Systematic lymph node dissection in lung metastasectomy of renal cell carcinoma: an 18 years of experience.

BACKGROUND: Pulmonary metastasectomy of renal cell carcinomas (RCC) remains controversial. Thoracic lymph node involvement (LNI) is a known prognostic factor. The aim of our analysis is to evaluate whether patients with LNI, and particularly N2 patients, should be excluded from surgical treatment. METHODS: We retrospectively reviewed data from 122 patients who underwent operations at two French thoracic surgery departments between 1993 and 2011 for RCC lung metastases. RESULTS: The population consisted of 38 women and 84 men; the average age at time of metastasectomy was 63.3 years (min: 43, max: 82). LNI was identified as a prognostic factor using univariate and multivariate analysis (median survival: 107 months vs. 37 months, P = 0.003; HR = 0.384 (0.179; 0.825), P = 0.01, respectively). Although differences in survival between metastases at the hilar and mediastinal locations were not significant (median survival: 74 months vs. 32 months, respectively, P = 0.75), length of survival time was associated with disease-free interval less than 12 months (median survival: 23 months vs. 94 months, P < 0.0001; HR = 3.081 (1.193; 7.957), P = 0.02). CONCLUSION: Although LNI has an adverse effect on survival; long-term survival can be achieved in pN+ patients. Consequently, these patients should not be excluded from surgery. Systematic lymphadenectomy should be performed to obtain more accurate staging and to determine appropriate adjuvant treatment.

FADD protein release mirrors the development and aggressiveness of human non-small cell lung cancer.

BACKGROUND: The need to unfold the underlying mechanisms of lung cancer aggressiveness, the deadliest cancer in the world, is of prime importance. Because Fas-associated death domain protein (FADD) is the key adaptor molecule transmitting the apoptotic signal delivered by death receptors, we studied the presence and correlation of intra- and extracellular FADD protein with development and aggressiveness of non-small cell lung cancer (NSCLC). METHODS: Fifty NSCLC patients were enrolled in this prospective study. Intracellular FADD was detected in patients' tissue by immunohistochemistry. Tumours and distant non-tumoural lung biopsies were cultured through trans-well membrane in order to analyse extracellular FADD. Correlation between different clinical/histological parameters with level/localisation of FADD protein has been investigated. RESULTS: Fas-associated death domain protein could be specifically downregulated in tumoural cells and FADD loss correlated with the presence of extracellular FADD. Indeed, human NSCLC released FADD protein, and tumoural samples released significantly more FADD than non-tumoural (NT) tissue (P=0.000003). The release of FADD by both tumoural and NT tissue increased significantly with the cancer stage, and was correlated with both early and late steps of the metastasis process. CONCLUSION: The release of FADD by human NSCLC could be a new marker of poor prognosis as it correlates positively with both tumour progression and aggressiveness.

Perioperative analysis of biopsies issued from mediastinoscopy.

BACKGROUND: The objective of this study was to evaluate frozen sections of samples obtained at mediastinoscopy for their clinical usefulness. METHODS: This study retrospectively reviewed the records of all patients who underwent mediastinoscopy with perioperative frozen sections in a 1-year period. RESULTS: A total of 123 consecutive patients underwent the procedure. There were no false-positive results. Of the 71 malignant proliferations, 67 were diagnosed from frozen sections. The technique never failed to establish the absence of mediastinal nodal involvement in patients with suspected or proven lung tumors and enlarged nodes (n = 18) who underwent immediate thoracotomy. Frozen sections allowed recognition (n = 36) or strong suspicion (n = 4) of N2 disease in patients subsequently treated by induction chemotherapy. The technique never failed to establish the nonresectability of lung cancer in patients for whom this condition was suspected perioperatively (clinical stage IIIb; n = 10). CONCLUSIONS: Mediastinoscopy with frozen sections remains an extremely useful tool for the management of paratracheal or subcarinal mediastinal disease.

[The place of routine immediate non-invasive ventilation following pulmonary resection in preventing pulmonary complications in patients with COPD (POPVNI Trial)]. / Intérêt de la ventilation non invasive (VNI) systématique en post-opératoire immédiat d'une résection pulmonaire pour prévenir les complications pulmonaires chez les patients BPCO (essai POPVNI).

