Evaluation of left atrial function and mechanical dispersion in breast cancer patients after chemotherapy

Abstract Background Left atrial (LA) function and mechanical dispersion changes in breast cancer patients treated with chemotherapy remain unclear. Hypothesis LA function and LA mechanical dispersion in breast cancer patients would be impaired after chemotherapy. Methods This single‐center retrospective study included 91 consecutive breast cancer patients treated with chemotherapy and 30 controls. Patients were examined by echocardiography three times at intervals. Conventional parameters, left ventricular strain, LA strain, and LA mechanical dispersion were evaluated and compared. Results LA strain during reservoir phase (LASr), conduit phase (LAScd), and contraction phase (LASct) all decreased markedly after chemotherapy and were lower than those of the controls (all p < .01). The standard deviation of time to peak positive strain during LA reservoir phase corrected by R‐R interval (LA SD‐TPSr) was significantly increased after chemotherapy and was higher than that of the controls (p < .001). The change of LA function was expressed as Δ. Multivariate linear regression analyses showed that LAVIp (0.399, 95% confidence interval [CI]: 0.610, 1.756, p = .000) was independently associated with ΔLASr, LAPEF (−0.325, 95% CI: −45.123, −10.676, p = .002) and age (0.227, 95% CI: 0.021, 0.350, p = .027) were independently associated with ΔLAScd, and LAVImax (0.341, 95% CI: 0.192, 0.723, p = .001) was independently associated with ΔLASct. LAVImax (0.505, 95% CI: 0.000, 0.001, p = .039) and mitral E (−0.256, 95% CI: 0.000, 0.000, p = .024）were independently associated with ΔLA SD‐TPSr. Conclusions Mechanical function of LA declined after chemotherapy in breast cancer patients. With the decrease of LA mechanical function, LA mechanical dispersion assessed by two‐dimensional speckle‐tracking echocardiography increased significantly, and its clinical value needs to be further studied.


| INTRODUCTION
Breast cancer remains the leading cause of mortality in women. In the United States, breast cancer affects nearly 3.32 million women. 1 Currently, advances in breast cancer treatment have led to improved survival in these patients. However, treatment can result in cancer therapeutic-related cardiac dysfunction (CTRCD) due to myocardial toxicity. 2 identified as a potential indicator of cardiac dysfunction and arrhythmias due to cancer treatments. 5 In addition, controversies remain regarding changes in LA strain in breast cancer patients after chemotherapy. Ana Teresa Timóteo et al. suggested that no significant change in LA strain in breast cancer patients was observed after chemotherapy. 6 In contrast, Hyukjin Park et al.
suggested that a significant decline in LA strain developed after chemotherapy for breast cancer. 7 The respective changes in LASr, LAScd, and LASct in these patients is still a matter of debate. LA mechanical dispersion is a parameter related to arrhythmia, especially atrial fibrillation (AF). 8 However, data on LA mechanical dispersion in breast cancer patients after chemotherapy remain limited. Echocardiography is a sensitive and reproducible technique for the assessment of LA function and LA mechanical dispersion.
Recently, GLS by two-dimensional speckle-tracking echocardiography (2D-STE) has been used for the assessment of regional and global LA function. Therefore, in this study, we aimed to assess LA function and LA mechanical dispersion in breast cancer patients after chemotherapy by echocardiography.

| Study population
This is a retrospective study with an initial sample of 100 patients with a histopathologically confirmed diagnosis of breast cancer at an early or locally advanced stage (Stage I-IIIC) between January 2016 and December 2019 at our institution. All patients were female, with a mean age of years (52.8 ± 9.8 years). The exclusion criteria were (1) prior history of chemotherapy, hormone treatment, or radiation; (2) LVEF < 50% before chemotherapy; (3) a previous history of heart failure (HF) and/or coronary heart disease, more than mild valve disease, arrhythmia (AF, atrial flutter, frequent ventricular/atrial premature beat, etc.) and/or cardiomyopathy; and (4) age <20 years or >80 years. All patients underwent chemotherapy one month after modified radical mastectomy. Hematological examination and echocardiography were performed in all patients who received follow-up after hospital discharge. Overall, 41 (45%) patients received epirubicin (360 mg/m 2 ) with concurrent cyclophosphamide, followed by docetaxel (EC-D); 22 (24%) received trastuzumab with docetaxel and either cyclophosphamide or carboplatin (TCH/TCbH); and 28 (31%) received epirubicin (360 mg/m 2 ) with concurrent cyclophosphamide, followed by trastuzumab and docetaxel (EC-DH). Radiation treatment occurred at a median of 5 months after the operation. CTRCD was defined by a reduction of 10% points in LVEF to a value below 50% (lower limit of normal) or by a relative percentage reduction of more than 15% of LVGLS from baseline .5 All patients signed an informed consent. The Dalian Medical University Ethics Committee approved this protocol.

