Acute Hemodynamic Effects and Tolerability of Phosphodiesterase 1-Inhibition with ITI-214 in Human Systolic Heart Failure

Abstract

Background:

Phosphodiesterase type 1 (PDE-1) hydrolyzes cyclic adenosine and guanosine monophosphate. ITI-214 is a highly selective PDE-1 inhibitor that induces arterial vasodilation and positive inotropy in larger mammals. Here, we assessed pharmacokinetics, hemodynamics, and tolerability of single-dose ITI-214 in humans with stable heart failure with reduced ejection fraction (HFrEF).

Methods:

HFrEF patients were randomized 3:1 to 10, 30, or 90 mg ITI-214 single oral dose or placebo (n=9/group). Vital signs and electrocardiography were monitored pre- to 5-hours post-dose, and transthoracic echoDoppler cardiography pre- and 2-hours post-dose.

Results:

Patient age averaged 54 years; 42% female, and 60% African American. Mean systolic blood pressure decreased 3-8 mmHg (p<) and heart rate increased 5-9 bpm (p≤0.001 for 10, 30 mg doses, RM-ANCOVA). After 4 hours, neither blood pressure or heart rate significantly differed among cohorts (supine or standing). ITI-214 increased mean left ventricular power index (mPWRi), a relatively load-insensitive inotropic index, by 0.143 Watts/mL²·10⁴ (P=0.03, a +41% rise; 5-71 CI) and cardiac output by 0.83 L/min (P=0.002, +31%, 13-49 CI) both at the 30 mg dose. Systemic vascular resistance declined with 30 mg (−564 dynes·s/cm⁻⁵, P<0.001) and 90 mg (−370, P=0.016). Diastolic changes were minimal, and no parameters were significantly altered with placebo. ITI-214 was well-tolerated. Five patients had mild-moderate hypotension or orthostatic hypotension recorded adverse events. There were no significant changes in arrhythmia outcome and no serious adverse events.

Conclusions:

Single-dose ITI-214 is well-tolerated and confers inodilator effects in humans with HFrEF. Further investigations of its therapeutic utility are warranted.

INTRODUCTION

The cyclic nucleotides cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) are intracellular second messengers that mediate multiple signal transduction pathways^{1, 2}. They play critical roles in cardiovascular homeostasis including the regulation of vascular tone and smooth muscle proliferation, adrenergic and cholinergic signaling, nitric oxide and natriuretic peptide modulation, metabolism, cardiac contractility and relaxation, and fibroblast function to name a few^2–5. Once synthesized, cyclic nucleotides are exquisitely regulated within cellular nanodomains by members of the 11-member superfamily of phosphodiesterases (PDE). Six PDE members are known to be expressed in cardiac myocytes including PDE1, PDE2, PDE3, PDE4, PDE5, and PDE9³. Of these, PDE3 is best known as a regulator of cAMP, and its inhibition by small molecules such as milrinone is approved for treating acute decompensated heart failure⁶. PDE5 and PDE9 selectively hydrolyze cGMP and the former is widely used to treat pulmonary hypertension and erectile dysfunction. PDE9 inhibition is currently in clinical trials for heart failure with reduced ejection fraction (HFrEF) pursuant to pre-clinical data showing it confers anti-hypertrophic and anti-fibrotic responses in pressure-overloaded hearts⁷ and can improve hemodynamics and renal function in HFrEF ⁸.

