Defining Late Onset Occult Asymptomatic Cardiotoxicity in Childhood Cancer Survivors Exposed to Anthracycline Therapy

Part A: Patients who received a high dose of anthracyclines and who completed chemotherapy for a minimum of 2 year and are still in remission we gathered information about their previous heart disease, radiation dose to the heart, bone marrow transplant, age at treatment, gender, elevation of serum cardiac troponin-T or N-terminal pro-brain natriuretic peptide measurements during anthracycline therapy, systolic dysfunction and congestive heart failure during anthracycline therapy. Participants had a Cardiac MRI (CMRI), echo and collection of serum for the the measurement of biomarkers of inflammation (C Reactive Protein (CRP), TNF-α, myocyte injury (Cardiac troponin (cTn), extracellular matrix remodeling, (Matrix Metalloproteinases (MMP), Tissue inhibitors of metalloproteinases (TIMPs), C terminal propeptide of type I procollagen (PICP), C terminal telopeptide of collagen type I (CITP), Bone Alkaline Phosphatase, apoptosis (Caspase-3 P17 peptide), and ventricular derived B-type natriuretic peptide (BNP) once during the study. Part A has been completed.

Part B: For those patients enrolled in part A they had another CMRI to validate CMRI findings and to study progression of cardiomyopathy overtime by comparing all CMRI derived imaging parameters between the first and second subject visits. Results of their standard of care echo will be post-processed to include the same parameters as the subjects' previous research echo that took place 2 years ago for comparison. Subjects had a blood draw (3mL) for for DNA and RNA analysis. Part B has been completed.

Part C: Will include an additional 60 patients who received low dose anthracycline therapy (<300 mg/m2) and who were also identified through the cardio-oncology registry. These patients will undergo a CMRI, study echo, and a blood sampling: Cardiac Troponin (cTn) , Matrix Metalloproteinases (MMP), Tissue inhibitors of metalloproteinases (TIMPs), C terminal propeptide of type I procollagen (PICP), C terminal telopeptide of collagen type I (CITP), Bone Alkaline Phosphatase, Caspase, C Reactive Protein (CRP), TNF-α, hematocrit, and NT-proBNP (a more sensitive measurement of BNP). Blood draw for DNA/microRNA analysis. Results of the standard of care echo will be post-processed to include the same parameters as the original 60 subjects' previous research echocardiogram for comparison purposes. All subjects recruited will serve either as a "case" (because they have cardiotoxicity) or as a control, if they are free of cardiotoxicity. Currently enrolling for part C patients who received low dose anthracyclines.

Cardiotoxicity for the overall group will be defined as global systolic dysfunction (indicated by an ejection fraction < 2SD below normal (Z score <-2), or decrease in global circumferential or longitudinal strain magnitude less than 2SD below normal. CMRI is considered to be the standard of care for assessment of cardiomyopathy. Based on the proven utility of CMRI as a screening tool for cardiomyopathy, CMRI studies will be performed as standard of care for assessment of global and regional myocardial function in cancer survivors. T1 pulse sequences for estimation of native T1 and ECV are continually being developed and refined, and will require further validation prior to clinical application. These sequences will still be considered be part of research.

STATISTICAL ANALYSIS Part A. CMRI parameters will be correlated with echo parameters of systolic and diastolic function. Data on peak circumferential strain will be compared to published historic normative data. T1 relaxation values will be compared to a group of age matched controls with normal SF and no exposure to anthracyclines. Results are expressed as mean and SD for continuous data and as percentages and numbers for categorical data. Subjects will be classified into 3 groups as detailed previously. Differences in the means between the groups and between exposed and controls for all normally distributed data will be assessed by analysis of variance. For skewed data, the Mann-Whitney U test will be performed. Image quality scores for all patients will be averaged for corresponding T1 maps. Mean segmental T1 and ventricular myocardial peak circumferential strain values and standard deviations will be calculated. The literature shows that T1 value is subject to a mild systematic heart rate dependency and a patient's height and weight has some impact on T1 values. To find and correct for these potential confounding factors, the relationships between T1 and heart rate, height, and weight will be investigated by using linear regression analysis of section data. In addition, the relationship between T1 mapping-derived relaxation time, left ventricular myocardial peak circumference strain and echocardiography parameters of systolic and diastolic function will be investigated by using linear regression analysis of section data. All statistical analysis will be performed with SAS 9.2. All tests are two tailed, and P < .05 is considered to indicate a significant difference. Interobserver and intraobserver variability of HARP strain will be performed using a paired t test and reported as a mean difference and 95% confidence interval.

Serum levels of MMP-1, MMP-2, MMP-9 and MMP-1/TIMP ratio, PICP, CITP, Bone Alkaline Phosphatase and that of Caspase-3 P17 peptide will be measured as indicators of extracellular matrix remodeling and apoptosis, respectively and compared to CMRI and echocardiographic parameters of systolic and diastolic function. Serum level of a cleaved p17 fragment (p17) of Caspase-3, the end effector caspase for apoptosis, will be compared to published normative data from a group of 167 healthy subjects 68. Circulating levels of MMP-1, MMP-2, MMP-9 and TIMP-1 will be compared to published historic healthy controls 69 Serum biomarkers N-terminal pro-BNP (NT-pro-BNP) and Cardiac troponin will also be obtained given their known association with cardiac dysfunction.

