|
|
||||||||
Ann Thorac Surg 2003;75:610-619
© 2003 The Society of Thoracic Surgeons
a Division of General Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts, USA
b Department of Surgery, Harvard Medical School, Boston, Massachusetts, USA
* Address reprint requests to Dr Grillo, Massachusetts General Hospital, Blake 1570, 55 Fruit St, Boston, MA 02114, USA
e-mail: pguerriero{at}partners.org
| Introduction |
|---|
|
|
|---|
By 1990, resection rates for tracheal tumors reached 63% for squamous carcinoma, 75% for adenoid cystic carcinoma, and 90% for other tumors [6]. The developments that led to such advances in 30 years between 1960 and 1990 deserve to be related in more detail than the page and a half devoted to airway surgery in another recent History of Thoracic Surgery [7].
The following, not a comprehensive review of the literature on tracheal surgery, is a selective account of tracheal surgical development. Emphasis is on beginnings and early development of important concepts and procedures. Current references are not necessarily included, unless they report progress in fundamental aspects of tracheal surgery or significant evolution of techniques. For historical reasons, an authors earlier publication may be cited rather than more complete later reports. Regrettably, omissions from this account are inevitable.
This review is divided into two parts, the first of which traces the evolution of techniques of tracheal surgery. The second part records the acquisition of information about characteristics and treatment of specific diseases of the trachea. There is, of course, considerable overlap. Part 2, "Treatment of Tracheal Diseases," will appear in the next issue of The Annals. References for both parts are numbered sequentially. See "Selected References" at the end of this article for further explanation.
| Techniques of tracheal surgery |
|---|
|
|
|---|
Over the centuries, few reports of successful tracheostomy are found [8]. A drowning victim was treated with tracheostomy by Detharding in 1714. De Garengot in 1720 described the division of the thyroid isthmus to accomplish tracheostomy, using a long vertical incision that extended almost from chin to sternum. He further argued that failure of tracheostomy was often due to its belated performance. Lorenz Heister in 1718 is said to have first used the word "tracheostomy." Martin in 1730 described an inner cannula for the tracheostomy tube, a device suggested to him by a layman. Chovell in 1732 performed tracheostomy at the request of a patient who faced death by hanging. Unfortunately, this tactic did not save the accused. Goodall [8] found reports of 28 successful tracheostomies carried out before 1825, when Bretonneau in Tours used tracheostomy with success in treating "croup." His pupil, Trousseau, applied the technique in 1831 for the management of diphtheria, saving about 25% of 200 children in Paris who were dying from the disease.
Tracheostomy changed little technically, although controversy continued about its indications, locations, and hazards [10]. Chevalier Jackson [13] largely cast the procedure in its modern form, cautioning against high tracheostomy. He believed that "tracheotomy is the worst done of all operations" [14]. Tracheostomy found application in general anesthesia, but was soon displaced by endotracheal intubation [15]. As diphtheria waned tracheostomy was used in poliomyelitis, to prevent infection, in head and chest injuries, after major surgery, and to reduce dead space. The endotracheal tube largely replaced tracheostomy as a preferred method to establish an emergency airway. Later, tracheostomy vied with endotracheal intubation for management of secretions, and subsequently was used as a route for mechanical positive pressure ventilation. A high incidence of complications was recognized even before the frequent appearance of postintubation injuries [9, 16, 17]. Plastic surgical closure of a persistent tracheostomy by a cutaneous inversion technique was described by Lawson and Grillo [18] in 1970.
Repair and healing of the airway
An ancient concern that cast a shadow on tracheal surgery into the 20th century was that cartilage healed poorly. Hippocrates [19] had cautioned that, "The most difficult fistulas are those which occur in the cartilaginous areas. ..." In the second century C.E. Aretaeus pronounced that "The lips of the wound do not coalesce, for they are both cartilaginous and not of a nature to unite" [8]. As late as 1990, Naef [20] repeated that: "Tracheo-bronchial tissue, as compared with the stomach, intestine, or even skin, does not heal well... both the rigidity and the poor blood supply of the cartilaginous structure are definitely major handicaps."
Nonetheless, examples of early attempts and sometimes success in bronchial and tracheal repairs after trauma are recorded. Indeed, the Rigveda, a book of Hindu medicine dating from between 2000 and 1000 B.C.E., noted that the trachea can reunite "when the cervical cartilages are cut across, provided they are not entirely severed" [10]. Ambroise Paré described suture of tracheal lacerations in the mid-1500s in 3 patients, the first from a sword wound and the latter 2 from knife wounds [21]. The first survived despite a concomitant injury to the internal jugular vein. The second suffered division of both trachea and esophagus and died. We do not know the outcome of the third. Brasavola [8] observed recovery after a suicide attempt cut five tracheal rings.
Eventually, cumulative clinical experience in the 20th century established that the trachea healed firmly with suture repair after laceration [22, 23] or rupture [2428]. Jackson and colleagues [29] had demonstrated firm healing of experimental bronchial anastomosis. Daniel and coworkers [30] in 1950 confirmed fibrous tissue repair of linear tracheal incisions in the laboratory, as Rob and Bateman [31] did in 1949 in the clinic. Quinby and Morse [32] in 1911, for the first time experimentally, had highlighted the importance of peribronchial tissue in bronchial closure. Rienhoff and associates [33] in 1942 made fundamental observations that bronchial healing after pneumonectomy was accomplished by new connective tissue that grew over the ends of the stump rather than by mucosal healing alone.
