Impact of dosing regimen of custirsen, an antisense oligonucleotide, on safety, tolerability and cardiac repolarization in healthy subjects

Abstract

Aims

Custirsen (OGX-011/TV-1011), a second-generation antisense oligonucleotide (ASO) that reduces clusterin production, is under investigation with chemotherapy in patients with solid tumours. Custirsen is associated with constitutional symptoms (CS) that may interfere with clinical pharmacology investigations, such as QT interval studies. Experience with other ASOs suggests NSAID premedication may ameliorate CS, but we observed suboptimal outcomes in healthy subjects given custirsen and NSAIDs. We sought to establish a custirsen regimen for future clinical pharmacology studies in healthy subjects.

Methods

Subjects received custirsen (640 mg intravenously over 120 min) with dexamethasone premedication or increasing doses (320, 480, 640 mg over 6 days) of custirsen with dexamethasone premedication, then one full custirsen dose without premedication on day 8. Incidence/severity of adverse events (AEs) and extensive electrocardiogram readings were evaluated. Pharmacokinetic parameters were estimated.

Results

AEs included CS, elevated transaminases and prolonged activated partial thromboplastin time (aPTT) that were predominantly grade 1/2. Administration of increasing custirsen doses and dexamethasone premedication reduced the incidence of CS associated with full dose custirsen. Transaminase elevation showed a dose-dependent effect (0% at days 2, 4, 27% at day 6) with the highest custirsen doses. Increasing doses of custirsen may have mitigated the severity but not incidence of aPTT prolongation. Neither regimen was associated with cardiac repolarization changes in QT values or concentration–effect analyses. The custirsen pharmacokinetic profile was consistent with previous experience.

Conclusion

Escalation of custirsen dose combined with dexamethasone premedication reduced CS associated with full dose custirsen and should be considered in future clinical pharmacology studies of custirsen.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• Infusions of antisense oligonucleotides (ASOs) may be associated with the development of constitutional symptoms.

• Ameliorating potential ASO-induced adverse events may improve clinical safety and tolerability of this therapeutic class.

• Evaluation of different dosing regimens for reduction of constitutional symptoms may facilitate future clinical development of ASOs.

WHAT THIS STUDY ADDS

• This study evaluated two dosing regimens for reduction of constitutional symptoms following full dose custirsen.

• Increasing doses of custirsen along with dexamethasone premedication reduced the incidence and severity of adverse events associated with the administration of full dose custirsen.

• Future clinical pharmacology studies in healthy subjects may be feasible following increasing doses of ASOs with low dose dexamethasone premedication.

Introduction

Clusterin is a cytoprotective chaperone protein overexpressed in response to stress induced by cytotoxic agents (e.g. chemotherapy or radiation therapy) that confers chemoresistance on cancer cells [1–5]. Custirsen (OGX-011/TV-1011) is a second generation antisense oligonucleotide (ASO), a 21-mer with a 2'-O-methoxyethyl (2'-MOE) conjugate developed to bind clusterin mRNA and reduce clusterin protein expression as a potentially effective strategy for reducing development of treatment resistance in various cancer types [1,6]. The 2'-MOE modification enhances binding to the target mRNA and resists degradation by nucleases, thus prolonging tissue half-life and reducing dose frequency when compared with first generation ASOs [6–8].

Constitutional symptoms (e.g. fever, chills, headaches) have been the main tolerability concern observed with custirsen and other ASOs as a drug class in patients with cancer or other diseases [9–13] and in healthy subjects [14–16]. With custirsen administration, these symptoms appear predominantly during the initial loading dose period when 640 mg is given three times over 5–9 days [17,18]. Other types of adverse events (AEs) include increases in aspartate transaminase (AST) and alanine transaminase (ALT) levels, and dose-dependent, transient prolongation of activated partial thromboplastin time (aPTT) [9–11,13–16].

Inflammatory effects (induced by cytokine release from T helper type 1 [Th1] cells) are thought to underlie these class effect AEs [7,18–22]. Previous studies suggested that constitutional symptoms may be countered by premedication [13,21] and that constitutional symptoms may diminish with repeated administration of ASOs [15,20]. In a previous, unpublished work with custirsen in healthy subjects, ASO-induced AEs occurred despite premedication with non-steroidal anti-inflammatory drugs (NSAIDs), but their incidence and severity appeared to decrease with repeat dosing (unpublished data on file, Teva Pharmaceuticals Industries Ltd., Netanya, Israel). The poor tolerability observed halted additional clinical investigation in healthy subjects and suggested a potential complication for evaluating custirsen cardiac safety [23,24]. Another potential pretreatment option is dexamethasone, an agonist of the glucocorticoid receptor, which has demonstrated dose-dependent inhibition of inflammatory effects related to Th1-mediated cytokine release in a preclinical study [25].

