Detecting Cancer Treatment-Related Cardiac Dysfunction
By Michael H. Crawford, MD, Editor
SYNOPSIS: When searching for breast cancer-related cardiac dysfunction, a sequential algorithm using echo ejection fraction and strain parameters produced an area under the receiver operating curve of 89%. Adding biomarkers did not improve the ability to diagnose cardiac dysfunction.
SOURCE: Esmaeilzadeh M, Urzua Fresno CM, Somerset E, et al. A combined echocardiography approach for the diagnosis of cancer therapy-related cardiac dysfunction in women with early-stage breast cancer. JAMA Cardiol 2022;7:330-340.
Left ventricular ejection fraction (LVEF) by cardiac magnetic resonance (CMR) imaging is perhaps the best way to detect cancer therapy-related cardiac dysfunction (CTCD), but serial studies for surveillance are not feasible. An alternative might be EF by echocardiography, along with global longitudinal strain (GLS) or global circumferential strain (GCS). Biomarkers, such as troponin and brain natriuretic peptide (BNP), could be a useful adjunct.
Investigators from the University of Toronto recruited 160 women with human epidermal growth factor receptor 2-positive (HER2) early-stage breast cancer scheduled to receive anthracycline and trastuzumab therapy, with or without adjuvant radiotherapy, from 2013-2019. The authors excluded those with contraindications to CMR, low life expectancy, history of cancer therapy, or known cardiac disease (n = 24). All patients underwent transthoracic echocardiograms, CMR, and biomarker measurements before anthracyclines and after — but before trastuzumab therapy — and at three, six, nine, and 12 months after trastuzumab therapy. The biplane Simpson method was employed to calculate LVEF from 2D echo and a semiautomated algorithm (GE Healthcare) for 3D echo EF. Peak systolic global longitudinal strain (GLS) was obtained from apical four-, three-, and two-chamber views of the LV using EchoPAC version 202 (GE Healthcare). Peak systolic GLS and global circumferential strain (GCS) were obtained from the three short axis views using Q analysis version 202 (GE Healthcare). CTCD by EF was defined as a > 10% reduction from baseline without heart failure symptoms and 5% with symptoms or to less than 55% at any time. GLS- and GCS-defined CTCD was a reduction > 15% from baseline. An abnormally high sensitivity troponin I or BNP was defined as a value greater than the 99th percentile.
Of 136 patients (mean age, 51 years), epirubicin was given to 73%; the rest took doxorubicin. All received trastuzumab, and 88% received radiation therapy. Baseline mean EF by CMR was 63%, and changes in echo EF (2D and 3D) GLS and BNP were associated with changes in CMR EF; changes in GCS and troponin were not. CTCD occurred in 27% of patients by CMR EF, 23% by 2D echo EF, 22% by 3D echo EF, 42% by GLS, 50% by GCS, 24% by BNP, and 10% by troponin. GLS is more sensitive than echo EF, and 3D echo was more sensitive than 2D EF for CMR CTCD. A sequential algorithm using 3D EF, GLS, and GCS was optimal for diagnosing CTCD by CMR (area under the receiver operating curve = 89%). When all three tests were negative, the probability of CTCD was 1%. Replacing 2D echo for 3D, the algorithm still performed well, but mainly because of a high negative predictive value. Biomarkers did not improve the algorithm. The authors concluded this sequential algorithm using 3D echo EF, GLS, and GCS was more accurate than using any measure alone for detecting CTCD.
COMMENTARY
The results of this study seem to substantiate the practice of using EF to define CTCD, and strain parameters and biomarkers to predict future CTCD. To wit, strain-based CTCD and elevated biomarkers were seen after one to two months of initiation of cancer therapy vs. four to five months after starting therapy to detect reduced LVEF by any of the three imaging techniques.
Overall, 27% of HER2-positive breast cancer patients receiving anthracyclines and trastuzumab developed CMR-defined CTCD. Of these patients, 75% met 2D echo criteria for CTCD, 81% by 3D echo, 91% by GLS, 97% by GCS, 24% by BNP criteria, and 8% by troponin criteria. GLS was more sensitive than echo and was more accurate than GCS. BNP and troponin were sensitive, but specificity was poor. Thus, the authors’ sequential algorithm leverages these results to come up with the timeliest diagnosis of CTCD during surveillance.
First, 3D echo EF should be considered. If CTCD criteria are met, there is a high probability of CTCD (77%). If LVEF criteria are not met, then strain should be considered. If both GLS and GCS are low, there is a 16% probability of CTCD. These patients may warrant closer follow-up, or perhaps CMR should be performed to ensure EF is normal. If none of these three measures are abnormal, then the probability of CTCD is 1%. Adding BNP or troponin did not improve the algorithm. Using 2D echo EF or employing just one circumferential LV slice for determining GCS did not affect the performance of the algorithm much. This study was conducted in a single center using a modest population. Also, poor echo images were discarded, so the value of echo indicated here may be inflated. In addition, the authors did not use a validation cohort. Nevertheless, the approach detailed here probably results in more objective cancer therapy decisions, such as determining how often to follow up.
When searching for breast cancer-related cardiac dysfunction, a sequential algorithm using echo ejection fraction and strain parameters produced an area under the receiver operating curve of 89%. Adding biomarkers did not improve the ability to diagnose cardiac dysfunction.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.