| 2D-STE
Echocardiographic examination was performed before chemotherapy (T0), after approximately 6 months of chemotherapy (T6) and after 12 months of chemotherapy (T12) in our department using a Vivid Segments in which tracking was inadequate were excluded from the analysis despite manual adjustment. If more than three segments were excluded, the subject was removed from the study.
The strain curves of the global and regional LA wall were automatically generated by the software, and the reference point for image analysis was taken at the onset of the QRS complex. 10 There are two peaks that correspond to LA reservoir function (first peak-LASr) and

| Intra-and interobserver variability
Both intra-and interobserver reproducibility were assessed by calculating the difference between the LA strain values of 20 randomly selected patients measured by one observer twice and by a second observer within 48 h.  The baseline characteristics of the patients are presented in Table S1.

| Statistical analysis
The time interval between T0 and T6 was 5.09 ± 1.66 months, and T12 was performed 6.05 ± 2.63 months after the second exam. One chemotherapy plan took 21 days as one cycle. The mean duration of F I G U R E 1 Left ventricle (LV) parameters in the study population. Em, early diastolic lateral mitral annular tissue doppler velocity; LVESV, left ventricular end systolic volume; LVGLS, global longitudinal strain of left ventricle. a, compared with T0, p < .05; *, compared with controls, p < .05 chemotherapy was 6.57 ± 1.9 cycles. The median time elapsed from the last chemotherapy to T12 was 190 ± 35days. None of the patients had developed cardiac complications, including AF.

| Univariate and multivariate linear regression analyses
After univariate analysis of clinical confounders, all measures of cardiac structure and function, the factors with p < .10 were included in multivariate regression analysis (   that compared with controls, LAVImax and LAVIp increased significantly after chemotherapy. 13 In breast cancer patients receiving chemotherapy, LA dilatation has been proven to be related to the occurrence of cardiac dysfunction. 14 However, in a study conducted among long-term survivors of childhood cancer treated with anthracyclines, the LAVI of the chemotherapy group decreased notably compared with that of the control group.
However, LAEF did not change significantly. The author explained that this might be due to fibrosis and cardiac remodeling. 15 Another study showed that among children exposed to anthracyclines, the short-term effects on LA function were small for patients with preserved LVEF. 16 In our study, the LAV of breast cancer patients increased after chemotherapy, but there was no significant difference. The shorter follow-up time, the younger subjects, the smaller sample sizes, vomiting or inadequate intake due to chemotherapy, and different

| LA strain
LA modulates LV filling pressure and cardiovascular performance by functioning as a reservoir, conduit, and booster pump, which plays an integral pathophysiological role in LV diastolic dysfunction. Atrial strain showed a good correlation with pulmonary capillary wedge pressure, even better than the E/e' ratio in advanced HF. 17 Theoretically, a reduction in LA performance mirrors diastolic dysfunction.
Atrial strain has been evaluated in multiple conditions, such as hypertension, diabetes, AF, HF, ischemic and valvular heart disease, and has been included for the assessment of prognostic implications. 9 In our study, we found that LA strain could detect atrial dysfunction earlier than the parameters of volume. Substantial reductions were observed in all the parameters of LA strain in this study. Li

| LA dispersion and arrhythmia
Chemotherapy is a frequent cause of arrhythmias including AF. 19 Several anticancer agents were found to induce AF, with a reported incidence for anthracycline of 2%-10% 20 significance of these phenomena still needs further study. The effects of chemotherapy on atrial myocardium fibers over time remain unclear, though they may result in autonomic nervous dysfunction through oxidative stress, decreases in intracardiac conduction and a heterogeneous dispersion of repolarization, leading to the dyssynchrony of mechanical movement of the atrium. 24

| Limitations
There are several limitations in the present study. The sample size was relatively small. No subgroup analysis was conducted according to the chemotherapy regimen. The follow-up period was only up to one year. The relationship between atrial parameters and possible arrhythmia was not monitored, and paroxysmal arrhythmias may have been missed. We used software designed for LV strain analysis to obtain LA strain because of the lack of dedicated atrial software. Vendor differences arising from differences between edge tracking and speckle tracking may affect the reliability of LA strain and SD-TPS. Further prospective studies will be required to determine the clinical significance of our observed findings.

| CONCLUSION
The mechanical function of LA changed after chemotherapy in breast cancer patients. The decrease in functional indicators measured by 2D-STE occurred before the changes in LAV parameters. With the decrease of LA mechanical function, LA mechanical dispersion assessed by 2D-STE increased significantly, and its clinical value needs to be further studied.

CONFLICTS OF INTEREST
The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT
The data used to support the findings of this study are available from the corresponding author upon request.