[Cardiac compression of hydatid origin]. / Compression cardiaque d'origine hydatique.

INTRODUCTION: Cardiac hydatid disease is uncommon and occurs in 0.5 to 2% of patients with hydatidosis. Isolated intrapericardial hydatid cystic disease is extremely rare. OBSERVATION: We report the case of a young woman with cardiac compression due to multiple primary intrapericardial hydatid cysts. Since 1 year, she had gradual general health deterioration including dyspnoea, sweats and weight loss of 8kg. A widening of the mediastinum was observed on chest X-ray. The CT-scan, echocardiography and the dynamic IRM showed multiple mediastinal cysts with mass effect on the heart and main pulmonary artery. The size of the main pulmonary artery was reduced to 5 mm in diameter and the right upper pulmonary vein was nearly closed by posterior cysts. The right and left ventricular ejection fractions were estimated at about 34%. A complete resection of the cysts was performed by sternotomy. The surgical procedure was technically difficult because of major local inflammatory process. The postoperative outcome after an initial pulmonary embolism event was finally favourable. CONCLUSION: Hydatidosis can lead to severe cardiac involvement. These rare forms of hydatid cystic disease must be known even in non endemic regions by surgeons because of increasing mobility of the world's population.

[Evolution of species: Lung cancer evolution over one third of a century]. / De l'évolution des espèces : évolution du cancer du poumon en un tiers de siècle.

INTRODUCTION: Management of non-small cell lung cancer (NSCLC) is getting better and results on long-term survival have improved. We reviewed the modifications observed in surgery over a 32-year time period. PATIENTS AND METHOD: Data of 6105 patients who underwent surgery from 1979 to 2010 were analyzed over three equal time-periods: gender, age, type of surgery, histology, pTNM, tobacco addiction, comorbidity and time periods. RESULTS: Age, number of females and high-risk patients with comorbidity (including the history of a previous cancer) increased with time periods. Number of exploratory thoracotomy (7.7 % to 1.6 %) and pneumonectomy (48 % to 18 %) decreased. Number of wedge resection (0.5 % to 6 %) and lobectomy (42 % to 64 %) increased. Rates of the other types of resection were unchanged. Neoadjuvant treatments accounted for more than 20 % of patients in the last time period. Postoperative mortality (4 %) did not vary but non-lethal complication rates increased (16.9 % to 27.7 %). Global 5-year survival rates dramatically increased with time going from 37.4 % to 49.8 % (P<10(-6)). Survival improvement was observed in the different components of the pTNM and whatever the type of treatment. However, survival was affected by increasing age and multiplication of comorbidities but without impairing the general better outcome trend. CONCLUSION: NSCLC itself, its diagnostic and therapeutic management, and patient's characteristics evolved with time. Survival improved in most studied prognosis factors. Time period factor was of paramount importance and might be included in research dealing with NSCLC.

[Lung cancer measuring 1cm or less: A miniature subset requiring surgery]. / Cancers du poumon de 1 à 10millimètres : des cancers miniatures à opérer.

INTRODUCTION: Lung cancer measuring 1cm or less has an apparently very good outcome. However, the characteristics permitting their oncological management are unappreciated. PATIENT AND METHOD: We reviewed 187 patients with such a cancer (145 men and 42 women, mean age 60.2years) and studied the type of surgery performed, the pTNM, and the histological features. RESULTS: Surgery (19 wedge-resections, 12 segmentectomies, 136 lobectomies, 20 pneumonectomies) was complete (R0) in 97.3%. The tumors, each precisely defined among 98 adenocarcinomas (52.4%), 83 squamous cell carcinomas (44.4%), and 6 others, measured 1mm to 10mm:<5mm (n=41), 6 to 9mm (n=43), and 10mm (n=103). There were 161 pT1 (86.1%), 22 pT2 (11.8%) and 4 pT3; 148 pN0 (79.6%), 18 pN1 (9.7%) and 20 pN2 (10.7%). pN1 and pN2 were present in tumors<5mm (12/41, 29.3%) as well as in the others (26/146, 17.8% P=0.11). Histological examination frequently discovered visceral pleura involvement (tumors:<5mm 12.2% (5/41), 6 to 9mm 7% (3/43), 10mm 13.6% (14/103), P=0.53) and lympho-vascular invasion (12.9%). Five-year survival rate (66.4%) was adversely influenced by age, type of resection, pN and histological features. The survival rate was not better in tumor<5mm. CONCLUSION: Surgical resection allows the local control of lung cancers<1cm and their complete histological study, a key issue in the therapy of the future, which renders surgery an absolute must even in very small tumors.

