Tubular carcinoids of the appendix: the CK7/CK20 immunophenotype can be a diagnostic pitfall
- 1Division of Gastrointestinal Pathology, The Johns Hopkins Hospital, Baltimore, Maryland, USA
- 2Department of Pathology, The Johns Hopkins Hospital, Baltimore, Maryland, USA
- Correspondence to Dr Karen E Matsukuma, Division of Gastrointestinal Pathology, The Johns Hopkins Hospital, Weinberg 2242 Pathology, 401 North Broadway, Baltimore, MD 21287, USA;
Contributors KEM and EAM reviewed and agree with the contents of the manuscript.
- Accepted 21 February 2012
- Published Online First 29 March 2012
Aims Tubular carcinoid is a rare variant of appendiceal well-differentiated neuroendocrine tumour. Although considered benign lesions, the small infiltrating tubules that characterise the tumour may raise concern for metastatic adenocarcinoma. To our knowledge, the cytokeratin 7 (CK7)/cytokeratin 20 (CK20) expression profile of these neoplasms remains unexplored.
Methods The authors characterised the CK7/CK20 immunophenotype and Ki-67 expression of the eight available tubular carcinoids seen at their institution from 1991 to 2011.
Results CK7 and CK20 staining was variable, ranging from none to focal staining for either or both CK7 and CK20, to diffuse expression of CK7 or CK20.
Conclusions The CK7/CK20 expression profile is of limited value when the differential diagnosis includes primary tubular carcinoid and well-differentiated metastatic adenocarcinoma. In such cases, careful attention to the location of the neoplasm, mitotic count and presence or absence of an associated classic carcinoid component are more useful for arriving at the correct diagnosis.
- GI neoplasms
- gastrointestinal disease
- soft tissue tumours
- soft tissue
Tubular carcinoid is a rare variant of appendiceal well-differentiated neuroendocrine tumour. Unlike the classic carcinoid tumours of the appendix, tubular carcinoids are characterised by small, widely separated tubules composed of cuboidal to low columnar cells with basally oriented nuclei, indistinct nucleoli and luminal mucin droplets. In addition, they often stain weakly and/or focally for neuroendocrine markers.1 ,2 Tubular carcinoids are a biologically distinct form of appendiceal carcinoid tumour and, unlike goblet cell carcinoids, follow a uniformly benign course. However, the small infiltrating tubules that characterise the tumour may on occasion raise concern for metastatic adenocarcinoma.
In the largest study to date, Burke and colleagues reviewed 17 appendiceal tubular carcinoids and demonstrated their unique capacity to express glucagon, whereas the other carcinoids (ie, classic, goblet cell) generally did not.3 ,4 Although this finding was substantiated by follow-up studies showing specific expression of glucagon mRNA in tubular carcinoids,5 other investigators, using a different glucagon antibody, were not able to demonstrate glucagon expression in their cases.6 Thus, immunostaining for glucagon may not be a reliable method for identifying tubular carcinoids. Furthermore, not all diagnostic pathology laboratories offer glucagon in their immunohistochemistry menus.
Given the paucity of tumour-specific markers for tubular carcinoids, a surgical pathologist may consider using CK7 and CK20 immunostains to help pinpoint the origin of the tumour, if the differential diagnosis includes well-differentiated metastatic adenocarcinoma. However, to date, no study has determined the CK7/CK20 expression profile of appendiceal tubular carcinoids.
Materials and methods
This study was approved by the Johns Hopkins Medicine Institutional Review Board. Hospital electronic pathology files were queried for ‘appendiceal carcinoid tumour with tubule formation’ or ‘tubular carcinoid.’ Ten cases were identified from 1991 to 2011: two routine surgical specimens and eight consult cases. Paraffin blocks or unstained slides were available for eight cases. CK7, CK20 and Ki-67 stains were performed according to standard protocols using the following antibodies: CK7 (clone OV-TL, 1:500 dilution; DAKO (Carpinteria, CA, USA)), CK20 (clone Ks20.8, 5 ml prediluted; DAKO) and Ki-67 (clone 30-9, prediluted; Ventana (Tucson, AZ, USA)).
Demographic information is summarised in table 1. No patients were known to have a history of adenocarcinoma at another body site.
The average size of the appendiceal tumours was 5.2 mm (range 3–10 mm). In the six cases in which the location of the tumour was documented, all were centred in the tip or the distal one-half of the appendix.
