[Value of a long duration temporary caval filter in critical thrombo-embolic situations]. / Intérêt d'un filtre cave temporaire de longue durée dans les situations thrombo-emboliques difficiles.

In the event of proximal venous thrombosis with a risk of pulmonary embolism, contraindications to or complications of anticoagulant treatment are not uncommon in surgical or intensive care units. These are worrying but temporary situations and represent a classic indication for partial interruption of the inferior vena cava, for which permanent caval filters are not usually suitable. Temporary filters are an attractive option in this context, as long as they are safe and stable, can be left in place long enough to permit normalization of the thrombosis and anticoagulation problems and can then be removed in all circumstances, whether or not they have trapped a thrombus while in place. Most temporary filters do not meet all these criteria and nor do permanent filters with a removal option. We tested the new Tempofilter II filter with increased stability, which has a smooth geometry and can be implanted for up to 6 weeks, in 13 patients. We selected two documented cases concerning, firstly, a contraindication to anticoagulants and, secondly, recurrent thrombosis in heparin-induced thrombopenia. In both cases, the filter trapped a thrombus and prevented a pulmonary embolism, which would have been poorly tolerated and difficult to treat. The filters were removed without any difficulty after 4 and 6 weeks. We did not observe any complications related to infection or migration. Monitoring is recommended throughout the implantation period, in order to identify any clots trapped in the filter and to monitor their lysis or non-emboligenic fibrous structure, authorising removal of the filter.

Factors affecting long-term survival after en-bloc resection of lung cancer invading the chest wall.

OBJECTIVE: Several reports emphasize the importance of en-bloc resection as the optimal surgical treatment of lung cancer with chest wall invasion. We investigated possible factors which could affect long-term survival following radical resection of these tumors. METHODS: Between 1981 and 1998, 100 patients (90 male; ten female), with a median age of 60 years (36-84), underwent radical en-bloc resection of non-small cell lung cancer (NSCLC) with chest wall involvement. Patients with superior sulcus tumors invading the thoracic inlet were excluded from this series. There were 43 squamous and 57 non-squamous tumors. The median number of resected ribs was three (1-5). Lung resection included 73 lobectomies, two bilobectomies, 18 pneumonectomies and seven segmentectomies. Chest wall resection also extended to the sternum in one patient, the transverse process in one, the costotransverse foramen and hemivertebrae in two. All patients had a complete resection. Sixty-three patients received postoperative radiotherapy and 12 received chemotherapy. Histological data, including differentiation and depth of chest wall invasion, were carefully reviewed. The effect of various factors on survival were studied. RESULTS: There were four in-hospital deaths. Lymph node involvement was negative on surgical specimens in 65 patients, and 28 patients had positive N1 nodes; the final histology revealed seven N2 diseases. Chest wall invasion was limited to the parietal pleura in 29 patients and included intercostal muscles, bones and extrathoracic muscles in 67, 24 and seven cases, respectively. The overall 2-year survival rate was 41%. The 5-year survival for patients with N0, N1 and N2 disease was 22, 9 and 0%, respectively. A local recurrence occurred in 13 patients, with four having a new resection and 45 patients developing systemic metastases. The nodal status (N0-1 vs. N2; P=0. 026) and the number of resected ribs(<2 vs. >2; P=0.03) were survival predictors in univariate analysis. By multivariate analysis, the two independent factors affecting long-term survival were the histological differentiation (well vs. poorly differentiated; P=0. 01) and the depth of chest wall invasion (parietal pleura vs. others; P=0.024). CONCLUSIONS: Histological differentiation and depth of chest wall involvement were the main factors affecting long-term survival in this series. The role of induction chemotherapy for tumors with poor prognosis should be investigated.

Experimental tracheal and tracheoesophageal allotransplantation. Paris-Sud University Lung Transplantation Group.