[Consequences of tobacco smoking on lung cancer treatments]. / Le tabac et ses conséquences dans le cadre du traitement du cancer bronchique.

In France, in 2010, tobacco induced 81% of deaths by lung cancer corresponding to about 28,000 deaths. Continued smoking after diagnosis has a significant impact on treatment. In patients with lung cancer, the benefits of smoking cessation are present at any stage of disease. For early stages, smoking cessation decreases postoperative morbidity, reduces the risk of second cancer and improves survival. Previous to surgery, smoking cessation of at least six to eight weeks or as soon as possible is recommended in order to reduce the risk of infectious complications. Tobacco could alter the metabolism of certain chemotherapies and targeted therapies, such as tyrosine kinase inhibitors of the EGF receptor, through an interaction with P450 cytochrome. Toxicity of radiations could be lower in patients with lung cancer who did not quit smoking before treatment. For patients treated by radio-chemotherapy, overall survival seems to be better in former smokers but no difference is observed in terms of recurrence-free survival. For advanced stages, smoking cessation enhances patients' quality of life. Smoking cessation should be considered as full part of lung cancer treatment whatever the stage of disease.

[Clinical and paraclinical prognostic factors in non-small cell lung cancer surgery]. / Facteurs pronostiques cliniques et paracliniques dans la chirurgie du cancer du poumon non à petites cellules.

INTRODUCTION: Lung cancer prognosis is mainly based on the TNM, histology and molecular biology. Our aim was to analyze the prognostic value of certain clinical and paraclinical variables. PATIENTS AND METHODS: We studied among 6105 patients operated on, divided during 3 time-periods (1979 to 2010), the following prognostic factors: type of surgery, pTNM, histology, age, sex, smoking history, clinical presentation, and paraclinical variables. RESULTS: Postoperative mortality was 4% (243/6105), rate of complications was 23.3% (1424/6105). The 5-year overall survival was 43.2% and 10-year was 27%. Best survival was observed after complete resection (R0) (P<10(-6)), lobectomy (P<10(-6)), lymph node dissection (P=0.0006), early pTNM stages (P<10(-6)), absence of a solid component in adenocarcinoma. Other pejorative factors were: male gender (P=10(-5)), age (P=0.0000002), comorbidity (P=0.016), history of cancer (P<10(-5)), postoperative complications (P=0.0018), FEV lower than 80% (P=0.0000025), time-periods (P<10(-6)). All these factors were confirmed by multivariate analysis, except gender. Smoking was not poor prognostic factor in univariate analysis (P=0.09) but became significant in the multivariate one (P=0.013). CONCLUSION: Medical and human factors, and the general physiological state, play an important role in prognosis after surgery. We do not know their exact meaning and, like studies on chemotherapy, they justify special research.

[Life threatening haemoptysis associated with partial abnormal pulmonary venous return and pulmonary arterial hypertension]. / Hémoptysie grave, retour veineux pulmonaire anormal partiel et hypertension artérielle pulmonaire.

A 73-year-old man with chronic respiratory insufficiency was referred to the intensive care unit because of life threatening haemoptysis and hypoxemia. The patient was in respiratory failure with pulmonary arterial hypertension. The bleeding was issuing from the left inferior lobe. A small arterio-venous fistula was suspected on tomodensitometry. Arterial embolization was contraindicated because of diffuse vascular disease. A left inferior lobectomy was performed. The left superior pulmonary vein was draining into the brachiocephalic vein forming a partial anomalous pulmonary venous return. An anatomic restoration of the venous return to the left atrium was performed. The postoperative course was uneventful. No recurrence of haemoptysis was observed, after 9-month follow-up the vascular anastomosis was patent. Despite the pulmonary vascular bed reduction following lobectomy, echocardiography showed a globally mild improvement of cardiac status that might be attributed to right heart preload amelioration.

[Place of limited resections and prognostic factors in non-small lung cancer]. / Place des résections limitées et leurs facteurs pronostiques dans le cancer bronchique non à petites cellules.