BACKGROUND: Respiratory complications are common following pulmonary resection and cause a significant mortality. The use of non-invasive ventilation (NIV) in acute respiratory insufficiency (ARI) is now well recognised. The prophylactic use of NIV in the absence of ARI and/or hypercapnia may be equally justified for the physiological benefits expected in the post-operative period following pulmonary surgery. The aim of our study therefore is to evaluate the effectiveness of NIV in the prevention of pulmonary complications in the immediate post-operative care of patients with moderate and severe COPD. METHODS: It will be a multicentre, prospective, randomised, parallel, open ended study of patients with moderate and severe COPD admitted to hospital for pulmonary resection. EXPECTED RESULTS: To determine whether the setting up of NIV immediately post-operatively reduces the incidence of acute respiratory events (acute respiratory insufficiency) and to identify any sub-groups who receive greater benefit from NIV. This study should establish the place of NIV in the immediate post operative care following pulmonary resection.

[Results of FDG-PET scanning in the pre-operative staging of broncho-pulmonary tumors]. / Résultats de la TEP [18F]-FDG dans la stadification préopératoire des tumeurs pulmonaires.

BACKGROUND: Positron emission tomography (PET) with [18F]fluorodeoxyglucose (FDG) has recently established itself as an important imaging strategy in the management of respectable non-small cell bronchial carcinoma (NSCLC). In this study we report our experience of the impact of FDG-PET in the pre-operative assessment of NSCLC. METHODS: In a single centre retrospective study between 01 January 2000 and 31 Dec 2002, 108 FDGPET scans were performed during the preoperative assessment of histologically proven or strongly suspected NSCLC. RESULTS: The sensitivity, specificity and accuracy of FDG-PET for the characterization of a parenchymatous opacity were 96%, 71% and 92% respectively (4 false negatives, 5 false positives). The sensitivity, specificity and accuracy for mediastinal node involvement were 62%, 94% and 84% respectively (10 false negatives and 4 false positives). The sensitivity, specificity and accuracy for the characterization of adrenal nodules were 88%, 100% and 97% (1 false negative) and for satellite pulmonary nodules 50%, 75% and 64% (2 false negatives and 3 false positives). CONCLUSION: FDG-PET is a useful imaging modality in the pre-operative management of NSCLC but is limited particularly in the characterization of lesions less than 10 mm in diameter and in the evaluation of mediastinal lymph nodes.

Reverse ventilation-perfusion mismatch in lung cancer suggests intrapulmonary functional shunting.

We report on a patient with squamous cell cancer of the left lung who was first considered ineligible for surgery because of severe hypoxemia. A ventilation-perfusion scan showed "reverse" ventilation-perfusion mismatch, with 20% of the total lung perfusion going to the left lung, which showed no ventilation with radioactive aerosols. This pattern suggested that the hypoxemia was due to intrapulmonary functional shunting and could therefore be improved by surgical resection of the tumor. Balloon occlusion of the left pulmonary artery resulted in an immediate rise in PaO2, indicating a right-to-left intrapulmonary shunt. After left pneumonectomy, PaO2 levels were normal. This patient provides an example of dysregulation of the pulmonary hypoxic vasoconstriction response in a non-small cell lung cancer. Lung cancer patients with severe hypoxemia should undergo ventilation-perfusion scanning to look for reverse ventilation-perfusion mismatch suggestive of intrapulmonary functional shunting.

Results and prognostic factors in resections of primary tracheal tumors: a multicenter retrospective study. The French Society of Cardiovascular Surgery.

To determine long-term survival and prognostic factors, 208 patients with primary tracheal tumors were evaluated in a retrospective multicenter study including 26 centers. Ninety-four patients had squamous cell carcinoma, four had adenocarcinoma, 65 had adenoid cystic carcinoma, and 45 patients had miscellaneous tumors. The following resections were performed: tracheal resection with primary anastomosis, 165; carinal resection, 24; and laryngotracheal resection, 19. Postoperative mortality rate was 10.5% and correlated with the length of the resection, the need for a laryngeal release, the type of resection, and the histologic type of the cancer. Fifty-nine percent of patients with tracheal cancer and 43% of patients with adenoid cystic carcinomas had postoperative radiotherapy. The 5- and 10-year survivals, respectively, were 73% and 57% for adenoid cystic carcinomas and 47% and 36% for tracheal cancers (p < 0.05). Among patients with tracheal cancers, survival was significantly longer for those with complete resections than for those with incomplete resections. On the other hand, the presence of positive lymph nodes did not seem to decrease survival. Postoperative radiotherapy increased survival only in the case of incompletely resected tracheal cancers. Long-term prognosis was worsened by the occurrence of second primary malignancies in patients with tracheal cancers and by the occurrence of late pulmonary metastases in patients with adenoid cystic carcinomas.

