—  SLIDE SEMINAR #17  —

Pulmonary Tumors
Moderators: Dr. Elizabeth Brambilla, Dr. John English and Dr. Donald Guinee

Case 2 - Atypical carcinoid tumour associated with adenocarcinoma

Dr. Philip Hasleton
Department of Histopathology
Manchester Royal Infirmary


Lung tumours with neuroendocrine (NE morphology) include typical carcinoid (TC), atypical carcinoid (AC), large cell neuroendocrine carcinoma (LCNEC), small cell lung cancer lung cancer (SCLC) and non-small cell lung cancer with NE morphology

Clinical History:
68-year-old male, ex smoker, presented with hemoptysis. On CT scan, a mass was identified in the lingula and on bronchoscopy, there was a mass obliterating the left lingular bronchus. A left upper lobectomy was performed with a tumour protruding through the inferior bronchial resection margin, within 2 mm of the cut edge. The tumour was well demarcated, cream coloured and measures 6.0x3.5x3.5 cm. the uninvolved lung was unremarkable apart from emphysema at the apex.


Case 2 - Slide 1
Click to view with ImageScope
Click to view with a Web-Based Viewer

Incidence
Recently the incidence of pulmonary NE tumours has been described Denmark. [1] This was a cancer registry-based analysis of patients, studied over the period 1978-1997. The patients were followed up to 31 December 1999. There was no histological verification of all of their cases. The recorded incidence of pulmonary NE tumours, other than SCLC, increased two-fold amongst men (0.24-0.53 per 100,000 inhabitants per year) and three-fold in women (0.14-0.41 per 100, 000 inhabitants per year) during the study period. The incidence of SCLC decreased amongst men and leveled of amongst women. Reasons for this increase are not available but on the twentieth anniversary of Chernobyl, it is fascinating to speculate if the fallout from this disaster may have played a part. Studies from other countries in the Baltic and suffering from exposure would be welcome to compare with rates from the rest of the world.

Definition
Carcinoid tumours are characterised by growth patterns (organoid, trabecular, insular, palisading, ribbon, rosette-like arrangements) that suggest NE differentiation. [2] Tumour cells have uniform cytologic features but moderate eosinophilic, finely granular cytoplasm and nuclei with a finely granular chromatin. Typical carcinoids (TC) are carcinoid tumours with fewer than 2 mitoses per 2mm2 and lacking necrosis. Atypical carcinoid (AC) is a carcinoid tumour with 2-10 mitoses per 2mm2 and/or foci of necrosis.

Macroscopic Features
TC and AC are both firm well-demarcated tan to yellow tumours. Typically TC are central tumours and frequently endobronchial, the overlying mucosa may be intact or ulcerated and there maybe squamous metaplasia. Association with airways may not be readily evident in peripheral tumours. [3]

Histology
Nuclear atypia and pleomorphism may be quite marked, even in TC. These are unreliable criteria for distinguishing TC from AC. [4] At presentation, 10-15% of TC have metastasised to regional lymph nodes and 5-10% of cases eventually spread to distant sites, such as liver or bone. TC is classified by the WHO as a malignant tumour and it is frequent that surgeons have to be appraised of this fact, since there is a belief this tumour behaves in a benign fashion. The figures for lymph node spread are higher in AC, where 40-50% metastasised to regional nodes and beyond, with 57% stage I disease, 21% stage II, 14% stage III and 8% stage IV at presentation. [5]

