Zhendan Yao1, Hong Yang1, Ming Cui1, Jiadi Xing1, Chenghai Zhang1, Nan Zhang1, Lei Chen1, Maoxing Liu1, Kai Xu1, Fei Tan1, Xiangqian Su2. 1. Department IV of Gastrointestinal Cancer Center, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital and Institute, Beijing 100142,China. 2. Department IV of Gastrointestinal Cancer Center, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital and Institute, Beijing 100142,China, Email: suxiangqian@bjmu.edu.cn.
Abstract
OBJECTIVE: To investigate the risk factors of postoperative pulmonary infection (PPI) in patients over 60 years of age with gastric cancer after radical gastrectomy. METHODS: Clinicopathological data of 373 patients over 60 years of age who underwent radical gastrectomy at Department IV of Gastrointestinal Cancer Center, Peking University Cancer Hospital, from April 2009 to December 2016 were retrospectively collected in this case-control study. The clinicopathological characteristics of patients with postoperative pulmonary infection (including postoperative atelectasis) and those without pulmonary infection were compared. A Student t-test (reported as Mean±SD if data matching normal distribution) or Mann-Whitney U test [reported as median (quartile) if data did not conform to normal distribution] was used to analyze continuous variables. A χ² test or Fisher exact tests (reported as number and percentage) was used for categorical variables. Multivariable logistic regression was used to analyze the risk factors for pulmonary infection after operation of gastric cancer.PPI was defined as postoperative patients with elevated body temperature (>38.0 degrees centigrade) for more than 24 hours; cough and expectoration; positive sputum bacteria culture;recent infiltration, consolidation or atelectasis confirmed by chest imaging examination. RESULTS: Among 373 patients, 50 cases had PPI(13.4%, PPI group), 323 cases had no PPI(86.6%, non-PPI group). There were 39 (78.0%) and 178(55.1%) patients with comorbidities (including hypertension, diabetes and cardiopulmonary disease) preoperatively in PPI and non-PPI group, respectively. The difference between two groups was statistically significant (χ²=9.325,P=0.002). The incidence of preoperative hypoalbuminemia in PPI group was also significantly higher than that in non-PPI group [10.0%(5/50) vs. 3.1% (10/323),χ²=4.098, P=0.048]. Compared to non-PPI group, the rate of total gastrectomy [54.0%(27/50) vs. 34.4% (111/323), χ²=12.501, P=0.002], postoperative wound pain [34.0%(17/50) vs. 11.8% (38/323),χ²=16.928, P<0.001], secondary operation [6.0%(3/50) vs. 0.6% (2/323), χ²=6.032, P=0.014] and the rate of gastric tube removal later than 7 days postoperatively [96.0%(48/50) vs. 84.5%(273/323),χ²=4.811, P=0.028] were significantly higher in PPI group, respectively. The postoperative hospital stay was also prolonged in PPI group [16.0(9.5) days vs. 12.0(5.0) days, U=4 275.0, P<0.001]. Multivariate logistic regression analysis showed that preoperative comorbidities (OR=4.008, 95%CI:1.768-9.086, P=0.001), abdominal infection (OR=3.164, 95%CI:1.075-9.313, P=0.037), and wound pain (OR=3.428, 95%CI:1.557-7.548, P=0.002) were independent risk factors for PPI in patients over 60 years of age with gastric cancer. Furthermore, 50 patients with pulmonary infection were classified according to the length of latency and the type of infection. The patients with PPI latency ≤ 3 days were classified as early onset (34 cases, 68.0%), and those with latency ≥ 4 days as delayed onset (16 cases, 32.0%); PPI combined with surgical infection (including anastomotic leakage, abdominal infection, duodenal stump leakage, wound infection, etc.) was classified into mixed infection group (13 cases, 26.0%), with non-surgical infection as simple infection group (37 cases, 74.0%). The results showed that the pulmonary infection occurred 0 to 12 days (median 3 days) before surgical infection in mix infection group. The incidence of previous chronic obstructive pulmonary disease (COPD) in patients with early onset was significantly higher than that in patients with delayed onset [17.6%(6/34) vs. 0, χ²=5.005, P=0.025], and the incidence of mixed infection in patients with delayed onset was significantly higher than that in patients with early onset [50%(8/16) vs. 14.7%(5/34), χ²=6.730, P=0.009],but there was no significant difference in postoperative hospital stay between the two groups[17.0(9.8) days vs. 14.0(9.5) days, U=224.0, P=0.317]. CONCLUSIONS: Postoperative pulmonary infection is common in gastric cancer patients over 60 years of age. Preoperative comorbidities, abdominal infection and wound pain are independent risk factors for postoperative pulmonary infection. Pulmonary infection within 3 days after operation is associated with preoperative COPD. For patients suffering from PPI after the 4th day,attentions should be paid to abdominal infection and anastomotic leakage.
