Iatrogenic Ruptures of the Tracheobronchial Tree*

Transcription

1 Iatrogenic Ruptures of the Tracheobronchial Tree* Herbert Kaloud, MD; Freya-Maria Smolle-Juettner, MD; Gerhard Prause, MD; and Werner Franz List, MD We did a retrospective study in 12 patients with iatrogenic tracheal or tracheobronchial ruptures treated since Ten female subjects, one male subject, and one child (age range, 8 to 72 years), all of whom had undergone intratracheal intubation, were admitted to the hospital. Four patients had been intubated with a double-lumen catheter (two Carlens type with carinal spur, two Robertshaw without spur), and seven had had "high volume-low pressure" tubes, placed under emergency conditions in three of those seven cases. In one further case, an unsuccessful attempt of percutaneous tracheostomy had been made. The localization of the ruptures (all of them longitudinally in the membranaceous wall; length, 2 to 13 em; mean, 7 em) comprised both cervical and intrathoracic trachea in seven, the intrathoracic trachea in three instances, and the left main stem bronchus in two cases. Ten patients had mediastinal and subcutaneous emphysema, seven presented with a, and nine had intratracheal bleeding. The interval until the onset of symptoms and diagnoses differed widely: twice diagnoses were made intraoperatively, during thoracic surgery. The longest interval until diagnosis was 5 days; only then did the patient show subcutaneous emphysema and have retrosternal pain. All patients had surgical repair. Nine recovered without sequelae, and three died of septic multiorgan failure. (CHEST 1997; 112:774-78) Key words: iatrogenic tracheobronchial rupture; intubation; mucosal tear; surgical repair; upper airways trauma problems during endotracheal intubation may cause iatrogenic trauma of the upper airways. Depending on their extent, iatrogenic injuries to the larynx or to the trachea are graded into more and less severe ones. 1 Superficial mucosal tears in the mouth, pharynx, or larynx are frequently found. In a large series of postmortem examinations following emergency endotracheal intubations, they were present in 18% of cases. 2 Transmural trauma of the membranaceous part at the level of larynx or tracheobronchial tree, such as tracheal or tracheobronchial disruption, however, is extremely rare. The occurrence of such injuries following intubation is infrequently reported; severe, transmural trauma after percutaneous tracheostomy has not been reported so far (to our knowledge). Tracheobronchial disruption is life threatening to the patient: tension and mediastinal emphysema, as well as a prolapse of the esophageal *From the Departments of Anaesthesiology (Drs. Kaloud, Prause, and List), and Thoracic and Hyperbaric Surgery (Dr. Smolle-Juettner), University Medical School of Graz, Austria. Manuscript received May 30, 1996; revision accepted February 13, Reprint requests: H. Kaloud, MD, or Werner Franz List, MD, Dept of Anaesthesiology, University of Graz Medical School, Auenbruggerplatz 29, A-8036 Graz, Austria 774 wall into the tracheal lumen, can cause acute asphyxia, whereas mediastinitis and sepsis can occur in the later course. We present a series of 12 patients with various types of tracheobronchial iatrogenic trauma treated since MATERIALS AND METHODS Twelve patients (10 female, 1 male, 1 child; mean age, 43.6 years; range, 8 to 72 years ) were operated on for iatrogenic tracheobronchial disruption (Table 1). Eight patients had sustained the trauma during routine anesthesia for elective surgery. In three cases, intubation had been done under emergency conditions and in one patient, an unsuccessful attempt of percutaneous tracheostomy had been made. In seven patients, the intubation had been done in peripheral hospitals. We evaluated predisposing factors such as duration of intubation, type and size of the endotracheal tube, unusual findings during intubation, as well as the time interval until the onset of symptoms, localization and length of the rupture, therapy, and outcome. RESULTS Female patients (ten) were more common than male patients (two), one of whom was an 8-yearold boy (Table 2).