The histology has been referred to above but there are variations, oncocytic tumours have abundant eosinophilic cytoplasm. Rarely the cytoplasm maybe be clear and can contain melanin. There may be a spindle cell, papillary, pseudo-glandular, follicular or rosette-like patterns but true gland formation is rare. The stroma is highly vascularized, due to TGF-a secretion by the tumour. In some cases the stroma is hyalinized or may show cartilage or bone. Bone formation is seen in up to 33% of carcinoid tumours, both typical and atypical. It is most probably related to secretion of IGF-1 and TGF-β. [6] In central tumours, there is, if anything, a decrease in NE cells in the bronchus in the region of the tumour. [7] However, some cases may show NE hyperplasia, sometimes associated with airway fibrosis or there may be diffuse idiopathic neuroendocrine hyperplasia. [8] Such changes are most commonly seen in associated with peripheral carcinoids. Rare cases may also show multiple tumourlets or multiple carcinoids tumours. [9] Apart from focal necrosis and a mitotic rate, numbering between 2-10 mitoses per 2mm2, AC show the histological features referred to above.

Prognosis
Prognosis in AC was studied in 106 cases by Beasley et al. [5] Using clinico-histological features, analysed by Kaplan-Meier and Cox regression analyses, they showed a worse prognosis was associated with a higher stage (p =0.003) and a tumour size of 3.5 of greater (p = 0.003). The features histologically unfavourable by univariate analysis were mitotic rate (p =0.002), pleomorphism (p =0.018) and aerogenous spread (p =0.007). Histologically favourable features by univariate analysis were palisading (p =0.008), papillary (p =0.039), pseudoglandular (p =0.026), and rosette (p =0.022) patterns. The female gender showed a trend towards poorer prognosis (p =0.085).

Multivariate analysis in this study, stratified by stage, showed mitoses (p < than 0.001), tumour size of 3.5cm or greater (p = 0.017) and female gender (p =0.012) were the only negative independent factors of prognosis. The presence of rosettes (p = 0.016) was the only independent positive factor. These authors also divided AC into subgroups with low (2-5 mitoses/2mm2) and high (6-10 mitoses/2mm2) mitotic counts and compared their survival with TC and LCNEC.

Patients with AC having a higher mitotic count had a significantly worse survival rate than those with a lower mitotic count (p < than 0.001) when stratified for age. 5 - and 10 -year survival rates for AC (61% and 35% respectively) stratified by stage was significantly worse than for AC but better than those for LCNEC and SCLC. The data for response to chemo- or radio- therapy were insufficient to evaluate tumour response.

A recent prognostic study, [10] using the 1999 WHO criteria, of resected pulmonary carcinoid tumours had material from 121 TC and 42 AC. The TC were predominantly N0 status (107/121) and 14 were N1. In the AC 15/42 were N0, 18 were N1 and 9 N2. All patients were included in follow-up, with a median of 54 months and mean of 58 months, (range 4-150 months). This follow-up included yearly total body CT and bronchoscopy. The overall 5-year survival rate was 90.3% with a mean survival time of 139 months (95% confidence interval, 133-145). In N0 patients with either TC or AC, no disease-related mortality was reported (100% 5-year survival). In N1 patients 5-year survival was 90.0% in TC and 78.8% in AC (p = 0.394). In AC with N2 disease, 5-year survival was 22.2%. In this study prognosis was related more to nodal status than histological subtype. No TC's had T2 disease, making true comparisons difficult. The authors found no statistical significance between typical and atypical subtype in the N0 and the N1 patients.

Mezzetti et al. [11] collected 98 pulmonary resections for primary bronchial carcinoid tumours, of which 88 had TC and 10 AC. The 5-year survival rate for TC was 91.9% and 71% for AC. The 10-year overall survival rate was 89.7% for TC and 60% for AC. The 5-year TNM-related survival rates in the TC group were 1A-B, 100%, 2A-B, 75% and 3A-50%. At 10 years they were 1A-B 100%, 2A-B 75% and 3A 0%.

The 5-year survival rate in the AC group was 1A-B 100%, 2A-B 100% and 3A 0%. At 10 years they were 1A-B 100%, 2A-B 66% and 3A-B 0%. Prognosis was favourable in both subtypes in early stage disease but in advanced stage it was better in TC. The recurrence rate was worse in AC.

