In this case a poorly differentiated eSCC was associated with florid ISM originating from esophageal gland ducts.
Two types of glands exist in the esophagus: esophageal cardiac-type glands located in the lamina propria which have a duct lined by gastric foveolar-like cells, and submucosal glands which are drained by ducts lined by a single layer of cuboidal epithelium in their proximal part but becoming stratified squamous when penetrating the muscularis mucosae and the epithelium before opening into the esophageal lumen [7]. In our case, stratification of ductal squamous epithelium of submucosal glands was already observed in the submucosa illustrating an early step of ISM. Furthermore, we illustrated that ISM displayed a ductal pattern. ISM with overlying mucus-secreting cells was only found in the lamina propria supporting the fact that this type of ISM developed from esophageal cardiac-type glands located in the lamina propria because the duct-lining cells may extend over the stratified squamous epithelium for variable distances [7].
Esophageal ISM is poorly reported in literature. About three decades ago, Takubo described reserve cell hyperplasia in 24% of 110 cases of which he analyzed histologically the complete mucosa of the esophagogastric junction [6]. Esophageal ISM could originate from stem cells that are present in esophageal glands and ducts; several studies support the notion that stem or progenitor cells within esophageal submucosal glands and/or ducts can generate both squamous and Barrett’s columnar cells [8]. We observed diffuse positivity for Ber-EP4 immunostaining in ISM while tumoural cells as well as normal esophageal squamous epithelium were negative. In squamous epithelium of endodermal origin Ber‐EP4 could not be demonstrated in mature squamous epithelium, but was focally present in immature, metaplastic and dysplastic squamous epithelium and occasionally also in hyperplastic squamous epithelium [9].
Esophageal ISM should not be confounded as intraductal spread of eSCC which can occur in nearly the quart of eSCC [10]. Intraductal spread of eSCC that extends to the submucosa should not be staged as submucosal infiltration and it has no impact on 5-year survival [11]. Similarly, esophageal ISM should not be considered as invasive cancer and we demonstrated in our case a preserved basal membrane surrounding the islets of ISM. Being aware of this pitfall can avoid overstaging and, so, overtreatment.
Classification of esophageal ISM as preneoplastic or dysplastic lesion (as in the cervix) is at this moment difficult due to lack of data in the literature.
In conclusion, extensive description of esophageal ISM is lacking in the literature; however, this entity needs to be recognized in order to avoid misinterpretation as cancer, and Ber-EP4 immunohistochemistry is a helpful immunohistochemical staining. Furthermore, data with regard to classification of esophageal ISM as possible preneoplastic or dysplastic entity are lacking. Co-occurence of ISM with eSCC as in the present case suggests that this subject merrits further investigation.