S pneumonia may be the presenting signs of advanced

Transcription

1 ORIGINAL ARTICLES Endoscopic Relief of Malignant Airway Obstruction Douglas J. Mathisen, MD, and Hermes C. Grillo, MD Department of Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts The ability to manage acute airway obstruction can be life-saving. Airway relief should be expeditious and immediate, with low morbidity and mortality. It should not interfere with future definitive therapy. In patients with terminal malignancy, it should be economical in cost and should minimize hospitalization. We used biopsy forceps and the rigid bronchoscope to core out 56 patients with obstructing airway neoplasms. The location of the obstruction was trachea in 16 patients, carina in 24, main bronchi in 8, and distal airway in 8. Improvement in the airway was accomplished in 90% of patients. A single bronchoscopy was sufficient in 96%. Nineteen complications occurred in 11 patients: pneumonia in 5, bleeding in 3, pneumothorax in 2, hypoxidhypercarbia in 2, arhythmias in 6, and laryngeal edema in 1. There were four deaths within 2 weeks of core-out related to respiratory failure. Further therapy consisted of resection in 28.6% (tracheal in 9, carinal in 3, pulmonary in 41, irradiation alone or in combination with chemotherapy in 60.7%, and no therapy in 10.7%. Palliation of symptoms and establishment of an airway in acute obstruction is the goal. Survival depends on the effectiveness of the proposed treatment. We find this time-honored method superior to use of the laser. (Ann Thorac Surg 1989;48:469-75) tridor, acute respiratory distress, and obstructive S pneumonia may be the presenting signs of advanced airway neoplasms. These neoplasms may be rapidly growing or slow-growing. Symptoms depend on location and extent of the tumor. Neoplasms of the trachea and carina may cause obstructive airway symptoms as opposed to neoplasms of the distal airway, which are more likely to show initial symptoms of postobstructive pneumonia. With the bronchoplastic techniques currently available, neoplasms of the trachea, carina, main bronchi, and lobar orifices may be amenable to surgical resection. In patients with such obstructions, it may be necessary to open the airway to allow obstructive pneumonia to clear or to facilitate airway management at the time of operation. Opening an obstructed airway in patients who have acute respiratory distress with potentially resectable tumors will allow time for a thorough workup, correction of underlying medical problems, or tapering of steroid administration if the patients have been treated for asthma. Many patients with advanced tumors have obstructed airways, and no surgical therapy is available. In such patients, palliative relief of symptoms is necessary to provide a good airway, allow clearing of obstructive pneumonia, and improve quality of life. It may also be necessary to open the airway to allow palliative treatment such as radiation therapy to be performed. Many methods of opening obstructed airways are available. The ideal method should be simple, expeditous, safe, and readily available to many physicians. It should allow immediate establishment of an airway and should not require multiple procedures. Many of these patients do not tolerate multiple procedures or general anesthetic Presented at the Twenty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Sep 26-28, Address reprint requests to Dr Mathisen, Massachusetts General Hospital, Warren 1109, Boston, MA agents. The procedure should not interfere with proposed resection or other therapy such as irradiation or chemotherapy. It must be cost-effective and time-effective for patients with advanced malignant disease, as long-term survival is not likely. It must be a procedure that can be accomplished with low morbidity and low mortality. We have used the time-honored method of endoscopic removal of tumor to achieve these goals. The combination of biopsy forceps and the tip of the rigid bronchoscope to core out the tumor has been very effective in our hands and satisfies many of the requirements for opening an obstructed airway. Material and Methods We reviewed the charts of 56 patients who underwent endoscopic removal of tumors of the airway between 1977 and 1988 at Massachusetts General Hospital. This is a consecutive series of patients. During this time period, 2 patients were unsuitable for our technique and were referred for laser therapy. Technical Considerations RADIOLOGICAL ASSESSMENT. If time permits, careful radiological assessment is a valuable asset to operative management. Tomography of the trachea and canna often provides information about extent of involvement, configuration, and axis of the airway and patency of the distal airway. This information can be invaluable when first gaining control of the airway. Obviously, patients in respiratory distress are unable to have detailed radiological assessment of their airways; a plain chest roentgenogram often will suffice. ANESTHESIA. The anesthetic management of patients with a compromised airway is crucial to a successful outcome. In Virtually every Case, We have used the same technique used for patients with tracheal stenosis (the deep by The Society of Thoracic Surgeons /89/$3.50