INTRODUCTION: Results of surgery for non-small-cell lung cancer (NSCLC) are poorer after limited resection, wedge and segmentectomy, than after lobectomy. Guidelines recommend avoiding wedge-resection, which new techniques (radiofrequency ablation and cyberknife) tend to replace. This work aimed to study the wedge-resection carcinological value. PATIENTS AND METHODS: NSCLC without previous other cancer history and neoadjuvant therapy measuring less than 31 millimetres and operated from 1980 to 2009 were reviewed. Analyzed variables were: location, gender, age, FEVS, type of resection, histology, pT and pN. RESULTS: There were 66 wedge-resections (10.9%), 32 segmentectomies (5.3%), 507 lobectomies (83.8%), nine postoperative deaths (1.5%), 136 complications (22.5%), 557 complete resections (R0=92%); 72.2% of NSCLC upper lobe location (437/605). Age was more advanced in wedge-resection and segmentectomy, FEVS lower and NSCLC most often a squamous cell pN0 and pStage I carcinoma than in lobectomy. Lymphadenectomy was not performed in half the wedge-resections. Five-year survival rates were poorer after wedge-resection: 50% versus segmentectomy 59.8% (P=0.09), and lobectomy 66% (P=0.0035), but the number of recurrences was similar. Multivariate analysis demonstrated that age, FEVS, type of surgery and lymphadenectomy, pN in pTNM were the only prognosis factors. CONCLUSION: Wedge-resection is less carcinological than segmentectomy when the patient-status and NSCLC location allow performing the latter, but more than the new techniques, because of its pathological yield, when the patient-status and nodule peripheral location allow wedging.

[Obesity and lung cancer: incidence and repercussions on epidemiology, pathology and treatments]. / Obésité et cancer bronchique: incidences épidémiologiques, pathologiques et thérapeutiques.

INTRODUCTION: Obesity and lung cancer are major public health problems. The purpose of this work is to review the data concerning this association. METHOD: We report clinical and epidemiological data on obesity and discuss the impact on the incidence of lung cancer, as well as the safety and efficiency of anti-tumor treatments. RESULTS: Obesity does not contribute to the occurrence of lung cancer, unlike other malignancies. Patients may be more likely to undergo treatment at lower risk. Regarding surgery, obesity makes anaesthesia more difficult, increases the operative duration but does not increase postoperative morbidity and mortality. Chemotherapy and radiotherapy seem to be administered according to the same criteria as patients with normal weight. Paradoxically, survival rates of lung cancer are better in obese patients as well after surgery than after non-surgical treatment. CONCLUSION: Obesity is related to many neoplasms but not to lung cancer. Regarding long-term survival all treatments combined, it has a favorable effect: this is the "obesity paradox".

[Lung cancer surgery and cirrhosis]. / Chirurgie du cancer du poumon chez le cirrhotique.

INTRODUCTION: Lung cancer is the leading cause of death by cancer and cirrhosis is the fourteenth, all causes included. Surgery increases postoperative risks in cirrhotic patients. Our purpose was to analyze this point in lung cancer surgery. METHODS: We collected, among 7162 patients, the data concerning those operated for lung cancer (n=6105) and compared patients with hepatic disease (n=448) to those presenting other medical disorder (n=2587). We analyzed cirrhotic patients' characteristics (n=49). RESULTS: Five-year survival of patients with hepatic disease was lower (n=5657/6105): 35.3% versus 43.8% for patients with no hepatic disease, P=0.0021. Survival of cirrhotic patients was not statistically different from the one of patients with other hepatic disorder, but none survived beyond 10 years (0% versus 26.4%). Surgery in cirrhotic patients consisted in one explorative thoracotomy, three wedges resections, two segmentectomies, 33 lobectomies and 10 pneumonectomies. Postoperative mortality (8.2%; 4/49) was not different for patients without hepatic disease (4.2%; 239/5657) (P=0.32), as well as the rate of complications (40.8%; 20/49 and 24.8%; 1404/5657, P=0.11). Only one postoperative death was associated to a hepatic failure. Multivariate analysis pointed age, histological subtype of the tumour and stage of disease as independent prognosis factors. CONCLUSION: When cirrhosis is well compensated, surgical resection of lung cancer can be performed with acceptable postoperative morbidity and satisfactory rates of survival. Progressive potential of this disease is worse after five years.

[Mucoepidermoid tracheo-bronchial tumors in adulthood. A series of 22 cases]. / Tumeurs muco-épidermoïdes trachéo-bronchiques chez l'adulte. À propos d'une série de 22 cas.