PDE1 was first discovered in 1970⁹ and is unique among all PDEs by requiring calcium and calmodulin for its activation. It is expressed as one of three isoforms (PDE1A, 1B, and 1C), with PDE1B primarily expressed in brain¹⁰. Both PDE1A and 1C are constitutively expressed in human myocardium¹¹, with PDE1C predominating. PDE1C has near equal affinity for both cyclic nucleotides whereas PDE1A, the dominant isoform expressed in the hearts of small rodents, favors cGMP over cAMP by >20-fold. PDE1C is upregulated in human heart failure¹², and so has garnered attention as a potential therapeutic target for this syndrome. ITI-214 is a highly potent (IC₅₀ for PDE1A and 1C at 35 pM)¹³ non-isoform selective PDE1 inhibitor that increases load-independent measures of ventricular contractility while concomitantly inducing systemic arterial vasodilation in intact rabbits and dogs with or without HFrEF¹⁴. Both species express predominantly PDE1C as found in humans. ITI-214 also increases heart rate, consistent with PDE1 modulation of sinus node automaticity¹⁵ although this was less pronounced in the HFrEF condition¹⁴. Importantly, the inodilator effects from ITI-214 are unaltered by high-degree beta-adrenergic blockade¹⁴. In isolated rabbit myocytes, inhibiting PDE1 under rest conditions does not alter contraction whereas it is enhanced by PDE3 inhibition. Furthermore, while in the presence of low-level background cAMP stimulation from adenylate cyclase stimulation, both PDE3 and PDE1 inhibition similarly increase myocyte contraction, only PDE3 inhibition concomitantly significantly increases intracellular calcium¹⁴. This work and a study by Zhang et al¹⁶ has further shown PDE1C couples to adenosine type 2 receptors to enhance cAMP signaling, playing a key role in inodilator responses¹⁴ and cyto-protection against apoptosis induced by doxorubicin¹⁶. Given prominent expression of PDE1 in the brain, ITI-214 was first explored in neuropsychiatric and neurodegenerative disorders such as patients with Parkinson’s disease^{17, 18}. These studies confirmed tolerability in individuals with normal cardiovascular function, though hemodynamic profiling was very limited. The cardiovascular effects of ITI-214 in patients with HFrEF have not been previously determined, but could pave the way for its therapeutic use.

We hypothesized that in humans with HFrEF, PDE1 inhibition increases cardiac systolic contraction and decreases systemic vascular resistance (SVR), much as observed in preclinical models. In the present study, we examined the acute hemodynamic profile of escalating single-doses of ITI-214 in stable HFrEF subjects and their tolerability in a randomized, double-blinded, placebo-controlled trial conducted at two centers. Primary hemodynamic outcomes were cardiac power and output, blood pressure, and systemic vascular resistance.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study population

Adult patients age 18-80 years old with chronic stable HFrEF (New York Heart Association class II-III symptoms and left ventricular ejection fraction ≤35%) were enrolled at Johns Hopkins and Duke University Medical Centers (NCT03387215). Patients were all treated with heart failure guideline-directed medical therapy, at stable doses at least 14 days prior to screening. They were screened by medical history, physical examination, laboratories, electrocardiogram, and transthoracic echocardiogram. Inclusion/exclusion criteria are itemized in Supplemental Table I. Exclusions were significant hypotension at screening and randomization (systolic blood pressure [SBP]<95 mm Hg), persistent atrial fibrillation or complete heart block, a significant cardiovascular event within 90 days (myocardial infarction, coronary artery bypass graft surgery or other major cardiovascular surgery, or cerebrovascular accident), or chronic renal or liver disease. All subjects provided informed consent and the study was approved by each site’s Institutional Review Board which includes ethical approval.

Study design and procedures

Participants meeting entry criteria were enrolled into sequential cohorts evaluating 10 mg, 30 mg, and 90 mg of ITI-214. Within each of these 3 dose cohorts, patients were randomized 3:1 to a single oral dose of ITI-214 or placebo (thus n=9 per dose group). Blinded pharmacodynamic, hemodynamic, and tolerability/safety data were reviewed for each cohort by study leadership prior to advancing to the next higher dose. On the day of randomization (Day 1), participants were instructed to fast for ≥ 8 hours. Physical examination, urine pregnancy test (females), and urine drug test were performed, and eligibility confirmed. Baseline vital signs (including supine and orthostatic heart rate and blood pressure), 12-lead electrocardiogram (ECG), blood laboratories and urine sample, and Doppler echocardiogram were obtained. Participants were connected to continuous ECG monitoring via bedside telemetry as well as recorded continuous monitor (Mortara H12+ Holter devices). A liquid solution of study drug or placebo was prepared by an unblinded pharmacist and provided to the study teams for oral administration; antiseptic mouthwash strips were used to fully blind participants.

