The Predictive Value And Clinical Application Of Peripheral Blood Biomarkers In Immunotherapy Of Advanced Non-small Cell Lung Cancer

Lung cancer is one of the most common neoplasms in the world and the main cause of cancer-related death.Non-small cell lung cancer(NSCLC)accounts for approximately 85%,of which more than half are diagnosed at an advanced stage.The emergence of immunotherapy has greatly changed the treatment mode of lung cancer,especially for patients at advanced stage,and immunotherapy targeting programmed death-1(PD-1)or its ligand(programmed death-ligand 1,PD-L1)has significantly prolonged the overall survival(OS)of patients with advanced NSCLC.However,only some patients can achieve durable responses after receiving immunotherapy,while some patients suffer significant immune-related adverse effects(ir AEs),in which checkpoint inhibitor-related pneumonitis(CIP)need paying close attention to in NSCLC patients because of its high incidence and mortality.In view of this,it is very important to find effective biomarkers to predict the efficacy and ir AEs of patients receiving immunotherapy.In terms of efficacy prediction,the level of PD-L1 and tumor mutational burdern(TMB)in tumor tissue are two biomarkers approved by FDA to screen the population who can benefit from ICIs.However,the two biomarkers are not perfect.Due to differences between immunologic preparations,test platforms,the use of different expression thresholds,and spatiotemporal heterogeneity PD-L1 expression in tumors,PD-L1 detection still has certain limitations in guiding clinical practice.As a biomarker,TMB also has limitations,including the lack of standardization among the test platforms and clearly fixed defining thresholds of specific TMB levels.In terms of ir AE prediction,research on biomarkers mainly focuses on the endocrine system,gastrointestinal tract,skin and lung,including the excessive release of inflammatory cytokines,the overactivation of the immune system,elevated levels of pre-existing autoantibodies,and presence of common antigens between tumors and normal tissues,many of which are organ-specific biomarkers.To predict ir AEs,many biomarkers have been used for research.Regrettably,none of them is perfect.The research on ir AE is currently at an early stage,as there are some deficiencies and little progress has been made.Some biomarkers are not routine tested in clinical diagnosis and treatment,with high cost,thus seriously limiting their clinical application.Therefore,developing simple,measurable and precise biomarkers is of paramount importance for personalized patient management.Compared with tumor tissue,blood samples are relatively easy to obtain and less traumatic,which is convenient for repeated real-time dynamic detection.It can not only be a powerful supplement to tissue detection,but also a way to overcome tumor patiotemporal heterogeneity to a certain extent.For a predictive role,peripheral blood biomarkers are one of the important research directions to explore from different levels such as DNA,proteins and cells.It has shown a certain predictive ability and may become a new dynamic predictive marker of immunotherapy.At the level of DNA,ct DNA testing may serve as a noninvasive,tumor specific marker that plays an important role in the process of population screening and efficacy monitoring of immunotherapy.Blood tumor mutational burden(b TMB),calculated using ct DNA from peripheral blood,which is simple and noninvasive,represents a more potential approach,for the advantages of correcting sampling bias introduced by tumor heterogeneity or low tumor proportion,and enabling dynamic monitoring of early treatment.In the exploratory analysis of the large randomized controlled phase III Came L-sq study led by Shanghai Pulmonary Hospital,reseachers found that b TMB before treatment and its dynamic changes(Δb TMB)could serve as predictive biomarkers for patients with advanced lung squamous cell carcinoma receiving camrelizumab plus chemotherapy.In Came L-sq study,biomarker analysis also revealed that ct DNA clearance after two cycles of treatment was independently associated with dramatically