Wang, Hanlin L. MD, PhD; Dhall, Deepti MD
Goblet cell carcinoid tumor is a unique mixed endocrine-exocrine neoplasm, almost exclusively seen in the appendix. It accounts for less than 5% of primary tumors of the appendix. Although recognized as a distinct clinicopathologic entity for 35 years, its histogenesis, diagnostic criteria, biologic behavior, and clinical management remain controversial.1 It is generally believed, however, that goblet cell carcinoid has the biologic behavior intermediate between classic carcinoid tumor and conventional adenocarcinoma, with a reported 5-year survival rate of 60% to 84%.2,3 At the time of diagnosis, over 50% of the tumors may have already invaded through the appendiceal serosa or into the mesoappendix. Approximately 15% to 30% of the cases may have already metastasized to the regional lymph nodes or distant organs with the ovaries being the most common metastatic site. Owing to its unpredictable malignant potential, many researchers advocate right hemicolectomy, and even prophylactic bilateral salpingo-oophorectomy for female patients regardless of age to reduce the risk of developing metastases,3–5 while others believe that appendectomy alone with a negative resection margin is sufficient if the tumor is confined to the appendix and exhibits low-grade histology.6 In cases with obvious spread of the disease, chemotherapy, mostly with 5-fluorouracil and leucovorin, is advised. Cytoreductive surgery with adjuvant intraperitoneal chemotherapy may also offer improved survival in cases with advanced peritoneal dissemination.3
There has been a lack of established clinicopathologic parameters that can be reliably used to predict which goblet cell carcinoid tumor will behave more aggressively especially when it is inadequately treated. As a general rule for all appendiceal carcinoids, metastasis usually occurs in tumors greater than 2 cm in size, and is almost never seen in those less than 1 cm. However, the size criterion may be difficult to apply to goblet cell carcinoid because it is usually a diffuse and infiltrative process and only rarely forms a discrete mass lesion. Other described parameters for predicting an aggressive behavior may include a high mitotic count (more than 2 mitotic figures per 10 high-power fields), a high Ki-67 labeling index, serosal or mesoappendiceal extension, angioinvasion, nodal involvement, increased number of Paneth cells, increased mucin secretion, and production of pancreatic polypeptide.3
In a recent retrospective study of 63 goblet cell carcinoids of the appendix, Tang et al7 showed that careful histologic evaluation is critically important in predicting the clinical outcome. On the basis of morphologic analysis, the researchers classified these tumors into 3 subgroups: typical goblet cell carcinoid (group A); adenocarcinoma ex goblet cell carcinoid, signet ring cell type (group B); and adenocarcinoma ex goblet cell carcinoid, poorly differentiated carcinoma type (group C).
Group A (typical goblet cell carcinoid) is characterized by well-formed goblet cells that are arranged in tight or ill-defined clusters or a cohesive linear pattern. The tumor cells exhibit no significant or only minimal cytologic atypia. Mitotic figures are rarely identifiable. There is minimal individual cell infiltration of the stroma. There is minimal distortion of the appendiceal wall with no or minimal desmoplasia (Figs. 1A to 1C). This group consisted of 30 cases in this study. Nine tumors presented T4 disease and 19 presented T3 disease at diagnosis if the staging parameters of the American Joint Committee on Cancer for appendiceal adenocarcinoma were followed. Four had nodal metastasis and 10 (33%) had distant metastasis.
Group B (adenocarcinoma ex goblet cell carcinoid, signet ring cell type) is characterized by goblet or signet ring cells that are arranged in irregular large clusters, discohesive single file or single cell infiltrating pattern. There is a lack of confluent sheets of cells. The tumor cells exhibit significant cytologic atypia with irregular and hyperchromatic nuclei. Desmoplasia is present with associated destruction of the appendiceal wall. There is a partial or near complete loss of the goblet cell clusters (Figs. 2A to 2C). This group consisted of 26 cases in this study. Twenty-three tumors presented T4 disease and the remaining 3 presented T3 disease at the diagnosis. Eleven had nodal metastasis including 2 with N2. Twenty-three (88%) had distant metastasis.
Group C (adenocarcinoma ex goblet cell carcinoid, poorly differentiated carcinoma type) is characterized by the presence of poorly differentiated adenocarcinoma. The adenocarcinomatous component should be larger than 1 low-power field or 1 mm2 in size, and can be either gland forming, confluent sheets of signet ring cells, high-grade neuroendocrine carcinoma, or undifferentiated carcinoma. It typically constitutes the major portion of the tumor but there should be at least focal evidence of goblet cell morphology (Figs. 3A to 3C). This group consisted of 7 cases in this study. All showed T4 disease at diagnosis and all had distant metastasis. Three also had nodal metastasis.