An improved administration protocol is necessary to reduce ASO-induced AEs in patients and in healthy subjects. We conducted this study to compare the safety, tolerability and pharmacokinetic (PK) profile of two custirsen dosing regimens, one with dexamethasone premedication alone and another with increasing doses of custirsen and dexamethasone premedication. In addition, an exploratory electrocardiogram (ECG) analysis was performed to explore custirsen cardiac safety associated with each regimen.

Methods

Study design

This was a single site, open label, two parallel group, one sequence (placebo followed by custirsen) crossover phase 1 study. All enrolled subjects were confined at the research centre during the study treatment period. Subjects were assigned to receive one of two dosing regimens: a premedication only regimen (group A) consisting of one full custirsen dose with dexamethasone premedication or an increasing dose regimen (group B) consisting of three escalating doses of custirsen with concurrent reductions in dexamethasone premedication, followed by a full therapeutic dose of custirsen without premedication (Figure 1). Subjects in the two groups were matched based on age (± 5 years) and weight (± 10 kg). No formal randomization was used. The Independent Ethics Committee provided written approval of the study protocol and consent form. Each study subject was provided with a consent form with details of the proposed study, study procedures and potential risks prior to study initiation. Subjects were instructed that they were free to obtain further information from the investigator or to withdraw from the study at any time.

Figure 1.

Treatment schema. Schedule of treatment for groups A and B. The two groups become comparable only on day 8.

Subjects

Healthy males aged 18–40 years with body mass index ranging from 18–29 kg m^–2 were enrolled. All subjects were in good health as determined at screening and provided written informed consent before participation. Subjects were excluded for the presence of clinically significant ECG abnormalities, resting QT interval corrected for heart rate using the Fridericia formula (QT_cF) ≤320 ms or ≥450 ms, abnormal heart rate or blood pressure or a history of/current risk factors for torsades de pointes (e.g. heart failure, hypokalaemia, or a family history of arrhythmia, sudden death or long QT syndrome).

Treatment

In both treatment groups, custirsen and placebo were administered by intravenous (i.v.) infusion over 2 h in 250 ml normal saline. Dexamethasone was administered as an i.v. bolus injection 1 h prior to custirsen or placebo. Group A received dexamethasone premedication (5 mg) followed by placebo on day 1, and on day 8 received the same dexamethasone premedication regimen followed by an infusion of 640 mg custirsen. Group B received placebo (without premedication) on day 1 and then on days 2, 4, and 6 received increasing doses of custirsen, 320 mg (day 2), 480 mg (day 4) and 640 mg (day 6) with dexamethasone premedication (5 mg on day 2 and 3 mg on days 4 and 6). On day 8, group B received a full custirsen 640 mg dose without dexamethasone premedication.

Assessments

Baseline assessments were collected on day 1 prior to any study medication. Tolerability of a custirsen dose of 640 mg was assessed on day 8 for both regimens, including assessments of safety, PK, pharmacodynamics (PD) and ECG. Subjects were discharged 48 h after the last custirsen dose (day 10) and a follow-up visit occurred 7 days after the last dose (day 15 ± 1).

AEs were recorded at each study visit and the day 15 follow-up visit and were graded according to National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. For subjects with one of the predefined constitutional symptom AEs (fever, chills, fatigue, myalgia, arthralgia, etc.) after the final dose of custirsen (on day 8), the overall impact of the incidence and severity of these symptom(s) on the subject's well-being was graded as mild, moderate or severe at 6, 12, 18, 24 and 36 h after the start of the custirsen infusion.

Laboratory evaluations included complete blood count with differential and coagulation studies (PT/PTT, INR) on day 8 at 0 (pre-dose), 2, 12, 24, 36 and 48 h after the start of the custirsen infusion (group A) and on days 4 and 6 at 0 (pre-dose) and on day 8 at 0 (pre-dose), 2, 12, 24, 36 and 48 h after the start of the custirsen infusion (group B). Serum chemistries were evaluated on day 8 at 0 (pre-dose), 2, 24 and 48 h after the start of custirsen infusion (group A and B). Safety ECG (not centrally read and used solely for the subjects’ safety assessment) and vital signs were taken at screening, on dosing days, at the beginning and end of confinement and at the follow-up visit.

