Pharmacokinetics and pharmacodynamics of the cathepsin S inhibitor, LY3000328, in healthy subjects

Abstract

Aim

The aim of this study was to assess the safety and tolerability, pharmacokinetics and pharmacodynamics of LY3000328 when administered as single escalating doses to healthy volunteers.

Methods

This was a phase 1, placebo-controlled, dose escalation study with LY3000328 in 21 healthy male volunteers. Subjects were administered escalating LY3000328 doses up to 300 mg with food in this single dose study. Blood samples were collected at set times post-dose for the assessment of LY3000328 pharmacokinetics and the measurement of cathepsin S (CatS) activity, CatS mass and calculated CatS specific activity.

Results

All doses of LY3000328 were well tolerated, with linear pharmacokinetics up to the 300 mg dose. The pharmacodynamic activity of LY3000328 was measured ex vivo showing a biphasic response to LY3000328, where CatS activity declines, then returns to baseline, and then increases to a level above baseline. CatS mass was also assessed post-dose which increased in a dose-dependent manner, and continued to increase after LY3000328 had been cleared from the body. CatS specific activity was additionally calculated to normalize CatS activity for changes in CatS mass. This demonstrated the increase in CatS activity was attributable to the increase in CatS mass detected in plasma.

Conclusion

A specific inhibitor of CatS which is cleared quickly from plasma may produce a transient decrease in plasma CatS activity which is followed by a more prolonged increase in plasma CatS mass which may have implications for the future clinical development of inhibitors of CatS.

What is already known about this subject

1. Impacts of specific cathepsin S inhibitors on animals have been reported.

2. We are not aware of a report in the peer-reviewed literature of administration of a specific cathepsin S inhibitor to human subjects.

What this study adds

1. A specific inhibitor of cathepsin S administered to healthy human subjects produced a transient decrease in plasma cathepsin S activity which was followed by a prolonged increase in plasma cathepsin S mass.

Introduction

Cathepsin S (CatS) is a cysteine protease which has been implicated in playing an important role in human disease 1–3. CatS has the ability to degrade many extracellular elements, such as elastins, collagens and proteoglycans which provide structure and support for tissues including vessel walls. It has been associated with the pathogenesis of cardiovascular disease such as atherosclerosis and abdominal aortic aneurysm 4–9. Inhibitors of CatS have been developed and are being evaluated as drug candidates for diseases including cardiovascular disease and autoimmune disorders 10–12.

LY3000328 is a potent and specific inhibitor of CatS, with a molecular weight of 484.52. Its chemical structure, physio-chemical properties and in vivo pharmacology will be detailed elsewhere 13. When exposed to a high concentration of HCl (pH <2.0), an oxetane ring in LY3000328 may open to form a chloroalcohol.

It was hypothesized that inhibition of CatS activity by LY3000328 would slow or stop abdominal aortic aneurysm (AAA) expansion and/or reduce the risk of AAA rupture through inhibition of CatS-mediated degradation of the extracellular matrix proteins, elastin and collagen 14,15.

Plasma CatS activity was measured as the primary pharmacodynamic (PD) biomarker in this study. In vitro experiments suggested that inhibition of CatS activity in plasma would be 50% of maximal when LY3000328 plasma concentration was approximately 60 ng ml⁻¹. It was assumed that a reduction in plasma CatS activity would be accompanied by a reduction in CatS activity in extravascular extracellular fluid, although the latter was not measured in this study. Plasma CatS mass and plasma cystatin C (CysC) concentrations were also measured as PD biomarkers, in order to explore the possibility that either might change in response to administration of LY3000328. CysC is a cysteine protease inhibitor produced by all nucleated cells at a constant rate and catabolized primarily by proximal renal tubules after glomerular filtration. It is a high affinity inhibitor of CatS 16.

CatS is also postulated to be involved in immune function and antigen presentation 17,18. As such, total immunoglobulins and lymphocyte counts were measured in this clinical trial.

