmine dye spraying for colonic lesions of 1 cm or less [107]. There was no difference in sensitivity or specificity for the FICE group compared with the chromoendoscopy group (93 versus 97 percent and 82 versus 89 percent, respectively).

•A study of 142 patients being screened for esophageal malignancy found that NBI performed well when compared with chromoendoscopy as the gold standard [108]. NBI had a sensitivity of 91 percent for detecting squamous cell carcinoma or high-grade intraepithelial neoplasia, with a specificity of 95 percent.

•A study of 13 patients with familial adenomatous polyposis compared chromoendoscopy with NBI, autofluorescence endoscopy, and white light (standard) colonoscopy [109]. Chromoendoscopy detected a higher mean number of lesions per patient compared with NBI, autofluorescence, or white light colonoscopy (43 versus 20, 21, and 12 lesions, respectively).

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Gastric cancer" and "Society guideline links: Esophageal cancer" and "Society guideline links: Barrett's esophagus" and "Society guideline links: Colorectal cancer".)

SUMMARY AND RECOMMENDATIONS

●Background – Chromoendoscopy involves the topical application of stains or pigments to improve tissue localization, characterization, or diagnosis during endoscopy. Several agents have been described that can broadly be categorized as absorptive (vital) stains, contrast stains, and reactive stains (table 1). (See 'Introduction' above.)

●Lugol's solution – Lugol's solution is most commonly used in the esophagus for detecting high-grade squamous dysplasia and early squamous cell carcinoma and for delineating the borders for endoscopic resection or dissection. Lugol's solution is readily available, easily applied, and clinically useful. It can also be combined with other mucosal enhancement techniques, such as narrow band imaging. (See 'Lugol's solution' above.)

●Indigo carmine – Indigo carmine is used primarily in the colon for the detection and evaluation of colorectal neoplasia. It is the most common form of dye-based chromoendoscopy applied in the colon. Indigo carmine may be used alone or with crystal violet for the detection of early colorectal cancers. (See 'Indigo carmine' above and 'Crystal violet' above.)

●Methylene blue – Methylene blue chromoendoscopy is a simple, inexpensive dye-based technique with high sensitivity and specificity for Barrett's esophagus and dysplasia detection. However, due to cost and concern with possible association with DNA damage, it has been performed infrequently. (See 'Methylene blue' above.)

In clinical practice, chromoendoscopy using vital stains (eg, indigo carmine, methylene blue) with standard-definition endoscopy has been largely replaced by virtual chromoendoscopy for detecting colonic dysplasia. However, data comparing diagnostic yields and outcomes are limited.

●Acetic acid – Acetic acid enhancement has been used for detecting Barrett's esophagus and dysplasia, and it is an inexpensive, simple, and highly sensitive and specific alternative to digital techniques that enhance endoscopic imaging. (See 'Acetic acid' above.)