Supine heart rate and blood pressure were measured by automated cuff (Dinamap V100) every 10 minutes starting 30 minutes before to 120 minutes after study drug administration, followed by every 30 minutes to 5 hours post-dose (end-of-study). Orthostatic heart rate and blood pressure were measured at 60 ± 5 minutes and 240 ± 15 minutes post-dose, and at end of study visit (Day 5). At 120 ± 15 minutes post-dose, Doppler echocardiogram measurements and 12-lead ECG were repeated. A follow-up visit on Day 5 ± 1 day included a physical exam, 12-lead ECG, blood and urine samples for hematology, chemistry, and urinalysis, plasma sample for ITI-214/metabolite levels, and pregnancy test (if applicable). Adverse event reporting occurred at each study visit per protocol.

Pharmacokinetics

Blood samples at time 0 (pre-dose) and 5, 15, 30, 45, 90, 120 minutes after dosing on Day 1, and on Day 5 post-visit were obtained for ITI-214 pharmacokinetics. The following parameters were calculated: total dose (area under the curve) from 0 to 2 hours, maximal concentration (C_max), time to maximal concentration (T_max), and mean lag time (T_lag). Concentrations of cAMP and cGMP were determined from frozen (−700C) human plasma and urine samples using a validated and sensitive liquid chromatography with tandem mass spectrometric (LC/MS/MS) method. For both cyclic nucleotides the lower and upper limits of quantitation were established between 20 ng/ml and 5000 ng/ml, respectively.

Noninvasive hemodynamic measurements and outcomes

Comprehensive 2-dimensional Doppler transthoracic echocardiograms were performed by a dedicated research cardiac sonographer (GE Vivid e95, GE Healthcare, Chicago, Illinois) at Johns Hopkins and Duke University for each participant at baseline and at 120±15 minutes post-dosing. Blood pressure and heart rate were also recorded contemporaneous with the imaging and combined with the echoDoppler results for hemodynamic analysis. Quantitative image analysis was performed offline on a dedicated workstation with standard protocols in the Johns Hopkins Echocardiography Core, over-read by a single investigator with expertise in echocardiography (AH) to reduce interobserver variability. Recording of data and analysis was blinded to clinical data and treatment.

Aortic blood flow in mL/min was calculated from the velocity-time waveform from pulsed-wave Doppler in the left ventricular outflow tract multiplied by cross-sectional aortic diameter. Stroke volume was the integral of this waveform, and maximal flow its peak value. Cardiac output was the product of heart rate and stroke volume; SVR was the ratio of mean arterial pressure (Systolic Pressure x 1/3 + Diastolic Pressure x 2/3) to cardiac output. Effective arterial elastance was equal to end-systolic pressure (estimated by systolic arterial pressure × 0.9) divided by stroke volume. Mean ventricular power index was calculated by the equation: mPWRi = 10^{4 *} [1333×10⁻⁷ × cardiac output × SBP]/EDV² in units of Watts/mL² × 10⁴. Left ventricular end-diastolic volume (EDV) was measured from apical 2 and 4 chamber views using Simpson’s rule, and end-systolic volume calculated as EDV – stroke volume and left ventricular ejection fraction as stroke volume/EDV, using Doppler derived stroke volume for overall consistency. Pulsed-wave Doppler spectra of trans-mitral inflow and tissue Doppler imaging of the lateral mitral annular (E’) velocities were used to assess diastolic function including E/E’ ratio and isovolumic relaxation time.

Statistical analysis

The sample size (n=9/dose) was chosen to detect a 20% change in mean left ventricular power index (mPWRi) with a correlation coefficient of 0.9 between pre- and post-drug mean ventricular power, and difference between ITI-214 and placebo with a two-sided alpha of 0.05 with at 80% power. Cardiac and hemodynamic variables were analyzed two ways. Absolute changes in hemodynamic parameters were determined between paired baseline and 2 hr post-dose values, and analyzed by analysis of covariance that included treatment versus placebo and baseline values as co-variates, and change in the parameter as the dependent variable. The percent change in both mPWRi and cardiac output was also determined, using pre-post paired data within each dose group, and a Wilcoxon test with no multiple comparisons correction. For clinical variables among the cohorts, we used Kruskal-Wallis or Chi-square tests for categorical variables. Time-course data for blood pressure and heart rate were compared between drug and placebo using ANCOVA. Change in these parameters between supine and upright position after 1, 3, and 5 minutes of standing were examined by ANOVA. Analysis was performed using Graphpad Prism Version 8 or Systat Version 10. Study investigators had full access to all the data in the study and take responsibility for its integrity and the data analysis.