Biomarkers statistical analyses results are expressed as mean and SD for continuous data and as percentages and numbers for categorical data. Subjects will be classified into three groups as detailed previously. Differences in the means between the groups for all normally distributed data will be assessed by analysis of variance. For skewed data, the Mann-Whitney U test will be performed. The relationship between levels of Troponin, Caspase-3 P17 peptide, MMPS, TIMPs, PICP, CITP, Bone Alkaline Phosphatase, CRP, CK-MB, and echo parameters of systolic and diastolic function will be investigated by using linear regression analysis of section data. All statistical analysis will be performed with SAS 9.2. All tests are two tailed, and P < .05 is considered to indicate a significant difference.

The interdisciplinary nature and the potential for bedside to bench translational research should position us to obtain pilot data based on the projected changes in our patient cohorts. For example, we will test whether serum P17 level in any patient groups (groups 1-3) is higher than that in healthy controls. However, it is possible that only the late groups (Group 2 or 3) will show a rise in P17 level. Similarly, only the late groups may show changes in serum MMPs or T1 relaxation time. If any of the change is too small to be detected given the sample size, we will focus on parameters exhibiting larger differences vs. healthy controls. This will also guide us in designing future study. The change in MMPs may reflect presence of diastolic vs. systolic dysfunction in the early vs. late groups. For example, it is possible that the early group (Group 1) will exhibit diastolic problem, manifested as a reduced MMP-1/TIMP-1 ratio. When significant systolic dysfunction ensues, the ratio may be reversed.

Part B: All CMRI imaging biomarkers of cardiotoxicity demonstrated in the baseline pilot study will be compared among the first and second study visits among the high dose (≥240 mg/m2) subjects. Frequencies and percents of the patients who exhibit the phenotype of cardiotoxicity as measured by cMRI will be calculated for each study visit. If the frequencies/percents are the same or greater in the second visit as compared to the first, then the use of cMRI to indicate anthracycline induced cardiotoxicity can be validated.

Specific Aim 2: All CMR imaging parameters of athracycline induced cardiomyopathy will be compared between the first and second study visits among high dose (≥240 mg/m2) subjects. All CMR volumetric parameters including volumes, mass, and measurements of end systolic fiber stress as well as measurements of global and regional myocardial function, T1 maps and ECV will be compared between the two study visits using paired t-tests, p<0.05 will be used to indicate a statistically significant difference.

Part C:

Specific Aim 1: To compare CMR parameters of global, and regional systolic function and microscopic fibrosis as indicated by changes in T1 mapping-derived relaxation time and increase in extracellular volume matrix (ECV) to those of subjects in the high-dose group.

Overview: The objective of this aim is to test the hypothesis that CMR imaging biomarkers of AIC will be less frequent among subjects exposed to low dose anthracyclines. The proportion of subjects that possess the anthracycline-induced cardiotoxicity phenotype in both the high and low dose groups will be compared using Fisher's exact test. P< 0.05 will be used to indicate a statistically significant difference.

Specific Aim 2: To quantitate serologic biomarkers of inflammation, myocyte injury, extracellular matrix remodeling and apoptosis among childhood cancer survivors treated with low-dose anthracyclines.

This aim will tell the hypothesis that elevation of biomarkers of inflammation, myocyte injury, extracellular matrix remodeling and apoptosis will be less prevalent in survivors of childhood cancer exposed to low-dose anthracyclines, as reflected by lower levels of serum CRP, high sensitivity troponin, MMPs and Caspase-3 P17 levels.

Overview: The objective is to characterize changes in serum levels of markers for extracellular matrix remodeling and apoptosis in patients at various intervals after anthracycline chemotherapy. Serum levels of MMP-1, MMP-2, MMP-9 and MMP-1/TIMP ratio, PICP, CITP, Bone Alkaline Phosphatase and that of Caspase-3 P17 peptide will be measured as indicators of extracellular matrix remodeling and apoptosis, respectively and compared to CMRI and echocardiographic parameters of systolic and diastolic function. Serum level of a cleaved p17 fragment (p17) of Caspase-3, the end effector caspase for apoptosis, will be compared to published normative data from a group of 167 healthy subjects (Agosto, Azrin, Singh, Jaffe, \& Liang, 2011). Circulating levels of MMP-1, MMP-2, MMP-9 and TIMP-1 will be compared to normal controls. Serum biomarkers N-terminal pro-BNP (NT-pro-BNP) and Cardiac troponin will also be obtained given their known association with cardiac dysfunction. A literature review of six small studies suggest that NT-pro-BNP, and Cardiac troponin might also be useful markers in the early detection of anthracycline-induced cardiotoxicity (Mavinkurve-Groothuis, Kapusta, Nir, \& Groot-Loonen, 2008). Of interest, in a larger study performed in 122 asymptomatic survivors of childhood cancer, none had abnormal cTnT levels and only thirteen percent had abnormal NT-pro-BNP levels (Mavinkurve-Groothuis et al. 631-36).

Specfic Aim 3: To compare the results of the DNA/RNA analysis between all subjects with and without anthracycline induced cardiomyopathy regardless of cumulative dose. DNA and RNA analysis will be collected for future analysis by the Jackson Lab bioinformatics core unit, to be analyzed based on the results of a parallel murine model of cardiotoxicity.