The clinical effects of levosimendan are not attenuated by sulfonylureas

Figures & data

Table I. Studies included in the analyses.

Study	N (total/LS)	LS infusion rate*	Comparator	NYHA class	Primary end-point
Invasive hemodynamic studies
Dose-finding (12)	151/95	0.1–0.6 μg/kg/min	placebo/dobutamine	III	hemodynamics
Dose escalation & withdrawal (14)	146/98	0.1–0.4 μg/kg/min	placebo	III-IV	hemodynamics
LIDO (15)	203/103	0.1–0.2 μg/kg/min	dobutamine	(III)-IV	hemodynamics
Outcome studies
RUSSLAN (16)	504/402	0.1–0.4 μg/kg/min	placebo	IV	Safety
SURVIVE(17)	1327/664	0.1–0.2 μg/kg/min	dobutamine	IV	mortality
REVIVE I & II (18)	700/350	0.1–0.2 μg/kg/min	placebo	IV	clinical composite

LS, levosimendan.

*In all studies, levosimendan was administered starting with a 10-minute bolus followed by an infusion which duration was generally 24 hours, except 42 of the patients in Dose escalation & withdrawal study received 48-hour infusion and in the RUSSLAN study the infusion duration was 6 hours.

Table II. Change in hemodynamic variables [mean (SD)] after a 24-hour infusion of levosimendan by the use of sulfonylurea treatment in studies with invasive hemodynamic assessment (12–15).

Variable	Time	SU+ n = 42	SU– n = 227–233*	p-value
PCWP (mmHg)	Baseline	21.8 (8.7)	22.3 (8.7)
	Change at 24 hour	−6.3 (6.7)	−7.3 (6.7)	0.401
Cardiac output (l/min)	Baseline	4.1 (0.9)	4.0 (1.1)
	Change at 24 hour	+1.0 (0.8)	+1.3 (1.1)	0.108
Systolic BP (mmHg)	Baseline	121 (18)	115 (19)
	Change at 24 hour	−6.3 (15.6)	−4.2 (12.7)	0.761
Heart rate (bpm)	Baseline	81 (14)	78 (15)
	Change at 24 hour	+9.4 (10.5)	+7.0 (11.1)	0.118

*The number of patients varies depending on the availability of data both at baseline and at 24 hour.

Table III. Baseline characteristics in all studies (12–18) combined (in sulfonylurea users and non-users and in levosimendan- and comparator-treated patients).

	All patients			SU+			SU−
	SU+ (N = 394)	SU− (N = 2610)	p	LS (N = 226)	Comparator (N = 168)	p	LS (N = 1474)	Comparator (N = 1136)	p
Age mean (SD), years	67 (11)	64 (13)	<0.001	67 (10)	66 (11)	NS	64 (13)	64 (13)	NS
BMI mean (SD, kg/m2)	29 (6)	27 (6)	<0.001	29 (6)	30 (6)	NS	28 (6)	27 (6)	NS
Male, %	65%	72%	0.010	64%	67%	NS	71%	73%	NS
Diabetes mellitus, %	97%	23%	<.001	99%	96%	NS	21%	26%	0.005
Coronary artery disease, %	55%	52%	NS	53%	58%	NS	51%	54%	NS
Hypertension, %	72%	57%	< .001	71%	73%	NS	57%	57%	NS
ACEi/ARB use, %	75%	69%	0.008	76%	74%	NS	66%	71%	0.004
B-blocker use, %	52%	48%	NS	55%	50%	NS	46%	51%	0.005
Aldosterone antagonist use, %	60%	48%	<0.001	53%	68%	0.003	42%	56%	<0.001
Systolic BP mean (SD), mmHg	123 (19)	117 (19)	<.001	124 (20)	121 (18)	NS	117 (19)	116 (19)	NS
Diastolic BP mean (SD), mmHg	71 (12)	70 (12)	NS	71 (12)	71 (13)	NS	71 (12)	70 (13)	NS
Heart rate mean (SD), bpm	80 (15)	82 (16)	0.038	81 (15)	80 (14)	NS	82 (16)	82 (17)	NS
LV ejection fraction mean(SD), %	23.7 (6.4)	23.1 (6.3)	NS	23.7 (6.5)	23.7 (6.3)	NS	23.0 (6.2)	23.3 (6.3)	NS
BNP median(IQ), pg/ml	776 (377–1545)	995 (474–1863)	0.002	797 (334–1456)	759 (421–1610)	NS	976 (473–1858)	1029 (478–1870)	NS
P-crea mean (SD), μmol/l	121 (54)	119 (50)	NS	119 (42)	126 (68)	NS	118 (48)	120 (53)	NS
P-potassium mean (SD), mmol/l	4.4 (0.5)	4.3 (0.6)	NS	4.4 (0.6)	4.3 (0.5)	NS	4.3 (0.6)	4.3 (0.6)	NS
P-sodium mean (SD), mmol/l	138 (4)	137 (5)	NS	138 (4)	138 (4)	NS	138 (5)	138 (5)	NS
B-gluc mean (SD), mmol/l	9.8 (4.4)	7.3 (3.3)	<0.001	10.0 (4.6)	9.5 (4.2)	NS	7.2 (3.2)	7.4 (3.3)	NS
WBC mean (SD), E9/l	8.4 (3.0)	8.3 (3.6)	NS	8.4 (2.8)	8.3 (3.2)	NS	8.4 (3.8)	8.2 (3.3)	NS
B-Hb mean (SD), g /dl	13.2 (2.0)	13.3 (1.9)	NS	13.2 (2.0)	13.1 (2.0)	NS	13.1 (2.0)	13.2 (1.9)	NS

LS, levosimendan-treated.