INTRODUCTION: Mucoepidermoid tumours (TME) are rare tumours arising from the submucosal glands of the tracheobronchial tree. The majority of these tumours develop in a benign fashion but some of them are malignant. The latter can be easily mistaken for adenosquamous carcinomas. PATIENTS AND METHOD: We have reviewed 22 patients suffering from TME observed over a period of 25 years. Two arose from the trachea and 20 from the cartilaginous bronchi; 12 of these tumours had macroscopic and histological criteria of low-grade malignancy, 4 had macroscopic and 6 macroscopic and microscopic criteria of high grade malignancy. RESULTS: Prognosis of the latter was very poor and no survival observed after 6 years follow-up, a behavior similar to that observed in non-small cell lung carcinomas and adenosquamous carcinomas. CONCLUSION: The best treatment of these orphan tumours remains surgery.

[Pneumonectomy for benign disease: indication and factors affecting the postoperative course]. / Pneumonectomie pour lésions bénignes: indications et facteurs affectant les suites opératoires.

Pneumonectomy for benign disease is rare but may generate more postoperative morbimortality than when performed for lung cancer. We questioned this assessment and retrospectively reviewed 1436 pneumonectomies and 54 completions of which 82 and 10 performed for benign disease (5.7% and 18.5%, respectively): left n=65 and right n=27. Indications were: post-tuberculosis destroyed lung (n=37), aspergilloma (n=18), bronchiectasis (n=19), infection (n=5), congenital malformations (n=5), inflammatory pseudotumor (n=3), trauma (n=2), post-radiation (n=2) and mucormycosis (n=1). Pneumonectomy consisted of 48 standard and 44 pleuro-pneumonectomies. Stump coverage by flaps was performed in 66.3% (61/92). Complications occurred in 21.7% (20/92) and postoperative deaths in 7.6% (7/92, of which 5 with fungal infections), which was not different than what was observed in lung cancer. There was no difference in fistula formation and mortality regarding the side, the type of resection and the protective role of stump coverage. Considering patients with fungal infections versus others, mortality was 26.3% (n=5/19) and 2.7% (n=2/74), respectively (P=0.0028). Pneumonectomy for benign disease achieves cure with acceptable mortality and morbidity. However, presence of fungal infection should raise the attention for possibility of increased postoperative risks.

Prognostic factors after surgical resection of N1 non-small cell lung cancer.

OBJECTIVES: Non-small cell lung carcinoma (NSCLC) with N1 involvement is associated with 5-year survival rates ranging from 7% to 55%. Numerous factors have been independently reported to explain this heterogeneous prognosis, but their relative weight on long-term survival is unknown. METHODS: Patients who underwent surgical resection for NSCLC in two French centers from 1993 to 2010 were prospectively recorded and retrospectively reviewed. The overall survival (OS) of patients undergoing first-line surgery for pN1 disease was analyzed according to the type of extension, number of metastatic LN, number and anatomic location of metastatic stations. RESULTS: The study group included 450 patients (male 80.2%, mean age 63.3 ± 9.9 years, 5-year overall survival 46%). The number of metastatic station was 1 in 340 (75.6%, single-station disease) and ≥2 in 110 patients (24.4%, multi-station disease). The number of metastatic stations was correlated with the number of metastatic LN (p < .001), and associated with adverse OS (p = .0014). The presence of intralobar metastatic LN (station 12-13-14) was associated with a mechanism of direct extension (p < .001), but did not impact OS (p = .71). The location of metastatic stations was of prognostic significance only in case of multi-station disease, with hilar (station 10) involvement being associated with adverse OS (p = .005). The 110 patients with multi-station pN1 disease and the 134 patients operated on for single-station pN0N2 (skip-N2) disease during the study period yield comparable outcome (p = .52). CONCLUSIONS: In patients with resected pN1 NSCLC, the number of metastatic stations and their location in case of multi-station disease have a prognostic value.

[Lung cancer surgery in a single-lung]. / Chirurgie du cancer bronchique sur poumon unique.

INTRODUCTION: The diagnosis of a second lung cancer in a patient with a previous medical history of lung cancer is no longer a rarity. Also, it is possible to observe a new location in a patient who underwent pneumonectomy in the past. Surgery remains the best treatment. Our objective was to overview this subject. PATIENTS AND METHODS: Among 5611 patients operated in our institution, 186 (3.3%) had metachronous cancer and 17 had previous pneumonectomy (0.7% of pneumonectomies and 0.2% of NSCLC treated in our department). The procedure was diagnostic and therapeutic in 88% of cases (n=15). RESULTS: There were 16 males and 1 female, mean age was 62.5-years. All were smokers (11 were former smokers) and 6 had other medical history. Mean FEV was 52% (range 35-95%). Types of resection were 2 lobectomies, 4 segmentectomies, and 11 wedge resections. There were no postoperative deaths, but two complications. Histological subtype of the first and second cancer was the same in 11 patients. All patients were pN0 after second surgery. The long-term survival (median 33 months) was 35.3% at 5-years and 14.1% at 10-years. Two patients treated with pneumonectomy for their first cancer were pN2. Patients who underwent upper right lobectomy for treatment of their second cancer survived longer than 5-years. CONCLUSION: Surgical resection for lung cancer on single-lung is associated with acceptable morbidity and mortality. Prolonged survival can be achieved in selected patients.

