Cellular innate immune responses to lung resection via video- assisted thoracoscopic surgery (VATS) and thoracotomy : predictors of post-operative pneumonia
Jones, Richard Oliver
Background and Objectives: The pathophysiology of post-operative pneumonia following lung resection is poorly understood despite it being the most common complication which may lead to death. The role of the acute inflammatory response following lung resection, in particular innate immune cells, was investigated and used to identify biomarkers for post-operative pneumonia. Comparison of inflammatory responses to resection undertaken by video-assisted thorascopic surgery (VATS) and thoracotomy was also evaluated. Methods: Patients undergoing lung resection for suspected bronchogenic carcinoma were recruited. Objective pre-defined criteria were used to diagnose pneumonia. Bronchoalveolar lavage (BAL) was conducted in the contra-lateral lung pre- and postoperatively to measure cellular composition and cytokines. Blood was sampled preoperatively and 6-, 24- and 48-hours post-operatively primarily to assess neutrophil phagocytic capacity, monocyte subsets, monocyte cytokine responses to lipopolysaccharide (LPS) stimulation and serum cytokine responses. Exhaled nitric oxide (eNO) was also measured at these time points. Patient groups were compared using paired or student t-tests together with ANOVA/ANCOVA modelling. The predictive strength of the biomarkers identified was tested. Results: 40 patients were recruited. 26 patients (65%) underwent major lung resection using VATS and 14 (35%) thoracotomy. There was a post-operative blood monocytosis (p<0.0005) with an absolute expansion of classical and intermediate monocytes (p=0.001) and a relative fall in non-classical monocytes (p<0.005). Post-operatively blood monocytes became more pro-inflammatory with an overall significant increase in IL-8 (p=0.034) and TNF-α (p=0.028) together with an increase in IL-6 (p=0.028) and IL-10 by 48 hours (p=0.010). VATS was associated with a smaller release of IL-10 only (p=0.011). There was a general trend towards post-operative reduction in neutrophil phagocytosis of zymosan (in suspension) on ANOVA modelling (p=0.047). Lung resection led to an increase in serum cytokines IL-6, IL-8 and IL-10 which peaked at 24hrs before falling (p<0.0005). ANOVA modelling confirmed significantly lower levels of serum cytokines in VATS patients compared with thoracotomy (p=0.026 for IL-6, p=0.018 for IL-8 p=0.047 for IL-10). No significant post-operative change was found for IL-1β, TNF-α and IL-12p70 (p>0.05). Bronchoalveolar lavage fluid (BALF) and blood samples demonstrated a relative post-operative leucocytosis due principally to neutrophilia. A relative blood lymphopenia and thrombocytopenia developed postoperatively (p<0.0005). VATS was associated with a lower fall in serum albumin (p=0.001). BALF from the non-operated lung became more pro-inflammatory immediately post-operatively with an increase in IL-6 (p<0.0005), IL-8 (p=0.017), IL- 10 (p=0.018) and IL-1β (p=0.002). eNO tended to fall post-operatively which reached significance at 48 hrs (p=0.029). 14 patients developed pneumonia. Pre-operatively, a blood neutrophil count above 5.04x109/L had a relative risk (RR) for pneumonia of 3.3 (95% confidence interval (CI95) 1.1-10.1), and a BAL cell count of greater than 1.04x105/ml had a RR of 3.4 (CI95 1.3-9.0), whilst LPS-stimulated monocyte secretion of IL-12 of less than 0.15 pg/ml/μg protein had a RR of 3.0 (CI95 1.2-7.3). At 24 hours post-operatively, LPS-stimulated release from monocytes of IL-10 greater than 1.99 pg/ml/μg protein (RR 4.1, CI95 1.3- 12.3) and IL-6 greater than 414 pg/ml/μg protein (RR 3.1, CI95 1.2-8.1) were predictive of pneumonia. Conclusion: Lung resection is associated with significant early pro- and antiinflammatory responses. VATS resection invoked significantly lower levels of serum cytokines and albumin changes compared with thoracotomy suggesting VATS lobectomy should be the surgical treatment strategy of choice for early stage lung cancer. No difference in neutrophil function or monocyte function was however observed between the surgical groups. Clinical benefits of this reduced inflammation need to be evaluated in a larger cohort of patients. Relative pre-operative leucocytosis in blood and BAL together with monocyte hyper-responsiveness in the early postoperative period is associated with the development of pneumonia. These findings warrant further investigation for their predictive power in accurately identifying postoperative pneumonia. Ultimately, they may be incorporated into a risk stratification model enabling targeted prophylactic or earlier therapeutic intervention.