Data on left ventricular ejection fraction and on other laboratory parameters except creatinine were not available in the RUSSLAN study. BNP data were available only in REVIVE I & II and SURVIVE studies.

Table IV. Mortality during 31-days post-infusion in individual studies and all studies combined.

	All patients			SU+			SU−			Interaction SU/levo
	SU+	SU−	Hazard ratio (95% CI)	LS	Comparator	Hazard ratio (95% CI)	LS	Comparator	Hazard ratio (95% CI)	P
Dose-finding (12)	6.9% (2/29)	0% (0/121)	NA	5.0% (1/20)	11.1% (1/9)	0.459 (0.029, 7355) p = 0.582	0% (0/75)	0% (0/46)	NA	NA
Dose escalation & withdrawal (14)	0% (0/18)	4.7% (6/128)	NA	0% (0/11)	0% (0/7)	NA	3.8% (3/87)	7.3% (3/41)	0.459 (0.093,2.275) p=0.341	NA
LIDO (15)	12.5% (3/24)	12.0% (21/175)	1.058 (0.316,3.548) p=0.927	8.3% (1/12)	16.7% (2/12)	0.457 (0.041,5.044) p=0.523	7.8% (7/90)	16.5% (14/85)	0.456 (0.184,1.129) p=0.090	0.978
RUSSLAN (16)	11.6% (5/43)	16.3% (75/460)	0.696 (0.281,1.720) p=0.432	12.8% (5/39)	0% (0/4)	NA	14.9% (54/362)	21.4% (21/98)	0.650 (0.393,1.076) p=0.094	NA
SURVIVE (17)	6.7% (11/163)	13.7% (158/1157)	0.478 (0.259,0.881) p=0.018	4.9% (4/82)	8.6% (7/81)	0.540 (0.158,1.847) p=0.326	12.9% (75/578)	14.3% (83/579)	0.894 (0.654,1.222) p=0.482	0.435
REVIVE I & II (18)	4.3% (5/117)	5.3% (30/569)	0.793 (0.208,2.044) p=0.631	4.8% (3/62)	3.6% (2/55)	1.356 (0.227,8,116) p=0.739	5.7% (16/282)	4.9% (14/287)	1.195 (0.583,2.448) p=0.6265	0.604
COMBINED	6.6% (26/394)	11.1% (290/2610)	0.585* (0.392,0.847) p=0.009	6.2% (14/226)	7.1% (12/168)	0.870 (0.402, 1.881) p=0.723	7.8% (155/1474)	11.9% (135/1136)	0.897 (0.713,1.130) p=0.357	0.943

LS, levosimendan; NA, Unable to estimate due to limited number of events.

*Adjusted analysis: Hazard ratio 0.710 95% CI (0.473, 1.067), p=0.100. Multivariate Cox proportional hazard model included effect for SU use and significant baseline factors observed in univariate analysis (Table III) using forward selection method. In addition to SU use the final model included effects for age, BMI, HR, SBP ACEi/ARB use and aldosterone antagonist use.

Propensity score adjusted Cox-regression model to evaluate effect of SU use for mortality gave similar results (HR 0.660 95% CI (0.441, 0.987), p = 0.043) compared to unadjusted results. Also, matched propensity score adjusted model gave similar results but not significant (HR 0.642 95% CI (0.393, 1.042), p = 0.077) compared to unadjusted results.

Table V. Adverse events of myocardial ischemia* during 31-days post-infusion in SURVIVE and REVIVE I & II.

	All patients			SU+			SU−			Interaction SU/levo
	SU+	SU−	p	LS	Comparator	p	LS	Comparator	p	p**
SURVIVE (17)	11.0% (18/163)	11.5% (133/1157)	>0.999	9.8% (8/82)	12.4% (10/81)	0.627	10.4% (60/578)	12.6% (73/579)	0.269	0.933
REVIVE I & II (18)	11.1% (13/117)	12.7% (72/569)	0.769	9.7% (6/62)	12.7% (7/55)	0.770	11.7% (33/282)	13.6% (39/287)	0.530	0.831
COMBINED	11.1% (31/280)	11.9% (205/1726)	0.765#	9.2% (14/144)	12.5% (17/136)	0.568	10.8% (93/860)	12.9% (112/866)	0.181	0.846

LS, levosimendan.

*Myocardial ischemia terms: acute coronary syndrome, acute myocardial infarction, angina pectoris, unstable angina, cardiac enzymes increased, chest discomfort, chest pain, coronary artery disease, electrocardiogram ST segment elevation, electrocardiogram ST-T change, electrocardiogram T wave inversion, myocardial infarction, myocardial ischemia, postinfarction angina.

**Breslow-Day Test for Homogeneity of the Odds Ratios.

#Adjusted analysis: p = 0.222 (Logistic regression model with SU use and forward selection for significant baseline variables observed in univariate analysis) In addition to SU use the final model included effects for age, gender, SBP, hypertension, and aldosterone antagonist use.

Figure 1. Change in systolic blood pressure (mmHg) in the combined data from SURVIVE and REVIVE I & II studies divided by the sulfonylurea use in levosimendan- and control-treated subjects (*p<0.05).

Figure 2. Percent change in BNP values in the combined data from SURVIVE and REVIVE I & II studies divided by the sulfonylurea use in levosimendan and control-treated subjects.