Study I5U-MC-ANBB (Study ANBB) was a first-in-man study of LY3000328 to investigate the safety, tolerability, pharmacokinetics (PK) and PD of single escalating oral doses of LY3000328 administered to healthy subjects. The study sought to establish a maximum tolerated dose of LY3000328 in order to support further clinical studies and assess CatS activity as the primary marker of target engagement.

Methods

Study design

This was a single centre, investigator- and subject-blind, randomized, placebo-controlled, single dose, dose escalation study evaluating the safety, tolerability and PK/PD of LY3000328 in healthy subjects (ClinicalTrials.gov Protocol Registration Number: NCT01515358), conducted at the Lilly-National University of Singapore Centre for Clinical Pharmacology, Singapore.

Two cohorts of healthy subjects (nine subjects per cohort) each received escalating doses of LY3000328 during three alternating study periods (Supplementary Table S1). In each study period, six subjects received LY3000328 and three subjects received matching placebo. During the study, each subject was randomly assigned to receive two doses of LY3000328 and one dose of placebo. Specifically, subjects in cohort 1 were randomly assigned to receive two escalating doses of either 1, 10 or 100 mg of LY3000328 plus one dose of placebo. Subjects in cohort 2 were randomly assigned to receive two escalating doses of either 3, 30 or 300 mg of LY3000328 plus one dose of placebo, alternating with cohort 1 dosing. This design was considered robust as the preclinical data suggested PD effects were reversible and an adequate washout period was employed between dosing occasions.

Plasma (EDTA) samples for CatS activity, CatS mass, CysC and LY3000328 concentration were collected at 0 (predose), 0.5, 1, 2, 4, 8, 12, 24 and 48 h after each dose and frozen at −70°C until assayed.

LY3000328 contains a strained four-membered ring which has the potential to open and form a chloroalcohol in the low pH of the stomach in a fasted state. As it is known that the fed state raises the stomach pH 19, LY3000328 was dosed in healthy human subjects in the fed state to avoid the degradation of LY3000328 and the formation of the chloroalcohol degradant.

Safety was assessed prior to each dose escalation decision. Medical assessments, routine clinical laboratory tests (including haematology, clinical chemistry and urinalysis), vital signs, electrocardiograms (ECGs), treatment-emergent adverse events (TEAEs), immunoglobulins and concomitant medications were reviewed. The next LY3000328 dose level, in alternating cohorts, was initiated only if safety results from the preceding dose level were determined acceptable.

Each study period was completed over approximately 7 days, with an additional washout duration of at least 7 days between each period, resulting in at least 14 days between doses for each subject.

Dose selection

The initial dose level (1 mg) of LY3000328 chosen for this study was predicted to produce no detectable PD effect. Higher dose levels (such as 10 to 300 mg) were projected by PK/PD modelling and allometric scaling of pre-clinical data to result in LY3000328 concentrations which might eventually show efficacy in the treatment of AAA. In a murine model of AAA, the estimated average plasma effective concentration (EC₅₀) was 31 ng ml⁻¹ (90% confidence interval [CI] 14, 69 ng ml⁻¹). A predicted human EC₅₀ was obtained by multiplying the mouse EC₅₀ by the ratio of human to mouse CatS in vitro binding affinity (mouse = 1.38 nm, human = 5.48 nm). This enabled us to predict that a human 24 h exposure of approximately 3000 ng ml⁻¹ h (90% CI 1000, 10000 ng ml⁻¹ h) would be required to reduce AAA expansion rate by 50%. With an estimated human oral clearance of 6.8 l h⁻¹ (90% CI 1.6, 29 l h⁻¹), the projected human dose predicted to reduce AAA expansion rate by 50% (ED₅₀) was 20 mg (90% CI 5, 100 mg). The starting dose of 1 mg in this study was approximately five-fold below the lower bound of the uncertainty range providing an additional margin of safety for the first administration to healthy subjects. The highest planned dose of 300 mg was approximately three-fold higher than the upper bound of the uncertainty in the estimated clinical ED₅₀ and approximately two-fold higher than the estimated dose predicted to reduce AAA expansion rate by 90% (ED₉₀). The 300 mg top dose was also the maximum allowed by the specifications supporting the clinical trial material, which were set by the level of impurities in this early formulation.