longer PFS and OS in camrelizumab plus chemotherapy group.Several other studies have proved that there was also some association between ct DNA changes during ICIS therapy and treatment response.At the level of proteins,tumor associated autoantibodies(TAAbs)have an important role in the early diagnosis of lung cancer.NMPA approved the adjuvant application of this seven autoantibodies(CAGE,SOX2,MAGE A1,GAGE7,p53,GBU4-5 and PGP9.5)serum detection kit for the early diagnosis of lung cancer.In addition,TAAbs can indirectly reflect the response situation of the body’s immune system to tumors and have a large potential as predictive markers for the efficacy of immunotherapy.At the cellular level,blood tests routinely taken during treatment of patients have been shown to serve as potential predictive biomarkers.The role of eosinophils in tumor immunotherapy is being actively explored.A retrospective study found that high absolute eosinophil count(AEC)levels in peripheral blood at baseline were significantly associated with prolonged PFS and OS in NSCLC patients treated with ICIs.Previous studies have reported that factors such as blood eosinophils may be associated with the development of ir AEs,but the correlation with the occurrence of specific ir AEs still lacks valid research evidence.This suggests that,as a clinical routine examination,the blood routine may hide tumor related information,and it is simple and feasible to explore its potential biomarkers,which will not bring additional examination risk and economic burden to patients..In conclusion,the emergence of ICIS has brought a new dawn to the survival and prognosis of patients with advanced NSCLC,however,not all patients can benefit,and the approved efficacy prediction markers are not perfect and limited by tissue biopsy.CIP has attracted much attention because of its high incidence and mortality in NSCLC,but there is a lack of effective predictive biomarkers,and the application of some biomarkers not routinely detected in clinical diagnosis and treatment is limited by relatively high costs.In the first part,we observed the efficacy and ir AE(focusing on CIP)of immunotherapy for advanced NSCLC in the real world;In the second part,we explored the markers that can be used to predict CIP from the blood routine examination indexes,and discussed its correlation with clinical efficacy;In the third part,we screened the markers that can predict the efficacy of immunotherapy in advanced NSCLC from the protein level(taabs),and explored the correlation with CIP.Part I: Real-world outcomes and checkpoint inhibitor-related pneumonitis of immunotherapy-based regimens in advanced non-small cell lung cancer[Objective] This part was to observe the outcomes and adverse events of immunotherapy for advanced non-small cell lung cancer in the real world.[Methods] 1.Patients with advanced NSCLC who received PD-1 inhibitor treatment in the department of Oncology in our center from June 2015 to June 2019 were collected.All patients had continuous imaging data.2.Clinical information(age,gender,smoking history,histological type,ECOG PS score,basic disease,hormone medication history,immunotherapy scheme and line number,etc.)was recorded.3.The efficacy was evaluated according to the RECIST 1.1.4.We recorded the occurrence time of CIP and graded it according to CTCAE 4.0.[Results]1.Baseline characteristics of study population A total of 300 patients with advanced NSCLC treated with PD-1 inhibitors were included in this part,with median age of 62 years(range 20～82 years),women accounting for 20%,ECOG PS score of 0～1 accounting for 92.3%;192 cases of adenocarcinoma(64.0%),82 cases of squamous cell carcinoma(27.3%),6 cases of large cell carcinoma(2.0%),and 21 cases of unspecified type(7.0%);134 cases(44.7%)were treated with single drug and 166 cases(55.3%)were treated with combined therapy;102 patients(34%)received ICIs as first-line treatment and 198patients(66%)received ICIs as second-line or later treatment.2.Clinical efficacy Of the 300 patients,none obtained CR,103 obtained PR,127 obtained SD and 70 obtained PD,with the overall ORR of 34.3 and DCR of 