2 470 MATHISEN AND GRILL0 Ann Thorac Surg 1989;48:46%75 halothane inhalation technique described in detail by Wilson [l]). This technique allows the patient to maintain spontaneous ventilation throughout the procedure. Induction of anesthesia can be time-consuming and requires patience on the part of the anesthesiologist and surgeon. Any disadvantages this may impose are greatly outweighed by the tremendous advantage of having the patient spontaneously ventilating throughout the procedure, which is especially important before the exact nature of the obstruction is known. In such circumstances, the surgeon will have more time to establish an airway than if the patient were apneic. During the course of tumor removal, maintaining the airway may be difficult, and patients ability to breathe is often the only thing that sustains them through this period. We have deliberately avoided use of muscle relaxants and agents that depress respiratory drive. Patients need to be able to clear secretions, debris, and blood from their airway postoperatively. Anesthetic agents that interfere with this function should be avoided. ENDOSCOPIC TECHNIQUE. It is most important that the surgeon be in attendance during induction of anesthesia. Sudden compromise of the airway may occur, and immediate insertion of a rigid bronchoscope may be required to secure the airway. This demands that the surgeon and assistants be prepared and have all necessary instruments available to establish an emergency airway. We had no occasion to perform emergency tracheostomy as all patients could be intubated with a rigid bronchoscope. Knowledge of the location of the tumor makes this task much easier. The technique of inserting a rigid bronchoscope should be familiar to all surgeons. It is important to have available rigid bronchoscopes of graded sizes (3.5 to 9 mm). A small pediatric rigid bronchoscope may be the only bronchoscope that can be inserted through very proximal, tight malignant strictures. This allows establishment of an airway; subsequently, larger bronchoscopes can be inserted. For more distal tumors, insertion of a larger bronchoscope in the proximal airway allows careful inspection and suctioning of any secretions. In most cases a No. 8 rigid bronchoscope can be passed along one side of the tumor to establish an airway and remove retained secretions. It is important to establish an airway, clear retained secretions, and stabilize the patient before proceeding with tumor removal. This allows the patient to be maximally ventilated at the start of the procedure. Inspection of the distal airway will aid the bronchoscopist in reestablishing the airway once the actual procedure has started. It is important to establish the axis of the airway when coring-out is started. The endoscopist should stay parallel to the axis at all times to avoid penetrating the wall and creating a pneumothorax or injuring the pulmonary artery. The tumor should first be biopsied to determine its consistency and vascularity. Once this has been established, the surgeon can decide whether to attempt repeated removal of the tumor with bronchoscopic forceps or to use the tip of the bronchoscope to core out the tumor. If the tumor is believed to be of sufficient consis- tency and vascularity to allow a large portion of it to be removed, the rigid bronchoscope is used in a corkscrew fashion. The beveled tip of the bronchoscope is used to shave off a large piece of the tumor. Most tumors are of such consistency that shaving is easily accomplished and does not require excessive force. The anesthesiologist should discontinue assisted ventilation of the patient to avoid forcing debris and tissue more distally into the airway. The piece of tumor can then be grasped with forceps and removed or may be aspirated. Occasionally, large pieces of tumor must be held against the end of the bronchoscope with biopsy forceps and tumor and bronchoscope must be removed from the airway at the same time. The endoscopist should intermittently cease removing tumor, reestablish the airway, and allow the anesthesiologist to ventilate the patient. This will also allow the bronchoscope to tamponade any bleeding that might occur from the surface