Influence of Simultaneous Neoadjuvant Radiotherapy ... - ATS Journals

2 downloads 0 Views 187KB Size Report
Neoadjuvant therapy did not induce radiation pneumonitis or changes in lung ...... JB, Johnson DH, Turrisi AT, editors. Lung cancer: principles and practice.
Influence of Simultaneous Neoadjuvant Radiotherapy and Chemotherapy on Bronchoscopic Findings and Lung Function in Patients with Locally Advanced Proximal Esophageal Cancer MARTIN RIEDEL, HUBERT J. STEIN, LEONARD MOUNYAM, FRANK ZIMMERMANN, ULRICH FINK, and JÖRG R. SIEWERT Chirurgische Klinik und Poliklinik, Klinik für Strahlentherapie, and Pneumologie der I. Medizinischen Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität, München, Germany

To assess the bronchoscopic and lung function changes induced by preoperative radiochemotherapy (30 Gy radiation and 5-fluorouracil) in patients with proximal esophageal cancer, we prospectively compared the findings in 77 consecutive patients before and after the therapy. All patients completed the radiochemotherapy protocol; toxicity was minimal. Sixty-four patients underwent surgery, 48 had total gross removal of disease, and six had a complete histologic response. Of the 13 patients who developed apparent direct macroscopic signs of tumor invasion into the airways during therapy, histologic proof of cancer was obtained in only one of the abnormalities. Bronchoscopy was falsely negative in six patients in whom airway invasion of the cancer was found at surgery. Neoadjuvant therapy led to no systematic changes in the appearance of the uninvolved tracheal mucosa; microscopically, an increase in postinflammatory changes, hyperplasia, and metaplasia was found. There was no significant change in the values of lung function parameters after the therapy. No patient developed symptoms suggestive of radiation-induced lung changes, although in one of them, subtle radiologic features consistent with radiation pneumonitis were found. No patient died of postoperative pulmonary complications. The interpretation of bronchoscopy in the assessment of airway invasion of esophageal cancer after radiochemotherapy is more difficult than at baseline staging; the positive predictive value of macroscopic abnormalities without microscopic proof of cancer is low, and even with extensive sampling for histology and cytology, the procedure was falsely negative in 9.4%. Neoadjuvant therapy did not induce radiation pneumonitis or changes in lung function that could be of concern at the following operation.

Esophagectomy is the standard treatment in patients with resectable esophageal cancer but is associated with major morbidity and mortality. Neoadjuvant (“induction”) radiotherapy and chemotherapy given before surgery may eliminate micrometastases, induce downstaging of tumor, render the tumor resectable, and ultimately improve survival (1–3). However, this approach has brought to light new problems such as patient selection and preoperative strategy. Compared with primary resection, a multimodal approach does not result in a survival benefit in patients with loco-regional, i.e., potentially resectable, tumors. In contrast, in patients with locally advanced tumors, i.e., tumors for which complete removal with primary surgery appears unlikely, neoadjuvant therapy increases the chance for subsequent complete tumor resection. Compared with preoperative chemotherapy alone, combined

radiochemotherapy increases the rate of response, particularly in squamous cell esophageal cancer, but it may increase postoperative morbidity and mortality (4–11), possibly because of radiation-induced pulmonary problems. Some patients may develop an esophagotracheal fistula while receiving neoadjuvant treatment (12). When external radiation is contemplated for the treatment of esophageal cancer, it is practically impossible to prevent normal airways and lung parenchyma closely adjacent to the radiated area from some exposure to radiation. Acute pneumonitis develops in 5 to 15% of patients a few weeks to months after radiation (13–16). The radiation-induced damage of the pulmonary parenchyma could make the lung more susceptible to postoperative problems such as acute respiratory distress syndrome (10). Alterations within airways after radiation may be due to superimposed bacterial infection, because the bronchial epithelium is relatively radioresistant. Nevertheless, focal necrosis of bronchial mucosa, increased mucus production, shedding of ciliated epithelial cells, and squamous metaplasia have been attributed to direct radiation effect (14, 16). Telangiectasia in the airways within the irradiated field occasionally gives rise to hemoptysis. Many cytotoxic agents have been reported to cause interstitial pneumonitis, usually between 3 wk and 3 mo after the treatment (15). The effects of radiation are increased by concurrent treatment with some agents, e.g., 5-fluorouracil (5-FU) (17, 18). The possible etiologic role of 5-FU in the development of interstitial pneumonitis is, however, questionable. Bronchoscopy and lung function tests are important in the assessment of patients with proximal esophageal cancer (i.e., located at or above the level of the tracheal bifurcation), and in evaluating their response to neoadjuvant therapy. The aim of this study was to evaluate prospectively the effects of neoadjuvant radiochemotherapy of proximal esophageal cancer on the bronchoscopic findings, chest radiographs and computed tomography (CT) scans, and lung function tests, and to assess the accuracy of bronchoscopy in the diagnosis of airway invasion after radiochemotherapy. In order to minimize the radiotherapy-related increase in postoperative morbidity and mortality (4) and to permit an outpatient treatment during the preoperative period, we choose a rather mild neoadjuvant regimen with only 30 Gy radiation and single drug therapy (5-FU).

