Abstract 4966


Doxorubicin (DOX) and related drugs are incompletely cleared from the heart, persist for months to years (Stewart et al Anticancer Res 1993), and slowly transform to longer-lived and more toxic secondary alcohol metabolites (eg, DOXOL) that are associated with an increased lifetime risk of congestive heart failure (Minotti et al J Pharmacol Exp Ther 2010). Anthracycline-related cardiotoxicity can be precipitated by enhanced stress or exposure to drugs that increase DOXOL formation. PIX is a novel aza-anthracenedione that is under development in patients with NHL that relapsed or was refractory after DOX-containing combination therapy. In preclinical models, PIX produced substantially less cardiotoxicity than DOX or MITOX even when animals were pretreated with DOX. Compared to anthracyclines and the anthracenedione analogue, mitoxantrone (MITOX), PIX lacks a hydroquinone moiety that binds iron and facilitates cardiotoxic free radical reactions. To further establish that it is safe in patients with NHL and prior DOX exposure, PIX should also have no effect on or diminish DOXOL formation from the cardiac residues of first-line DOX.


To characterize the effects of PIX on DOXOL formation in a translational cardiac model of DOX administration followed by PIX administration. MITOX was used as the comparator.


Myocardial samples that were routinely discarded during aorto-coronary bypass grafting were dissected into strips and loaded with 10 μM DOX in plasma. After 30 min the strips were subjected to 2h multiple washouts to simulate post-treatment clearance. The strips were then incubated for 1.5h in fresh anthracycline-free plasma w/wo 1 μM PIX or MITOX. After the experiments, DOX(OL), PIX, and MITOX were extracted from the soluble fractions of the strips and assayed by HPLC. Pharmacometabolic interactions of 10 μM or 50 μM DOX with increasing concentrations of PIX or MITOX were also studied in NADPH-supplemented isolated soluble fractions.


After sequential DOX loading/clearance and PIX or MITOX administration, PIX:DOX ratios were significantly higher than MITOX:DOX ratios in soluble fractions of human myocardial strips. This correlated with inhibition of DOXOL formation by both PIX and MITOX, but PIX decreased DOXOL levels in a significant manner when compared to MITOX (Table). In isolated soluble fractions, both PIX and MITOX, in a concentration-dependent manner, inhibited the metabolism of 10 μM DOX to DOXOL; however, PIX caused stronger inhibition (Figure). Increasing DOX to 50 μM abated inhibition by MITOX but not by PIX.


PIX inhibited DOXOL formation in a translation model of human heart used to simulate sequential DOX and PIX administration. This finding was obtained under pharmacokinetically relevant conditions that probed DOX and PIX at their clinically documented plasma Cmax values of 10 μM or 1 μM, respectively. Studies with isolated soluble fractions suggest that PIX diminished DOXOL formation by noncompetitive inhibition of NADPH-dependent reductases. MITOX was less effective than PIX at diminishing DOXOL formation, which correlated with the lower MITOX:DOX ratios in treated human myocardial strips and with the competitive mode of action of MITOX. By inhibiting formation of toxic and long-lived DOXOL, PIX potentially is a cardiac tolerable therapeutic agent for patients with NHL that failed or relapsed after DOX treatment.

TreatmentDOX (nmol/g)Anthracenedione (nmol/g)Anthracenedione:DOX (%)DOXOL (nmol/g)
DOX 1.8 ± 0.2 − − 0.005 ± 0.001 
DOX + MITOX 1.5 ± 0.1 0.13 ± 0.02 11.3 ± 1.8 0.004 ± 0.002 
DOX + PIX 2.0 ± 0.2 0.35 ± 0.03* 16.9 ± 1.1* 0.002 ± 0.001 
TreatmentDOX (nmol/g)Anthracenedione (nmol/g)Anthracenedione:DOX (%)DOXOL (nmol/g)
DOX 1.8 ± 0.2 − − 0.005 ± 0.001 
DOX + MITOX 1.5 ± 0.1 0.13 ± 0.02 11.3 ± 1.8 0.004 ± 0.002 
DOX + PIX 2.0 ± 0.2 0.35 ± 0.03* 16.9 ± 1.1* 0.002 ± 0.001 

n= 5-9 (triplicate); * P<0.01 vs MITOX; P<0.05 vs DOX alone

Inhibition of DOXOL formation by PIX and MITOX in isolated soluble fractions of human myocardium incubated with NADPH (0.25 mM) and DOX for 4h. Control DOXOL was 0.15–0.19 nmoles/mg of protein at 10 μM DOX and 0.34–0.84 nmoles/mg of protein at 50 μM DOX. Each inhibition curve is the mean and SE of three experiments.


Salvatorelli:Cell Therapeutics, Inc: Research Funding. Gonzalez Paz:Cell Therapeutics, Inc: Research Funding. Singer:Cell Therapeutics, Inc: Employment. Menna:Cell Therapeutics, Inc: Research Funding. Minotti:Cell Therapeutics, Inc: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution