Background
Gastrointestinal neuroendocrine tumor (NET) is a well differentiated neuroendocrine neoplasm commonly occurring in the rectum and appendix. In the latest World Health Organization (WHO) Classification, NET Grade 1 (G1) is defined as low-grade malignancy [
1]. Although adenocarcinomas showing neuroendocrine differentiation or those mixed with high-grade neuroendocrine carcinoma (NEC) are sometimes encountered, composite tumors comprising NET G1 and adenocarcinoma are exceedingly rare. To the best of our knowledge, only 10 cases of possible composite NET and adenocarcinoma arising in the colorectal region have been reported in the English literature [
2‐
6].
Here we report a case of composite NET G1 and adenocarcinoma of the rectum in which the two components showed histological transition.
Discussion
NET is a neuroendocrine neoplasm with low-grade malignancy, being common in the rectum and appendix [
7]. Histologically, the tumor cells have uniform rounded nuclei and granular cytoplasm, and show a solid, insular, acinar or trabecular growth pattern. Sometimes, they show rosette-like and/or glandular growth patterns and a rich vascular stroma. In the WHO Classification, NET G1, NET G2 and NEC are all classified as neuroendocrine neoplasms. However, NET and NEC have been considered to be different entities because of differences in histogenesis, malignant potential and gene alterations [
7]. Although association of NEC with adenoma or adenocarcinoma is not uncommon, composite tumors comprising NET G1 and adenocarcinoma are exceedingly rare. In the present case, NET G1 and a tubular adenocarcinoma were intermixed without a distinct boundary, and showed histologic transition between them. Accordingly, we concluded that this case was a composite tumor of NET G1 and adenocarcinoma. This case appeared to be different from “adenocarcinoid” or “goblet cell carcinoid”, most of which arise in the appendix. Adenocarcinoid exhibits histological features of both adenocarcinoma and NET, but is composed of mucin-containing goblet-shaped cells or signet-ring-like cells [
8‐
10].
To our knowledge, about 10 cases of composite NET and adenocarcinoma arising in the colorectal region have been reported [
2‐
6]. In addition to these colorectal cases, composite tumors of the stomach, small intestine, anal canal and gallbladder have also been documented [
4,
11,
12]. In some cases, however, it was not clear if the two components were intermixed, and therefore the possibility of collision tumor of NET and adenocarcinoma could not be excluded. Furthermore, some reports did not give information about the grade of NET, including details of the mitotic count and the Ki-67 labeling index, and thus the possibility of mixed adenoneuroendocrine carcinoma (MANEC) could not be ruled out. According to the WHO Classification (2010), MANECs have both gland-forming epithelial and neuroendocrine components, with one component exceeding 30%, and both of components are defined as carcinoma with high-grade malignancy [
1]. The present case was not a MANEC because the NET component was apparently a well differentiated NET G1 with a low mitotic index and a low Ki-67 labeling index. In addition, composite tumors comprising NET and adenocarcinoma should be distinguished from adenocarcinoma with focal neuroendocrine differentiation. Differences in clinical outcome can be expected between the present type of tumor, MANEC, and adenocarcinoma with focal neuroendocrine differentiation.
Another differential diagnosis that needs to be considered is glandular differentiation of NET. NET is known to show glandular differentiation or mucin production [
13]. In the present case, the glandular component showed apparent nuclear atypia, discrete gland formation, and positivity for CEA and CDX-2 without a “salt-and-pepper” chromatin pattern. In addition, the Ki-67 labeling index was apparently higher than that of a typical NET. Accordingly, we were able to exclude the possibility of glandular differentiation of NET. Furthermore, the possibility that the atypical glandular component could be an epithelial cell inclusion or ectopic glands, rather than adenocarcinoma, was ruled out because this component showed moderate atypia and a much higher Ki-67 labeling index than normal epithelial cells.
The histogenesis of composite NET and adenocarcinoma has not been fully elucidated. At least three possible hypotheses can be suggested: (a) Partial differentiation of adenocarcinoma into NET, (b) partial transformation of NET into adenocarcinoma, and (c) bidirectional transformation of common putative precursor cells into both endocrine cells and glandular epithelial cells. In the present case, neither of the specimens obtained by biopsy and ESD showed atypia in the surface epithelium, being incompatible with hypothesis (c). If the adenocarcinoma had partially differentiated into NET, then for former would have been observed in the surface epithelium. In addition, it has been reported that endocrine cells of the digestive tract are derived from local multipotent gastrointestinal stem cells, rather than migrating from the neural crest as reported previously [
14,
15]. Taken together, the findings suggest that this composite tumor might have differentiated from common putative precursor cells bidirectionally at an early stage of tumorigenesis.
The prognosis of composite NET and adenocarcinoma seems to be determined by the carcinoma component, although the number of reported cases is limited [
4]. In the present case, the adenocarcinoma had invaded to the superficial submucosal layer at its deepest point, but there was no vascular or lymphatic invasion. Eighteen months after ESD without additional therapy, the patient is doing well without any recurrence or metastasis.