Abstract
Objective: Pulmonary carcinoid tumors are rare low-grade malignant neoplasms and constitute 2–5% of all lung tumors. In this study, we aimed to determine the clinical presentations, types of surgery, long-term treatment outcomes and survival rates of patients diagnosed with carcinoid tumor treated surgically in our clinic. Methods: Patients operated in our clinic between 1992 and 2008 with confirmed or suspected diagnoses of carcinoid tumors were included in our study. Patients’ hospital records were retrospectively analyzed. Results: A total of 104 patients (age 19–71 years, mean 44 ± 13 years, 2 SD) with pathologically confirmed carcinoid tumor were analyzed. A total of 84 patients (81%) were diagnosed as typical and the remaining 20 (19%) being atypical carcinoid tumor. As many as 24 patients (23%) were asymptomatic. The most frequent symptom was recurrent respiratory infection (35%). The most used surgical procedures were lobectomy (47%) and bilobectomy (16%). Mean postoperative follow-up period was 72 months (6–190 months). No surgery related mortality was noticed. As many as 15 (14%) patients died during the follow-up period. Overall 5- and 10-year survival rates were 89% and 72%, respectively. For typical carcinoid tumors, the 5- and 10-year survival rates were 92% and 83%, and for atypical carcinoids 73% and 46%, respectively (p ≪ 0.001). Conclusions: In our study, we noticed histological subtype, stage of the disease and the type of surgery performed to be prognostic factors of carcinoid tumors. Atypical carcinoid tumors tend to be more metastatic and had worse prognosis when compared with typical carcinoid tumors. We conclude that surgery is the best treatment of choice for carcinoid tumors, especially parenchyma-sparing procedures, because of their good mid- and long-term survival rates.
1 Introduction
Carcinoid tumors are the second most frequent tumors of the tracheobronchial tree and constitute 2–5% of all lung tumors [1,2]. Although they are rare low-grade malign neoplasms, there has been a report showing significant increase in their incidence in the past 30 years [3]. The carcinoid tumor was first described as bronchial adenoma by Muller in 1882 and later named ‘Karzinoid’ by Oberndorfer in 1907 due to its similarity to lung cancer [4]. The first differentiation between typical carcinoid (TC) and atypical carcinoid (AC) was described in 1944 [5]. In 1972, Arrigoni and colleagues described the histological differentiation criteria for these tumors; because these criteria did not have a wide international usage, they were later updated by the World Health Organization (WHO) based on the strict criteria defined by Travis and colleagues in 1999 [6,7]. From the beginning of 2004, carcinoid tumors were included in the classification of bronchopulmonary neuroendocrine tumors (BP-NET). Even though carcinoid tumors have structural, morphological and immunochemical features similar to large-cell neuroendocrine tumors and small-cell carcinomas both belonging in the same group according to the latest classification, the biological characteristics of carcinoid tumors are considerably variable. While TC is considered a low-grade neuroendocrine tumor, AC is considered to be an intermediate-grade neuroendocrine tumor [4,8]. With the exception of atypical histological type and a small group with more aggressive clinical behavior, carcinoid tumors generally display benign growth and clinical behavior.
In our study, we aimed to determine the clinical, pathological, and radiological bronchoscopic characteristics, long-term outcomes and prognostic factors of patients with pulmonary carcinoid, who were treated surgically between the years 1992 and 2008.
2 Materials and methods
Clinical, surgical, and pathological records of 104 patients, with surgical treatment for carcinoid tumor at the Thoracic Surgery Clinics of Ataturk Training and Research Hospital for Chest Diseases and Chest Surgery, were retrospectively reviewed. The age, gender, smoking habits, presenting symptoms, method of diagnosis, localization, size, endoscopy, postoperative staging, histological type, and survival rates of patients were recorded. All pathological specimens were reviewed and reclassified according to the 1999 WHO criteria.
