br Limitations br Our study
Our study is limited by its retrospective and observational design. Accordingly, we cannot establish that the relationship between statin exposure and preservation of LVEF is causal. We matched on age and anthracycline exposure, then per-formed an adjusted analysis to demonstrate that statin use remained independently associated with LVEF decrement. However, residual confounding cannot be ruled out, as we
were unable to adjust for other factors that may influence LVEF measurements such as blood pressure, volume status, and anemia. However, we have no reason to believe that these changes would have occurred differentially in statin-treated patients vs controls. Moreover, we had to rely on clinical notes to determine patients’ statin exposure status. Therefore, it 220556-69-4 was not possible to confirm adherence during cancer ther-apy or to explore the impact of the duration of statin use prior to cancer therapy on preventing LVEF decrement. Further-more, as patients receiving statin treatment had a higher prevalence of cardiovascular comorbidities, it is possible that they were more likely to have been followed by cardiologists, which may hypothetically reduce the risk of cardiotoxicity.
Echocardiography (specifically 3D echocardiography) is currently considered the preferred imaging modality for sur-veillance during cancer treatment.26 However, in our patients treated between 2002 and 2013, MUGA scans were still the most commonly used modality. Although the temporal reproducibility of LVEF measurements with MUGA in patients receiving cancer therapy is not well established, it still remains widely used in routine clinical practice.2,21 Also, previous work has demonstrated that MUGA-measured LVEF has similar agreement with MRI measurements as 3D echo-cardiography in patients receiving breast cancer therapy.27 Furthermore, as the same imaging modality was used in our statin-treated and control groups, the intergroup comparisons are still valid.
There was a higher proportion of patients with metastatic cancer in the control group; however, after adjustment for this variable, the overall effect persisted. Specific cardiovascular outcomes, such as cardiovascular death or cardiovascular dis-ease hospital admission, were not recorded and likely warrant further exploration in longer-term follow-up studies. The number of confounders included in our logistic model was limited by the total number of cardiotoxicity events; however, we included the most important confounders in our model with 1 variable for every 6 to 7 events.28 Given the wide confidence interval around the odds ratio for statins in the multivariable logistic model, we consider this finding to only be hypothesis generating. We used the cardiac risk score from the NSABP B-31 study as a confounder in the logistic model. This risk score was created in patients who were treated with anthracyclines followed by trastuzumab, which constituted only 56% of our cohort. However, as this was a secondary analysis, and the score incorporates 2 important confounders (age and baseline LVEF), we decided to include this score in the model, considering the low number of cardiotoxicity events. Finally, a relatively small sample size from a single centre is a limitation that can be addressed by validating these results prospectively in large multicentre studies.
The concomitant use of statins in women with HER2-positive breast cancer receiving trastuzumab therapy was associated with a lower magnitude of reduction in LVEF during treatment compared with those not receiving statin therapy. The benefit was seen in those who did and did not receive previous anthracycline therapy. Furthermore, the attenuation of LVEF decline in the statin group occurred despite a higher baseline cardiovascular risk profile than the Canadian Journal of Cardiology Volume 35 2019
non-statin users. Prospective clinical studies are warranted to confirm our findings and to determine whether concomitant administration of statins can prevent cardiotoxicity in trastuzumab-treated patients.
This work is supported by the Heart and Stroke Polo Research Grant. Dr Thavendiranathan is supported by a CIHR New Investigator Award (147814).
Disclosures The authors have no conflicts of interest to disclose.
1. Thavendiranathan P, Abdel-Qadir H, Fischer HD, et al. Breast cancer therapy-related cardiac dysfunction in adult women treated in routine clinical practice: a population-based cohort study. J Clin Oncol 2016;34:2239-46.
2. Thavendiranathan P, Abdel-Qadir H, Fischer HD, et al. Risk-imaging mismatch in cardiac imaging practices for women receiving systemic therapy for early-stage breast cancer: a population-based cohort study [published online May 23, 2018]. J Clin Oncol https://doi.org/10.1200/ JCO.2018.77.9736.