- Email: [email protected]
Delivery of Intraoperative Radiation Therapy after Pneumonectomy: Experimental Observations and Early Clinical Results
Delivery of Intraoperative Radiation Therapy after Pneumonectomv: Experimental Observations and Early Clinical RGsult; Harvey I. Pass, M.D., William F. Sindelar, M.D., Timothy J. Kinsella, M.D., Anne-Marie DeLuca, B.S., Margaret Barnes, M.D., Scott Kurtzman, M.D., Harold Hoekstra, M.D., Zelig Tochner, M.D., Jack Roth, M.D., and Eli Glatstein, M.D. ABSTRACT Intraoperative radiation therapy (IORT) is capable of delivering high doses of radiation to mediastinal structures while sparing lung parenchyma, heart, and other locoregional tissues. A canine model of pulmonary resection and IORT was investigated by performing a pneumonectomy in 15 adult foxhounds followed by 0 cGy, 2,000 cGy, 3,000 cGy, or 4,000 cGy. No clinical complications developed in 4 animals in the 2,000-cGy group. However, 2 of the 8 animals given a high dose died of esophageal hemorrhage or carinal necrosis. Esophagitis occurred in 10 of 12 animals, and none of the animals experienced bronchial stump dehiscence. In a limited Phase I protocol, 4 patients with non-small cell lung cancer were treated with resection and 2,500 cGy of IORT to two separate ports encompassing the superior and inferior mediastinum. Two patients experienced lifethreatening bronchopleural fistulas, and 2 patients died as a consequence of esophageal problems. One patient had recurrence with brain metastases, and the 1 long-term survivor is free from disease. As opposed to the animal model of thoracic IORT, the clinical study demonstrated major toxicity with respiratory and esophageal morbidity. The therapeutic usefulness of thoracic IORT in the management of lung cancer must be questioned in view of this small but consistent series of patients. Further carefully designed clinical studies using lower doses of IORT are needed. Intraoperative radiation therapy (IORT) may be a potentially useful adjunct combined with operation or external beam irradiation or both in the treatment of locally advanced tumors [l].Using teletherapy techniques such as electron or orthovoltage x-ray beams, a single large dose of ionizing radiation can be delivered to the tumor bed at the time of operation. This procedure can theoretically minimize the extent of normal tissue irradiation by From the Surgery Branch and Radiation Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD. Presented at the Thirty-third Annual Meeting of the Southern Thoracic Surgical Association, White Sulphur Springs, WV, Oct 30-Nov 1, 1986. Address reprint requests to Dr. Pass, Thoracic Oncology Section, Surgery Branch, National Cancer Institute, Building 10, Room 2807, Bethesda, MD 20892.
14 Ann Thorac Surg 44:14-20,July 1987
excluding radiosensitive normal tissues from the radiation field by surgical manipulation or by custom lead shielding placed intraoperatively. Clinical trials of IORT in the treatment of gastric [2], pancreatic [3], and rectal carcinomas [4] as well as retroperitoneal sarcomas [5] have been performed at various centers including the National Cancer Institute (NCI). The use of IORT in the management of thoracic neoplasms has yet to be defined. Moreover, the dose tolerance of normal thoracic viscera has not been investigated with IORT in a large animal model to help establish safe guidelines for treatment portals as well as dose. This report describes the effects of thoracic IORT in a canine model of pulmonary resection at various radiation doses as a function of time, and serves as the basis for clinical investigation of the efficacy of IORT as an adjunct to resection in 4 patients with non-small cell carcinoma of the lung.
Material and Methods
Animal Investigations Fifteen purebred male and female adult American foxhounds ranging in age from 9 to 18 months and weighing 24 to 41 kg were maintained in sheltered runs and fed dog chow and water ad libitum. All animals received humane care in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 8023, revised 1978). Large animals were used to allow for adequate surgical exposure and placement of IORT cones for field size comparable to what might be used in a human research setting. SURGERY AND DELIVERY OF INTRAOPERATIVE RADIATION
A left pneumonectomy was performed with staple closure of the left main bronchus. In this study, a single IORT field was treated using a cylindrical acrylic treatment cone with a 5-cm inner diameter. The treatment cone was sheathed in a stainless steel shield to decrease lateral electron scatter beyond the area of interest [6]. This field encompassed the bronchial stump, pulmonary artery and vein, esophagus, aorta, pericardium, left atrium, phrenic nerve, and a portion of the left ventricle. At the completion of treatment, the portal perimeter was delineated with surgical clips.
THERAPY.
15 Pass, Sindelar, Kmsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
IORT. Four dogs were treated at each of three dose levels of IORT: 2,000 cGy, 3,000 cGy, and 4,000 cGy. Radiation treatment was performed using a Scanitronix Microtron located in the NCI Radiation Oncology Branch IORT suite. The microtron has a full 360-degree rotational gantry, and beam entry angles were determined at the time of IORT cone placement. A 13-meV electron beam was used at a dose rate of 800 to 900 cGy per minute, and doses were quoted at the 90% isodose line. Three additional dogs underwent pneumonectomy and placement of the IORT acrylic cone but were not given treatment; they served as controls. FOLLOW-UP STUDIES. The dogs were monitored by weekly body weights and monthly chest roentgenograms. Contrast esophagograms were performed before the animals were killed. Fiberoptic bronchoscopy and esophagoscopy were performed monthly to document gross pathological changes at the earliest possible time. One animal from each treatment group as well as 1 sham-irradiated control was to be killed electively at one, three, and twelve months following IORT. The remaining animal from each treatment group would be maintained for up to five years for documentation of late IORT effects. All dogs that died or were killed were subjected to a complete autopsy including gross examination of thoracic and abdominal viscera and microscopic examination of the thoracic viscera and tissues. Particular attention was paid to segments of tissues included within the radiation portals. Sections of both irradiated and grossly normal tissues were submitted for histological review, including irradiated bronchial stump, trachea, right mainstem bronchus, right lung, left atrium and ventricle, aorta, phrenic nerve, esophagus, and ligated irradiated stump of the pulmonary artery and vein. All tissues were prepared and stained with hematoxylin and eosin, and special stains including van Gieson’s for connective tissue, Verhoeff‘s plus van Gieson for elastin, and Mallory’s trichrome were used to aid in the examination of connective tissue changes.