Chest computed tomography (CT) was used for non-invasive intrathoracic diagnosis and staging. Preoperative and/or intraoperative bronchoscopy to determine the air way part affected, and the degree of abnormal bronchial tissue with endoscopic biopsy and brushing for histological diagnosis was performed for all patients. Whenever appropriate, transthoracic needle-aspiration biopsy was performed for peripherally located carcinoids. Intraoperative frozen section to verify carcinoid tumor diagnosis was performed for patients for whom preoperative diagnosis was difficult. All surgery was performed by posterolateral thoracotomy. In suitable cases, parenchyma-sparing procedures such as segmentectomy or sleeve lobectomy were applied. Hilar and mediastinal lymph-node sampling were performed for all cases.
Periodic clinical and radiological follow-up was performed (3, 6, 12, and 24 months). Routine chest X-ray was done on all patients during every examination, and chest CT was used when necessary. Fiberoptic bronchoscopy was performed on postoperative 3rd and 12th months for patients with suspected recurrence. For the past 2 years, positron emission tomography–computed tomography (PET–CT) scanning was also used during check-up.
Statistical analysis of continuous variables is presented as mean ± SD or median (minimum–maximum); the categorical variables were expressed as number of cases and percentage (%). Student’s t-test was used to determine the mean significant difference between the groups and the Mann–Whitney U-test to determine the median significant difference. Pearson’s chi square (χ2) test or Fisher’s exact test was used to determine the categorical variables. Kaplan–Meier survival analysis applying log-rank test statistics was used to determine whether disease type, type of surgery performed, and the stages among the disease types had any effect on the general survival. Five- and 10-year survival rates were calculated for each case.
3 Results
Preoperatively diagnosed or intraoperative frozen section confirmed 104 carcinoid patients, who were treated surgically, were identified. Out of the 104 patients with postoperative pathological confirmed diagnosis of carcinoid, 52 were male and 52 female. Their ages ranged between 19 and 71 years (median age 44 ± 13.2 years). In terms of histological findings, 84 patients (39 male, 45 female, median age 42 ± 13 years) had TC, and AC was observed in 20 patients (13 male, seven female, median age 50 ± 11 years). There were statistically significant differences between the TC and AC groups in terms of median age with patients in the AC group tending to be older (p = 0.013). There were no significant differences in performance, spirometric values, and diffusion capacity between gender and histological subtypes.
As many as 24 patients (23%) were asymptomatic; carcinoid tumor was coincidentally found in these patients. The most frequent symptoms were recurrent respiratory infection (36) followed by hemoptysis (21), cough (13), and dyspnea (10). Patients with these symptoms had a centrally located tumor. Carcinoid syndrome was not noticed in any of the patients.
Out of the 104 carcinoid tumor cases, 69 patients (66%) were non-smokers, 35 patients (34%) were current or ex-smokers. In the current or ex-smoker group, 24 of the patients were TC (29% of all TCs) and 11 of them were AC (55% of all ACs). Comparison of smoking habits between the TC group and the AC group showed significant differences with more smoking history in the AC group (p = 0.025).
All patients underwent preoperative chest X-ray, chest CT scanning, and pulmonary function tests. Cardiac investigation was carried out for selected patients with positive physical examination or cardiologic complaints. Only 9 (9%) of the 104 patients had a normal chest X-ray finding. The most frequent radiological findings were atelectasis and pulmonary infiltration (58 patients); other radiological findings noticed with decreasing frequency were solitary pulmonary nodule (32 patients), and isolated hilar mass. In our series, 80 patients (77%) had a centrally located tumor and 24 patients (23%) presented with peripheral tumor localization. Out of the 80 centrally located tumors, 66 patients (82.5%) were typical and 14 (17.5%) of them were atypical; and of the 24 peripheral tumors, 18 (75%) were typical and six (25%) atypical.
We performed preoperative bronchoscopy and, when necessary, intraoperative bronchoscopy for all patients. Tumors seen directly under the bronchoscope, or tumors with atelectasis or postobstructive pneumonia were classified as ‘central’ and tumors not visible under the bronchoscope were classified as ‘peripheral’. Bronchoscopies of 18 patients were normal, seven were asymptomatic, and 11 showed symptoms. As many as 17 of the 86 patients, who had abnormal bronchoscopy findings, were asymptomatic, and 59 were symptomatic.