End-to-end tracheal and bronchial anastomosis
Gluck and Zeller [34] in 1881 demonstrated healing after end-to-end tracheal anastomosis in dogs and believed the technique could be applied in humans. Colley [35] in 1895, to avoid stenosis, tried elliptical and bayonet anastomoses in dogs after resecting five rings. Primary anastomosis of the cervical trachea after limited resection for posttraumatic stenosis followed in 1886 by Küster [36], apparently the first in human. Bruns [37] in 1898 performed an extended lateral excision of a papillary tumor in cervical trachea but managed the tracheal defect by packing and using a cannula. Complex methods for repair of cervical tracheal defects, with skin or fascia lata, were also explored in the early 20th century by Nowakowski [38] in 1909, Levit [39] in 1912 and by others. Eiselsberg [40] successfully performed a second tracheal resection of 1.5 cm in 1 patient. Mathey and associates [41] commented in 1966: "This type of radical tracheal surgery was then forgotten for half a century."
The era of open thoracic surgery had arrived. By 1936, Churchill [42] had refined the technique of lobectomy to achieve a 2.6% mortality rate. As interest in bronchial and tracheal surgery grew by the mid-20th century, laboratory experiments confirmed that healing followed end-to-end anastomosis of both bronchi and trachea, although sometimes with stenosis [2931, 4346].
Bronchial repair after trauma proved the feasibility of airway reconstruction. Sanger [47] described bronchial repair in patients during World War II. In 1949, Griffith [48] resected a stricture and anastomosed the bronchus 3 months after rupture. Other late repairs of ruptured bronchi followed [49]. Scannell [50] first performed immediate repair of a bronchus ruptured during closed injury in 1951. Belcher [51] in 1950 and Mathey and Oustrieres [52] in 1951 reported reanastomosis of main bronchi after accidental division during surgery.
Earlier cautious enlargement of bronchial stenosis by wire-supported dermal grafts [53, 54] were replaced by resection and reconstruction. The technique was applied to low-grade tumors [5557] and to carcinoma [58, 59], as sleeve lobectomy evolved. Concurrent vascular sleeve resection was also pursued by Pichlmaier and Spelsberg [60]. Main bronchial resection without removal of lung tissue was extensively described by Newton and associates [61].
Impediments to tracheal reconstruction
With retrospective wisdom, we may ask, "What were the barriers to application of the bronchoplastic and tracheal anastomotic techniques just noted to clinical tracheal resection and reconstruction?" One obstacle was the persistent suspicion that tracheal cartilage healed poorly. A more insistent concern was that only a limited segment of trachea could be removed and reanastomosis accomplished. Nowakowski [38] in 1909 placed the limit of resection at 3 to 4 cm from cadaver studies. Colley [35] and Küster [36] reported resections of three rings and 2 to 4 cm, respectively. Rob and Bateman [31], based on cadaver dissection, placed the limit at 2 cm. Belsey [1] believed that three or four rings, about 2 cm, was the limit in humans. Cantrell and Folse [62] placed the limit at two rings if the patient was older than 80 years. Nicks [63] cited "1 inch or more" as a limit in the cervical trachea.
These presumed limits led to devising complex methods of cervical tracheal reconstruction with available tissue flaps and transfers, and, further, to a century-long search for a means of tracheal replacement. This search ranged over foreign material in many forms, autogenous tissue constructions, tissue and foreign material complexes, fixed or "tanned" tissues, transplantation, and, recently, tissue engineering. Success has eluded investigators to date. The story of this frustrating pursuit and the reasons for overall failure thus far are detailed in a recent review [64].
An additional difficulty of reconstruction was maintenance of safe, continuous, stable ventilation throughout the procedure, especially for intrathoracic tracheal operations. The evolution of anesthetic techniques is discussed later. Finally, primary tumors of the trachea remained rare, as can be seen from earlier chronicles of their occurrence [65, 66]. Stenoses from traumatic, iatrogenic, or inflammatory causes were not seen frequently before 1960. Thus, any single thoracic surgeon was not often challenged. Each case was largely dealt with in an ad hoc fashion.
Primary resection and reanastomosis: initial experiences
In the mid-20th century, a recrudescence of interest in tracheal surgery was marked by experiments in tracheal healing and replacement and by renewed clinical efforts. Earliest attempts at reconstruction of the cervical trachea were still most often by staged, complex repairs, typified by Crafoord and Eindgrens [67] cutaneous reconstruction after tumor removal in 1945. Belsey [1] seems first to have dared to remove intrathoracic tracheal tumors, but his repair was with wire-supported fascia, leaving a residual strip of mucosa for continuity and for epithelial regeneration. Clagett and associates [68] and others followed, using polyethylene tubes or patches to repair the defects. Efforts to replace the trachea partially or circumferentially have been described in detail recently [64].
Despite continued concerns [68] about the feasible length of tracheal resection and lingering doubts about cartilaginous healing, a number of successful resections and reconstructions with primary anastomosis were described in the 1950s and early 1960s, most often for shorter, benign lesions such as stricture. Conley [69] successfully resected the second and third rings for scar in 1944, with end-to-end anastomosis. Kay [70] removed four rings of proximal trachea for leiomyoma without event in 1951. Sweet [71] in 1952 resected a cervical "cylindroma" with end-to-end anastomosis and questioned whether this procedure might be possible intrathoracically. Macmanus and McCormick [72] in 1954 excised a three-ring segment for the same tumor, located about 2 cm above the carina, with end-to-end repair. An anastomotic leak was patched with fascia lata and a protective tracheostomy added. Forster and colleagues [73, 74] in 1957 and 1958 reported a series of three successful cervical and cervico-mediastinal tracheal resections with primary anastomosis of 1.5, 4, and 3 cm for tumor, posttraumatic stenosis, and postintubation stenosis. Other similar resections were reported by Binet and Aboulker [75] and Miscall and colleagues [76]. Flavell [77] in 1959 had successfully corrected a postintubation stricture at the thoracic outlet by resection, but carried out the procedure from a difficult, transthoracic approachan error to be repeated later by other surgeons. Mattes [78] performed a 4-cm transthoracic lower tracheal resection for cylindroma in 1958, wrapping the anastomosis with pleura.
Indicative of revived interest in tracheal surgery were extensive reports in 1960 by Baumann and Forster [79] of worldwide experiences in tracheal surgery. They pointed out that improvements in diagnosis (endoscopy) and technical and ventilatory methods had served to widen the field beyond tracheostomy and endoscopic treatment alone. Simultaneously, the potential for surgery of the thoracic trachea was exciting interest.
Anatomic mobilization of trachea
These midcentury experiences in tracheal reconstruction, chiefly in the upper trachea and most often of limited extent, clarified that the basic techniques of tracheal anastomosis could achieve sound healing. The "2-cm rule," which had served to inhibit advances in tracheal surgery, was now challenged by experimental studies reinvestigating the extent of trachea that could be removed and approximation achieved by anatomic tracheal mobilization, without use of prosthetic replacement. Clinical experiences, especially with intrathoracic and carinal lesions, contributed to widening the possibilities for more extended resection.
Ferguson and associates [44] determined the extensibility of human trachea from cadavers to be 35% at 29 years and 17% at 76 years, with the most stretch reached with 200 grams (g) of tension. In living dogs, the majority of resectable length was obtained at 450 g of tension at the anastomosis, about 30% to 35% of tracheal length. Michelson and colleagues [80], in an effort to increase the length of resectable trachea, freed the right main bronchus in dogs by incising the right pulmonary ligament and dividing the left main bronchus at carina, then reanastomosing it to the bronchus intermedius. This strategy permitted resection of 12 rings in the dog. They found that the human trachea could be stretched 4 to 6 cm by mobilization and that an additional 2.5 to 5 cm could be obtained by the maneuver described in dogs. Tracheal elongation in fresh human cadavers with the same dissection and 450 g of upward pull, allowed 2.5 to 3 cm elevation after division of the pulmonary ligament and 5 to 6 cm after freeing the left main bronchus in 4 cadaver subjects younger than 50 years of age. One to 1.5 and 2 to 3 cm, respectively, were measured in 4 subjects aged 50 to 75 years [44]. Cantrell and Folse [62] sought to determine the limits of feasibility of primary anastomosis during the repair of circumferential defects. In resection of 20% to 58% of dog trachea, suture line tension ranged from 400 to 2,750 g. Tension required for anastomosis varied markedly between flexed and extended neck positions. Disruption of anastomosis occurred between 1,700 and 3,100 g, at resection lengths of 46% to 63% of the trachea. However, they noted in human tracheas obtained at autopsy that resection of more than 2 cm in subjects older than 80 years produced unacceptable anastomotic tension, based on experimentally derived standards.
Harris [81] in 1959 showed radiologically that neck extension elongated the trachea by 2.6 cm. Som and Klein [82] extended the length of human cadaver trachea by only 1.6 cm by circumferential incision of the intercartilaginous annular ligaments.
Grillo and colleagues [83] reported in 1964, from autopsy studies in humans, that more than half of the adult trachea could be resected and continuity reestablished by full mobilization of limiting structures. The steps to mobilization were as follows: (1) right hilar dissection and division of the right pulmonary ligament; (2) division of the left main bronchus; and (3) freeing pulmonary vessels from the pericardium. With the subjects neck in a neutral position, these steps permitted tracheal excisions averaging 3 cm (three to eight rings), 2.7 cm (three to 12 rings), and 0.9 cm (0.5 to three rings), for a total of 6.4 cm (11 to 18 rings). Anastomotic tension rose exponentially with resection of successive 1-cm segments from 25 g at 1 cm to 675 g at 7 cm. Age did not prove to be seriously limiting. These tensions were considerably below the biologically dangerous limit of 1,700 g determined by Cantrell and Folse [62]. Division of the left main bronchus allowed the advancement of the distal tracheal stump and right main bronchus.
In addition, if an even more extended resection were necessary, division of the cervical trachea two to three rings below the cricoid allowed this segment of cervical trachea to be devolved into the mediastinum with intact lateral vascular supply [83]. This maneuver proposed to allow reconstruction of intrathoracic trachea by simple anastomosis while later permitting staged reconstruction of cervical trachea, which would be safer. Baumann and Forster [79] had made a similar suggestion previously. Because of the complexity of this last approach, division and reimplantation of the left main bronchus was later applied clinically only in the case of carinal reconstruction and then rarely.
Stimulated by Grillos clinical experiences with cervical tracheal resection for postintubation stenosis, in 1968 Mulliken and Grillo [84] investigated the amount of trachea that might be resected by cervical and mediastinal mobilization and still permit anastomosis, leaving the pleural cavity intact. They performed pretracheal mobilization down to the carina with division of the thyroid isthmus in cadavers. With the neck in 15 to 35 degrees of flexion, 1,000 to 1,200 g of tension applied to the divided tracheal ends permitted an average resection and reapproximation of 4.5 cm (7.2 rings). Right hilar mobilization with the pleura open allowed an increment of resection of 1.4 cm (2.5 rings), a total of 5.9 cm. Average tracheal length was 11 cm. Cervical flexion permitted a gain of 1.3 cm (2.5 rings) over the neutral position. Thus, cervical flexion and pretracheal mobilization alone appeared to allow significant cervical or cervicomediastinal resection and anastomosis, especially important for the postintubation lesions that were increasing in frequency and often occurred in patients who could not tolerate thoracotomy.
