| Background and purpose:In recent years,immune checkpoint inhibitors(ICIs)have become one of the most promising anticancer drugs,whose efficacy has been confirmed in a series of clinical trials.These monoclonal antibodies Body has the ability to reactivate the immune system of tumor cells,but it can also cause many autoimmune side effects,which are called immune related adverse events(irAE).irAE can affect all systems of the whole body,and its occurrence characteristics are quite different from traditional chemotherapy and targeted drugs.In these immune related adverse reactions,pneumonia is rare,and as a serious clinical event,it can endanger the life of patients.Even though there are many reports about the incidence rate of PD-1/PD-L1 inhibitor associated pneumonia,there are limited numbers of individuals and cohort acquired in the above clinical trials.In addition,although some related meta-analyses also studied the incidence of immune-related pneumonia,they also faced the problem of the limited number of studies included in the analysis.In addition,PD-1/PD-L1 inhibitors have been shown to significantly prolong OS and PFS in patients with advanced cancer.Based on this,we conducted a meta-analysis on the clinical trials of PD-1/PD-L1 inhibitors in patients with advanced cancer.First,we analyzed the overall incidence of pneumonia associated with the use of immune checkpoint inhibitors,and the correlation between the incidence of pneumonia and immune checkpoint inhibitors,and compared the incidence of pneumonia in different tumor types and types At the same time,we compared the differences between the first-line and second-line immune checkpoint inhibitors in the incidence of immune-related pneumonia.Finally,we analyzed the incidence of PD-1 and PD-L1 related pneumonia,and compared the differences between them in the incidence of immune pneumonia.Such analysis can provide important information for us to diagnose and manage immune-related pneumonia better and faster.Methods:PubMed,Cochrane Library,EMBASE and China HowNet databases were searched by computer,and the relevant literature about the incidence and risk of pneumonia related to PD-1 / PD-L1 inhibitors in patients with advanced cancer was searched.Because avelumab and durvalumab published few reports,the search term in English was: "immune checkpoint?inhibitor?ICIs?nivolumab?pembrolizumab?atezolizumab?cancer?tumor?phase II?phase III?pneumonia ",Chinese: " PD-L1 inhibitor?PD-1 inhibitor?immunocheckpoint inhibitor ",the search time is from January 2000 to January 2020,the literature type is limited to clinical trials,in order to prevent literature omission,and search relevant references at the same time.This article is not retrieved manually.Then two assessors select documents,evaluate document quality and extract relevant original data according to the established standards.If there are differences between the two sides in the process of checking,the third assessor needs to intervene or discuss with each other to solve them.Finally,the data were processed by statistical software Revman 5.3,and the incidence of each study was shown in the small forest area.In order to more intuitively and effectively compare the incidence of immune pneumonia under different conditions,we processed the single incidence of immune pneumonia according to the data.The meta-analysis of the incidence of pneumonia was performed by using the metaprop function of meta package in r3.6.2 software.If the incidence does not obey the normal distribution,the logarithm conversion is performed.Single factor Logistic regression model was used to test the incidence of immune pneumonia among different groups.Result:After eliminating duplicate literature,topic screening,abstract screening and full-text screening,15 randomized controlled trials were finally included,a total of 8642 patients.The meta analysis results are as follows.1.With immune checkpoint inhibitors,the total incidence of all levels of pneumonia: 3.0%(95% CI:2.0%—4.0%).2.To evaluate the incidence of immune pneumonia in the study of treatment group as PD-1 inhibitor or PD-L1 inhibitor and control group as chemotherapy group,the incidence OR of all levels of immune pneumonia was 6.65(95% CI: 4.24-10.44,P < 0.00001),and the incidence OR of high-level immune pneumonia was 4.76(95% CI: 2.54-8.91,P < 0.00001).3.Compared with the control group,the application of nivolumab increased the risk of all levels of immune pneumonia(OR: 6.27,95% confidence interval: 2.35-16.73,P = 0.0002),but did not increase the risk of high-level immune pneumonia(OR: 2.73,95% confidence interval: 0.56-13.39,P = 0.22).Compared with the control group,pembrolizumab treatment group increased the risk of all levels of pneumonia(OR: 6.76,95% confidence interval: 3.98-11.50,P < 0.00001),and increased the risk of high-level pneumonia(OR: 5.13,95% confidence interval: 2.49-10.57,P < 0.00001).Compared with the control group,atezolizumab treatment group increased the risk of all pneumonia(OR: 6.75,95% confidence interval: 1.22-37.45,P = 0.03),but did not increase the risk of high-grade pneumonia(OR: 5.74,95% confidence interval: 0.69-47.86,P = 0.11).4.The or of pneumonia in patients with NSCLC was higher than that in patients with chemotherapy(OR: 7.11,95% confidence interval: 4.11-12.28,P < 0.00001)and high-level(OR: 6.17,95% confidence interval: 2.73-13.95,P < 0.0001).5.The incidence of pneumonia in all grades of NSCLC in the treatment group was 1.624 times higher than that of other tumors(95% CI:1.171 – 2.286,P = 0.0044).The incidence of high-grade pneumonia in non-small cell lung cancer in the treatment group was 1.989 times higher than that of other tumors(95% CI:1.154-3.618,P = 0.0175).6.The incidence of pneumonia at all levels in first-line ICIS was 0.494 times higher than that in second-line and above(95% CI: 0.363-0.675,P < 0.0001).The incidence of high-grade pneumonia in the first-line application of ICIS was 0.533 times higher than that in the second-line application(95% CI: 0.325-0.889,P = 0.0138).7.The incidence of pneumonia at all levels was 3.143 times(95% CI: 1.742-6.393,P = 0.0005).The incidence of high-grade pneumonia was 2.255 times higher than that of PD-L1(95% CI: 0.996-6.470,P = 0.0818).There was no statistical significance.Conclusion:1.Compared with the control group(chemotherapy),the risk of all levels and high-level immune-associated pneumonia is higher when PD-1 / PD-L1 inhibitors are used.2.Compared with the control group,the application of nivolumab and atezolizumab increased the risk of all levels of immune pneumonia,but did not increase the risk of high-level immune pneumonia.The risk of all levels of pneumonia and high-level pneumonia in pembrolizumab treatment group increased.3.High risk of non-small cell lung cancer and first-line application in ICIS group4.Compared with the use of PD-L1 inhibitors,the risk of all levels of immune-related pneumonia was higher,and the risk of high-level immune pneumonia was not increased. |
PDF Full Text Request