The researchers further demonstrated that the histologic classification correlated with patient survival (Table 1). The overall disease-specific survival was 77% for the entire study group, with a mean follow-up of 49±5 months (range: 8 to 191 mo), but was 96%, 73%, and 14% for groups A, B, and C, respectively. The 3-year and 5-year disease-specific survival rates were 100% and 100% for group A, 85% and 36% for group B, and 17% and 0% for group C. The mean survival time was close to 10 years for group A, but was 43±6 months for group B, and 31±6 months for group C. Most group A patients (86%) survived without evidence of disease but many group B patients (58%) lived with disease. When only the patients with distant metastasis (stage IV) were analyzed, the 3-year and 5-year survivals were still 100% for group A, with only 1 patient died after 119 months. The 3-year and 5-year survivals were 82% and 38% for group B. Again, group C patients had the worst outcome, comparable with those with stage-matched conventional adenocarcinoma of the appendix.
This study demonstrates that histologic stratification of appendiceal goblet cell carcinoid tumors into different subgroups is critically important in determining the prognosis and thus in guiding the clinical management (Table 2). It is not totally surprising, however, that the group C patients would do poorly because these tumors are essentially mixed carcinoid-adenocarcinomas as classified by the World Health Organization.8 It has been shown that the adenocarcinomatous component of the tumor determines the prognosis, which would be worse than for a pure goblet cell carcinoid.9 The immunohistochemical findings and stage-matched survival data presented in the study by Tang et al7 provide more evidence to support this notion.
More importantly, the study by Tang et al demonstrates that the goblet cell carcinoid tumors without a clear-cut component of conventional adenocarcinoma should be stratified into 2 subgroups: typical goblet cell carcinoid (group A) and adenocarcinoma ex goblet cell carcinoid, signet ring cell type (group B), because group A patients clearly have a far more favorable long-term prognosis than group B patients, regardless of the presence of distant metastasis. The separation is purely on histologic grounds. According to the researchers, the hallmarks of group B are partial or near complete loss of clustered goblet cell architecture with single discohesive infiltrating cells, large irregular clusters of signet ring (or goblet) cells, and significant cytologic atypia. Unfortunately, there is no immunomarker that can be used to help the separation. The Ki-67 proliferative index, which has been recommended as a useful prognostic indicator for gastrointestinal carcinoids,10,11 was similarly low in both groups (11%±2 and 16%±4 for groups A and B, respectively; P=0.27), relative to 80% or greater than 80% for group C tumors and conventional appendiceal adenocarcinomas.
Therefore, the distinction between groups A and B is much based on histologic interpretation by pathologists whether the tumor cells are goblet or signet ring cells. Like any other challenging diagnosis in surgical pathology, this would certainly create interobserver and even intraobserver variations because the separating histologic features are quite subtle. Apparently, the researchers recognized the difficulty and an interobserver variability analysis was performed, which showed a high degree of concordance.7 This suggests that with the knowledge of the histologic features, an accurate classification can be achieved. It will be interesting to see, however, if the classification can be reproduced by other investigators.
As part of a gastrointestinal neuroendocrine tumor referral center, we have encountered a relatively large number of goblet cell carcinoids of the appendix and frequently face the challenge of how to classify these tumors. We found a general lack of agreement among gastrointestinal pathologists on whether a given cluster of tumor cells should be interpreted as signet ring cells or goblet cells. This difficulty is illustrated in the following 4 examples that we encountered recently.
Case 1 was a 62-year-old man who underwent appendectomy and cecectomy for an appendiceal tumor. Sections showed that most of the tumor consisted of small and regular clusters of goblet cells without significant cytologic atypia, typical of goblet cell carcinoid. However, there were several small areas where tumor cells lost their clustered architecture and were distributed as single cells (Fig. 4). Cytologically, these cells were not dissimilar to those in the clusters, without a sharp increase in the degree of cytologic atypia. Some pathologists still considered these cells as goblet cells because of the lack of cytologic atypia and felt that the single cell pattern probably resulted from sectioning of the clusters at different angels, but others believed that they were no different from signet ring cells and felt that the presence of that many single cells could not be explained by tissue sectioning. This case presented as T3 disease with no clinical evidence of distant metastasis. No lymph node was retrieved in the specimen.
Case 2 was a 43-year-old woman who underwent appendectomy that showed what appeared to be a goblet cell carcinoid in the major portion of the tumor with prominent extracellular mucin production (Fig. 5A). However, there were large irregular clusters of tumor cells that did not form confluent sheets (Fig. 5B). The tumor cells in these clusters exhibited goblet or signet ring morphology and appeared to show slightly more significant nuclear atypia (Fig. 5C). Interestingly, immunostaining for Ki-67 showed a relatively high proliferative index (∼50%) throughout the tumor (Fig. 5D), which did not distinguish the irregular clusters from other parts of the tumor. This case was interpreted as adenocarcinoma ex goblet cell carcinoid signet ring cell type by some but simply goblet cell carcinoid or frank signet ring cell adenocarcinoma by others. This case also presented as T3 disease with no clinicopathologic evidence of nodal or distant metastasis.