of the tumor. It is important to clean the airway of debris periodically. This may be aided by irrigation. Use of the flexible bronchoscope through the rigid bronchoscope allows suctioning at the subsegmental level. CONTROL OF HEMORRHAGE. Major hemorrhage is uncommon and did not occur in this series of patients. Minor bleeding is easily controlled with simple measures. Inigation with saline is the first maneuver to dimipish bleeding, and frequently is all that is necessary. Dilute epinephrine (0.1 mg/ml) or epinephrine-soaked pledgets on long applicators will stop persistent oozing. Various methods of tamponade are also successful in managing any persistent bleeding that does not respond to the previous methods. Use of the rigid bronchoscope to tamponade the raw surface of a tumor is effective. For bleeding from the main bronchus or more distal airway, dental packs or Fogarty venous occlusion catheters can be used to tamponade bleeding. Fogarty venous occlusion balloons can be left in place if necessary. Use of a carefully placed endotracheal tube is also effective in controlling bleeding that persists beyond the previously mentioned measures. Placing the endotracheal tube so that the balloon is resting against the raw surface and then inflating the balloon will serve to tamponade and control any active bleeding. This measure is rarely necessary. Insulated electrodes are available to coagulate bleeding points if needed. Any coagulation abnormalities should be corrected before the procedure is performed. Use of fresh-frozen plasma and platelets at times may be necessary to correct these abnormalities. Great care must be used to clear the airway if the patient is intubated at the completion of the procedure. The endotracheal tube may be plugged easily by blood clot. It is preferable to have the patient extubated and awake to clear the airway by coughing. Age and Sex Distribution There were 28 men and 28 women in this series. The median age for men was 59 years (range, 22 to 79 years), and the median age for women was 54 years (range, 20 to 81 years).

3 Ann Thorac Surg 1989:48: MATHISEN AND GRILLO 471 Table 1. Histological Classification of Tumors Treated Histology Trachea Canna Bronchus Squamous carcinoma Adenoidcystic Adenocarcinoma Thyroid Sarcoma Carcinoid Mucoepidermoid Lymphoma Small cell carcinoma Melanoma metastatic Metastatic acinar carcinoma 8 0 Symptoms All patients were symptomatic when first seen. The majority of patients experienced symptoms of shortness of breath or dyspnea on exertion (88%). The next most common symptom was hemoptysis (45%). Obstructive pneumonia was present in 18 patients. Nineteen patients had acute respiratory distress or stridor. Five patients had been treated for long periods of time for adult onset asthma and were on heavy doses of steroids when the diagnosis of obstruction was made. These patients have gradual onset of shortness of breath and wheezing with a clear chest roentgenogram. Tracheal tumors are sufficiently rare that little attention is paid to the tracheal air column, or the tumor is subtle enough and cannot be seen. Steroid administration is a contraindication to tracheal or carinal resection and must be tapered before resection. Location The trachea was the site of primary tumor involvement in 16 patients. Twenty-four patients had tumor involving the carina. The right main bronchus was involved in 6 patients, and the left main bronchus was involved in 2 patients. The distal airways were involved in 8 patients at the lobar or segmental level. Histology Squamous cell carcinoma was the most common histology in the trachea, canna, and bronchus (Table 1). Adenoid cystic carcinoma was the next most common histological finding in the canna and the trachea. The miscellaneous categories included thyroid cancers, carcinoids, mucoepidermoid tumors, sarcomas, lymphomas, and various metastatic tumors. Initial Treatment The majority of patients were evaluated and underwent bronchoscopy electively (61.8%). Core-out was performed at the time of resection to provide an adequate airway in some patients. The remainder of the patients had to be treated emergently to establish an airway or control symptoms (23.2%) or urgently because of postobstructive pneumonia (16%). Results Fifty-one of the 56 patients had an improved airway after endoscopic removal of tumor. The improvement was corroborated by subjective improvement in symptoms or radiographic appearance. Five patients had no improvement, all of whom had tumor at the lobar or segmental level. Only 2 patients required repeat bronchoscopy within 2 months of the initial successful core-out. Both patients underwent repeat bronchoscopy at seven days because of subsequent airway