Serum and plasma sphingolipids as biomarkers of chemotherapy-induced cardiotoxicity in female patients with breast cancer

Abstract

Although effective as a chemotherapeutic, the utility of Doxorubicin (Dox) is hampered by cardiotoxicity. Despite this, the ability to predict and guide monitoring of patients receiving Dox is hampered by a lack of effective biomarkers to identify susceptible patients and detect early signs of subclinical cardiotoxicity. Based on their well-established roles in the response to Dox and other chemotherapies, we performed a retrospective analysis of serum and plasma sphingolipids (SLs) from female patients with breast cancer (BC) undergoing anthracycline-containing therapy, correlating with cardiac parameters assessed by echocardiography. Results showed substantial changes in both plasma and serum SL species during therapy including ceramide (Cer), deoxydihydroCer, and dihydrosphingosine with reversion toward baseline after treatment. Linear mixed-effects model analysis revealed that baseline levels of a number of SLs correlated with adverse cardiac outcomes. Here, serum sphingosine-1-phosphate (S1P), dihydroS1P, and plasma Cer performed comparably to the prognostic value of pro-NT-BNP, an established biomarker of cardiotoxicity. Intriguingly, while pro-NT-BNP had no predictive value at mid- and post-therapy timepoints, serum S1P and dihydroS1P, and plasma Cer levels showed a correlation with adverse outcomes, particularly at the post-therapy timepoint. Finally, analysis of plasma and serum C16:C24-Cer ratios-previously linked with adverse cardiac outcomes-showed no correlation in the context of chemotherapy treatment. Overall, this pilot study provides initial evidence that plasma and serum SLs may have benefits as both prognostic and diagnostic biomarkers for female BC patients undergoing anthracycline-containing chemotherapy. Consequently, diagnostic SL measurements-recently implemented for metabolic-associated cardiac disorders-could have wider utility.

Keywords: biomarkers; cardiotoxicity; doxorubicin; sphingolipids.

Fig. 1

Alterations in plasma and serum sphingolipid levels in patients undergoing anthracycline-containing therapy. Lipids were extracted from 100-200ul of (A) plasma; and (B) serum, and analyzed for the sphingolipid classes shown as described in “Materials and Methods”. Data were normalized to ml plasma/serum and are expressed as pmol/ml at baseline (BL), midpoint (Mid), Post-therapy (post) and 6-months follow up (6-months). Data are expressed as mean ± 95% confidence interval.

Fig. 2

Baseline plasma and serum sphingolipid levels correlated with changes in LVEF in patients receiving anthracycline-containing therapy. Linear mixed-effects models were used to correlate baseline lipid levels for (A) serum and (B) plasma with changes in LVEF from baseline to 6 months follow up. Here, a decrease in LVEF would be considered an adverse outcome. Trendline is shown in red with 95% confidence interval shown in light gray. Abbreviations: Deoxydhcer = deoxydihydroceramide; dhS1P = dihydrosphingosine 1-phosphate; S1P = sphingosine 1-phosphate.

Fig. 3

Baseline plasma and serum sphingolipid levels correlated with changes in GLS in patients receiving anthracycline-containing therapy. Linear mixed-effects models were used to correlate baseline lipid levels for (A) serum and (B) plasma with changes in GLS from baseline to 6-month follow-up. Here, an increase in GLS would be considered an adverse outcome. The trendline is shown in red with 95% confidence interval shown in light gray. Abbreviations: Deoxydhcer = deoxydihydroceramide; dhS1P = dihydrosphingosine 1-phosphate; S1P = sphingosine 1-phosphate.

Supp Fig 1