76.7%.The median follow-up time was 23.8 months(0.47～43.67 months),with the median PFS of 6.6m and the median OS of 23.8m.Among 102 patients receiving ICIs as first-line treatment,ORR and DCR was52.0% and 87.3%,respectively,with m PFS of 12.4m,and m OS was not reached;among 198 patients receiving ICIs as the second-line or later treatment,ORR was and DCR was 25.3% and 71.2%,respectively,with m PFS of 5.3m,and m OS of 20.4m.Among 134 patients receiving ICIs monotherapy,ORR and DCR was 31.3% and67.9%,respectively,with m PFS of 4.7m,and m OS of 21.9m;among 166 patients receiving combined therapy,ORR and DCR was 36.7% and 83.7%,respectively,with m PFS of 9.7m,and m OS was not reached.Multivariate analysis showed that patients who received ICIs as second-line or later therapy had worse outcome than those as first-line therapy(ORR: HR=3.879,P<0.001;DCR: HR=2.599,P=0.014;PFS: HR=1.428,P=0.041;OS: HR=1.831,P=0.032).Compared with ICIs monotherapy,patients treated with ICIs combined therapy were conducive to higher DCR(HR=0.35,P=0.003)and longer PFS(HR=0.682,P=0.014).3.Checkpoint inhibitor-related pneumonitisNo unexpected adverse events were observed in 300 patients.Fifty-four patients(18%)experienced CIP,with the median age of 63.5 years,and the median occurrence time was 3.8 months(ranging from 0.7 to 21.0 months).In 54 patients with CIP,grade 1 was experienced by 25 patients(46.3%),grade 2 was experienced by 21 patients(38.9%),grade 3 was experienced by 7 patients(13.0%),grade 4 was experienced by one patient(1.9%),and none of the patients experienced grade 5.The radiologic features of CIP were classified into five patterns,with the following distribution in our population: COP(15/54,27.8%),GGO(2/54,3.7%),NSIP(9/54,16.7%),HP(21/54,38.9%),and a final group mixing other patterns(7/54,13.0%).There was no significant difference in the incidence of CIP among patients with different clinical characteristics,such as gender,age,ECOG PS score,smoking history,histology,treatment lines,treatment scheme,and whether there was pulmonary disease.[Conclusions] In the real world,some patients with advanced NSCLC obtain long-term survival from immunotherapy;early and combined application of immunotherapy could bring more benefits.The incidence of CIP is higher than that in clinical researches.It is urgent to explore biomarkers predicting benefit population and high-risk population of CIP.Part II: The prediction and clinical application of eosinophils in immunotherapy of advanced non-small cell lung cancer[Objective] This part was designed to explore the biomarkers correlated with CIP and clinical outcomes in patients with advanced NSCLC receiving ICIs from baseline blood tests routinely taken in clinical practice,providing clinical data for relevant basic transformation research.[Methods] The peripheral blood routine results of patients with advanced NSCLC treated with ICIs in Part I were collected retrospectively.ROC curve was used to identify the optimal cutoff value for analyzing risk of CIP.Multivariate logistic analysis was performed to determine risk factors of CIP.Clinical characteristics and treatment outcomes were collected and compared according to the optimal cutoff value.[Results]1.Associations between baseline peripheral blood routine results and CIPBaseline peripheral-blood absolute eosinophil count(AEC)was higher in patients with CIP than in patients without CIP(P=0.013).2.Associations between baseline AEC and CIPBased on the ROC curve,the optimal threshold for predicting baseline peripheral-blood AEC in CIP was 0.125×109 cells/L.Incidence of CIP was higher in high-AEC group compared with low-AEC group(27.7 % vs.9.8 %,P<0.001).Among the patients with high AEC(n=38),17(44.7%)had grade 1 CIP,15(39.5%)had grade 2,five(13.2%)had grade 3,and one(2.6%)had grade 4.Among those with low AEC(n=16),eight(50%)had grade 1 CIP,six(37.5%)had grade 2,and two(12.5%)had grade 3.There was no significant difference in the distribution of severity between the high-AEC group and the low-AEC group(P=0.918).The distributions of radiologic types of CIP were similar in the high-AEC group and the low-AEC group(P=0.266).3.Association between baseline AEC and clinical outcomesPatients with high AEC had significantly higher ORR(40.9% vs.28.8%,P=0.029)and longer median PFS(8.93 m vs.5.87 m,P=0.038).