Selepressin, a novel selective vasopressin V_1A agonist, is an effective substitute for norepinephrine in a phase IIa randomized, placebo-controlled trial in septic shock patients

This was a multicenter, double-blind, randomized, placebo-controlled phase IIa trial investigating three ascending doses of selepressin in patients with early septic shock. Patients were recruited into the trial between 2009 and 2011 in Belgium, Denmark, and the United States in accordance with the Declaration of Helsinki and the principles of good clinical practice. The study protocol and informed consent documents were approved by the independent ethics committees or research ethics boards of all participating institutions, and written informed consent was obtained from all patients, their next of kin, or another surrogate decision maker as appropriate prior to enrollment. The study was approved by the competent regulatory authorities of each country participating in the trial. An independent data safety and monitoring committee evaluated the safety of the dose regimens prior to escalating to the next dose level.

At each dose level, patients were randomized to constant intravenous infusion of selepressin or placebo in a ratio of 2:1. Open-label norepinephrine was concomitantly administered to maintain the treatment target MAP of ≥65 mmHg. Study drug infusion continued as long as arterial pressure support was deemed necessary, but no longer than 7 days. Patients needing vasopressor support after 7 days were switched from study drug infusion to norepinephrine or another vasopressor. Assessments were performed during study drug treatment and up to 4 weeks after study drug initiation.

Patients with hypotension in early septic shock, defined as hypotension not responding to infusion of fluid and requiring at least 0.1 μg/kg/minute norepinephrine for at least 2 h, with a proven or suspected site of infection and at least one sign of tissue hypoperfusion (oliguria, decreased Glasgow Coma Scale score, decreased ratio of partial pressure arterial oxygen to fraction of inspired oxygen, or increased arterial blood lactate) could be included. To be eligible, patients had to be shifted to the open-label norepinephrine and randomized to a constant intravenous infusion of selepressin or placebo within 24 h of meeting the inclusion criteria. Briefly, exclusion criteria (see Additional file 1: Table S1 for details) were acute coronary syndrome; hypovolemia suspected on clinical grounds; cardiac failure; pregnancy or breastfeeding; hypotension other than septic shock; use of vasopressin or terlipressin during the current hospital admission; acute mesenteric ischemia; episode of septic shock within 1 month; death anticipated within 24 h; chronic heart disease, including heart failure and second- and third-degree atrioventricular block without pacemaker; hyponatremia; brain injury; burn; peripheral vascular disease; previously randomized in this trial; intake of an investigational drug within the last 3 months; participation in another clinical trial; and considered unsuitable to participate in the trial for any other reason.

The study design comprised four treatment cohorts at three ascending infusion rate levels, the first three cohorts with ten receiving active treatment and five receiving placebo. The last cohort comprised active and placebo to finally reach 20 patients at the maximum tolerated infusion level and 20 patients receiving placebo. The randomization process was a computer-generated random listing of the treatment allocations, stratified by center and in variable permuted blocks of 2, 4, or 6.

The investigated starting and maximal infusion rates of selepressin were 1.25 ng/kg/minute, 2.5 ng/kg/minute, and 3.75 ng/kg/minute, with the patient’s body weight being measured or estimated. Selepressin or placebo was infused via a central venous catheter at the constant initial rate in addition to open-label norepinephrine targeting a mean arterial pressure (MAP) of ≥65 mmHg. When patients were hemodynamically stable, open-label norepinephrine was tapered while maintaining target MAP with study drug. When the MAP was stable for 4 h without norepinephrine, study drug was weaned stepwise according to the protocol (Additional file 1: Table S2). If weaning resulted in hemodynamic instability, the study drug was reinstituted, and, if needed, open-label norepinephrine was restarted. Open-label norepinephrine was adjusted to maintain the MAP ≥65 mmHg if blinded study drug was insufficient. Patients received study drug until shock resolution (i.e., no vasopressor support) or a maximum of 7 days unless discontinued for safety reasons (Additional file 1: Table S3).

Open-label norepinephrine was supplied in the United States and Canada as norepinephrine bitartrate 1 mg/ml (Levophed, norepinephrine base; Hospira, Lake Forest, IL, USA) diluted in 5% dextrose and in Europe as Norepinephrine 1:1000 (norepinephrine tartrate; (Cardinal Health Ltd., Basingstoke, UK) 2 mg/ml (1 mg/ml norepinephrine base) diluted in 5% glucose. A tight protocol and accurate pumps (Braun Perfusor® Space; B. Braun Melsungen AG, Melsungen, Germany) were used to calculate open-label norepinephrine delivered.

The co-primary endpoints were stabilization of MAP as determined by proportion of patients maintaining a MAP >60 mmHg without open-label norepinephrine at 12, 24, 48, and 96 h, and day 7; infusion rates and cumulative dose of open-label norepinephrine; and proportion of patients maintaining a MAP >60 mmHg at 12, 24, 48, and 96 h, and day 7, regardless of open-label norepinephrine administration. The limit for maintained MAP was defined as 60 mmHg to prevent small variations around the clinical treatment target of 65 mmHg having disproportional impact on the overall evaluation.