2 Table!-Patients Characteristics, Conditions of Intubation, and Types of Tube Duration of No.I Age, Intubation, Peculiarities During yr/sex Conditions of Intubation/Type of Surgery min 1/31/F Percutaneous tracheostomy for long-term 45 ventilatory support 2/61/F Lobectomy 180 3/8/M Tonsillectomy 20 4/34/F Laparoscopy 20 5/50/F Suture of the talofibular ligament 25 6/43/F Esophagectomy (over right-sided 180 thoracotomy) 7/61/M Esophagectomy (right thoracotomy), poor 300 general condition 8/45/F Heller's myotomy (left thoracotomy) 90 9/57/F Felon of the index finger 15 10/72/F Emergency flexible bronchoscopy with 10 guided intubation of the left main bronchus 11122/F Emergency intubation (LeFort Il 1,460 fracture) 12/40/F Emergency intubation (LeFort III 360 fracture), diabetes mellitus *HiLa= high volume-low pressure. Type of Tube (Size)* Intubation mask (9) Positioning of the tube difllcult Robertshaw (35) None HiLa (5) None HiLa (7.5) None HiLa (7.5) None Robertshaw (35) Tracheal lumen narrow L m ) ~1 g e a l Carlens with spur (37) None Carlens with spur (35) Larynx narrow, use of stylet HiLa (7.5) Poor vision to the larynx, use of stylet HiLa (7.5) Left main bronchus narrow HiLa (7) Poor vision due to bleeding, use of stylet HiLa (7.5) Poor vision due to bleeding, use of stylet The tentative diagnosis of a tracheobronchial injury made by the clinical features and by the findings from routine chest radiographs was confirmed by an emergency fiberendoscopy. In all but patient 1 (attempt of percutaneous tracheostomy), patient 10 (emergency intubation of the left main bronchus), patient 11 and patient 12 (both emergency intubation following craniocerebral injury), standard inhalative anesthesia using nitrous oxide and muscle relaxants had been applied. Monitoring of the cuff pressure was not reported in any case. In four patients, the duration of anesthesia was short, ranging from 15 to 25 min. Another four patients had major thoracic surgery with essentially longer duration of intubation (90 to 300 min). In those four patients, double-lumen tubes of the Carlens type (n=2) No. 37 with carina! spur, and of the Robertshaw type (n=2) No. 35 without spur, respectively, had been used. Six patients had been intubated using "high volume-low pressure" tubes with sizes between 7 and 7.5 mm. In the child, a latex tube with 5-mm diameter had been inserted. In this case, the patient's record did not give evidence of whether the cuff had been inflated. The types of operation and the endotracheal tubes used are listed in Table 1. In eight instances, the anesthetists reported difficulties during intubation; in the remaining patients, no problems whatsoever were encountered (Table 1). Symptoms All patients showed at least one of the typical symptoms of a rupture of the large airways such as, pneumomediastinum, tracheobronchial hemorrhage, subcutaneous emphysema, dyspnea, or difficulties during artificial ventilation, respectively. In those cases diagnosed with a delay, retrosternal pain and leukocytosis (12,000 to 15,000 WBCs/mm 3 ) became evident. The interval until the manifestation of symptoms or until diagnosis, respectively, differed widely (Table 1). In two patients, the diagnosis was made during thoracic surgical interventions because of contralateral tension. In one patient who had sustained a rupture of the left main broncl1us after selective left-sided intubation for rightsided bleeding, the lesion was detected during the same endoscopy session after the tube had been pulled back. In the 8-year-old child, first symptoms of an injury to the upper airways became apparent immediately after extubation. The patient in whom an attempt to install a percutaneous tracheostomy had been made developed sudden tension as well as subcutaneous and mediastinal emphysema 4 h later. In one case of emergency intubation,, subcutaneous emphysema, and pneumomediastinum evolved after 6 h. In the other one, almost identical features, accompanied by pneumoperitoneum, developed after an interval of CHEST I 112 I 3 I SEPTEMBER,

3 Table 2-Symptoms, Preoperative Delay, Endoscopic Findings, and Outcome No. Symptoms Interval Until First Symptoms Endoscopic Findings* Treatment/Outcome hemoptysis, subcutaneous e mphysema Cough, dyspnea, retrosternal pain hemoptysis, subcutaneous emphysema Cough, hemoptysis, nasal voice, dyspnea, retrosternal oppression, pne umomediastinum Pneumomediastinum, subcutaneous emphysema, hemoptysis Intraoperative left-sided Intraoperative subcutaneous emphysema at the neck nasal voice, subcutaneous emphysema, hemoptysis Nasal voice, cough, hemoptysis, pneumomediastinum, retrosternal pain, subcutaneous emphysema Pneumomediastinum, he moptysis, subcutaneous emphysema, hemoptysis, pneumoperitoneum Pneumomediastinum, hemoptysis, cuff does not prevent air leak, 4h 3d Immediately after extubation 4d Intraoperatively Intraoperatively 5d 20 min *All ruptures were localized in the membranaceous wall of the trachea. 