Optimising Screening for the Cardiotoxic Effects of Cancer Therapy

BACKGROUND Cancer is a major worldwide health problem, with around 330 000 new cases diagnosed annually in the UK alone. Improved survival rates due to better treatments have led to an increase in the number of cancer survivors, with over 60% of patients diagnosed expected to survive 5 years. There is also increasing recognition of the cardiotoxic effects of many of the cancer therapies (particularly anthracyclines and Trastuzumab), with subclinical cardiotoxicity found in 18% of subjects receiving anthracyclines of whom one-third have clinically overt heart failure. With trastuzumab, rates of heart failure are even higher, with more than 15% of subjects having to suspend or discontinue treatment. Early detection of cardiotoxicity however permits treatment and (particularly with trastuzumab) recovery of left ventricular function. Patients should therefore have their cardiac function monitored before, during and after treatment (as per published recommendations) to permit early detection and treatment of resultant cardiotoxicity.

The key parameter used for screening for myocardial effects of chemotherapy is the ejection fraction. Cancer therapeutics related cardiac dysfunction is defined by an asymptomatic reduction >= 10% points of the left ventricular ejection fraction (LVEF) to a value <55% or a reduction of LVEF <= 5% to 55% with symptoms of heart failure7. Conventionally this has been measured by nuclear techniques such as Multiple Gated Acquisition Cardiac Blood Pool imaging (MUGA). More recently, echocardiography or cardiac MRI have also been used to measure ejection fraction, as they are considered to be able to detect cardiac dysfunction at an earlier stage.

Each of the three cardiac imaging modalities has its strengths and limitations:

Transthoracic echocardiography (TTE) is widely used due to its availability, ease of use and absence of radiation. It has the advantage of evaluating not only systolic, but also diastolic and longitudinal function, valve and pericardial disease. However the poor test-retest reproducibility with 2D echocardiography (2DE) reported in previous studies raises concern about the accuracy of the technique, and the risk of erroneously stopping chemotherapy due to changes on LVEF that are only due to measurement noise. Indeed, some studies showed disappointing results, with 11% being the smallest change in EF that can be recognized with 95% confidence. Its limitation of relying on 2D geometrical assumptions in a single plane to estimate ventricular ejection is overcome by three dimensional echocardiography (3DE) which acquires more precise anatomic datasets and has been shown to be more accurate than 2DE for LVEF and volumes.

MUGA, is also a 3D technique that has consistently outperformed 2DE with respect to accuracy, reproducibility, intra and inter-observer variabilities for LVEF measures . It also has the advantage of few technical limitations (obesity, pacemakers and poor windows don´t affect image quality). However, radiation exposure is a common concern. 20-30 mCi of Tc99 exposes patients to 5-10 mSv of radiation, although the linkage of such low levels to increase cancer risk has not been shown.

CMR, the reference standard for cardiac function and volumes, also assesses myocardial mass, pericardium and detects subclinical cardiac disease. However, its high cost, low availability and long exam duration preclude its use for serial follow up. The current general recommendation is to use CMR when considering stopping chemotherapy.

For all three modalities, since therapeutic decisions are based on the results of these serial examinations (with therapies being withheld if deteriorations in cardiac function are detected on serial imaging), the imaging technique used must have optimal test-retest, intraobserver and interobserver reproducibility, with minimal background noise. The optimal imaging modality for screening for cardiotoxicity however remains unknown.

To date, there is no head-to-head comparison of the reproducibility (and accuracy) of all three modalities performed with no temporal gap between them. The investigators will perform a multi-modality imaging study, where the CMR, echocardiography and MUGA departments of Barts Heart Centre will work together to assess which is the best modality for cardiac surveillance in oncology patients. If one technique outperforms the others, implementation of this modality for screening for cardiotoxicity would clearly improve our cardio-oncology practice, and potentially ensure earlier detection of cardiac dysfunction and prevent inappropriate withdrawal of potentially life saving cancer treatments.

Risks and Benefits This study is not primarily designed to benefit the current research participants, however since some of the imaging techniques used can detect subclinical heart disease, in the case of any significant findings being detected, The investigators will inform subjects and offer life-style recommendations and, when necessary, medical treatment. All research data will be anonymised and only the researcher will be able to link data to individual patients.