Human Investigations A Phase I protocol (85-C-251), approved by the Clinical Research Subpanel of the Clinical Center of the NCI, investigated the use of resection and IORT in 4 patients. These initial studies in human beings were to establish the levels of toxicity at a given radiation dose, and were performed in patients with non-small cell lung cancer believed preoperatively to be Stage I1 or Stage 111. PREOPERATIVE TREATMENT. Patients were staged preoperatively with computed tomograms of the chest and abdomen, bone scan, and liver function studies. Pulmonary function tests and preoperative cardiac evaluation indicated that all 4 patients were acceptable candidates for resection and IORT. Informed consent was obtained; it carefully detailed the possible complications of pulmonary resection as well as known (from animal data, i.e., esophagitis) or theoretical complications of thoracic irradiation. The theoretical complications included pul-
Fig l . View through typical portal used for intraoperative radiation therapy of the superior mediastinum after right pneumonectomy.
monary fibrosis, transverse myelitis, pleural effusion, hemorrhage, bronchial stump dehiscence, and empyema. None of the patients had received previous treatment of the lung cancer. Baseline preoperative barium esophagography as well as fiberoptic bronchoscopy and esophagoscopy with photographic documentation of pretreatment status was performed. OPERATIVE TREATMENT. Pulmonary resection and delivery of IORT were done in the NCI Radiation Oncology Branch IORT suite. After resection, the patients received two opposing ports of electron irradiation of 2,500 cGy, one to the superior mediastinum and the other to the inferior mediastinum (Fig 1).The esophagus was manipulated out of the radiation fields without further shielding; however, it was not dissected free from behind the membranous portion of the trachea. “Through-theportal” viewing of the irradiation field was documented by video camera at the time of operation and stored on tape for future reference. The appropriate resection was performed for the location of the carcinoma, and complete nodal dissection was performed in every patient. Patients were followed with monthly FOLLOW-UP. clinic visits for documentation of toxicity. Repeat fiberoptic bronchoscopy and esophagoscopy were performed along with barium swallows every three months. Further investigations were done as dictated by patient symptomatology and the necessity for postoperative staging of disease.
Results
Animal lnvestigations CONTROL DOGS. Neither acute nor late complications were noted. Postmortem examination revealed normal healing of the bronchial stump in all dogs, and there was no evidence of esophageal complications. Histological examination of all tissues was normal.
16 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
I 4 I
I
I
I
1
1
1
1
S 1 2 3 4 5 6 7 8 9101112 T i (months) Fig 2 . Weight curves as a function of time for the animals receiving intraoperatiue radiation therapy. Arrows indicate first detection of esophagitis and arrows with circles, first detection of tracheobronchial abnormalities.
IORT DOGS EFFECTIVELY KILLED. All 3 of the dogs that had pneumonectomy and radiation treatment with 2,000 cGy were electively killed at designated intervals up to 12 months following IORT. One animal was retained for long-term follow-up. None of the animals receiving 3,000 or 4,000 cGy could be maintained longer than one year; there was 1 early death (esophagoaortic fistula) and 1early humane death (carinal necrosis with obstruction). Weight curves for the three groups of irradiated animals are depicted in Figure 2. After initial postoperative weight loss, the animals tended to regain and maintain their preoperative weight until death or inception of esophagitis. Once esophagitis was documented, the animals consistently lost weight over a two-week to twelveweek period, after which all the long-term survivors were able to maintain or surpass the weight measured before esophagitis occurred. Only the animals given a low dose (2,000 cGy) were able to gain weight despite clinical documentation of healed esophagitis. None of the animals electively killed exhibited any manifestation
of pulmonary insufficiency after pneumonectomy and IORT. Serial bronchoscopic examinations of the treated animals revealed healing of the bronchial stump without any instance of bronchial dehiscence. Late radiationinduced changes with decreased vascularity and smooth, glistening white mucosa could be appreciated in 4 of the 5 long-term survivors between 5.5 and 10 months after IORT. These changes were more pronounced and appeared earlier in the animals treated with 3,000 and 4,000 cGy (Table 1). Endoscopic verification of esophagitis was confirmed as early as 3 months in all the irradiated groups. Mucosal ulceration was less severe and there was evidence of late healing in the 2,000-cGy group. Stricture formation with circumferential ulceration was a later manifestation of progressive esophagitis in the animals given a higher dose (see Table 1). Postmortem examination and histological examination of irradiated and nonirradiated tissues in the animals electively killed disclosed a greater degree of esophageal involvement in the animals receiving a higher dose (2.8 vs 0.5 cm, mean length of affected segment), as well as a greater depth of submucosal and muscularis involvement. The severity of the radiation-induced changes was greater in the animals given a higher dose, and there was earlier documentation of these changes (Table 2). Moreover, aortic, cardiac, and contralateral lung changes were seen only with the higher doses of IORT. These changes consisted of cystic medial necrosis, frank myocardial necrosis in the left atrium, and consolidative pulmonary changes. ANIMALS WITH LIFE-THREATENING COMPLICATIONS. Two animals in the higher-dose IORT groups could not be electively killed per the prescribed schedule. One dog in the 3,000-cGy group was documented to have severe circumferential esophagitis 3 months after treatment. One week later it sustained massive exsanguination. Autopsy revealed esophagoaortic fistula with full-thickness necrosis of the esophagus within the radiation portal. The bronchial stump was intact despite histological confirmation of cartilaginous cell death. Frank necrosis of the aorta, with cellular dropout of the nuclei of the media, contributed to the development of the fistula (Fig 3). The vasa vasorum had medial sclerosis and hyaline degeneration with perivascular fibrosis. These microscopic changes in the aorta were also seen in the dogs in the 4,000-cGy group that were electively killed, yet they did not result in clinically obvious problems or affect animal survival. Bronchoscopy revealed radiation-induced changes in 1 animal in the 4,000-cGy group 5 months after treatment; at that time, the animal appeared toxic and had labored breathing. A chest roentgenogram revealed right middle lobe consolidation, and bronchoscopy demonstrated carinal degeneration with heaped-up, inflamed mucosa of the right mainstem bronchus and circumferential narrowing of the lumen by 80%. Elective
17 Pass, Sindelar, Kinsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
Table 1. Tolerance of Canine Mediastinal Viscera to lORT after Pneumonectomy ______
~
No. Electively No. Dylng of Killed for No. Alive Clinicopathological Findings Pathological Long Term Dose Level No. of Dogs Treatment (>24 mo) Bronchoscopy Esophagoscopy Treated Complications Evaluation" (GY)
0
3
0
3 (1, 3, 12)
0
20
4
0
3 (1, 3, 12)
1
30
4
1
3 (1, 3, 12)
0
40
4
1
3 (1, 3, 12)
0
No evidence of esophagitis
Normal healing of stump at 1, 3, 12 mo Normal healing of stump; tracheobronchial radiation changes 10 mo (1 dog) Normal healing of stump; tracheobronchial radiation changes at 9 mo (1 dog) Normal healing of stump; tracheobronchial radiation changes at 5 mo (1 dog) and 8 mo (1 dog); carinal necrosis at 9 mo (1 dog)
Esophagitis (ulcer) at 6 mo (2 dogs), no late stricture
Esophagitis (ulcer) at 3 mo (2 dogs) and 5 mo (1 dog); esophagoaortic fistula at 3 mo (1dog) Esophagitis (ulcer) at 3, 4, 5 mo (3 dogs); stricture at 12 mo
"Numbers in parentheses indicate month of death of each dog. IORT
=
intraoperative radiation therapy.
