Abstract
Objective: To evaluate type of surgery, long-term survival and factors influencing outcome in pulmonary carcinoid tumors. Patients and methods: We reviewed data of 252 patients who underwent surgery for carcinoid tumor in 1968–1989 (Group A) and in 1990–2005 (Group B). All cases were reviewed and classified as typical (TC) or atypical carcinoid (AC) according to WHO criteria (1999). Results: There were 174 (69%) patients with TC (167 N0, 6 N1 and 1 N2) and 78 (31%) with AC (56 N0, 13 N1, 9 N2). Surgery consisted of 163 (64.7%) formal lung resections (121 lobectomies, 18 bilobectomies, 14 segmentectomies, 10 pneumonectomies), 76 (30.1%) sleeve or bronchoplastic resections and 13 (5.2%) wedge resections. No perioperative mortality occurred, 17 (6.7%) patients experienced complications. Overall 5, 10 and 15-year survival rate was 90%, 83% and 77%. TC showed a more favourable prognosis than AC (10-year survival rate 93% and 64%; p = 0.00001) as well as N0 patients in comparison with N1-2 patients (10-year survival rate 87% and 50%; p = 0.00005). Group A received lymph-node sampling, Group B received a systematic lymphadenectomy. No difference was found between Group A and B in detection of nodal metastases (10.9% versus 11.9%; p = 0.79), but in Group A we observed 2 lymph-node relapses. In Group B number of sleeve resections significantly increased (2.7% versus 20.4%; p = 0.0001) and number of pneumonectomies showed a significant reduction (7.2% versus 1.4%; p = 0.01). Conclusions: Typical histology and N0 status were important prognostic factors in carcinoid tumors. Parenchyma-sparing procedures must be considered the treatment of choice with systematic lymphadenectomy.
1 Introduction
Bronchial carcinoids are rare, well-differentiated neuroendocrine malignant tumors and account for 1–2% of all lung neoplasms [1]. They are classified in two distinctive forms: typical (TC) and atypical (AC) with different histologic features, clinical course and prognosis. The difference in histological criteria between TC and AC was first described by of Arrigoni et al. [2], later modified by Travis et al. [3] and finally was fixed in 1999 by World Health Organization [4]. TC is a variant of neuroendocrine tumors with a low-grade histological malignancy profile (rare mitosis and pleomorphism, absence of necrosis) and it rarely metastasizes [5,6]. TC shows an excellent long-term prognosis with a 5-year survival rate ranging from 87% to 97% [3,7]. Patients with AC have a greater risk to develop metastases and show a 5-year survival rate ranging from 56% to 77% [5,7]. Surgery represents the treatment of choice for bronchial carcinoid and parenchyma saving resections (sleeve or bronchoplastic procedures) have been suggested in central carcinoid tumors [6,8]. In several Centres surgery comprises, as in lung cancer, a systematic lymphadenectomy because of the risk of lymph-nodes metastases in both type of carcinoid tumors, but the opportunity of a systematic lymphadenectomy is still controversial [7–12].
Aim of the study was to analyze our experience in two consecutive period (1968–1989, Group A and 1990–2005, Group B) in the management of bronchial carcinoids to better define the evolution of surgical strategy and the factors influencing outcome and long-term survival.
2 Patients and methods
We retrospectively reviewed data of 252 patients surgically treated for carcinoid tumors at the Division of Thoracic Surgery, University of Padua from 1968 to 1989 (Group A) and from 1990 to 2005 (Group B). In the first period we operated on 110 patients (43.6%) and in the second one 142 (56.4%) patients underwent surgery. All pathological specimens were reviewed and classified as typical (TC) or atypical carcinoid (AC) according to WHO criteria [4].
Clinically, 187 (84.2%) patients were symptomatic, presenting the following scenario: obstructive pneumonia (74 patients, 29.4%), persistent cough (42 patients, 16.7%), haemoptysis (34 patients, 13.5%), recurrent fever (9 patients, 3.6%), chest pain (8 patients, 3.2%), dyspnoea (6 patients, 2.4%) and other symptoms (12 patients, 4.8%). Carcinoid syndrome was observed only in two cases (0.8%). Sixty-five (25.8%) patients were asymptomatic. Preoperative evaluation included radiological investigations (Chest X-Ray, Chest and upper abdomen computed tomography), tracheo-bronchial endoscopy and, since 2000 the analysis of the expression of Somatostatin receptors in the primary tumor and/or in secondary localizations by using Octreotide marked with Indium 111 (Octreotide-scan).
The tumors were classified as ‘central’ if visualized directly at bronchoscopy or if associated with atelectasis or obstructive pneumonia and ‘peripheral’ when the tumor was not visible at endoscopy.