Holter ECG monitoring

To explore the potential effect of custirsen on cardiac intervals, extensive ECG monitoring was conducted using continuous 12-lead Holters. ECGs were centrally read by an external laboratory (eResearch Technology, Inc., Philadelphia, PA, USA). Baseline ECG data were collected in triplicate on day 1 at 0 (pre-dose), 2, 4, 6, 8, 12, 16, 20 and 24 h (after placebo infusion) and on day 8 at 0 (pre-dose), 2, 4, 6, 8, 12, 16, 20, 24, 30, 36 and 48 h (after custirsen infusion). Pre-dose ECGs were collected after dexamethasone administration (in the premedication group) and before custirsen or placebo infusion. Data collected after custirsen infusion were matched with those collected at the same time points on day 1 to estimate the change from baseline and to compare treatment groups.

Pharmacokinetics/pharmacodynamics

Blood samples were drawn to determine custirsen concentrations following ECG data collection on day 8. Samples were collected within 15 min prior to the start of custirsen infusion and at 2, 4, 6, 8, 12, 16, 20, 24, 30, 36 and 48 h after the start of infusion to match Holter ECG data collection, and on day 15 follow-up. PK parameters (area under the plasma concentration vs. time curve from time 0 to the time of the last measurable concentration [AUC(0,t)], area under the plasma concentration vs. time curve from time 0 to infinity [AUC(0,∞)], area under the plasma concentration vs. time curve from time 0 to 48 h [AUC(0,48)], maximum observed plasma concentration [C_max], time to maximum plasma concentration [t_max], apparent first order plasma terminal phase half-life [t_1/2], total plasma clearance [CL], apparent volume of distribution during the terminal phase [V_z]) were calculated using standard non-compartmental approaches.

Statistical methods

Enrolled subjects who received at least one dose of any study drug were included in safety analyses. The ECG analysis population included subjects with at least one baseline ECG and at least one time-matched post-baseline ECG. Descriptive statistics were used to summarize all ECG parameters (for heart rate; PR, QRS and QT intervals) at each time point and change from baseline per treatment group. Each parameter was calculated from three replicate ECGs at each relevant time point. QT_cF values were derived using Fridericia's correction (QT_cF = QT/RR^1/3). Two-sided 90% confidence intervals were calculated for the change from baseline for all ECG parameters. Time-matched ECG analysis was performed by subtracting each baseline time point from the matching time point on day 8 for each subject.

An outlier analysis was also performed using categorical cut points to determine whether subjects demonstrated a cardiac repolarization effect not manifested in the central tendency data. For that purpose, the most extreme value across all post-dose time points was compared with the baseline ECG value (average across baseline values), to define whether any subject fell into the outlier criteria. Outlier criteria included abnormal U waves, heart rate <50 beats min^–1 and ≥25% decrease from baseline mean or >100 beats min^–1 and ≥25% increase from baseline mean, PR interval >200 ms and ≥25% increase over baseline, QRS interval >100 ms and ≥25% decrease over baseline mean, QT interval >500 ms when baseline mean value was ≤500 ms, QT_cF >450 or >480 ms when baseline mean values were ≤450 or ≤500 ms, respectively or a 30–60 ms increase over baseline. Concentration–effect mixed model analysis was performed for baseline-corrected QT_c in all subjects with paired ECG and plasma custirsen concentration data. Safety findings were summarized by treatment group using descriptive statistics. No formal statistical comparisons between treatment groups were performed.

Results

Disposition

Twenty-two subjects were enrolled and assigned to group A (n = 10) or B (n = 12). All subjects in group A completed all planned study treatments and were included in PK and extensive ECG endpoint analyses. Eleven of 12 subjects in group B completed study treatment and extensive ECG endpoint assessments. One subject in group B received study drug up to day 4 (two doses of custirsen), but withdrew consent on day 6 for personal reasons and was excluded from the extensive ECG endpoint analyses. Baseline characteristics for the two groups were similar (Supplemental Table S1).