6h 13-cm t ear down to the 7 -em tear 1 em above 8-cm tear down to the 10-cm tear 2 em above 11-cm tear down to the carina 2-cm central tear in the left main bronchus beginning at 5-cm tear at the thoracic inlet 4-cm tear, 3 em above th e ca1ina 5-cm tear, 2 em above 3-cm tear from 1 em above into the left main bronchus 12-cm tear down to the 10-cm tear, 1 em down to Operation/died of septic ARDS on day 21 Suture using the actual surgical access/uneventful Operation/died of unrelated multiorgan failure on day 23 Operation using the right pectoralis muscle as reinforcement for the suture/uneventful Operation/died of sepsis due to abdominal trauma on day 54. Another two patients showed first symptoms after uneventful routine anesthesia. Both had slight bronchial hemorrhage and subcutaneous as well as mediastinal emphysema, complicated by a in one of them. Two patients reported cough and dyspnea at exertion and retrosternal pain, 3 and 4 days after uneventful anesthesia for lobectomy or laparoscopy, respectively. In one of them, the radiograph revealed a pneumomediastinum. The longest interval until the appearance of typical symptoms was 5 days following surgical revision of felon of the index finger. Only then subcutaneous and mediastinal emphysema, nasal voice, and hemoptysis led to the diagnosis, though cough had been present for 3 days. Localization of the Ruptures The localization of the ruptures comprised both the cervical and the intrathoracic portion of the 776 trachea in seven, the intrathoracic portion only in three, and the left main-stem bronchus in two cases. The injuries had a length between 2 and 13 em, four of them comprising practically the whole length of the trachea from the cricoid cartilage down to the carina. All ruptures were longitudinal and afflicted the membranaceous wall (Table 2). Treatment and Outcome All patients underwent surgical repair of the lesion with either a thoracotomy or both thoracotomy and cervical approach, depending on both the length and the localization of the rupture (the tears were closed by a one layer continuous suture line using absorbable material). In one patient who was operated on 5 days after the trauma, the extent of the mediastinitis did not allow a clear-cut suturing of the tear. In this patient, a pedicled pectoralis major muscle flap had to be used both to cover the suture line and to

4 prevent the formation of an esophagotracheal fistula, as the esophageal wall already showed signs of severe inflammatory damage. Nine of the 12 patients experienced an uneventful. Three patients died of a septic condition which only in one case could be attributed directly to the tracheobronchial trauma: this patient developed septic ARDS following mediastinitis. In one instance, multiorgan failure developed following esophageal resection in a patient with impaired liver function; in the other, the fatal sepsis was due to severe abdominal trauma. ARDS was present in these two cases as well, necessitating high ventilatory pressure. Eventually, the suture line yielded in both cases and an esophagotracheal fistula developed. Another repair was impossible due to the preterminal general condition (Table 2). DISCUSSION Once the tentative diagnosis of an iatrogenic tracheobronchial disruption has been made, an immediate fiberendoscopic evaluation has to be done in the operating room, since this diagnostic procedure may precipitate "wide open" injury, thus requiring emergency operation. Surgical repair ought to be done in a specialized unit with both anesthetists and surgeons experienced in the operative therapy of tracheobronchial lesions. This will increase the basically good prognosis of such injuries. An absolute indication for surgical repair is present whenever a transmural tear with a length exceeding 1 em causes and/or pneumomediastinum. Especially in the presence of long disruptions of the membranaceous wall, the esophageal wall will prolapse into the tracheal lumen, causing acute respiratory distress. Short lacerations in the upper third of the trachea, especially if they do not involve the whole thickness of the tracheal wall, can sometimes be treated by antibiotics and intubation with the cuff inflated distal to the tear. 3 The therapist has to be aware, however, that conservative treatment may result in cicatrization and stenosis. As very small, incomplete lacerations will heal unnoticed, the true incidence of iatrogenic trauma to the trachea and the bronchi remains unknown. In a variety of studies, the detrimental effect of high cuff pressures has been demonstrated. 4 Normally, the capillary perfusion pressure at the tracheal mucosa ranges from 25 to 30 mm Hg. If the cuff pressure exceeds 32 mm Hg, the capillmy circulation within the mucosa stops. Depending on the duration of the intubation, damage to the mucosa will result. 5 Another possible cuff-related injury is simple mechanica! disruption as soon as the diameter of the cuff exceeds the maximum diameter of the airway. As all cuffs are permeable to nitrous oxide and thus \vill expand during anesthesia, fatal overinflation can occur, even if initially the pressure is normal. 6 As there are no recordings of the cuff pressure in our patients, we can only presume that at least some injuries, especially those localized in the left main bronchi, could have been due to cuff overinflation of the left sided double lumen tube. Inappropriate use of a stylet is another notorious cause of iatrogenic tracheal laceration. 7 In four of our patients, the use of a stylet was documented, in two instances because of very poor sight due to profuse bleeding from cranial wounds. In four of our cases, double-lumen catheters had been used. The size of the catheter and inflation of the cuff may have contributed to the resulting trauma. Moreover, however, two patients had undergone esophagectomy: it has been shown that due to the close anatomic relationship of the esophagus and the tracheobronchial tree, a weakening of the membranaceous wall may result. It can only be speculated whether the spur of the catheter did cause the injury in two cases, though this mechanism has been documented by other investigators. 