Appreciation of the possible degree of tracheal mobilization based on anatomic principlespretracheal mobilization, cervical flexion, hilar dissection, including intrapericardial freeing, and mobility of detached main bronchimade possible a systematic and aggressive approach to tracheal resection and reconstruction not previously conceived. The episodic ad hoc approach that produced single case reports, at times almost expressing a surgeons surprise at the results, yielded to more confident and planned approaches. Using such principles, surgeons reported their successes with series of resections and reconstructions of cervical and thoracic trachea for stenosis and tumor [8595].
Laryngeal release
An additional dividend for extended upper tracheal resection came from otolaryngology. Ogura and colleagues [96, 97] had suggested dividing hyoid muscles to help close the gap produced by resection of a subglottic laryngeal stenosis. In 1969, Dedo and Fishman [98] offered thyrohyoid laryngeal release as a necessary adjunct to tracheal resection for stenosis. Division of the thyrohyoid muscles, superior cornua of the thyroid cartilage, and the thyrohyoid membrane, with care to preserve superior laryngeal nerves, allowed the larynx to drop about 2.5 cm. Montgomery [99] described an alternative method for laryngeal releaseknown as suprahyoid release. Muscle attachments to the superior surface of the hyoid bone, the stylohyoid muscles, and the hyoid bone anterior to the digastric slings were all divided allowing the larynx to drop. It was the opinion of both Grillo and Pearson (unpublished) that less severe deglutitional disorders of shorter duration more often followed suprahyoid release than thyrohyoid release. Release is not routinely necessary for upper tracheal resection [100]. Grillo [101] observed clinically that laryngeal release did not transfer effective relaxation for lower tracheal or carinal resections. Valesky and associates [102] confirmed this observation through cadaver studies.
Tracheal blood supply
The detailed arterial supply of the trachea was described as a necessary corollary to tracheal surgery. Grillo [86] emphasized the entry of small segmental arteries through "lateral pedicles" of tissue attached to either side of the trachea. Miura and Grillo [103] precisely defined the blood supply to the upper trachea in 1966, usually from three principal branches of the inferior thyroid artery, with the firstor lowestbranch most often predominant. Salassa and coworkers [104] completed a definitive study of tracheal blood supply in 1977, confirming Miura and Grillos description of the cervical tracheal supply, and mapping the arterial supply of the thoracic trachea from bronchial, supreme intercostal, subclavian, right internal thoracic, and innominate arteries. Segmental tracheoesophageal arteries connected often to lateral longitudinal arteries and then to transverse intercartilaginous arteries. Three to seven tracheal arteries were found in the "lateral pedicles." Unlike the intramural collateral of tracheal blood supply in the dog [62], which maintain tracheal viability despite complete circumferential dissection and subsequent division and anastomosis, the same procedure in humans has led to necrosis.
Carinal resection and reconstruction
At the lower end of the trachea loomed the special problems of carinal reconstructionanatomic, technical, and anesthesiologic. Lesions, most often neoplastic, were centered in the carina, extended to the carina from low in the trachea, or to the carina from main bronchi or lungs. Experimentally, Grindlay and colleagues [105] in 1949 resected right lung and carina in dogs, with end-to-end anastomosis of the trachea to the left main bronchus. Ferguson and coworkers [44] in 1950 also performed right and left pneumonectomies in dogs with resection of the carina and end-to-end anastomosis. Juvenelle and Citret [106] in 1951, working at the University of Buffalo, showed experimentally the feasibility of lateral implantation of the bronchus into the trachea, without loss of blood supply and without interference in ventilation. They further described experiments in which they resected carina with a three- to four-ring segment of trachea, then anastomosed the trachea to the right or left main bronchus and implanted the other main bronchus into the side of the trachea. Freeing the trachea was necessary to reduce otherwise excessive tension. Additionally, they remarked [107] that freeing the trachea permitted anastomosis of trachea directly to the right and left main bronchi without excessive tension, after short segment resection. Meyer and associates [108] experimentally implanted the right upper lobe and right main bronchus into the trachea in 1951. Ehrlich and colleagues [109] in 1952 transposed right main bronchus to lateral tracheal wall and later resected the left lung and carina in dogs. Kiriluk and Merendino [46] in 1953 described a variety of experimental tracheal, bronchial, and carinal reconstructions, including reapproximation of both main bronchi to the carina and tracheobronchial anastomosis after carinal pneumonectomy. Nicks [63] similarly reconstructed the carina after resection in pigs in 1956, but under hypothermia. In 1958, Björk and Rodriguez [110] described experiments in reconstruction by direct anastomosis after resection of carina and 12 tracheal rings in dogs. The right main bronchus was sutured end-to-end to trachea and left main end-to-side to the intermediate bronchus. This attempt followed a similar successful clinical procedure carried out by Barclay and colleagues [111] as described below. The same anastomoses after carinal resection were performed in dogs in 1969 in confirmatory studies [112].
Clinically, Abbott [113] in 1950 repaired large oval defects created by right pneumonectomy and right carinal lateral excision for bronchogenic carcinoma in 5 patients, by transverse closure; 2 died. Other patching techniques were used to repair such lateral defects, including dermal grafts, synthetic materials, and patches or flaps of retained bronchial wall [114, 115]. These complex and frequently unsuccessful patch techniques have been reviewed previously [64].