RESULTS

Trial participants

Fifty-two participants were enrolled between August 2018 and March 2020, of which 36 met inclusion criteria at screening and were randomized. One participant randomized to placebo failed to meet study blood pressure inclusion criteria on the day of study drug administration and was not dosed or included in the analysis. The remaining 35 completed the study; 8 received placebo and 27 received one of three doses of ITI-214 (10 mg, 30 mg, 90 mg).

Baseline characteristics by treatment assignment group are shown in Table 1. Mean age was 54 years, 42% were female and 60% African American. The majority (69%) had non-ischemic dilated cardiomyopathy, and there were no significant differences in clinical status, co-morbidities, or percent with an implantable cardioverter defibrillator. Medical treatment was similar with the exception of higher angiotensin receptor neprilysin inhibitor use in placebo and ITI-214 30 mg cohorts. Baseline hemodynamics including echo Doppler measures of cardiac contractile function were similar among the cohorts.

Table 1.

Baseline demographic, clinical and hemodynamic characteristics of patients with heart failure randomized to phosphodiesterase-1 inhibitor ITI-214 versus placebo.

	ITI-214 Dose
	10 mg (N=9)	30 mg (N=9)	90 mg (N=9)	Placebo (N=9)
Female	4 (44)	4 (44)	4 (44)	3 (38)
Age (years)	48 ± 14	57 ± 5	55 ± 15	55 ± 11
Race
Caucasian	2 (22)	5 (56)	3 (33)	3 (38)
African American	6 (67)	4 (44)	6 (67)	5 (62)
Other	1 (11)	0 (0)	0 (0)	0 (0)
Body mass index (kg/m2)	40.4 ± 9.5	33.6 ± 6.3	33.7 ± 9.7	31.5 ± 5.6
New York Heart Association Class
II	8 (89)	8 (89)	9 (100)	6 (75)
III	1 (11)	1 (11)	0 (0)	2 (25)
Ischemic Cardiomyopathy	4 (44)	2 (22)	3 (33)	2 (25)
LV Ejection Fraction (%)	27.1 ± 4.5	25.4 ± 6.2	23.9 ± 7.4	25.0 ± 6.6
ICD	6 (67)	7 (78)	8 (89)	7 (88)
Comorbidities
Coronary artery disease	5 (56)	2 (22)	4 (44)	1 (13)
Chronic kidney disease	0 (0)	3 (33)	1 (11)	1 (13)
Hypertension	7 (78)	6 (67)	7 (78)	5 (63)
Diabetes mellitus	3 (33)	4 (44)	5 (56)	2 (25)
Heart failure medications
Beta blocker	9 (100)	9 (100)	9 (100)	8 (100)
Ace inhibitor / ARB	6 (67)	0 (0)	6 (67)	2 (25)
ARNI	2 (22)	8 (89)	3 (33)	6 (75)
Diuretics	8 (89)	9 (100)	9 (100)	8 (100)
Hemodynamics
Heart Rate (min−1)	65 ± 13	66 ± 12	70 ± 12	72 ± 10
Systolic Blood Pressure (mmHg)	121 ± 14	116 ± 12	119 ± 17	114 ± 17
Diastolic Blood Pressure (mmHg)	71 ± 11	68 ± 13	71 ± 9	69 ± 11
LV End-diastolic Volume (mL)	208 ± 63	196 ± 43	202 ± 63	224 ± 90
LV End-systolic Volume (mL)	147 ± 66	137 ± 47	140 ± 63	159 ± 90
Cardiac Output (L/min)	3.9 ± 0.8	3.8 ± 0.9	4.3 ± 0.6	4.6 ± 1.0
Systemic Vascular Resistance (dynes • sec/cm−5)	1905 ± 408	1933 ± 696	1640 ± 351	1513 ± 390
Effective Arterial Elastance (mmHg/mL)	1.9 ± 0.4	1.9 ± 0.5	1.7 ± 0.3	1.6 ± 0.3
Mean Power Index (watts/mL2 •104)	0.38 ± 0.4	0.29 ± 0.1	0.33 ± 0.2	0.32 ± 0.2
Isovolumic Contraction Time (msec)	105 ± 24	91 ± 23	71 ± 16	93 ± 20
Mitral Valve E/A Ratio	1.2 ± 0.6	1.4 ± 1.7	1.7 ± 1.0	1.36 ± 1.0

Values are mean ± SD for continuous variables and n (%) for categorical variables. ARB: angiotensin receptor blocker; ARNI: angiotensin receptor/neprilysn inhibitor; ICD: implantable cardioverter defibrillator; LV: left ventricular.

