(Chest. 2003;124:141-146.)
© 2003
American College of Chest Physicians
A Clinicopathologic Study of 12 Neuroendocrine Tumors Arising in the Thymus*
Olivier Tiffet, MD;
Andrew G. Nicholson, DM;
Georges Ladas, MD;
Mary N. Sheppard, MD and
Peter Goldstraw, MB ChB
* From the Departments of Thoracic Surgery (Drs. Tiffet, Ladas and Goldstraw) and Histopathology (Drs. Nicholson and Sheppard), Royal Brompton Hospital, London, UK.
Correspondence to: Peter Goldstraw, MB ChB, Department of Thoracic Surgery, Royal Brompton Hospital, Sydney St, London, SW3 6NP, United Kingdom; e-mail: p.goldstraw{at}rbh.nthames.nhs.uk
 |
Abstract
|
|---|
Study objectives: To determine whether the new classification system for thymic carcinoid tumors/neuroendocrine carcinomas provides prognostic data, and to study the presentation, diagnosis, treatment, and prognostic factors of these rare tumors.
Design: Retrospective analysis.
Setting: Royal Brompton Hospital, London, UK.
Results: Eight men and four women with a median age of 58 years. Complete excision was possible in nine patients. Postoperative staging revealed two stage I tumors, two stage II tumors, three stage III tumors, one stage IVA tumor, and four stage IVB tumors. All tumors demonstrated the histopathologic features of neuroendocrine tumors, which were confirmed by positive immunohistochemical staining for chromogranin A in 11 of 12 tumors and for CD56 in 12 of 12 tumors, and the presence of dense core granules on ultrastructural analysis in 9 of 9 tumors. All 12 tumors did not stain positively for somatostatin receptors. Three tumors were grade 1, six cases were grade 2, and three cases were grade 3. Follow-up was available in all patients. One patient died 1 month postoperatively. Distant metastasis developed in nine patients (82%). Local recurrence was evident in six patients, of whom five had not received postoperative radiotherapy. Seven patients died of distant metastasis (22 to 83 months after surgery). Two are alive and disease-free (at 67 and 81 months), and two are alive with disease (at 60 and 86 months)
Conclusions: Neither grading as neuroendocrine carcinomas nor any individual histologic parameter showed a significant association with prognosis. Initial aggressive treatment, including complete surgical excision and adjuvant radiotherapy, appears to offer the best hope for prolonged survival. Adjuvant chemotherapy also should be considered, since the incidence of distant relapse is high.
Key Words: mediastinum • neuroendocrine carcinoma • thymic carcinoid tumor
|
Introduction
|
|---|
Neuroendocrine tumors arising in the thymus are extremely rare. In 1972, Rosai and Higa1
differentiated them from other tumors such as thymomas and tumors of the ectopic parathyroid gland. Subsequently, there have only been a few studies2
3
4
5
reporting substantial numbers of cases, with most concluding that these tumors behave in a malignant fashion and have histopathologic features of malignancy when assessed with the criteria used for assessing primary pulmonary neuroendocrine tumors. Most series have used the term carcinoid to describe these tumors, noting atypical features when present, although De Montpreville et al3
suggested that they should all be regarded as belonging to a spectrum of neuroendocrine carcinomas. There also have been occasional reports of small cell carcinomas (SCCs),6
7
8
and, more recently, large cell neuroendocrine carcinomas (LCNECs),9
arising at this site. Because of the apparent difference in behavior and prognosis of thymic neuroendocrine tumors and those arising in the lung, an alternative classification system has been proposed of well, moderately, and poorly differentiated neuroendocrine carcinomas, based on mitotic count, amount of necrosis, preservation of neuroendocrine architecture, and degree of cytologic atypia,10
which has been reported as correlating with biological behavior.11
We have therefore retrospectively reviewed a series of 12 patients, some of which have been reported previously,6
to determine whether this classification system provides prognostic data, and to look at the presentation, diagnosis, and treatment of these rare tumors.
|
Materials and Methods
|
|---|
The records of 12 cases diagnosed between 1977 and 1994 of thymic neuroendocrine tumors were studied. Two patients originated from Hospital Nord in Saint-Etienne, 10 originated from the Royal Brompton Hospital in London, where during the same period 135 thymic epithelial tumors were resected. Thymic neuroendocrine tumors accounted for 7.4% of the total number of tumors. Clinical data and follow-up data were obtained from the patients’ records, referring physicians, or patients and their families. The classification of Masaoka et al12
for thymomas was used for postoperative staging (Table 1
).