Open window thoracostomy followed by intrathoracic flap transposition in the treatment of empyema complicating pulmonary resection.

OBJECTIVE: Successful treatment of postoperative empyema remains a challenge for thoracic surgeons. We report herein our 12-year experience in the management of this condition by means of open window thoracostomy. METHODS: Open window thoracostomy was used in the treatment of 46 patients with empyema complicating pulmonary resection. A bronchopleural fistula was associated in 39 of 46 cases. Previous operations included pneumonectomy (n = 30), bilobectomy (n = 5), lobectomy (n = 9), and wedge resection (n = 2) performed for benign (n = 10) or malignant (n = 36) disease. In 10 patients open window thoracostomy was definitive because of patient death (n = 2), concomitant major illness (n = 2), tumor recurrence (n = 4), spontaneous closure (n = 1), or patient choice (n = 1). In 36 cases intrathoracic flap transposition was eventually performed. Muscular (n = 29), omental (n = 5), or combined muscular and omental (n = 2) flaps were used to obliterate the thoracostomy cavity and to close a possibly associated bronchopleural fistula. In 9 patients with postpneumonectomy cavities too wide to be filled by the available flaps, a limited thoracoplasty represented an intermediate step. RESULTS: Among patients treated with definitive open window thoracostomy, local control of the infection was achieved in all the survivors (8/8). After open window thoracostomy and subsequent flap transposition, success (definitive closure of the thoracostomy and, if present, of the bronchopleural fistula) was achieved in 27 (75. 0%) of 36 patients. Four initial failures could be salvaged by means of reoperation (initial reopening of thoracostomy and subsequent muscular or omental transposition). CONCLUSION: Open window thoracostomy followed by intrathoracic muscle or omental transposition represents a valid therapeutic option in patients with empyema complicating pulmonary resections.

Congenital respiratory-esophageal fistula in the adult. Report of nine cases and review of the literature.

Congenital tracheoesophageal or bronchoesophageal fistulas, if not associated with esophageal atresia, may not appear initially until adult life. Nine such cases (two tracheoesophageal and seven bronchoesophageal) are reported. The chief presenting symptoms were recurrent bouts of coughing, after drinking, and hemoptysis. In the majority of cases the duration of symptoms exceeded 15 years. The diagnosis was confirmed in seven patients by esophagography, in one patient by bronchoscopy, and in one patient the fistula was discovered incidentally during thoracotomy. The esophageal opening of the fistula was in the lower third in seven patients and in the middle third in two. Bronchoesophageal fistulas communicated with a segmental bronchus in four patients and with a main or lobar bronchus in three. Treatment involved excision of the fistula (five patients) or division and suturing (four patients). Postoperative follow-up revealed no long-term sequelae except persistent chronic respiratory failure in one patient. The respiratory failure had developed before treatment of the fistula. The analysis of this series and a review of the literature underline the high index of suspicion required in all cases of chronic cough and lung suppuration, to diagnose this benign condition before life-threatening complications occur.

Completion pneumonectomy: experience in eighty patients.

OBJECTIVE: Because completion pneumonectomy is a procedure reputed to place patients at risk, we reviewed our results with the objective of identifying factors that influence complications and survival. METHODS: In a 25-year period, 80 completion pneumonectomies were performed after first operations for 17 cases of benign disease and 63 cases of lung cancer (89% stages I and II), with 7 of the latter patients receiving postoperative radiotherapy. Completion pneumonectomy was performed in 18 cases of benign disease and 62 cases of lung cancer: 28 second primary cancers, 26 recurrent cancers, 3 metastases, and 5 primary cancers in patients previously operated on for benign disease. RESULTS: No intraoperative deaths occurred. Postoperative mortality rates were 5% for the entire series, 6.4% for patients operated on for cancer, and 0% for patients operated on for benign diseases. Patients previously irradiated and those operated on for infectious disease were at risk for postoperative empyema and fistula formation. In the cancer treatment group the actuarial 5-year survival was 36%, without significant difference between patients with recurrent and second primary lung cancers. The actuarial 5-year survivals were 51% for patients with stage I disease, 42% for patients with stage II disease, and 18% for patients with stage IIIA disease (P <.05). CONCLUSIONS: Completion pneumonectomy can be performed with an acceptable operative mortality rate and offers a second chance for cure to patients with cancer. Patients previously irradiated and those requiring completion pneumonectomy for infectious benign disease are at risk for postoperative complications.