The study was conducted in full accordance with the Declaration of Helsinki, principals of good clinical practice and local laws regarding the protection of the rights and welfare of human participants in biomedical research. The protocol was approved by an independent institutional review board and informed consent for all participants was obtained prior to screening.

Subject selection

Males and females, 35 to 70 years of age, with a body mass index (BMI) of between 18.5 and 32.0 kg m⁻² were eligible for participation in this study if they were judged to be in good health based on their medical history and physical examination including vital signs, ECG and clinical laboratory test results. Women were of non-childbearing potential due to surgical sterilization or were post-menopause. Subjects were also required to have normal renal function as renal clearance was postulated to be a major clearance pathway for LY3000328. This was defined by an estimated creatinine clearance calculated by the Cockroft−Gault formula of >80 ml min⁻¹ for subjects aged 35 to 50 and >70 ml min⁻¹ for those aged over 50 years. Non-study over-the-counter and prescription medications were not allowed for 7 and 14 days, respectively, prior to the first dose of study medication and until the completion of the study, unless deemed necessary and acceptable by the investigator.

Bioanalytical methods

Pharmacokinetics

Human plasma and urine samples obtained during this study were analyzed at Covance Laboratories, Inc., Madison, WI, USA. Plasma samples were analyzed for LY3000328 using a validated liquid chromatography tandem mass spectrometry (LC/MS/MS) method. The lower limit of quantification was 1 ng ml⁻¹ and the upper limit of quantification was 1000 ng ml⁻¹.

Human urine samples were analyzed for LY3000328 using a validated LC/MS/MS method. The lower limit of quantification was 250 ng ml⁻¹ and the upper limit of quantification was 25000 ng ml⁻¹.

Pharmacodynamics

Both CatS mass and CatS activity were measured with validated assays (enzyme-linked immunosorbent assay [ELISA] and mass spectrometry, respectively) 20.

The study protocol specified that CatS activity, CatS mass, CysC and LY3000328 concentrations would be measured up to 48 h after each dose, with LY3000328 additionally measured at 72 and 120 h after each dose.

Residual plasma from samples collected at 72 and 120 h after each dose for the measurement of LY3000328 concentration were also assayed for CatS mass, because upon review of the 0–48 h CatS mass data, it was apparent CatS mass had not returned to a pre-dose baseline. Plasma samples collected at 72 and 120 h post-dose were not collected under appropriate conditions to measure CatS activity, however.

Pharmacokinetic analysis

Pharmacokinetic analyses were conducted on data from subjects who received LY3000328 and had samples collected. PK parameter estimates for LY3000328 were calculated by standard non-compartmental methods of analysis using WinNonlin Professional Edition, Version 5.3 (Pharsight, Cary, NC, USA). The PK samples collected within ± 10% of the scheduled post-dose sampling times were included in the presentation of mean concentration−time plots. Mean concentrations were only calculated when quantifiable values for at least two-thirds of the subjects were reported. Plasma concentrations below the quantitation limit (BQL) that occurred prior to the first quantifiable concentration were assigned a value of 0 ng ml⁻¹. All other post-dose BQL concentrations were excluded from analysis.

