PHARMACOKINETICS OF SILDENAFIL IN PATIENTS WITH A LEFT VENTRICULAR ASSIST DEVICE: A WORD OF CAUTION

INTRODUCTION

Patients with heart failure have physiologic changes associated with altered drug pharmacokinetics (e.g. hepatic dysfunction, hypoalbuminemia, or drug absorption). Drug pharmacokinetics in the heart failure population with LVADs remain unknown.

Sildenafil citrate is a phosphodiesterase-type-5 inhibitor used in patients with LVADs for treatment of pulmonary hypertension. Sildenafil is metabolized by hepatic cytochrome 3A4 (major) and 2C9 (minor) and has an active metabolite (n-desmethyl sildenafil).¹ Clinically significant adverse events (syncope, hypotension, visual changes, priapism) are associated with elevated sildenafil maximal plasma concentrations (C_max).^1–3 The purpose of this study was to measure and compare the C_max of sildenafil and n-desmethyl sildenafil in patients with LVADs relative to the reference range.

METHODS

This research was approved by the Vanderbilt institutional review board. We prospectively identified adults with LVADs who were hospitalized and prescribed sildenafil 20 mg orally three times daily with 4 previous doses documented to ensure steady-state. Exclusion criteria included hepatic cirrhosis, right ventricular mechanical support, or concomitant interacting medications.^1,2 A 5 mL blood sample was drawn to capture maximal plasma concentrations of sildenafil and n-desmethyl sildenafil 60–90 minutes after administration similar to previous methods.^4,5 The reference ranges for sildenafil and n-desmethyl sildenafil C_max were 59.6±31.4 ng/mL and 34.1±14.1 ng/mL respectively⁴, which were validated in the RELAX trial (sildenafil C_max 78 (IQR 35–130) ng/ml).⁵ For sildenafil and metabolite concentration assessment, a calibration curve across 2–1000 ng/mL was prepared. Samples were analyzed on a Thermo Quantum Ultra AM mass spectrometer with a Thermo Accela HPLC pump. The HPLC method used an Acquity BEH C18 1.7μm 2.1×100 mm column at 40°C. Data was processed with a coefficient of variation (CV) not to exceed 20% and a minimum Coefficient of determination (r²) of 0.95.

RESULTS

Patient characteristics are shown in Table 1. The mean age was 55 ± 9 years. Patients were predominantly white (n=7/12, 58.3%). The average body mass index was 28.8 ± 6.7. The majority of patients were implanted as a bridge to transplant (n=7/12, 58.33%). Sildenafil was initiated post-LVAD in 75% of patients, while 3 patients were treated before and after LVAD for pulmonary hypertension. All patients had continuous flow LVADs (n=5/12 HeartMate II, n=7/12 Heartware) implanted a mean of 290 days (range 6 – 671 days) prior. Two patients were treated with intravenous inotropes at the time of sampling due to right ventricular failure or proximity to surgery.

Table 1:

Sildenafil C_max, Demographic, and Laboratory Values

Patient	Sildenafil Cmax (ng/mL)	Gender	Etiology	Body Weight (kg)	SCr	CrCl (mL/min)	AST	ALT	Tbili	Days from LVAD surgery	RV function	Tricuspid Regurgitation
5	467.1	M	NICM	84.5	1.80	51	32	19	1.9	406	Normal	Moderate
2	339.5	M	NICM	75.8	0.83	107	31	19	1.9	26	Mildly depressed	Trace
11	217.2	M	NICM	126	2.24	36	28	15	0.5	9	Moderately depressed	Trace
12*	197.2	M	NICM	75.2	2.07	33	15	12	1.8	6	Severely depressed	Mild
1*	140.0	F	NICM	80.9	1.92	25	67	27	1	486	Moderately depressed	Moderate to Severe
3	139.1	F	NICM	75.1	1.12	57	38	19	0.3	259	Mildly depressed	Trace
8	131.5	F	ICM	79.4	1.53	46	38	16	1.2	9	Moderately depressed	Moderate
4	129.3	M	NICM	106.9	1.95	53	52	30	0.5	671	Mildly depressed	Trace
9	46.7	M	NICM	117.2	0.85	112	30	28	1.4	553	Moderately depressed	Moderate to severe
10	37.7	F	NICM	95.0	1.54	54	47	31	1.4	378	Moderately depressed	Trace
7	18.7	M	NICM	113.3	0.76	121	36	21	0.8	7	Severely depressed	Mild
6	17.9	M	NICM	78.0	1.45	60	23	17	0.6	666	Mildly depressed	Mild

Patients listed in descending order by sildenafil C_max.