compromise. Additional tumor was removed in both, with substantial improvement. There were no intraoperative deaths in this series. Four patients died within 2 weeks of endoscopy of causes unrelated to the procedure. All 4 patients had received no benefit from previous radiation therapy and were in desperate condition with postobstructive pneumonias. All patients died of respiratory failure or sepsis and could not be weaned from the ventilator. Length of Procedure The exact duration of the procedure was difficult to determine from review of records, but information regarding the duration of anesthesia, which did not correlate with duration of procedure, is available. Taking this factor into account, however, median length of anesthesia was 60 minutes (range, 20 to 150 minutes). Further Therapy Sixteen patients (28.6%) underwent resection of the tumor. There were nine tracheal resections, three canna1 resections, and four pulmonary resections, including one sleeve lobectomy. The indications for endoscopic resection of the tumor varied. Some patients were originally seen with postobstructive pneumonia and had relief of the obstruction and subsequent clearing of pneumonia before definitive resection. A few patients were first examined while on high-dose steroids for presumed asthma, thus precluding definitive surgical resection at that time. Coring out the tumor allowed the patients to return home and be weaned from the steroids; resection was performed later. A few patients had acute respiratory distress and were initially treated by endoscopy for relief of airway obstruction. Subsequent workup and successful resection were then performed. The remainder of patients had endoscopic removal of the tumor at the time of operation to allow safe management of the airway. Most patients had unresectable disease. Thirty-four patients had other forms of treatment (60.7%). Radiation therapy was the most common additional treatment; a few patients were treated with chemotherapy or a combination of chemotherapy and radiation therapy. Six patients had no further therapy. Four of these patients had failure of previous radiation therapy or chemotherapy and subsequently died. Two patients refused additional therapy.

4 472 MATHISEN AND GRILL0 Ann Thorac Surg 1989;48: Complications Nineteen major complications were recorded in 11 patients. Five patients developed pneumonia in previously unaffected lungs after relief of postobstructive pneumonia. All of these patients responded to antibiotics and chest physiotherapy. Virtually all patients had minor amounts of bleeding. Three patients had more than minor bleeding. None lost more than 500 ml of blood, and the bleeding was easily controlled with the methods previously mentioned. Two patients developed a pneumothorax postoperatively. Only 1 of them required a chest tube. Hypoxia and hypercarbia were poorly documented in patients before the advent of CO, monitoring and oximetry. Two patients had documented hypercarbia and hypoxia during the course of the operation that persisted and required intubation at the end of the procedure. These 2 patients were extubated within 24 hours. Hypoxia and hypercarbia were related to blood, retained secretions, and debris. Minor arrhythmias were noted during the course of the operation in many patients. Only 6 patients required intervention beyond routine measures to control their arrhythmias. Four patients with ventricular premature beats required intravenous Xylocaine (lidocaine) for 24 hours. Atrial arrhythmias requiring digoxin were documented in 2 patients. One patient developed postoperative laryngeal edema. This responded to racemic epinephrine and steroids. Tracheostomy was avoided in this patient. Survival Short-term survival is directly dependent on the success of the endoscopic core-out of tumor. Long-term survival depends on the effectiveness of proposed treatments. For patients who have an obstructing lesion that requires immediate removal, coring out will allow future resection when appropriate. For inoperable malignancies, coring out will establish an airway until appropriate therapy is started. For patients who have failed previous therapies, endoscopic removal of tumor is a temporary measure that allows a patent airway. Use of tracheotomy tubes and T tubes may ultimately be effective in maintaining an airway in this group of patients who have no other options. Because of the diverse patient population reported in this series, no conclusions can be drawn regarding survival. Survival of patients who underwent resection has ranged from 5 months to more than 8 years with no evidence of disease. Survival of inoperable patients after irradiation or chemotherapy, or both, has ranged from 2 months to 82 months