AEs

All subjects experienced at least one AE during the study. AEs are summarized in Table 1. Six subjects (60%) in group A and three subjects (25%) in group B had at least one grade ≥2 AE. All events considered related to treatment were grade 1 or 2. No AEs were attributable to placebo or dexamethasone alone. None of the AEs led to study discontinuation. No serious AEs or deaths occurred.

Table 1.

Adverse events (AEs) reported throughout the study period

	Group A(Premedication) (n =10)		Group B(increasing dose) (n =12)
AEs, n (%)	Any grade	Grade≥2	Any grade	Grade≥2
Constitutional symptoms
Pyrexia	5 (50)	1 (10)	3 (25)	2 (17)
Myalgia	4 (40)	0	7 (58)	0
Chills	2 (20)	0	4 (33)	0
Dizziness	2 (20)	0	1 (8)	0
Headache	2 (20)	0	4 (33)	1 (8)
Flushing	2 (20)	0	1 (8)	0
Fatigue	1 (10)	0	1 (8)	0
Other AEs
Activated partial thromboplastin time prolonged	9 (90)	6 (60)	11 (92)	1 (8)
Alanine aminotransferase increased	2 (20)	0	3 (25)	1 (8)
Nausea	3 (30)	0	0	0
Folliculitis	3 (30)	0	0	0
Blurred vision	0	0	2 (17)	0
Eye pain	0	0	1 (8)	0
Vomiting	1 (10)	0	0	0
Injection site extravasation	1 (10)	0	0	0
Injection site reaction	1 (10)	0	1 (8)	0
Contusion	1 (10)	0	0	0
Excoriation	1 (10)	0	0	0
Aspartate aminotransferase increased	1 (10)	0	1 (8)	0
Blood creatine phosphokinase increased	1 (10)	1 (10)	0	0
Gamma-glutamyl transferase increased	1 (10)	0	1 (8)	0
Lymphocyte count decreased	1 (10)	1 (10)	0	0
Decreased appetite	1 (10)	0	0	0
Muscle spasms	0	0	1 (8)	0
Paresthesia	1 (10)	0	0	0
Presyncope	1 (10)	1 (10)	0	0
Haematuria	1 (10)	0	0	0
Dermatitis contact	1 (10)	0	0	0
Hyperhidrosis	1 (10)	0	0	0
Hot flush	1 (10)	0	0	0

Following study treatment in group A, the most common AEs (any grade) were aPTT elevation (n = 9, six grade 2 events) and AEs pre-specified as constitutional symptoms including pyrexia (n = 5; one grade 2 event), myalgia (n = 4) and chills, dizziness, headache and flushing (n = 2 each) (Table 1). Folliculitis and nausea occurred in three subjects each and ALT elevation was experienced by two subjects. In group A, two subjects (20%) experienced grade 3 AEs that were considered unrelated to treatment (creatine phosphokinase increase and presyncope, both of which resolved).

The incidence of AEs by day of treatment in group B is presented in Table 2. Grade ≥3 events were not observed. On day 2, when 320 mg custirsen was administered with dexamethasone (5 mg) pretreatment, five subjects (42%) had at least one AE. The most common were myalgia (n = 4) and chills (n = 2). On day 4, when custirsen dose increased from 320 mg to 480 mg while dexamethasone decreased from 5 mg to 3 mg, the incidence and severity of constitutional symptoms (as reflected in chills, fatigue, pyrexia, headache, myalgia, dizziness and flushing) did not, in general, change. Thereafter, the proportion of subjects who reported AEs declined from 50% (n = 6) on day 4 to 9% (n = 1) on day 6, and to none on day 8 when the full dose of 640 mg of custirsen was administered without dexamethasone. In contrast, the proportion of subjects with aPTT prolongation increased from 0 on days 2, 4 and 6 to 100% (n = 11) on day 8, and the proportion of subjects with any elevation of AST or ALT increased from 0 on days 2 and 4 to 27% (n = 3) on day 6, but none was reported on day 8.

Table 2.

