Strain Echocardiography to Predict Cardiotoxicity in Patients Receiving Chemotherapy Containing Doxorubicin

Primary Objective:

1. Determine if a 20% reduction in global longitudinal strain by echocardiography can predict subsequent composite cardiotoxicity in a collection of patients with either HER2(-) breast cancer, sarcoma, or lymphoma receiving modern chemotherapy regimens containing anthracyclines.

Secondary Objectives:

1. Determine if a troponin I level of at least 0.08 ng/mL using high-sensitivity cardiac troponin I assay (Hs-cTn-I) can predict composite cardiotoxicity.
2. Determine if a serum B-type natriuretic peptide (BNP) level of at least 200 pg/mL can predict composite cardiotoxicity.
3. Determine independent predictors of cardiotoxicity among clinical parameters, serial echocardiography strain measurements, and biomarkers.

Exploratory Objectives:

1. Determine if various pharmacogenomic factors can predict composite cardiotoxicity.
2. Determine independent predictors of cardiotoxicity among clinical parameters, serial echocardiography strain measurements, biomarkers, and pharmacogenomics markers.

Study Procedures:

Enrollment will not begin until all regulatory approvals have been obtained. Potential patients will be identified by the treating oncologist. Eligibility criteria must be confirmed by the site staff. Patients who meet the eligibility criteria will be registered with the Indiana University Simon Cancer Center Clinical Trials Office (IUSCC-CTO) prior to first dose of planned chemotherapy. Applicable regulatory documents and registration must be completed and on file prior to the registration of any patient. A signed informed consent, completed eligibility checklist, and supporting source documentation will be sent to IUSCC-CTO for eligibility verification.

Written informed consent will be obtained from all subjects before initiation of any study-specific procedures. Procedures performed as part of the subject's routine clinical management and obtained prior to signing the informed consent may be utilized for screening or baseline purposes provided the procedure was performed within the timeframe specified in the protocol.

Echocardiograms with strain will be performed to determine cardiac structure and mechanical function during the course of treatment. All echocardiograms will include the addition of a new signal (sound wave) processing method, called strain echocardiography. This technology allows precise analysis of the individual regions of the heart to determine cardiac function. Echocardiograms with strain will be done on four visits using GE strain-capable technology. Echocardiograms will be performed in a manner compatible with strain calculation by either GE or Tomtec software depending on which software the echocardiography lab has experience using. The baseline echocardiogram will be within a window of 28 days prior to starting doxorubicin-based chemotherapy regimen. The second will be obtained between 1-4 weeks after receiving a cumulative dose of at least 240mg/m2 and before another dose is given. The third echocardiogram is obtained 6 months (+/-4 weeks) after the start of C1D1. The final echocardiogram will be obtained 12 months (+/-4 weeks) after the start of C1D1.

The first and last echocardiogram will be complete studies with strain calculation and 3D volumes. The middle two studies (after receiving at least a cumulative 240mg/m2 of doxorubicin and at 6 months after the start of treatment) will be limited studies consisting of seven 2D images (PLAX, APLAX, 4CH, 2CH, SAX-base, SAX-mid, SAX-apex) with strain calculation.

Blood pressures will be recorded by the sonographers at the time of every exam.

Strain analysis at IUH facilities will be done online (i.e. on-cart) by the sonographer at the conclusion of each echocardiogram. Strain analyses at other institutions may be done off-line if the institution has the necessary software. As is standard for strain studies, intraobserver and interobserver reproducibility of strain measurements will be each be performed on a portion of the test population. 20-25% of the patients' will have reproducibility studies performed on their initial echocardiograms. For IUH studies, following the initial examination sonographers will recalculate strain on the same images or they will have another strain sonographer recalculate strain on the same images. Efforts will be made to blind the 2nd measurer to the results of the previous strain calculations. For non-IUH facilities, this same process may be performed offline using either sonographers or staff echo readers.

A maximum of 11 mL of blood will be drawn for biomarkers at visits up to 28 days prior to beginning chemotherapy, after receiving at least a cumulative 240 mg/m2 of doxorubicin, 6 months (+/-) 4 weeks after starting chemotherapy, and 12 months (+/-) 4 weeks following starting chemotherapy. Serum Troponin I and BNP levels will be drawn at the same intervals as the echocardiograms.

Blood samples will also be obtained from patients for pharmacogenomics. Preferably these samples will be drawn at approximately the same time as the second echocardiogram/second set of blood draws, but can be obtained at any time during the study.