[Squamous cell carcinoma of the trachea: imaging lymph node mapping]. / Cancer épidermoïde de la trachée: imagerie de l'atteinte ganglionnaire.

The anatomy of the trachea lymphatics is poorly understood and the only researches date back to more than one century. Tracheal tumors are very rare, miscellaneous and variously lymphophilic. The cancers of the trachea have no TNM and their lymph node metastases are little studied despite their poor prognosis. We observed 2 cases of squamous cell carcinoma, one in the cervical and the other in the intrathoracic trachea. TDM-3D reformats demonstrated metastatic lymph nodes of the right para-tracheal lymph node chain (2R and 4R) in both patients and in the cervical lymph nodes (right recurrent nerve lymph node chain) in the patient with cervical tumor. Right location of the mediastinal metastases may be explained by the anatomy of the lymph node chain drainage of the lung segments, the right para-tracheal chain being the only one to regularly possess lymph nodes at that level. The right recurrent nerve lymph node metastases of the cervical tumor are explained by common lymph drainage of the cervical trachea towards larynx lymph centres. Besides lymph node metastases, cancers prognosis may also depends on its location in the trachea. Thus, the tracheal tumors are complex and constitute quite as many orphan tumors. Multicentric studies are mandatory to better understand their behavior. Means provided by new imaging techniques might permit establishing a veritable TNM lymph node mapping of these tumors.

[Place and role of the pleura in non-small cell lung cancer dissemination]. / Place et rôle de la plèvre dans la dissémination des cancers du poumon non à petites cellules.

INTRODUCTION: The pleural involvement (PLI) in non-small cell lung cancer (NSCLC) has a poor prognosis, even though it might be very heterogeneous. PATIENTS AND METHODS: A multicentric retrospective descriptive study was performed over 2329 patients who were operated for NSCLC between 1979 and 2010. The patients with PLI were classified in P(Parietal)PLI and V(visceral)PLI and then each subdivided : VPLI to peripheric (VPLI-P) and fissural (VPLI-F) and PPLI to mediastinal (PPLI-M) and costal (PPLI-C). Characteristics and survival were compared between the subgroups as well as with patients without PLI (WPLI, n=1439). RESULTS: The sex-ratio was 2.8 (males: n=1713). The PLI patients were significantly younger, with a less sex-ratio, less R0 resections (96% versus 98.7%, P=0.000076), and less N0 (60% vs 70%, P<10(-6)) as their 5-year survival (45.7% vs 55.5%, P<10(-6)). The PLI was related to the size of NSCLC (P<10(-6)) and N2 involvement (P=0.0020). It was less frequent after neoadjuvant treatment (36.2% vs 39.1% P=0.03). In the VPLI-F or PPLI-M, pneumonectomies were more frequent (P<10(-6)). In VPLI-P (n=196/561), there were more pN1 and pN2 (P=0.0065) with a 5-year survival of 42.9% vs 54.4%, P=0.013. In multivariate analysis, the PLI was not an independent prognostic factor contrary to age, sex, type of resection, pT and pN. CONCLUSION: The pleura play a major role in NSCLC dissemination. Its involvement affects pN, the type of surgical resection and justifies the use of neoadjuvant treatment.

[Lung cancer surgery in solid organ transplanted patients]. / Chirurgie du cancer du poumon chez les greffés d'organe solide.