We investigated the effects of allograft perfusion with a preservative technique and of combined thyrotracheoesophageal implantation on airway epithelium of long segments of thyrotracheal grafts allotransplanted on their own vascular pedicles into immunosuppressed pigs. Four groups of five animals each underwent heterotopic (into the neck) thyrotracheal (group 1) and thyrotracheoesophageal (group 2) and orthotopic thyrotracheal (group 3) and thyrotracheoesophageal (group 4) allotransplantation. Allograft revascularization included (1) interposition of donor right subclavian artery--incorporating the inferior thyroid artery--to recipient right carotid artery (end-to-end fashion) and (2) end-to-side anastomosis of donor anterior vena cava to recipient right external jugular vein. All thyrotracheoesophageal blocks were harvested after inferior thyroid artery perfusion with 4 degrees C Euro-Collins solution. The overall lengths of tracheal and esophageal grafts were 10.7 +/- 2.7 cm and 13.4 +/- 3.6 cm, respectively. In the heterotopic groups, all allografts were viable and histologically normal at postmortem examination and the incidence and severity of airway ischemia and rejections (at equal residual levels of cyclosporine) were not different between groups 1 and 2. In the orthotopic groups, the first two pigs died of airway collapse with histologically normal grafts. In the remaining pigs, temporary airway stenting was inserted and allografts remained viable and histologically intact for their entire length 30 days after transplantation. Transplanted tracheal smooth muscles had concentration-dependent contractions and relaxations similar to those of nontransplanted (native) tracheas. This study documents the feasibility of allotransplanting long tracheal and esophageal segments on their own vascular pedicles and demonstrates that allograft preservation and thyrotracheoesophageal transplantation are equally effective in minimizing airway ischemia. Thyrotracheoesophageal transplantation does not enhance recipient alloimmune response compared with thyrotracheal transplantation alone.

Heterotopic pig model for direct revascularization and venous drainage of tracheal allografts. Paris-Sud University Lung Transplantation Group.

A macrosurgical technique of thyrotracheal harvesting and direct revascularization with and without venous drainage in a heterotopic thyrotracheal and immunosuppressed allograft in the pig model is described. Harvesting included en bloc cervicothoracic exenteration of the aortic arch and its supraortic trunks, anterior vena cava, jugular veins, subclavian vessels, thyroid gland, cervicothoracic trachea, and esophagus. This technique conserves the tracheal arterial supply provided by either the right or left subclavian artery, directly or indirectly via the inferior thyroid artery, and venous return provided by the anterior vena cava, directly or indirectly via the descending cervical vein. In recipients, implantation included (1) arterial end-to-end anastomoses of the proximal and postscalenic stumps of donor's subclavian artery to the proximal and prescalenic stumps of recipient's subclavian artery; (2) end-to-side venous anastomosis of the donor's anterior vena cava to the recipient's brachiocephalic venous trunk; and (3) heterotopic implantation of the proximal and distal orifices of the grafted trachea into the neck. Ten adult Large White pigs underwent direct revascularization of a thyrotracheal allograft with (n = 6, group 1) and without (n = 4, group 2) venous drainage. All grafts of group 2 exhibited a venous infarction, extensive inferior thyroid artery thrombosis, and ischemic and suppurative thyrotracheal necrosis 1 to 2 days after transplantation. In group 1, the length of the grafted trachea and number of rings were 9.75 +/- 1.5 cm and 22.1 +/- 3.3, respectively; ischemic time was 236.3 +/- 338.3 minutes. Group 1 pigs were put to death 4 (n = 4) and 3 (n =2) weeks after transplantation. All tracheal grafts had histologically normal airway epithelium; isolated areas of necrotic ischemia of the chorion and submucosa lasted for the first 7 days after transplantation but disappeared after epithelial regeneration. Premortem angiograms showed that all vascular anastomoses were patent. Grafts were histologically normal at postmortem examinations and all but one had no rejection. This large animal model demonstrates that long tracheal allografts might be transplanted by means of this direct revascularization and venous drainage technique.

Transforming growth factor beta in normal human lung: preferential location in bronchial epithelial cells.