Slides were reviewed for all 12 patients, and immunohistochemistry was performed using MNF116 (1:80 dilution; Dako; Cambridge, UK), neurone-specific enolase (1:1,000 dilution; Dako), chromogranin (1:20 dilution; Dako), synaptophysin (1:200 dilution; Dako), CD56 (1:200 dilution; Novocastra; Newcastle upon Tyne, UK), somatostatin (1:400 dilution; Dako), p53 protein using DO7 antibody (1:500 dilution; Dako); and Ki67 (1:100 dilution; Dako) using an immunostainer (Horizon; Dako). Slides stained for p53 protein, Ki67, and CD56 were pretreated by pressure cooking, and slides stained for MNF116 were pretreated by trypsinization to facilitate epitope retrieval. Ultrastructural studies were undertaken using a transmission electron microscope (model 7000; Hitachi; Tokyo, Japan), as previously described.6
As well as immunophenotype, the following histopathologic features were assessed: mitotic rate (number per 2 mm2), presence or absence of necrosis, architecture, presence of local invasion, features of thymic differentiation (accompanying lymphocytes and perivascular lacunae). Tumors then were graded initially into typical carcinoid (TC), atypical carcinoid (AC), SCC, or LCNEC, according to the World Health Organization criteria for neuroendocrine thymic tumors,13
and subsequently by the classification system proposed by Klemm and Moran10
(Table 2
).
|
Results
|
|---|
Table 3
summarizes the clinical features. The median age was 58 years (age range, 35 to 78 years). Eight of 12 patients (66.6%) were men. Four patients were asymptomatic and had their tumor discovered on a routine chest radiograph. Seven patients presented with local symptoms related to the mediastinal mass ranging from cough and chest pain to superior vena cava syndrome and hoarseness. One patient (case 8) presented with lethargy, weight loss, and polyuria due to ectopic adrenocorticotrophic hormone (ACTH) secretion, which was diagnosed by the presence of raised 24-h urinary cortisol levels, elevated cortisol levels in the blood, and raised serum ACTH levels (urine cortisol level, 32,000 nmol per 24 h [normal level, 110 to 440 nmol per 24 h]; serum cortisol level, 1,700 nmol/L [normal level, 165 to 660 nmol/L at 8:00 AM]; and ACTH level, 242 pg/mL [normal level, 9 to 52 pg/mL at 8:00 AM]). The findings of an MRI of the pituitary gland and a CT scan of adrenal glands in this patient were normal. Cortisolemia, which had been noted preoperatively, resolved with complete surgical excision of the tumor. Two cases were associated with multiple endocrine neoplasia, type 1 (MEN-1) [16.6%]. Case 3, a 78-year-old woman, had asymptomatic hyperparathyroidism (parathyroid hormone level, 92 ng/L [normal level, 10 to 68 ng/L]). Case 2 was an affected member of a Bangladeshi MEN-1 family. One sibling had hyperparathyroidism and a prolactin- secreting tumor, and one first-degree relative had hyperparathyroidism and an adrenal carcinoma. Chest CT scans (nine scans) showed an anterior mediastinal mass that appeared to be homogeneous in five patients and heterogeneous with central necrosis and cystic degeneration in four patients. Only three tumors were well-circumscribed, with the other nine tumors showing invasion into the surrounding tissue. No tumors showed calcification. Bronchoscopy was performed in eight patients, and the findings were normal in seven patients. In case 5, bronchoscopy revealed extrinsic compression of the left main bronchus. In case 11, there was dense localized uptake in the primary neoplasm using 123I-MIBG.
Two core biopsy specimens (in cases 3 and 11) yielded thymic carcinoid tumors, while one fine needle aspiration specimen (in case 8) was diagnosed initially as thymoma. Three of four surgical biopsy specimens, which were obtained during anterior mediastinotomy (in cases 4, 6, 10, and 12), revealed a neuroendocrine tumor, whereas the tumor in case 7 was initially identified as a thymoma.
In order to obtain complete macroscopic clearance, part of the pericardium was resected in five cases, the phrenic nerve was resected in four cases, and the innominate vein was resected in three cases (one with reconstruction). The resection of lung parenchyma was necessary in five patients, by wedge excision in three cases and by lobectomy in two cases. Resection was microscopically incomplete in two cases. One patient (case 12) did not undergo surgery because local extension excluded complete resection of the tumor.