Bronchioloalveolar lung carcinoma: results of surgical treatment and prognostic factors.

STUDY DESIGN: To determine the long-term results after surgical treatment of bronchioloalveolar lung carcinoma (BALC) and to identify prognostic factors. PATIENTS AND METHODS: A retrospective study of 70 patients (49 men, 21 women), mean age 61+/-10 years, was carried out. Their carcinomas were classified into three clinicopathologic types: nodular or tumoral, pneumonic, and diffuse types. All the diagnosed BALC cases were reviewed and were classified into histologic types: mucinous, nonmucinous (including fibrotic center), and mixed tumors. Univariate and multivariate analyses were carried out. RESULTS: The nodular or tumoral type was identified in 42 patients, pneumonic in 21, and diffuse in seven. Histologically, there were 36 mucinous, 25 nonmucinous, and nine mixed tumors. Resection was complete in 61 instances (87%) and incomplete in five. The 5-year survival rate was 34% in patients with curative resections. Five prognostic factors were identified by univariate analysis, but in multivariate analysis, only three factors remained significant: the absence of symptoms, the TNM stage, and completeness of resection. Thirty-six patients with curative resection (59%) developed recurrences (in the lung in 26 patients; mediastinal lymph nodes, four; distant metastases, nine). The frequency of recurrence was significantly greater in patients with pneumonic-type BALC than in nodular or tumoral types (p<0.01), and pulmonary recurrences were significantly more frequent in pneumonic than in tumoral types (p<0.02). CONCLUSIONS: This study confirmed that the overall prognosis of BALC is not significantly different from that of the other non-small cell lung cancers. We found that the lungs are the predominant site of recurrence in BALC, especially in the pneumonic types. The complete surgical resection of localized BALC offers the best chances of long-term survival.

Surgical management of non-small cell lung cancer with synchronous brain metastases.

OBJECTIVES: Published series on the synchronous combined resection of brain metastases and primary non-small cell lung cancer are small and scarce. We therefore undertook a multicenter retrospective study to determine long-term survival and identify potential prognostic factors. DESIGN: Our series includes 103 patients who were operated on between 1985 and 1998 for the following tumors: adenocarcinomas (74); squamous cell carcinomas (20); and large cell carcinomas (9). Three patients had two brain metastases, and one patient had three metastases; the remaining patients had a single metastasis. Ninety-three patients presented with neurologic signs that regressed completely after resection in 60 patients and partially, in 26 patients. Neurosurgical resection was incomplete in six patients. Seventy-five patients received postoperative brain radiotherapy. The time interval between the brain operation and the lung resection was < 4 months. Pulmonary resection was incomplete in eight patients. RESULTS: The survival calculated from the date of the first operation was 56% at 1 year, 28% at 2 years, and 11% at 5 years. Univariate analysis showed a better prognosis for adenocarcinomas (p = 0.019) and a trend toward a better prognosis for patients with small pulmonary tumors (T1 vs T3, p = 0.068), N0 stage disease (N0 vs N+, p = 0.069), and complete pulmonary resection (p = 0.057). In a multivariate analysis, adenocarcinoma histology also affected the survival rate (p = 0.03). CONCLUSIONS: It seems legitimate to proceed with lung resection after complete resection of a single brain metastasis, at least in patients with an adenocarcinoma and a small lung tumor and without abnormal mediastinal lymph nodes seen on the CT scan or during mediastinoscopy.

Lung resection after high doses of mediastinal radiotherapy (sixty grays or more). Reinforcement of bronchial healing with thoracic muscle flaps in nine cases.