Table 2 . Histological Changes in Tissues Treated with Thoracic lORT and Resectiona ~
20 Gy
30 Gy
40 Gy
Time (mo)
Tiqe (mo)
Time (mo)
Organ
1
3
12
1
3
12
1
3
9
12
Trachea Bronchus (L) Bronchus (R) Esophagus Aorta Heart
0 0 0 0 0 0
0 0 0 Mild 0 0
Mild Mild Mild Severe 0 0
0 Minimal Minimal 0 0 0
0 Mqd Mild Severe Severe 0
Mild Mild Mild Severe 0 0
0 0 0 Mild Minimal 0
0 Mild Mild Mild Mild 0
Mild Mild Severe Severe Severe Severe
Mild Mild Mild Severe Severe 0
"Minimalindicates acute inflammatory changes but no loss of architecture; mild indicates sclerosis of vessels, glandular loss, and submucosal thickening; and severe indicates necrosis and ulceration, vessel obliteration, and loss of normal architecture.
IORT
=
intraoperative radiation therapy.
death at the time confirmed severe radiation-induced fibrosis, with necrotic cartilaginous bars, surrounded by a dense, chronic inflammatory infiltrate. There was no evidence of bronchial stump dehiscence.
Clinical Trial Ten patients with non-small cell carcinoma of the lung were evaluated preoperatively for possible resection and IORT. Six were excluded either because they had a poor preoperative physiological reserve (3 patients) or the cancer was unresectable at operation (3 patients). Clinical and staging characteristics of the 4 patients who underwent resection and IORT are given in Table 3. Preoperatively, all patients were believed to have either Stage I1 or Stage I11 disease.
There were no early deaths. However, 2 patients had early bronchial stump dehiscence (Patients 2 and 3). Patient 2 had an acute bronchial disruption on the tenth postoperative day and required immediate reexploration and bronchial stump closure with an intercostal muscle flap. Retrospective review of the videotape recording of the IORT portals revealed an overlap of the bronchial stump closure, which doubled the dose to 5,000 cGy. The patient remains free from disease 18 months after resection. She had only a transient episode of mild esophagitis 5 months after treatment. A postpneumonectomy empyema developed in Patient 3, and subsequently a small but symptomatic bronchopleural fistula was diagnosed within 45 days after resection and IORT. It necessitated a right-sided tho-
18 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
Fig 3. (A) Uninvolved normal aorta in an animal administered 3,000 cGy. (B)Stain in necrotic aorta from the same dog revealed dissolution of elastic fibrils and cystic medial necrosis. (van Gieson; x 10 before 15% reduction.)
Table 3. Summary of Data on Patients Having Pulmonary Resection and IORT” Patient No.
Age (yr)
1
2
Complications
Resection
Surgical Stage
Early
Late
72
Left upper lobectomy
TzNo
None
4
61 42
Right pneumonectomy Right pneumonectomy
T2N1 T3No
Stump dehiscence, d. 10 Empyema, bronchopleural fistula
4
57
Right pneumonectomy
T2NZ
Esophagitis (reversible)
Treatment-related death at 9 mo: irreversible radiation esophagitis Esophagitis (reversible) Contralateral esophagobronchial fistula; death at 9 mo after IORT None; death due to systemic disease at 8 mo after IORT
”All patients received two 25-Gy mediastinal IORT ports
IORT = inhaoperative radiation therapy.
racoplasty, which successfully closed the fistula and controlled the empyema. Repeat studies at 7 months after resection and IORT revealed esophagitis at 25 cm, which was within the superior mediastinal IORT portal. One month later, the patient was acutely ill with swallowing-induced paroxysms of coughing. Thin barium swallow revealed a fistula from the esophagus to the left (contralateral to the treatment portal) bronchus. Substernal esophagogastrostomy and esophageal exclusion were performed, and the patient was discharged from the hospital. He died within a month of Serratia pneu-
monia. Autopsy revealed marked radiation-induced changes of the left bronchus and bronchoesophageal fistula. The left upper and lower lobes revealed bacterial and fungous pneumonitis. The resection margin of the right mainstem bronchus was free from tumor, as was the remainder of the area examined. Esophagitis developed in Patient 1 6 months after left upper lobectomy and IORT. The esophagitis was irreversible and progressive despite systemic and topical antibiotics, topical and intravenous administration of analgesics, and a prolonged course of intravenous hy-
19 Pass, Sindelar, Kinsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
Fig 4. (Patient 1 .) (A)Barium swallow obtained 3 months after intraoperative radiation therapy (ZORT) revealed no esophagitis. ( B ) Barium swallow revealed localized segmental esophagitis corresponding to the treatment portal for IORT.
peralimentation (Fig 4). The patient died 9 months after resection and IORT while undergoing treatment for the intractable esophagitis. There was no clinical evidence of tumor recurrence at the time of death, and all esophageal biopsy specimens prior to death revealed radiationinduced esophagitis. No autopsy was performed. Esophagitis developed in Patient 4 three weeks after right pneumonectomy and IORT. The esophagitis responded to topical treatment and resolved within a twoweek period. Three months after resection and IORT, neurological evaluation disclosed brain metastases, for which the patient received systemic chemotherapy. He died of systemic disease 8 months after treatment. There was no clinical evidence of local recurrence at the time of death. No autopsy was performed.