1. TRIAL OBJECTIVES

Main clinical question: Which is the best modality (MUGA, echocardiography or CMR) in terms of test-retest reproducibility for cardiac screening in patients undergoing chemotherapy?

Primary Endpoint: Test-retest reproducibility of MUGA, echocardiogram and CMR.

Secondary Endpoints: Correlation between imaging modalities for assessment of ejection fraction and Intra-observer and inter-observer reproducibility for each of the 3 imaging modalities.

1. METHODOLOGY

Inclusion Criteria We will include 100 patients, aged 18 or above, who are already booked for cardiac imaging investigations (MUGA scan, echocardiography (TTE) or CMR) to assess their left ventricular systolic function, as part of their routine clinical care. The majority of these patients will be patients who are being screened for cardiac toxicity resulting from cancer therapies. Some patients will be pre-treatment but many will be patients who have started treatment or completed treatment.

Exclusion Criteria:

• pregnancy and breast feeding.
• standard contraindications to cardiac MRI including pacemakers/implantable defibrillators, claustrophobia.
• atrial fibrillation.

1. STUDY PROCEDURES

Patient recruitment Potential participants will be identified from the lists of patients already booked for MUGA scans and TTE at Barts Heart Centre, as part of their standard clinical care.

Study plan The investigators will recruit 100 oncology/ haematology patients who are already scheduled for a MUGA cardiac scan or an echocardiopgraphy at Barts Heart Centre to assess cardiac function, as part of their standard clinical investigations.

Patients will be invited to participate in the study, which would require them to also undergo cardiac assessment with other imaging modality on the same day (see diagram below).

Group 1 will undergo their planned MUGA plus a repeat MUGA scan plus 2 CMR scans.

Group 2 will undergo their planned MUGA plus a repeat MUGA scan plus 2 TTE. Group 3 will undergo their planned TTE, plus a repeat TTE plus 2 CMR scans.

No patient will receive more than the standard radiation dose because the repeat MUGA acquired on the same day does not require a second injection of tracer, but simply makes a repeat measurement of activity levels in the heart using the tracer from the initial injection.

Acquired images will be interpreted by two blinded, independent observers. Analysis will be conducted to obtain reliable information on intra-observer, interobserver and test- retest variabilites of each technique.

Imaging Techniques

CMR will be performed using a modified focused protocol in a Siemens 1.5T Avanto scanner. This will take only 10 minutes and includes the localizers, black and white blood sequences and cine images (4 chamber view, 2 chamber view and short axis stack). No contrast agent will be administered.

Analysis will be performed using CVI42 customised software (Circle Cardiovascular) and will be performed by 2 independent, blinded observers with at least SCMR Level 2 accreditation.

2D and 3D Echocardiography will be performed using a GE Vivid9 machine, according to EACVI / BSE protocols. Analysis will be performed using ECHO-PAC software, by two blinded observers with EACVI/BSE accreditation.

MUGA scan will be performed according to the local standard operating procedures (SOP). The patient will be given an injection of stannous medronate reducing agent 20 minutes prior to the injection of 800 MBq of Tc-99m Pertechnetate. Images will be acquired with a dual-head gamma camera (Bright View XCT; Philips healthcare) using a low-energy high-resolution collimator, with energy window of 20% placed over a photopeak of 140 keV.

The LV and RV regions of interest, as well as background activity, will be delineated automatically by the QBS (Quantitative Blood Pool SPECT) software available on the gamma camera processing workstation, with manual correction by the interpreting fellow, if necessary.

Cardiac function measurements LVEF, end-diastolic and end-systolic volumes of the LV (EDV/ESV), global and regional contractility, myocardial mass, as well as early markers of ventricular dysfunction such as longitudinal function, myocardial strain and diastolic parameters (when possible according to the technique).

Acquired images will be anonymized and then interpreted by two blinded, independent observers.

Analysis will be conducted so as to obtain reliable information on test-retest reproducibility for each of the 3 modalities, as well as intraobserver and interobserver variabilites of each technique.

End of Study Definition The study will end after the last patient´s visit and scanning had been performed. At the end of this study, an anonymized summary of the results will be offered to the participants.

REC Reference: 16/LO/1815, IRAS Project ID: 200038