with a median of 6 months. Success of treatment depends on histology, location, and extent of disease. Comment Use of the rigid bronchoscope and biopsy forceps has been recognized as an effective method of removing endobronchial tumor since the earliest days of endoscopy. In 1915, Jackson [2] removed a fibroma from the lower margin of the left upper lobe bronchus. He concluded that the removal of benign neoplasms of the bronchi can readily be effected by bronchoscopy. For many years, patients with unresectable neoplasms, elderly patients, or poor-risk patients were palliated reliably with this technique. Concern about hemorrhage as a result of this technique has been expressed, but such hemorrhage has been poorly documented. Houston and colleagues in 1969 [3] and Nakratzas and co-workers in 1974 [4] reported their results with endoscopic removal of obstructing tumors of the airway. The patients in neither series experienced profuse bleeding, but these investigators perpetuated the misconception of the possibility of hemorrhage by citing earlier reports in which this complication was mentioned. No other reports in the literature have countered this unfavorable opinion. When notoriously vascular tumors such as hemangiomas, arteriovenous malformations, and some carcinoids are avoided, massive bleeding from endoscopic removal of tumors is exceedingly unusual, as in our series. Even some carcinoids can be managed in this fashion, as evidenced by the 2 patients successfully treated in our series. Adherence to the principles of the technique and an awareness of the methods to control hemorrhage should minimize the threat of hemorrhage. Although few studies report endoscopic removal of tumor, many studies describe large series of patients treated with lasers [ Initially, most reports were from pulmonologists using yttrium-aluminum-garnet lasers through flexible bronchoscopes. Pulmonologists are now rediscovering the rigid bronchoscope as the preferred way to use the laser at a time when many thoracic surgeons are abandoning rigid bronchoscopy in favor of flexible bronchoscopy. There is little doubt that the laser is an alternative to biopsy forceps and the rigid bronchoscope. Both achieve airway patency rates of approximately 90%. Critical comparison of the instruments, however, shows that the laser has few if any advantages and, in fact, may have disadvantages for the patient. The laser was touted as a major advance because of its ability to control hemorrhage. Hemorrhage, as we have shown, is quite infrequent and easily controlled by much simpler and less expensive methods. Lasers may even be responsible for major hemorrhage, as evidenced by the 0.5% to 10% incidence of moderate hemorrhage and 1% to 3% incidence of fatal hemorrhage in most series [7-9, 12, 13, 15, 181. Our main objections to the use of the laser are the expense (equipment, training personnel, physician, and operating room costs), length of procedure, need for repeated procedures to open airways, inferior anesthetic techniques, and unique risks (fires, hemorrhage, potential eye injuries). We consider all of these to be important factors in choosing our technique rather than the laser. It is ironic that many laser enthusiasts now report their technique of using the laser to coagulate the mass and mechanically scoop out the tumor with the bronchoscope. The airway wall is photocoagulated for hemostasis (8, 91. This statement tends to support our view that the laser is unnecessary, because bleeding is rarely a problem with our technique. Complicated airway obstructions are indeed a challenge to physicians. Complications are inherent in dealing with these problems. Strict attention to details of technique,

5 Ann Thorac Surg 1989;48: MATHISEN AND GRILL0 473 however, should minimize complications. Our anesthetic technique of using halothane to maintain spontaneous ventilation eliminates the risks to patients posed by barbiturates, narcotics, and muscle relaxants needed for laser therapy [18-22]. The final concern about lasers is the potential for inappropriate treatment. A wide spectrum of physicians is now involved in treating airway neoplasms with lasers. Laser therapy for tracheobronchial malignancy is practiced by thoracic surgeons, otolaryngologists, and pulmonary medicine specialists. All who are involved in the care of such patients must be aware of the surgical options available to patients with malignant disease of the airway. Tracheal resections, carinal resections, and sleeve resections of the tracheobronchial tree are accepted methods of treatment and can be curative. In most situations, laser therapy is palliative only and offers no chance of cure. Patients with potentially curable lesions should