Comment General Considerations The chief advantage of IORT in cancer therapy is the ability to deliver precisely directed radiotherapy to areas of malignant disease while avoiding potentially damag-
ing irradiation of surrounding normal tissues and viscera. There has been minimal investigation of the efficacy of IORT in regard to thoracic malignancies [7,8]. Previous work from this laboratory [9] included investigation of the normal tissue tolerance of intact lung and mediastinum to single-dose IORT of 2,000, 3,000, 4,000 cGy. The studies revealed that all tissues within the thorax receiving single-fraction high-dose IORT demonstrate histological evidence of radiation-induced damage, which tends in most organs to increase with increasing dose. Although 2,000 cGy was well tolerated, doses in excess of 3,000 cGy resulted in clinically significant radiation-induced injury. Before proceeding to Phase I clinical trials of the use of IORT for thoracic malignancies, a canine pneumonectomy model with IORT was investigated, and there was careful documentation of the normal tissue tolerances. As described earlier, the complications seen in the animal investigations were dose related. Esophagitis, either gross or microscopic, was noted in all dogs but 2 in the 2,000-cGy group, and a spectrum of radiationinduced changes from mucosal thickening to stricture formation was seen over the dose range and times examined. Late tracheobronchial changes with frank necrosis of cartilaginous bars and mucosal fibrosis were seen both grossly and histologically. The absence of acute or late effects on the bronchial stump was surprising, but could probably be attributed to the fact that, once the bronchus healed, an ongoing radiation-induced fibrosis occurred, thereby preventing stump “blowout.” The remainder of the tracheobronchial tree, however, specifically the distal trachea and contralateral bronchus, was affected by the progressive radiation-induced fibrosis and submucosal thickening. The radiation-induced changes noted in the aorta of the animals receiving a higher dose were previously seen in our initial set of animal investigations for mediastinal IORT, and ranged from an initial sclerosis of the vasa vasorum to fibrosis. Cystic medial degeneration of the aorta was seen in the 3,000-cGy and 4,000-cGy animals, and led to the frank aortic necrosis responsible for the esophagoaortic fistula in 1animal. This degeneration of the medial elastin fibrils dictated the use of IORT doses of less than 3,000 cGy in any human investigation. The rationale behind a Phase I trial in patients with non-small cell lung cancer was based on the radiosensitivity of squamous cell carcinoma and the theoretical ability to give regional field treatment in a dose-directed manner under direct vision. The ability to shield directly in the operating room or to manipulate vital structures out of the radiation port would theoretically decrease radiation-induced toxicity. However, the animal model at 2,000 cGy did not predict human responses to 2,500 cGy. It was indeed feasible to perform the IORT with pulmonary resection, and the use of specially designed acrylic cones with stainless steel shielding facilitated delivery. Nevertheless, immediate life-threatening bronchial stump problems occurred in 2 of the 4 patients. In 1 patient, inadvertent overlap of the IORT ports at the
20 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
level of the divided bronchus may have been instrumental in the dehiscence. The bronchial problems, however, though life-threatening, were amenable to corrective procedures, as opposed to the esophageal problems. The 2 late deaths caused by esophageal problems would not have been predicted from the animal data if doses of less than 3,000 cGy were thought to be appropriate. In both patients, the esophageal problem was located posteriorly to the carina at a point in the superior mediastinum where the esophagus had not been dissected circumferentially and removed from the radiation beam because this is the area of its apposition to the membranous trachea. To overcome this problem, lead shielding could be used over the esophagus and subcarinalregion. This, however, would probably defeat the purpose of the IORT, which is to encompass the local tissues at risk for regional disease.
Future Considerations for Thoracic lORT Initial examination of these data might lead to the conclusion that there is no place for IORT in the management of locally advanced lung cancer. Although this clinical study does show major toxicity, it is possible that the dose used is higher than necessary to still be tumoriudal and that less toxicity could have been realized with 1,500 to 2,000 cGy. More importantly, these investigations stress that further studies of thoracic IORT should be conducted only at centers strictly devoted to Phase I clinical research, and require close collaboration between the radiation oncologist and the surgeon. By limiting these investiga-
tions to those institutions with special expertise in this technique, the economic impact of premature IORT use will be avoided and patient safety will be optimized. Only through closely controlled studies will the risk/ benefit ratio of this experimental innovation be defined.
References 1. Tepper J, Sindelar W: Summary of the workshop on intraoperative radiation therapy. Cancer Treat Rep 65:911, 1981 2. Abe M, Yabumoto E, Takahashi M, et al: Intraoperative radiotherapy of gastric cancer. Cancer 342034, 1974 3. Wood WC, Shipley WU, Gunderson LL, et al: Intraoperative irradiation for unresectable pancreatic carcinoma. Cancer 491272, 1982 4. Cohen AM, Gunderson LL, Wood WC: Intraoperative electron beam radiation therapy boost in the treatment of recurrent rectal cancer. Dis Colon Rectum 23453, 1980 5. Sindelar WF, Kinsella TJ, Tepper J, et al: Experimental and clinical studies with intraoperative radiotherapy. Surg Gynecol Obstet 15725, 1983 6. Fraass BA, Miller BW, Kinsella TJ, et al: Intraoperative radiation therapy at the National Cancer Institute:technical innovations and dosimetry. Int J Radiat Oncol Biol Phys 11:1299, 1985 7. Goldson AL: Past, present, and prospects of intraoperative radiotherapy (IOR). Semin Oncol 8:59, 1981 8. Abe M, Yabumoto E, Nishidai T Trials of new forms of radiotherapy for locally advanced bronchogenic carcinoma. Strahlentherapie 153:149, 1977 9. Barnes M, Pass HI, DeLuca A-M, et al: Response of the mediastinal and thoracic viscera of the dog to intraoperative radiation therapy (IORT). Int J Radiat Oncol Biol Phys 13:371, 1987
Notice from the American Board of Thoracic Surgery The American Board of Thoracic Surgery began its recertification process in 1984. Diplomates interested in participating in this examination should maintain a documented list of the cardiothoracic operations they performed during the year prior to application for recertification. They should also keep a record of their attendance at thoracic surgical meetings, and other continuing medical education activities pertaining to thoracic surgery and thoracic disease, for the two years prior to application. A minimum of 100 hours of approved CME activity is required. In place of a cognitive examination, candidates for recertification will be required to complete both the general thoracic and cardiac portions of the SESATS I11 syllabus (Self-EducatiodSelf-Assessment in Thoracic Surgery). It is not necessary for candidates to purchase
SESATS 111booklets prior to applying for recertification. SESATS I11 booklets will be forwarded to candidates after their applications have been accepted. Diplomates whose 10-year certificates will expire in 1989 may begin the recertification process in 1987. This new certificate will be dated 10 years from the time of expiration of the original certificate. Recertification is also open to any Diplomate with an unlimited certificate and will in no way affect the validity of the original certificate. The deadline for submission of applications was July 1, 1987. A recertification brochure outlining the rules and requirements for recertification in thoracic surgery is available upon request from the American Board of Thoracic Surgery, One American Plaza, Suite 803, Evanston, IL 60201.