Urine PK evaluation

Total urine output was collected and measured from subjects at approximately 0 to 12 h and 12 to 24 h post LY3000328 dose. Renal clearance of LY3000328 was calculated as the ratio of total LY3000328 amount excreted in the urine to the appropriate plasma area under the plasma concentration–time curve from zero to the time of the last quantifiable concentration (AUC(0,t)). The estimated unbound glomerular filtration rate (GFR) was calculated from the pre-dose serum creatinine for each subject using the Cockroft–Gault formula. The ratio of renal clearance to unbound GFR was calculated as the ratio of renal clearance to unbound GFR.

Statistical analyses

It was planned that at least 18 subjects (nine per cohort) would complete the study. Subjects who withdrew or were withdrawn from the study for reasons other than a drug-related adverse event (AE) were replaced so that the targeted numbers of subjects for safety review and data collection could be achieved. The sample size that was selected is customary for phase 1 studies evaluating safety, tolerability, PK and/or PD parameters, and provided adequate placebo control for each subject and each dosing occasion.

All analyses were pre-specified in the statistical analysis plan. PK and PD analyses were conducted on the full analysis set. This set included all data from all randomized subjects receiving at least one dose of study drug according to the treatment the subjects actually received. Safety analyses were conducted for all enrolled subjects, whether or not they completed all protocol requirements. The minimum observed CatS activity and time to minimum observed CatS activity of each subject were summarized for each treatment group. A linear mixed effects model was applied for the inference of percent change from baseline in CatS activity. The response variable of the model was the percent change from baseline in CatS activity and the independent variables were baseline CatS activity, time from dosing, dose, and time-by-dose interaction. Correlations among within-subject repeated measures were considered by assigning subject as random effect within an appropriate covariance structure.

Post hoc analyses for CatS mass and CatS specific activity were conducted and these were considered as additional PD parameters. The analysis method was the same as the method used for CatS activity. No adjustment for multiple testing was performed. Data analysis was performed using SAS® Version 8.2 or greater.

Results

Subject disposition

Data were analyzed for a total of 21 healthy male subjects of Asian descent who were randomly assigned to treatment and received at least one dose of study drug. Nine subjects were enrolled into cohort 1 and 12 subjects were enrolled into cohort 2 ( Table 1). Of the 21 total subjects enrolled, 18 completed the study. All three subjects who discontinued treatment were in cohort 2 and received at least one dose of LY3000328. Two of the discontinuations were due to subject decision (unable to commit further to the study). The other subject was discontinued by the sponsor because he did not meet the required entry criteria.

Table 1.

Demographic and baseline characteristics

Demographic and baseline characteristics	Cohort 1	Cohort 2	Overall
Number of subjects studied	9	12	21
Age (years), mean (SD)	47 (10)	40 (5)	43 (8)
Body mass index, (kg m−2)	23	25	24
Gender, male, n (%)	9 (100)	12 (100)	21 (100)
Smoking status, yes, n (%)	3 (33.3)	7 (58.3)	10 (47.6)
Alcohol use (units per week)*	1.4	0.3	0.8
Ethnicity, Asian, n (%)	9 (100)	12 (100)	21 (100)

n = number of patients; SD = standard deviation; ^*1 unit = 12 oz or 360 ml of beer; 5 oz or 150 ml of wine; 1.5 oz or 45 ml of distilled spirits.

Safety

No deaths or other serious adverse events occurred during this study and no subjects discontinued the study because of an AE. Of the 21 subjects who received LY3000328, five subjects (including one subject who received placebo) reported a total of eight AEs that were possibly related to study drug as judged by the investigator. All eight of these AEs were mild in severity and included isolated incidences of dizziness, vomiting, abdominal distension, headache, nausea and pyrexia.

There were no clinically significant alterations in laboratory values (including clinical chemistry, haematology, immunoglobulin assessments and urinalysis), vital signs or ECGs.

PK evaluations

Plasma

The arithmetic mean plasma concentration vs. time profiles on a log scale are shown in Figure 1. Table 2 summarizes the non-compartmental PK parameter calculations by treatment group.

Figure 1.