In 1951, Mathey [116] locally resected a "cylindroma" of the back wall of the trachea at the carina, including posterior walls of both proximal and main bronchi. The repair was effected by longitudinal suture of the medial bronchial margins and transverse suture of the remaining defect. Surgeons struggled with the problem of tracheobronchial anastomosis at the carina. In 1954, Crafoord and associates [117] reported anastomosis of the bronchus intermedius to the trachea at the site of main bronchial origin after upper lobectomy and bronchial excision. The next year, Björk [118] obtained access to the carina from the left chest, mobilizing the aorta after division of four pairs of intercostal arteries, to successfully resect left main bronchus and anastomose its lobar bifurcation to the prior origin of the bronchus at the trachea. In 1959, he presented follow-up of 16 patients who had undergone bronchotracheal anastomosis, with four stenoses [119]. Abbey-Smith and Nigan [120] later described a similar left-sided approach in 1979 for amputation of the left main bronchus at the carina for pneumonectomy in a case of proximal lung tumor.
Barclay and coworkers [111] in 1957 resected about 5 cm of trachea and carina to remove recurrent adenoid cystic carcinoma. Division of the pulmonary ligament allowed anastomosis of the trachea to the right main bronchus. The left main bronchus was anastomosed end-to-side to the bronchus intermedius. Intermittent ventilation sufficed for the second anastomosis. A second patient was handled identically; both recovered. The authors reported that dissection in fresh cadavers before operation permitted resection of 6 cm of trachea, using this technique. They also proposed, where carinal resection was not required, to close the left main bronchial stump. Eschapasse and colleagues [121] used this technique in 1961. Archer and associates [122] excised a granular cell myoblastoma at the carina similarly in 1963. The procedure was a major step in carinal surgery. Grillo and colleagues [123] in 1963 described resection of carina and trachea for a length of 4 cm to remove an adenoid cystic carcinoma. The right hilum was mobilized and the trachea anastomosed to the right main bronchus. The left main bronchus reached the trachea easily enough to be anastomosed there end-to-side. The patient did well. Cross-field intubation and ventilation were used. Temporary occlusion of the pulmonary artery to the nonventilated lung eliminated shunting, but later rarely proved to be necessary. In 1966, Mathey and associates [41] reported results in 7 patients who underwent carinal excision with or without bronchial resection, using thoracotomy. However, they believed that sternotomy might be preferable. Three patients had pneumonectomy and 2 had partial lung resection. The following anastomoses were done: trachea to main bronchus; side-to-side left main and intermediate bronchus and both end-to-end to trachea; 2 had dermal graft patches. Two patients died after the operation. Eschapasse and colleagues [124] in 1967 cited 3 patients who had circumferential resection of the entire carina with primary reconstruction. Anastomoses were right main bronchus to trachea with left main end-to-side to bronchus intermedius in 2 patients and left main to trachea with intermediate bronchus to left main. One patient died after the operation. Eschapasse favored right thoracotomy, cross-table ventilation, avoidance of prostheses, and primary reconstruction. Anesthesia for carinal resection, which had initially seemed formidable, was easily enough managed in patients by cross-table ventilation of trachea and bronchi as the procedure progressed, so that ventilation was not interrupted or uncontrolled at any point.
Nonetheless, the anesthetic challenge of carinal resection suggested the use of extracorporeal circulation to some. Nissen [125] removed a "malignant adenoma" thus. Under bypass, Woods and associates [126] excised recurrent "cylindroma" from the carina with a very limited margin. Reconstruction was by suture and a patch of skin supported by wire mesh. Adkins and Izawa [127] performed lateral resection of the carinal wall for cylindroma, patching the defect with Marlex (Chevron Phillips Chemical Company LP, Houston, TX) and mediastinal tissues. As might be expected, granulation tissue formed at the patch site. The considerable potential for hemorrhage because of the need for heparinization during bypass was not encountered in these technically limited cases.
Experience with carinal resection and reconstruction grew slowly. In 1974, Eschapasse [128] collected 19 cases from several French teams. Perelman and Koroleva [129] in 1980 recorded 29 carinal resections with reconstruction. By 1982, Grillo [101] had performed 36 carinal reconstructionsfor 23 primary tracheal neoplasms, five bronchogenic carcinomas, and eight inflammatory lesions. Eleven of these reconstructions were accomplsihed without loss of lung tissue. Based on this experience, Grillo [101] presented a comprehensive schema for carinal reconstruction. For short resections, carinal restoration was by side-to-side main bronchial anastomosis, which was then joined end-to-end to the trachea; for longer lesions, the trachea was placed end-to-end to the left main bronchus (if the gap was less than 4 cm) and the right main end-to-side to the trachea; for still more extensive tracheocarinal removal, the "Barclay" anastomosis of the right main bronchus to the trachea and left main end-to-side to the intermediate bronchus was used. Other special problems were also presented, including the difficulty of lesions involving a long length of trachea and also of left main bronchus. Recent exploration of problems with carinal reconstruction has been reported for 143 resections [130].
Approach for carinal resection by right thoracotomy has been preferred by most surgeons [101, 123, 128, 131, 132]. Left thoracotomy with subaortic dissection was used for specific lesions, principally those involving the left main bronchus and the carina but little of the tracheal length [41, 101, 128, 133, 134]. Left thoracotomy with retroaortic dissection was also explored early [118, 120] but failed to gain acceptance. Median sternotomy for carinal access was described by Goeltz [135] in 1907 for foreign body removal, by Padhi and Lynn [136] in 1960 for bronchopleural fistula, by Abruzzini [137] in 1961 for treatment of postpneumonectomy tuberculous fistulas, and was reintroduced with anterior and posterior pericardial opening by Perelman in 1976 [132]. Pearson and coworkers [138] favored this approach for carinal resection. Maeda and associates [134] added left anterior thoracotomy to sternotomy to improve access. Grillo [101] used bilateral thoracotomy ("clamshell" incision) for free access to carina and to both thoraces for treatment of complex lesions, especially those involving the left main bronchus, carina, and a long extent of the lower trachea.
