Use of integrated imaging and serum biomarker profiles to identify subclinical dysfunction in pediatric cancer patients treated with anthracyclines

Abstract

Background: Anthracycline induced cardiomyopathy is a major cause of mortality and morbidity among pediatric cancer survivors. It has been postulated that oxidative stress induction and inflammation may play a role in the pathogenesis of this process. Accordingly, the present study performed an assessment of biomarker profiles and functional imaging parameters focused upon potential early determinants of anthracycline induced cardiomyopathy.

Methods: Patients (10-22 years) were prospectively enrolled between January 2013 and November 2014. Thirteen subjects completed the study and underwent serial cardiac magnetic resonance imaging and plasma biomarker profiling performed 24-48 h after the first anthracycline dose and at set dose intervals. In addition, we collected plasma samples from 62 healthy controls to examine normal plasma biomarker profiles.

Results: Left ventricular ejection fraction (LVEF) decreased from 64.3 ± 6.2 at the first visit to 57.5 ± 3.3 (p = 0.004) 1 year after chemotherapy. A decline in longitudinal strain magnitude occurred at lower cumulative doses. A differential inflammatory/matrix signature emerged in anthracycline induced cardiomyopathy patients compared to normal including increased interleukin-8 and MMP levels. With longer periods of anthracycline dosing, MMP-7, a marker of macrophage proteolytic activation, increased by 165 ± 54% whereas interleukin-10 an anti-inflammatory marker decreased by 75 ± 13% (both p < 0.05). MMP7 correlated with time dependent changes in EF.

Conclusions: Asymptomatic pediatric patients exposed to anthracycline therapy develop abnormal strain parameters at lower cumulative doses when compared to changes in EF. A differential biomarker signature containing both inflammatory and matrix domains occur early in anthracycline treatment. Dynamic changes in these domains occur with increased anthracycline doses and progression to anthracycline induced cardiomyopathy. These findings provide potential prognostic and mechanistic insights into the natural history of anthracycline induced cardiomyopathy.

Trial registration number: NCT03211520 Date of Registration February 13, 2017, retrospectively registered.

Keywords: Anthracyclines; Cardiotoxicity; Global systolic function; Myocardial strain.

Conflict of interest statement

Competing interests The authors declare that they have no competing interests.

Figures

Fig. 1

Study Cohort and Diagram. CMRI: cardiac magnetic resonance imaging; ALL: acute lymphocytic leukemia; AML: acute myeloid leukemia. Twenty patients (10–22 years) were prospectively enrolled between January 2013 and November 2014. Seventeen subjects completed visit 1 (30–60 mg/m2) and thirteen completed the remaining study visits (125–175 mg/m2, 200–250 mg/m2, 275–325 mg/m2, after max AC dose and 1 year after max AC dose)

Fig. 2

Schematic diagram demonstrating the three dimensional circumferential-radial-longitudinal coordinate system used for strain calculation. ERR; Peak mid radial strain magnitude; ECC: Peak global circumferential strain magnitude; ELL: Peak global longitudinal strain magnitude. Each element of strain is, simply, a measurement of the fractional or the percent change of length in a specific direction where Lo is the original fiber length before tag deformation and L is the current length

Fig. 3

Plasma profiling for specific determinants of inflammation and signaling. a Plasma profiling for specific determinants of inflammation and ECM remodeling were examined in referent normal subjects and those undergoing initial AC treatment, designated here as Visit 1 in tabular format. The plasma TIMP-1 and 2 levels at this AC treatment time point were not different from referent controls, and while IL-10 and IL-8/IL-10 ratios fell in the AC treatment group at Visit 1, these did not reach statistical significance. b Due to the high incidence of plasma TIMP-3 and TIMP-4 levels were fell below analytical detection limits, a categorical analysis was performed using a Chi Square analysis. A significant shift in the proportion of detectable TIMP-3 and TIMP-4 in plasma samples taken from the initial AC treatment time point occurred (*p < 0.05). c In contrast to the analyte shown in Table A, significant shifts in plasma MMP-2, − 7 and − 9 occurred at AC treatment Visit 1 when compared to referent control values (*p < 0.05)

Fig. 4

Plasma profiling for specific determinants of inflammation and signaling. A number of inflammatory signaling pathways were profiled in plasma samples taken from age matched referent normal subjects and at the initial visit following AC treatment- identified as Visit 1. While robust signals for the soluble IL, TNF and VEGF receptors were detected in both referent normal and Visit 1 samples many soluble receptor analytes were similar between groups and summarized in Table A. On the other hand, specific soluble receptors/pathways were significantly different from referent control values and are shown in Figure 2-2B. These included the soluble receptors for IL-1, -2, and -4. A relative reduction in sRage and sVEGFR3 occurred in Visit 1 samples with an increase in Sgp130. (*p < 0.05 vs Visit 1 values). Normal: referent normal; MMP-2: Matrix metallopeptidase 2; MMP-7: Matrix metallopeptidase 7; MMP-9: Matrix metallopeptidase 9; IL-8: Interleukin 8; sVEGFR3: Fms-related tyrosine kinase 4; sgp130: Interleukin 6 signal transducer; sIL-1RII: Interleukin 1 receptor, type II; sIL-2Ra: Interleukin 2 receptor, alpha; sIL-4R: Interleukin 4 receptor

Fig. 5

Plasma profiling changes from Visit 1 to last available visit in AC treated patients. The relative change in ECM and inflammatory pathways were computed as a function of Visit 1 and the final visit in the AC treated patients. Those with a significant shift from Visit 1 are shown here and identify that continued shifts in determinants of inflammatory signaling and ECM remodeling occurred at later AC treatment time points. (*p < 0.05 vs Visit 1 values)