8 As a consequence, double-lumen catheters with carina! spur should be used with caution. Percutaneous tracheostomy is a technique developed only recently. To our knowledge, a full-length rupture of the membranaceous wall has not been reported up to this time. In our patient, who had already been receiving artificial ventilation for 15 days, the tube was removed and a laryngeal mask was inserted prior to performing the percutaneous tracheostomy. We presume that the posterior wall of the trachea had been perforated when inserting the guide wire. The subsequent stepwise dilation enlarged this perforation. When finally the tracheostomy tube was inserted, the therapist found that it could not be placed in a way ensuring sufficient ventilation. Probably the tentative inflation of its cuff further enlarged the rupture. Eventually, the attempt of percutaneous tracheostomy was stopped and the patient was reintubated transorally. Deliberately, the cuff was placed far distally, close to the carina, to preserve the tracheostomy site from further trauma. This explains why the symptoms of rupture evolved hours later. An endotracheal tube might have protected the posterior tracheal wall during insertion of the guide wire. Only one of our patients was an adult man. The prevalence of female patients vvith iatrogenic tracheal rupture has already been observed b y other authors. This fact has lead to the assumption that the membranous trachea is less finn in women and CHEST I 112 I 3 I SEPTEMBER,

5 children than in men. 9 Other factors contributing to tracheal rupture such as poor general condition, old age, and diabetes mellitus were present in three patients. 10 There were no hints about coughing or other signs of the patients moving during anesthesia with the cuff inflated, factors that also have been found to cause ruptures: 9 any small longitudinal tracheal tear tends to enlarge in both directions, once a high intraluminal pressure is present. For this reason, it is advisable to do bronchoscopy only with the patient in full relaxation once a tracheal disruption has been suspected. Reintu bation in these patients must be done under fiberscopic control to ensure a correct placement of the tube, if possible distal to the lesion. In presence of a tear reaching down to, it can be necessary to perform a selective intubation of the nonafflicted main bronchus. Anesthesia for the surgical repair of tracheobronchial disruption is demanding. It includes the use of jet ventilation, apneic oxygenation, and if necessary, selective intubation over the operative site, as well as a meticulous monitoring. A close cooperation with the thoracic surgeon is mandatory. Ruptures extending into both main bronchi may even require the use of a cardiopulmonary bypass. CONCLUSIONS Even if care is applied during intubation, superficial lesions to the larynx and to the tracheobronchial tree are unavoidable, especially under emergency conditions. To prevent large, iatrogenic disruptions, the use of small rather than large tracheal or tracheobronchial tubes is advisable. Double-lumen catheters with carina! spur should be generally avoided. Both initial and continuous monitoring of the cuff pressure during anesthesia is mandatory to overcome the problem of unnoticed cuff distention due to diffusion of nitrous oxide Even in a case of the slightest suspicion for iatrogenic tracheobronchial rupture, immediate further investigation by chest radiographs and fiberendoscopy under full relaxation is necessary. If a reintubation is necessary, it should be performed under bronchoscopic guidance and if possible in the presence of a thoracic surgeon. REFERENCES 1 B ein Th, Lenhart FP, Berger H, e t a!. Ruptur der Trachea bei erschwerter Intubation. Anaesthesist 1991; 40: Maxeiner H. Weichteilverletzungen an kehlkopf bei notfallmaessiger intubation. Anaesth Intensivmed 1988; 29: d' Odemont JP, Ptingot J, Goncette L, e t a!. Spontaneous favorable outcome of tracheal laceration. Chest 1991; 99: Van Klarenbosch J, Meyer J, de Lange J. Tracheal rupture after tracheal intubation. Br J Anaesth 1994; 73: Striebel HW, Pinkwart LU, Karavias Th. Trachealruptur durch zu stark geblockte tubusmanschette. Anaesthesist 1995; 44: Messahel BF. Total tracheal obliteration with a low-pressure cuffed tracheal tube. Br J Anaesth 1994; 73: Wagner A, Roeggla M, Hirsch! M, et al. Tracheal rupture after emergency intubation during cardiopulmonary resuscitation. Resuscitation 1995; 30: Wagner LD, Gammage GW, Wong ML. Tracheal rupture following the insertion of a disposable double-lumen endotracheal tube. Anesthesiology 1985; 63: Marty-Ane CH, Picard E, Jonquet 0, et al. Membranous tracheal rupture after endotracheal intubation. Ann Thorac Surg 1995; 60: Chilla R, Chilla-Wubbena U. Intubationsschaeden von kehlkopf und trachea-ursachen, formen, therapie und endoskopische frueherfassung zur kontrollierten prophylaxe. Anaesthesist 1983; 32: Abbey NC, Green DE, Cicale MJ. Massive tracheal necrosis complicating endotracheal intubation. Chest 1989; 95: Luna CM, Legarreta G, Esteva H, et al. Effect of tracheal dilatation and rupture on mechanical ventilation using a low-pressure cuff tube. Chest 1993; 104: