ObjectiveTo evaluate the impact of empiric tissue flaps on bronchopleural fistula (BPF) rates after pneumonectomy. MethodsPatients who underwent pneumonectomy between January 2001 and December 2019 were included. Primary end point was development of BPF. Secondary end points were impact of flap type on BPF rates, time to BPF development, and perioperative mortality. ResultsDuring the study period, 383 pneumonectomies were performed; 93 were extrapleural pneumonectomy. Most pneumonectomy cases had empiric flap coverage, with greater use in right-sided operations (right: 97%, 154/159; left: 80%, 179/224, P < .001). Empiric flaps harvested included intercostal, latissimus dorsi, serratus anterior, omentum, pectoralis major, pericardial fat/thymus, pericardium, and pleura. BPF occurred in 10.4% of the entire cohort but decreased to 6.6% when extrapleural pneumonectomy cases were excluded; 90% (36/40) of BPFs occurred on the right side (P < .001). Median time to develop BPF was 63 days, and 90-day mortality was greater in patients with BPF (12.5% BPF vs 7.4% non-BPF, P < .0001). Intercostal muscle had the lowest rate of BPF (4.5%), even in right-sided operations (8.7%). In contrast, larger muscle flaps such as latissimus dorsi (21%) and serratus anterior (33%) had greater rates of BPF, but the sample size was small in these cohorts. ConclusionsEmpiric bronchial stump coverage should be performed in all right pneumonectomy cases due to greater risk of BPF. In our series, intercostal muscle flaps had low BPF rates, even in right-sided operations. Coverage of the left pneumonectomy stump is unnecessary due to low incidence of BPF in these cases.
Read full abstract