Occult dysplasia is disclosed by lugol chromoendoscopy in alcoholics at high risk for squamous cell carcinoma of the esophagus

Background and Study Aims: Squamous cell carcinoma of the esophagus (SCCE) becomes symptomatic at a late stage when the disease is already advanced, a nd this contributes to its poor prognosis. Esophagoscopy of asymptomatic in dividuals exposed to known risk factors associated with the development of this cancer may facilitate the diagnosis of early cancerous or precancerous lesions; however, conventional esophagoscopy is not accurate enough. The a im of this study was to measure the value of Lugol chromoendoscopy of the e sophagus (LCE) as an endoscopic technique to detect dysplasia in patients a t risk. Patients and Methods: We studied 190 male patients older than 35 attending an outpatient unit for alcoholics who consumed more than 80 g of alcohol, m ore than 10 cigarettes and more than 500 ml 'mate' (a hot infusion of herbs ) per day over 10 years. All underwent conventional upper gastrointestinal endoscopy followed by LCE, a spraying of Lugol 3% on the entire esophagus. All patients denied dysphagia. Biopsies were obtained from any unstained ar eas larger than 5 mm and also from stained areas in all individuals. Biopsi es were analyzed independently by two pathologists unaware of the biopsy si tes. All conventional esophagoscopies showed normal mucosa, except for two suspicious small elevated lesions, confirmed histologically to be SCCE. The se two cases were excluded from the statistical analysis. Results: The LCE found unstained areas in 23 patients and a uniformly stain ed esophageal mucosa in the remaining 165. Biopsies taken from these 23 uns tained areas showed dysplasia in six (two high grade and four low grade), a nd the ones from the 165 stained areas taken at the middle esophagus showed low-grade dysplasia in seven. There was a high prevalence (6.9%) of dyspla stic lesions in these individuals and occult dysplasia was significantly mo re frequent in unstained than stained areas (p=0.0017). LCE showed a sensit ivity of 46%, a specificity of 90%, a positive predictive value of 26% and a negative predictive value of 96% when unstained areas were compared to st ained ones. Agreement between two independent pathologists was high, with a kappa coefficient of 0.64. Conclusion: We concluded that individuals who abuse alcohol, smoke and cons ume 'mate' have a high prevalence of dysplastic lesions that can be better detected by LCE. Esophagi with unstained areas had an eight-fold higher cha nce of revealing dysplasia than the uniformly stained ones. LCE is an easy and inexpensive method. It improves the detection of dysplasia and should b e added to conventional upper GI endoscopy in patients at risk for SCCE.