Histopathology
All 12 tumors showed cytologic and architectural features of neuroendocrine differentiation (Tables 4
and 5
), supported by positivity for CD56 (12 of 12 tumors), neuron-specific enolase (12 of 12 tumors), chromogranin (11 of 12 tumors), and synaptophysin (10 of 12 tumors) [Table 5
]. Staining was usually diffuse and strong in intensity, although the tumor in case 7 showed only focal staining for CD56. Ultrastructural analysis confirmed the presence of dense core granules in nine of nine tumors. Only two tumors were single masses with recognizable capsules on microscopy, and only one of these showed no evidence of capsular invasion. Four tumors showed no evidence of necrosis, four tumors had individual cell necrosis, three tumors showed discrete islands of necrosis, and one tumor had geographic areas of necrosis. Other than the tumor with geographic necrosis, there was no correlation with mitotic rate. Lymphocyte numbers were not marked in association with the tumor cells, and all tumors except one lacked the characteristic perivascular lacunae seen in thymomas. Staining for Ki67 showed a rough correlation between the percentage of positive cells and the mitotic rate. Staining for p53 showed a wide variation of both nuclear and cytoplasmic staining, independent of other parameters (Table 4)
. From the above, three tumors were classified as grade 1 neuroendocrine carcinomas (ie, TCs), six tumors were classified as grade 2 neuroendocrine carcinomas (ie, ACs), and three tumors were classified as grade 3 neuroendocrine carcinomas (two LCNECs and one SCC). The assessment of all histopathologic parameters and the overall classification by both grade and neuroendocrine tumor subtype showed no association with survival.
Full follow-up was available for all patients. One patient (case 3) died 1 month postoperatively of a hemopericardium. She was receiving therapy with oral anticoagulants for atrial fibrillation and arrhythmia prior to undergoing a simple thymectomy, and a perturbation of her hemostasis may have contributed to her death. Metastasis (lung and pleura, 6 patients; bone, 5 patients; liver, 4 patients; brain, 1 patient) developed in 9 of the remaining 11 patients (82%). A CT scan of patient 8 showed bilateral lung metastasis at 13 months, but the octreotide scintigraphy finding was negative. Local recurrence was evident in six patients, of whom five had not received postoperative radiotherapy (Fig 1 ). The treatment of recurrence included chemotherapy and radiotherapy, except for the patient in case 1, who underwent repeat complete surgery plus radiotherapy. This patient was still alive and free of disease 35 months after undergoing the repeat surgery. Over the 18-year period of data collection, different combinations of chemotherapy were employed, with progressive disease seen in some patients despite treatment (eg, cisplatin- etoposide, ifosfamide, cyclophosphamide-adriamycin- vincristine-prednisone, and interferon-
) and partial response seen in others (eg, therapy with continuous 5-fluorouracil, 5-fluorouracil-cisplatin, or cisplatin-VP16). Seven patients died of distant metastases (median survival time, 52 months; range 22 to 83 months), two patients are alive and disease-free (at 67 and 81 months), and two patients are alive with disease (at 60 and 86 months). The median survival time was 82 months for patients with stage I disease (two patients), 76 months for patients with stage II disease (two patients), 67 months for patients with stage III disease (three patients), 27 months for patients with stage IV-A disease (one patient), and 39 months for patients with stage IV-B (four patients) [p = 0.015].
|
Discussion
|
|---|
Primary neuroendocrine carcinomas of the thymus (previously called TCs) are rare among all anterior mediastinal tumors, accounting 7.4% of the tumors found in our study. Wick et al14
in 1980 identified 7 primary neuroendocrine carcinomas of the thymus among 173 primary anterior mediastinal tumors (4%). A search of the literature indicated that approximately 200 cases of neuroendocrine tumors of the thymus have been reported to date, and most of them are case reports. The median age at diagnosis ranged from 40 to 58 years, and men predominate in all series (67 to 91%).2
3
4
5
6
14
There is a frequent association with endocrine disorders. First, there is an association with Cushing syndrome due to ectopic ACTH production, which has been reported in 20 to 35% of patients,5
although it was found in only 8.3% of patients in our study. Second, MEN-1 syndrome has been reported in 25% of thymic neuroendocrine tumors,15
with 2 of our 12 patients having a history of parathyroid disease. Interestingly, MEN-1 and Cushing syndrome were seen to be associated only with TCs and ACs, and not with higher grade tumors. This study therefore emphasized the importance of screening for such disorders once a diagnosis of thymic neuroendocrine carcinoma is made.
