| Part I Circulating micro RNAs as diagnostic and prognostic biomarkers for ALK-positive lung cancerBackground Micro RNA is widely exited and stably expressed in human body fluids.Aberrant levels of circulating micro RNAs are potential biomarkers for diagnosing tumors early and predicting therapeutic response.However,there has been no relevant research on patients with anaplastic lymphoma kinase(ALK)-positive lung cancer.We sought to identify potential plasma biomarkers for ALK-positive non-small cell lung cancer(NSCLC)and to find the correlation.Another aim of this study was to investigate the correlation between plasma mi RNA levels and the efficacy of crizotinib in advanced ALK-positive NSCLC patients treated with crizotinib.Furthermore,we sought to explore the correlation between the dynamic changes of plasma mi RNA expression levels and the efficacy and prognosis of patients with ALK-positive advanced NSCLC treated with crizotinib.Methods 1.We used the microarray method to detect micro RNA expression from three groups of plasma samples,including three ALK-positive advanced NSCLC patients,three ALK-negative advanced NSCLC patients,and three healthy subjects.This may allow us to understand the differences in plasma micro RNA expression between lung cancer patients and healthy subjects,and the differences in plasma mi RNA expression between ALK-positive NSCLC patients and ALK-negative NSCLC patients.The ALK gene status of all enrolled patients in this study was detected by fluorescence in situ hybridization(FISH).2.We used the real-time reverse transcriptase quantitative polymerase chain reaction(RT-q PCR)method to detect micro RNA expression levels in plasma from five group of subjects,including eight healthy subjects,eight lung benign disease patients,six squamous lung cancer patients,six adenocarcinoma lung patients,and six small cell lung cancer patients.We used ge Norm and Norm Finder software to screen out the most stable gene from the 12 candidate internal reference genes as the internal reference gene.The 12 candidate internal reference genes include U6,5s r RNA,18 s r RNA,RNU38 B,RNU43,RNU44,RNU48,mi R-93-5p,mi R-103-3p,mi R-191-5p,mi R-197-3p,and mi R-221-3p.3.We used RT-q PCR method to detect micro RNA expression levels in plasma from 41 ALK-positive advanced NSCLC patients and 32 ALK-negative advanced NSCLC patients.We validated the difference of mi RNA expression in plasma between ALK-positive advanced NSCLC patients and ALK-negative advanced NSCLC patients.4.We assessed the sensitivity and specificity of micro RNAs for ALK gene status diagnosis by the area under the receiver operating characteristic curve(ROC).We calculated the optimal cut-off values by the Youden’s index and calculated the positive predictive value(PPV)and negative predictive value(NPV)for each mi RNA and mi RNA combination.5.We used mi RDB,mi Randa,and Target Scan website to predict the target genes of each mi RNAs,and used KEGG pathway to predict the related signaling pathways.6.We used Kaplan-Meier survival assay to analyze the correlation between plasma mi RNA expression levels of ALK-positive NSCLC patients(n=31)and progression-free survival(PFS)pre-crizotinib.At the same time,the relevant clinical pathological features of multivariate were analyzed.7.We used RT-q PCR method to detect the level of plasma micro RNA expression in patients with ALK-positive advanced NSCLC treated with crizotinib.We followed the efficacy and PFS of these patients to explore the correlation between the dynamic changes of plasma mi RNAs and the efficacy of crizotinib during the treatment period.Results 1.As a result,more micro RNAs were detected in the plasma of the healthy subjects than in the plasma of the NSCLC patients.409 micro RNAs were detected in the plasma of healthy subjects,whereas 116 micro RNAs in the plasma of ALK-positive NSCLC patients and 209 micro RNAs in the plasma of ALK-negative NSCLC patients.The plasma levels of 21 micro RNAs were differentially expressed for ALK-positive and ALK-negative NSCLC(fold chang >2 or < 0.5),including 14 down-regulated and 7 up-regulated micro RNAs.2.We identified 5s r RNA as the most stable endogenous control gene using ge Norm and Norm Finder algorithms.Based on calculations performed with ge Norm,the expression levels of 5s r RNA and 18 s r RNA were more stable than other candidate genes.Based on Norm Finder,the stability of endogenous candidate genes from high to low were: 5s r RNA,18 s r RNA,U6,RNU43,mi R-103-3p,mi R-197-3p,mi R-93-5p,RNU38 B,RNU48,mi R-191-5p,RNU44,and mi R-221-3p.Thus,5s r RNA was identified as the most stable endogenous gene in the present study.3.The levels of three micro RNAs were significantly different between ALK-positive NSCLC patients and ALK-negative NSCLC patients,including mi R-28-5p,mi R-362-5p,and mi R-660-5p(p<0.0001,p=0.0091,and p=0.0030,respectively).The three micro RNAs were all down-regulated in ALK-positive