The primary tumor diameter was between 9 and 52 mm (median 24 mm). A total of 70 patients had no nodal invasion (N0) (61 typical, nine atypical); in this group, the median diameter of the primary tumor was 22.5 mm. Nodal invasions were noticed in 34 patients (23 typical, 11 atypical), 20 with hilar (N1) and 14 with mediastinal lymphadenopathic (N2) nodal invasions. The median diameter of the primary tumor in the N1 group and the N2 group was 33.5 mm and 38.0 mm, respectively.
Carcinoid tumor was located in the right lung (64%) in 67 patients. While 23 lesions were located in the main bronchus (22.5%), 37 were located in the lower lobe (35.5%), 32 in the upper lobe (30.5%), and 12 were located in the intermediate lobe (11.5%) (Fig. 1 ).
Anatomic distribution of lesions.
The aim of surgical treatment was to achieve a complete resection of the entire primary tumor with negative margins; hence, no endobronchial treatment was applied to any of our patients. During thoracotomy, hilar (N1) and mediastinal (N2) lymph-nodes sampling was done. Since 1999, our clinic performs systematic radical mediastinal lymphadenectomy for positive frozen-section cases.
Surgical procedures carried out for the 104 patients included 49 lobectomies, 15 pneumonectomies, 17 bilobectomies (eight with intermediate lobe lesion, nine with lower lobe lesion), seven sleeve lobectomies, four wedge resections, one anatomic segmentectomy, and 11 sleeve bronchial procedures without lung resection. Standard resection was performed in 82 (79%) carcinoid-tumor patients and parenchyma-sparing resection for 18 patients (17%). Wedge resection was performed for the remaining four patients (4%) (Table 1 ).
Applied surgical procedures in accordance with tumor lesions.
None of the patients received neoadjuvant treatment. No operative or postoperative mortality was observed.
Postoperative complications were observed in 17 patients (16%). Most of these complications were related to pulmonary functions as shown in Table 2 . Postoperative hospitalization stay was 10.8 ± 7 days (5–47 days).
Postoperative morbidity.
Follow-up evaluations were performed for 6–190 months (mean 72 months). Pathological staging, histological subtypes, and patient’s characteristics are shown in Table 3 .
Pathological subtypes of resection-applied patients with bronchial carcinoid in accordance with gender and stage.
The median staging of disease, Stage Ia for both male and female groups, was not significant between the groups (p = 0.057). The male group had higher rates of Stages II–III–IV when compared with females (p = 0.037). While the median stage was Stage Ia in the typical group, the median stage of cases in the atypical group was Stage Ib. Therefore, a significant difference in median stages between groups was observed (p = 0.002).
Follow-up was performed by periodic clinical visits and controls until death of the patients. During the follow-up, 15 relapses (14%) were noticed. The most common recurrence was in the lung (40%) followed by the pericardium (20%), mediastinal lymph node (20%), bones (13%), and liver (7%). Bronchial recurrence was not observed. The recurrence rate was higher in ACs than TCs (40% and 8%, respectively). Significant difference in overall survival was noticed between the two groups (p ≪ 0.001).
As many as 7 of the 84 patients in the typical subgroup died from disease recurrence (one N0, two N1, and four N2). The 5-year survival rates for typical carcinoid tumors with respect to staging were 100% for Stage Ia–b, 100% for Stage IIa–b, and 33% for Stage IIIa–b. Significant difference in overall survival was noticed between stages (p ≪ 0.001). The 5-year overall survival rate was 92%. As many as 8 of 20 patients in the atypical subgroup died from disease recurrence (one N0, three N1, and four N2). The 5-year survival rates for AC tumors with respect to staging were 100% for Stage Ia–b, 100% for Stage IIa–b, and 28% for Stage IIIa–b. Significant difference in overall survival was noticed between stages (p ≪ 0.023). The overall survival rate of Stage I was higher than for Stages II and III. The 5-year overall survival rate was 73% (Fig. 2 ).
Overall survival curves of atypical and typical cases, in accordance with stages.
The 10-year survival rate was 93% (n = 70) for the N0 group, 30% (n = 20) for the N1 group and 21% (n = 14) for the N2 group.