Time course of adverse events (AEs) utilizing multiple increasing doses of custirsen (group B only)

	Day 2	Day 4	Day 6	Day 8
n (%)	320 mg(n =12)	480 mg(n =12)	640 mg(n =11)	640 mg(n =11)
Subjects with at least one AE*	5 (42)	6 (50)	4 (36)	11 (100)
Constitutional AEs
Myalgia	4 (33)	4 (33)	0	0
Chills	2 (17)	2 (17)	0	0
Fatigue	1 (8)	0	0	0
Pyrexia	1 (8)	2 (17)	0	0
Dizziness	1 (8)	0	0	0
Headache	1 (8)	3 (25)	1 (9)	0
Flushing	0	1 (8)	0	0
Other AEs
Activated partial thromboplastin time prolonged	0	0	0	11 (100)
Alanine aminotransferase increased	0	0	3 (27)	0
Aspartate aminotransferase increased	0	0	1 (9)	0
Gamma glutamyl transferase increased	0	0	1 (9)	0
Muscle spasms	0	1 (8)	0	0
Eye pain	0	1 (8)	0	0
Blurred vision	1 (8)	0	0	0

^*Events with onset on or after first dose of study drug (on-study events). All events listed according to MedDRA Version 15.0 preferred term.

The frequency and intensity of constitutional symptoms assessed at fixed time intervals after dosing on day 8 are summarized in Table 3. Overall, constitutional symptoms manifested initially at 6 h, peaked during the 12 to 24 h timeframe and began a moderate decline in severity thereafter. After the 6 h post-dose time point, four to six subjects in group A (40%–60%) experienced mild to moderate constitutional symptoms during the remaining 30 h of monitoring. In contrast, one group B subject experienced mild constitutional symptoms throughout the 36 h assessment.

Table 3.

Frequency of constitutional symptoms at fixed time intervals following initiation of custirsen infusion on day 8 by treatment group and investigator-assigned intensity

	Group A(Premedication) (n =10)			Group B(increasing dose) (n =11)
Interval post-dose, n (%)	None	Mild	Moderate	None	Mild	Moderate
0–6 h	9 (90)	1 (10)	0	10 (91)	1 (9)	0
6–12 h	6 (60)	4 (40)	0	10 (91)	1 (9)	0
12–18 h	5 (50)	2 (20)	3 (30)	10 (91)	1 (9)	0
18–24 h	5 (50)	1 (10)	4 (40)	10 (91)	1 (9)	0
24–36 h	4 (40)	3 (30)	3 (30)	10 (91)	1 (9)	0

ECG analyses

No clinically significant ECG abnormalities were observed with the safety 12-lead ECG. Extensive ECG analysis by Holter monitoring showed no indication of a custirsen effect on cardiac repolarization, as reflected in the time-matched mean change from baseline QT_cF for the two treatment groups (Figure 2A). Additionally, concentration–effect analysis in all subjects indicated that custirsen plasma concentration had no significant effect on QT_cF, as represented by the flat slope of the custirsen plasma concentration vs. QT_cF curve (Figure 2B). Of note, group A showed an increased heart rate when compared with group B in the 12–24 h post-dose interval (Figure 2C), and peaked at a mean change of approximately 20 beats min^–1 at 20 h post-dose compared with a mean change of approximately 5 beats min^–1 in group B at the same time point. The time-averaged heart rate change from day 1 to 8 was +8.4 beats min^–1 for group A and +4.8 beats min^–1 for group B. The time-averaged mean change in QT_cF from baseline was −4.5 ms and −4.4 ms for group A and B, respectively. In the outlier analysis, one subject in group A exhibited a >60 ms increase in QT_cF from baseline. Six (60%) tachycardic outliers were observed in group A (none in group B). Bradycardic outliers were not identified in either treatment group. No new abnormal U waves were observed in either group, and no subject exhibited a >60 ms or new >500 ms change in QT_cF from baseline.

Figure 2.

A) QT_cF as determined from extensive ECG cardiac assessment (time-matched) for mean change from baseline (with 90% two-sided confidence intervals) by treatment group. B) Concentration–effect analysis of custirsen plasma concentration vs. QT_cF change from baseline, all subjects. C) Heart rate as determined from extensive ECG cardiac assessment (time-matched) for mean change from baseline (with 90% two-sided confidence intervals) by treatment group. A) and C) treatment Group A; Group B.

PK

The PK profile for custirsen was consistent with previous studies (unpublished data on file, Teva Pharmaceuticals Industries Ltd., Netanya, Israel) [17,18]. The highest mean plasma custirsen concentration on day 8 after custirsen infusion was observed at the end of the infusion (at 2 h), and was followed by a biphasic elimination pattern (Figure 3). Summaries of PK parameters for both groups are presented in Supplemental Table S2.

