Small cell carcinomas in gastrointestinal tract: immunohistochemical and clinicopathological features
- Anna Fen-Yau Li1,2,
- Alice Chia-Heng Li3,
- Chih-Yi Hsu2,
- Win-Yin Li2,
- Han-Shui Hsu4,5,
- Jeou-Yuan Chen6
- 1School of Medicine, National Yang-Ming University, Taipei, Taiwan
- 2Department of Pathology, Taipei Veterans General Hospital, Taiwan
- 3The Webb School, Claremont, California, USA
- 4Institute of Emergency and Critical Care Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan
- 5Division of Thoracic Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan
- 6Institute of Biomedical Sciences, Academia Sinica, Taiwan
- Correspondence to Dr Anna Fen-Yau Li, Department of Pathology, Taipei Veterans General Hospital, No. 201, Sec. 2 Shih-Pai Road, Taipei 112, Taiwan;
- Accepted 29 March 2010
- Published Online First 7 June 2010
Aims To test the incidence of the expression of the immunohistochemical markers that aid diagnosis of gastrointestinal tract small cell carcinoma (GI-SmCC) and to evaluate the incidence of mixed endocrine–exocrine carcinomas in GI-SmCC.
Methods Immunohistochemical studies of three antibodies against epithelial markers (CK8, AE1/AE3, EMA), four neuroendocrine differentiation markers (synaptophysin (Syn), neuron specific enolase (NSE), neuronal cell adhesion molecules (CD56), chromogranin A (CgA)), and a transcription factor (thyroid transcription factor 1 (TTF-1)) were performed. The incidence of non-endocrine carcinoma component was evaluated in 42 GI-SmCCs (11 in the oesophagus, 15 in the stomach, 15 in the colon, and 1 in the small intestine).
Results The percentages of GI-SmCC with positive immunoreactivity were: CK8 92.9%, AE1/AE3 76.2%, EMA 71.4%, Syn 100%, NSE 100%, CD56 90.5%, CgA 61.9%, TTF-1 21.4%. The low molecular weight cytokeratin CK8 is more commonly expressed in GI-SmCC than is the expression of AE1/AE3 or EMA. Synaptophysin and NSE are expressed in all GI-SmCCs studied. Non-endocrine carcinoma components were demonstrated in 8 patients (4 in the oesophagus and 4 in the stomach).
Conclusion In detecting GI-SmCC, epithelial marker CK8 is more sensitive than AE1/AE3 or EMA, and neuroendocrine differentiation markers synaptophysin and NSE are the most useful markers. TTF-1 positivity is not uncommon in GI-SmCC, but cases with negative TTF-1 staining may indicate an extra-pulmonary primary. Non-endocrine carcinoma components were demonstrated in about 30% of oesophagus and stomach SmCC; the neoplasms should be diagnosed as mixed endocrine–exocrine carcinoma.
- Gastrointestinal tract
- small cell carcinoma
- immunohistochemical study
- gastrointestinal disease
- neuroendocrine tumours
- surgical pathology
Small cell carcinoma (SmCC) is a highly aggressive malignancy that occurs most frequently in the lung. Gastrointestinal tract small cell carcinoma (GI-SmCC) is rare and may cause difficulties in diagnosis.1 The diagnosis of SmCC is aided by immunohistochemical staining. Systemic treatment with chemotherapy recommended for GI-SmCC (cisplatin and etoposide or other chemotherapeutic agents) is different from that recommended for adenocarcinoma or squamous cell carcinomas.2 Therefore, it is important to make the correct diagnosis for GI-SmCC patients. A panel of immunohistochemical studies may be necessary in differentiating poorly-differentiated tumours, such as poorly-differentiated squamous cell carcinoma, poorly-differentiated adenocarcinoma, undifferentiated carcinoma, malignant melanoma, lymphoma, and secondary tumours, from SmCCs.3–6 Cytokeratin (AE1/AE3), EMA, leucocyte common antigen (LCA), and neuroendocrine differentiation markers, including synaptophysin (Syn), chromograin A (CgA), neuron specific enolase (NSE) and neuronal cell adhesion molecule (CD56) are commonly used in this situation. In two studies, immunohistochemical detection of AE1/AE3 was reported in only 57.1% and 71.4%, respectively, of oesophageal small cell carcinomas.7 8 We would like to establish the incidence of immunohistochemical detection of CK8 (cytokeratin 8, low molecular weight cytokeratin, 52.5 kDa), AE1/AE3, EMA, neuroendocrine differentiation markers, and a transcription factor (TTF-1) in GI-SmCC, and to determine which epithelial antigen and neurendocrine differentiation markers are more sensitive in detecting GI-SmCC.