The prognosis of patients with TC in Denmark [1] ranged from 87% for patients with TC to 44% in patients with AC but these authors did not have histological verification of their cases.

Immunohistochemistry
The Beasley study [5] had immunohistochemistry available in 38 cases. Such studies are not required for the diagnosis of AC using the WHO criteria. Cytokeratin "cocktail" stains were available in 31 cases, 26 (84%) showed at least some degree of positive staining, 11/13 (85%) cases stained with CAM 5.2 were positive, 31/33 cases (94%) chromogranin-positive, 30/33 (94%) cases were positive for chromogranin, 30/33 (91%) cases were positive for synaptophysin and 21/28 (75%) cases were positive for Leu-7. All the cases with immunohistochemical studies available showed some degree of staining with at least one of the NE markers. When a percentage of cells staining with a particular marker was analysed, a trend towards a better prognosis was identified with tumours 75% of cells or greater staining with chromogranin, compared with tumours with less than 75% of cells staining (p = 0.082). No other significant results were identified with regard to percentage of cells staining or the intensity of staining.

Most carcinoid tumours stain for cytokeratin but 20% may be keratin negative. [2] NE markers, such as chromogranin, synaptophysin and CD56 (NCAM) are typically strongly positive, especially in TC. In AC staining for these markers may be patchy or focal. Results with TTF-1 have given varying results. Some authors have shown both TC and AC are negative. [12] Others have shown approximately a third of TC and most AC are positive. [13, 14, 15, 16] The differences between the papers may be explained by the fact that Sturm et al. [12] did not use the same antibody as the other authors. The Du et al. paper [16] showed all extra-pulmonary NE tumours were uniformly negative for TTF-1. In their TTF-1 positive cases of TC and AC, 12/14 had a peripheral location and a spindle cell morphology. 11 tumourlets were positive with this antibody. They concluded that TTF-1 expression was 100% specific, though not so sensitive, for a lung primary in TC and AC.

Recently Pelosi et al [17] have shown that normal and hyperplastic bronchial neuroendocrine cells expressed E-cadherin/beta-catenin along the cell membrane. NE tumours retained beta-catenin expression in all tumours and E-cadherin in most tumours, with the exception of 9% of AC, 2% of LCNEC and 3% of SCLC. E-cadherin showed membrane-associated, linear immunoreactivity in TC, whereas "membrane-disarrayed" (defined as a wrinkled staining of the membrane as well as variable cytoplasmic accumulation of staining) and cytoplasmic staining (prevalent cytoplasmic staining with only minimal or absent membrane labelling) were seen in most AC, LCNEC and SCLC (p <0.001). Beta-catenin exhibited similar immunoreactivity patterns according to tumour type and a close association was seen with E-cadherin subcellular distribution (p < 0.001). In AC, disarrayed immunoreactivity for E-cadherin and/or beta catenin was associated with a non-trabecular growth pattern, altered expression of the cell-motility marker fascin and lymph node metastases. In addition, a disarrayed E-cadherin distribution pattern was associated with the pathologic lymph node classification and the number of involved lymph nodes. Multivariate analysis confirmed a disarrayed E-cadherin or beta-catenin pattern was an independent predictor for lymph node metastases in patients with AC.

A further paper by Pelosi et al. [18] showed that fascin immunoreactivity might identify subsets of pulmonary carcinoid patients with different metastatic potential to regional lymph nodes. Fascin-immunoreactivity was detected in only 5% of the 38 TC but 35% of the 23 AC. The figure increased in LCNEC (83% of 40 cases) and 100% of 27 SCLC). Normal NE cells or hyperplastic NE tumourlets were non-reactive. In both TC and AC but not in high-grade NE malignancy, fascin immunoreactivity closely correlated with the presence of lymph node metastases.