14 Ann Thorac Surg 44:14-20,July 1987
excluding radiosensitive normal tissues from the radiation field by surgical manipulation or by custom lead shielding placed intraoperatively. Clinical trials of IORT in the treatment of gastric [2], pancreatic [3], and rectal carcinomas [4] as well as retroperitoneal sarcomas [5] have been performed at various centers including the National Cancer Institute (NCI). The use of IORT in the management of thoracic neoplasms has yet to be defined. Moreover, the dose tolerance of normal thoracic viscera has not been investigated with IORT in a large animal model to help establish safe guidelines for treatment portals as well as dose. This report describes the effects of thoracic IORT in a canine model of pulmonary resection at various radiation doses as a function of time, and serves as the basis for clinical investigation of the efficacy of IORT as an adjunct to resection in 4 patients with non-small cell carcinoma of the lung.
Material and Methods
Animal Investigations Fifteen purebred male and female adult American foxhounds ranging in age from 9 to 18 months and weighing 24 to 41 kg were maintained in sheltered runs and fed dog chow and water ad libitum. All animals received humane care in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 8023, revised 1978). Large animals were used to allow for adequate surgical exposure and placement of IORT cones for field size comparable to what might be used in a human research setting. SURGERY AND DELIVERY OF INTRAOPERATIVE RADIATION
A left pneumonectomy was performed with staple closure of the left main bronchus. In this study, a single IORT field was treated using a cylindrical acrylic treatment cone with a 5-cm inner diameter. The treatment cone was sheathed in a stainless steel shield to decrease lateral electron scatter beyond the area of interest [6]. This field encompassed the bronchial stump, pulmonary artery and vein, esophagus, aorta, pericardium, left atrium, phrenic nerve, and a portion of the left ventricle. At the completion of treatment, the portal perimeter was delineated with surgical clips.
THERAPY.
15 Pass, Sindelar, Kmsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
IORT. Four dogs were treated at each of three dose levels of IORT: 2,000 cGy, 3,000 cGy, and 4,000 cGy. Radiation treatment was performed using a Scanitronix Microtron located in the NCI Radiation Oncology Branch IORT suite. The microtron has a full 360-degree rotational gantry, and beam entry angles were determined at the time of IORT cone placement. A 13-meV electron beam was used at a dose rate of 800 to 900 cGy per minute, and doses were quoted at the 90% isodose line. Three additional dogs underwent pneumonectomy and placement of the IORT acrylic cone but were not given treatment; they served as controls. FOLLOW-UP STUDIES. The dogs were monitored by weekly body weights and monthly chest roentgenograms. Contrast esophagograms were performed before the animals were killed. Fiberoptic bronchoscopy and esophagoscopy were performed monthly to document gross pathological changes at the earliest possible time. One animal from each treatment group as well as 1 sham-irradiated control was to be killed electively at one, three, and twelve months following IORT. The remaining animal from each treatment group would be maintained for up to five years for documentation of late IORT effects. All dogs that died or were killed were subjected to a complete autopsy including gross examination of thoracic and abdominal viscera and microscopic examination of the thoracic viscera and tissues. Particular attention was paid to segments of tissues included within the radiation portals. Sections of both irradiated and grossly normal tissues were submitted for histological review, including irradiated bronchial stump, trachea, right mainstem bronchus, right lung, left atrium and ventricle, aorta, phrenic nerve, esophagus, and ligated irradiated stump of the pulmonary artery and vein. All tissues were prepared and stained with hematoxylin and eosin, and special stains including van Gieson’s for connective tissue, Verhoeff‘s plus van Gieson for elastin, and Mallory’s trichrome were used to aid in the examination of connective tissue changes.
Human Investigations A Phase I protocol (85-C-251), approved by the Clinical Research Subpanel of the Clinical Center of the NCI, investigated the use of resection and IORT in 4 patients. These initial studies in human beings were to establish the levels of toxicity at a given radiation dose, and were performed in patients with non-small cell lung cancer believed preoperatively to be Stage I1 or Stage 111. PREOPERATIVE TREATMENT. Patients were staged preoperatively with computed tomograms of the chest and abdomen, bone scan, and liver function studies. Pulmonary function tests and preoperative cardiac evaluation indicated that all 4 patients were acceptable candidates for resection and IORT. Informed consent was obtained; it carefully detailed the possible complications of pulmonary resection as well as known (from animal data, i.e., esophagitis) or theoretical complications of thoracic irradiation. The theoretical complications included pul-
Fig l . View through typical portal used for intraoperative radiation therapy of the superior mediastinum after right pneumonectomy.
monary fibrosis, transverse myelitis, pleural effusion, hemorrhage, bronchial stump dehiscence, and empyema. None of the patients had received previous treatment of the lung cancer. Baseline preoperative barium esophagography as well as fiberoptic bronchoscopy and esophagoscopy with photographic documentation of pretreatment status was performed. OPERATIVE TREATMENT. Pulmonary resection and delivery of IORT were done in the NCI Radiation Oncology Branch IORT suite. After resection, the patients received two opposing ports of electron irradiation of 2,500 cGy, one to the superior mediastinum and the other to the inferior mediastinum (Fig 1).The esophagus was manipulated out of the radiation fields without further shielding; however, it was not dissected free from behind the membranous portion of the trachea. “Through-theportal” viewing of the irradiation field was documented by video camera at the time of operation and stored on tape for future reference. The appropriate resection was performed for the location of the carcinoma, and complete nodal dissection was performed in every patient. Patients were followed with monthly FOLLOW-UP. clinic visits for documentation of toxicity. Repeat fiberoptic bronchoscopy and esophagoscopy were performed along with barium swallows every three months. Further investigations were done as dictated by patient symptomatology and the necessity for postoperative staging of disease.