METHODS Patient Population and Initial Staging (Received in original form March 20, 2000 and in revised form June 13, 2000) Correspondence and requests for reprints should be addressed to Martin Riedel, M.D., FESC, 1. Medizinische Klinik und Poliklinik, Klinikum rechts der Isar, Ismaninger Strasse 22, D-81675 München, Germany. E-mail: [email protected] Am J Respir Crit Care Med Vol 162. pp 1741–1746, 2000 Internet address: www.atsjournals.org

Seventy-seven consecutive patients with newly diagnosed locally advanced squamous cell carcinoma of the proximal esophagus who received neoadjuvant radiochemotherapy were included in the study. They underwent initial staging, including contrast swallow esophagography, posteroanterior and lateral chest radiography, spiral CT thoracoabdominal scanning with intravenous and oral contrast enhance-

1742

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

ment, percutaneous sonography, esophageal endoscopic ultrasound (EUS), pulmonary function tests, and bronchoscopy. The tumor was considered proximal if its upper margin was at or above the level of the tracheal bifurcation on radiographic studies. T and N categories were determined by EUS (19, 20).

Bronchoscopy Fibreoptic bronchoscopy was performed as previously reported (21). First, 20 ml of physiological saline were instilled separately into the left and right main bronchus, and washings for cytologic examination were obtained by suction from all lung lobes. The complete tracheobronchial tree was examined. All direct tumor signs (exophytic intraluminal growth or infiltration of the airway wall) and indirect signs (localized distortion or compression of normal structures, altered structure of the mucosa, increased vascularity, focal thickening, protrusion at the posterior wall of the trachea or a major bronchus, widened bifurcation, rigid and fixed tracheobronchial structures at breathing or coughing maneuvers) were recorded, and three to five biopsies, as well as a brush cytology sample, were taken from these areas. Additionally, any endobronchial pathology such as mucosal atrophy, airway instability, or hypersecretion of mucus were noted. Signs of atrophy due to chronic inflammation (dilated ducts of mucosal glands, transverse ridges, longitudinal light bands, and sharp carinas) were semiquantitatively described as slight, moderate, or severe. If no abnormalities were seen at bronchoscopy, brush cytology and three to four biopsies were routinely taken from the pars membranacea of the right main bronchus; left main bronchus; and the distal, medial and proximal third of the trachea. The quality of biopsies was considered optimal when at least two pieces larger than 2 mm were obtained from the region of interest (81.5% of cases), and adequate when at least two specimens were larger than 1 mm (17.2% of cases); 1.3% of bioptic samples were considered inadequate and discarded from the analysis. The brush and washings samples for cytology were considered adequate when at least four cell-rich slides, consisting predominantly of columnar epithelial cells, could be analyzed microscopically.

Chest Radiology and Pulmonary Function Tests Plain radiographs and CT scans were reviewed at the time of acquisition by a radiologist, without prior knowledge of each patient’s clinical status or outcome. However, in assessing the postradiochemotherapy preoperative radiologic findings, the observers were provided with all the previous films. Any new interstitial or alveolar haziness or infiltrate in the absence of another evident cause was considered indicative of radiation pneumonitis (14). Lung volume and airway resistance (Raw) measurements were carried out in a constant volume body plethysmograph. Maximally forced expiratory parameters were derived from the best of three adequately performed maneuvers, registered as flow-volume loops. Results were expressed as percentage of predicted, calculated from the reference values (22). Raw was expressed as specific resistance (sRaw ⫽ Raw ⫻ volume of thoracic gas). The lung transfer factor for carbon monoxide (TLCO) was determined in duplicate by the single-breath method (23). TLCO was normalized to a standard hemoglobin value and expressed in percent predicted and in absolute values corrected for body surface. Arterialized blood sampling from the earlobe was performed with the subjects in the sitting position. All testing was done in room air at an altitude of 540 m above sea. Duplicate determinations were required to check within 1 mm Hg for PaO2 and PaCO2.