Carinal pneumonectomy
Early on, surgeons conceived of extension of pneumonectomy for bronchogenic carcinoma to include tumor that also involved the carina [113, 115, 139]. Carinal pneumonectomy with anastomosis of terminal end of the trachea to left main bronchus was reported in 1958 by Mattes [78] and in 1959 by Gibbon [140]. Hardin and Fitzpatrick [141] in 1959 excised carina for bronchogenic carcinoma, reconstructing the carina by direct suture with the aid of free cartilage graft, using ventilatory anesthesia delivered by a tube placed in a distal left main bronchial aperture. The graft was taken from an uninvolved portion of the right bronchus. MacHale [142] repeated this procedure in 1972. Thompson [143] in 1966 described anastomosis of the left main bronchus to a tailored trachea after right pneumonectomy for squamous cell carcinoma that included a sleeve of carina. Also in 1966, Mathey and associates [41] expressed preference for circumferential carinal resection over lateral resection and patching. Grillo [101] included carinal pneumonectomy with circumferential resection in the spectrum of techniques of carinal resection and reconstruction.
Carinal pneumonectomy for bronchogenic carcinoma became further established with significant series reported by Jensik [144] in 1972, Ishihara [145] in 1977, Deslauriers [146] in 1979, Dartevelle [147] in 1988, Tsuchiya [148] in 1990 and their colleagues, and Mathisen and Grillo [149] in 1991. The initially high operative mortality of nearly 30% proved to be primarily a result of a form of adult respiratory distress syndrome labeled postpneumonectomy pulmonary edema, of noncardiogenic origin. Mathisen and colleagues [150] found this condition responded favorably to prompt treatment with nitric oxide. Believed to be the result of barotrauma, this dread complication has become less frequent with close attention to ventilatory volumes and pressures; mortality has decreased to about 10% or less [147, 151, 152].
Anesthesia for tracheal surgery
McClish and colleagues [153] noted that concern about anesthesia for major airway reconstruction "stems from the complexity of simultaneously controlling the airways, maintaining satisfactory gas exchange, and ensuring good surgical exposure to the trachea." The technique of ventilation across the operative field with direct insertion of endotracheal tubes into the trachea and bronchi during phases of surgery when the airway is open, developed early and is described with variations in reports of tracheal reconstructions cited earlier. Tubes across the operative field were used in experimental work [43, 105, 106] and Ravitch [154] mentioned clinical usage back in 1951. Friedmann and Emma [155] in 1951 described a catheter insufflation technique for carinal resection in 1 patient. Grigor and Shaw [156] working with Barclay and his colleagues used cross-field ventilation in combination with endotracheal intubation, depending on intermittent ventilation during the implantation of the left main bronchus into the bronchus intermedius. They recognized that the preceding development of one-lung anesthesia provided the groundwork for carinal anesthesia. Baumann and Forster [157] in 1960 described systematic approaches to anesthesia for tracheal surgery (cervical, intrathoracic, and carinal), including intubation through distal tracheostomy and also across the operative field. Grillo and associates [123] in 1963 detailed a similar technique for carinal resection and reconstruction. Cross-table anesthetic technique was fully described by Grillo [85] in 1965 and expanded in 1970 [131]. The potential use of two anesthesia machines for complex carinal reconstruction was also noted. Geffin and colleagues [158] summarized cross-table anesthetic techniques for tracheal and carinal reconstruction in 1969 based on experience with 31 operations. Theman and colleagues [159] confirmed these techniques.
Lee and English [160] in 1974 described the use of a Saunders-type bronchoscopic injector through a catheter passed beyond a tracheal stenosis. El Baz and colleagues [161] favored high-frequency positive pressure ventilation for tracheoplasty to permit better visualization and access. McClish and associates [153] expressed similar convictions in a series of 18 patients. Wilson [162] thoroughly updated these anesthesiologic approaches. Although Wilson preferred cross-table ventilation, high-frequency ventilation was valued as a useful adjunctive technique in special circumstances, such as complex carinal reconstruction. These choices remained a matter of preference, with both techniques proving satisfactory in experienced hands. All agree that close communication and cooperation between surgeon and anesthesiologist are uniquely demanded for this type of surgery, preoperatively, intraoperatively, and, optimally, postoperatively.
The use of cardiopulmonary bypass for tracheal and especially carinal resection has been mentioned earlier. Through their extensive experience with tracheal surgery, Eschapasse and associates, Grillo, Pearson, and Perelman found bypass to be unnecessary. The use of bypass in complicated cases, in which the procedure might have been theoretically desirable, but in which extensive dissection and manipulation of the lung was required, led to fatal pulmonary parenchymal hemorrhage in 2 patients due to necessary anticoagulation (Grillo, Pearson, unpublished).
Laryngotracheal resection and reconstruction
Just as reconstruction of the carina presented unusual difficulties, so did the proximal end of the airway. When tracheal lesions also affect the subglottic larynx, the anatomic and functional characteristics of that structure offer special problems. Many otolaryngologic procedures were developed to manage inflammatory stenosis at this level, when conservative measures failed. "Conservative measures" included dilation, stents, intubation, steroid injection, cryotherapy, and laser surgery. Surgical procedures that were devised included anterior and posterior cricoid splits, placement of stents, mucosal and cutaneous grafts, free grafts of cartilage and hyoid, pedicled hyoid, cutaneous flaps variously supported with cartilage, and multistage "trough" procedures. These many operations will not be reviewed here [163], but, in general, success was limited.