From a histopathologic viewpoint, the tumors comprised a spectrum ranging from well-differentiated to poorly differentiated neuroendocrine carcinomas (ie, TC to SCC), dependent on the classification system being used. However, neither system showed an association either with stage of disease at the time of resection or survival. This is in contrast to the larger series of Moran and Suster,11
which demonstrated an inverse correlation between survival and tumor differentiation. Furthermore, no individual macroscopic parameters (ie, tumor size) or microscopic parameters (ie, mitotic rate, the presence of necrosis, Ki67 staining, or p53 staining) that were of potential value in regard to prognosis showed an association with survival, which is consistent with the findings of other groups analyzing proliferation rates in these tumors.16
We found that the median survival time was 82 months for patients with stage I disease, 76 months for patients with stage II disease, 67 months for patients with stage III disease, 27 months for patients with stage IVA disease, and 39 months for patients with stage IVB, which suggests that stage may be important for survival. Gal et al,17
in reviewing their own cases and some previously published cases, showed that unresectability and advanced clinical stage were statistically significant independent prognostic factors that were associated with poor outcome. In the literature,2
3
17
the overall survival rate at 5 years ranged from 31 to 82%, compared to 50% in our study. From our results, the major prognostic factor remains how radical the excision may be. The only patients alive and free of disease at 5 years underwent total excision regardless of stage, sometimes requiring surgical excision of a local recurrence.2
Furthermore, adjuvant radiotherapy appears to prevent local recurrence after total excision. Local recurrence occurred in 6 of the 11 patients undergoing surgery. Five of these six patients did not receive postoperative radiotherapy, and the sixth patient had microscopic residual disease after surgery, while three patients who had undergone complete excision and postoperative radiotherapy had no local relapse. The high rate of distant metastasis (82%; 9 of 11 patients) suggests that chemotherapy also should be considered, but its modalities remain to be defined.
Because several authors have underlined the accuracy of indium-111-diethylenetriamine pentaacetic acid-D-Phe1-octreotide scintigraphy in detecting thymic neuroendocrine tumor,18
19
20
21
we also assessed our patients for the presence of somatostatin receptors (SRs). This was undertaken to determine whether SR status might provide useful data, both for staging using octreotide scintigraphy and for adjuvant treatment in selecting those patients who might benefit from treatment with octreotide. However, in one case with a lung metastasis larger than 1.5 cm, an octreotide scan failed to demonstrate any recurrence. Furthermore, none of the tumors stained positively for somatostatin. Therefore, this study provides no evidence that scanning for SRs may help in the preoperative staging of neuroendocrine tumors in the thymus.
In conclusion, thymic neuroendocrine tumors carry a poor prognosis and appear to have no prognostic relationship with the size of tumor or histologic grading in this small series. The best chance for a cure seems to be complete surgical excision plus adjuvant radiotherapy. Adjuvant chemotherapy also should be considered, since the incidence of distant relapse is high.
|
Footnotes
|
|---|
Abbreviations: AC = atypical carcinoid; ACTH = adrenocorticotrophic hormone; LCNEC = large cell neuroendocrine carcinoma; MEN-1 = multiple endocrine neoplasia, type 1; SCC = small cell carcinoma; SR = somatostatin receptor; TC = typical carcinoid
Received for publication January 15, 2002.
Accepted for publication January 24, 2003.
|
References
|
|---|
- Rosai, J, Higa, H (1972) Mediastinal endocrine neoplasm of probable thymic origin, related to carcinoid tumor: clinicopathologic study of 8 cases. Cancer 29,1061-1074[CrossRef][ISI][Medline]
- Fukai, I, Masaoka, A, Fujii, Y, et al Thymic neuroendocrine tumor (thymic carcinoid): a clinicopathologic study in 15 patients. Ann Thorac Surg 1999;67,208-211[Abstract/Free Full Text]
- De Montpreville, VT, Macchiarini, P, Dulmet, E Thymic neuroendocrine carcinoma (carcinoid): a clinicopathologic study of fourteen cases. J Thorac Cardiovasc Surg 1996;111,134-141[Abstract/Free Full Text]
