Small Barrett's adenocarcinoma, 3 mm in size, in long-segment Barrett's esophagus detected after 4 years of follow up

Authors


Corresponding: Dai Hirasawa, Department of Gastroenterology, Sendai City Medical Center, 5-22-1, Tsurugaya, Miyagino-ku, Sendai, Miyagi 983-0824, Japan. Email: hirasawa@openhp.or.jp

Abstract

We herein report a rare case of very small Barrett's adenocarcinoma. A 65-year-old man underwent surveillance esophagogastroduodenoscopy (EGD) 3 years after endoscopic submucosal dissection (ESD) for superficial Barrett's adenocarcinoma, which revealed a very small reddish area in the mucosa, 2 mm in diameter, in long-segment Barrett's esophagus. The EGD carried out 1 year later confirmed slight enlargement of the lesion, from 2 mm to 3 mm in diameter. Macroscopic type changed from flat type to slightly depressed type. On narrow-band imaging with magnifyingendoscopy, an irregular microstructure and irregular microvasculature became recognizable. It was resected by ESD and diagnosed as mucosal adenocarcinoma with a diameter of only 3 mm.

Introduction

Barrett's esophagus (BE) has high carcinogenic potential to develop into Barrett's adenocarcinoma (BAC). Surveillance in cases of BE is recommended to detect BAC in its early stage.[1, 2] However, it is sometimes quite difficult to diagnose the presence[3] and the lateral extent of BAC[4] because BE has inflammation and intestinal metaplasia of the columnar epithelium caused by reflux of gastric acid into the esophagus. Due to the above-mentioned reason, many cases of BAC are not detectable until their advanced stage. To detect BAC in its early stage, in Western societies, the guidelines for BE surveillance recommend random four-quadrant biopsies every 2 cm for the entire length of the BE.[1, 5] However, this system has problems such as economic efficacy[6] and safety. Although it is highly desirable to detect early-stage BAC by endoscopic observation only, there have been few reports of small BAC detected during surveillance endoscopy.

We herein report a case of very small BAC in BE, only 3 mm in diameter. This lesion was detected by surveillance esophagogastroduodenoscopy (EGD) after endoscopic submucosal dissection (ESD) for BAC and considered to have occurred and developed over a period of 4 years.

Case Report

A 68-year-old man underwent EGD for heartburn in April 2005 and Barrett's esophagus, 7 cm in length, was detected. The squamocolumnar junction and the esophagogastric junction were located at 31 cm and 38 cm from the incisor teeth, respectively. A flat reddish elevated lesion was found at 34 cm from the incisor teeth in the right-anterior wall of the lower thoracic esophagus (Fig. 1). Histological examination of the biopsy sample taken from this lesion verified well-differentiated adenocarcinoma. The lesion was resected by ESD in May 2005. Final histological diagnosis based on the findings of the resected specimen of ESD (Fig. 2) was well-differentiated adenocarcinoma with slight submucosal invasion (tumor depth of submucosal invasion was 500 μm) without lymphovascular involvement and free of tumor cells in the lateral/vertical histological margin. Lymph node metastasis was not recognized on computed tomography (CT). Surveillance with EGD and CT every 6 or 12 months was planned without additional operation after discussion with the patient.

Figure 1.

Conventional endoscopic view. A well-demarcated reddish depressed lesion was observed in the right-anterior wall of the lower thoracic esophagus.

Figure 2.

Fresh specimen resected by endoscopic submucosal dissection. The size of the specimen is 25 × 23 mm. A reddish depressed area was recognized in the center of the specimen. The area of adenocarcinoma corresponds to this depressed area and the size is 13 × 10 mm.

EGD carried out 3 years after ESD (May 2008) revealed a small reddish area at the 1-cm oral side of the ESD scar, which was located 33 cm from the incisor teeth (Fig. 3). Observation using narrow-band imaging with magnifying endoscopy (NBI-ME) revealed that the lesion was approximately 2 mm in diameter, and well demarcated with brownish coloration without an uneven surface (Fig. 4). The micro-surface structure of neighboring normal mucosa contained small round pits and that of the reddish area contained small villi. The lesion was considered to have inflammatory change because there were no endoscopic findings indicating malignancy such as inhomogeneity in the size of the villi. It was decided to follow up the lesion without biopsy because it might have been lost by biopsy.

Figure 3.

Conventional endoscopic view 3 years after endoscopic submucosal dissection (ESD), May 2008, taken at surveillance esophagogastroduodenoscopy. The ESD scar is indicated by a yellow arrow. Small reddish mucosa is located 1 cm oral to the scar.

Figure 4.

Narrow-band imaging with magnifying endoscopy view of the lesion shown in Figure 3. The lesion is well demarcated with brownish coloration and its size is approximately 2 mm.