Mean LY3000328 plasma concentration−time plot (log linear). , LY3000328 1 mg; , LY3000328 3 mg; , LY3000328 10 mg; , LY3000328 30 mg; , LY3000328 100 mg; , LY3000328 300 mg; , LOQ = 1 ng ml⁻¹

Table 2.

Summary of non-compartmental LY3000328 plasma pharmacokinetic estimates

	Geometric mean (%CV geometric mean)
	LY3000328 1 mg	LY3000328 3 mg	LY3000328 10 mg	LY3000328 30 mg	LY3000328 100 mg	LY3000328 300 mg
n	6	6	6	6	6	6
Cmax	8.73	30.1	81.4	287	908	3870
(ng ml−1)	15	21	19	26	17	20
tmax*	4.05	3.04	3.05	3.53	4.05	2.05
(h)	(1.05–4.05)	(1.05–4.05)	(1.05–4.05)	(1.05–6.00)	(2.05–4.07)	(0.77–4.13)
t1/2†	4.68	5.16	5.80	7.62	5.65	5.96
(h)	(3.82–5.31)	(4.85–5.52)	(5.09–6.61)	(6.25–15.6)	(4.38–7.70)	(5.06–8.16)
AUC(0,∞)	72.8	231	830	2480	7990	24900
(ng ml−1 h)	7	15	10	18	16	15
AUC(0,24 h)	69.9	219	773	2280	7560	23700
(ng ml−1 h)	7	15	10	18	14	15
CL/F	13.7	13.0	12.0	12.1	12.5	12.0
(l h−1)	7	15	10	18	16	15
Vz/F	92.8	96.7	101	133	102	104
(l)	14	15	14	44	20	18

AUC(0,∞) = area under the concentration−time curve from time zero to infinite time; AUC(0,24 h) = area under the concentration−time curve from time zero to time 24 h; CL/F = apparent total body clearance of drug calculated after extra-vascular administration; C_max = maximum observed concentration; CV = coefficient of variation; n = number of subjects; t_max = time to maximum concentration; t_1/2 = half-life; V_z/F = apparent volume of distribution during the terminal phase after extra-vascular administration.

^*Median (range).

^†Geometric mean (range).

A longer half-life (7.62 h) was observed in the 30 mg group than in other groups. This observation was due to a single subject with a half-life (15.6 h) longer than other subjects given this dose. This subject did not receive other doses of LY3000328 because this subject was a replacement for another subject who discontinued early.

Urine

The estimated renal clearance of LY3000328 ranged from 120 to 148 ml min⁻¹ (%CV ranged from 8.5% to 20%) in the healthy subjects of this study. The mean unbound fraction of LY3000328 in human plasma (f_u) was approximately 81%. On average, across the oral dose range of 1 to 300 mg, the percent of the dose recovered as unchanged LY3000328 in urine ranged from 48.4% to 66.5% (%CV ranged from 5.9% to 15.1%), indicating that as predicted urinary excretion of LY3000328 is a major human clearance pathway. Additionally, a large proportion of the oral dose was recovered in urine, indicating that oral bioavailability was at least 48.4% to 66.5%.

PD evaluations

CatS activity

The mean time to minimum observed CatS activity across all LY3000328 dose levels ranged from 3.3 to 4.7 h, (Table 3). The means of the minimal observed CatS activity of LY3000328 were shown in Table 3. The least-square mean and standard error of percent change from baseline in plasma CatS activity is shown in Figure 2. Plasma CatS activity decreased in a dose-dependent manner between 2 to 24 h. A 13% to 20% inhibition was seen with the 1 and 3 mg doses, which was only marginally greater than predicted from in vitro and animal in vivo data, compared with a maximal inhibition of 98% at the 300 mg dose. Plasma CatS activity increased at 48 h for the 100 and 300 mg doses.

Table 3.