A one-stage approach to subglottic stenosis characterized by cricoid involvement developed slowly. Initial work was carried out by otolaryngologists, but full development of techniques was by thoracic surgeons who faced the problem of subglottic stenosis as presented in the spectrum of postintubation tracheal stenosis. Conley [69] removed the entire cricoid in 1944 for chondroma, preserving the mucoperichondrium, which was held in place by a foam rubber stent. Great care was taken to avoid injury to the recurrent laryngeal nerves. Shaw and associates [28] resected damaged or stenotic cricoid in 2 patients with anastomosis to the thyroid cartilage. Existing vocal cord paralysis simplified the problem in these patients. Ogura and Roper [96] apposed the second tracheal ring to thyroid cartilage after subtotal excision of traumatically scarred and stenotic cricoid in 2 patients. The recurrent nerves were paralyzed, arytenoidectomy was done, and a stent was used postoperatively. The distal trachea was mobilized and the thyrohyoid muscles and constrictors that are attached to the thyroid cartilage were divided to assist in approximation. Subperichondrial cricoid resection avoided injury to recurrent nerves [97]. Six of 7 patients with chronic subglottic stenosis were helped.
Gerwat and Bryce [164] in 1974 placed the upper line of resection for stenosis at the lower border of the thyroid cartilage anteriorly and through the posterior cricoid lamina below the cricothyroid joints posteriorly. Thyrohyoid release was added and believed to be important. Four patients were so treated. In 1975, Pearson and colleagues [165] followed the same line of cricoid resection, but rongeured all but a thin shell of posterior lower cricoid plate, sutured the ends of the first intact cartilaginous ring of trachea together, and inset this narrowed piece into the rongeured groove to form the laryngotracheal anastomosis. Recurrent nerves were preserved. Superior laryngeal release was done, and a splinting T-tube was added postoperatively. Six patients were treated successfully. Couraud and associates [166] in 1979 added 4 patients; all but one operation was successful. They indicated that disturbing the recurrent nerves was unnecessary, that sometimes the posterior cricoid cartilage did not need to be tailored, and that tracheostomy was not regularly necessary. Grillo [163] in 1982 described 18 patients with subglottic stenosis treated with a somewhat modified procedure. In patients with anterolateral stenosis, simple beveled cricoid resection was sufficient, and the tracheal cartilage to be anastomosed was tailored obliquely to fit easily. For circumferential stenosis, scar tissue over the posterior cricoid plate was excised and the raw area resurfaced with a broad-based flap of posterior membranous tracheal wall shaped for this purpose. Neither laryngeal release nor tracheostomy was routinely needed.
Grillo and colleagues [167] in 1992 reviewed the cases of 80 patients who underwent one-stage laryngotracheal resection and reconstruction for subglottic stenosis by these techniques50 from postintubation lesions, 7 from trauma, 19 idiopathic, and 4 others. Thirty-one patients required circumferential resection with posterior flap resurfacing. Two resections failed. Glottic correction, if needed, was done initially as a separate procedure. Maddaus, with Pearsons group [168], proposed synchronous glottic reconstruction where that was also required, reporting 15 such cases out of 53 subglottic repairs. They also adopted the posterior tracheal membranous wall flap described by Grillo and colleagues [163, 167]. Monnier and colleagues [169] proved this type of repair to be useful in children.
Cervicomediastinal exenteration and mediastinal tracheostomy
Rarely, after extensive resection of larynx and upper trachea for neoplasms, such as thyroid carcinoma, adenoid cystic carcinoma, or recurrent laryngeal carcinoma after laryngectomy, mediastinal tracheostomy will be necessary, well below the sternal notch. Watson [170] in 1942 devised a procedure to treat a patient with squamous carcinoma 4 cm above the carina. The patient had undergone laryngectomy for cancer 15 years earlier followed by radium treatment. A "V" of sternum was resected and skin flaps mobilized to allow closure of the margins of the tracheal stoma. Sloan and Cowley [171] in 1951 met the problem of tracheal compression by an aortic aneurysm by establishing a side tracheostomy, the tube of which emerged from the back medial to the right upper scapula, after removal of proximal rib segments. After wrapping the aneurysm, the tube was removed. The authors discussed earlier proposals and even attempts to establish transpleural bronchial fistulas for this purpose and a proposal, not acted upon, by Gluck in 1907, to perform posterior bronchotomy.
Kleitsch [172] in 1952 removed upper sternum and inserted a polythene tube. A sequence of irradiation necrosis and recurrent tumor frustrated plans to line the opening with skin grafts. In the same year, Minor [173], after removal of recurrent carcinoma of the tracheal stoma, brought skin flaps as a tube through a sternal opening to connect with trachea. Healing failed, and the patient bled to death 4 months later. Waddell and Cannon [174] in 1959 pulled a short tracheal stump to the right of the ascending aorta and created a skin tube from crossed anterior chest skin flaps that passed through a hole rongeured in the sternum and were anastomosed to the tracheal end. Two of 4 patients, all with squamous cell carcinoma, died of massive hemorrhage.