Results
Animal lnvestigations CONTROL DOGS. Neither acute nor late complications were noted. Postmortem examination revealed normal healing of the bronchial stump in all dogs, and there was no evidence of esophageal complications. Histological examination of all tissues was normal.
16 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
I 4 I
I
I
I
1
1
1
1
S 1 2 3 4 5 6 7 8 9101112 T i (months) Fig 2 . Weight curves as a function of time for the animals receiving intraoperatiue radiation therapy. Arrows indicate first detection of esophagitis and arrows with circles, first detection of tracheobronchial abnormalities.
IORT DOGS EFFECTIVELY KILLED. All 3 of the dogs that had pneumonectomy and radiation treatment with 2,000 cGy were electively killed at designated intervals up to 12 months following IORT. One animal was retained for long-term follow-up. None of the animals receiving 3,000 or 4,000 cGy could be maintained longer than one year; there was 1 early death (esophagoaortic fistula) and 1early humane death (carinal necrosis with obstruction). Weight curves for the three groups of irradiated animals are depicted in Figure 2. After initial postoperative weight loss, the animals tended to regain and maintain their preoperative weight until death or inception of esophagitis. Once esophagitis was documented, the animals consistently lost weight over a two-week to twelveweek period, after which all the long-term survivors were able to maintain or surpass the weight measured before esophagitis occurred. Only the animals given a low dose (2,000 cGy) were able to gain weight despite clinical documentation of healed esophagitis. None of the animals electively killed exhibited any manifestation
of pulmonary insufficiency after pneumonectomy and IORT. Serial bronchoscopic examinations of the treated animals revealed healing of the bronchial stump without any instance of bronchial dehiscence. Late radiationinduced changes with decreased vascularity and smooth, glistening white mucosa could be appreciated in 4 of the 5 long-term survivors between 5.5 and 10 months after IORT. These changes were more pronounced and appeared earlier in the animals treated with 3,000 and 4,000 cGy (Table 1). Endoscopic verification of esophagitis was confirmed as early as 3 months in all the irradiated groups. Mucosal ulceration was less severe and there was evidence of late healing in the 2,000-cGy group. Stricture formation with circumferential ulceration was a later manifestation of progressive esophagitis in the animals given a higher dose (see Table 1). Postmortem examination and histological examination of irradiated and nonirradiated tissues in the animals electively killed disclosed a greater degree of esophageal involvement in the animals receiving a higher dose (2.8 vs 0.5 cm, mean length of affected segment), as well as a greater depth of submucosal and muscularis involvement. The severity of the radiation-induced changes was greater in the animals given a higher dose, and there was earlier documentation of these changes (Table 2). Moreover, aortic, cardiac, and contralateral lung changes were seen only with the higher doses of IORT. These changes consisted of cystic medial necrosis, frank myocardial necrosis in the left atrium, and consolidative pulmonary changes. ANIMALS WITH LIFE-THREATENING COMPLICATIONS. Two animals in the higher-dose IORT groups could not be electively killed per the prescribed schedule. One dog in the 3,000-cGy group was documented to have severe circumferential esophagitis 3 months after treatment. One week later it sustained massive exsanguination. Autopsy revealed esophagoaortic fistula with full-thickness necrosis of the esophagus within the radiation portal. The bronchial stump was intact despite histological confirmation of cartilaginous cell death. Frank necrosis of the aorta, with cellular dropout of the nuclei of the media, contributed to the development of the fistula (Fig 3). The vasa vasorum had medial sclerosis and hyaline degeneration with perivascular fibrosis. These microscopic changes in the aorta were also seen in the dogs in the 4,000-cGy group that were electively killed, yet they did not result in clinically obvious problems or affect animal survival. Bronchoscopy revealed radiation-induced changes in 1 animal in the 4,000-cGy group 5 months after treatment; at that time, the animal appeared toxic and had labored breathing. A chest roentgenogram revealed right middle lobe consolidation, and bronchoscopy demonstrated carinal degeneration with heaped-up, inflamed mucosa of the right mainstem bronchus and circumferential narrowing of the lumen by 80%. Elective
17 Pass, Sindelar, Kinsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
Table 1. Tolerance of Canine Mediastinal Viscera to lORT after Pneumonectomy ______
~
No. Electively No. Dylng of Killed for No. Alive Clinicopathological Findings Pathological Long Term Dose Level No. of Dogs Treatment (>24 mo) Bronchoscopy Esophagoscopy Treated Complications Evaluation" (GY)
0
3
0
3 (1, 3, 12)
0
20
4
0
3 (1, 3, 12)
1
30
4
1
3 (1, 3, 12)
0
40
4
1
3 (1, 3, 12)
0
No evidence of esophagitis
Normal healing of stump at 1, 3, 12 mo Normal healing of stump; tracheobronchial radiation changes 10 mo (1 dog) Normal healing of stump; tracheobronchial radiation changes at 9 mo (1 dog) Normal healing of stump; tracheobronchial radiation changes at 5 mo (1 dog) and 8 mo (1 dog); carinal necrosis at 9 mo (1 dog)
Esophagitis (ulcer) at 6 mo (2 dogs), no late stricture
Esophagitis (ulcer) at 3 mo (2 dogs) and 5 mo (1 dog); esophagoaortic fistula at 3 mo (1dog) Esophagitis (ulcer) at 3, 4, 5 mo (3 dogs); stricture at 12 mo
"Numbers in parentheses indicate month of death of each dog. IORT
=
intraoperative radiation therapy.