Mixed endocrine–exocrine carcinoma should be considered in the presence of non-endocrine carcinoma components with endocrine carcinoma. This condition constitutes a diagnostic challenge, as often only one component of the tumour is identified, which leads to an incomplete diagnosis and suboptimal treatment. A second histology component is common in small cell lung carcinoma (SCLC), and it has been reported that discrete areas of non-small-cell morphological components can be found in up to 30% of SCLCs.9 10 The incidence of mixed endocrine–exocrine carcinomas in GI-SmCCs is not well established yet.
We retrospectively examined the records and pathology sections of GI-SmCC patients treated at Taipei Veterans General Hospital, Taiwan, between 1989 and 2009. The following issues were investigated: (1) the immunohistochemical staining incidence of three epithelial markers (CK8 (CAM-5.2), AE1/AE3, EMA), four neuroendocrine differentiation markers (NSE, synaptophysin, chromogranin A, CD56), and TTF-1; and (2) the incidence of associated second histological components in GI-SmCC. Two possible pitfalls in the diagnosis of GI-SmCC—negative staining for AE1/AE3, EMA in some cases, and presence of a non-endocrine carcinoma component with the possibility of mixed endocrine–exocrine carcinoma—were emphasised.
Materials and methods
Patients and clinical data
The case records of GI-SmCC treated between January 1989 and December 2009 were retrieved from the surgical pathology files and medical records of the Taipei Veterans General Hospital. Sections were reviewed according to the World Health Organization (WHO) classification of tumours of the lung, pleura, thymus and heart, and the WHO classification of tumours of the digestive system.11 12 Small cell carcinomas (poorly differentiated endocrine carcinomas) of the oesophagus, stomach and colon are indistinguishable from their counterparts in the lung in histological and immunohistochemical features. Large cell neuroendocrine carcinomas of the gastrointestinal tract have not been well described because of their apparent low frequency.12 In order to avoid confusion, we decided to focus on GI-SmCC in this study. Cases of small cell carcinoma were diagnosed based on both histological and immunohistochemical features; at least two of the four neuroendocrine differentiation markers had to be positive in order to confirm the diagnosis and be included in this study.
For inclusion as cases of primary GI-SmCC, a normal chest radiograph, CT scan of the chest, sputum cytology, and/or negative bronchoscopy was required. In the absence of an established staging system for extrapulmonary small cell carcinomas, lesions were staged according to the Veterans' Administration Lung Study Group (VALSG) staging system for primary SCLCs. This staging system consists of two categories: (1) limited disease (LD), defined as a tumour contained within a localised anatomic region, with or without regional lymph node involvement; and (2) extensive disease (ED), defined as a tumour outside loco-regional boundaries.13 14 Patients were divided into LD and ED groups for each primary organ site. Mixed endocrine–exocrine carcinomas were defined as tumours containing both SmCC and non-endocrine carcinoma components, regardless of the proportion of the latter.15 Mixed SmCCs should be suspected in a careful and thorough searching of the routine H&E staining sections. All patients with an adenocarcinoma component had either a distinct glandular component on H&E stained sections or mucin as demonstrated by mucicarmine or Periodic acid-Schiff stains. Cases of suspected squamous component were confirmed by cytokeratin 5/6 and p63 immunohistochemical staining. For the retrieved cases the following factors were considered: age, gender, anatomic location, extent of disease, treatment, survival and second histological components.
Sections (3 μm) of formalin-fixed paraffin-embedded tissue were cut onto silane-coated slides, incubated at 70°C for 20 min, dewaxed and rehydrated by successive immersion in xylene, graded ethanol and water. Immunohistochemical analysis was performed with antibodies against the following proteins: CK8, AE1/AE3, epithelial membrane antigen (EMA), synaptophysin, neuron-specific enolase, CD56, chromogranin A, TTF-1, p63, and cytokeratin 5/6. Table 1 lists clones, antigen retrieval methods, dilution and commercial sources of the antibodies. The stains were done on a Leica Bond automated system. Endogenous peroxidase was quenched by hydrogen peroxide, and non-specific adsorption was reduced by blocking against swine serum. The bound antibodies were detected by the biotin-free Novolink Max Polymer Detection System (Novocastra). GI-SmCCs may express AE1/AE3, CK8 and EMA in a diffuse strong pattern; some may express it in a patchy fashion in focal areas, and if the stain looks convincing, we would regard the staining as positive result. Tumors were diffusely strong positive in immunostaining for neuroendocrine markers, and were different from carcinoma with neuroendocrine differentiation.