Differential diagnosis
The differential diagnosis of carcinoid tumours is large, since many tumours may show an NE pattern. Bronchoscopic and frozen section samples can give rise to false-positive diagnoses. This author has wrongly diagnosed a SCLC on a superficial biopsy from an AC, which subsequently came to resection. Similar results have been found by others. [19, 20] Tumourlets may resemble TC but are distinguished by size, being less than 5mm in diameter. The higher grade NE tumours (LCNEC and SCLC) have a mitotic rate greater than 10 per 2mm 2. Spindle-cell lesions may be confused with spindle cell squamous carcinomas, pleomorphic carcinomas or mesenchymal tumours. This includes mesotheliomas, especially as there is a case report of two AC presenting clinically as mesotheliomas. [21] Immunohistochemistry may resolve these issues. Occasional carcinoids may have a prominent papillary pattern and these may be confused with sclerosing haemangioma or secondary bladder cancer. The former tumour is TTF-1 positive but is negative with NE stains.

The commonest problem is differentiating pseudoglandular or gland-like patterns in carcinoid tumours. These can give rise to misdiagnoses of adenocarcinomas, mucoepidermoid carcinomas, acinic cell tumours, and adenoid cystic carcinomas. Usually adenocarcinomas show more cytological atypia and mucin production, though both may be seen in carcinoids. Adenocarcinomas may show NE-positivity but it is less diffuse than in carcinoids.

A very rare differential is paraganglioma. S100-positive, sustentacular cells may be identified in carcinoids. However, paragangliomas do not stain with cytokeratin, which is frequently positive in carcinoids. [22] Glomus tumours may also resemble carcinoids but are positive for smooth muscle actin and negative for NE markers. [23] Secondary carcinoid tumours from extra-pulmonary sites, as mentioned above, are TTF-1 negative.

The epithelial pattern of a carcinoid tumour maybe mimicked by metastatic breast or prostate carcinoma, the appropriate markers (PSA, TTF-1, ER and PR) may well help in this situation. Another diagnostic problem may be posed by metastatic medullary carcinoma of thyroid. This is NE-positive and though it shows amyloid, occasionally this material may be identified in carcinoid tumours. Usually there are multiple nodules from the rare cases of metastatic medullary carcinoma of thyroid in the lung.

The present case is interesting in two respects. The NE has a mitotic count above that seen in AC but on histology strongly resembles an AC. This author has seen this "problem" several times before. However the work of Travis in establishing the numbers of mitoses to differentiate the NE subtypes is statistically robust. This case indicates how important mitotic count is as prognostic factor in NE tumours. The second unusual feature is the combination of the tumour with an adenocarcinoma. In this instance TTF-1 may be of little in differentiating the two tumours and as always reliance should finally be placed on routine H & E sections.

References
  1. Skuladottir H, Hirsch FR, Hansen HH, Olsen JH. Pulmonary neuroendocrine tumours: incidence and prognosis of histological subtypes. A population-based study in Denmark. Lung Cancer 2002; 37: 127-135.

  2. Travis E, Brambilla E, Muller-Hermelink HK and Harris CC. Tumours of the lung, pleura, thymus and heart. WHO classification of tumours. IARC Press. Lyon 2004.

  3. Soga J, Yakuwa Y. Bronchopulmonary carcinoids: An analysis of 1875 reported cases with special reference to a comparison between typical carcinoids and atypical varieties. Ann Thorac Cardiovasc Surg 1999; 5: 211 - 219.

  4. Travis WD, Rush W, Flieder DB, Falk R, Fleming MV, Gaal AA, Koss MN. Survival analysis of 200 pulmonary neuroendocrine tumours with clarification of criteria for atypical carcinoid and its separation from typical carcinoids. Am J Surg Pathol. 1998; 22: 934-944.

  5. Beasley MB, Thunnissen FB, Brambilla E, Hasleton P, Steele R, Hammar SP, Colby TB, Sheppard M, Shimosato Y, Koss MN, Falk R, Travis WD. Pulmonary atypical carcinoid: predictors of survival in 106 cases. Hum Pathol 2000; 31:1255-1265.