NSCLC patients.4.The area under the curve(AUC)of mi R-28-5p,mi R-362-5p,and mi R-660-5p of diagnosis for ALK status were 0.873(95%CI,0.795-0.952),0.673(95%CI,0.549-0.796),and 0.760(95%CI,0.649-0.870),respectively.The AUC of combination of mi R-28-5p,mi R-362-5p,and mi R-660-5p was 0.876(95%CI,0.801-0.951).The sensitivity of mi R-28-5p,mi R-362-5p and mi R-660-5p to ALK gene positive were 65.8%,51.22% and 63.41%,the specificity were 96.88%,80.77% and 83.87%,PPV were 96.43%,80.77% and 83.87%,NPV were 68.89%,57.45% and 64.29%.Among the three micro RNAs,mi R-28-5p was the most valuable.5.Mi R-28-5p may be involved in the ALK downstream pathway m TOR and PI3K-AKT by targeting GSK3 B.Mi R-660-5p may be involved in the PI3K-AKT pathway by targeting Bcl-2,MDM2,IRS1,and ECM,involved in the MAPK pathway by targeting Elk-1,Rap1,and CACN,involved in the m TOR pathway by targeting IRS1 and Rictor.6.Kaplan-Meier survival analysis showed that the median PFS of 31 patients with ALK-positive NSCLC treated with crizotinib was 6.0 months(range,1-28 months).The expression level of mi R-362-5p has a predictive effect on the PFS of crizotinib.The PFS of the patients with lower expression levels of mi R-362-5p were significantly longer than that with higher levels,HR=0.285(95%CI,0.104-0.780),p=0.015.The expression levels of mi R-28-5p and mi R-660-5p were not significantly correlated with PFS(p = 0.712,0.769,respectively).7.We analyzed the changes of micro RNAs level in plasma pre-crizotinib and post-crizotinib using Wilcoxon matched-paired signed ranks sum test.There were total 19 patients,including 16 patients acquired partial response(PR),and 3 patients acquired progressive disease(PD).In the 16 patients with PR,the levels of mi R-660-5p were increased in 15 patients(p=0.001).There had no the similar trend in the patients with PD.The changes of mi R-28-5p and mi R-362-5p levels in plasma had no correlations with the crizotinib efficacy(p=0.776 and 0.796).Conclusions 1.There were more micro RNAs in plasma with lung cancer patients than health subjects by microarray methods.There were 409 micro RNAs in plasma with health subjects cohort,116 micro RNAs in plasma with ALK-positive NSCLC cohort,and 209 micro RNAs in plasma with ALK-negative NSCLC cohort,respectively.2.Using microarray method,we identified 21 micro RNAs that were differentially expressed in plasma between patients with ALK-positive NSCLC and those with ALK-negative NSCLC,and validated the results with RT-q PCR method.The results showed that the differential expression of plasma mi RNAs by microarray was a feasible and reliable method.3.This study showed that the expression levels of mi R-28-5p,mi R-362-5p and mi R-660-5p were significantly down-regulated in plasma of ALK-positive advanced NSCLC.The AUC analysis shows that the three micro RNAs and the 3-micro RNA combinations showed high sensitivity and specificity.The plasma micro RNAs had the ability to be the potential diagnostic biomarkers for ALK status with NSCLC patients.4.In the present study,we analyzed the dynamic changes of micro RNAs in plasma with ALK-positive NSCLC patients pre-crizotinib and post-crizotinib,and we identified that mi R-660-5p was significantly correlated with the efficacy of crizotinib,increased expression levels of mi R-660-5p demonstrated longer PFS.This suggests that plasma micro RNAs are expected to be a predictor of the efficacy of crizotinib.5.The expression level of mi R-362-5p in plasma pre-crizotinib has a predictive effect on PFS of crizotinib.Patients with low expression levels of mi R-362-5p had longer PFS.Part II Analysis of the therapeutic efficacy of crizotinib on ALK-positive advanced non-small cell lung cancerBackground In NSCLC patients,the prevalence of ALK rearrangement was reported to be 3% to 7%.Crizotinib,as the first ALK tyrosine kinase inhibitor(TKI),plays an important role in the treatment of ALK-positive advanced NSCLC.The clinical data on the treatment of crizotinib were mostly from international multi-institution clinical studies,of which the proportion of patients in China is limited.There was little research on the treatment of crizotinib in Chinese lung cancer patients.The correlation between brain metastases and extracranial metastases during the treatment of crizotinib is rarely reported.In Chinese patients with ALK-positive NSCLC,the treatment of crizotinib,population selection,the impact of brain metastases are to be verified by clinical data.This study aims to observe the efficacy,PFS,and overall survival(OS)of crizotinib in patients with ALK-positive advanced NSCLC in China,analyze the correlation between the crizotinib administration,brain metastases at crizotinib-baseline with PFS and OS,and explore the correlation between PFS-intracranial and PFS-systemic.In addition,clinical prognostic factors were identified by Cox proportional hazards modeling.Methods We conducted a single-institution,retrospective,observational