BACKGROUND: The incidence of lung cancer is reputed to be higher and prognosis worse in solid organ transplant recipients than in the general population. Our purpose was to review the results of surgery in this group of patients. METHODS: We retrospectively reviewed 49 male and 6 female patients; mean aged 60.6 years (38-85). Transplanted organ was heart (n = 37), kidney (n=12), liver (n = 5) and both-lungs (n = 1); 48 patients had smoking habits and 42 heavy comorbidities (76.4%). Lung cancer was diagnosed during surveillance (78.2%, n = 43) or because of symptoms (21.8%, n = 12). We reviewed TNM and other main characteristics, among them histology (squamous-cell-carcinoma n = 23, adenocarcinomas n = 24, others n = 8). RESULTS: Surgery consisted of: exploratory thoracotomy (n = 2), wedge resections (n = 6), segmentectomy (n = 1), lobectomy (n = 42), pneumonectomy (n = 4). Postoperative mortality was 7.4% (n = 4) and complication rate 34.5% (n = 19). Five-year survival rate was 46.4% (65.4% for stage I patients, n = 25). Among the 35 dead patients during follow-up, 14 died of their lung cancer (40%). Two had been re-operated from another lung cancer: one after 3 and 8 years who survived 16 years, and the other after 2 years who survived 70 months. CONCLUSIONS: Surgery results are good and postoperative events acceptable despite theoretically increased risks. This also supports performing a close follow-up of transplanted patients and particularly those with smoking history in view of detecting lung cancer appearing at an early stage.

[Place of bilobectomy in pulmonary oncology and prognostic factors in NSCLC]. / Place des bilobectomies en oncologie pulmonaire et leurs facteurs pronostiques dans le cancer bronchique non à petites cellules.

INTRODUCTION: Bilobectomy may be performed for different reasons and lung tumors. There are still controversies regarding the results of this procedure. We reviewed our experience of bilobectomy to evaluate the particularities of this resection. METHODS: The clinical files of patients operated on for lung tumors in two French centers between 1980 and 2009 were prospectively recorded and retrospectively analyzed. The characteristics, management, pathology, and survival after right-sided resections for non-small cell lung cancer (NSCLC) were then compared. RESULTS: During the study period, 3280 right-sided resections were performed, including 235 bilobectomy (7%), for NSCLC in 192 cases (82%). Lower-middle lobectomy (LML) represented 60% of bilobectomy, with carcinoid tumors and squamous cell carcinoma being more frequent in this group. Upper-middle lobectomy (UML) represented 40% of bilobectomy, with less postoperative complications and mortality in this group. In N0-NSCLC, the rate of postoperative mortality and 5-year survival rates after bilobectomy (4.7% and 46.1%, respectively) were intermediate between lobectomy (2.7% and 52.6%) and pneumonectomy (9.6% and 31.7%, P<10(-6) for both comparisons). There was no significant difference in 5-year survival rates according to the type of bilobectomy and the performance of any induction therapy. CONCLUSION: Bilobectomy is associated with acceptable in-hospital mortality and encouraging 5-year survival rates despite an increased incidence of postoperative complications. Approximation in survival of UML and pneumonectomy and of LML and lobectomy may be due to differences in histologic features with different fissure extension and interlobar node involvement.

[pT4 non-small cell lung cancer: Surgical characteristics in present practice]. / Cancers du poumon non à petites cellules pT4 : aspects chirurgicaux en pratique courante.

INTRODUCTION: pT4 is a group of miscellaneous tumors: our goal was to revisit their surgical reality. METHODS: The different characteristics and prognostic factors of lung pT4 (n=403) were analysed according to three subgroups: G1 - by direct extension; G2 - by nodule in other ipsilateral lobe; G3 - because of both. RESULTS: There were 332 males and 71 females mean aged 61.5 years. Surgery [exploratory: 89 (22.1 %), lobectomy: 149 (37 %), pneumonectomy: 169 (41.9 %)] was followed by 26 postoperative deaths (6.5 %), 82 complications (20.3 %) and concerned few pN0 (47.6 %). G1 (n=196) and G3 (n=53) were not different. By comparison with them, G2 (n=53) were mainly females (24\13 %), with less explorative thoracotomy (2.6\34 %), more complete R0 resections (77\29 %), less pneumonectomy (31\47 %), more small sized tumors (mean: 37\57 mm), more adenocarcinoma (67\32 %), more N0 tumors (48\31.7 %) and stages IIIA disease (46.7\56 %). G2 5-year survival rates were higher (G2: 22 %; G1: 13 %; G3: 15 %); G1 rates depended of the invaded structure (20.9 % for the vertebra down to 0 % for the esophagus and carina). pN2 rates were not very high but not different between groups (G1: 13.6 %; G2: 15.6 %; G3: 14.3 %; P=0.52). Multivariate analysis demonstrated completeness and type of resection, stage and age as independent factors of prognosis. CONCLUSION: Surgery for pT4 is justified provided rigorous selection of extension forms. However, assimilating extension and ipsilateral lobe nodule in a same group does not obey to surgical reality.