Pharmacokinetics of ITI-214

Following single oral dosing of ITI-214, mean and peak plasma ITI-214 concentrations increased over the first 45 minutes in a dose-dependent manner (Figure 1). The mean time to peak concentration ranged between 0.75 to 1.1 hours for the three dosing cohorts, and median lag time between 0.0 to 0.08 hours. After five days, drug levels at each dose declined to 2-3% of peak concentration, consistent with a half-life of ~22 hours.

Figure 1. Pharmacokinetics of phosphodiesterase-1 inhibitor ITI-214 in patients with heart failure with reduced ejection fraction.

Time course of plasma ITI-214 concentrations in each dosing cohort shows dose dependent levels. Peak concentration was achieved within 45 minutes after oral dosing, remaining at this level out to 2 hours post dose. Data are displayed as mean ± 95% CI. Results of 2W repeated measures ANOVA : p<10⁻⁹ for dose, time, and dose*time interaction. Tukey multiple comparisons test shown: 10 vs 30 mg: * p≤0.007, # p=0.15, † p<0.001; 10 vs 90 mg: $ p=0.01, ‡ p<0.001; 30 vs 90 mg: § p<0.001.

Effect of ITI-214 on arterial pressures and heart rate

Baseline arterial blood pressure and heart rate were similar in all cohorts prior to dosing (Table 1). Figure 2 shows time-course results for change in each parameter over the primary 2-hour observation period and then after 4-hour post-dose follow-up. By repeated measures regression analysis, SBP declined significantly in each ITI-214 dosing cohort, rising slightly in placebo, such that by 2 hours, SBP had fallen by 3-8 mmHg (95% CI −4 to −22.7; 8.7 to 14) varying with PDE1-inhibitor dose. Repeated measures ANCOVA found each dose impacted the slope of the SBP-time plot compared with placebo (p=0.023, 0.007, and <0.001, for 10, 30, and 90 mg, respectively). Diastolic BP also declined with ITI-214 at each dose and was significantly different from placebo (drug effect: p=0.004, <0.001, and 0.078, respectively). Heart rate increased over time in the 10 and 30 mg cohorts but was not significantly altered in 90 mg or placebo, though the slope of each ITI-214 heart rate-time relation was different from placebo (p<0.001, <0.001, and 0.029, respectively). At 4 hours post dosing, vital signs did not significantly differ between the groups based on Kruskal Wallis test.

Figure 2. Blood pressure and heart rate changes from baseline through 2-hour post-dosing monitoring period as well as at 4-hours post-dosing with ITI-214 or placebo in patients with heart failure with reduced ejection fraction.

Data are mean ± 95% CI. P-values for within dose-group regressions over 2-hour period are displayed in each plot. See main text for covariance results versus placebo at each ITI-214 dose.

Cardiac and Systemic Hemodynamic Response to ITI-214

Hemodynamic results derived from echo-Doppler and vital sign data measured at the time of the imaging procedure are shown in Figure 3. Absolute change from baseline to 120 minutes following dosing for the primary outcomes of mPWRi, cardiac output, SVR, and mean arterial BP (mABP), as well as heart rate, ejection fraction, total ventricular afterload (Ea), and diastolic parameters are shown in Figure 3A. Median PWRi increased by 0.018 (IQR 0.122), 0.143 (IQR 0.154), and 0.083 (IQR 0.126) Watts/mL²·10⁴ at 10, 30 and 90 mg doses respectively (p=0.03 for 30 mg). Median cardiac output rose 0.717 (IQR 1.804; p=0.049), 0.830 (IQR 1.172; p=0.002) and 0.666 L/min (IQR 0.844; p=0.09) in each cohort respectively. These reflected significant increases of 32% or 33% in mPWRi, and 35% or 18% in CO at 30 and 90 mg doses, respectively (Figure 3B). ITI-214 also reduced systemic vascular resistance (−564, [IQR 312; P<.001]; −370 [IQR 189; P=.016] dynes·s/cm⁻⁵ at 30 and 90 mg, respectively), and significantly reduced total LV afterload (Ea) at all doses. There was little effect on end-diastolic volume or isovolumic relaxation time, but a small rise in E/E’ at the 30 mg dose. Viewed together, these hemodynamic responses support inodilator activity by ITI-214 in HFrEF patients.

Figure 3. Hemodynamic responses comparing baseline to 2-hours post-dosing with single dose of phosphodiesterase-1 inhibitor - ITI-214, or placebo.

A) Panels display violin plots for absolute change in each parameter before and after dosing, with all raw data, and median, 25^th, and 75^th percentile range. Groups received placebo (P), or 10, 30, 90 mg ITI-214. P values above each group are from analysis of covariance with placebo comparison and baseline values for each patient serving as covariates, and the dependent variable being change between pre-post measurements. B) Percent change for mean LV power index (mPWRi) and cardiac output for pre versus post dose in each patient. P-values show results for Wilcoxon test of pre-post percent change for each patient within a dose cohort, without a multiple comparisons correction. Ea: effective arterial elastance; EDV: end diastolic volume; EF: ejection fraction; E/E’: ratio of early transmitral flow/tissue velocity; IVRT: isovolumic relaxation time; mPWRi: mean LV power index; SVR: systemic vascular resistance.

Tolerance outcomes

Mild adverse events were reported in 11 of the 27 subjects treated with ITI-214 and in 1 placebo subject (Supplemental Table II). These included parosmia (n=4, 15%), hypotension (n=3, 11%), orthostatic hypotension (n=3, 11%), headache (n=2, 7%), taste disorder (n=2, 7%), and feeling hot (n=2, 7%). The 6 adverse events related to blood pressure change occurred in 5 different participants, all having received ITI-214 and occurring on study randomization day (Supplemental Table III). One participant in the 10 mg cohort developed hypotension at 4 hours and orthostatic hypotension at 5 hours after dosing. Two in the 30 mg cohort developed orthostatic hypotension (one each at 1 hour and 4 hours post-dose) and two in the 90 mg cohort developed hypotension (one each at 30 minutes and 4 hours post-dose). All events self-resolved on study randomization day without interventions.

Orthostatic blood pressure and heart rate changes are provided in Supplemental Figure I. No SBP changes were significantly different than what was observed with placebo. Heart rate increased upon standing in all groups and was not different from placebo at any of the ITI-214 doses although heart rate increases were numerically larger in the treated participants. Mean QT interval corrected for heart rate did not increase significantly at 120 minutes post-dose when compared to baseline in any of the cohorts (change in QTc in placebo: 6 ± 26; 10 mg: 1 ± 9; 30 mg: −4 ± 18; 90 mg: −4 ± 24 msec) as well as on Day 5 (change in QTc in placebo: 6 ± 21; 10 mg: −13 ± 17; 30 mg: −19 ± 22; 90 mg: −16 ± 19 msec). There was also no clinical change in arrhythmia outcomes including number of hourly supraventricular or ventricular arrhythmia runs and percent burden of supraventricular or ventricular ectopic beats, compared to pre-dosing, as calculated from continuous recorded Holter monitoring (Supplemental Table IV). There were no serious adverse events reported at any time in this study.

Effect of ITI-214 on Plasma cAMP and cGMP

Plasma cGMP and cAMP levels were measured at baseline and 2 hours after dosing. Overall, treatment with ITI-214 did not significantly change either cyclic nucleotide level in the blood (p=0.9 and 0.5 for dose x time interaction, and 0.5 and 0.25 for dose alone by 2WANOVA). This is similar to results we reported in conscious dogs, and likely reflects compartmentation of the regulated cyclic nucleotides that are not extruded into extracellular/vascular space. However, cyclic nucleotide levels did decline slightly over the 2-hour observation period, independent of treatment group (Supplemental Figure II). The cause is uncertain but could reflect the subject’s being at rest during this period.

DISCUSSION

In this first study of a PDE1 inhibitor (ITI-214) in patients with HFrEF we report that a single oral dose of 30 or 90 mg results in short-term (2-hour) increases in mean left ventricular power index, cardiac output and heart rate while inducing systemic arterial vasodilation. We find in this study of 27 dosed human subjects, single-dose ITI-214 is well-tolerated with acute hemodynamic effects similar to those first reported in pre-clinical models, and without serious adverse effects including new or sustained arrhythmia. A single dose of ITI-214 resulted in self-resolving mild-moderate hypotension or orthostatic hypotension in approximately one-fifth of participants.