[Conclusions] Among patients treated with ICIs,a baseline feature of high AEC(≥0.125×109 cells/L)was associated with an increasing risk of CIP,and with a better clinical outcome.The conclusion that "patients with high peripheral AEC before ICIs treatment are at high risk for CIP" was written in the guidelines for toxicity management related to immune checkpoint inhibitors issued by CSCO in the 2021 edition.Part III: Predictive role and clinical application of autoantibodies in immunotherapy of advanced non-small cell lung cancer[Objective] This part was designed to determine a panel of TAAbs in plasma which can predict the clinical outcome of ICIs-based treatment in patients with advanced NSCLC and to investigate correlation between the TAAbs-penel and CIP.[Methods] 1.Baseline plasma of advanced NSCLC patients before ICIs treatment was collected,and the concentration of autoantibodies in peripheral blood was detected by ELISA.The cutoff optical density(OD)value for positive result was calculated as the average OD value of healthy control subjects plus twice the standard deviation.2.In the discovery cohort,43 TAAbs shown in previous studies with high correlation with lung cancer were detected to screen the ones related to ORR and PFS of ICIs.3.In the validation cohort,the predictive value of the screened autoantibodies on the efficacy and ir AE of ICIs was verified by the method of "double-blind"(that is,the clinical information was unknown to the testers and the clinical data collection personnel did not participate in the sample detection).[Results] 1.Baseline characteristics of study populationA total of 166 advanced NSCLC patients treated with ICIs were included into this part,of which 34 patients had EGFR mutation.The positive rate of TAAbs is about 50%.The positive rate of TAAbs in patients receiving ICIs as the first-line therapy was higher than that in patients who received ICIs as subsequent-line treatment(68.2% vs.38.8%,P=0.002).2.Determination of the reactivity of the 5-TAAbs panel in the discovery cohortIn the discovery cohort(n=37),43 TAA proteins were selected for testing,and five TAAbs(p53,BRCA2,HUD,trim21 and NY-ESO-1)related to the best efficacy of immunotherapy were selected.A group of 5-taab was evaluated and the test result was considered positive if at least one autoantibody was positive.In the discovery cohort,the PR rate of patients with positive 5-TAAbs panel was 47.4%,while the PD rate was only 15.8%.The positive rate of the 5-TAAbs panel was also related to better PFS(not reached vs.5.32 m,HR=0.383,95%CI: 0.140~1.046,P=0.05).3.Predictive value of the 5-TAAbs panel in the validation cohortIn the validation cohort(n=129),patients with positive 5-TAAbs panel had significantly higher ORR(44.4% vs.13.6%,P<0.001)and longer PFS(7.6 vs.3.3m,P<0.001).In the subgroup population receiving ICIs monotherapy,the positivity of this 5-TAAbs panel was associated with better efficacy(ORR: 46.7% vs.13.3%,P=0.003;PFS: 19.7 vs.2.2m,P<0.001).In addition,this 5-TAAbs panel remained a good predictive biomarker for patients receiving combination therapy(ORR: 43.8%vs.13.7%,P=0.004;PFS: 6.7 vs.3.7m,P=0.017).With regard to the treatment line,subgroup analysis showed that our 5-TAAs panel performed better in patients receiving ICIs-based treatment as subsequent lines(ORR: 42.4% vs.7.7%,P=0.001;PFS: 6.2 vs.3.0 m,P=0.004).4.Association between 5-TAAbs panel and CIPFor patients followed up for at least 3 months(n=122),the incidence of CIP was12.3%,and the median time of onset was 2.3 months.The incidence of CIP was significantly higher in 5-TAAbs panel positive patients than that in negative patients(20.4% vs.5.9%,P=0.015).However,for all patients diagnose with CIP,the grades were similar between 5-TAAbs positive and negative patients.[Conclusion] Our 5-TAAbs panel(p53,BRCA2,HUD,TRIM21 and NY-ESO-1)can is a potential predictive biomarker for responses and CIP to ICIs-based treatment in patients with advanced NSCLC.