Summary of minimum cathepsin S (CatS) activity

LY3000328 dose (mg)	Mean (SD) minimum CatS activity (μU ml−1)	Mean (SD) time of minimum CatS activity (h post-dose)
Placebo	18.24 (4.47)	5.1 (11)
1	18.93 (2.13)	4.3 (0.8)
3	13.53 (2.11)	4.3 (0.8)
10	13.35 (1.73)	4.3 (0.8)
30	6.33 (1.69)	4.7 (1.0)
100	3.77 (0.83)	3.3 (1.0)
300	0.97 (0.38)	3.3 (1.0)

Figure 2.

Least-square mean (standard error) cathepsin S activity percent change from baseline vs. time. , placebo; , LY 1 mg; , LY 3 mg; , LY 10 mg; , LY 30 mg; , LY 100 mg; , LY 300 mg

The least-square mean of CatS activity was statistically significantly different compared with placebo following 30, 100, 300 mg LY3000328 from 2 to 6 h, 2 to 8 h and 0.5 to 12 h post-dose, respectively.

CatS mass

The least-square mean and standard error of CatS mass percent change from baseline (0 to 120 h) is shown in Figure 3. There was little change in CatS mass following 1, 3 or 10 mg doses of LY3000328 compared with placebo. CatS mass increased after 8 to 12 h following 30, 100 and 300 mg LY3000328 doses, and continued to increase after 48 h in the 100 and 300 mg treatment groups. By 48 h, a dose-dependent increase in CatS mass was observed, with a 100% increase compared with baseline in CatS mass in the 300 mg treatment group. This trend toward increasing CatS mass continued to 120 h in the 100 mg and 300 mg cohorts.

Figure 3.

Least-square mean (standard error) cathepsin S mass percent change from baseline vs. time (0–120 h). , placebo; , LY 1 mg; , LY 3 mg;, LY 10 mg; , LY 30 mg; , LY 100 mg; , LY 300 mg

CysC

The least-square mean and standard error of CysC percent change from baseline (0 to 48 h) is shown in Figure 4. In all dose groups, CysC remained relatively constant, within a band ± 10% of the pre-dose value.

Figure 4.

Least-square mean (standard error) cystatin C percent change from baseline vs. time. , placebo; , LY 1 mg; , LY 3 mg; , LY 10 mg; , LY 30 mg; , LY 100 mg; , LY 300 mg

CatS specific activity

Since CatS activity may reflect changes in CatS mass, CatS specific activity was calculated to normalize activity for changes in mass. Specific activity was calculated for time points up to 48 h post-dose by dividing the CatS activity by CatS mass. The least-square mean and standard error of CatS specific activity percent change from baseline is shown in Figure 5. CatS specific activity decreased in a dose-dependent fashion between 2 to 24 h. At 48 h, most doses returned to the pre-dose level. Maximal decrease in CatS specific activity occurred at 4 h across all doses. The decrease was relatively small for the 1, 3, and 10 mg LY3000328 dose groups at 4 h (≤30%, P ≤ 0.02). For the 30, 100 and 300 mg LY3000328 dose groups, the decrease was >60% at 4 h (P < 0.001).

Figure 5.

Least-square mean (standard error) cathepsin S specific activity percent change from baseline vs. time. , placebo; , LY 1 mg; , LY 3 mg; , LY 10 mg; , LY 30 mg; , LY 100 mg; , LY 300 mg

Detailed numerical values for CatS activity, CatS mass and CatS specific activity least-square means and standard errors as depicted in Figures 3, 4, and 6 are provided in Supplementary Table S2.

Figure 6.

Least-square mean (SE) cathepsin S activity percent change from baseline vs. time including predicted cathepsin S activity at 72 and 120 h. , placebo; , LY 1 mg; , LY 3 mg; , LY 10 mg; , LY 30 mg; , LY 100 mg; , LY 300 mg

Discussion

The doses planned for this first-human-dose study were selected based on non-clinical toxicology studies and analysis of preclinical PK/PD data to cover a dose range expected to produce a full range of minimal to maximal reduction of CatS activity. The alternating cohort, escalating dose study design employed enabled us to see AEs at each dose level prior to exposing subjects to the next higher dose. However, this design does not support the analysis of any possible order effects.