- Economopoulos, GC, Lewis, JW, Lee, MW, et al Carcinoid tumors of the thymus. Ann Thorac Surg 1990;50,58-61[Abstract]
- Wang, DY, Chang, DB, Kuo, SH, et al Carcinoid tumours of the thymus. Thorax 1994;49,357-360[Abstract]
- Valli, M, Fabris, GA, Dewar, A, et al Atypical carcinoid tumour of the thymus: a study of eight cases. Histopathology 1994;24,371-375[Medline]
- Rosai, J, Levine, G, Weber, WR, et al Carcinoid tumors and oat cell carcinomas of the thymus. Pathol Annu 1976;11,201-226[Medline]
- Wick, MR, Scheithauer, BW Oat-cell carcinoma of the thymus. Cancer 1982;49,1652-1657[Medline]
- Chetty, R, Batitang, S, Govender, D Large cell neuroendocrine carcinoma of the thymus. Histopathology 1997;31,274-276[Medline]
- Klemm, KM, Moran, CA Primary neuroendocrine carcinomas of the thymus. Semin Diagn Pathol 1999;16,32-41[Medline]
- Moran, CA, Suster, S Neuroendocrine carcinomas (carcinoid tumor) of the thymus: a clinicopathologic analysis of 80 cases. Am J Clin Pathol 2000;114,100-110[CrossRef][ISI][Medline]
- Masaoka, A, Yasumasa, M, Nakahara, K, et al Follow-up study of thymomas with special references to their clinical stages. Cancer 1981;48,2485-2492[CrossRef][ISI][Medline]
- Rosai, J, Sobin, LH World Health Organisation histological classification of tumors: histological typing of tumors of the thymus 2nd ed. 1999,15-18 Springer Verlag. Berlin, Germany:
- Wick, MR, Scott, RE, Li, CY, et al Carcinoid tumor of the thymus: a clinicopathologic report of seven cases with a review of the literature. Mayo Clin Proc 1980;55,246-254[ISI][Medline]
- Teh, BT, Zedenius, J, Kytola, S, et al Thymic carcinoids in multiple endocrine neoplasia type 1. Ann Surg 1998;228,99-105[CrossRef][ISI][Medline]
- Goto, K, Kodama, T, Matsuno, Y, et al Clinicopathologic and DNA cytometric analysis of carcinoid tumors of the thymus. Mod Pathol 2001;14,985-994[Medline]
- Gal, A, Kornstein, M, Cohen, C, et al Neuroendocrine tumors of the thymus: a clinicopathological and prognostic study. Ann Thorac Surg 2001;72,1179-1182[Abstract/Free Full Text]
- Silva, F, Vazquez-Selle, J, Aguilo, F, et al Recurrent ectopic adrenocorticotropic hormone producing thymic carcinoid detected with octretide imaging. Clin Nucl Med 1999;24,109-110[CrossRef][ISI][Medline]
- Lastoria, S, Vergara, E, Palmieri, G, et al In vivo detection of malignant thymic masses by indium-111-DTPA-D-Phe- octreotide scintigraphy. J Nucl Med 1998;39,634-639[Abstract/Free Full Text]
- Satta, J, Ahonen, A, Parkkila, S, et al Multiple endocrine neoplastic-associated thymic carcinoid tumour in close relatives: octreotide scan as a new diagnostic and follow-up modality; two case reports. Scand Cardiovasc J 1999;33,49-53[CrossRef][Medline]
- Nilsson, O, Kolby, L, Wangberg, B, et al Comparative studies on the expression of somatostatin receptor subtypes, outcome of octreotide scintigraphy and response to octreotide treatment in patients with carcinoid tumours. Br J Cancer 1998;77,632-637[ISI][Medline]
This article has been cited by other articles:
|
|
|
|
 
B. V. Duwe, D. H. Sterman, and A. I. Musani
Tumors of the Mediastinum
Chest,
October 1, 2005;
128(4):
2893 - 2909.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. Ferolla, A. Falchetti, P. Filosso, P. Tomassetti, G. Tamburrano, N. Avenia, G. Daddi, F. Puma, R. Ribacchi, F. Santeusanio, et al.
Thymic Neuroendocrine Carcinoma (Carcinoid) in Multiple Endocrine Neoplasia Type 1 Syndrome: The Italian Series
J. Clin. Endocrinol. Metab.,
May 1, 2005;
90(5):
2603 - 2609.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
E. K. Hansen, R. M. Krieg, R. J. Azar, D. M. Jablons, and T. M. Jahan
Unusual Thoracic Problems in Patients With Malignancies: CASE 3. Carcinoid Tumor of the Thymus
J. Clin. Oncol.,
April 20, 2005;
23(12):
2859 - 2861.
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Whitaker
Thymic neuroendocrine carcinoma (carcinoid): What if octreotide scanning is negative?
J. Thorac. Cardiovasc. Surg.,
November 1, 2004;
128(5):
801 - 801.
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Whitaker and J. Dussek
PET Scanning in Thymic Neuroendocrine Tumors
Chest,
June 1, 2004;
125(6):
2368 - 2369.
[Full Text]
[PDF]
|
|
|
|
|
|
|
F. C. Detterbeck and A. M. Parsons
Thymic tumors
Ann. Thorac. Surg.,
May 1, 2004;
77(5):
1860 - 1869.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
Abstract
Introduction