Figure 3.

Mean custirsen plasma concentrations on day 8 through end of study (group A and group B) on a semi-logarithmic scale. Group A: placebo followed by 640 mg custirsen. Group B: placebo followed by 320 mg, 480 mg, 640 mg and 640 mg custirsen.

Discussion

Clinical pharmacology studies in healthy subjects are particularly relevant where observable effects may be small and a clearer interpretation may be derived from a more homogeneous population. ASOs represent a distinct class of therapeutics with a unique safety profile. However, clinical pharmacology studies of ASOs may be complicated by the occurrence of constitutional symptoms associated with this drug class. Identification of a dosing regimen that minimizes constitutional symptoms and other ASO-induced AEs is necessary to enable future clinical studies. This phase 1 study was conducted to evaluate two dosing regimens of custirsen for safety and tolerability and identify a regimen suitable for future studies in patients and in healthy subjects.

We evaluated two potential approaches to decreasing the incidence of ASO-related AEs on the day when the 640 mg dose was administered, followed by PK and ECG evaluations. The first was to administer a full dose of custirsen with dexamethasone premedication, and the second was to gradually increase the custirsen dose combined with dexamethasone premedication to allow administration of the full custirsen dose without premedication. Dexamethasone has a rapid onset but a short duration of action that makes it suitable for treatment of acute disorders responsive to adrenocortical steroid therapy. Preclinical studies suggested that bolus infusion of dexamethasone 5 mg prior to custirsen may have a rapid effect by reducing cytokine release via inhibition of 60%–100% of glucocorticoid nuclear receptors [25,26]. Dexamethasone premedication at full dose (5 mg) prior to the first custirsen dose was expected to curtail constitutional symptoms, while the reduced dexamethasone dose (3 mg) was expected to maintain this effect when increasing custirsen doses. The premedication regimen in group A was designed to abrogate constitutional symptoms caused by custirsen. Custirsen stimulates cytokine transcription as early as 30 min after infusion and peak constitutional symptoms are expected within 4–6 h of infusion (unpublished data on file, Teva Pharmaceuticals Industries Ltd., Netanya, Israel). The regimen in group B was designed to diminish cytokine release with repeated administration by gradually increasing the custirsen dose along with dexamethasone administration.

This study confirmed previous clinical experience with custirsen in patients and with other ASOs, including the incidence, dose relationship and timing of constitutional symptoms [17]. Of interest, no moderately severe infusion-related AEs were reported in the first 12 h after administration, confirming clinical studies in which AEs were reported after the patient left the clinical site, not during or immediately following custirsen administration. The timing of constitutional symptoms associated with custirsen is similar to the timing of flu-like symptoms that occurred during a phase 3 study of mipomersen, another second-generation ASO with 2′-MOE modifications that targets apo-B-100 mRNA, administered via subcutaneous injection [27].

This study was limited by the small cohort of healthy subjects and the lack of cytokine measurements to correlate the effects of the two dosing regimens. In addition, control arms were not included to evaluate the potential effects of dexamethasone premedication alone with increasing custirsen doses or removal of 5 mg dexamethasone premedication after administration of 640 mg custirsen. We were able to show that constitutional symptoms associated with the administration of 640 mg custirsen were ameliorated following three increasing doses combined with dexamethasone premedication in group B compared with a single dose of 640 mg custirsen with dexamethasone premedication in group A (see Table 3). However, the administration of 5 mg dexamethasone on day 2 at the 320 mg dose and 3 mg dexamethasone on day 4 at the 480 mg dose were not sufficient to prevent constitutional symptoms in all patients (see Table 2). Moreover, despite good tolerability on day 8 (Table 3), group B subjects experienced AEs throughout the study treatment phase (Table 2). The administration of dexamethasone may be scheduled later in the increasing dose schedule in future studies to further decrease constitutional symptoms in healthy subjects. No correlation was observed between the timing of peak custirsen plasma concentration at 2 h and the timing of constitutional symptom development (Table 3).

The incidence of elevated AST/ALT observed with the first 640 mg custirsen administration in group B did not recur with the second 640 mg dose, indicating that the increase in AST/ALT levels may be transient and that tolerance develops with repeated exposure to custirsen. It remains unclear if the mechanism by which second generation ASOs elevate transaminases is a dose- or exposure-dependent inflammatory effect as described in a preclinical study of first generation ASOs [19]. An increase in transaminases was generally absent with first and second generation ASOs [14,15], with few exceptions. However, the longer in-tissue half-lives and liver tissue tropism observed with second generation ASOs may contribute to the manifestation of these effects in healthy subjects.