not be denied the chance for cure because physicians are unfamiliar with these techniques. Input from knowledgeable thoracic surgeons should be mandatory to avoid this situation. Unnecessary delay caused by repeated laser treatments should be avoided as the opportunity for cure may be lost. Another misconception exists regarding the ability of laser therapy to treat benign tumors of the airway. Rarely does a tumor involve only the mucosa of the airway. Penetration deeper than the mucosa poses the risk of perforation of the airway wall by the laser. Resectional therapy for benign tumors is usually curative. Use of the laser to treat obstructing tumors of the airway is indeed an alternative to the technique we have described, but such use of the laser must be justified as we believe it to be inferior in terms of cost, time, success in achieving a patent airway initially, and risk of complications to the patient and physician. Even the most ardent supporters of the laser now recognize the ability of the rigid bronchoscope to scoop out tumor after treating it with the laser. We believe that the laser represents an unnecessary step. Our data support greater consideration of the technique we describe. References 1. Wilson RS. Anesthetic management for tracheal reconstruction. International trends in general thoracic surgery, Vol 2. Philadelphia: W.B. Saunders, 1987: Jackson C. Endothelioma of bronchus removed by peroral bronchoscopy. Am J Med Sci 1917;53: Houston HE, Payne WS, Harrison EG Jr, et al. Primary cancers of the trachea. AMA Arch Surg 1969;99: Nakratzas G, Wagenaar JPM, Reintjes J, et al. Repeated partial endoscopic resections as treatment for two patients with inoperable tracheal tumours. Thorax 1974;29: Strong S, Baughan CW, Polany T, Wallace R. Bronchoscopic carbon dioxide laser surgery. Ann Otol 1974;83: Toty L, Personne C, Colchen A, et al. Laser treatment of postintubation lesions. In: Grillo HC, Eschapasse H, eds. International trends in general thoracic surgery, Vol2. Major challenges. Philadelphia: W.B. Saunders, 1987;4: Brutinel WM, Cortese DA, McDougall JC, et al. A two-year experience with the neodymium-yag laser in endobronchial obstruction. Chest 1987;91: Cavaliere S, Foccoli P, Farina PL. Nd:YAG laser bronchoscopy; a five-year experience with 1,396 applications in 1,OOO patients. Chest 1988;94: Dumon JF. YAG laser bronchoscopy, Vol 5. New York Praeger, Dumon JF, Reboud E, Garbe L, et al. Treatment of tracheobronchial lesions by laser photoresection. Chest 1982;81: Goldberg M. Endoscopic laser treatment for bronchogenic carcinoma. Surg Clin North Am 1988;68: Kvale PA, Eichenhorn MC, Radke JR, Miks V. YAG laser photoresection of lesions obstructing the central airways. Chest 1985; McElvein RB, Zorn GL Jr. Indications, results, and complications of bronchoscopic carbon dioxide laser therapy. Ann Surg 1984;199: Parr GVS, Unger M, Trout RG, Atkinson WG. One hundred neodymium-yag laser ablations of obstructing tracheal neoplasms. Ann Thorac Surg 1984;38: Shapshay SW, Healy GB, Davis RK, Strong MS. Endoscopic management of airway obstruction from tracheobronchial neoplasia; use of the carbon dioxide laser. Lahey Clin Foundation Bull 1983;32: Toty L, Personne C, Colchen A, Vourc h G. Bronchoscopic management of tracheal lesions using the neodymium yttrium aluminum garnet laser. Thorax 1981;36: Wolfe WG, Cole PH, Sabiston DC Jr. Experimental and clinical use of the YAG laser in the management of pulmonary neoplasms. Ann Surg 1984;199:52& Vourc h G, Fischler M, Personne C, Colchen A. Anesthetic management during Nd-YAG laser resection for major tracheobronchial obstructing tumors. Anesthesiology 1984;61: Dumon JF, Shapshay S, Bourcereau J, et al. Principles for safety in application of neodymium-yag laser in bronchology. Chest 1984;86: Casey KR, Fairfax WR, Smith J, Dixon JA. Intratracheal fire ignition by the Nd-YAG laser during treatment of tracheal stenosis. Chest 1983;84: Warner ME, Warner MA, Leonard PF. Anesthesia for neodymium-yag (Nd-YAG) laser resection of major airway obstructing tumors. Anesthesiology 1984;60:23C Vourc h G, Tannieres ML, Toty L, Personne C. Anesthetic management of tracheal surgery using the neodymiumyttrium-aluminium-garnet laser. Br Anaesthesiol 1980;52: DISCUSSION DR RICHARD B. McELVEIN (Birmingham, AL): Dr Mathisen describes the anesthetic technique to include deep halothane anesthesia with spontaneous ventilation; however, he also describes having the anesthesiologist stop ventilation while tissue is removed. Dr Mathisen, can you clarify whether these patients are being ventilated or assisted in their ventilation? Also, we have found the peripheral pulse oximeter to be an excellent monitoring instrument; do you find it useful?

