Childhood cancer prognosis has significantly improved in the last decades, thanks to earlier diagnosis and more effective treatments [1, 2]. The increase in survival rates, however, has also resulted in long-term sequelae due to anti-neoplastic drugs [1]. Cardiovascular (CV) diseases, in particular heart failure (HF), represent a leading cause of non-oncologic morbidity and mortality in this highly-vulnerable population [3, 4]. Anthracyclines, used in a great variety of hematological and solid pediatric malignancies, are associated with cardiomyocyte damage, that can lead to impaired cardiac function, even decades after chemotherapy [4, 5]. Moreover, it is well known that childhood cancer survivors (CCSs) are at increased risk of hypertension, diabetes and obesity [6, 7], which may negatively impact on an already fragile system and predispose to CV diseases.

The early detection of subclinical left ventricle (LV) dysfunction can be crucial in these patients, as it might promote a closer clinical follow-up and treatment aimed to prevent or reduce the evolution toward overt LV dysfunction and HF. At present, however, no reliable marker of risk of LV dysfunction has been identified [4, 5, 8]. In a previous paper on the same population, our group demonstrated that three-dimensional (3D) LV ejection fraction (LVEF) was able to detect subclinical changes of ventricular function in CCSs [9].

Although CV risk assessment has traditionally focused on the heart, vascular dysfunction may also occur and persist for years after chemotherapy. However, whether vascular alterations happen and influence cardiac function in CCSs remains poorly investigated. Left Ventricular-Arterial Coupling (LVAC) is becoming a valid surrogate of myocardial mechanics, integrating arterial load and LV end-systolic elastance, hence quantifying chamber stiffness and contractility [8, 9, 10]. It provides an assessment of CV performance and efficiency, with higher values of LVAC ratio reflecting a compromised ventricular-vascular matching [10, 11, 12]. Importantly, LVAC has been found to have prognostic value in HF adult patients [10, 11, 12] but, despite its importance, it has never been investigated in CCSs previously treated with anthracyclines.

Aim of this study was to assess LVAC, performed with different echocardiographic modalities, as well as vascular function, reported as flow mediated dilatation (FMD), in CCSs and matched controls.

To our knowledge, this is the first study that extensively reports, through multiple echocardiographic modalities, LVAC in CCSs previously treated with anthracyclines, relating these parameters with endothelial function assessed by FMD. Several findings hence emerge from our investigation: (1) Ea, Ees and Ea/Ees did not differ between CCSs and controls; (2) 3D SV was significantly lower in CCSs; (3) 3D ESV/SV ratio was significantly higher in CCSs; (4) lack of remarkable difference between survivors and controls was detected in terms of endothelial function; (5) among CCSs, 3D LVAC parameters showed a significant correlation with FMD values, anthracycline cumulative doses, as well as with years after drug exposure.

The population of cancer survivors is increasing over time. This has been paralleled by the growing evidence of long-term adverse effects of cancer therapies, the need for a better understanding of the mechanisms underlying CV toxicities and early identification of these negative effects [4, 22, 23]. Anthracycline cardiotoxicity may occur early after the first cycles or become manifest even several years later, highlighting the necessity of detecting subgroups of patients at increased risk of developing HF [4, 22, 23]. However, in cancer survivors’ programs, evaluation of cardiac function is almost exclusively assessed through LVEF [1].

Non-invasive measurement of LVAC is feasible in children [24], may provide comprehensive assessment of LV performance and has been demonstrated to give incremental information to LVEF in the characterization and management of patients with HF, pulmonary hypertension and coronary artery disease (CAD) [10]. In the setting of cardio-oncology, however, evidence on LVAC remains sparce. Narayan et al. [25] showed, in 135 adult patients affected by breast cancer treated with anthracyclines, that LVAC alteration may precede LVEF deterioration [26]. Moreover, a recent study demonstrated that LVAC decreases early after anthracycline exposure, primarily because of a change in LV elastance relative to arterial elastance [27]. In a further report on breast cancer patients previously exposed to chemotherapy, LVAC was not altered at rest, but it was impaired during exercise echocardiography at 7 years of follow-up [28].

Long-term effects of anthracyclines on LVAC have never been investigated in a highly vulnerable population as survivors of pediatric malignancies. In these subjects, we found no significant differences neither in terms of arterial and ventricular elastance assessed with 2D echocardiography, nor relating to their ratio. On the other hand, lower 3D stroke volume was detected in CCSs. As a consequence, significantly higher 3D ESV/SV values were observed in our population than in controls, with a positive significant correlation between 3D LVAC and both anthracycline cumulative doses administered and time after exposure.