The hemodynamic profile of acute ITI-214 in humans with HFrEF is analogous to that reported in conscious dogs with normal or dilated failing hearts and in anesthetized rabbits¹⁴. In this preclinical study, cardiovascular effects were determined by pressure-volume analysis to separate load-independent increases in contractility from concomitant afterload decline. The rise in cardiac output in the current trial could be ascribed to arterial vasodilation and/or positive inotropy, but the rise in mPWRi provides a less load-sensitive indication that some contractility change likely contributed^{19, 20}. While heart rate also increased in normal dogs and rabbits, dogs with HF had a blunted response, and in rabbits, beta-blockade or fixed-rate atrial pacing prevented HR change yet inodilator activity by ITI-214 remained intact. While all HFrEF patients in the present study were chronically receiving beta-blockade, we still observed a heart rate increase at higher doses of ITI-214. This could be caused by cAMP-dependent activation of sinoatrial automaticity by PDE1 inhibition¹⁵ or a baroreflex response as mABP also declined, though our pre-clinical data suggest neither are required for inotropic effects of PDE1 inhibition. Chronically elevated HR, whether intrinsic or coupled to pharmaco-therapy has been associated with worse outcomes in HFrEF. Whether the increase in HR observed in the current study following short-term PDE1 inhibition persists in HFrEF patients upon chronic dosing remains to be determined.

With the exception of Hashimoto et al¹⁴, all prior studies of PDE1 inhibitors in vivo have been performed in small rodents. Two recent PDE1 inhibitors, LuAF41228 and LuAF58027, were tested in rats that express primarily PDE1A in the heart and found to lower blood pressure by ~15 or more mmHg while elevating heart rate²¹. Mice also primarily express PDE1A in myocardium, and we observed much smaller changes in both parameters in response to acute ITI-214¹⁴. ITI-214 is >1000x more potent than either Lu-compound, and more selective for PDE1 over PDE3 (94,000-fold) or PDE5 (19,000 fold)¹³. Another selective PDE1 inhibitor is IC86340, though virtually all its reported data is in vitro^{22, 23}. To our knowledge, the only in vivo data is a 1-week study testing its impact on isoproterenol stimulated hypertrophy and fibrosis²². While effective, this was coupled to cGMP not cAMP signaling, and its impact on systemic resistance or cardiac contractility were not reported²².

The only PDE inhibitor currently established as a therapy for heart failure targets PDE3²⁴. In cardiomyocytes, PDE3 inhibition primarily increases cAMP to activate protein kinase A, enhancing excitation-contraction coupling and contractility^{3, 25–27}. This is primarily mediated by the PDE3A isoform, though inhibition of this protein is also pro-apoptotic²⁸ and pro-arrhythmic^{29, 30}. Furthermore, PDE3 inhibitors also suppress PDE3B that regulates a cAMP-dependent signalosome protective against ischemic stress, and its inhibition can have adverse consequences³¹. Small molecule PDE3 inhibitors cannot distinguish these isoforms as their ATP-binding catalytic sites are too similar, so this poses a therapeutic limitation^{31, 32}. To our knowledge, only one study has directly compared the effects of PDE3 versus PDE1 inhibition in the cardiac myocyte¹⁴, and none to date in an intact animal. In rabbit myocytes, PDE3 but not PDE1 inhibition augments cell contraction and calcium transients under basal conditions and with isoproterenol co-stimulation. However, when added to a low dose of the adenylate cyclase stimulator forskolin, PDE1 and PDE3 inhibition similarly augment contraction, but only PDE3 inhibition raises calcium transients¹⁴. While we cannot assess myocyte calcium changes in patients, the lack of arrhythmia change in our study is intriguing, and contrasts to results from acute exposure to PDE3 inhibitors³⁰.

