Continuous variables, reported as mean ± SD or as median and minimum-maximum range for non-normal distributions, were compared by Student's t-test or Mann-Whitney test, as appropriate. Categorical variables, reported as percentages, were compared by chi-squared test. When feasible, Cox multivariable regression analysis (variable selection method: backward stepwise elimination) was performed including all candidate variables (p < 0.10 in univariate analysis). Incidence rates of events during follow-up were compared with Poisson regression analysis. A two-sided p < 0.05 was considered statistically significant. Data were analyzed with SPSS software version 25 (SPSS Inc., Chicago, Illinois).

At the time of the baseline Cardio-Oncologic evaluation, 295 (58%) patients were in an adjuvant cancer setting (Table 1). Two-hundred eighty-seven (57%) were scheduled to receive anthracycline chemotherapy and 165 (33%) anti-HER2 targeted therapy, with trastuzumab alone (129, 25%) or trastuzumab and pertuzumab (36, 6%). Thirty-nine (8%) patients had a previous exposition to anthracyclines.

Over a mean follow-up of 2.5 ± 1 years, accounting for 1251.4 patient/years, 15 (3%) BC patients developed LVD, 10 (2%) suffered from other CV events and 55 (11%) died (Figure 1).

In this BC cohort, prevalence of CV risk factors was significant, and higher as compared to the European general population (18, 19). Most importantly, CV risk factors control was suboptimal, in particular in the case of dyslipidaemia and arterial hypertension. Overall, the baseline Cardio-Oncologic evaluation allowed to implement CV prevention, recognize unknown CV conditions (even severe valvular heart disease or CAD) and optimize CV profile in the vast majority of BC patients. In a contemporary American BC population receiving trastuzumab therapy, cardiologic involvement during oncologic treatment was performed in <30% of cases (20). BC patients undergoing cardiologic evaluations, however, showed improvements in CV risk factor control.

The usefulness of a baseline cardiologic consultation, in terms of assessment of LVEF, in BC patients scheduled to receive anthracycline treatment, has been questioned given that detection of reduced LVEF, and therefore indications for changing chemotherapy strategy, is generally low (21). However, in contemporary care, Cardio-Oncology should not be intended as a simple act of cardiologic clearance to anticancer treatments, rather it should encompass a thorough assessment of CV risk profile (6). Moreover, for many oncologic patients, a baseline Cardio-Oncologic evaluation may represent the first (and even the only, as was the case for several of the BC women in our cohort) occasion to undergo a cardiologic consultation. With non-adherence to CV medications being detrimental for long-term outcome of BC patients, the baseline Cardio-Oncologic evaluation may also be the chance for CV health education and motivational support (20, 22). It is undeniable that such an “holistic” cardiologic approach requires resources and may be perceived as time consuming. Nevertheless, it is reasonable to assume that the implementation of CV prevention in the oncologic setting by the means of Cardio-Oncology would be of great potential, given the high prevalence of CV comorbidity and risk factors, especially in women undertreated with little awareness of their CV risk (1, 2, 4, 11). Moreover, a well-delivered baseline cardiologic assessment may avoid unplanned CV evaluations during the anticancer treatment period.

Few LVD events occurred in our cohort. Contrary to previous reports (23, 24), classic modifiable CV risk factors, and in particular arterial hypertension, were not associated with the occurrence of LVD. It might be speculated that such finding was the straight consequence of CV risk profile optimization secondary to the baseline Cardio-Oncologic evaluation. Since the risk of cardiotoxicity is mainly related to the individual baseline CV risk profile and the intrinsic toxicity of a given drug (9), “blunting” the CV profile may result in reducing cardiotoxicity risk. In a cohort of oncologic patients treated with anti-vascular endothelial growth factor agents, we have shown that an approach based on a structured Cardio-Oncology programme with baseline evaluation and tailored recommendations for BP management resulted in a lack of association between both controlled and uncontrolled arterial hypertension at baseline and the risk of CV events (25). Similarly, in this BC population, modifiable CV risk factors were not associated with LVD occurrence and its risk was mostly driven by the inherent toxicity of anticancer treatments, such as combined anti-HER2 therapy. Taken together, these findings highlight the importance of baseline CV evaluation of BC women, and that of adequate monitoring of patients scheduled to receive at-risk treatments, even in the long-term (6, 7). However, it should be noted that BC patients with LVD in the overall study population underwent a greater number of Cardio-Oncologic evaluation as compared to those without LVD (Figure 2), while this result was not found in the anti-HER2 subgroup, likely due to the regular echocardiographic monitoring these patients undergo. Moreover, only a third of patients with LVD had overt HF, and more than a half had complete recovery of LVEF. Analyses from the Cardiotoxicity of Cancer Therapy study showed that BC patients treated with anthracyclines and/or anti-HER2 therapies frequently display subclinical, modest, but persistent indexes of LVD, with only partial recovery over time (26, 27). However, association of LVD with overt HF is unclear, and studies to understand the long-term significance of modest-but-stable systolic dysfunction are needed (27). Moreover, a recent case-control study of BC patients treated with trastuzumab showed no association between adherence to echocardiographic monitoring and risk of HF (23). Thus, it may be worth to re-evaluate clinical practice in the light of novel findings, in order to understand what would be the best strategy to follow-up BC patients, especially if treated with anti-HER2 therapy, prior, during and after anticancer treatment. In other words, the risk of cardiotoxicity related to this anticancer treatment does not appear to be reduced by a careful echocardiographic monitoring. Strategies differing from the 3/6 months echocardiographic LVEF screening, derived from seminal trials (28), may be worth of investigation in future studies.

Finally, it is of note that the only CV risk factor significantly associated with LVD was family history of CAD. It has been found that sarcomeric gene variants contribute to the risk of developing cancer-therapy induced cardiomyopathy (mostly due to anthracyclines) (29). Our data may indicate that at some extent a genetic background predisposition might also contribute to the development of early LVD due to anti-HER2 therapy, warranting further investigations.

Beyond LVD, there is a broad spectrum of CV issues that may affect BC women. These CV events are not of secondary importance in the clinical history of BC women, as they may be severe and happen with a similar rate to that of LVD (Figure 1). Some of these CV events were clearly related to anticancer therapy (as elevation of BP with anti-vascular endothelial growth factor agents), while other may have been not, and just have resulted from an unfavorable CV profile. Indeed, characteristics of BC patients experiencing these CV events were different from those of patients with LVD. BC women with CV events other than LVD had more commonly advanced cancer (and as such had more frequently a previous exposition to anthracyclines) and a worse CV risk profile, characterized by arterial hypertension with organ damage. The occurrence of these CV events may be the result of the interaction of cardiologic and oncologic comorbidities, which may intersect and favor occurrence of adverse clinical outcomes (2, 30). Consistently, BC patients who did not survive, as compared to survivors, more frequently experienced a CV event other than LVD during their clinical course. In other words, the burden of CV risk factors and CVD extend beyond the risk of LVD, and has a significant impact on the clinical course of oncologic patients (31). As a plausible consequence, in our cohort, AF was the only predictor of mortality – together with advanced cancer, as expected –, reflecting the fact that AF represents a proxy of frailty (32).

All these aspects underline the importance of CV prevention in oncologic patients, and specifically in BC women. Adequate CV risk profile optimization and management exerts beneficial effects not only in the short-term to minimize cardiotoxicity, but also in the long-term clinical course, both from the cardiologic and the oncologic standpoint (2, 11). This consideration acquires even more importance when one considers that mortality was overall low, and therefore the vast majority of our BC cohort would experience extended life expectancy. In such a perspective, CV prevention becomes of paramount importance.