These findings merit further discussion: SV at rest was reduced in a limited group of breast cancer survivors almost 10 years post-anthracycline chemotherapy completion and this alteration was conceivably responsible for impaired exercise capacity in these patients [29]. In our report, the reduction in 3D SV, which assessment is feasible and accurate both in aldults and in the pediatric population [30], may corroborate the alteration of 3D LVAC; moreover, this last metric is also a surrogate of 3D LVEF, which in a previous paper on the same population [9], resulted to be the only parameter significantly altered between groups.

Yet, in this study we failed to detect marked differences in endothelial function, as assessed by FMD. Endothelial dysfunction is an early step in the pathogenesis of many CV diseases, in particular atherosclerosis, and plays an important role in their development and progression [21]. Of importance, endothelial dysfunction has been suggested to participate in the pathogenesis of myocardial damage related to anthracycline therapy, possibly mediated by increased oxidative stress, decreased NO production and increase in NO inactivation [28]. Besides these solid basis, previous studies showed contrasting data on the effects of anthracycline therapy on vascular endothelial function, as assessed by FMD, probably because of differences in anthracycline doses administered, follow-up times and coexistence of traditional CV risk factors [31, 32, 33, 34, 35]. Our data demonstrated that, in this relatively small cohort, endothelial-dependent function appears normal in the years after treatment, suggesting an inherent endothelial plasticity after removal of the toxic perturbation.

Significant negative correlations between FMD, 3D ESV/ESV and anthracycline dose were detected. However, the same trend could not be confirmed when arterial and ventricular elastance, expression of CV reserve capacity (CVRC) [10], were evaluated. Therefore, 3D LVAC may represent a precocious marker of impaired cardiac performance, considering its significant correlation with endothelial function, cumulative anthracycline dose and time after chemotherapy exposure. Nevertheless, our data still underscore the necessity to find early modifications in CV physiology that can predict adverse long-term events.

Limitations should be acknowledged: the small sample-size, deriving from a previously published analysis, limits the generalization of our results. Hence, all conclusions should be considered hypothesis-generating. We also decided to exclude patients exposed to radiotherapy, that may have important effects on arterial and ventricular performance and may prevent a clear interpretation of direct anthracycline effects on the CV system. Second, in our cohort the lack of CV events or availability of cardiac biomarkers prevents attribution of a prognostic role to LVAC values. The follow-up of our patients was limited and therefore we cannot exclude that some events may be found at a later stage. Finally, we did not perform endothelial-independent estimation of brachial vasodilatation.

Limitatations should be weighted with our study’s strengths: this is the first report investigating LVAC, obtained by multiple echocardiographic methods, in CCSs, and correlating results with endothelial function. Our results pave the way to future directions: the study of ventricular-vascular coupling, both at rest and during exercise echocardiography may represent a new field of imaging in cardio-oncology, not only in CCSs but also in adults undergoing recently-introduced therapies. Nevertheless, considering the acknowledged limitations, verification of our findings in larger populations, in particular with high comorbidity burden, is hence warranted. Overall, our study highlights the critical need for more detailed characterisation of arterial properties and their impact on outcomes in cardio-oncology.

Echocardiographic LVAC and its correlation with endothelial function have never been investigated in survivors of childhood malignancies [36]. In our population, previously exposed to anthracycline therapy, we could not detect significant alterations in terms of arterial and ventricular elastance estimated by echocardiography, as well as of endothelial function. However, 3D SV and consequently 3D ESV/SV were significantly impaired in CCSs when compared to controls, this last one with a positive correlation with chemotherapy cumulative doses and years after exposure. Whether these parameters may portend for clinical implications over longer follow-up deserve careful assessment in large prospective studies.

All data used or analysed during the current study are available from the corresponding author on reasonable request.

MC—conceptualization-writing-statistics-revision; LB—analysis-interpretation of data-writing-revision; ABD—analysis and interpretation of data-revision; PL—conceptualization-revision; ADV—conceptualization-writing-revision; VM—conceptualization-writing-revision; ARom—conceptualization-revision; ARug—conceptualization-revision; GA—conceptualization-revision; GAL—statistics-revision; MM—conceptualization-revision; FC—conceptualization-revision; AL—conceptualization-revision. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The study was performed in accordance with the Helsinki declaration; all subjects provided written informed consent. The study was approved by the Ethics Committee of the Fondazione Policlinico Universitario Agostino Gemelli IRCSS; the approval number is DIPUSVSP-15-11-2239.

The authors thank “Fondazione per l’Oncologia Pediatrica” for their dedicated patient care.

This research received no external funding.

The authors declare no conflict of interest.