Secondary endpoints included pharmacodynamics, pharmacokinetics, safety (vital signs, central venous pressure, central venous oxygen saturation, electrocardiographic and cardiac function, respiratory function, clinical chemistry, hematology, hemostasis, and urinalysis), organ dysfunction, an indirect measure of vascular leakage (i.e., fluid balance), and morbidity. Treatment-emergent adverse events (occurrence from start of study drug to 48 h after infusion was stopped) were collated and evaluated. Morbidity and organ dysfunction were evaluated by time course of Sequential Organ Failure Assessment (SOFA) scores; days alive and free of organ dysfunction (using SOFA); proportion of patients off all vasopressors; days alive and out of intensive care unit (ICU); days alive and free of vasopressors; corticosteroids for sepsis treatment, dialysis, or mechanical ventilation at days 7, 14, and 28; ICU and hospital (up to day 28) lengths of stay; and plasma C-reactive protein (CRP), tumor necrosis factor (TNF)-α, interleukin (IL)-6, IL-10, and IL-1ra levels.

Because the study drug infusion was administered according to need, the pharmacokinetic parameters steady-state concentration, total systemic clearance, distribution volume at steady state, and half-life were calculated by modeling using a two-compartment population pharmacokinetic model with random subject effects on clearance and distribution volume using WinNonlin®Pro (Pharsight Corp., Cary, NC, USA). Actual blood sampling time points were used for the calculations.

All statistical analyses were performed using SAS version 9.2 for Windows software (SAS Institute Inc., Cary, NC, USA). The first two co-primary endpoints were compared using a logistic regression model. The rates and cumulative amounts of open-label norepinephrine administered were compared between treatment groups by using a repeated measures analysis of covariance (ANCOVA) model with treatment, time, and treatment by time interaction as factors; baseline rate of norepinephrine as a covariate; and subject as the experimental unit. The analyses were done for both the full analysis set and the per-protocol analysis set. For both analysis sets, the analyses were presented for the whole analysis set and those alive (and not discontinued) at the time of the measurement. The selepressin groups were compared individually with the placebo group in an analysis of variance model.

Changes in vascular leakage endpoints (i.e., fluid balance), as well as changes from baseline in cytokines and SOFA scores, were compared between treatment groups using the same ANCOVA model as for norepinephrine. Patients were categorized as free of organ dysfunction if all six individual SOFA scores were 0. Percentage days alive and free of organ dysfunction/ICU/hospital (i.e., number of days/7 observation days × 100) were analyzed between treatment groups using nonparametric Wilcoxon tests. The treatment differences between the selepressin groups and placebo were estimated using the Hodges-Lehmann estimator for independent samples. The proportions of patients alive were analyzed between treatment groups on days 7, 14, and 28 using a logistic regression model. Confidence intervals (Clopper-Pearson) were calculated for mortality rates and for the ORs of mortality rates.

Fifty-three patients were randomized, and 52 subjects were dosed; 10 subjects received 1.25 ng/kg/minute, 19 subjects received 2.5 ng/kg/minute, 2 subjects received 3.75 ng/kg/minute, and 21 subjects received placebo. All randomized patients were included in the intention-to-treat dataset. Two patients were infused at 3.75 ng/kg/minute, one of whom had the study drug infusion discontinued for possible safety reasons, with subsequent discontinuation of this dose group. Owing to the small sample size and short duration of infusion in this group, efficacy analyses were not possible. A Consolidated Standards of Reporting Trials diagram of the study is shown in Fig. 1.

There were essentially no differences between selepressin- and placebo-treated patients in baseline characteristics, apart from a lower baseline norepinephrine dose in the 1.25 ng/kg/minute than in the 2.5 ng/kg/minute and placebo groups (Table 1). The most common underlying organ dysfunctions were gastrointestinal, metabolic and nutritional, respiratory, and renal (Table 1). The primary infection was predominantly abdominal (40%) or pulmonary (31%).

MAPs during the 7-day assessment period were similar between groups at approximately 70 mmHg initially and increasing over 2 days to approximately 80–85 mmHg (Additional file 1: Figure S1).

The mean total selepressin infused doses were 6.1 and 8.7 μg/kg infused over 3.6 and 3.2 days in the 1.25 and 2.5 ng/kg/minute dose groups, respectively. The swift onset of action of selepressin was illustrated by the high proportion of patients receiving 2.5 ng/kg/minute selepressin who early on maintained MAP >60 mmHg without norepinephrine (about 50% at 12 h and 70% at 24 h) (Fig. 2). In contrast, in the placebo and 1.25 ng/kg/minute selepressin groups, no patient was free of norepinephrine at 12 h and ≤20% were free of norepinephrine at 24 h (p < 0.01). However, over time, the differences decreased as more patients recovered also in the latter groups.

The 7-day baseline adjusted mean cumulative doses of open-label norepinephrine were 761, 659, and 249 μg/kg for the placebo, 1.25 ng/kg/minute, and 2.5 ng/kg/minute groups, respectively (p < 0.001 for 2.5 ng/kg/minute vs. placebo groups) (Fig. 3a). Furthermore, the norepinephrine mean infusion rate was initially reduced more rapidly in the selepressin 2.5 ng/kg/minute group than in the placebo group; at 24 h, the infusion rates were 0.04 vs. 0.18 μg/kg/minute (p < 0.005) (Fig. 3b).

As expected, there were no differences between treatment groups in the proportions of patients who maintained MAP >60 mmHg regardless of norepinephrine, meaning that the patients were generally treated equally (Additional file 1: Table S4). Selepressin at 2.5 ng/kg/minute appeared to result in a faster recovery from shock (off vasopressors, including selepressin) than 1.25 ng/kg/minute and placebo, but without statistical significance (58% vs. 29%, p = 0.11, at 48 h) (Additional file 1: Figures S2 and S3). There was no significant difference between groups in 28-day mortality rates (placebo 4 [21%] of 19, selepressin 1.25 ng/kg/minute 5 [50%] of 10, selepressin 2.5 ng/kg/minute 1 [5%] of 19).

The proportion of days alive and free of ventilation was greater in the selepressin 2.5 ng/kg/minute group than in the placebo group (54% vs. 23%, p < 0.02) over the 7-day treatment period. There was no difference between the 1.25 ng/kg/minute (31%) and placebo groups (Additional file 1: Figure S4).

Selepressin at 2.5 ng/kg/minute decreased the cumulative net fluid balance over the treatment period compared with the 1.25 ng/kg/minute and placebo groups (from about 9 L to 6.5 L, p = 0.1) (Fig. 4), and compared with placebo, the difference was significant (p < 0.05) from day 5 (94 h) onward. The differences in fluid balance appeared to be due to fluid input rather than to urine output because there were no differences between groups in urine output (Additional file 1: Figure S5).

There were no significant differences between groups in length of stay in the ICU or hospital up to 28 days; in plasma CRP, TNF-α, IL-6, IL-10, or IL-1ra levels; or in any other secondary morbidity endpoint.

The mean steady-state concentrations of selepressin (0.50 and 0.99 ng/ml) were proportional to the initial infusion rates of 1.25 and 2.5 μg/kg/minute, with a time to steady-state concentration of approximately 7 h. The modeled mean total systemic clearance values were 10.0 and 13.1 L/h, respectively, increasing with body weight and being higher in men than in women. The terminal half-life was approximately 2.5 h in both dose groups, with an initial distribution/elimination phase half-life of approximately 10 minutes. The distribution volume at steady state was 18–31 L, indicating extravascular distribution (Additional file 1: Table S5).

Selepressin was well tolerated, with no difference between selepressin dose groups and placebo in terms of treatment-emergent adverse events. The high-dose group was stopped after two patients because of potential adverse events in the second patient. The most frequent adverse events were atrial fibrillation, bradycardia, and hypertension (six subjects on seven occasions), equally distributed across treatment groups. The severe treatment-emergent adverse events were generally single observations attributable to the underlying disease (Additional file 1: Table S6). Nine treatment-emergent adverse events (in eight subjects) were regarded by the investigator to be related to treatment; one, four, and four treatment-emergent adverse events were regarded as mild, moderate, and severe, respectively (Table 2). Four of the adverse drug reactions in three patients were judged as serious: myocarditis and peripheral ischemia (one patient in the 2.5 ng/kg/minute group), myocardial ischemia (3.75 ng/kg/minute group), and atrial fibrillation (placebo group). There were no deaths related to ischemic event(s) attributable to selepressin.

There were no regional or global signs of hypoperfusion, as suggested by similar decreases of lactate levels (Additional file 1: Figure S5) and serum creatinine (Additional file 1: Figure S6), with no significant differences between study groups during the treatment period.

Strengths of this trial include the randomized, concealed, placebo-controlled design; generalizability (multicenter in Europe and North America); precision of the co-primary endpoints; the inclusion of patients with relatively early septic shock; the novel selective V_1A agonist selepressin; and administration of a tight protocol in critically ill septic shock patients. Limitations of this trial include the small sample size due to the early stage in selepressin’s clinical development in a limited number of centers (that may limit the overall impact), but they were suitable for initial assessment of selepressin in human septic shock. The fluid input was not controlled, and we do not have information about the different solutions used (crystalloids, colloids, blood products), which might, at least in part, have influenced the amount of infused fluids, a potential limitation in interpreting the fluid balance. Although the addition of selepressin allowed sparing of norepinephrine and some would argue that one is simply substituting one vasopressor (selepressin) for another (norepinephrine), the beneficial effects on ventilation and fluid balance suggest additional nonvasopressor benefits of selepressin vs. norepinephrine. The pharmacodynamic effects of selepressin should be interpreted with caution owing to the small sample size of this phase IIa trial. We did not measure cardiac output, so we cannot comment on selepressin vs. norepinephrine effects on this parameter. Of note, vasopressin and norepinephrine had similar effects—no decrease in cardiac output in VASST [28]. Interaction of vasopressin and corticosteroids has been reported elsewhere, but owing to the small sample size, it was not possible to assess the potential interaction of corticosteroids and selepressin in this trial.