Case 3 was a 66-year-old woman who presented with pseudomyxoma peritonei. Appendectomy and right hemicolectomy showed a primarily mucinous adenocarcinoma consisting predominantly of pools of mucin lined by columnar mucinous epithelium intermixed with well-formed infiltrating glands (Fig. 6A) and numerous signet ring cells floating in the mucin pools (Fig. 6B). In addition, a focus of classic carcinoid tumor and interspersed clusters of what appeared to be goblet cell carcinoid were noted (Fig. 6C). The tumor cells that displayed a goblet cell architecture exhibited cytologic atypia similar to those with a signet ring morphology (Fig. 6D), and occasionally showed positive staining for chromogranin and synaptophysin. This case was interpreted as adenocarcinoma ex goblet cell carcinoid signet ring cell type by some but as composite carcinoid-adenocarcinoma by others who felt that the cells with a goblet cell architecture were also signet ring cells. This tumor presented as T4, N1, and M1 disease with a wide spread to the ovaries and the peritoneum.
Case 4 was a 78-year-old man who underwent appendectomy for acute appendicitis. Sections showed that in addition to acute suppurative appendicitis, there was a diffuse infiltrative growth of neoplastic cells forming small tight clusters in a linear concentric pattern (Fig. 7A). Intracytoplasmic mucin was frequently noted in tumor cells, giving rise to a goblet cell morphology. Mitotic figures were not detected and Ki-67 immunostain showed a ∼5% proliferative index throughout. This case was interpreted as typical goblet cell carcinoid by some because of lack of large irregular clusters, lack of excessive single cells and lack of significant cytologic atypia, but as adenocarcinoma ex goblet cell carcinoid signet ring cell type by others because of distorted and irregular goblet cell clusters, scattered individual signet ring cells and too much cytologic atypia to be categorized as type A (Fig. 7B). This tumor presented as T3 disease with a negative appendectomy margin. It was unclear whether the patient had nodal or distant metastasis but right hemicolectomy was recommended.
As our patients have been followed up for only a short period of time, it is impossible at present to determine which diagnosis is most appropriate in predicting their clinical outcome. However, these cases highlight the great challenge to separate goblet cells from signet ring cells. It is anticipated that with the rapid advance in molecular pathology, useful molecular and immunohistochemical markers will emerge to facilitate the separation which is apparently critically important according to the study by Tang et al.7
On a separate note, the terms adenocarcinoma ex goblet cell carcinoid signet ring cell type and adenocarcinoma ex goblet cell carcinoid poorly differentiated carcinoma type seem somewhat less ideal because the presence of signet ring cells generally signifies a poorly differentiated histology. In addition, adenocarcinoma ex goblet cell carcinoid poorly differentiated carcinoma type also includes cases with confluent sheets of signet ring cells, which can be confused with adenocarcinoma ex goblet cell carcinoid signet ring cell type. As a result, some of the group B tumors may be inadvertently classified as group C tumors which carry a much worse prognosis, or vice versa. In contrast, the adenocarcinomatous component in group C tumors can be gland forming, which theoretically should also include those with a well or moderately differentiated morphology.
1. Volante M, Righi L, Asioli S, et al. Goblet cell carcinoids and other mixed neuroendocrine/nonneuroendocrine neoplasms. Virchows Arch. 2007;451:S61–S69.
2. McCusker ME, Coté TR, Clegg LX, et al. Primary malignant neoplasms of the appendix: a population-based study from the surveillance, epidemiology and end-results program, 1973-1998. Cancer. 2002;94:3307–3312.
3. Pahlavan PS, Kanthan R. Goblet cell carcinoid of the appendix. World J Surg Oncol. 2005;3:36.
4. Goede AC, Caplin ME, Winslet MC. Carcinoid tumour of the appendix. Br J Surg. 2003;90:1317–1322.
5. Toumpanakis C, Standish RA, Baishnab E, et al. Goblet cell carcinoid tumors (adenocarcinoid) of the appendix. Dis Colon Rectum. 2007;50:315–322.
6. Varisco B, McAlvin B, Dias J, et al. Adenocarcinoid of the appendix: is right hemicolectomy necessary? A meta-analysis of retrospective chart reviews. Am Surg. 2004;70:593–599.
7. Tang LH, Shia J, Soslow RA, et al. Pathologic classification and clinical behavior of the spectrum of goblet cell carcinoid tumors of the appendix. Am J Surg Pathol. 2008;32:1429–1443.
8. Capella C, Solcia E, Sobin LH, et al. Endocrine tumors of the appendix. In: Hamilton SR, Aaltonen LA, eds. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of the Digestive System. Lyon: IARCP Press; 2000: 99–101.
9. Carr NJ, Sobin LH. Neuroendocrine tumors of the appendix. Semin Diagn Pathol. 2004;21:108–119.
10. Rorstad O. Prognostic indicators for carcinoid neuroendocrine tumors of the gastrointestinal tract. J Surg Oncol. 2005;89:151–160.
11. Cunningham JL, Grimelius L, Sundin A, et al. Malignant ileocaecal serotonin-producing carcinoid tumours: the presence of a solid growth pattern and/or Ki67 index above 1% identifies patients with a poorer prognosis. Acta Oncol. 2007;46:747–756.
© 2009 Lippincott Williams & Wilkins, Inc.