Treatment After the initial staging, potentially operable patients with nonmetastatic locally advanced disease (defined as T3 or T4) received neoadjuvant therapy. Patients with histologic proof of tumor invasion into the airways on bronchoscopy were excluded from entry into the neoadjuvant protocol. Exclusion criteria further included age older than 70 yr, compromised general status, a white blood cell count of less than 4 ⫻ 109/L, a platelet count ⬍ 10 ⫻ 109/L, serum creatinine ⬎ 1.5 mg/dl, prior or concurrent other malignancy, and prior chest irradiation or chemotherapy. Patients received 5-FU intravenously at a dose of 300 mg/m2/d, infused over 24 h, for 3 wk. Radiotherapy was administered over a 3-wk

VOL 162

2000

period, beginning concurrently with chemotherapy. A total dose of 30 Gy in 2 Gy daily fractions, 5 d a week, was received by each patient. The target volume consisted of the primary and a safety margin of 4 cm in craniocaudal and 2 cm in lateral direction beyond the visible gross tumor volume, as defined by the endoscopic and CT examinations. Opposite anteroposterior mediastinal portals were used to obtain an optimal sparing of lung tissue. The locoregional lymphatic drainage, even with enlarged lymph nodes, was not included into the target volume. After completion of the radiochemotherapy, the patients underwent restaging in exactly the same manner as the initial staging. The interval between the initial and the restaging bronchoscopy and lung function testing was 69.5 ⫾ 24.4 d (range, 41 to 174; median, 63). Surgery was undertaken in 64 patients 12.4 ⫾ 9.0 days (range, 1 to 44 d) after restaging. The patients had routine preoperative physiotherapy and inclusive incentive spirometry for 3 d before the operation. Subtotal transthoracic en bloc esophagectomy with extended two-field lymphadenectomy was the procedure of choice. Tumor extent and tracheobronchial invasion were assessed intraoperatively and by histopathologic examination of the resection specimens. Pathologic stage was determined on the basis of the UICC TNM classification guidelines (19). Operated patients were seen every 3 mo for the first year and at 6-mo intervals thereafter. Episodes suggestive of cancer invasion into the airways were specifically looked for. The fate of all patients was ascertained by October 31, 1999, through telephone calls with the patients, their relatives, or their primary physicians.

Statistical Analysis Descriptive information is summarized as mean ⫾ SD for continuous variables and as frequencies or percentages for categorical variables. Results before and after treatment, on the same patients, were compared using Student’s two-sided paired t test and Wilcoxon’s rank test for paired data. All statistical tests were performed using a 5% level of significance.

RESULTS The mean age of the patients was 53.9 ⫾ 6.9 yr (range, 38 to 70 yr). There was a male predominance of 62:15. Most patients had tumors more than 4 cm in length (mean, 5.9 ⫾ 2.0 cm). Forty-two were current smokers, 19 were past smokers, and 16 were nonsmokers; 69 patients had no pulmonary symptoms, two patients had dyspnea, six had dry cough, and four had hoarseness. No patient had hemoptysis. None had an upper respiratory infection within 2 mo prior to study. All patients completed the neoadjuvant protocol; eight patients required a dose adjustment of their chemotherapy (four because of stomatitis, two because of a herpes simplex infection, one because of diarrhea, and one because of transient leukopenia and thrombocytopenia) and these adjustments were considered to be minimal. Three patients developed new symptoms of hoarseness (none showed vocal-cord paralysis at the restaging bronchoscopy), one developed nonproductive cough (tracheal infiltration of the esophageal cancer was proved at the restaging bronchoscopy in this patient), and three complained of dyspnea at the restaging (in one of them, new metastases were found radiographically); 67 patients remained without pulmonary symptoms. Lung Radiology

Initially, most patients had either a normal chest radiograph or a radiograph consistent with an obstructed esophagus, i.e., an air-fluid level in esophagus as the only abnormality. After neoadjuvant therapy, no changes on the plain chest radiograph were observed in 74 patients. Minor changes within the radiation field (pleural effusion and atelectasis in the right upper and middle lung field) were observed in one patient. Two

1743

Riedel, Stein, Mounyam, et al.: Neoadjuvant Therapy in Esophageal Cancer

patients developed multiple pulmonary nodules, which proved to be metastases of the esophageal cancer. None of the patients had lung parenchymal abnormalities on the initial thoracic CT. At the restaging, minor dystelectasia within the radiation field was observed in one patient. In two patients, multiple pulmonary metastases appeared on the CT at restaging. There was no case of an esophagotracheal fistula on either the initial or the restaging plain chest radiograph, CT, or swallow esophagograph. Tumor-related Bronchoscopic Findings

No patient showed a direct tumor sign at the initial bronchoscopy (Table 1). Of the indirect signs, mobile protrusion of the pars membranacea was the most frequent, occurring in 23 (29.9%) of the patients. In 61% of patients the airways were macroscopically normal. At the restaging, there was no change in the bronchoscopic findings in 31 patients with normal macroscopy and in eight patients with some macroscopic abnormalities at the initial bronchoscopy (Figure 1). In 20 patients the bronchoscopic findings were downstaged, with complete normalization of the macroscopic appearance and normal histologic and cytologic samples. Fifteen patients with initially normal macroscopic appearance of the central airways and two patients with mobile protrusion of the pars membranacea presented new tumorsuspicious macroscopic abnormalities at the restaging bronchoscopy: exophytic growth (frank intraluminal mass with white tissue of soft consistency) in 10, tumorous infiltration in three, suspicious mucosa (mucosal irregularities with telangiectasia) in two, mobile protrusion of the pars membranacea in one, and localized rigidity of the pars membranacea in one. Of these 17 patients, proof of cancer by biopsy and brushing was achieved in only one patient, and cancer-positive brush cytology but normal biopsy was obtained in two (one of them was rejected from surgery, and the other underwent an R0-resection and is well 39 mo after surgery); the tissue samples from the other 14 contained only nonspecific inflammatory and postinflammatory changes. One patient with normal macroscopy at both examinations had a strongly cancer-suspect routine brush cytology at the restaging bronchoscopy. This patient thereafter underwent an R0-resection and is well 19 mo after surgery; the cytologic diagnosis of airway invasion at bronchoscopy was thus falsely positive. No airway fistula was found after the radiochemotherapy. In five of the 41 operated patients with normal both macroscopic and microscopic examination at the restaging bronchos-

Figure 1. Bronchoscopic findings at the initial and the restaging examination and patient outcome.

copy, and in one of the 21 operated patients with macroscopic abnormality but normal microscopic examination at the restaging bronchoscopy, tracheobronchial cancer infiltration was diagnosed at surgery and confirmed by the histologic examination of the resected specimens (Figure 1). Thus, the bronchoscopy was falsely negative in these six patients. Mucosal Findings at Bronchoscopy

The general macroscopic appearance of the mucosa in the trachea and main bronchi (outside the possible tumor-related abnormalities) was normal in 25 patients at the initial bronchoscopy (Table 1). Thirty-four patients showed slight and 18 patients moderate mucosal atrophy. There were only insignificant changes of the overall appearance of the mucosa after the neoadjuvant therapy; 64 patients showed no change, in six the previously normal mucosa was described as slightly atrophic, and in 2 the initially slightly atrophic mucosa as moderately atrophic. No scarring leading to significant airway stenosis occurred in any of the patients. In Figure 2, the results of histologic examination of the tumor-uninvolved tracheal biopsies before neoadjuvant therapy are compared with those after the therapy. Induction therapy resulted in an increase in postinflammatory fibrosis, as well as hyperplasia and metaplasia. However, these abnormalities were frequently present even before neoadjuvant therapy. Brush cytology of the posterior tracheal wall was normal in 28 patients before and 25 after therapy. Cell metaplasia was described in 17 patients before and 16 after the therapy. Acute

TABLE 1 BRONCHOSCOPIC FINDINGS IN THE CENTRAL AIRWAYS BEFORE AND AFTER NEOADJUVANT THERAPY Abnormality Possibly tumor-related None Mobile protrusion Rigid protrusion Suspect mucosa Sole rigidity of posterior wall Endoluminal exophytic mass Tumorous infiltration Other mucosal changes None Slight atrophy Moderate atrophy Increased vulnerability

Initial Bronchoscopy

Restaging Bronchoscopy

47 23 5 2 0 0 0

52 (one positive brush cytology) 6 1 4 1 10 (one positive biopsy and brush) 3 (two positive brush cytologies)

25 34 18 6

22 37 18 4

1744

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 162

2000

TABLE 2 LUNG FUNCTION TESTS BEFORE AND AFTER NEOADJUVANT THERAPY Parameter VC, % pred FEV1, % pred FEV1/VC, % TLC, % pred RV, % pred RV/TLC, % Vtg, % pred sRaw, kPas TLCO, % pred TLCO/BSA, mmol/min/kPa/m2 PaO2, mm Hg PaCO2, mm Hg

Figure 2. Histology of the tumor-uninvolved trachea before (open bar) and after (closed bar) neoadjuvant radiochemotherapy.

inflammation occurred only sporadically and without any clear relation to neoadjuvant therapy. Lung Function Studies

Before therapy, pulmonary function was only minimally abnormal for the entire group of patients (Table 2). Only four patients had abnormal FEV1 and FEV1/VC values indicating mild obstructive disease. Significant increase in PaCO2 occurred after radiochemotherapy. There was no statistically significant change in the mean values of any other respiratory parameter examined. Five patients had a ⬎ 10% decrease in some lung volumes after the radiochemotherapy; in only one of them had the VC and FEV1 declined to less than 65% predicted. Only one patient experienced a decrease in TLCO to less than 80% predicted; the lung volumes and PaO2 remained in the normal range in this patient (Table 3). Esophageal Resection

After restaging, 64 patients proceeded to surgery, and 13 were treated palliatively. The reasons for palliation were distant metastases in four, tracheal infiltration in two, no change in tumor mass in three, poor physiologic status in three, and patient’s refusal of surgery in one. The complete resectability (R0) rate was 48 of 64 (75%) (Figure 1). In six patients, there was airway infiltration found at surgery or by histologic examination of the resected specimens (predictive value of a negative bronchoscopic result was 90.6%; 95% CI, 80.7 to 96.5%). Histopathologic examination of the surgical specimens dem-

Before

After

p Value

100.4 ⫾ 13.7 97.2 ⫾ 15.4 75.9 ⫾ 6.7 107.8 ⫾ 14.4 130.0 ⫾ 33.7 39.9 ⫾ 8.1 122.4 ⫾ 25.2 1.12 ⫾ 0.48 114.0 ⫾ 19.2 5.85 ⫾ 1.07 76.5 ⫾ 7.4 36.3 ⫾ 3.3

99.2 ⫾ 13.9 95.0 ⫾ 15.9 75.4 ⫾ 5.8 108.3 ⫾ 13.6 131.3 ⫾ 31.1 40.7 ⫾ 7.2 120.2 ⫾ 22.1 1.07 ⫾ 0.53 111.0 ⫾ 19.6 5.77 ⫾ 1.11 78.7 ⫾ 7.2 37.1 ⫾ 3.3

NS NS NS NS NS NS NS NS NS NS NS 0.01

Definition of abbreviations: BSA ⫽ body surface area; RV ⫽ residual volume; TLCO ⫽ lung transfer factor for carbon monoxide; Vtg ⫽ volume of thoracic gas.

onstrated no evidence of viable tumor (pT0 N0) in six (9.4%) of the 64 operated patients. The in-hospital mortality was 4 of 64 patients (6.3%). The causes of death were bleeding from an abscess, mediastinitis, multiorgan failure after peritonitis, and circulatory failure in a patient with coronary disease. Apart from 24 postoperative pleural effusions, postoperative pulmonary complications were noted in seven patient (pneumonia in three, empyema in two, tracheal lesion in two, pneumothorax in one, and acute respiratory failure in one). No patient died of pulmonary complications. There was no case of acute respiratory distress syndrome postoperatively. Follow-up

The median follow-up for all 77 patients was 15.5 mo, with 17.5 mo for the 60 operated and discharged patients. Of the 60 operated and discharged patients, 29 are alive 22.7 ⫾ 11.1 mo after surgery without pulmonary problems; 31 discharged patients died on average 15.7 ⫾ 10.6 mo after surgery, 29 of them of progressive or recurrent disease (in nine with tracheal invasion shortly before death) and two of other causes unrelated to the airways or lung parenchyma. Of the 11 palliatively treated patients, excluded from surgery for reasons other than airway infiltration, nine died 9.8 ⫾ 7.2 mo after the restaging bronchoscopy. Eight of them died of causes unrelated to the tracheobronchial tree, and in one, airway invasion was diagnosed 7 mo after the bronchoscopy. Two patients are surviving 18 and 19 mo after the bronchoscopy, respectively, without pulmonary problems.

TABLE 3 PATIENTS WITH DECLINE IN LUNG FUNCTION PARAMETERS AFTER THERAPY Patient No. 1 2 3

4 5 6

Postoperative Pulmonary Complication

Decline in Lung Function

Radiology at Restaging

FEV1 from 109 to 89% pred VC from 77 to 63% pred FEV1 from 79 to 63% pred VC from 117 to 81% pred FEV1 from 97 to 80% pred TLC from 135 to 114% pred VC from 116 to 102% pred FEV1 from 120 to 106% pred VC from 131 to 106% pred FEV1 from 128 to 91% pred TLC from 119 to 106% pred TLCO from 92 to 67% pred

Normal, no change Normal, no change

None None

Living 20 mo after resection Died 4 mo after resection

Normal, no change

Pneumonia

Living 8 mo after resection

Normal, no change

None

Died 20.5 mo after resection

Normal, no change



Not operated, died after 6 mo

Multiple metastases



Not operated, died after 6 mo

Definition of abbreviation: TLCO ⫽ lung transfer factor for carbon monoxide.

Outcome

Riedel, Stein, Mounyam, et al.: Neoadjuvant Therapy in Esophageal Cancer

DISCUSSION Tumor staging after neoadjuvant treatment is important both for identifying the degree of response and for planning subsequent treatment. Although EUS and CT are the best procedures to assess the local extent of esophageal cancer (20, 24), bronchoscopy is the best procedure to assess possible invasion of the cancer into the airways. With regard to the use of bronchoscopy in tumor staging after neoadjuvant treatment, few reports are available (21, 25). Our study was undertaken to evaluate the role of bronchoscopy and lung function testing after radiochemotherapy at the time of restaging before surgery. The patients served as their own controls. It is a firm policy at our institution to perform neoadjuvant therapy in all potentially operable patients with locally advanced nonmetastatic squamous cell esophageal cancer. All patients receiving neoadjuvant therapy underwent complete staging both at diagnosis and after the therapy. Our patient group is therefore representative of patients with this condition. Because neoadjuvant therapy with two- or three-drug chemotherapy (mainly cisplatinum-based) combined with 36 to 40 Gy radiation may increase postoperative morbidity and mortality (4), a single-drug chemotherapy (5-FU) in combination with only 30 Gy of radiation administered to a narrow field was chosen as the preoperative regimen. Although such regimen is not very aggressive, its efficacy and safety is documented by the absence of serious side effects, a complete histopathologic tumor response in 9.4%, and total gross removal of the disease in 75% of operated-on patients, as well as by a postoperative hospital mortality in the range reported for esophageal resection without preoperative radiochemotherapy. Diagnosis of Airway Invasion at Restaging

The evaluation of possible airway invasion by esophageal cancer after neoadjuvant therapy is likely to be more difficult than before treatment; first, if tumor regression has occurred, the volume of macroscopic tumor will be less than in the pretreatment bronchoscopy; second, tumor response may result in inflammation, granulation tissue, and fibrosis, all of which can be misinterpreted as residual active tumor. Macroscopic abnormalities in the airways were observed in 39% of our patients at the initial and in 32.5% at the restaging bronchoscopy (Table 1). Although the frequency of mobile protrusion of the pars membranacea decreased significantly, new apparent direct tumor signs (exophytic growths and suspected tumor infiltration) appeared in 13 cases. However, proof of cancer through histology could be obtained in only one of these 13 abnormalities. This suggests that macroscopic abnormalities after radiation must be interpreted even more carefully than at baseline staging, because altered mucosal structure (contracting scars, pallor, and telangiectasia) or white soft intraluminal masses could be the result of radiationinduced inflammatory or fibrous reactions alone and not of tumor invasion. Comparing the bronchoscopy results in patients receiving neoadjuvant therapy with those obtained in patients undergoing resection alone (21), a distinct worsening of diagnostic accuracy (overstaging) in the evaluation of direct tumor signs without histologic proof was noted. We believe that only histologic examination of all suspect abnormalities in the central airways is capable of proving or ruling out airway invasion of esophageal cancer after neoadjuvant therapy with acceptable accuracy. None of the biopsies or brushings from areas of protrusion of the posterior airway wall with normal motility during breathing at the initial and restaging bronchoscopies were can-

1745

cer-positive. This indicates that such protrusion may only be due to the presence of a bulky tumor in the vicinity and does not preclude radical resection. Regression of this protrusion at restaging might be a valid additional indicator of a good tumor response to neoadjuvant therapy. Brush cytology was falsely positive in two of three patients, as documented by an R0-resection and long-term survival of these patients. Some investigators have reservations concerning the usefulness of sampling for cytology from the airways in the presence of a primary aerodigestive tract tumor because of shedding of malignant cells from the tumor and their aspiration into the airways (26). It is now our policy in the presence of a positive cytology but negative biopsies to repeat the bronchoscopy, utilizing very extensive bioptic sampling. The results of bronchoscopy were falsely negative in six of the 64 operated patients. The extension of the tumor into the airway was probably limited in them and did not result either in appreciable macroscopic findings or in positive samples taken at bronchoscopy. Interestingly, in our previous study there were no falsely negative results in patients proceeding directly to resection without undergoing neoadjuvant therapy (21), probably because in our protocol only patients with early tumors are considered suitable for primary resection. Bronchopulmonary Effects of Neoadjuvant Therapy

The tumor-uninvolved mucosa of the trachea and main bronchi did not undergo any significant changes in the macroscopic appearance, especially no scarring leading to airway stenosis. Histology of the uninvolved mucosa demonstrated slightly more fibrosis, hyperplasia, and metaplasia, but these abnormalities were already present before radiochemotherapy in a high proportion of patients. In early radiation lung injury, pulmonary function tests are typical for an alveolar-based disease, with a decrease in lung volumes and TLCO (14, 15). In this series of patients with irradiated esophageal cancer and concurrent mild chemotherapy, the study of functional and imaging parameters did not detect any consistent impairment of the lung. Apart from a clinically unimportant but statistically significant increase in PaCO2, there was no significant change in the mean values of lung volumes or TLCO (Table 2). This may be in part due to the good general status of the patients and the great effort undertaken to avoid radiation to lung parenchyma uninvolved in the neoplastic process. Our findings are in agreement with reports on tangential irradiation of breast cancer (27), and we believe that, with adequate fractionation, the inclusion of a small portion of lung in the irradiated volume is acceptable. Malignant esophagotracheal fistula is a serious incurable complication of esophageal cancer. Radiotherapy and chemotherapy are thought to be contraindicated because cytoreduction enlarges the fistula. However, it has also been reported that patients who are responsive to radiochemotherapy may have actually closure of their fistulas (12). According to selection criteria for neoadjuvant therapy, none of the patients had a fistula before the therapy. No patient developed a fistula during the therapy. The etiologic role of 5-FU in the development of interstitial pneumonitis has not been described. The intensity of chemotherapy given in this study was less than that preferred for systemic therapy. Rather, the chemotherapy may be primarily radiosensitizing. In some experimental studies, a deleterious effect of high fractional doses of radiation on lung tissue has been described, but no increase in the toxicity of radiation has been observed when chemotherapy (cisplatin) has been added (28–30). However, in other studies on lung toxicity a deleterious effect of cytotoxic drugs administered simultaneously or

1746

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

sequentially with irradiation has been described. Severe pulmonary reactions on irradiation have been increased by alkylating agents, antimetabolites, antitumor antibiotics, and others (31). Although a radiosensitizing effect of 5-FU is discussed, no increased lung toxicity has been described either from interaction with irradiation or as a primary direct lung damage. This suggests that the primary factor for the development of lung disease, if any, would be radiation rather than 5-FU. Esophagectomy was performed in 83% of patients after completion of neoadjuvant therapy, with an in-house mortality rate of 6.3% and an operative mortality rate of 0%. These results are similar to those reported with surgery alone and to those from other studies using induction radiochemotherapy with moderate radiation doses (3, 7, 9, 11). This suggests that the addition of radiochemotherapy does not appreciably increase the mortality associated with esophagectomy. The low toxicity of our neoadjuvant regimen on one hand and the incomplete clinical response on the other raises the question about possible benefits of much stronger regimens. The ability to safely deliver curative intent radiochemotherapy in a neoadjuvant setting may have clinical advantages. Improved long-term survival appears to correlate with final pathologic stage; patients with a complete pathologic response have the best survival rates (5). Utilizing the best-proven medical therapy to produce the greatest degree of complete pathologic response preoperatively may further improve long-term outcomes. In the future, patients with advanced local disease may be treated uniformly with aggressive radiochemotherapy, rather than being tracked early to operative versus nonoperative therapy. This may allow more effective therapy upfront while still maintaining the ability to restage patients after induction therapy. In summary, the results of this study suggest that in most patients, neoadjuvant therapy of proximal esophageal cancer with 5-FU and 30 Gy conventionally fractionated radiation using tumor-confined treatment portals does not induce pneumonitis or changes in lung function that could be of concern at the following operation. This treatment method, such as described and with the dose and fractionation such as they were applied, appears safe and reliable with regards to the lung.

9.

10.

11.

12.

13. 14.

15. 16.

17.

18.

19. 20.

21.

22.

23.

Acknowledgment : The writers thank Dr. R. W. Hauck for performing some of the bronchoscopies in this study and Mrs. E. Amaseder, O. Felbermayr, E. Hopp, and I. Reinheimer for excellent technical assistance with the bronchoscopies. 24.

References 1. Tsujinaka T, Shiozaki H, Yamamoto M, Inoue M, Yano M, Doki Y, Monden M. Role of preoperative chemoradiation in the management of upper third thoracic esophageal squamous cell carcinoma. Am J Surg 1999;177:503–507. 2. Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N, Hennessy TPJ. A comparison of multimodal therapy and surgery for esophageal adenocarcinoma. N Engl J Med 1996;335:462–467. 3. Malhaire J, Labat J, Lozach P, Simon H, Lucas B, Topart P, Volant A. Preoperative concomitant radiochemotherapy in squamous cell carcinoma of the esophagus: results of a study of 56 patients. Int J Radiat Oncol Biol 1996;34:429–437. 4. Fink U, Stein HJ, Bochtler H, Roder JD, Wilke HJ, Siewert JR. Neoadjuvant therapy for squamous cell esophageal carcinoma. Ann Oncol 1994;3:S17–S26. 5. Fink U, Stein HJ, Siewert JR. Multimodale Therapie bei Tumoren des oberen Gastrointestinaltrakts. Chirurg 1998;69:349–359. 6. Siewert JR, Stein HJ, Fink U. Multimodality therapy for esophageal cancer. The Oncologist 1996;1:210–218. 7. Jones DR, Detterbeck FC, Egan TM, Parker LA, Bernard SA, Tepper JE. Induction chemoradiotherapy followed by esophagectomy in patients with carcinoma of the esophagus. Ann Thorac Surg 1997;64:185–192. 8. Bosset J-F, Gignoux M, Triboulet J-P, Tiret E, Mantion G, Elias D, Lo-

25. 26.

27.

28.

29. 30.

31.