study.A total of 91 patients with ALK-rearrangement who were treated with crizotinib were enrolled.Patients clinical and treatment information were extracted from electronic medical records at our institution.We observed the objective response rate(ORR),disease control rate(DCR),PFS,and OS of all enrolled patients.The difference of ORR,DCR,PFS,and OS between first-line and second and further-line treatment of crizotinib were compared.The correlation between brain metastases status at crizotinib baseline and PFS and OS were evaluated.PFS-intracranial and PFS-systemic were observed in the patients who accuired brain metastases during treatment with crizotinib.The correlation between PFS-intracranial and PFS-systemic were evaluated.The clinicalpathological factors were identified by Cox proportional hazards modeling.Results 1.A total of 91 ALK-positive NSCLC patients with a median age of 48 years(range,24-77 years)were evaluated for the present study,including 43 male patients(47.3%)and 48 female patients(52.7%),PS score were all 0-1.Among the 91 patients,most of them were adenocarcinoma patients(88/91,96.7%),G3(36/91,39.6%)and Gx(43/91,47.3%)differential grade,with no or light smoking history(83/91,91.2%),detected with FISH(80/91,87.9%).28 patients(30.8%)were treated with crizotinib for first-line treatment,30 patients(33.0%)for second-line treatment,33 patients(36.3%)for third and further-line treatment.There is no patient with complete response(CR),65 patients with PR(71.4%),18 patients with stable disease(SD)(19.8%),8 patients with PD(8.8%).As of the last visit,there were 34 patients(37.4%)dead,57 patients(62.6%)were alive or lost of visit.2.The ORR and DCR in all 91 patients were 71.4% and 91.2%.The ORR in the first-line(78.6%)setting was higher than that in the second-line setting(68.3%),the DCR in the first-line setting(96.4%)was also higher than that in the second-line setting(88.9%),the difference was no significant(p=0.445).3.The median PFS of crizotinib in all patients was 9.5 months(range,1.0-65.0 months),median OS was 45.0 months(range,6.0-84.0 months).4.At baseline,a total of 33 patients(36.3%)had brain metastases,and 58 patients(63.7%)had no brain metastases.The median PFS in the brain metastases setting and no brain metastases setting were 11.0 months and 9.0 months(p=0.9106),median OS were 45.0 months and 44.0 months(p=0.6695).5.The median PFS in the first-line setting and second and further-line setting were 18.0 months and 8.0 months,hazard ratio was 0.5701(95% CI: 0.3340-0.9450),p=0.0293.The median OS in the first-line setting was not reached,that in the second and further-line setting was 44.0 months,hazard ratio was 0.7367(95% CI: 0.3262-1.664),p=0.4623.6.At baseline,a total of 58 patients had no brain metastases,of whom 20 patients acquired brain metastases during the treatment of crizotinib.Among them,there were 12 patients with intracranial PD,followed by extracranial PD,the other 8 patients with intracranial,extracranial synchronization PD.The median PFS-systemic was 10.0 months(95% CI: 7.61-23.56 months),median PFS-intracranial was 4.5 months(95% CI: 1.96-14.54 months),median PFS-expanded was 6.5 months(95% CI: 3.53-11.14 months).There was no significant difference in PFS-intracranial and PFS-expanded(p=0.906).7.Multivariate analysis was performed using the Cox proportional hazards regression model,treatment line of crizotinib usage(HR=2.357,95% CI: 1.208-4.598,p=0.012)were independent predictor for PFS.Other clinicopathologic factors were no significantly different with PFS and OS(p>0.05).Conclusions 1.In the total 91 patients with ALK-positive advanced NSCLC,more patients were younger,adenocarcinoma and non-smokers.2.The ORR and DCR in all 91 patients were 71.4% and 91.2%.The ORR in the first-line setting was higher than that in the second-line setting,the DCR in the first-line setting was also higher than that in the second-line setting,the difference was no significant.3.The median PFS and OS in all 91 patients were 9.5 months and 45 months.4.There was no significant correlation between brain metastases in baseline with PFS and OS.5.There was significant correlation between treatment-line of crizotinib and PFS,the earlier the treatment,the longer the PFS.The median PFS of first-line crizotinib was 10.0 months longer than second and further-line.There was no significant correlation between treatment-line of crizotinib and OS.6.For patients without brain metastases at baseline developing brain metastases during crizotinib treatment,the median PFS-systemic was 10.0 months,the median PFS-intracranial was 4.5 months,the median PFS-expanded was 6.5months.There was no significant correlation between PFS-intracranial and PFS-expanded.The PFS-expanded could not be predict by PFS-intracranial.7.Multivariate analysis showed that the earlier the treatment-line of crizotinib was,the longer the PFS was. |
PDF Full Text Request