Pulmonary carcinoid tumors comprise 1 to 2% of all lung malignancies. Historically called bronchial adenomas, they were thought to be benign tumors. The recognition of a more aggressive variety of carcinoid tumor, the atypical carcinoid, suggested that they might all be malignant. Currently, typical and atypical carcinoids are considered to be part of a spectrum of malignant neoplasms with neuroendocrine differentiation along with large cell neuroendocrine carcinoma and small cell carcinoma.
In general’ patients with typical carcinoids have good prognoses’ with > 87% of patients surviving for 10 years. In contrast’ approximately 25 to 69% of patients with atypical carcinoid tumors survive 5 years’ and many develop widespread disease. Multivariate analyses from four studies suggest that pathologic stage and atypical histology are the most important factors affecting survival; however’ not all studies have confirmed this. In a study of 93 patients with typical carcinoid tumors, the presence of lymph node metastases in 9 patients (all with N1 disease) did not have any prognostic significance. Eight patients survived a mean of 8 years, while one patient died of unrelated causes. In another report of 120 patients, there was insufficient information to allow any conclusions about the significance of either atypical features or lymph node metastases (only 7 patients had lymph node metastases). No prior studies specifically analyzed the subset of patients with typical or atypical pulmonary carcinoid tumors who presented with lymph nodal metastases at the time of surgical resection. As shown in Table 3, we compare the survival data for patients with typical or atypical pulmonary carcinoid tumors presenting with lymph node metastases with a historical series of patients with pulmonary carcinoid tumors.
In the present study, patients with typical pulmonary carcinoid tumors with thoracic lymph node metastases did well after surgical resection alone. The overall survival rate was excellent, with only two patients (8.7%) developing systemic disease at a median time of 66 months after diagnosis and two patients dying. This is in contrast to patients with atypical pulmonary carcinoid tumors with thoracic lymph node metastases. When treated with surgical resection alone, most of these patients (63.6%) developed SM at a median time of 17 months after diagnosis. The majority of patients with SM (54.5%) died shortly thereafter (median time, 25.5 months after initial diagnosis).
The results of this study suggest that the accurate classification of pulmonary neuroendocrine tumors is essential to determining prognosis, but until recently there have been no widely accepted criteria with which to make a reproducible or clinically meaningful separation between the various tumor types. The new WHO classification scheme, based on the work of Travis and Sobin, utilizes a combination of mitotic activity and the presence of necrosis to separate typical carcinoid tumors from atypical carcinoid tumors. This study supports the clinical utility of the WHO classification, demonstrating differences in the Kaplan-Meier survival curves comparing the patients’ original diagnoses with the WHO classification (Fig 4). Although our numbers are relatively small’ they do suggest that carcinoid tumors with mitotic counts between 2/2 mm2 and 10/2 mm2 of viable tumor or coagulative necrosis behave in a more aggressive fashion than ordinary carcinoid tumors and that patients with those tumors that have a very high mitotic rate (> 10/2 mm2) have the worst prognoses. Although much has been written about the difficulty in counting mitotic figures in several tumor systems’ there are equally difficult problems with basing the classification scheme on such things as “cellularity and atypia.”’ One problem encountered in this study was distinguishing between true coagulative necrosis and what might be called “incipient necrosis.” This latter finding consisted of foci of cells with somewhat pyknotic nuclei and shrunken-appearing cytoplasm’ in comparison to immediately adjacent cells’ and cell discohesion without frank karyorrhexis’ eosinophilic necrotic debris’ or apoptosis (Fig 1). Such foci were not counted as foci of coagulative necrosis and were not used to classify a tumor as an atypical carcinoid tumor, Because of the difficulty in evaluating small foci of necrosis and in counting mitotic figures on frozen sections’ it may be very difficult in borderline cases to make a definitive diagnosis without examining the permanent sections. For intraoperative surgical management’ however’ a diagnosis of carcinoid tumor should be sufficient’ and we reiterate the opinion of others that all such patients should be treated and staged similarly to other patients with lung cancer.
Finally’ this study supports the significance of recognizing the existence of a high-grade non-small cell tumor with neuroendocrine differentiation (ie. large cell neuroendocrine carcinoma). Although there is debate regarding the significance of nonsmall cell carcinomas the neuroendocrine differentiation of which is only suggested by the presence of chromogranin A or synaptophysin identification im-munohistochemically’ tumors that appear to be neuroendocrine by light microscopy and that have a mitotic rate > 10/2 mm2 behave significantly worse than those with mitotic rates < 10/2 mm2. We identified two patients whose tumors were reclassified from atypical carcinoid tumor to large cell neuroendocrine carcinoma. Both patients with large cell neuroendocrine carcinoma developed systemic disease and died shortly after diagnosis (median time’ 18 months after initial diagnosis).
We concur with others that all patients with any type of carcinoid tumor should be treated and staged similarly to other patients with malignant epithelial lung tumors. Because the majority of pulmonary carcinoid tumors occur centrally and involve the bronchial tree’ lung preservation surgery (ie’ bronchial sleeve resection) is recommended when possible. Tumors with greater involvement of the airway or ones that are located peripherally need more extensive surgery’ ranging from lobectomy to pneumonectomy.’’ Bronchoscopic removal is suboptimal treatment’ because the tumors can extend submucosally and adequate resection is not achieved by removing the intraluminal component. However’ a minority of patients have bronchial carcinoid tumors presenting as polypoid structures in the airway lumen’ and in these rare cases broncho-scopic resection alone has been curative. Broncho-scopic resection does not’ however’ allow for lymph node staging’ which the results of this study suggest is very important for prognostic information and may be beneficial in making therapeutic decisions.
Our data suggest that the subset of patients with atypical pulmonary carcinoid tumors and regional lymph node metastases have a high likelihood of developing recurrent disease if treated with surgical resection alone. Although there is a paucity of data available on the treatment of atypical pulmonary carcinoid tumors’ previous studies have demonstrated a poor response to standard chemotherapy and radiation therapy. Currently’ there is no standard therapy for treating atypical carcinoid tumors with any modality other than surgery. Further investigation is needed to determine the optimal treatment with chemotherapy’ radiotherapy’ or alternative promising new experimental agents.
ACKNOWLEDGMENT: We thank Darrell R. Schroeder from the Mayo Clinic Section of Biostatistics for performing the statistical analysis in this study.
Figure 1. Typical carcinoid tumor with central (top, A) and peripheral (bottom, B) foci (arrows) of cell discohesion, nuclear hyperchromatism, pyknosis, and condensed cytoplasm. Such foci were not accepted as tumor necrosis (hematoxylin-eosin, original X 100 [top, A], original X 200 [bottom, B]).
Figure 4. Kaplan-Meier survival analysis of study patients using the Arrigoni criteria (top, A) or the current WHO classification (bottom, B).