Re-appraisal of bronchus-first right upper lobectomy as an alternative routine procedure

Authors: Takahiro Utsumi, Haruaki Hino, Natsumi Maru, Hiroshi Matsui, Yohei Taniguchi, Tomohito Saito, Tomohiro Murakawa

Abstract

OBJECTIVES

Postoperative air leakage is a major complication of lung resection, particularly right upper lobectomy. However, various surgical procedures can reduce postoperative complications and shorten the drainage period. The current study aimed to analyse the utility of bronchus-first right upper lobectomy as an alternative routine procedure.

METHODS

We retrospectively analysed the data of 225 (53.7%) patients who underwent bronchus-first right upper lobectomy and 194 (46.3%) patients who underwent the conventional bronchus-last right upper lobectomy at our institution from 2015 to 2022. In patients with incomplete fissures who underwent bronchus-first right upper lobectomy, the bronchus was dissected 1st, followed by the pulmonary artery and vein, and then, the interlobar fissure was divided. We compared the outcomes of 2 procedures and analysed the surgical utility of bronchus-first right upper lobectomy.

RESULTS

The surgical outcomes and postoperative morbidity comparing bronchus-first and bronchus-last procedure were as median operation time (min) 103/126 (*P < *0.001), median bleeding amount (ml) 28/55 (*P = *0.003), incomplete lobulation rate (%) 35.1/24.2 (*P = *0.02), incidence of prolonged air leakage (%) 2.2/3.1 (P = 0.76) and rate of fellow surgeon’s operation (%) 28.0/4.6 (*P < *0.001). The procedure was associated with a decreased incidence of prolonged air leakage. The 4-year overall survival rates did not significantly differ between the 2 groups (*P = *0.24).

CONCLUSIONS

Bronchus-first right upper lobectomy can prevent postoperative air leakage in patients with incomplete fissure. Additionally, as an alternative routine procedure, it is associated with a shorter surgical duration and a lower volume of blood loss regardless of interlobar fissure and operator’s experience.

INTRODUCTION

Right upper lobectomy (RUL) is a major surgery performed by general thoracic surgeons for primary lung cancer. However, due to incomplete lobulation, RUL is more associated with prolonged air leakage (PAL) and bleeding compared with other types of lobectomies [1–3]. The incidence rate of PAL in pulmonary resection is 8–15%. PAL is associated with a longer length of postoperative stay, high incidence of postoperative complications and greater medical costs [4–6]. According to previous publications, bronchus-first right upper lobectomy (BF-RUL) is associated with a lower incidence of PAL and a shorter surgical duration and postoperative drainage period [7, 8]. However, previous studies did not evaluate factors such as lobulation (the degree of interlobar fissure), operator’s experience and postoperative survival. The primary outcome of this study was the BF-RUL procedure’s effect on perioperative outcomes, such as operation time, bleeding amount and complication including the incidence of postoperative air leakage as well as postoperative survival time compared with those of bronchus-last right upper lobectomy (BL-RUL).

MATERIALS AND METHODS

Ethical statement

This research was conducted based on the Declaration of Helsinki and was approved by the ethics committee of Kansai Medical University on 20 June 2023 (approval 2023007). The need for informed consent was waived. The data of this article can be requested from the corresponding author upon reasonable requests. This retrospective study used data from a clinical database of a single institution. We included 419 patients with 410 non-small-cell lung cancers (NSCLC), 6 metastatic lung tumours, and 3 pulmonary infections who underwent BF-RUL or BL-RUL from 2015 to 2022. Then, the participants were divided into the BF-RUL (n = 225, 53.7%) and BL-RUL (n = 194, 46.3%) groups. As for an inclusion criterion, especially for primary lung cancer, complete resection towards clinical stage I to IIIA was included; however, patients with incomplete resection and preoperative treatment were excluded in this study.