Single oral doses of LY3000328 ranging from 1 to 300 mg in healthy subjects were well tolerated by the healthy volunteers in this study. Hence, a maximum tolerated dose was not established. Analysis of vital signs indicated no consistent clinical effects, and statistical analysis of ECG characteristics demonstrated no significant prolongation of the corrected QT interval. No clinically significant or dose-related adverse trends in safety laboratory tests, including immunoglobulin levels, were observed.

LY3000328 was rapidly and well absorbed after oral administration. The t_max was approximately 2 to 4 h after dosing, and urine analysis results indicated that oral bioavailability was at least 48.4% to 66.5% in these study subjects. After single doses of 1 to 300 mg, the PK of LY3000328 were linear and plasma exposures by criterion of AUC(0,∞) were approximately dose proportional, and corresponding observed C_max values were slightly higher than dose proportional. The observed human oral clearance was from 12 l h⁻¹ to 13.7 l h⁻¹ which is approximately twice the value (6.8 l h⁻¹) projected for human oral clearance from animal PK data. After oral administration, LY3000328 is mainly cleared through urinary elimination. Renal clearance ranged from 1.3- to 1.9-times higher (%CV from 6.3% to 22.3%) than the estimated unbound GFR, indicating that LY3000328 renal clearance not only occurred predominantly by glomerular filtration but also that there was potential for an additional contribution from active tubular secretion.

LY3000328 concentration and CatS mass were assayed up to 120 h post-dose, whereas CatS activity and CysC were assayed only up to 48 h post-dose. CatS specific activity can thus be calculated only up to 48 h post-dose. Extrapolation of the observed CatS specific activity value at 48 h post-dose to 72 to 120 h post-dose were possible by assuming that in each subject the values at these later time points are identical to those at 48 h. This was possible because at 48 h post-dose, we did not expect CatS specific activity to decrease further as CatS specific activity had returned to near baseline in all groups at 48 h post-dose with LY3000328 plasma concentrations approaching BLQ. CysC concentrations showed minimal variation in the study and would not be expected to be significantly different at either 72 or 120 h post-dose, than at 48 h post-dose. If CatS specific activity had increased after 48 h, extrapolation of CatS activity at 72 and 120 h post-dose would be conservative and would continue to support the position that CatS activity at 72 and 120 h was significantly above baseline.

Therefore, we can make the assumption CatS specific activity was effectively constant between 48 and 120 h post-dose.

Using the extrapolated values of CatS specific activity at 72 and 120 h post-dose in each subject, CatS activity can be estimated at these time points from the CatS mass measurements. Measured CatS activity values and predicted CatS activity values are displayed together on Figure 6. We believe it is reasonable to examine both the measured and calculated values in order to obtain a greater understanding of the impact of LY3000328 administration on CatS activity over time. It is evident that there was a biphasic response to LY3000328, in that CatS activity declined, then returned to baseline, and then increased to a level above baseline. Interestingly, if the entire time period from 0 to 120 h post-dose was considered, it appeared that CatS activity following a 300 mg dose of LY3000328 was above baseline for a larger fraction of this time period than it was below baseline. As the final sample was stopped at 120 h post-dose, the total time that CatS activity remained above baseline cannot be determined. However, we hypothesize that if a 300 mg dose of LY3000328 was to be given once per week, the steady-state impact on time-averaged CatS activity over the course of a week would be to increase, rather than decrease, it relative to baseline or placebo administration. Our goal in the development of a CatS inhibitor was to decrease the activity of CatS in order to preserve integrity of extracellular matrix proteins in vessel walls. If intermittent administration of a CatS inhibitor increases CatS activity, then this may be hypothesized to produce an undesirable impact on vessels.