BACKGROUND: Transforming growth factor beta (TGF-beta) is an immunomodulatory cytokine regulating the proliferation and differentiation of various cell types. It also contributes to the maintenance of tissue architecture by influencing the production of extracellular matrix components. TGF-beta has been detected in bronchoalveolar lavage fluid from normal human lung, but the nature and distribution of cells containing TGF-beta in this organ remain unknown. METHODS: Fourteen normal human lung specimens were studied by immunohistochemistry with a monoclonal antibody recognizing TGF-beta 1, TGF-beta 2 and TGF-beta 3. RESULTS: TGF-beta was detected in all cases. Bronchial epithelial cells contained the largest amounts of TGF-beta. In these cells the staining was brightest at the apical pole. Macrophages and smooth muscle cells also contained TGF-beta, although less than epithelial cells. No TGF-beta was detected in other cell populations, including endothelial cells, fibroblasts, and pneumocytes. CONCLUSIONS: The bronchial epithelial compartment appears to be the main location of TGF-beta in the normal human lung, suggesting that this cytokine has a pivotal role in the immunological properties of the bronchial mucosa.

Extended operations after induction therapy for stage IIIb (T4) non-small cell lung cancer.

Twenty-three patients with stage IIIb (T4) non-small cell lung cancer received induction chemotherapy (median, 2 cycles) with (n = 12) or without (n = 11) radiation (median, 45 Gy) before operation. Nine tumors involved the carina (n = 8) or lateral tracheal wall (n = 1), 11 were located centrally and invaded the proximal pulmonary artery (n = 6), veins (n = 3), or both (n = 2), three were apical tumors involving T4 structures, and six were associated with histologically diseased mediastinal nodes. Five complete and 18 partial responses were observed after induction treatment. Resection of all residual tumor at the primary site and involved vestiges was possible in 21 patients (91%); in two apical tumors, tumor was left behind. Nine right tracheal sleeve and 11 intrapericardial pneumonectomies and three resections of apical tumors were performed; 11 patients (48%) had radical mediastinal lymph node dissection. Complete sterilization of the primary tumor was observed in 3 patients (13%). Mean operating time was 209.3 +/- 86.8 minutes, and mean blood loss was 896.9 +/- 1031 mL. Major postoperative complications occurred in 6 patients (26%), including hemothorax requiring drainage (n = 1) or reoperation (n = 1), acute distress syndrome (n = 2), and bronchopleural fistula (n = 2), and their incidence was significantly higher (p = 0.0003) among patients receiving induction chemoradiation than among those receiving chemotherapy alone (42 versus 9%). Early (< 1 month) postoperative mortality was 8.6% (n = 2). With a median follow-up of 25 months (range, 12 to more than 39 months), the projected 3-year overall survival was 54%.(ABSTRACT TRUNCATED AT 250 WORDS)

Chest wall reconstruction following resection of large primary malignant tumors.

Reconstructive procedures following radical resection of large primary malignant chest wall tumors (PMCWT) continue to evolve. Between 1982 and 1993, 32 consecutive patients (18 males/14 females) with a median age of 47 years (range, 12-77) underwent radical resection for large (median 10 +/- 5.4 cm) PMCWTs arising either from the bone (n = 15) or soft tissues (n = 17) of the chest wall. Nine (28%) had previous surgical resection before referral. Sixteen (50%) required extensive skin excision. Twelve sternectomies (5 total and 7 partial) and 20 lateral chest wall resections were performed. In this latter group, 16 patients (80%) had at least three ribs resected. Resection extended to the lung (10 wedge resections, 2 lobectomies and 1 pneumonectomy) in 13 patients, diaphragm in 3, abdominal wall in 2, brachiocephalic and subclavian vessels in 5, superior vena cava in 1 and upper limb in 1. Stability of the chest wall was obtained with prosthetic material in 27 patients, including Marlex (n = 21), polytetrafluoroethylene (PTFE) (n = 4) and polyglactin (n = 2) meshes. After sternectomy, six patients had a methyl methacrylate mesh reinforcement while soft tissue reconstruction was carried out using the pectoralis major muscle (PM), either alone with skin advancement (n = 8) or as a myocutaneous flap in three males (unilateral n = 2, bilateral n = 1) and by a latissimus dorsi (LD) myocutaneous flap in one female. Muscle transposition was used to reconstruct defects of the lateral chest wall and included 10 LD, 6 PM and 2 serratus anterior (SA) muscles, with associated advancement of the diaphragm in two cases.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparative outcome of heart-lung and lung transplantation for pulmonary hypertension.