Surgical procedure

In our general fashion, a 4-port approach is routinely adopted, 2-port in 4th or 5th intercostal space [one of anterior (3–4 cm) and one of dorsal position (2 cm)], and 2-port in 6th or 7th intercostal space (2 cm in each port). When 2-port approach was performed by TM, 4th or 5th intercostal space (3–4 cm) and 7th intercostal space (2 cm) were used as a surgical port. In BF-RUL, the right upper bronchus was resected 1st while simultaneously dissecting the hilar peribronchial lymph nodes (Fig. 1A). After the dissection between the anterior pulmonary artery trunk and the right upper bronchus, the pulmonary artery branch was easily detected (Fig. 1B). Next, the anterior pulmonary artery trunk or the ascending pulmonary artery, followed by the pulmonary vein and, finally, the interlobar fissure, was stapled (Fig. 1C; Video 1). If lobulation was poor and the pulmonary artery could not be identified easily from the interlobar fissure, the right upper bronchus and the pulmonary artery and vein were 1st dissected. Finally, the incomplete lobar fissure was resected, which was referred to as fissureless lobectomy [9]. If hilar peribronchial lymph nodes were challenging to dissect due to adhesion or metastasis, or pulmonary artery at the interlobar fissure was easily detected, or lobulation was almost completed, BL-RUL was performed. However, when the interlobar pulmonary artery could not be identified intraoperatively, BF-RUL was performed with the fissureless method. In BL-RUL, a fissureless method was used as much as possible. Meanwhile, lymph node dissection was performed at a similar range regardless of the order in which the bronchus was resected (BF-RUL or BL-RUL). All surgeries were performed by a team involving 2 fellows (Takahiro Utsumi and Natsumi Maru) and 5 attendant physicians (Tomohiro Murakawa, Haruaki Hino, Tomohito Saito, Yohei Taniguchi and Hiroshi Matsui) as a fellow surgeon and an attending physician or 2 attending physicians. The surgery was basically completed by 1 operator from beginning to end. However, the frequency of assistance provided by the attending physician was more likely to decrease based on the course of experiences on those procedures. The surgical outcomes of patients who underwent BF-RUL or those who underwent BL-RUL were compared and assessed.

Statistical analysis

The following data were age, sex, body mass index, smoking history (pack-years), serum albumin level, percentage of vital capacity, forced expiratory volume in 1 s (FEV1)/forced vital capacity, Charlson comorbidity index [10], carcinoembryonic antigen level, surgical duration, volume of blood loss, type of procedure, experience of the operator (fellow or attending physician), number of staplers used to dissect interlobar fissure, number of hilar lymph nodes dissected, degree of lobulation (complete or incomplete fissure), clinical and pathological stage, postoperative complications, chest tube detention period, length of postoperative stay and observation time. The degree of lobulation was assigned during the operation, which was obtained from our database, and the Craig’s definition was used to assess fissure grades, which were as grade I, complete fissure; grade II, lobes partially fused at the fissure but with complete visceral cleft; grade III, lobes partially fused at the fissure with incomplete visceral cleft; and grade IV, completely fused lobes and no evidence of fusion line [11]. Postoperative complications were defined via a chart view according to the Clavien-Dindo classification. PAL was defined as a pulmonary fistula lasting for ≥ 7 days or pleurodesis or need for additional surgical treatment for PAL [12]. Overall survival was defined as time from surgery to the date of death or last follow-up. Disease-free survival was calculated from the date of surgery to the date of lung cancer recurrence. Tumour stage was determined according to the 8h edition of the TNM staging system of the International Union against Cancer [13]. The histological tumour type was determined according to the 3rd edition of the World Health Organization Classification of Tumours [14]. Continuous and categorical variables between the BF-RUL and BL-RUL groups were analysed using the chi-square test. Survival time was calculated using the Kaplan–Meier method, and differences in survival were assessed using the log-rank test. Univariate and multivariate logistic regression analysis was performed to assess the risk of postoperative air leakage. All statistical analyses were performed by EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a modified version of R commander (The R Foundation for Statistical Computing, Vienna, Austria) [15]. A P value of <0.05 was considered statistically significant.

RESULTS

Surgical outcomes

Table 1 shows the preoperative clinical characteristics of the participants. The BF-RUL group was older (*P = *0.02) and had a higher proportion of patients with adenocarcinomas (*P = *0.01) than the BL-RUL group. Meanwhile, the BL-RUL group had higher pack-years (*P = *0.04) and a lower percentage of vital capacity (*P = 0.01) than the BF-RUL group. There were no significant differences between the 2 groups in terms of sex, body mass index, serum albumin level, FEV1/forced vital capacity, Charlson comorbidity index, carcinoembryonic antigen level and clinical stage (P > *0.05). Table 2 shows the perioperative surgical outcomes. The BF-RUL group had a significantly shorter surgical duration (*P < *0.001), lower volume of blood loss (*P = *0.003), higher percentage of incomplete fissures (*P = *0.02), greater number of resected hilar lymph nodes (*P = *0.003) and higher rate of surgery performed by fellow surgeons (*P < *0.001) than the BL-RUL group. However, the BF-RUL and BL-RUL groups had similar outcomes in terms of numbers of stapler used to divide the incomplete fissures (*P = *0.36), pathological stage (*P = *0.41) and rate of progression from clinical stage N0 to pathological stage N1 (*P = *0.87). Next, Table 3 depicts the postoperative complications. There were no significant differences in terms of the overall incidence of postoperative complications (BF-RUL vs BL-RUL) 38 (16.9%) vs 38 (19.6%) (*P = *0.53), pulmonary fistula 16 (7.1%) vs 15 (7.7%) (*P = *0.85), PAL 5 (2.2%) vs 6 (3.1%) (*P = *0.76), severity of pulmonary fistula 8 (3.6%) vs 10 (5.2%) (*P = *0.47), median drain detention period 2 (1–29) vs 2 (1–22) (*P = *0.12) and median length of postoperative hospital stay 6 (3–209) vs 10 (2–58) (*P = *0.08) (Table 3). Based on the current analysis, the BF-RUL and BL-RUL groups differed in terms of the interlobar lobulation. Therefore, a subgroup analysis of 126 patients with grade III and IV lobulation (poor lobulation) was performed. Results showed that 79 and 47 patients underwent BF-RUL and BL-RUL, respectively. Hence, BF-RUL was performed more commonly than BL-RUL in patients with poor lobulation (*P = *0.02). Further, the BF-RUL and BL-RUL groups were similar in terms of the incidence of 7-day postoperative air leakage (4 [3.2%] vs 3 [3.4%], *P = *0.75), overall incidence of grade III or higher complications (7 [5.6%] vs 3 [2.4%], *P = *0.62) and median drain retention period (2 [range: 1–21] vs 1 [range: 1–22] days, *P = *0.35). The median surgical duration, volume of blood loss and length of postoperative hospital stay were 120 (range: 65–270) and 135 (range: 70–266) min (*P = *0.02), 120 (range: 3–65) and 135 (range: 70–266) ml (*P = *0.02), and 7 (3–65) and 10 (4–58) days (*P = *0.04), respectively. Hence, BF-RUL can be possibly associated with a reduced length of hospitalization in patients with poor lobulation. Next, we analysed the learning curve benefit for the procedure of BF-RUL by 2 fellow surgeons. The cases were divided into the 1st half part (n = 30) and latter half part (n = 33) and analysed surgical outcomes. There were no significant differences between the 2 parts in terms of the median volume of blood loss (ml) (the 1st part vs the latter part), 35 (range: 0–150) vs 51 (range: 0–764) (*P = *0.53); overall incidence of postoperative complications, 0 (0.0%) vs 5 (15.2%) (*P = *0.05); and pulmonary fistula, 0 (0.0%) vs 2 (6.1%) (*P = *0.49). However, the latter half part had a significantly shorter surgical duration than the 1st half part; median surgical duration (min), 124.0 (range: 81–180) vs 104.5 (range: 62–174) (*P = *0.005). As a result, we realized that each surgical manoeuvre in BF-RUL is familiar and performed more precisely and quickly than beginning.

Multivariate analysis

Based on the result of the multivariate analysis for postoperative air leakage (Table 4), male sex [odds 4.694, 95% confidence interval (CI) 1.138–12.422, *P = *0.03], lower FEV1 (odds 0.962, 95% CI 0.926–1.000, *P = *0.048) and poor lobulation (odds 2.020, 95% CI 1.116–3.500, *P = *0.01) were independent risk factors for postoperative air leakage. Moreover, BF-RUL was marginal variable to reduce the complication (odds 0.688, 95% CI 0.302–1.560, *P = *0.37).

Survival analysis

Figure 2 shows the survival analysis results. Since BF-RUL was performed from 2019, survival time was analysed among 190 patients with pathological stage I NSCLC: 161 patients in BF-RUL and 29 patients with BL-RUL. The overall 4-year survival rates of patients with pathological stage I NSCLC in the BF-RUL and BL-RUL groups were 94.2% (95% CI: 88.6–97.1) and 100% (95% CI: NA–NA), respectively. The median overall survival time of the BF-RUL and BL-RUL groups was 27 and 17 months (*P = *0.24), respectively (Fig. 2A). Similarly, the 4-year disease-free survival rates of patients with pathological stage I NSCLC in the BF-RUL and BL-RUL groups were 89.1% (95% CI: 78.9–94.6) and 94.7% (95% CI: 68.1–99.2) (*P = *0.59), respectively. The median disease-free survival time was 24 and 15 months, respectively (Fig. 2B). Therefore, the BF-RUL and BL-RUL groups had similar survival outcomes.