Table 2 . Histological Changes in Tissues Treated with Thoracic lORT and Resectiona ~
20 Gy
30 Gy
40 Gy
Time (mo)
Tiqe (mo)
Time (mo)
Organ
1
3
12
1
3
12
1
3
9
12
Trachea Bronchus (L) Bronchus (R) Esophagus Aorta Heart
0 0 0 0 0 0
0 0 0 Mild 0 0
Mild Mild Mild Severe 0 0
0 Minimal Minimal 0 0 0
0 Mqd Mild Severe Severe 0
Mild Mild Mild Severe 0 0
0 0 0 Mild Minimal 0
0 Mild Mild Mild Mild 0
Mild Mild Severe Severe Severe Severe
Mild Mild Mild Severe Severe 0
"Minimalindicates acute inflammatory changes but no loss of architecture; mild indicates sclerosis of vessels, glandular loss, and submucosal thickening; and severe indicates necrosis and ulceration, vessel obliteration, and loss of normal architecture.
IORT
=
intraoperative radiation therapy.
death at the time confirmed severe radiation-induced fibrosis, with necrotic cartilaginous bars, surrounded by a dense, chronic inflammatory infiltrate. There was no evidence of bronchial stump dehiscence.
Clinical Trial Ten patients with non-small cell carcinoma of the lung were evaluated preoperatively for possible resection and IORT. Six were excluded either because they had a poor preoperative physiological reserve (3 patients) or the cancer was unresectable at operation (3 patients). Clinical and staging characteristics of the 4 patients who underwent resection and IORT are given in Table 3. Preoperatively, all patients were believed to have either Stage I1 or Stage I11 disease.
There were no early deaths. However, 2 patients had early bronchial stump dehiscence (Patients 2 and 3). Patient 2 had an acute bronchial disruption on the tenth postoperative day and required immediate reexploration and bronchial stump closure with an intercostal muscle flap. Retrospective review of the videotape recording of the IORT portals revealed an overlap of the bronchial stump closure, which doubled the dose to 5,000 cGy. The patient remains free from disease 18 months after resection. She had only a transient episode of mild esophagitis 5 months after treatment. A postpneumonectomy empyema developed in Patient 3, and subsequently a small but symptomatic bronchopleural fistula was diagnosed within 45 days after resection and IORT. It necessitated a right-sided tho-
18 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
Fig 3. (A) Uninvolved normal aorta in an animal administered 3,000 cGy. (B)Stain in necrotic aorta from the same dog revealed dissolution of elastic fibrils and cystic medial necrosis. (van Gieson; x 10 before 15% reduction.)
Table 3. Summary of Data on Patients Having Pulmonary Resection and IORT” Patient No.
Age (yr)
1
2
Complications
Resection
Surgical Stage
Early
Late
72
Left upper lobectomy
TzNo
None
4
61 42
Right pneumonectomy Right pneumonectomy
T2N1 T3No
Stump dehiscence, d. 10 Empyema, bronchopleural fistula
4
57
Right pneumonectomy
T2NZ
Esophagitis (reversible)
Treatment-related death at 9 mo: irreversible radiation esophagitis Esophagitis (reversible) Contralateral esophagobronchial fistula; death at 9 mo after IORT None; death due to systemic disease at 8 mo after IORT
”All patients received two 25-Gy mediastinal IORT ports
IORT = inhaoperative radiation therapy.
racoplasty, which successfully closed the fistula and controlled the empyema. Repeat studies at 7 months after resection and IORT revealed esophagitis at 25 cm, which was within the superior mediastinal IORT portal. One month later, the patient was acutely ill with swallowing-induced paroxysms of coughing. Thin barium swallow revealed a fistula from the esophagus to the left (contralateral to the treatment portal) bronchus. Substernal esophagogastrostomy and esophageal exclusion were performed, and the patient was discharged from the hospital. He died within a month of Serratia pneu-
monia. Autopsy revealed marked radiation-induced changes of the left bronchus and bronchoesophageal fistula. The left upper and lower lobes revealed bacterial and fungous pneumonitis. The resection margin of the right mainstem bronchus was free from tumor, as was the remainder of the area examined. Esophagitis developed in Patient 1 6 months after left upper lobectomy and IORT. The esophagitis was irreversible and progressive despite systemic and topical antibiotics, topical and intravenous administration of analgesics, and a prolonged course of intravenous hy-
19 Pass, Sindelar, Kinsella, et al: Intraoperative Radiation Therapy after Pneumonectomy
Fig 4. (Patient 1 .) (A)Barium swallow obtained 3 months after intraoperative radiation therapy (ZORT) revealed no esophagitis. ( B ) Barium swallow revealed localized segmental esophagitis corresponding to the treatment portal for IORT.
peralimentation (Fig 4). The patient died 9 months after resection and IORT while undergoing treatment for the intractable esophagitis. There was no clinical evidence of tumor recurrence at the time of death, and all esophageal biopsy specimens prior to death revealed radiationinduced esophagitis. No autopsy was performed. Esophagitis developed in Patient 4 three weeks after right pneumonectomy and IORT. The esophagitis responded to topical treatment and resolved within a twoweek period. Three months after resection and IORT, neurological evaluation disclosed brain metastases, for which the patient received systemic chemotherapy. He died of systemic disease 8 months after treatment. There was no clinical evidence of local recurrence at the time of death. No autopsy was performed.