Pharmacogenomic studies will be largely exploratory and will consist of TaqMan SNP genotype assays for genotyping single nucleotide polymorphisms (SNPs) annotated in UGT1A6, ABCB4, ABCC1, HNMT, SLC28A3, FMO2, SPG7, SLC10A2, SLC28A3 using germline DNA samples from the patients. In addition, we will employ next generation sequencing (NGS) approach using the TruSight One Sequencing Panel based on the Nextera Enrichment DNA Sample Prep kit (Illumina, San Diego, CA) which targets 4,813 genes at a 100X average coverage, and at 20X minimum coverage of 95% regions in the panel. This targeted enrichment approach will focus on genes such as, but not limited to HAS3, CBR1, CBR2, involved in cardiotoxicity, as well genes coding for proteins involved in heart cytoskeleton and metabolism. Other high-throughput approaches may also be used.

An electronic medical record review will be done by a research staff member for each patient enrolled in this study. The following information will be collected from the medical record prior to enrollment:

• demographics including age, sex, race and ethnicity
• diagnosis including stage of disease
• previous treatments including radiation therapy to the mediastinum
• previous medical and concomitant medication history
• performance status
• height and weight
• planned (or previous) chemotherapeutic regimen including dose (mg/m2) of doxorubicin
• extent of previous radiation therapy including site of radiation and dose delivered

Upon completion of the final echo, interim information will be collected by chart review of subsequent visits and will include the following:

• Vitals
• Height and weight
• Total amounts of all chemotherapy drugs given
• Medication changes
• ECOG or Karnofsky Performance Status
• New radiation therapy (if applicable)
• Presence of new medical conditions such as hypertension, hyperlipidemia, or tachycardia

A complete physical exam including vital signs, weight, and a measure of activity level will be completed at all visits. Height will be collected at the pre-treatment visit only. Body surface area will be calculated at each visit by the PI or his co-investigator to ensure a consistent equation is used. All concomitant medications will be recorded at each visit except the annual follow-up visit.

Follow-up will be obtained by the investigator and/or study staff each year beginning at 12 months (+/- 2 months) after C1D1 for up to 5 years. Follow-up information will be obtained from the patient's electronic medical record, by calling the patient's treating or primary care physician, or by contacting the patient directly.

Sample Size:

The primary objective is to show that change of strain echocardiography parameters (between "after 240mg/m2" and baseline) is associated with the 1-year event rate of the primary endpoint. Based on existing literature, the overall 1-year event rate is about 8% in the targeted cancer population. We projected that 30% of the population will experience deteriorated echocardiography parameters (the risk group) and the rest 70% (the normal group) will have stable values on these parameters. Previous studies have shown that the normal group very close to zero event rate. Assuming that the event rate for normal group is 3% and thus the event rate for the risk group is 20%, a total of 128 subjects allows us to have 85% power to detect the difference with type I error rate (one-sided) is controlled at 0.05. Alternatively, if the overall event rate is as low as 5 % with 1% in the normal group and 15% in the risk group, using χ2 test we will still have 85% power to detect the difference. With 15% attrition rate, we will recruit 128/0.85=150 patients. With a projected enrollment of 6 patients per month, we anticipate the recruitment period to be 25 months.

Statistical Analysis:

Continuous baseline variables will be summarized by typical parameters such as mean, standard deviation and range and compared using two-sample T test on the at-risk and non-risk groups (if the normality assumption holds) or Wilcoxon rank-sum test (if the normality assumption does not hold). Normality of distribution will be determined using the Kolmogorov-Smirnov goodness-of-fit test. Categorical data will be summarized by frequency and percentage and analyzed using the Chi-square or Fishers exact test, as appropriate. Logistic regression will be used to analyze the association of change (between "after 240 mg/m2" and baseline) of echocardiography parameters, biomarkers (BNP and troponin) and pharmacogenomics markers with the primary endpoint, separately. Pharmacogenomics measures will be considered exploratory. Logistic regression will be used to study dosage effect on the primary endpoint. Multi-variable logistic regression will be used to identify independent predictors among echocardiography parameters, biomarkers and pharmacogenomics markers. If the event rate is only 8%, we will not be able to perform a multivariate analysis. However, this event rate is highly variable, and a substantially higher event rate (from 16-20%) has been noted in more recent studies. If the event rate is only 8%, we will not be able to perform a multivariate analysis, but if it is 16-20%, we should be able to fit some multivariable models without overfitting. Hence, we will only consider performing a multivariate analysis if the event rate is at least 16-20%. All analysis will be performed in SAS and SPSS. Strain cut-off levels defining risk vs. non-risk will be a 20% relative decrease in global longitudinal strain. For the secondary objectives, the cutoff value for troponin will be ≥ 0.08 ng/mL at any time. For BNP the cutoff value will be ≥ 200 pg/mL at any time.

Data Safety Monitoring Plan:

Investigators will conduct continuous review of data and patient safety. Quarterly review meetings for low risk trials are required and will include the PI and other members per PI's discretion. Quarterly meeting summaries will be reviewed by the full IUSCC Data Safety Monitoring Committee at the time of study review.