Despite the development of several lung transplantation procedures, the most advantageous for pulmonary hypertension remains controversial. Between 1986 and February 1992, 30 patients with end-stage primary pulmonary hypertension (n = 24), chronic pulmonary embolism (n = 4), and hystiocytosis X (n = 2) underwent heart-lung (n = 21), double lung (n = 8), or single lung (n = 1) transplantation. Indications for double lung transplantation were similar to those for heart-lung transplantation, and the preoperative clinical and hemodynamic parameters were not significantly different between the two groups. There were no intraoperative deaths, but two reoperations were needed for pleural hematoma. Five early deaths were related to graft failure (two heart-lung transplantations), mediastinitis (one heart-lung transplantation), multiorgan failure (one double lung transplantation), and aspergillosis (one double lung transplantation). There was a similar improvement in early (days 0 and 2) and late (6 months postoperatively) right-sided hemodynamic function in patients undergoing heart-lung and double lung transplantation. Three double lung transplant recipients had early and reversible left ventricular-failure. The early postoperative course of the one patient who had single lung transplantation was characterized by severe pulmonary edema, left ventricular failure, and persistent desaturation and later on by moderate pulmonary hypertension and an important ventilation/perfusion mismatch. The pulmonary function results were also similar in the heart-lung and double lung transplantation groups. The overall projected 2- and 4-year survivals were 49% and 41%, respectively, and were not significantly different between the heart-lung and double lung recipients. Results demonstrate that heart-lung and double lung transplantation are equally effective in obtaining early and durable right-sided hemodynamic and respiratory improvement and similar respiratory function. In patients with pulmonary hypertension, double lung transplantation should be preferred to single lung transplantation because of the critical postoperative course and the uncertain long-term results of single lung transplantation.

Anterior transcervical-thoracic approach for radical resection of lung tumors invading the thoracic inlet.

We describe an original anterior transcervical-thoracic approach required for a safe exposure and radical resection of non-small-cell lung cancer that has invaded the cervical structures of the thoracic inlet. Through a large L-shaped anterior cervical incision, after the removal of the internal half of the clavicle, the following steps may be performed: (1) dissection or resection of the subclavian vein; (2) section of the anterior scalenus muscle and resection of the cervical portion of the phrenic nerve, if invaded; (3) exposure of the subclavian and vertebral arteries; (4) dissection of the brachial plexus up to the spinal foramen; (5) section of invaded ribs; and (6) en bloc removal of chest wall and lung tumor, either directly or through an extension of the cervical incision into the deltopectoral groove. An additional posterior thoracotomy may be required for resection of the chest wall below the second rib. Between 1980 and 1991, 29 patients underwent radical en bloc resection of the inlet tumor, chest wall (ribs 1 and 2), and underlying lung, either through the anterior transcervical approach alone (n = 9) or with an additional posterior thoracotomy (n = 20). The inferior root of the brachial plexus, either alone (n = 11) or with the phrenic nerve (n = 4), was involved and resected in 15 patients (52%). Twelve patients (41%) had a vascular involvement that included the subclavian artery alone (n = 3); subclavian artery and subclavian vein (n = 3); subclavian artery, subclavian vein, and vertebral artery (n = 2); subclavian artery and vertebral artery (n = 1); subclavian vein alone (n = 1); vertebral artery alone (n = 1), or subclavian artery and vertebral artery (n = 1). The subclavian artery was revascularized either with a prosthetic replacement (n = 7) or an end-to-end anastomosis (n = 2), and the median graft patency was 18.5 months (range, 6 to more than 73 months); only 1 patient had postradiotherapy graft occlusion in the revascularized artery 6 months after operation. We performed 14 wedge resections, 14 lobectomies, and 1 pneumonectomy. There were no operative or hospital deaths. Postoperative radiotherapy (median, 56 Gy) was given to 25 (86%) patients, either alone (n = 14) or in combination with adjuvant systemic chemotherapy (n = 11). With a median follow-up time of 2.5 years, overall 2- and 5-year survivals were 50% and 31%, respectively. This transcervical-thoracic approach affords a safe exposure and radical resection of non-small-cell lung cancer involving the thoracic inlet and results in encouraging long-term survival.