The 5- and 10-year survival rates for the 104 cases, irrespective of typical or atypical status, were 89% and 72%, respectively (Fig. 3 ).
Overall survival curves of patients with typical and atypical carcinoid.
The 5- and 10-year survival rates of patients with carcinoid tumor with respect to the type of surgical procedures performed were 75–0% for the wedge group, 84–72% for the segmentectomy–lobectomy–bilobectomy group, 100–57% for the pneumonectomy group and 100–92% for the parenchyma-sparing group. There was a significant overall survival difference between the type of resection performed (p = 0.007). Overall survival was lowest in the wedge group that consisted of only four cases.
4 Discussion
Pulmonary carcinoid tumors are the most frequent benign tumors of the tracheobronchial tree and constitute 2–5% of all lung cancers [1,2]. In addition to the increase in incidence over the past 30 years, histological classification, staging techniques, and surgery options have been developed over the past years; in parallel, there has been an increase in patient series and publications opting for Travis and colleagues’ method to re-classify carcinoid tumors in 1999 [7]. TC and AC have similar pathological and biological characteristics, but different clinical behaviors and prognosis. TC shows less local and distant metastasis with higher survival rates compared with AC [9,18]. TC constitutes 90% of all carcinoid tumors and shows a high degree of differentiation [10]. Rare mitosis (≪2 mitosis per 10 high-power fields) and pleomorphism with no necrosis are diagnostic. The 5-year survival rate is 87–100% and 10-year survival rate is 82–87% [11,16,19]. However, AC tumors have increased mitotic activity (2–10 mitosis), atypical cells, and increased pleomorphism with necrosis. AC patients have high-risk metastasis rate, and the 5-year survival rate is between 56% and 77% while the 10-year survival rate is between 35% and 56% [7,9].
Carcinoid tumors are most frequently seen between the ages of 45 and 55 years; TCs tend to be seen in the early decades compared with ACs. Female and male distribution is equal and in contrast to other lung cancers, these tumors have no correlation with exposure to carcinogens such as tobacco smoke and asbestos [10]. Initial symptoms are primarily cough, hemoptysis, recurrent pulmonary infection, dyspnea, chest pain, wheezing, or fever. However, symptoms vary with tumor location, diameter, and growth pattern. Some patients may be asymptomatic. Carcinoid syndrome is rarely seen together with bronchial carcinoids [10].
ACs constitute 10–35% of all cases [2,12]. Our prevalence for all carcinoid tumors was 19%. Patients with AC were older (median age: 50 years) than patients with TC (median age: 42 years); the difference was statistically significant (p = 0.013).
Our data showed demographic features similar to those observed by other authors with equal gender distribution of 1:1 [2,10]. In our study, the median age of patients with carcinoid tumor was 44 years. When compared with other pulmonary malignancies, most of the patients were relatively younger (in their 4th–6th decades); this was similar to the findings of other authors [2,7,10,11].
There was a significant difference between the typical group and atypical group in our series with respect to smoking history with a higher number of patients with smoking history in the atypical group (p = 0.025). There are reports showing no relation between smoking and carcinoid tumor [10,11] as well as reports showing the possible relationship between AC and smoking history [2].
In most of the published series, the majority of the patients (53–83%) with carcinoid tumors have symptoms [10,11,13,14]. In our study, one-fourth of the patients (23%) were asymptomatic and diagnosis was coincidental. The remaining patients had respiratory symptoms (recurrent pulmonary infection, hemoptysis, cough, and dyspnea).
Chest X-ray, chest CT scanning, and preoperative or intraoperative bronchoscopy were performed in all patients for the purpose of diagnosis. The chest X-rays of nine patients (9%) with carcinoid were normal. The most common anomalies in our chest X-rays findings were similar to those reported by other authors: atelectasis and pulmonary infiltration (56%). Bronchoscopies of 18 patients with carcinoid were normal; the remaining 86 patients (17 asymptomatic and 69 symptomatic) were diagnosed preoperatively as carcinoids. Because of the risk of bleeding, we used to perform our bronchoscopies in the operating environment recently. As suggested by some authors in their reports, we also believe that a more aggressive, accurate diagnostic, and more careful preoperative bronchoscopy is very critical in the selection of the appropriate surgical procedure and in determining the feasibility of the bronchoplastic procedure [9] (Table 4 ).