Statistical analysis was performed on the relationship between clinical parameters, including gender, age, tumour location, disease stage, patient survival, and eight immunohistochemical markers by the two-sample Kolmogorov-Smirnov test using SPSS V.10.0, or Fisher's exact test or the χ2 test when appropriate. Overall survival was calculated as the time from diagnosis to death or the time to the last follow-up appointment for surviving patients. Statistical significance was defined as p<0.05.
Forty-two GI-SmCCs were retrieved and included in this study (11 oesophageal SmCCs, 15 gastric SmCCs, 15 colorectal SmCCs and one small intestine SmCC). There were 37 men and 5 women, with a male:female ratio of 7.4:1, and median age of 64, ranging from 32 to 84 years. All the GI-SmCC patients underwent combined surgery and chemotherapy, with the exception of 1 LD gastric SmCC patient and 3 ED colorectal patients, who underwent surgery only. GI-SmCC is a highly aggressive malignancy, and the prognosis is poor. The median survival for SmCCs in the oesophagus, stomach and small intestine were 27.7±14.8 months, 21.3±29.0 months and 8.00 months, respectively. All the colorectal SmCC patients had ED at the time of diagnosis, with a median survival of 5.1±3.1 months, and all of the colorectal SmCC patients died within 1 year (table 2).
Immunohistochemical staining results
Immunohistochemical staining for three epithelial markers (CK8, AE1/AE3, EMA), four neuroendocrine differentiation markers (NSE, Syn, CD56, CgA) and TTF-1 were performed on the 42 GI-SmCCs; table 3 summarises the results. Immunological reactivity of the samples was CK8 92.9%, AE1/AE3 76.2%, EMA 71.4%, Syn 100%, NSE 100%, CD56 90.5%, chromogranin-A 61.9% and TTF-1 21.4%. The incidence for CK8 staining in GI-SmCC is significantly higher than AE1/AE3 (p=0.027), and also significantly higher than EMA (p=0.0086). Immunological reactivity of AE1/AE3, EMA and CK8 was negative in 23.8%, 28.6% and 7.1%, respectively, of GI-SmCC lesions. No immunological relativities with any of the CK8, AE1/AE3 or EMA antibodies in GI-SmCCs were unusual and were only demonstrated in two cases, one oesophageal SmCC and one gastric SmCC. The incidence for immunoreactivity of the Syn and NSE is 100% in our patients. The incidence for immunoreactivity of the TTF-1 is relatively common in oesophageal SmCCs, and less common in stomach and colon SmCCs. In the cases of mixed endocrine–exocrine carcinomas, the non-neuroendocrine compartment showed positive AE1/AE3, CK8 and EMA staining, but negative neuroendocrine marker staining. Most of the SmCC compartment showed positive staining for both epithelial and neuroendocrine markers (figure 1), but some was negative for epithelial markers, such as AE1/AE3 staining (figure 2).
Associated second histology component
Eight mixed endocrine–exocrine carcinomas were demonstrated out of the 42 GI-SmCCs. A second histology component was shown in 8 cases, including a squamous cell carcinoma component in 4 of the 11 oesophageal SmCCs, and an adenocarcinoma component in 4 of the 15 gastric SmCCs. Composite mixed endocrine–exocrine carcinoma with intimate admixture of neoplastic glandular exocrine and neuroendocrine cell types was demonstrated in one gastric mixed endocrine–exocrine carcinoma. Small cell carcinoma was the dominant component, and the non-endocrine component was present in limited areas close to the epithelium and sub-epithelial area in the other 7 mixed endocrine–exocrine carcinomas (figure 1 and figure 2). No second histology component was demonstrated in the SmCCs in the small intestine or colorectum.