Microscopically, the tumours showed widely dispersed, small round tubules composed of cytologically bland cuboidal to low columnar epithelial cells (figure 1). In 9 of 10 cases, intraluminal mucin was present. No mitotic figures were identified in any of the cases. In six of eight cases, a minor classic carcinoid component composed of small nests and trabeculae was identified interspersed among the neoplastic tubules. In all cases, the tumours stained at least focally for chromogranin and/or synaptophysin. (Neuroendocrine stains were performed prior to this study.) No surface epithelial component was seen in any of the cases. Seven of nine tumours extended into the muscularis propria, with one infiltrating into the subserosal fat.
The clinical and immunohistochemical profile for the tubular carcinoids is summarised in table 1. The expression of CK7 ranged from none, to focal, to diffuse; CK20 expression also ranged from none, to focal, to diffuse. Most cases showed at least focal positivity for either CK7 or CK20 (figure 2). The Ki-67 expression in most cases was ≤2%; however, in one case the proliferative index focally reached 30% (figure 3).
There was no evidence of recurrent or disseminated disease in the eight patients with available clinical follow-up (mean 42 months; table 1).
Tubular carcinoids are rare low-grade appendiceal neuroendocrine tumours often discovered as incidental findings in young adults presenting with acute appendicitis. Although there are few reports in the literature, each with a small number of cases, there is overall agreement that tubular carcinoids are benign neoplasms. To date there have been no reports of metastasis or tumour-related deaths.1 ,3 ,4
From a clinical and morphological standpoint, the 10 tubular carcinoids in our study mirrored those reported in the literature in that they were all small incidental neoplasms that occurred in the distal half of the appendix, showed a low mitotic index and had a uniformly excellent prognosis. We found that expression of CK7 and CK20 varied from none, to focal, to diffuse and showed no predominant pattern. These immunohistochemical findings, while somewhat intellectually dissatisfying, make clear the danger of relying on CK7 and CK20 immunolabeling to characterise a tubular neoplasm in the appendix.
In case 7, the patient's older age at presentation, the tumour's involvement of the subserosa and the Ki-67 proliferative index of up to 30% in one focus were somewhat more worrisome features for metastatic adenocarcinoma than those of the other nine cases. However, although subserosal involvement was unusual in our cohort, it was a common feature in the cases described by Warkel and colleagues (six of nine cases),1 and, importantly, none of those resulted in recurrence or metastasis. This is particularly noteworthy in that mesoappendiceal invasion has been shown to be an independent risk factor for metastasis in classic appendiceal carcinoid.7 While the Ki-67 proliferative rate of the aforementioned tumour was also focally increased, no mitotic figures were identified. One possible explanation for the increased proliferative rate is that the significant acute appendicitis (unique to this case) with neutrophils infiltrating the neoplastic epithelium and tubular lumens caused a reactive proliferative response. Furthermore, because it is recommended that at least 500 neoplastic cells be reviewed to assign a Ki-67 proliferative index8 and because most tubular carcinoids measure less than 1 cm and form only a minority of the tumour mass, Ki-67 may not be an appropriate measure of proliferation in these tumours. Additional studies and longer clinical follow-up will be necessary to determine whether Ki-67 is a reliable prognostic marker in tubular carcinoids. With regard to the clinical outcome, no history of adenocarcinoma at another site was elicited, and the patient has had no evidence of disease at last follow-up (7 months). While this would be an inadequate follow-up period to detect an adverse outcome attributable to a small primary appendiceal neuroendocrine neoplasm, it is a fully adequate period to exclude a metastasis to the appendix from another source.
Although metastasis to the appendix is exceedingly rare, several cases have been documented in the literature,9–11 and in most of those cases, the patient came to clinical attention for ‘appendicitis’. Given the infiltrative pattern of tubules in appendiceal tubular carcinoid on the one hand, and the prevalence of breast and prostate cancer on the other, it is appropriate to keep metastatic adenocarcinoma in the differential diagnosis of neoplasms of the appendix. However, the practising pathologist should be aware that the CK7/CK20 expression pattern of an appendiceal neoplasm is of limited value when the differential diagnosis includes primary tubular carcinoid and well-differentiated metastatic adenocarcinoma. Instead, features such as the location of the neoplasm (ie, primarily in the mucosa/submucosa vs primarily in the subserosa/serosa), mitotic count and presence or absence of an associated classic carcinoid component are more useful for arriving at the correct diagnosis.
Tubular carcinoids are benign, often incidental neoplasms of the vermiform appendix.
The CK7/CK20 expression profile of tubular carcinoids varies widely and shows no predominant pattern.
When the differential diagnosis includes well-differentiated metastatic adenocarcinoma, a location primarily in the mucosa/submucosa (vs subserosa/serosa), a low mitotic count, and the presence of a classic carcinoid component favor the diagnosis of tubular carcinoid over well-differentiated metastatic adenocarcinoma.
Competing interests None.
Ethics approval Ethics approval was provided by Johns Hopkins Medicine Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.