Technique for resecting primary and metastatic nonbronchogenic tumors of the thoracic outlet.

Although primary or metastatic nonbronchogenic tumors infrequently arise in or involve the thoracic outlet, they represent a major surgical challenge because of their tendency to encapsulate outlet structures. Fourteen patients with a histologically proven primary (n = 8) or metastatic (n = 6) nonbronchogenic outlet tumor were treated by an anterior transcervical approach, including an L-shaped cervicotomy extended into the deltopectoral groove, resection of the internal half of the clavicle, and, in the case of tumor involvement, resection of the jugular and subclavian veins, phrenic nerve, subclavian artery, brachial plexus, and ribs. All patients underwent a radical resection. Tumors extended to bony (usually the first rib), muscular (usually the anterior scalenus muscle), and nerve (usually the phrenic nerve) outlet structures in 8, 10, and 7 patients, respectively. Ten patients had involvement of outlet vessels: 6 had simple ligature (n = 5) or wedge resection (n = 1) of the subclavian vein and related branches, 1 had revascularization of both the subclavian vein by an end-to-end anastomosis and the subclavian artery by a ringed polytetrafluoroethylene graft, 1 had revascularization of the subclavian artery alone, and 2 had revascularization between the left brachiocephalic vein and superior vena cava (ringed polytetrafluoroethylene graft). Follow-up venograms showed complete patency of the anastomoses. There was one postoperative death (7%) due to multiorgan system failure. Other complications were mild and short-lasting. With a median follow-up of 3.4 years, all patients but 1 (who had systemic progression) are alive and disease free 3 to 127 months postoperatively.(ABSTRACT TRUNCATED AT 250 WORDS)

Tracheal allograft replacement. An unsuccessful method.

Replacement of the tracheal conduit remains a difficult and unresolved surgical problem. We present an experimental study of 20 pigs undergoing replacement of the cervical trachea using glutaraldehyde (n = 5), glycerol (n = 5), lyophilized (n = 5) and cryopreserved (n = 5) pretreated allogenic grafts (AGs). In the lyophilized group, a stainless steel spiral endoprosthesis was used as stent. A segment of native trachea measuring 3.2 +/- 0.3 cm (range: 1-5 cm) in length and 1.4 +/- 0.02 cm in diameter was resected and replaced with AGs measuring 3.8 +/- 0.2 cm in length and 1.4 +/- 0.2 cm in diameter. Neither immunosuppressive agents nor steroids were given. Animals were followed up with weekly bronchoscopy, and trachea and chest roentgenography. Those receiving glutaraldehyde AG (21.6 +/- 6.4 days) and lyophilized AG (19.5 +/- 7.8 days) had a longer survival than those receiving glycerol AG (6.8 +/- 0.3 days) and cryopreserved AG (5 +/- 0.5 days). At postmortem, grafts were examined grossly and with light microscopy. The cause of death was always airway obstruction, and the underlying processes were: 1) collapse due to cartilaginous microscopic necrosis in cryopreserved and glycerol AGs; 2) necrosis of allograft which crossed the spiral stent for lyophilized AGs; 3) granulation formation, tissue necrosis and anastomosis leakage in glutaraldehyde AGs. Results demonstrate that passage of necrotic tissue across the grafts represent the common failure denominator, making the different AGs studied unsuitable for long-segment tracheal replacement.

Laryngotracheal resection and reconstruction for postintubation subglottic stenosis. Lessons learned.