Comment General Considerations The chief advantage of IORT in cancer therapy is the ability to deliver precisely directed radiotherapy to areas of malignant disease while avoiding potentially damag-
ing irradiation of surrounding normal tissues and viscera. There has been minimal investigation of the efficacy of IORT in regard to thoracic malignancies [7,8]. Previous work from this laboratory [9] included investigation of the normal tissue tolerance of intact lung and mediastinum to single-dose IORT of 2,000, 3,000, 4,000 cGy. The studies revealed that all tissues within the thorax receiving single-fraction high-dose IORT demonstrate histological evidence of radiation-induced damage, which tends in most organs to increase with increasing dose. Although 2,000 cGy was well tolerated, doses in excess of 3,000 cGy resulted in clinically significant radiation-induced injury. Before proceeding to Phase I clinical trials of the use of IORT for thoracic malignancies, a canine pneumonectomy model with IORT was investigated, and there was careful documentation of the normal tissue tolerances. As described earlier, the complications seen in the animal investigations were dose related. Esophagitis, either gross or microscopic, was noted in all dogs but 2 in the 2,000-cGy group, and a spectrum of radiationinduced changes from mucosal thickening to stricture formation was seen over the dose range and times examined. Late tracheobronchial changes with frank necrosis of cartilaginous bars and mucosal fibrosis were seen both grossly and histologically. The absence of acute or late effects on the bronchial stump was surprising, but could probably be attributed to the fact that, once the bronchus healed, an ongoing radiation-induced fibrosis occurred, thereby preventing stump “blowout.” The remainder of the tracheobronchial tree, however, specifically the distal trachea and contralateral bronchus, was affected by the progressive radiation-induced fibrosis and submucosal thickening. The radiation-induced changes noted in the aorta of the animals receiving a higher dose were previously seen in our initial set of animal investigations for mediastinal IORT, and ranged from an initial sclerosis of the vasa vasorum to fibrosis. Cystic medial degeneration of the aorta was seen in the 3,000-cGy and 4,000-cGy animals, and led to the frank aortic necrosis responsible for the esophagoaortic fistula in 1animal. This degeneration of the medial elastin fibrils dictated the use of IORT doses of less than 3,000 cGy in any human investigation. The rationale behind a Phase I trial in patients with non-small cell lung cancer was based on the radiosensitivity of squamous cell carcinoma and the theoretical ability to give regional field treatment in a dose-directed manner under direct vision. The ability to shield directly in the operating room or to manipulate vital structures out of the radiation port would theoretically decrease radiation-induced toxicity. However, the animal model at 2,000 cGy did not predict human responses to 2,500 cGy. It was indeed feasible to perform the IORT with pulmonary resection, and the use of specially designed acrylic cones with stainless steel shielding facilitated delivery. Nevertheless, immediate life-threatening bronchial stump problems occurred in 2 of the 4 patients. In 1 patient, inadvertent overlap of the IORT ports at the
20 The Annals of Thoracic Surgery Vol 44 No 1 July 1987
level of the divided bronchus may have been instrumental in the dehiscence. The bronchial problems, however, though life-threatening, were amenable to corrective procedures, as opposed to the esophageal problems. The 2 late deaths caused by esophageal problems would not have been predicted from the animal data if doses of less than 3,000 cGy were thought to be appropriate. In both patients, the esophageal problem was located posteriorly to the carina at a point in the superior mediastinum where the esophagus had not been dissected circumferentially and removed from the radiation beam because this is the area of its apposition to the membranous trachea. To overcome this problem, lead shielding could be used over the esophagus and subcarinalregion. This, however, would probably defeat the purpose of the IORT, which is to encompass the local tissues at risk for regional disease.
Future Considerations for Thoracic lORT Initial examination of these data might lead to the conclusion that there is no place for IORT in the management of locally advanced lung cancer. Although this clinical study does show major toxicity, it is possible that the dose used is higher than necessary to still be tumoriudal and that less toxicity could have been realized with 1,500 to 2,000 cGy. More importantly, these investigations stress that further studies of thoracic IORT should be conducted only at centers strictly devoted to Phase I clinical research, and require close collaboration between the radiation oncologist and the surgeon. By limiting these investiga-
tions to those institutions with special expertise in this technique, the economic impact of premature IORT use will be avoided and patient safety will be optimized. Only through closely controlled studies will the risk/ benefit ratio of this experimental innovation be defined.
References 1. Tepper J, Sindelar W: Summary of the workshop on intraoperative radiation therapy. Cancer Treat Rep 65:911, 1981 2. Abe M, Yabumoto E, Takahashi M, et al: Intraoperative radiotherapy of gastric cancer. Cancer 342034, 1974 3. Wood WC, Shipley WU, Gunderson LL, et al: Intraoperative irradiation for unresectable pancreatic carcinoma. Cancer 491272, 1982 4. Cohen AM, Gunderson LL, Wood WC: Intraoperative electron beam radiation therapy boost in the treatment of recurrent rectal cancer. Dis Colon Rectum 23453, 1980 5. Sindelar WF, Kinsella TJ, Tepper J, et al: Experimental and clinical studies with intraoperative radiotherapy. Surg Gynecol Obstet 15725, 1983 6. Fraass BA, Miller BW, Kinsella TJ, et al: Intraoperative radiation therapy at the National Cancer Institute:technical innovations and dosimetry. Int J Radiat Oncol Biol Phys 11:1299, 1985 7. Goldson AL: Past, present, and prospects of intraoperative radiotherapy (IOR). Semin Oncol 8:59, 1981 8. Abe M, Yabumoto E, Nishidai T Trials of new forms of radiotherapy for locally advanced bronchogenic carcinoma. Strahlentherapie 153:149, 1977 9. Barnes M, Pass HI, DeLuca A-M, et al: Response of the mediastinal and thoracic viscera of the dog to intraoperative radiation therapy (IORT). Int J Radiat Oncol Biol Phys 13:371, 1987
Notice from the American Board of Thoracic Surgery The American Board of Thoracic Surgery began its recertification process in 1984. Diplomates interested in participating in this examination should maintain a documented list of the cardiothoracic operations they performed during the year prior to application for recertification. They should also keep a record of their attendance at thoracic surgical meetings, and other continuing medical education activities pertaining to thoracic surgery and thoracic disease, for the two years prior to application. A minimum of 100 hours of approved CME activity is required. In place of a cognitive examination, candidates for recertification will be required to complete both the general thoracic and cardiac portions of the SESATS I11 syllabus (Self-EducatiodSelf-Assessment in Thoracic Surgery). It is not necessary for candidates to purchase
SESATS 111booklets prior to applying for recertification. SESATS I11 booklets will be forwarded to candidates after their applications have been accepted. Diplomates whose 10-year certificates will expire in 1989 may begin the recertification process in 1987. This new certificate will be dated 10 years from the time of expiration of the original certificate. Recertification is also open to any Diplomate with an unlimited certificate and will in no way affect the validity of the original certificate. The deadline for submission of applications was July 1, 1987. A recertification brochure outlining the rules and requirements for recertification in thoracic surgery is available upon request from the American Board of Thoracic Surgery, One American Plaza, Suite 803, Evanston, IL 60201.