The assessing the left bronchial invasion of esophageal cancer in computed tomography

Background and aims: The prognosis of esophageal squamous cell carcinoma varies according to the degree of invasion of the tumor in the organ wall. When the trachea and the main bronchus are involved in esophageal carcinoma, the disease is classified as incurable and the mortality is lower than that for patients who can be treated with curative surgery. To evaluate the carcinoma invading the trachea and main bronchus, retrospectively, we investigated computed tomography (CT) findings. Methods: Of 74 patients with or without proven tracheobronchial involvement who underwent thoracotomy for esophagectomy and Surgery disclosed tumor invasion in pericardium, aorta, pleura, trachea, left main bronchus and parenchyma. Ten patients were not operated upon but exhibited invasion to respiratory tract by broncho-fiber scope. Results: Lt. main bronchus (NMS angle) were statistically significant in patients with and without invasion. (P<0.05) Conclusion: It is our conclusion that preoperative evaluation by CT plays an important role whether the tumor is resectable or not. Correspondence to: Hiroshi Makino, MD, Department of Surgery, Nippon Medical School, Tama Nagayama Hospital, Tokyo, Japan. 1-7-1, Nagayama, Tama-city, Tokyo, Japan; Tel: +81-42-371-2111; Fax: +81-42-372-7384; E-mail: himiyumo@nms.ac.jp


Introduction
Patients with esophageal carcinoma have a poor prognosis because many esophageal tumors are detected at advanced stages. Many cases of esophageal carcinoma involving the trachea or main bronchus are unresectable or resected with high risk. Preoperative diagnosis to determine possible tumor invasion to a major airway is therefore important.
Numerous reports have described the usefulness of CT in the preoperative staging of esophageal carcinomas [1][2][3]. Several authors have been disappointed by its poor sensitivity and accuracy [4,5]. Vilgrain et al. reported that endoscopic sonography was superior to CT, and many authors reported broncho -fiber scope or ultrasonography were useful for evaluating cancer invasion into the tracheal bronchus [6]. Takashima et al. reported that magnetic resonance imaging (MRI) and CT had nearly the same accuracy in predicting resectability of tumors in patients with esophageal carcinoma [7].
But CT is a popular method to evaluate the invasion of major airways. Matsubara et al. mentioned that CT and broncho-fiber scope played a significant role in selecting resectable patients [8]. Picus et al. [9] reported that the angle to aorta was useful to evaluate the invasion to aorta. To evaluate the carcinoma invading the trachea, main bronchus and aorta we analyzed 6 matters including the angles and distance from the center of the esophageal lumen, the sectional area and the diameter of the esophagus in CT, retrospectively.

Cases
We analyzed 74 patients (male: female 59:15) with thoracic esophageal carcinoma who had undergone radical esophagectomy at the Nippon Medical School Hospital or Nippon Medical School, Chiba Hokusoh Hospital between January 1994 and December 2005. The median age is 61.4 years old ranged from 41 to 82 years (Table 1).

CT Scan
All patients had CT scans of the neck, chest, and abdomen. Ten millimeter continuous scans were obtained in areas from the neck to the bottom of the liver. The CT scans were performed using intravenous contrast medium. 6 matters were investigated via CT (computed tomography) for findings of tracheo-bronchial involvement.

Depth
Diameter t3 t4 Figure 3. The frequency of group t4 in diameter tended to be higher than that of t3.
We analyzed the frequencies of tumors in the 6 matters via CT (Figure 2-5). The patients were divided into 2 groups; involvement of other organs (t4) and non-involvement (t3),and also separated into the 2 groups; involvement to other organs (t4) and non-involvement (t3) based on the pathological findings after surgery.
We compared the frequencies of tumors in the 6 matters of both groups (t4 and t3). In the sectional area of the tumor (1) (Figure 2), diameter (4) (Figure 3), Picus angle (5) (Figure 4), angles of lt. main bronchus (NMS angle 6) ( Figure 5), the frequencies in group t4 tended to be higher than that of t3.
We compared CT findings (t4 and t3) and surgical-pathological findings (t4 and t3). There were significant differences of Picus angle and angles of lt. main bronchus (NMS angle) between the 2 groups by statistical analysis. (P<0.05) There were no significant differences in other matters.
The criteria is as follows: (1) aortic infiltration diagnosed the existence of an angle greater than 90 degrees between tumor and aorta (2) lts main bronchus infiltration was diagnosed when the tumor developed an intimate contact with the lt. main bronchus that was greater than 135 degrees. Sensitivity, specificity and accuracy of (1) were 71%, 77% and 67%, respectively and those of (2) were 100%, 85 % and 75%, respectively (   On the other hand, other studies have obtained very poor results with CT in the evaluation of esophageal carcinoma [14].
Many investigators reported that in staging the depth of tumor growth, endoscopic ultrasonography was significantly more accurate [6]. Broncho-fiber scope allows the direct visualization of changes, such as hyperemia of the mucosa, compression of the airway, and invasion itself, and is essential for the evaluation of the respiratory tree in esophageal cancer carcinoma. But Halvorsen et al. [10] summarized a statistically significant advantage for spiral CT over bronchoscopy in advanced esophageal masses with a sensitivity of 88.3%.
Takashima et al. concluded that MR and CT have nearly the same accuracy in predicting resectability of tumors in patients with esophageal carcinoma [7].
Endoscopic ultrasonography was an excellent method for evaluating t3 or t4, but sometimes the endoscopy cannot pass the tumor because of stenosis caused by esophageal tumor especially in the case of the tumor extending beyond adventitia. CT was applied to all the patients even when the esophageal lumen was stenotic. Schimer et al. described no statistically significant difference between CT and broncho-fiber scope because an accuracy of 85.1% was obtained with regard to impingement, displacement and invasion of trachea and bronchus [1]. Lehr et al. also mentioned poor results with CT in the evaluation of esophageal carcinoma due to small sample size and that one-third of the cases had trans-mediastinal esophageal resection, that were unhelpful in evaluating invasion [15]. The larger the tumor the greater the diagnostic capacity of CT, justifying the accuracy of 85.1% obtained [16][17][18].
Another problem is evaluation by surgical technique. A transmediastinal esophageal dissection technique is not available in evaluating tumors because dissection at the superior mediastinum is perfomed without direct vision [1,18].
It is our conclusion that esphagectomy is preferable when the tracheo-bronchial involvement is assessed as technically, resectable and preoperative assessment by CT plays an important role whether the tumor is resectable or not.