EGD carried out 1 year later (May 2009) confirmed slight enlargement of this reddish area (Fig. 5). On NBI-ME, it was recognized as a well-demarcated slightly depressed brownish area, about 3–4 mm in diameter. The microstructure of the neighboring normal mucosa was homogeneous with small round pits, the same as the previous finding, and that of the reddish area contained small villi of heterogeneous size and irregular micro-vascularity (Fig. 6). Small adenocarcinoma arising from BE was suspected based on the endoscopic findings and biopsy revealed well-differentiated adenocarcinoma. ESD was done. Based on the findings of the resected specimen (Fig. 7), the size of the specimen was 20 × 15 mm and the lesion was confirmed as a slightly depressed area, 3 mm in diameter, in its center (Fig. 8). The microstructure of the neighboring mucosa in the fixed specimen contained small round pits the same as observed by NBI-ME and that of the depressed area was irregular with intermingling of small pits and small villi. The final histological diagnosis was well-differentiated adenocarcinoma limited to the upper muscularis mucosae (MM) in the double MM layer, without lymphovascular involvement and free of tumor for lateral/vertical histological margins (Fig. 9). Specialized columnar epithelium was not observed in the whole resected specimen.

Figure 5.

Conventional endoscopic view 4 years after endoscopic submucosal dissection (ESD), May 2009. The ESD scar is indicated by a yellow arrow. The reddish lesion is slightly enlarged.

Figure 6.

Narrow-band imaging with magnifying endoscopy view of the lesion shown in Figure 5. The size of this lesion is approximately 3 mm. The lesion was observed as a well-demarcated brownish area and contained irregular micro-surface and irregular microvascular structures.

Figure 7.

Fixed specimen resected by endoscopic submucosal dissection. The size of the specimen is 20 × 15 mm. A slightly depressed area was recognized in the center of this specimen.

Figure 8.

Magnified view of the depressed area in Figure 7. The depressed area contains heterogeneous structures compared with neighboring normal mucosa.

Figure 9.

Histopathological findings with hematoxylin-eosin staining of the depressed area. The area shown by the yellow line was diagnosed as well-differentiated adenocarcinoma. The muscularis mucosae (MM) is double layered, and the depth of cancer invasion is the shallow layer of the MM.

Review of the endoscopic images of the area where the minute cancer developed was carried out retrospectively. Figure 10 was taken 18 months after ESD (November 2006). Although the lesion of small BAC was located 1 cm from the oral side of the previous ESD scar (yellow arrow), there were no findings suggesting a malignant lesion in the area 1 cm oral to scar (white arrows). On the EGD 2 months after ESD (July 2005), there were also no changes (Fig. 11).

Figure 10.

Conventional endoscopic view 1.5 years after the first endoscopic submucosal dissection (ESD), November 2006. The scar after the first ESD is indicated by a yellow arrow. White arrows are considered to correspond to the area of the second cancer, located 1 cm oral to the scar after the first ESD.

Figure 11.

Conventional endoscopic view 2 months after the first endoscopic submucosal dissection (ESD), July 2005, taken by an oblique-viewing endoscope (GIF-XK240; Olympus Co., Tokyo, Japan). The ESD scar is indicated by a yellow arrow. It is impossible to identify the area where the second minute cancer developed.

Discussion

The incidence of BAC is increasing because of the decrease in the prevalence of Helicobacter pylori infection followed by the increasing risk of gastroesophageal reflux disease (GERD) and BE, which is considered to be a premalignant condition of BAC.[7] To improve the prognosis of BAC, it is important to detect it when it is amenable to radical treatment.[2] To achieve this, in Western countries, random biopsy by the surveillance endoscopy of BE is recommended.[1, 5] In Japan, screening endoscopy for gastric cancer is widespread and endoscopists have experience in detecting small adenocarcinoma. Early BAC is, in general, also detected through endoscopic observation.[7] However, most BAC found in Japan arises in short-segment BE and there is little experience in BAC detection in long-segment BE.[8]

We experienced a rare case of very small BAC, 3 mm in diameter, which developed in long-segment BE and was treated by ESD. The lesion was detected in surveillance EGD for BE after ESD for the first BAC. Endoscopy carried out 1.5 years after the previous ESD showed no change at the site where the second BAC developed retrospectively. Three years after the first ESD, the lesion could be recognized as a small reddish area of the mucosa about 2 mm in diameter. One year later, the lesion increased to 3 mm in diameter and its macroscopic type changed from flat type to depressed type. This particular case suggests that small BAC can be detected by surveillance EGD conducted at 1-year intervals with careful observation in combination with NBI-ME. To detect small BAC, it is most important to pay attention to endoscopic findings such as faint coloration and/or difference of microstructure.

Conflict of Interests

Authors declare no conflict of interests for this article.

Ancillary