Effect of Tadalafil on cardiac function and left ventricular dimensions in Duchenne muscular dystrophy: safety and cardiac MRI substudy results from a randomized, placebo-controlled trial

Abstract

Background: Inhibition of phosphodiesterase 5 (PDE5) was hypothesized to slow disease progression in Duchenne muscular dystrophy (DMD). Tadalafil, a once-daily PDE5 inhibitor, did not slow loss of ambulation in a phase 3 placebo-controlled trial. This report details the cardiac findings from this study.

Methods: Patients with DMD (N = 331) aged 7 to 14 years on stable glucocorticoids were randomized to tadalafil 0.3 mg/kg/day, 0.6 mg/kg/day, or placebo. Ejection fraction (EF), fractional shortening, and M-mode ventricular dimensions were measured on echocardiograms. 12-lead ECGs were centrally evaluated for heart rate and intervals, and qualitative diagnoses. Vital signs and unsolicited adverse events were collected throughout the study. Cardiac MRI (CMR) was collected in a subset of 27 patients. Z-scores for ventricular dimensions and volumes were calculated based on published age-normative reference values. Treatment differences for change in continuous ECG parameters and vital signs were compared using Wilcoxon rank-sum tests. Echocardiogram and CMR parameters were analyzed with an ANCOVA model.

Results: Tadalafil had no adverse effects on echocardiographic left ventricular (LV) EF or fractional shortening, ECG findings, or vital signs. Mean diastolic LV internal dimension (LVIDd) was increased in the tadalafil 0.6 mg/kg group versus placebo at Week 24 (+ 0.13 cm, p =.019) and Week 48 (+ 0.18 cm, p =.008), with a similar pattern observed for LV systolic dimensions (LVIDs). Mean LV end diastolic volume (EDV) measured by CMR also increased at Week 48 in the tadalafil 0.3 mg/kg (+ 13.0 ml, p =.047 vs. placebo) and 0.6 mg/kg (+ 12.0 ml, p =.08 vs. placebo) groups, with numerically smaller increases in LV EDV and commensurate increases in stroke volume and cardiac output. Z-scores for LVIDd and LV EDV were generally below the normal range at baseline and increased toward or within the normal range in the tadalafil groups but not in the placebo group.

Conclusions: No adverse effects of tadalafil on cardiovascular function were evident based on adverse events, echocardiograms, ECG, or vital sign measurements through 48 weeks in patients with DMD. The small mean increases in LVID and LV volume observed with tadalafil are consistent with PDE5 inhibitor pharmacology, but their clinical relevance in the context of LV tonic contraction in DMD is unknown and deserve further study.

Gov identifier: NCT01865084 (first registration date: 24-May-2013).

Keywords: Cardiac; Duchenne muscular dystrophy; PDE5 inhibitor; Tadalafil.

Fig. 1

Change from baseline in left ventricular internal dimension (LVID) at diastole (left panel) and systole (right panel) measured by M-mode echocardiography. Points represent the least squares (LS) mean change with standard error. *P <.05 for the difference between each of the tadalafil groups minus placebo are derived using ANCOVA with factors for treatment and baseline echocardiogram measure as covariate

Fig. 2

Change from baseline to endpoint (Week 48) or last observation in left ventricular CMR parameters. Points represent the least squares (LS) mean change with standard error. End systolic volume (A), end diastolic volume (B), cardiac output (C), stroke volume (D), left ventricular mass (E), left ventricular ejection fraction (F). ETD = estimated treatment difference for the difference of tadalafil minus placebo with 95% confidence interval from an analysis of covariance (ANCOVA) with change from baseline as dependent variable, baseline value as covariate, and treatment as fixed effect

Fig. 3

Vector plot of participant-level Z-scores versus age for end diastolic volume (top panels) and end systolic volume (bottom panels) measured by CMR. Each arrow represents an individual participant, with beginning of arrow = baseline Z-score plotted against baseline age (years) and end of arrow = endpoint Z-score (after 48 weeks) plotted against endpoint age

Fig. 4

Diastolic left ventricular internal dimension (LVID) Z-score measured by M-mode echocardiography at baseline by age

Fig. 5

Retrospective analysis on previously published GRMD data. Measures of (A) stroke volume, (B) cardiac output, (C) end diastolic volume, and (D) fractional shortening are shown as individual data points for each dog (dashed lines show the individual control dogs that increase SV values with age). Data were analyzed using paired T-tests (α = 0.05). Effect size is shown as Cohen’s d (d)