GI-SmCC is a highly aggressive malignancy and there is no recommended standard treatment at this time.16–18 With advances in imaging techniques, more of the endocrine tumours of the gastrointestinal tract are being removed and sent for pathological evaluation. It is important for the diagnostic pathologists to be aware of these tumours and the pitfalls of the diagnosis. Immunohistochemical studies of three epithelial markers (CK8, AE1/AE3, EMA), four neuroendocrine markers (NSE, synaptophysin, chromogranin A, CD56) and TTF-1, were performed in 42 GI-SmCCs in this study. Epithelial marker CK8, expressed in over 92% of GI-SmCC, is more sensitive than AE1/AE3 or EMA in detecting GI-SmCC. Neuroendocrine differentiation markers synaptophysin and NSE are almost always positive in GI-SmCC. Expression of TTF-1 in GI-SmCCs is not uncommon. A non-endocrine carcinoma component is demonstrated in about 30% of oesophagus and stomach SmCC, and is uncommon in colorectal SmCC. Due to the small number of cases in our study, statistical analysis was performed, but without significance in the clinical features of SmCCs of various locations in the gastrointestinal tract. The discussion will focus on immunohistochemical analysis and the second histological component in the GI-SmCCs.
Immunohistochemical staining results
The diagnosis of GI-SmCC is aided by immunohistochemical staining. The incidence of expression of epithelial marker CK8 is much higher than is the expression of AE1/AE3 or EMA in GI-SmCCs. Immunohistochemical studies for leucocyte common antigen (LCA) and AE1/AE3 were used routinely as the markers in the differential diagnosis between lymphoma and poorly differentiated carcinoma. However, about 25% of GI-SmCC lesions were negative for AE1/AE3. Immunohistochemical studies for CK8 and the neuroendocrine markers should be performed in cases with neuroendocrine tumour morphology, even with negative reactivity with AE1/AE3 or EMA. This is also very useful in endoscopic biopsy specimens (figure 3). The incidence of positive immunohistochemical reactivity for synaptophysin, NSE, CD56 and CK8 was high, and was useful in the diagnosis for GI-SmCC. TTF-1 expression has been reported to be positive in nearly all small cell lung carcinomas. A negative TTF-1 staining in SmCC may indicate an extra-pulmonary primary.
Associated second histology component and mixed endocrine–exocrine carcinoma
Recognising both endocrine and non-endocrine carcinoma components, and employing immunohistochemical studies are important for a correct diagnosis and optimal treatment for mixed endocrine–exocrine carcinoma. Mixed endocrine–exocrine carcinoma in the gastrointestinal tract is a well known entity; frequency of the associated second histology component has been reported as rare to up to 40% in oesophageal SmCCs.19 20 The possible reasons for the discrepancy could be neglect or failure to identify the minor non-endocrine carcinoma component, or mistaking the poorly differentiated endocrine carcinoma component as a poorly differentiated squamous cell carcinoma or adenocarcinoma. Mixed tumour histology is common (37%) and may affect therapy of small cell carcinoma in the GI tract.21 22 It has been widely believed that the histological heterogeneity indicates origin from multi-potential reserve cells.23 24 We only had a limited number of mixed endocrine–exocrine carcinomas, therefore the effects of the second histology component on survival were not analysed in this study.
GI-SmCC is a rare but highly aggressive malignancy. About 25% of GI-SmCC cases may fail to express detectable AE1/AE3 or EMA. The incidence for positive immunohistochemical reactivity for CK8, synaptophysin, NSE and CD56 in GI-SmCC was over 90%, and useful in diagnosis. TTF-1 positivity is not uncommon in GI-SmCC. SmCCs with negative TTF-1 staining may indicate an extra-pulmonary primary. Histological heterogeneity was demonstrated in 36% of oesophageal SmCC, and 27% of gastric SmCCs, and may be diagnosed as mixed endocrine–exocrine carcinoma.
Gastrointestinal tract small cell carcinoma (GI-SmCC) is a rare, highly aggressive malignancy.
The incidence for positive immunohistochemical reactivity for CK8, synaptophysin, NSE and CD56 in GI-SmCC was over 90%, and useful in diagnosis.
Histological heterogeneity was shown in 36% of oesophageal SmCCs and 27% of gastric SmCCs, and may be diagnosed as mixed endocrine–exocrine carcinoma.
Funding The work was funded by the National Science Council (NSC 98-2320-B-010-015) and Taipei Veterans General Hospital (V99C1-126).
Competing interests None.
Ethics approval Obtained from the Taipei Veterans General Hospital Research Ethics Committee (VGHIRB No. 97-12-42A).
Provenance and peer review Not commissioned; externally peer reviewed.