Mediastinal radiotherapy of more than 60 Gy highly compromises bronchial and wound healing after lung resection. Nine patients with primary lung cancers underwent radical resection after high radiation doses. Eight patients had primary lung cancer previously treated by radiotherapy alone (n = 2) or associated with chemotherapy (n = 6). One patient had a tracheal cancer involving the carina that was previously treated by radiotherapy. Seven patients underwent pneumonectomy and one patient underwent lobectomy with reinforcement of bronchial stump closure with use of the serratus anterior muscle. One patient underwent a sleeve lobectomy with bronchial reconstruction wrapped with an intercostal pedicle flap. Five patients had no postoperative complications and four patients had empyema, one associated with a small bronchial fistula. All except one patient were successfully treated by thoracostomy and immediate or secondary transposition of the pectoralis major muscle and the omentum to fill the cavity. These results show that lung resections can be done after high doses of radiotherapy without a high rate of bronchial fistula by using thoracic muscle flaps to reinforce bronchial stumps and anastomoses. In this procedure, surgical dissection is more time-consuming and increases the postoperative empyema rate (4/9). However, the higher long-term survival may justify this choice in selected cases.

Primary mediastinal germ cell tumors. Results of a French retrospective study.

Eighty-seven patients with primary mediastinal germ cell tumors treated between 1983 and 1990 were studied. Among the 23 patients classified as pure seminoma, eight (35 percent) underwent surgery followed by radiotherapy (n = 6), radiotherapy and/or chemotherapy (n = 2); two patients underwent radiotherapy; 13 patients (57 percent) underwent induction cisplatin-based chemotherapy (ten complete responses) followed by radiotherapy (n = 9), second line chemotherapy (n = 2) and surgical resection of residual tumor (n = 2). On completion of treatment, 22 patients (96 percent) with seminoma were free of disease. The two-year Kaplan-Meier survival rate of these patients was 86 percent. Among the 64 patients with nonseminomatous germ cell tumor, 19 patients (30 percent) underwent surgery as first treatment (ten complete resections) followed by chemotherapy (n = 17) and radiotherapy (n = 5). On completion of treatment, 12 of 19 patients were disease free. Forty-five patients (70 percent) underwent induction cisplatin-based chemotherapy (ten complete responses), and 22 of them underwent resection of residual tumor (19 complete resections). Twenty-three patients were treated with first line chemotherapy without postchemotherapy surgery (three complete responses). In summary, 33 patients (52 percent) with nonseminomatous germ cell tumors became free of disease, and seven patients (21 percent) relapsed after achieving a complete response. The two-year Kaplan-Meier survival rate of the nonseminomatous germ cell tumor patients was 53 percent (87 percent if a complete response), with a median survival of 28 months. Despite a worse prognosis than nonseminomaous tumors from other primary sites, this series of mediastinal germ cell tumors has confirmed the efficacy of therapy.

Prognostic factors and long-term results after thymoma resection: a series of 307 patients.

Three hundred seven cases of patients who underwent operation for thymoma (196 of whom had myasthenia gravis) were analyzed to assess the prognostic values of Masaoka clinical staging, completeness of resection, histologic classification, history of myasthenia gravis, and postoperative radiotherapy. According to the Masaoka staging system, 135 thymomas were stage I, 70 were stage II, 83 were stage III, and 19 were stage IV. According to the Verley and Hollmann histologic classification system, 67 thymomas were type 1, 77 were type 2, 139 were type 3, and 24 were type 4. Two hundred sixty patients underwent complete resection, 30 underwent incomplete resection, and 17 underwent biopsy. Postoperative radiotherapy was performed mainly in cases of invasive or metastatic thymoma. Mean follow-up was 8 years; eight patients were unavailable for follow-up. The overall 10- and 15-year survivals were 67% and 57%, respectively. In univariate analysis, three prognostic factors were established: completeness of resection, Masaoka clinical staging, and histologic classification. Furthermore, among patients with stage III thymomas, survival was significantly higher for patients with complete resection than for patients with incomplete resection (p < 0.001). Completeness of resection should therefore be taken into account in clinical-pathologic staging. We did not find any significant difference with respect to disease-free survival between patients who had postoperative radiotherapy and those who did not. In multivariate analysis, the sole significant prognostic factor was completeness of resection. On the basis of these findings, a new clinical-pathologic staging system is proposed.

Peroperative frozen section analysis of TTF-1 antigen expression.