OBJECTIVE: To investigate the risk factors of postoperative pulmonary infection (PPI) in patients over 60 years of age with gastric cancer after radical gastrectomy. METHODS: Clinicopathological data of 373 patients over 60 years of age who underwent radical gastrectomy at Department IV of Gastrointestinal Cancer Center, Peking University Cancer Hospital, from April 2009 to December 2016 were retrospectively collected in this case-control study. The clinicopathological characteristics of patients with postoperative pulmonary infection (including postoperative atelectasis) and those without pulmonary infection were compared. A Student t-test (reported as Mean±SD if data matching normal distribution) or Mann-Whitney U test [reported as median (quartile) if data did not conform to normal distribution] was used to analyze continuous variables. A χ² test or Fisher exact tests (reported as number and percentage) was used for categorical variables. Multivariable logistic regression was used to analyze the risk factors for pulmonary infection after operation of gastric cancer.PPI was defined as postoperative patients with elevated body temperature (>38.0 degrees centigrade) for more than 24 hours; cough and expectoration; positive sputum bacteria culture;recent infiltration, consolidation or atelectasis confirmed by chest imaging examination. RESULTS: Among 373 patients, 50 cases had PPI(13.4%, PPI group), 323 cases had no PPI(86.6%, non-PPI group). There were 39 (78.0%) and 178(55.1%) patients with comorbidities (including hypertension, diabetes and cardiopulmonary disease) preoperatively in PPI and non-PPI group, respectively. The difference between two groups was statistically significant (χ²=9.325,P=0.002). The incidence of preoperative hypoalbuminemia in PPI group was also significantly higher than that in non-PPI group [10.0%(5/50) vs. 3.1% (10/323),χ²=4.098, P=0.048]. Compared to non-PPI group, the rate of total gastrectomy [54.0%(27/50) vs. 34.4% (111/323), χ²=12.501, P=0.002], postoperative wound pain [34.0%(17/50) vs. 11.8% (38/323),χ²=16.928, P<0.001], secondary operation [6.0%(3/50) vs. 0.6% (2/323), χ²=6.032, P=0.014] and the rate of gastric tube removal later than 7 days postoperatively [96.0%(48/50) vs. 84.5%(273/323),χ²=4.811, P=0.028] were significantly higher in PPI group, respectively. The postoperative hospital stay was also prolonged in PPI group [16.0(9.5) days vs. 12.0(5.0) days, U=4 275.0, P<0.001]. Multivariate logistic regression analysis showed that preoperative comorbidities (OR=4.008, 95%CI:1.768-9.086, P=0.001), abdominal infection (OR=3.164, 95%CI:1.075-9.313, P=0.037), and wound pain (OR=3.428, 95%CI:1.557-7.548, P=0.002) were independent risk factors for PPI in patients over 60 years of age with gastric cancer. Furthermore, 50 patients with pulmonary infection were classified according to the length of latency and the type of infection. The patients with PPI latency ≤ 3 days were classified as early onset (34 cases, 68.0%), and those with latency ≥ 4 days as delayed onset (16 cases, 32.0%); PPI combined with surgical infection (including anastomotic leakage, abdominal infection, duodenal stump leakage, wound infection, etc.) was classified into mixed infection group (13 cases, 26.0%), with non-surgical infection as simple infection group (37 cases, 74.0%). The results showed that the pulmonary infection occurred 0 to 12 days (median 3 days) before surgical infection in mix infection group. The incidence of previous chronic obstructive pulmonary disease (COPD) in patients with early onset was significantly higher than that in patients with delayed onset [17.6%(6/34) vs. 0, χ²=5.005, P=0.025], and the incidence of mixed infection in patients with delayed onset was significantly higher than that in patients with early onset [50%(8/16) vs. 14.7%(5/34), χ²=6.730, P=0.009],but there was no significant difference in postoperative hospital stay between the two groups[17.0(9.8) days vs. 14.0(9.5) days, U=224.0, P=0.317]. CONCLUSIONS:Postoperative pulmonary infection is common in gastric cancerpatients over 60 years of age. Preoperative comorbidities, abdominal infection and wound pain are independent risk factors for postoperative pulmonary infection. Pulmonary infection within 3 days after operation is associated with preoperative COPD. For patients suffering from PPI after the 4th day,attentions should be paid to abdominal infection and anastomotic leakage.