Usefulness of NT-Probnp Levels and Fresco Cardiovascular Risk Scale for Prediction of Myocardiotoxicity Induced By Anthracyclines in Patients with Diffuse Large B-Cell Lymphoma

Introduction: Anthracyclines are cytotoxic antibiotics used in the treatment of lymphomas. Myocardiopathy is a well-known toxicity of anthracyclines, and both acute/subacute and late-onset presentations have been described. Early detection of asymptomatic cardiac dysfunction and identification of biomarkers for anthracycline cardiotoxicity risk may be important to prevent irreversible heart damage.

Objective: To prospectively evaluate the incidence, time of appearance and clinico-biological variables associated with the development of myocardiotoxicity in patients with diffuse large B-cell lymphoma (DLBCL) treated with R-CHOP/R-CHOP-like regimens.

Patients and Methods: one hundred and sixty six patients with DLBCL from 05/2004 to 05/2014. Excluded: HIV infection (10), treatment without anthracyclines (21) and other causes (5) (Table 1). Left ventricular ejection fraction (LVEF) determined by high resolution echocardiography and the N-terminal brain natriuretic peptide (NT-proBNP) fragment were done prior initiation of treatment, at the end and every 6-12 months thereafter. Myocardiotoxicity was defined as: LVEF <55% or decrease >15% if prior LVEF <55% and/or clinical manifestations of heart failure (CHF). The FRESCO scale that accurately estimates 10-year cardiovascular risk in the adult European population was used to estimate cardiovascular risk. The C statistic for NT-proBNP and FRESCO scale was determined and competitive risk analysis was performed in the evaluation of myocardiotoxicity.

Results: median age 68 years (IQR: 54-75), 51% male, median NT-proBNP 251.8 pg/mL (IQR: 76.2-560.1), median LVEF 64% (IQR: 60-69%) and median FRESCO scale 4.5 (IQR 2.1-7.2). NT-proBNP was correlated with FRESCO scale, but not with LVEF. With a median follow-up of 64.7 months (95% CI 56.4-73.6) 24 cardiac events were observed (5 clinical CHF + normal LVEF, 7 clinical CHF + LVEF low and 12 low LVEF without clinical CHF). Cumulative incidence of myocardiotoxicity was 8.8% at 6 months (95% CI 5.0 to 15.3), 12.3% at 12 months (95% CI 7.6 to 19.6), 14.2% at 24 months (95% CI 9.1 to 21.9) and 17.6% at 5 years (95% CI 11.6 to 26.1). C statistic for myocardiotoxicity was: FRESCO scale 0.7192 (95%CI: 0.6284- 0.8099) and FRESCO scale+NT-proBNP 0.807 (95%CI: 0.7309-0.8832). Competing risks analysis showed a significantly increased cumulative incidence of myocardiotoxicity in patients with NT-proBNP >600 +/- FRESCO >4.5 (Figure 1). Finally, we performed a computational modeling using 12 Bayesian networks to analyze the connections between demographic characteristics, cardiovascular risk factors, treatment, occurrence of cardiotoxicity and death (Figure 2 illustrates one of the networks).

Conclusions: myocardiotoxicity induced by anthracyclines remains an important problem in the daily practice. The FRESCO cardiovascular risk function is useful for predictingmyocardiotoxicity and levels of NT-proBNP improved accuracy myocardiotoxicity risk. We propose that patients with NT-proBNP >600 pg/mL or FRESCO > 4.5 should have a specific cardiologic surveillance and follow-up in cardio-oncology units from the start of their anthracyclin-containing chemotherapy.