We assume the activity of CatS in plasma depends on the quantity of CatS protein in plasma (CatS mass), the specific activity of the detected CatS protein molecules (possibly impacted by modifications of the protein catalyzed by other enzymes), and the presence of endogenous inhibitors such as CysC. The level of a protein cleavage product produced as a result of the enzymatic action of CatS or the level of a substrate specifically cleaved by CatS may somehow influence the production of CatS and/or the release of CatS from intracellular stores into plasma. Intracellular and/or extracellular CatS activity may thus be ‘sensed’ by the body and a homeostatic mechanism may exist to increase CatS activity when a decrease in it is detected.

One possible mechanism the body might employ to maintain CatS activity would be to increase CatS mass if/when an exogenous inhibitor reduces CatS activity. Alternatively, the body might regulate CatS activity by regulating the concentration of endogenous inhibitors. However, CysC concentrations were not seen to change in this study when LY3000328 was administered.

Other enzyme inhibitors have been shown to increase the mass of their respective enzymes including statins and CETP inhibitors 21,22. This likely occurs via a feedback loop as seen with statins and HMG-CoA reductase 23. However, it is common that the enzyme mass returns to baseline as the dosing of the inhibitor stops and drug concentrations decrease. In this study, we observed elevation of CatS mass long after LY3000328 was no longer detectable in plasma (limit of detection 1 ng ml⁻¹), and believe that CatS activity was also significantly elevated long after LY3000328 was no longer detectable. We do not know the mechanism behind these elevations. A generalized inflammatory response to LY3000328 or one of its metabolites seems unlikely, as neither general markers of inflammation (e.g. complete differential blood counts) nor immunoglobulin levels changed during the study.

It may be hypothesized that in response to LY3000328 administration, there was an increased synthesis of CatS, an increased release of CatS from intracellular stores and/or a reduction in the clearance of CatS from plasma. We are unaware of any reports detailing the clearance of CatS from plasma.

In healthy human subjects, we measured CatS activity and CatS mass in plasma and calculated CatS specific activity, at multiple time points after a single oral dose of a specific competitive inhibitor, LY3000328. As the dose of LY3000328 was increased, CatS specific activity decreased in a dose−response fashion. CatS specific activity returned to baseline as LY300328 was cleared from plasma. CatS mass began to increase from its baseline level as CatS specific activity was returning to baseline, and CatS mass was seen to be elevated above baseline 5 days after the inhibitor was administered. As with reduction in CatS specific activity, the increase in CatS mass was monotonically related to the dose of LY3000328 given. We believe that the combined effects of CatS specific activity reduction and CatS mass increase produced a biphasic change in CatS activity, such that in a dose−response fashion, CatS activity initially decreased, then returned to baseline and then increased to a level above baseline.

We conclude from our observations that a specific inhibitor of CatS which is cleared quickly from plasma may produce a transient decrease in plasma CatS specific activity which is followed by a more prolonged increase in plasma CatS mass which may have clinical implications for the development of inhibitors of CatS. In future clinical studies of enzyme inhibitors, study designers should consider prolonged post-dose measurement, beyond the projected exposure of the administered inhibitor, of both the quantity and activity of the target enzyme.

Competing Interests

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare no support from any organization for the submitted work, no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years and no other relationships or activities that could appear to have influenced the submitted work. CDP, MAD, MC, LHT, ESL, TS, and DJD are all employees of and hold stock in Eli Lilly and Company.

This analysis was sponsored by Eli Lilly and Company. The authors acknowledge Keri Poi, PhD for assistance with editing of the manuscript. The authors would like to acknowledge all investigators and their staff, as well as the participants for their contributions to this study.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's web-site:

Table S1 Dose allocation assignment for individual study cohorts

Table S2 Percentage change from baseline of CatS activity, CatS mass and Specific CatS activity per dose over time