DISCUSSION

Lung cancer surgery has been increasing worldwide, and PAL is a major complication associated with a prolonged hospital stay, impaired quality of life and possibly decreased survival. According to previous publications, the risk factors of PAL after primary lung cancer surgery are low FEV1/forced vital capacity, upper lobe resection, incomplete fissure, pleural adhesion, body mass index of <18.5 kg/m^2^ and steroid use [3, 16, 17], which are consistent with our results (Table 4). These factors are also correlated with postoperative morbidities such as pneumonia and empyema [6, 18–20]. In particular, RUL is associated with a high risk of PAL due to the higher incidence of incomplete lobulation and complications such as severe pulmonary emphysema and pulmonary interstitial fibrosis [3]. Temes et al. [9] 1st reported the procedure of fissureless lobectomy for resecting hilar component with the pulmonary artery and vein and bronchus prior to incomplete fissure plasty to prevent postoperative PAL. Thereafter, there have been several reports showing that fissureless RUL can shorten the postoperative drainage period and reduce the complication rate of PAL in clinical practice [21–23]. Furthermore, based on the studies of Refai et al. [21], Stamenovic et al. [22] and Igai et al. [23], the median surgical durations of fissureless and non-fissureless lobectomy were 183 and 180 min (*P = *0.80), 176 and 185 min (*P = *0.52), and 270 and 222 min (*P = *0.0045), respectively. Hence, fissureless lobectomy had almost similar or a longer surgical time than non-fissureless lobectomy. Our results showed that BF-RUL had a shorter surgical time, lower volume of blood loss and an acceptable morbidity. However, the procedure of BF-RUL did not significantly impact on postoperative complication of air leakage (Tables 3 and 4). Surgical indication of BF-RUL might affect the multivariate analysis. Hence, a greater number of BF-RUL may contribute to reducing a significant postoperative air leakage, which we need to clarify in future.

BF-RUL was 1st reported to be applied in the 1940s, and it aimed to control massive bronchial secretions during anaesthesia. Further, it was considered a superior technique for lymph node dissection and vascular exposure [24]. BF-RUL can have a significantly shorter operative time and drain detention period compared with BL-RUL [8]. However, the degree of lobulation and volume of blood loss were not clearly analysed. We experienced a case in which the anterior pulmonary artery trunk was not long enough to be safely resected because the hilar peribronchial lymph nodes were enlarged. Therefore, in such a case, the length of the anterior pulmonary artery trunk was secured by cutting the right upper bronchus prior to pulmonary artery, and then, the anterior pulmonary artery was safely dissected afterwards. Dissecting enough hilar peribronchial lymph nodes and achieving a sufficient separation between the right upper bronchus and the anterior pulmonary artery trunk can enable a safer division of the right upper bronchus without causing any injury in the pulmonary artery (Fig. 1A). After dissecting the right upper bronchus, the anterior pulmonary artery trunk and the ascending pulmonary artery or small A3 branch are easily identified and divided safely (Fig. 1B). In particular, the right upper lobe is placed in the anterior mediastinum during and after resecting the right upper bronchus, and the pulmonary artery branches are towed to the anterior direction. Then, a stapler is inserted intuitively from the caudal to the cranial direction to cut the pulmonary artery branch without any obstruction at the tip of liner stapler. In addition, this procedure can substantially provide an educational value to younger surgeons. Therefore, BF-RUL as an alternative routine method is considered a feasible procedure compared with conventional BL-RUL regardless of the interlobar fissure condition and operator’s experience.

The survival rates of BF-RUL and BL-RUL are similar in patients with pathological stage I NSCLC (Fig. 2). According to previous publications, as the number of circulating tumour cell is reduced, lung cancer surgery with pulmonary vein resection 1st may improve survival compared with pulmonary artery-first resection [25, 26]. By contrast, the order of resection of the pulmonary artery or vein did not affect postoperative survival [27]. During BF-RUL, the pulmonary vein was often resected next to the pulmonary arteries. Therefore, circulating tumour cells might affect survival. However, the right upper lobe is not commonly manipulated around a tumour, and a standardized method can improve the surgical duration and decrease the volume of blood loss with a greater number of resected lymph nodes. Therefore, BF-RUL and BL-RUL have similar survival rates even from an oncological perspective. However, the observation period for the 4-year survival rates was considered short. Therefore, a longer follow-up is required to validate sufficient oncological survival.

Limitations

The current study has several limitations. First, this retrospective study was conducted at a single institution and was not randomized. Hence, some biases of surgical indication might exist. Further, BL-RUL could have been performed in challenging cases such as hilar and chest wall adhesion, metastasis and bleeding. Second, the postoperative observation period was short; the actual 5-year survival rate could not be calculated. Third, the indication for BF-RUL was not clearly established. However, we performed a reasonably robust analysis showing that BF-RUL is a utility procedure regardless of interlobar fissure conditions and can detect the pulmonary artery branch comprehensively. Hence, BF-RUL may shorten the surgical time and reduce the volume of blood loss. Nevertheless, further multi-institutional cohort studies with a higher number of patients undergoing BF-RUL should be performed to obtain the generalizability of BF-RUL, thereby supporting its surgical utility and strengthening the overall interpretation of the study results in the future.

CONCLUSION

BF-RUL can be considered as a utility procedure regardless of interlobar fissure condition. Further, the procedure may have a shorter surgical duration and lower volume of blood loss as it is performed by an alternative routine method. Nevertheless, more studies should be performed to investigate the appropriate indication for BF-RUL.