6 474 MATHISEN AND GRILLO Ann Thorac Surg 1989;48:%9-75 Flow volume loops are a very sensitive indicator and an inexpensive, reproducible method of determining whether a patient has serious recurrent airway obstruction. Dr Mathisen, have you had any experience in using these? We have patients whom we retreat, not because of failure of the first treatment, but because of regrowth of the tumor. Dr Mathisen, how many patients have you retreated by your technique? We found that if we have established the extent of the disease and retreatment is necessary, many of these patients can be treated on an outpatient basis. In summary, Dr Mathisen has described one method of managing malignant airway obstruction. There are others. The important message is that people with airway obstruction should not be left to suffocate. Expedient care can relieve them of their obstruction and in some cases provide long-term survival. I have used the technique described by Dr Mathisen and have also had extensive experience using the laser; I find the laser technique preferable. DR DAVID J. COHEN (San Antonio, TX): I enjoyed Dr Mathisen s presentation, but I think, in a sense, that the argument he has set up is a straw man. I believe that the real issue is whether to use laser therapy through a rigid bronchoscope or through a flexible bronchoscope. Use of the rigid bronchoscope, with either the yttrium-aluminum-garnet or the CO, laser, allows one to use the coring-out technique that Dr Mathisen has so eloquently described, as well as to biopsy and to maintain the airway, but it also allows the advantage of using the laser for removal of tumor. The laser, with which we have experience, has the added advantage that it can be used for coagulation. I believe that using the laser with the rigid bronchoscope is better than using it with the flexible bronchoscope. The rigid bronchoscope does allow use of the coring-out technique described by Dr Mathisen and provides a much greater opportunity to reopen the bronchus with a single procedure rather than with the multiple procedures so often necessary when the laser is used through a flexible scope. DR PAUL A. KIRSCHNER (New York, NY): I enjoyed the paper very much. I would like to ask a question and then make a brief comment. Dr Mathisen, have you seen metastatic tumors of the trachea treated in this way? 1 have had 2 such patients. The first patient had a metastasis to the midtrachea 12 years after a right upper lobectomy for a stage I adenocarcinoma in He underwent rigid bronchoscopic removal in 1983, 1987, and twice in Radiotherapy was administered after the first bronchoscopy. He is asymptomatic now at the age of 94 years. The second patient had a polypoid metastasis to the same area which appeared 4 years after a sternotomy for a follicular thyroid carcinoma removed in The metastasis was removed with the rigid bronchoscope, and the airway has remained patent to date; she is now 84 years old. She received radioactive iodine. She has developed a metastasis to the manubrium. Dr Mathisen, have you ever had large pieces of tumor break off and fall distally into the airway on either side during the procedure? I had one such experience; a main bronchus was blocked on one side due to recurrent metastatic carcinoma. I removed a large piece which fell off the forceps into the opposite main bronchus, obstructing it. The situation was precarious until I was able to retrieve the piece of tumor and relieve what could have been a fatal asphyxis. We continue to use the rigid scope. DR ROBERT J. GINSBERG (Toronto, Ont, Canada): 1 support use of the rigid bronchoscope and the coring-out technique, which I think is applicable to most major airway endoluminal malignant obstructions. In Toronto, we have used this technique for many years. Approximately 6 years ago, we began using the laser as the primary tool and were initially quite enthusiastic. However, we are once again using the coring-out technique for most cases, and use either the CO, or neodymium:yttriumaluminum-garnet laser as a touch-up at the end of the procedure. We have found that coring is much faster and quite safe and agree with the remarks of Drs Mathisen and Grillo about the utility of this technique. DR KAMAL G. KHALIL (Houston, TX): I would like to know more about the 2 cases of pneumothorax that occurred in this series. What was the pathogenesis of those 2 cases with this technique? PRESIDENT GRILLO: Before Dr Mathisen closes the discussion, I d like to add that I have used this technique for 30 years yet I do not know the number of patients I have treated. I never thought to start recording such cases until the laser came along and a reply was