In 1962, Sisson and associates [175], operating for recurrent laryngeal carcinoma at the tracheal stoma, excised a large portion of surrounding skin with the specimen and removed the manubrium and the heads of the clavicles. Skin flaps were turned up to effect closure about the stoma, and an inferior defect was skin grafted. After 2 patients died after the operation from innominate artery hemorrhage, the pectoralis muscles were undermined and rotated between innominate and left carotid arteries and the trachea. Also in 1962, Ellis and colleagues [176] used a tube of heavy Marlex mesh to reach the surface after low transsection of the trachea. However, granulation tissue formation and the possibility of infection, erosion, and hemorrhage makes tubes of foreign material undesirable in this setting. In an effort to eliminate tension at the tracheal cutaneous anastomosis, which carried the threat of subsequent nonhealing and fatal innominate hemorrhage, Grillo [177] in 1966 proposed fashioning a broad full-thickness bipedicled flap of anterior chest wall skin and subcutaneous tissue formed with two long, horizontal incisions. This flap was depressed to meet the end of the trachea in the mediastinum, made accessible by resection of manubrium, heads of clavicles, and upper two costal cartilages. The stoma emerged in midflap, resulting in a simple suture line more likely to heal well. Two end stomas and one in-continuity stoma were reported. Grillo and Mathisen [178] subsequently offered further guard against vessel erosion in the event of deficits in peristomal healing by (1) advancing omentum routinely to separate trachea and great vessels, and (2) electively dividing the brachiocephalic artery under electroencephalographic monitoring, where the tracheal stump was very short, after preoperative angiography. One operative death occurred in 18 patients. Additional experiences have been recorded in this area by numerous authors [179182]. Withers and associates [183] suggested use of a pectoralis musculocutaneous flap, which has particular application to cases in which a wide resection is necessary around an existing stoma because of peristomal carcinoma or irradiation damage.
Complications of tracheal surgery
As tracheal surgery became more common, inevitably a pattern of complications appeared. Complications were analyzed by Levasseur and colleagues [94] in 1971, Couraud and associates [184] in 1982, and by Grillo and associates [185] in 1986. In 11 patients, Levasseur observed one restenosis and four lethal erosions of brachiocephalic artery. These findings led the authors to recommend cervical muscle or thymic interposition for prophylaxis against hemorrhage. Couraud and colleagues [184], reporting on 122 cases of resection with only four deaths and one failure, emphasized antiinfectious and antiinflammatory preparation. Grillo and associates [185] reviewed incidence, causes, treatment, and prevention of complications in 416 primary reconstructions, 279 for postintubation lesions. Suture line granulations occurred in 28 of 209 when various nonabsorbable sutures were used for anastomosis, but in none of 113 after the adoption of absorbable Vicryl (Ethicon Inc, Somerville, NJ) sutures. Brachiocephalic artery injuries were best avoided by avoiding any dissection of the artery, and where dissection was inevitable, interposing a strap muscle. From the first to the second halves of the series, deaths fell from four to one, failures from 13 to seven, and complications from 42 to 30.
The unreconstructible trachea: T-tubes and stents
T-tubes had been devised and variously used in the past, as by Bond [186] in 1891 and Falbe-Hansen [187] in 1955, citing Laurens earlier use of a T-tube. Aboulker and associates [188] treated postintubation injuries with a T tube in 1960. The silicone rubber T-tube developed by Montgomery [189] in 1965 proved widely useful in tracheal surgery, although the device was developed initially in the false hope that prolonged stenting would resolve tracheal stenosis. Cooper and associates [190] in 1981 and Gaissert and associates [191] in 1994 used the T-tube for permanent and temporary restoration of airway continuity when the trachea was not reconstructible, a lesion was not removable, or a temporary airway was needed. Westaby and associates [192] added a bifurcated T-tube for help with carinal problems.
The development and deployment of stents will not be reviewed here. However, caution needs to be raised against tendencies to use essentially permanent expandable stents where lesions might otherwise be readily and definitively corrected by surgery. The result too often is doubly negative: correction of the lesion is permanently prevented and severe complications may develop from the stent [193]. Removable silicone stents also hinder curative treatment and may readily cause granulations, especially in the subglottic region. Granulations are sometimes reversible, however, in contrast to problems caused by permanent stents.
Part 2 of this review, "Treatment of Tracheal Diseases," will appear in the next issue of The Annals.
Selected References (Part 1)[194214], [215235], [236256], [257277], [278298], [299325]*
| Footnotes |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. M. Abdelkafy, M. N. El Atriby, N. M. Iskandar, D. E. Mattox, and K. A. Mansour Slide Tracheoplasty Applied to Acquired Subglottic and Upper Tracheal Stenosis: An Experimental Study in a Canine Model Arch Otolaryngol Head Neck Surg, April 1, 2007; 133(4): 327 - 330. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Seguin, E. Martinod, M. Kambouchner, G. O. Campo, P. Dhote, P. Bruneval, J. F. Azorin, and A. Carpentier Carinal Replacement With an Aortic Allograft Ann. Thorac. Surg., March 1, 2006; 81(3): 1068 - 1074. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Walles, B. Giere, P. Macchiarini, and H. Mertsching Expansion of chondrocytes in a three-dimensional matrix for tracheal tissue engineering Ann. Thorac. Surg., August 1, 2004; 78(2): 444 - 448. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Ernst, D. Feller-Kopman, H. D. Becker, and A. C. Mehta Central Airway Obstruction Am. J. Respir. Crit. Care Med., June 15, 2004; 169(12): 1278 - 1297. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. S. Costache, C. Renaud, L. Brouchet, T. Toma, F. Le Balle, J. Berjaud, and M. Dahan Complete tracheal rupture after a failed suicide attempt Ann. Thorac. Surg., April 1, 2004; 77(4): 1422 - 1423. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.S. Pramesh, R. C. Mistry, R. K. Deshpande, and S. Sharma Is compromising on surgical margins in tracheal resection for cancer justified? Eur. J. Cardiothorac. Surg., June 1, 2003; 23(6): 1077 - 1078. [Full Text] [PDF] |
||||
![]() |
K. Hazama, S. Miyoshi, and H. Matsuda Reply to Pramesh et al. Eur. J. Cardiothorac. Surg., June 1, 2003; 23(6): 1078 - 1078. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |