Patient characteristics

Among 182 patients with LA-NSCLC who underwent radical resection after iCRT between 1999 and 2019, 30 did not undergo a covering procedure. Thus, we reviewed 152 patients. Table 1 summarizes the patients’ clinical characteristics (male, n = 114; female, n = 38; median age, 60 [range 31–79] years). Adenocarcinoma was the most common histological subtype (n = 85; 55.9%). Clinical stage III was the most common (stage IIIA, n = 82, 53.9%; stage IIIB, n = 50, 32.9%). The right side was operated on in 84 patients (55.3%). The bronchial stump/anastomotic site was sutured by hand-stitching in 58 cases (38.2%). The surgical approach in all cases was open thoracotomy, with lobectomy being the most common procedure (110 cases, 72.4%), followed by sleeve lobectomy (17 cases, 11.2%), bilobectomy (14 cases, 9.2%), and pneumonectomy (11 cases, 7.2%).

Table 1 Characteristics of patients with locally advanced non-small cell lung cancer undergoing induction chemoradiation followed by surgery with covering of the bronchial stump/anastomotic site (n = 152)Tissue flaps used to cover the bronchial stump/anastomotic site

Supplemental Table 1 summarizes the tissue flaps used to cover the bronchial stump/anastomotic site. The following tissues were used: pericardial fat pad and/or thymus, n = 102 (67.1%); omentum, n = 25 (16.4%); intercostal muscle, n = 20 (13.2%); both pericardial fat pad and intercostal muscle, n = 2; serratus anterior muscle, n = 2; and latissimus dorsi muscle, n = 1. The pericardial fat pad and/or thymus were collected as pedicled flaps in 101 of the 102 cases. We experienced 1 case of gastric outlet obstruction after covering the omentum, in which gastrojejunostomy was required after surgery. Subsequently, gastrostomies and jejunostomies were performed at the time of surgery in almost all cases, in which the omentum was used to cover the bronchial stump/anastomotic site. Although we experienced another case in which the patient presented with gastric outlet obstruction, it was managed conservatively. The choice of covering tissue is usually at the surgeon’s discretion. Recently, however, there has been a growing trend in our institution to use of pericardial fat pads.

Frequency of BPF and detailed information of each case

BPF was observed in 5 patients (3.3%) (Supplemental Table 2); however, no significant differences in the frequency of BPF were observed in any of the tissues. Table 2 shows the detailed information of each case with BPF. Five patients underwent right upper sleeve lobectomy with the sacrifice of pulmonary artery A6, right middle and lower lobectomy, right middle and lower sleeve lobectomy with the sacrifice of pulmonary artery A2, right upper and S6 sleeve lobectomy, and right middle and S6 sleeve lobectomy; that is, 4 of 5 cases had bronchial anastomosis. The tissue flaps used for covering were the omentum in 2 cases, and the intercostal muscle, pericardial fat pad, and pericardial fat pad plus thymus in the remaining cases. Each case had factors that may have led to BPF. Two cases had the sacrifice of the pulmonary artery branch to the spared lobe, which we previously reported as a possible risk factor for BPF in sleeve lobectomy after iCRT; that is, residual S6 necrosis associated with the sacrifice of A6 occurred in 1 case, and organizing pneumonia and hemorrhagic infarction had been pathologically demonstrated in the residual upper lobe after completion pneumonectomy in another case with the sacrifice of A2 [24]. In addition, we had 1 case with low nutritional intake because of difficulty in oral intake due to the development of esophageal-bronchial fistula during iCRT and 2 cases of postoperative pneumonia. Furthermore, metastasis in lymph node #7 was found preoperatively in 4 of 5 cases, and the subcarinal area was included in the irradiation field in all 5 cases. After BPF, thoracic fenestration was performed in 3 cases, and intrathoracic omentum transposition was performed in 2 cases in which the omentum was spared for the first surgery. However, 4 of these patients died within 3 years after surgery, including 2 who died from acute respiratory distress syndrome and bronchovascular fistula that developed after BPF (Supplemental Table 3).

Table 2 Detailed information of each case with BPFThickness of the tissue flap used to cover the bronchial stump/anastomotic site

We previously encountered a patient with a history of iCRT followed by bilobectomy for LA-NSCLC, who developed BPF after completion pneumonectomy for recurrence [4]. We treated the patient conservatively with N-butyl-2-cyanoacrylate under bronchoscopy because the covering tissue was thick, and the fistula was small (Supplemental Fig. 1). Therefore, we focused on the thickness of the covering tissue in this study. We measured the thickness of the intercostal muscle, omentum, and pericardial fat pad, which are frequently used for covering, in a horizontal CT section in the available cases (n = 90) in which CT was performed within 1 month (Fig. 1A–C left side) and approximately 1 year (Fig. 1A–C right side) after surgery. The measurement was the shortest vertical distance of the covered tissue at the bronchial stump/anastomotic site. Figure 2 shows the distribution of the thickness of the intercostal muscle (n = 10), omentum (n = 13), and pericardial fat pad (n = 67). At the time of the operation, there was a significant difference in tissue thickness among the 3 groups, with the omentum being the thickest. There was no significant difference between the intercostal and pericardial fat pads at one year after surgery. In the cases that developed BPF, the covering tissue appeared to be slightly thin, as shown in (Fig. 1D) (cases 3 and 5). In contrast, there were some patients without BPF, irrespective of the presence of postoperative pneumonia or radiation pneumonitis on the operative side. The covering tissue in these cases appeared to be slightly thicker than that in the cases that developed BPF Fig. (1E, F). Thus, we also analyzed the relationship between the thickness of the covering tissue and the BPF. As shown in (Fig. 3A), when comparing the 127 cases with CT scans obtained within 6 months postoperatively, the thickness of the covering tissue was significantly greater in cases that did not develop BPF (median, 12.1 [range, 4.0–29.2] mm) relative to cases that developed BPF (median, median, 8.2 [range, 5.5–13.0] mm) (p = 0.0290). Furthermore, as shown in (Fig. 3B), even among the 50 cases in which BPF did not develop, irrespective of the presence of pneumonia or radiation pneumonitis on the operative side within 6 months after surgery, the thickness of the covering tissue (median, 12.3 [range, 6.7–23.8] mm) was abundant relative to the 5 cases with BPF (median, 8.2 [range, 5.5–13.0] mm) (p = 0.0077). As shown in (Table 3), there were no significant differences in perioperative risk factors for BPF between cases with and without BPF. However, in BPF cases, squamous cell carcinoma was histologically predominant, and all cases were managed using hand-stitch procedures.

Fig. 1

CT scans of patients in whom the bronchial stump/anastomotic site was covered by A intercostal muscle, B omentum, and C pericardial fat pad. The arrow shows the thickness of the tissue flap covering the bronchial stump/anastomotic site in the transverse plane on CT scans obtained within 1 month (left side) and approximately 1 year after surgery (right side). D In the cases in which BPF developed (left side: case No.3, right side: case No.5), the thickness of the covering tissue were 8.3 mm and 7.7 mm, respectively. E, F In the cases without BPF, irrespective of postoperative E pneumonia and F radiation pneumonitis of the operative side (considered to be risk factors for BPF), the thickness of the covering tissue was 15.3 and 15.2 mm, respectively (left side: mediastinal window setting, right side: lung window setting)

Fig. 2

The distribution of the thickness of the tissue flaps covering the bronchial stump/anastomotic site at the time of operation (left side) and 1 year after surgery (right side). We compared the thickness of the intercostal muscle (n = 10), omentum (n = 13), and pericardial fat pad (n = 67)

Fig. 3

The distribution of the thickness of the covering tissue in A cases that developed BPF (n = 5) and the others (n = 127), and in B cases that developed BPF (n = 5) and high-risk cases without BPF (n = 50), in which CT scans were obtained within 6 months postoperatively. The thickness of the covering tissue was based on the CT scan at the time at which postoperative complications (e.g., BPF, pneumonia, etc.) developed

Table 3 Background of patients with locally advanced non-small cell lung cancer undergoing induction chemoradiation followed by surgery with coverage of the bronchial stump/anastomotic site according to the presence or absence of BPF