BACKGROUND: The assessment of thyroid transcription factor 1 (TTF-1) expression is a useful way to investigate the origin of lung adenocarcinomas or large cell carcinomas when dealing with a solitary lung nodule in a patient with a history of extrathoracic cancer. However, if immunohistological analysis has not been performed before surgery, a peroperative frozen section may be insufficient to distinguish between a primary pulmonary tumour and a metastatic tumour. AIMS: To develop a technique for the rapid assessment of TTF-1 expression that could improve the ability of frozen section peroperative histological diagnosis to answer such questions. METHODS: A rapid immunohistochemical technique (lasting 30 minutes) to assess the expression of TTF-1 was developed and tested. RESULTS: Among the 45 interpretable cases, results of frozen section immunohistochemistry were similar to those found by the standard immunohistochemical technique for the expression of TTF-1. CONCLUSIONS: This technique enables TTF-1 to be analysed peroperatively, but further prospective studies are needed to assess its usefulness in routine practice.

Management of empyema complicating lobectomy with superior vena cava replacement.

We present the case of a 49-year-old man with right upper lobe adenocarcinoma invading the right brachiocephalic vein and the origin of the superior vena cava. En bloc resection of right upper lobe with the involved venous segments was carried out through a median sternotomy. Venous pathway was reestablished with a Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) prosthesis. Postoperative course was marked by right pneumonia complicated by empyema. The patient underwent thoracotomy with completion pneumonectomy and latissimus dorsi transposition to cover both the prosthesis and the bronchial stump, as well as to fill the cavity. A favorable outcome was observed and long-term survival achieved.

Leiomyosarcoma of the superior vena cava.

Leiomyosarcoma of the superior vena cava is exceptional. A case in a 52-year-old man is described. A treatment by means of neoadjuvant chemotherapy, operation, and adjuvant radiotherapy was performed. This aggressive treatment has permitted the patient to obtain a relatively long survival with a good quality of life.

Preoperative carcinoembryonic antigen level as a prognostic indicator in resected primary lung cancer.

The aim of this study was to evaluate the prognostic significance of elevated preoperative carcinoembryonic antigen (CEA) levels in cases of resected primary lung cancer. Between 1985 and 1989, 152 patients with tumors and CEA levels above 10 ng/mL underwent operation. One hundred twenty-five of them underwent resection of their tumors and the other 27 underwent exploratory thoracotomy only. Fifty-two percent of cancers were adenocarcinomas and 33% were epidermoid. Forty-two resected tumors were classified as stage I, 29 as stage II, 45 as stage IIIa, 7 as stage IIIb, and 2 as stage IV. The 3-year actuarial survival rate was 54% for patients with stage I tumors, 28% for those with stage II, 18% for those with stage IIIa, 44% for those with stage IIIb, and 0% for those with stage IV tumors. The 5 year actuarial survival was 40% for those with stage I tumors, 28% for those with stage II, 7% for those with stage IIIa, and 0% for those with stage IIIb tumors. Preoperative CEA levels increased from stage I to stage IIIa (p < 0.05). However, based on preoperative CEA levels we were not able to predict resectability, because levels were not significantly different between stage IIIa and exploratory thoracotomy-only groups. Adenocarcinoma was not significantly associated with higher CEA levels than was epidermoid, except in stage IIIa disease (p < 0.05). We found a critical unfavorable level of prognostic significance at 30 ng/mL.(ABSTRACT TRUNCATED AT 250 WORDS)

Continuous intercostal analgesia with 0.5% bupivacaine after thoracotomy: a randomized study.

This study was undertaken to evaluate the effectiveness of 0.5% bupivacaine (360 mg/day) as a continuous infusion through an indwelling intercostal catheter inserted intraoperatively in the management of pain after thoracotomy. Eighty-six patients were randomized into three groups: group 1 = intercostal bupivacaine, group 2 = intercostal saline solution, and group 3 = fixed-schedule intramuscular buprenorphine. Supplementary buprenorphine was given as required. Pain and pulmonary function were assessed throughout the first 5 days after operation. Pain score was lower in group 1 than in group 2 for the first 8 hours after operation (p < 0.02). During the first 3 postoperative days, mean postoperative pain scores of 5 or more were recorded in 9% of group 1 patients versus 40% of group 2 patients (p < 0.05) and 13% of group 3 patients (not significant). Total doses of buprenorphine were lower in groups 1 and 2 than in group 3 (p < 0.001). No between-group differences in pulmonary function were observed. Respiratory complications occurred in no patients in groups 1 and 3 versus 5 in group 2 (p < 0.05). Continuous intercostal bupivacaine provided similar early pain control as compared with fixed-schedule narcotics but induced better analgesia with fewer complications than on-demand narcotics alone (group 2).