needed. DR MATHISEN: I thank all of the discussants for their comments. I also recognize Dr McElvein s contributions in the area of laser use, and wish to underscore the fact that we recognize the laser as an alternative technique. One important point is that the surgeon must remain involved in the care of these patients. At a time when many societies and specialties are worried about the loss of patients to other disciplines, I think this is one example in which, because of the lack of interest of the thoracic surgeon, some of the other disciplines, such as pulmonary medicine and otolaryngology, are undertaking the care of these patients. Not only does this represent a loss of patients to those other disciplines, but it also represents a loss of input from thoracic surgeons. This is unfortunate because in certain situations a surgical option might be available and, because a physician who is unfamiliar with the condition may not be aware of it, a patient might be denied the opportunity for cure. Many proponents of laser bronchoscopy are beginning to reconsider its usefulness, as Dr Ginsberg has intimated. Frequently the laser is used to photocoagulate the tumor to char it, and the tip of the rigid bronchoscope is then used to scoop it out. Next, the laser is used to touch up the wall of the tumor to prevent any oozing. Yet the laser actually may not be necessary, as bleeding is not the major problem it was once believed to be. Dr McElvein asked about assisted ventilation. With halothane anesthesia, patients are able to ventilate spontaneously at all times. At times during the course of this procedure assisted ventilation is certainly necessary, and the only reason for the anesthesiologist to discontinue assistance is to prevent pushing debris or blood further into the distal airway at the time of the coring-out procedure. We agree that use of oximetry is very helpful in the management of these patients. I m certain that before oximetry was used, many patients became hypoxic and we were unaware of it, yet we continued to use the procedure, not really knowing the status of patients oxygenation. Clearly, the oximeter is of great benefit. We do not have extensive experience using flow volume loops in patients with airway obstruction. 1 think it certainly is a very effective method of following them, although 1 think that most patients tell you when they become symptomatic. In terms of the number of retreated patients, 2 patients were retreated at seven days because of persistent or recurrent symptoms and were found to have a tumor that could then be removed. Four other patients were retreated and represent the difference between the total number of patients and the total number of procedures. These treatments occurred at intervals ranging between 2 and 4 months. I believe that the patient who

7 Ann Thorac Surg 1989;48: MATHISEN AND GRILL0 475 had the most procedures lived for 6 to 7 years with an adenoidcystic tumor and required repeated endoscopic removal of her tumor at about 4-month intervals. We have used the technique to treat exuberant granulations. Using the tip of the scope or the rigid biopsy forceps removes the granulation tissue as effectively as the laser does. I think this underscores the fact that these two methods are alternative techniques for treating the same kind of problem. Dr Cohen raised the issue of the flexible bronchoscope and the rigid bronchoscope. If you use the laser, you should use the rigid bronchoscope, thus avoiding the risk of damaging the flexible bronchoscope by fire or melting the tip of the scope, which obviously can be very expensive. Use of the rigid bronchoscope allows the technique that we have described. If you are comfortable using the laser, I encourage you to use the technique that Dr Cohen described, including use of the rigid bronchoscope as a supplement. Dr Kirschner asked about treatment of patients with metastatic lesions. In this series 8 or 9 patients had metastatic tumors. However, in no patient in this series did we penetrate the airway while removing the tumor. The pathogenesis of pneumothorax in 2 patients may have been related either to intraoperative trauma or to overassisted ventilation and resultant further obstruction of the patient's airway. If one does not observe the axis of the airway and stay parallel to it, the wall of the tumor is easily penetrated and a pneumothorax is created. If the anesthesiologist is overvigorous in assisting ventilation, the airway becomes further obstructed by retained debris or secretions. High pressures might result, in turn causing a pneumothorax. We cannot tell from review of these patients' records which situation occurred.