Assessment And Prediction Of PD-1/PD-L1 Inhibitors Associated Early Cardiotoxicity In Patients With Lung Cancer Using Cardiac Magnetic Resonance

Part I Assessment of early cardiotoxicity in patients with lung cancer receiving programmed death-1 and its ligand(PD-1/PD-L1)inhibitors by cardiac magnetic resonance Objective To investigate the feasibility and value of cardiac magnetic resonance(CMR)imaging in the evaluation of early cardiotoxicity associated with programmed death-1 and its ligand(PD-1/PD-L1)inhibitors in patients with lung cancer.Methods This study prospectively recruited patients with lung cancer who were ready to receive PD-1/PD-L1 inhibitors.Patients were planned to undergo CMR examinations before the initiation of therapy(baseline),before the second therapy(the first follow-up),and after several times of therapy(3 months after the initiation of therapy;the second follow-up),respectively.Sixty-six patients who received the first follow-up were included for analysis,and 38 patients of them received the second follow-up.Qualitative analysis and quantitative post-processing of CMR images were performed.Quantitative parameters included: left ventricular ejection fraction(LVEF),right ventricular ejection fraction(RVEF),left ventricular global radial strain(LV-GRS),left ventricular global circumferential strain(LVGCS),left ventricular global longitudinal strain(LV-GLS),right ventricular global radial strain(RV-GRS),right ventricular global circumferential strain(RV-GCS),right ventricular global longitudinal strain(RV-GLS),average left atrial long-axis strain,right atrial longaxis strain,native T1 value,post-contrast T1 value,extracellular volume(ECV)value,T2 value,and epicardial adipose volume(EATV)value.Mixed linear models were performed to test for longitudinal changes in CMR parameters over time,and sidak adjustment was utilized for pairwise comparisons of the three time points.Chi-square test or Fisher’s exact test were applied to compare categorical variables.The change value of CMR quantitative parameters was defined as the difference between the follow-up value and the baseline value,and the change ratio was defined as the ratio of the change value to the baseline value.Results At the first follow-up(a median time of 22 days after the first therapy)and the second follow-up(a median time of 86 days after the first therapy),increased pericardial effusion was observed in 2(3.0%)patients and 3(7.9%)patients,respectively.Late gadolinium enhancement manifestations did not change significantly after ICI therapy.Compared with baseline values,LVEF(52.2% ± 6.0% vs.49.2% ± 6.0%)and LV-GRS(41.1% ± 10.8% vs.39.3% ± 13.3%)significantly decreased at the second follow-up(P < 0.05).LV-GCS(-19.6%± 2.5% vs.-18.8% ± 3.0% vs.-18.7% ± 2.4%)at two follow-up CMRs significantly decreased(P < 0.05).Compared with baseline values,RVEF(46.4% ± 7.1% vs.43.3% ±7.6% vs.40.8% ± 9.1%),RV-GRS(33.9% ± 12.8% vs.30.7% ± 11.1% vs.26.3% ± 9.5%)at two follow-up CMRs were significantly reduced(P < 0.05).Average left atrial long-axis strain didn’t show significant difference compared with baseline values(P > 0.05),while right atrial long-axis atrain,and EATV value showed decreasing trends after receiving PD-1/PD-L1 inhibitors,and the differences were significant compared with baseline values(P< 0.05).During follow-ups,tissue characteristic parameters of global left ventricular myocardium(native T1 value,post-contrast T1 value,ECV value,T2 value)showed no significant differences compared with baseline values(P > 0.05).Compared with baseline,the change ratios of LVEF at two follow-up CMRs both showed significantly positive correlations with the change ratios of LV-GRS,LV-GCS,and LV-GLS(P < 0.05).Conclusions PD-1/PD-L1 inhibitors related cardiotoxicity in patients with lung cancer mainly manifested as subclinical cardiac dysfunction in the early stage.LV-GCS can detect the impairment of left ventricular systolic function ealier than LVEF.Tissue characteristic parameters of left ventricular myocardial did not show significant changes in the subclinical stage after receiving PD-1/PD-L1 inhibitors.Part II The value of cardiac magnetic resonance combined with clinical features in predicting early cardiac dysfunction in patients with lung cancer receiving programmed death-1 and its ligand(PD-1/PD-L1)inhibitorsObjective The aim of this study was to explore the value of baseline cardiac magnetic resonance (CMR)parameters,clinical features,and changes in CMR parameters in predicting early cancer therapy-related cardiac dysfunction(CTRCD)after receiving programmed death-1 and its ligand(PD-1/PD-L1)inhibitors.Methods This study prospectively recruited patients with lung cancer who were ready to receive PD-1/PD-L1 inhibitors.Patients were planned to undergo CMR,12-lead electrocardiogram(ECG)and laboratory examinations before the initiation of therapy(baseline),before the second therapy(the first follow-up),and after several times of therapy(3 months after the initiation of therapy;the second follow-up),respectively.Sixty-six patients who received the first follow-up were included for analysis,and 38 patients of them received the second follow-up.Quantitative parameters included: left ventricular ejection fraction(LVEF),right ventricular ejection fraction(RVEF),left ventricular global radial strain(LV-GRS),left ventricular global circumferential strain(LV-GCS),left ventricular global longitudinal strain(LV-GLS),right ventricular global radial strain(RV-GRS),right ventricular global circumferential strain(RV-GCS),right ventricular global longitudinal strain(RV-GLS),average left atrial long-axis strain,right atrial long-axis strain,native T1 value,post-contrast T1 value,extracellular volume(ECV)value,T2 value,and epicardial adipose volume(EATV)value.CTRCD was defined as a reduction of more than 10% to less than 55% in LVEF compared to baseline.At the two follow-up CMRs,patients were divided into CTRCD group and non-CTRCD group,respectively.Baseline CMR parameters,clinical features,and changes in CMR parameters were further compared between groups.The change value of CMR quantitative parameters was defined as the difference between the follow-up value and the baseline value,and the change ratio was defined as the ratio of the change value to the baseline value.Multivariate logistic regression was performed using a stepwise method(backwards).Receiver operator characteristic curve was drawn and area under the curve(AUC)was calculated to assess the ability of parameters to predict CTRCD.Results The abnormal ECG results and laboratory indexes did not change significantly after receiving PD-1/PD-L1 inhibitors(P > 0.05).CTRCD was observed in 15.2%(n=10)and 26.3%(n=10)patients at two follow-up CMRs,respectively.At the first follow-up,the proportion of patients with body mass index(BMI)≥ 25 kg/m2 in CTRCD group(60.0% vs. 17.9%,P = 0.040)and baseline EATV values(84.7 ml ± 25.7 ml vs.66.1 ml ± 24.8 ml,P = 0.034)were both higher than those in non-CTRCD group.The change ratios of LV-GRS and LV-GCS showed differences between CTRCD group and non-CTRCD group at both time points during follow-up(P < 0.05).Parameters with significant differences between two groups at the two follow-ups were included into the multivariate regression analysis models.And the results showed that the change ratio of LV-GRS at both time points during follow-up showed as an independent predictor of early CTRCD,and had potential diagnostic ability for CTRCD(AUC values were 0.778 and 0.704,P values were 0.005 and 0.047,respectively).Conclusions Patients with lung cancer can develop CTRCD in the early stage of ICI therapy.The change ratio of LV-GRS was an independent predictor of CTRCD after ICI therapy within three months.