Virtually all subjects in the present study experienced prolonged aPTT associated with a 640 mg dose of custirsen. However, the longer coagulation time was not associated with bleeding events and generally resolved within 24 h (most within 10 h). The incidence of aPTT prolongation was similar between the two regimens, but the severity of those events was reduced with the multiple increasing doses compared with the single dose regimen. When placed in context with other studies of first and second generation ASOs, transient dose-related aPTT prolongation is a class effect common to all ASOs independent of the hybridization site [15,16,20,28,29]. A possible mechanism proposed for aPTT prolongation is a non-specific interaction of ASO with the intrinsic tenase complex [29].

Importantly, there was no indication from time-matched ECG analysis or concentration–effect analysis that custirsen was associated with any effect on cardiac repolarization. ECG analysis showed uninterrupted, normal cardiac conduction and all measured intervals were within normal values for duration (see Figure 2A). Clusterin had no effect on repolarization as represented by the flat slope for QT_cF in the concentration–effect analysis (see Figure 2B). QT_cF measurements were within normal limits, without evidence for a >60 ms increase over baseline or an absolute value exceeding 500 ms. A transient increase in heart rate was observed in group A at about 20 h post-dose. Although dexamethasone may be contributory, we cannot exclude a potential role for custirsen. It is possible that the heart rate increase is a manifestation of custirsen's pro-inflammatory effects, which may be mitigated by dose escalation in a manner similar to that observed with constitutional symptoms in group B. Additional studies may be required to distinguish the individual roles of dexamethasone and custirsen on heart rate increase.

The PK profile of custirsen was generally consistent with our previous experience in healthy subjects (unpublished data on file, Teva Pharmaceuticals Industries Ltd., Netanya, Israel) and in cancer patients [17,18]. The PK profile is characterized by a prolonged elimination phase, large volume of distribution and small plasma clearance. Systemic exposure was similar in both groups (see Figure 3, Supplemental Table S2).

In conclusion a regimen of multiple increasing doses of custirsen with low dose dexamethasone premedication is effective in diminishing the incidence and severity of events related to inflammatory effects, such as constitutional symptoms and transaminase elevation, on day 8 when full dose custirsen was administered prior to clinical pharmacology assessments. No significant heart rate elevation was observed with this regimen. Thus, it should be considered for future clinical pharmacology studies of custirsen or similar studies of other ASOs in healthy subjects. Custirsen had no apparent effects on cardiac repolarization as measured by the lack of a significant change in QT_cF or by concentration–effect analysis.

Competing Interests

All authors have completed the International Committee of Medical Journal Editors (ICMJE) Unified Competing Interest form and declare LRG, AE, OBG, AP and OS are employees of Teva Pharmaceutical Industries Ltd and received salary and/or stock options for the submitted work, ST is a former employee of Teva Pharmaceutical Industries Ltd and received salary and/or stock options during the conduct of the study. In addition, LRG and AE have a pending patent that is entitled ‘Custirsen treatment with reduced toxicity’, SKW is an employee of PRA Health Services, which was contracted by Teva Pharmaceutical Industries Ltd and received a salary from PRA Health Services during the conduct of the study and LE reported consultant fees from Teva Pharmaceutical Industries Ltd during the conduct of the study.

Financial support for medical editorial assistance was provided by Teva Branded Pharmaceutical Products R&D, Inc. The authors thank Ada Ao-Baslock, PhD, of Powered 4 Significance LLC for medical writing and editorial assistance.

Author Contributions

Sandra K. Willsie, DO, acted as Principal Investigator for this study.

• Study design: LRG, LE, SKW, OBG, AP

• Data collection: AE, SKW, OBG

• Data analysis: AE, LRG, LE, SKW, ST, OBG, AP, OS

• Review/revised manuscript: AE, LRG, LE, SKW, ST, OBG, AP, OS

• Other (please specify): study clinical leader (AE), principal investigator (SKW)

Supporting Information

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

Table S1 Demographics by treatment group

Table S2 Effects of custirsen dosing regimen on pharmacokinetic parameters

Supporting Information