PDE1C, the PDE1 isoform most expressed in human myocardium^{11, 22, 33} hydrolyzes both cyclic nucleotides with equal affinity. This isoform is primarily expressed in cardiomyocytes, whereas PDE1A is more prominent in vascular smooth muscle. However, both isoforms are also found in the brain and other organs. The cardiovascular role of PDE1 (primarily PDE1A) was first identified in vascular smooth muscle where it controls vascular tone, nitrate intolerance, and smooth muscle phenotype^34–36. PDE1 is also expressed in the pulmonary vasculature, and its upregulation in pulmonary hypertension has suggested a potential use of inhibitors in this disease ^{37, 38}. PDE1 inhibition in rat and mouse hearts first reported anti-fibrotic and anti-hypertrophic effects most likely related to enhanced cGMP-PKG signaling pathway^{22, 23}. However, PDE1C knockout mice also display attenuated myocardial hypertrophy, fibrosis and loss of contractility in a chronic pressure overload model¹². The potential effects of chronic PDE1 inhibition in the adult heart and other tissues remains to be determined. A mouse gene knockout of PDE1A demonstrated mild renal cystic disease but no cardiovascular toxicity ³⁹, and the PDE1C knockout also displays no basal cardiovascular dysfunction¹². Moreover, these models reflect embryonic deletion, and so may not predict pharmacological blockade in an adult disease state. Lastly, PDE1 inhibition has been linked to improved insulin signaling and may have a beneficial impact on diabetes ⁴⁰. Its inhibition was also recently found to protect against proteotoxicity in mice with misfolded protein disease ⁴¹. These effects from PDE1 inhibition are potentially relevant to various heart failure syndromes.

To date, the clinical experience with a putative PDE1 inhibitor is limited to vinpocetine, an extract from periwinkle that is sold in health food stores to enhance cognition and memory. It is a relatively weak PDE1 inhibitor (<10⁶ less potent than ITI-214) and also impacts ion channels and intracellular signaling ¹⁸. Though developed for neuropsychiatric and neurodegenerative disease, the current findings with ITI-214 suggest a potential application to heart failure. Several important issues remain to be addressed. First, does it have utility when administered chronically in heart failure patient, HFrEF, or those with preserved ejection fraction? Second, are inotropic effects dependent or largely independent of baroreflex stimulation, even when observed in beta-blocked heart failure patients? Third, does PDE1 inhibition alter myocardial energetics, and if so, how? Lastly, what if any are its chronic effects on neuropsychiatric and neurodegenerative behavior.

Our study limitations include a small cohort in each dose group that limited robust dose-response analyses. Another is our reliance on noninvasive functional analysis that while reasonable given this was the first human-HFrEF study, also reduced our ability to fully dissect its complex hemodynamic effects. Importantly, the study and all data analysis were performed in a double blinded manner to reduce bias. Lastly, both the study cohort characteristics (majority NYHA Class II and non-ischemic cardiomyopathy etiology) and single-dose nature of the study precludes making conclusions about ITI-214 effects on acute decompensated or ischemic heart failure, or chronic heart failure. This awaits future studies.

In summary, we report the first study testing a selective PDE1 inhibitor in humans with heart failure. We show its inhibition by ITI-214 acutely reduces arterial vascular resistance while augmenting cardiac contraction, output, and heart rate. It is generally well tolerated with no acute toxicity. Further investigation of its potential use as a heart failure therapeutic is warranted.

Clinical Perspective.

What is New?

• This is the first study of phosphodiesterase (PDE) type 1 inhibition in human heart failure.

• Selective inhibition of phosphodiesterase (PDE) type 1 with ITI-214 induces inodilator effects in patients with systolic heart failure after a single oral dose.

• Hemodynamic effects of PDE1 inhibition via ITI-214 included increase in mean left ventricular power index, increase in cardiac output, and reduction in systemic vascular resistance.

What are the Clinical Implications?

• Acute therapies for systolic heart failure are limited.

• Acute PDE1-inhibition may offer a novel, hemodynamically favorable inodilator treatment for systolic heart failure. Chronic treatment requires further investigation.

Abbreviations

cAMP

cyclic adenosine monophosphate

cGMP

cyclic guanosine monophosphate

ECG

electrocardiogram

EDV

end-diastolic volume

HFrEF

heart failure with reduced ejection fraction

mPWRi

mean ventricular power index

PDE

phosphodiesterase

PDE1C

phosphodiesterase type 1C

SBP

systolic blood pressure

SVR

systemic vascular resistance