After the preoperative evaluations the patients underwent surgery: the extension of surgical resection was established on the basis of local growth of the tumor privileging a parenchyma-sparing resection, when indicated. The decision to choose a conservative surgery was not due to the differences in performance status, functional respiratory tests, diffusing capacity, but was taken on bronchoscopic presentation, CT scanning and intra-operative findings as the state of lung parenchyma and the involvement of the frozen sections. Group A received lymph-node sampling, Group B received a systematic lymphadenectomy for histological examination.
3 Statistical analysis
Survival was estimated according to the Kaplan–Meier method. The statistical difference between the survival curves was determined using the log-rank test. Association between categoric variables were determined by using the chi-square test. Multivariate analysis by Cox proportional hazard model was used to investigate the relative importance of different prognostic factors. The parameters evaluated were histology, N status, age, sex, localization (central or peripheral), type of surgery, time of surgery. A p-value lower than 0.05 was considered statistically significant.
4 Results
Table 1 summarizes the demographic, pathological and surgical data of patients in the two periods analyzed. There were 117 females (46.4%) and 135 males (53.6%) with a median age of 45 years (range 8–79 years). On the basis of bronchoscopic findings 195 carcinoids (77.3%) were centrally located and 57 (22.7%) were classified as peripheral. In central carcinoids an endoscopic biopsy permitted a correct pathological preoperative diagnosis in 117 patients. In group A we had a preoperative diagnosis in 40 of 97 patients (41.2%), while in group B in 77 of 98 patients (78.6%). No significant complications as bleeding occurred during endoscopic procedures. In 30 patients (11.9%) a laser-assisted resection of the endoluminal component of the tumor was useful to relief the obstructive symptoms and to better locate the base implant of carcinoid.
Demographic, clinical, pathological and surgical characteristics. Overall and classified by period of surgery
Surgery consisted of 163 (64.7%) formal lung resections (121 lobectomies, 18 bilobectomies, 14 segmentectomies, 10 pneumonectomies), 76 (30.1%) sleeve or bronchoplastic resections and 13 (5.2%) wedge resections. In Group B number of sleeve resections significantly increased (2.7% versus 20.4%; p = 0.0001) and number of pneumonectomies showed a significant reduction (7.2% versus 1.4%; p = 0.01).
No operative or postoperative mortality was seen. Seventeen patients (6.7%) experienced complications: 4 patients presented pneumothorax after chest tube removal, 4 had atrial fibrillation, 3 experienced prolonged air leaks, 2 had haemothorax that required blood transfusions, 2 had pleural empyema after lobectomy, one patient received a left lower lobectomy 2 days after an isolated bronchoplasty of the lower lobar bronchus for bronchial stenosis and one patient had a late cicatricial bronchial stenosis after sleeve lobectomy, treated successfully by laser therapy.
According to the histologic findings there were 174 (69%) patients with TC and 78 (31%) with AC. In Table 2 we show the differences in clinical and pathological characteristics between the two histotypes. Nineteen (7.5%) patients had the involvement of the ipsilateral hilar lymph-nodes (N1) and 10 (4%) patients presented N2 disease. TC had a lower rate of nodal involvement compared with AC (4% and 28.2%, respectively; p = 0.0001). No difference was found between Group A and B in detection of nodal metastases (10.9% versus 11.9%; p = 0.79), but in Group A we observed during follow-up 2 lymph-node relapses (carinal and paratracheal) that required a reoperation.
Differences in demographic and clinical characteristics on the basis of histotype
Follow-up evaluation ranged from 6 months to 36 years (median 121 months). During follow-up 20 (7.9%) relapses were diagnosed. The most common location of recurrences was the liver (55%), followed by lung (25%), bone (20%), adrenal gland (10%), pericardium (10%) and mediastinal lymph-nodes (10%). No bronchial recurrences were seen. Recurrent cancer developed preferentially in AC than TC (17.9% and 3.4%, respectively; p = 0.0001). Eight (40%) patients (7 AC and 1 TC) who manifested relapses of disease were N+.
Overall 5, 10 and 15-year survival rates were 90%, 83% and 77% (Fig. 1 ). The univariate analysis (Table 3 ) reached statistical value for histology (Fig. 2 ), nodal status (Fig. 3 ), sex, age, type of surgery and localization. No differences in survival were showed between Group A and B. Multivariate analysis showed a significant independent prognostic value for histology (p = 0.0004), nodal status (p = 0.004), age (p = 0.003) and sex (p = 0.005).
Overall survival.
Univariate analysis of prognostic factors
Survival by histology.
Survival by nodal status.
5 Discussion
Carcinoid tumors have been classified in the spectrum of neuroendocrine lung tumors although they have a lower grade of malignancy respect to large cell carcinoma and small cell carcinoma. TC and AC show similar pathologic characteristics and biologic markers, but the clinical behaviour and subsequently the prognosis are different for the two distinct form of carcinoids. The very favourable prognosis in long run for TC is justified by the low percentage of local or distant metastases after a radical surgical resection, while ACs have demonstrated in the most of studies [1,5,7,11,12] a trend to a poorer prognosis due to the greater oncological aggressiveness.
These characteristics are evident also in our experience: TC showed a significant better survival than AC (10-years survival 93% and 64%, respectively; p = 0.00001), having a lower percentage of recurrence (3.4% and 17.9%, respectively; p = 0.0001).
In several experiences [5,7,11,12], histology represents the most important factor determining the prognosis: in our series the univariate and multivariate analysis demonstrated the strong power of histologic subtype in conditioning the long-term survival.
Our data on the demographic characteristics demonstrated a small prevalence of carcinoids in males (53.6% versus 46.4%, male/female ratio: 1.15); instead, a significant difference in the age at presentation was noted in our and other experiences [7,13]. In fact patients affected by TC were younger (median 39 years) than those affected by AC (median age 56 years) who had a median age similar to that of patients with lung carcinoma. Male sex and older age have been found significant negative prognostic factors in statistical analysis that could also be explained by the strict association with histology: in fact AC affected preferentially male patients older than those affected by TC (Table 2).
Moreover the poor prognosis observed in patients with peripheral located tumors seems also related to the different preference for peripheral pulmonary localization of TC and AC (28 of 174; 16.1% and 29 of 78; 37.2%; p = 0.0002).
Endoscopy plays a central role in diagnosis and initial management of carcinoids: in our experience the majority of patients (77.4%) had a centrally located tumor visible at bronchoscopy. Out of 195 patients with central carcinoid, 117 (60%) had a preoperative diagnosis thanks to biopsy. Moreover, in the last period we observed a significant higher rate of preoperative diagnosis that can be explained by early acknowledgement of the linked symptoms and a more aggressive and accurate diagnostic approach, especially related to endobronchial biopsies, that in our hands showed no significant morbidity. Therefore, a careful preoperative endoscopic assessment is of fundamental importance in order to define the better surgical treatment and to determine the feasibility of a bronchoplastic procedure. In cases of central tumors obstructing the major bronchi, endoscopic debulking allows to look behind the tumor evaluating its base implant and it allows to treat airway obstruction avoiding the recurrent pneumonia that could damage irreversibly the lung parenchyma. In this line the use in the last years of endobronchial laser treatment has been useful to implement the rate of parenchyma-sparing operations (bronchoplastic and sleeve resections).
Surgery represents the cornerstone of treatment for pulmonary carcinoids and it is the treatment of choice, achieving a long-term survival in case of radical resection. Despite that, a number of surgical aspects are still controversial. Our review of surgical experience covering a long period of time demonstrated an evolution of surgical strategy. Last period of our experience was characterized by a significant increased number of sleeve resection or bronchoplastic procedures in carcinoids centrally located, that led to a significant reduction of pneumonectomies. These results have been obtained thanks to the improvement of surgical techniques, but also thanks to the popularization of parenchyma-sparing operations based on bronchoplastic or sleeve resections which do not alter the oncologic result and obviously guarantee a better quality of life [14]. In particular, in our experience sleeve resections did not increase the operative risk and, at long-term follow up, we did not notice local recurrences even in cases with reduced edges of healthy tissue. Therefore, we think that the modern management of central carcinoids should privilege, when possible, sleeve resections being the oncological results good and the rate of local recurrence low in most experiences [6,14–16]. The need to avoid pneumonectomy gets more weight when we consider that central TC affects preferentially young people.
Kurul et al. [17] found wedge bronchoplasty not a safe procedure, although easy and fast. They stated that wedge bronchoplasty with large margins can cause kinking in the bronchial system, which is one of the major reasons for complications such as atelectasis, stenosis and dehiscence. In our experience we observed one complication due to kinking of bronchus that required a reoperation. We believe that most of all in TC large free margins of resections as required for lung carcinomas are not mandatory being local recurrence rare. Regarding the peripheral tumors, Ferguson et al. [11] in a multicentric retrospective study found limited resections as wedge resection or segmentectomy to be justified in case of peripheral early stage TC tumors because of the unlikely local recurrence and the excellent survival. Mezzetti et al. [18] found an increasing number of relapses in AC treated with atypical resection. Our experience did not show an increased percentage of recurrence in case of limited resection.
Surgical resection of carcinoid tumors should always be combined with complete homolateral lymph node dissection. The necessity for lymph node dissection is justified by the possibility of lymph nodal metastases, which may have an incidence ranging from 6% to 25% [12,19]. The disparity of these data is certainly explained by a different surgical strategy of lymph node dissection: some authors [7,13] use sampling or reserve the radical lymphadenectomy only for AC [11], other authors [6,8,9,12] believe that a systematic mediastinal lymph node dissection is mandatory such as for non small lung cancer, for both types of carcinoids. In our experience we changed surgical strategy: before 1990 we preferred surgical excision combined with sampling and, after a revision of our data [20], we found that the patients with lymph nodes involvement had a worse prognosis for local or distant metastases. Since 1990 we start to treat the patients with surgery and complete lymphadenectomy: although we did not note an increased number of lymph node metastases in comparison with the first period (10.9% versus 11.9%, p = 0.79), no nodal relapses were recorded after 1990.
Prognostic relevance of lymph-nodes involvement has been underlined by several authors: Cardillo et al. [12] found a significant better survival for N0 patients compared with N1 and 2 patients (5-year survival 100%, 84.2% and 22.2%; p = 0.000). Garcia-Yuste et al. [21] and Filosso et al. [7] reported a significant difference in the rate of nodal involvement between TC and AC. Our findings corroborate the previous results: N0 patients had a statistical better survival than N1-2 and a lower percentage of lymph nodes metastases was detected in TC than in AC. Therefore, the investigation for lymph nodes metastases seems an unavoidable requisite to establish the prognosis and eventually to evaluate the opportunity of adjuvant therapies. The accuracy of pathological diagnosis can be increased as demonstrated by Mineo et al. [9], who described the relevance of lymph node micrometastases detected by immunohistochemical techniques.
In conclusion, histology seems the main prognostic factor in carcinoid tumors. AC affected preferentially patients older than TC and demonstrated a more aggressive oncologic behaviour with a high percentage of nodal metastases and distant recurrences. In central carcinoids sleeve and bronchoplastic parenchyma-saving resections should be considered the standard surgical procedure together with a systematic lymphadenectomy. A multidisciplinary treatment should be investigated for AC with lymph node metastases.
Presented at the joint 20th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 14th Annual Meeting of the European Society of Thoracic Surgeons, Stockholm, Sweden, September 10–13, 2006.
Appendix A
Conference discussion
Dr A. End(Vienna, Austria): What is your current strategy with adjuvant therapy in atypical carcinoids?
Dr Rea: We don’t use any kind of chemotherapy in these patients at the moment if we have no residual tumor. Only if we have relapse do we treat these patients medically.
Dr End: You applied a multivariate model, including several putative prognostic factors. The results always depend on the analysis which is used. Did you ever look at other multivariate models?
Dr Rea: Sorry?
Dr End: Did you only do one multivariate model or did you do other models?
Dr Rea: No, we just used the univariate and the multivariate analysis.
Dr G. Egri(Budapest, Hungary): I would like to ask if you have any data on the long-term results. Did you have distant metastases? Second, did you do an evaluation of somatostatin receptors and did you think of any therapy based on that?
Dr Rea: The first question, the distant relapse, yes, we had distant relapse mainly in atypical carcinoids. Your second question?
Dr Egri: Somatostatin receptors in the histology.
Dr Rea: The study of the receptors?
Dr Egri: Yes.
Dr Rea: Now I am trying to study the somatostatin receptors in these patients. We have a study to perform somatostatin receptors in every patient, but we have preliminary data.
Dr K. Athanassiadi(Hannover, Germany): Congratulations on your long series, but I’m a little bit concerned since the first group started in 1968 if you have the correct diagnosis. I mean we did immunohistochemistry for carcinoids in the late 1980s. Are you sure about your diagnosis in this period of time?
Dr Rea: Of course this is one of the limitations, the long period, but in all patients the histology was reviewed, and we excluded all patients who did not show a histologic diagnosis of a carcinoid.
Dr O. Kshivets(Siauliai, Lithuania): What is your personal opinion in the case of central typical carcinoid with using endoscopic cryosurgery and laser surgery as the radical procedure?
Dr Rea: Sorry?
Dr Kshivets: For central localized typical carcinoid, maybe cryosurgery and laser surgery will be enough in terms of completeness.
Dr Rea: The question is if we can use a radical laser resection? No, I do not believe this is a good option, because we have seen that even in typical carcinoid we can have lymph node metastases and we can have tumor relapse in the lymph nodes.