  6. Hasleton PS. Bostanci G. Pulmonary carcinoid and related tumours. Annales Academie Medicae Bialostociensis 1997; 42: 28-42.

  7. Bostanci G, Hasleton PS (Unpublished data)

  8. Miller RR, Muller NL. Neuroendocrine cell hyperplasia and obliterative bronchiolitis in patients with peripheral carcinoid tumors. Am J Surg Pathol 1995; 19: 653-658.

  9. Miller MA, Mark GJ, Kanarek D. Multiple peripheral pulmonary carcinoids and tumorlets of carcinoid type with restrictive and obstructive lung function. Am J Med 1978; 65: 373-378.

  10. Cardillo G, Sera F, Di Martino M et al. Bronchial carcinoid tumours: nodal status and long-term survival after resection. Ann Thorac Surg 2004; 77: 1781-1785.

  11. Mezzetti M, Raveglia, Panigalli T et al. Assessment of outcomes in typical and atypical carcinoids according to the latest WHO classification. Ann Surg 2003; 76: 1838-1842.

  12. Sturm N, Rossi G, Lantuejious et al. Expression of thyroid transcription factor 1 in the spectrum of neuroendocrine cell proliferations with special interest in carcinoids. Hum Pathol. 2002; 33:175-182.

  13. Folpe AL, Jown AM, Lamps LW et al. Thyroid transcription factor; Immunohistochemical evaluation in pulmonary neuroendocrine tumours. Mod Pathol. 1999;12: 528

  14. Oliveria AM, Tazelaar HT, Myers JL, Erickson LA, Lloyd RV. Thyroid transcription factor 1 distinguishes metastatic pulmonary from well differentiated neuroendocrine tumours of other sites. Am J Surg Pathol. 2001; 25: 815-819.

  15. Wick MR. Immunohistology of neuroendocrine and neuroectodermal tumours. Semin Diag Pathol 2000; 17: 194-203.

  16. Du EZ, Goldstraw P, Zacharias J et al. TT-1 expression is specific for lung primary in typical and atypical carcinoids; TTF-1 positive carcinoids are predominantly in predominantly in peripheral location. Hum Pathol. 2004; 35: 825-831.

  17. Pelosi G, Scarpa A, Puppa G et al. Alteration of the E-cadherin/beta-catenin cell adhesion system is common in pulmonary neuroendocrine tumours and is an independent predictor of lymph node metastasis in atypical carcinoids. Cancer 2005; 103: 1154-1164.

  18. Pelosi G, Passini F, Fraggetta F et al. Independent value of fascin immunoreactivity for predicting lymph node metastases in typical and atypical pulmonary carcinoids. Lung Cancer 2003; 43: 203-213.

  19. Marchevsky AM, Changsri C, Gupta I, Fuller C, Houck W, McKenna RG Jr. Frozen section diagnoses of small pulmonary nodules; accuracy and clinical implications. Ann Thorac Surg 2004; 78: 1760

  20. Pelosi G, Rodriguez J, Viale G, Rosai J. Typical and atypical pulmonary carcinoid tumour overdiagnosed as small-cell carcinoma on biopsy specimens: a major pitfall in the management of lung cancer patients. Am J Surg Pathol 2005; 29: 179-87.

  21. Van Hengel P, van Geffen F, Kazzaz BA, Heyerman HGM. Atypical carcinoid presenting as mesothelioma. Neth J Med 2001; 58: 185-190.

  22. Colby TV, Koss M, Travis WD. Tumors of the Lower Respiratory Tract. 3rd ed. Armed Forces Institute of Pathology: Washington DC.

  23. Gaertner EM, Steinberg DM, Huber M et al. Pulmonary and mediastinal glomus tumours - report of five cases including a pulmonary glomangioasarcoma: a clinicopathologic study with literature review Am J Surg Pathol 2000; 24:1105-1114.