VOL 162

2000

zach P, Ollier J-C, Pavy J-J, Merciar M, Sahmoud T. Chemoradiotherapy followed by surgery compared with surgery alone in squamouscell cancer of the esophagus. N Engl J Med 1997;337:161–167. Chidel MA, Rice TW, Adelstein DJ, Kupelian PA, Suh JH, Becker M. Resectable esophageal carcinoma: local control with neoadjuvant chemotherapy and radiation therapy. Radiology 1999;213:67–72. Tribble CG, Flanagen TL, Christensen CP, Daniel TM, Kron IL. The influence of preoperative radiation therapy on morbidity and mortality for transhiatal esophagectomy. Am Surg 1991;57:716–719. Yano M, Tsujinaka T, Shiozaki H, Inoue M, Doki Y, Yamamoto M, Tanaka E, Inoue T, Monden M. Concurrent chemotherapy (5-fluorouracil and cisplatin) and radiation therapy followed by surgery for T4 squamous cell carcinoma of the esophagus. J Surg Oncol 1999;70: 25–32. Ahmed HF, Hussain MA, Grant CE, Wadleigh RG. Closure of tracheoesophageal fistulas with chemotherapy and radiotherapy. Am J Clin Oncol 1998;21:177–179. Thar TL, Million RR. Complications of radiation treatment of Hodgkin’s disease. Semin Oncol 1980;7:174–183. Davis SD, Yankelevitz DF, Henschke CI. Radiation effects on the lung: clinical features, pathology, and imaging findings. Am J Roengenol 1992;159:1157–1164. Movsas B, Raffin TA, Epstein AH, Link CJ. Pulmonary radiation injury. Chest 1997;111:1061–1076. Hasleton PS. Lung injury due to radiation. In: Hasleton PS, editor. Spencer’s pathology of the lung, 5th ed. New York: McGraw-Hill; 1996. 537–549. Nakajima Y, Miyamoto T, Tanabe M. Enhancement of mammalian cell killing by 5-fluorouracil in combination with x-rays. Cancer Res 1979; 39:3763–3767. Rich TA, Lokich JJ, Chaffey J. A pilot study of protracted venous infusion of 5-fluorouracil and concomitant radiation therapy. J Clin Oncol 1985;3: 402–406. Union Internacional Contra la Cancrum. TNM Classification of malignant tumors, 5th ed. Berlin: Springer; 1997. Laterza E, de Manzoni G, Guglielmi A, Rodella L, Tedesco P, Cordiano C. Endoscopic ultrasonography in the staging of esophageal carcinoma after preoperative radiotherapy and chemotherapy. Ann Thorac Surg 1999;67:1466–1469. Riedel M, Hauck RW, Stein HJ, Mounyam L, Schulz C, Schomig A, Siewert JR. Preoperative bronchoscopic assessment of airway invasion by esophageal cancer: a prospective study. Chest 1998;113:687–695. Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows. Report Working Party Standardization of Lung Function Tests. Official Statement of the European Respiratory Society. Eur Respir J 1993;16(Suppl):5–40. Cotes JE, Chinn DJ, Quanjer PH, Roca J, Yernault JC. Standardization of the measurement of transfer factor (diffusing capacity): Report Working Party Standardization of Lung Function Tests. Official Statement of the European Respiratory Society. Eur Respir J 1993;16(Suppl): 41–52. Griffith JF, Chan ACW, Chow LTC, Leung SF, Lam YH, Liang EY, Chung SCS, Metreweli C. Assessing chemotherapy response of squamous cell oesophageal carcinoma with spiral CT. Br J Radiol 1999; 72:678–684. Baisi A, Bonavina L, Peracchia A. Bronchoscopic staging of squamous cell carcinoma of the upper thoracic esophagus. Arch Surg 1999;134:140–143. Dhooge I, Joos G, Van Cauwenberge P. Value of rigid bronchoscopy and cytodiagnosis of bronchial washings in detecting bronchial carcinoma in the presence of a carcinoma of the upper aerodigestive tract. J Otolaryngol 1996;25:219–222. Botterman J, Tasson J, Schelstraete K, Pauwels R, Van Der Straeten M, DeSchryver A. Scintigraphic, spirometric, and roentgenologic effects of radiotherapy on normal lung tissue: short-term observations in 14 consecutive patients with breast cancer. Chest 1990;97:97–102. Giri PGS, Kimler BF, Giri UP, Cox GG, Reddy EK. Comparison of single, fractionated and hyperfractionated irradiation on the development of normal tissue damage in rat lung. Int J Radiat Oncol Biol Phys 1985;11:527–534. Van Dyk J, Mah K, Keane TJ. Radiation-induced lung damage: dosetime-fractionation considerations. Radiother Oncol 1989;14:55–69. von der Maase J, Overgaard J, Vaeth M. Effect of cancer chemotherapeutic drugs on radiation-induced lung damage in mice. Radiother Oncol 1986;5:245–257. Phillips TL. Radiation-chemotherapy interactions. In: Pass HL, Mitchell JB, Johnson DH, Turrisi AT, editors. Lung cancer: principles and practice. Philadelphia: Lippincott-Raven; 1996. p. 251–270.