Patient characteristics and survival results in literature series.
We did not encounter any significant morbidity in our experiences. In recent years, the use of endobronchial laser treatment to perform endoscopic volume reduction of centrally located tumors to determine the base root of the tumor and prevent irreversible lung-parenchyma damage due to recurrent pneumonia related to tumor obstruction was not used as a surgical procedure in any of our patients [15,20].
A total of 82% of patients (66 patients) with centrally located and 75% of patients (18 patients) with peripheral localization were TC. The primary tumor diameter was between 9 and 52 mm and there was no nodal invasion (N0) in 70 of the patients. While the median tumor diameter was 22.5 mm in the N0 group, it was 33.5 mm in the N1 group, and 38 mm in the N2 group. Some authors reported no correlation of lymph node as a prognostic factor, while other authors found nodal invasion to be an important prognostic factor in addition to histological type [9,10]. In our findings, the survival rate was higher in the N0 group when compared with the N1 and N2 groups in both carcinoid types.
Surgery still remains the main treatment choice for pulmonary carcinoid tumors as it provides the longest long-term survival. However, the method of surgery to be selected is still controversial. During the last period of our experience, we preferred limited resections such as segmentectomy in peripherally located TC, and bronchial sleeve resection or sleeve lobectomy, if suitable, for proximal tumors because they have no local recurrence and have a good survival rate. While previously, we used lobectomy or pneumonectomy to a greater extent as the surgical treatment for bronchial carcinoid, after noting that sleeve resections did not increase operative risk and that our long-term follow-up did not reveal any significant local recurrence, we recently started performing more parenchyma-sparing resections.
Mezzetti and colleagues concluded that parenchyma-sparing resection should be considered in early stage TC, but a wide surgical resection is preferable for ACs, even though in the early stage, because of their high recurrence rates [5]. Rea and colleagues reported that their experience in the increased number of sleeve resections in centrally located carcinoids that led to a decreased need for pneumonectomy. They implicated that this type of surgery would be preferable for young people, who have central TCs [9]. Our experience supported these data.
Surgical resection of carcinoid tumors and complete homolateral lymph-node dissection is necessary to determine possible lymph-node metastasis, because of the metastasis rates’ (6–25%) variation [1,16]. The opinions of authors concerning lymph-node dissection vary. While some authors suggest lymph-node sampling or radical lymphadenectomy for only AC, others suggest systematic mediastinal lymph-node dissection as in non-small lung cancer for both carcinoid subtypes [12,17]. We also modified our lymph-node strategy after years of experience. We noticed that lymph-nodes involvement had an effect on local and distant metastasis, worsening the prognosis; hence, after 1999, we started performing surgery with complete lymphadenectomy.
In our experience, histological type and lymph-node involvement were the most important prognostic factors affecting long-term survival. As reported in many other series, we noticed that TC had lesser local or distant metastasis and better 5- and 10-year survival when compared with AC, and also that the major prognostic factor was histological type [8,16]. Moreover, we also found lymph-node involvement (especially N2) to be a bad prognostic factor.
We conclude that histology is the major prognostic factor in pulmonary carcinoid tumors. AC tumors tend to have more local and distant metastasis, more frequent recurrence, and worse prognosis than TC tumors. We assume that with the exception of wedge resections, limited surgical resection can be applied to patients with peripheral, small-sized TC tumor with no lymph-node metastasis; on the other hand, sleeve or bronchoplastic parenchyma-sparing surgery can decrease the need for pneumonectomy in indicated centrally located carcinoid tumors. We also defend more aggressive surgery such as lobectomy or pneumonectomy for the AC subtype and systematic lymphadenectomy for both carcinoid subtypes as a necessity. Furthermore, multidisciplinary treatments such as postoperative adjuvant chemotherapy and/or radiotherapy should be taken into consideration.