Between 1981 and June 1992, 26 consecutive patients with a postintubation subglottic stenosis (21 circumferential, 2 anterolateral) underwent the Pearson operation. Subglottic stenosis resulted from a complication of mechanical ventilation with endotracheal intubation with (n = 14) or without (n = 12) tracheostomy (median placement: 25 days). One patient had an associated laryngopharyngeal and tracheoesophageal fistula. Overall, the upper limit of the stenoses lay 1.8 +/- 0.3 cm below the vocal cords, falling in the range of 1 to 2 cm in 88% of patients; they measured 2.9 +/- 0.8 cm in length and the diameter at the level of the maximum stenotic process was 0.5 +/- 0.1 cm. Operations were performed without dissection of the recurrent nerves and plicature of the membranous trachea. Because of scarred mucosa at a higher level, one vertical section of the posterior cricoid plate with interposition of autogenous costal cartilage and 2 subtotal cricoid plate resections with stenting were necessary. The mean length of resection was 3.6 +/- 0.8 cm (range: 2-5 cm) and 88% of them ranged within 2.8 and 5 cm. Twelve thyrohyoid and 3 supralaryngeal releases were performed. Six patients required postoperative tracheostomy, but all were extubated within 24 h. Good results were obtained in 24 (96%) surviving patients; 1 failure and 1 postoperative death (sudden myocardial infarction) occurred. The results confirm that the Pearson operation is an adequate treatment for subglottic stenosis extending up to 1 cm below the vocal cords and measuring up to 6 cm in length. Dissection of both the recurrent nerves, plicature of the membranous trachea, postoperative decompressive tracheostomy and stenting are not necessary.

Human immunodeficiency virus and infected aneurysm of the abdominal aorta: report of three cases.

Three patients who were seropositive for human immunodeficiency virus underwent surgery for infected aneurysm of the abdominal aorta. Fever and abdominal pain were the principal presenting clinical features. None of the patients had any opportunistic infections or endocarditis. In two cases, a ruptured aneurysm was demonstrated radiographically. In the remaining case, sonograms were diagnostic. The organisms responsible were salmonella, Hemophilus influenzae, and Mycobacterium tuberculosis. In two cases, the infectious origin was evidenced by bacteriologic examination of the aortic wall, which revealed the presence of Salmonella enteritidis and Koch's bacillus. Although Hemophilus influenzae was not found in the aortic wall of the remaining case, the infectious origin of the aneurysm was established because preoperative blood cultures were positive for this pathogen, and pathohistologic examination of the specimen showed destruction associated with leukocyte infiltration of the aneurysmal wall. An in situ prosthetic graft replacement protected by omentum was performed in all three cases. Antibiotic therapy was continued for several weeks. All patients are well with follow-up ranging from 10 to 21 months. Infectious aneurysm associated with human immunodeficiency virus seropositivity results in bacterial infestation of an atheromatous aorta. Infected phenomena are promoted by cellular immunodeficiency. Surgery was justified in these cases because of the immediate threat of rupture.

[Isolated proximal revascularization for double aorto-iliac and femoral lesions]. / Revascularisation proximale isolée dans les lésions à double étage aorto-iliaque et fémoral.

From 1980 to 1990, 101 limbs were revascularized at the upper level only in 67 patients, while they presented with associated aortoiliac and femoral obstructive lesions. The symptoms disappeared after aortofemoral revascularization in 94% of the limbs operated on for claudication and 80% of those operated on for critic ischemia. Surgery of the deep femoral artery was associated in 51% of all cases. The average time lapse is 58 months. No complementary revascularization was needed in the cases of claudication. Out of the patients operated on for critic ischemia, upper revascularization was insufficient in 8 cases. Two of the operated patients were cured after secondary downstream revascularization (4%). Three operated patients still presented with intermittent claudication (6%), and 3 were amputed due to acute iliac obstruction seen at an advanced stage. As no reliable predictive test is available, we find it justified to carry out only upper revascularisation in most cases and to decide on the need for secondary downstream extension according to the clinical outcome. However, simultaneous revascularization at both levels is required in case of extensive involvement of the deep femoral artery, such as observed in only 5 of the patients operated during the same period.

Long-term follow-up after prosthetic replacement of the superior vena cava combined with resection of mediastinal-pulmonary malignant tumors.

The contraindication to curative excision of mediastinal and pulmonary cancers because of invasion of the superior vena cava is now challenged by the existence of vascular prostheses that are suitable for venous replacement. Between 1979 and 1990 22 patients underwent resection of lung cancer (n = 6) or malignant mediastinal tumors (n = 16) involving the superior vena cava. Resection was done with concomitant venous reconstruction, and polytetrafluorethylene grafts were used. All bronchogenic carcinomas necessitated right pneumonectomy, whereas the excision of mediastinal tumors had to include pulmonary resections in nine patients (five lobectomies and four sublobar resections) and the right phrenic nerve in 12 patients. Venous reconstruction was performed by interposition of a large polytetrafluoroethylene graft between the proximal and cardiac ends of the superior vena cava (n = 8), or between one (n = 10) or both brachiocephalic veins (n = 4) and the right atrium. One patient died postoperatively (4.5%), and another had mediastinitis that was successfully treated by omentopexy. Chemotherapy was administered preoperatively to five patients and postoperatively to seven patients; radiotherapy was administered to two and 10 patients, respectively. The overall actuarial survival rate is 48% at 5 years, with 11 patients presently alive. The survival rate of patients with mediastinal tumors is 60% at 5 years. Among the patients with lung cancer, two with N1 disease are alive at 16 and 51 months, and one died at 38 months; the two patients with N2 disease died at 6 and 8 months. Only one graft occlusion occurred in the postoperative period; another occurred 14 months after operation and was precipitated by insertion of a central venous catheter. The patency of all remaining grafts was demonstrated after an average time of 23 (1 to 98) months. On the basis of these results, polytetrafluoroethylene graft replacement of the superior vena cava should be part of the planning and execution of radical excision with curative intent of mediastinal and right pulmonary malignant tumors that are not present with other contraindications, such as pleural or distant metastasis and severe systemic disease.

Aortoiliac surgery and kidney transplantation.

Between January 1980 and December 1989, we performed 407 renal transplantations. Twelve of these patients (3%) underwent aortoiliac reconstruction before (Group I, two patients), concomitant to (Group II, five patients) or after (Group III, five patients) renal transplantation. The aortoiliac lesions treated included four aneurysms and seven occlusions of the abdominal aorta and one postarteriography dissection of the iliac artery. A prosthetic graft was inserted in nine cases (75%). Endarterectomy was performed in the three other cases (25%). Four of five patients in Group III were operated on without any particular protection for the transplant. There were no postoperative deaths in Groups I and III. In Group II, one patient died of infection secondary to a urinary tract fistula. Early and late vascular morbidity (renal artery stenosis, occlusion of aortoiliac reconstruction, anastomotic false aneurysm) occurred with equal frequency in the three groups. Renal transplantation in patients having already undergone aortoiliac surgery and, conversely, aortoiliac reconstruction in the renal transplant patient, are possible without any particular technical precautions with minimal mortality and kidney morbidity. Simultaneous renal transplantation and aortoiliac reconstruction carries a significant risk of infection and a two-stage procedure should be considered in this situation.

[Rupture of the diaphragm of late manifestation]. / Les ruptures du diaphragme de révélation tardive.

Twenty one cases of delayed diagnosis of ruptured diaphragm caused by closed trauma are reported: 14 on the left side and 7 on the right side. The clinical signs and the modes of presentation are non-specific. Although the history of thoraco-abdominal trauma and the chest x-ray are sufficient to establish the diagnosis of rupture of the left hemidiaphragm, they can only suggest the diagnosis in cases of righ-sided rupture. The mechanism of rupture is more often due to sudden reflex contraction of the diaphragm against a closed glottis than to excessive abdominal pressure caused by the trauma. Diaphragmatic rupture due to closed trauma causes large tears exposing the patient to a low risk of strangulation of intestinal structures in contrast with ruptures due to a penetrating injury, which causes small tears. The diaphragmatic domes must be systematically explored during laparotomy or thoracotomy performed for thoraco-abdominal trauma.