M: Male, F: Female; NICM: Non-ischemic cardiomyopathy, ICM: Ischemic cardiomyopathy

Laboratory values measured in mg/dL unless otherwise noted

^*Patient on inotrope therapy

Sildenafil C_max and n-desmethyl sildenafil C_max exceeded the reference range in 66% and 75% of patients respectively. Sildenafil C_max ranged from 37.7 ng/mL to 467.1 ng/mL, with a 2 to 5-fold increase over the upper limit of the reference range in 25% of patients (Figure 1). The mean sildenafil C_max was 156.8 ± 124.5 ng/mL. Plasma n-desmethyl sildenafil C_max ranged from 28.7 ng/mL to 366.1 ng/mL, with a 3 to 7-fold increase over the upper limit of the reference range in 40% of patients (Figure 1). The mean n-desmethyl sildenafil C_max was 133.3 ± 102.0 ng/mL. For the majority of patients (n=10/12), trends were consistent among sildenafil and n-desmethyl sildenafil. Two patients with low sildenafil concentrations had n-desmethyl sildenafil concentrations within the reference range or greater. Neither age, weight, BMI, MELD score, gender, ethnicity, inotrope use, RV function, tricuspid regurgitation severity, or level of care were associated with sildenafil C_max greater than the reference value (p>0.05). Patients who had sildenafil concentrations greater than the reference value had significantly lower creatinine clearance (p=0.028). All patients were tolerating therapy at the time of assessment without reported adverse effects.

Figure 1. C_max of Sildenafil and n-desmethyl Sildenafil.

Individual patient C_max are represented by black diamonds. Gray boxes represent the reference range (mean + 1 standard deviation) of sildenafil C_max as 59.6 ± 31.4 ng/mL and n-desmethyl sildenafil C_max as 34.1 ± 14.1 ng/mL⁴

DISCUSSION

The key finding of this study is that two-thirds of sildenafil-treated patients with LVADs had C_max elevated above the reference range, with one-fourth experiencing more than double the reference upper limit.⁴ This population may be at increased risk for bothersome and serious concentration-related adverse effects, such as hypotension whose incidence is already increased at standard concentrations. ^2,3,5 The greatest C_max (467.1 ng/mL, patient 5) exceeded the literature average by more than 7-fold. Likewise, the C_max for n-desmethyl sildenafil was higher than reported, and the highest value (366.1 ng/mL, patient 1) exceeded the literature average by more than 10-fold.⁴ Patients with pulmonary hypertension experience higher sildenafil concentrations compared to healthy volunteers; however the magnitude of the difference we found was greater.¹ Our data suggests that conservative and individualized dosage titration may be warranted in patients with LVADs.

From this single-point assessment, we are unable to determine which specific pharmacokinetic parameters led to the concentration changes. Severe renal dysfunction (CrCl<30 mL/min) has been associated with elevated sildenafil concentrations, but mild or moderate dysfunction has not.¹ Creatinine clearance was associated with elevated concentrations in our study. However, the average CrCl was 62.9 mL/min, and only one patient had a CrCl<30 mL/min. Thus, the elevation is greater than expected based on renal function. Hepatic dysfunction is also associated with higher sildenafil plasma concentrations.¹ Severe hepatic dysfunction was not present in our study by laboratory assessment. We hypothesize that despite having normal measures of hepatic function, there may have been subclinical dysfunction in hepatic metabolizing capacity. We are unable to discern the relative contributions of LVAD and chronic heart failure to the altered pharmacokinetics.

We collected a single plasma sample for C_max 60–90 minutes after the sildenafil dose¹, however some may have peaked outside of this window. Additionally, concomitant food can delay or decrease C_max,¹ and we measured plasma concentrations in hospitalized patients without fasting. However, both of these limitations would have diminished the observed C_max. We lacked a control group, which prohibits distinction between heart failure or LVAD as the primary cause of our findings. Conclusions regarding associations between patient characteristics and sildenafil concentrations are limited in this small sample size. We did not assess patients’ genetic status. Poor metabolizers CYP 3A4 metabolizers display increased sildenafil C_max; however, that phenotype only affects 14.3% of the population.⁶ It is possible that genetic polymorphisms contributed to the high concentrations observed in our study. Patients with heart failure and LVADs are increased risk of concentrated-related sildenafil adverse events. Future research should perform complete pharmacokinetic assessment of drugs commonly administered to patients with LVADs.