7/25/2019 Xian 2016

1/7

A randomized, double-blind, multicenter, placebo-controlled clinical

study on the efcacy and safety of Shenmai injection in patients with

chronic heart failure

Shaoxiang Xian a,n,1, Zhongqi Yang a,1, Jun Lee b, Zhiping Jiang c, Xiaohan Ye d, Luyi Luo e,Lili Jin f, Tianlun Yang g, Suilin Ye h, Dongfeng Lu i

a The First Afliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, Chinab The First Afliated Hospital, College of Medicine, Zhejiang University, Hangzhou, Chinac Yuebei People's Hospital, Shaoguan, Chinad Dongguan Hospital of Traditional Chinese Medicine, Dongguan, Chinae Shenzhen Hospital of Traditional Chinese Medicine, Shenzhen, ChinafGuangdong second Traditional Chinese Medicine Hospital, Guangzhou, ChinagXiangya Hospital, Central South University, Changsha, Chinah Guangzhou Hospital of Traditional Chinese Medicine, Guangzhou, Chinai The First Afliated Hospital of Guangzhou Medical University, Guangzhou, China

a r t i c l e i n f o

Article history:

Received 19 November 2015

Received in revised form

4 February 2016

Accepted 31 March 2016Available online 1 April 2016

Keywords:

Heart failureShenmai injection

Randomized controlled trial

Traditional Chinese medicines

Panax ginseng

Ophiopogon japonicas

a b s t r a c t

Ethnopharmacological relevance: Shenmai injection (SMI) is a traditional Chinese herbal medicine ex-

tracted from Panax ginseng (Panax ginseng C.A. Mey, steamed and dry) and Ophiopogon japonicus

(Ophiopogon japonicus (L.f.) Ker-Gawl, root). It has been widely used for the treatment of chronic heart

failure (CHF) in China. However, the evidence supporting its effects remains unclear due to lack of high

quality trials. The aim of this study was to investigate the efcacy and safety of SMI in CHF patients with

coronary artery disease (CAD).

Materials and methods: This double-blind, multicenter study randomized 240 eligible patients equally to

receive SMI or placebo (100 ml/day) in addition to standard medicines for the treatment of CHF. Theprimary endpoint was the New York Heart Association (NYHA) functional classication. The secondary

endpoints were 6-min walking distance (6MWD), short-form 36 (SF-36) hearth survey score, traditional

Chinese medicines (TCM) syndrome score, left ventricular ejection fractions (LVEF) and B-type natriuretic

peptide (BNP) level.

Results: During treatment of 1 week, the NYHA functional classication was gradually improved in both

groups, but the SMI group demonstrated a signicantly greater improvement compared with the placebo

group (p0.001). Moreover, the improvement in patients received SMI was superior to those in control

group with respect to 6MWD, SF-36 score and TCM syndrome score. Treatment with SMI within 1 week

was well tolerated with no apparent safety concerns.

Conclusions: The integrative treatment with standard medicines plus SMI can further improve NYHA

functional classication for patients with CHF and CAD. Therefore, SMI could be recommended in the

combination therapy for CHF accompanied with CAD.

& 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Chronic heart failure (CHF) is a major public health problem

and has been singled out as an epidemic ( Roger et al., 2004), as-

sociated with signicant mortality and heavy healthcare ex-

penditures (Go et al., 2014; Roger, 2013). Evidence from epide-

miology conrmed that coronary artery disease (CAD) have the

largest population attributable risk for HF (Owan et al., 2006;

Roger, 2013) and is correlated with poor prognosis for patients

with CHF (Murdoch et al., 1998). Treatment strategies have been

Contents lists available atScienceDirect

journal homepage: www.elsevier.com/locate/jep

Journal of Ethnopharmacology

http://dx.doi.org/10.1016/j.jep.2016.03.066

0378-8741/& 2016 Elsevier Ireland Ltd. All rights reserved.

n Correspondence to: The First Afliated Hospital of Guangzhou University of

Chinese Medicine, No. 16 Airport Road, Guangzhou 510000, China.

E-mail addresses: zhongfyy@126.com(S. Xian), zqyang66@163.com(Z. Yang),

junlee008@126.com(J. Lee), jzp7448@163.com(Z. Jiang),

xiaohanye666@126.com (X. Ye), luoluyi8@126.com(L. Luo),

wondergirl2@yeah.net (L. Jin), yangtl44@yeah.net(T. Yang),

suilinye52@163.com (S. Ye), ludf123@126.com(D. Lu).1 Dr. Shaoxiang Xian and Dr. Zhongqi Yang are the co-rst authors and con-

tributed equally to this work.

Journal of Ethnopharmacology 186 (2016) 136142

http://www.sciencedirect.com/science/journal/03788741http://www.elsevier.com/locate/jephttp://dx.doi.org/10.1016/j.jep.2016.03.066mailto:zhongfyy@126.commailto:zqyang66@163.commailto:junlee008@126.commailto:jzp7448@163.commailto:xiaohanye666@126.commailto:luoluyi8@126.commailto:wondergirl2@yeah.netmailto:yangtl44@yeah.netmailto:suilinye52@163.commailto:ludf123@126.comhttp://dx.doi.org/10.1016/j.jep.2016.03.066http://dx.doi.org/10.1016/j.jep.2016.03.066http://dx.doi.org/10.1016/j.jep.2016.03.066http://dx.doi.org/10.1016/j.jep.2016.03.066mailto:ludf123@126.commailto:suilinye52@163.commailto:yangtl44@yeah.netmailto:wondergirl2@yeah.netmailto:luoluyi8@126.commailto:xiaohanye666@126.commailto:jzp7448@163.commailto:junlee008@126.commailto:zqyang66@163.commailto:zhongfyy@126.comhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.03.066&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.03.066&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.03.066&domain=pdfhttp://dx.doi.org/10.1016/j.jep.2016.03.066http://dx.doi.org/10.1016/j.jep.2016.03.066http://dx.doi.org/10.1016/j.jep.2016.03.066http://www.elsevier.com/locate/jephttp://www.sciencedirect.com/science/journal/03788741

7/25/2019 Xian 2016

2/7

developed based upon the understanding of the pathophysiolo-

gical mechanisms in HF (Kemp and Conte, 2012). According to the

latest guidelines for HF (McMurray et al., 2012;Yancy et al., 2013),

standard medicines include diuretics, angiotensin converting en-

zyme inhibitors, angiotensin receptor blockers, beta blockers, al-

dosterone antagonist and digitalis. Despite advances in the drug

and surgical treatment, the survival estimates remain 50% and 10%

at 5 and 10 years, and rates of readmissions continue to rise (Ro-

ger, 2013). Therefore, improved treatments are imperative.In the theory of traditional Chinese medicine (TCM), Qi usually

refers to life energy that manifests simultaneously on the physical

and mental-spiritual level (Maciocia, 2005). Yin and Yang are

terms used to describe relative opposite qualities or manifesta-

tions of Qi. From the perspective of TCM, the principal problems of

CHF are deciency of Qi and inadequacy of either Yin or Yang,

which impair the circulation of blood (Li et al., 2013b;Mao et al.,

2009; Tang and Huang, 2013), leading to retention of body uid

and stasis of blood. Shenmai injection (SMI) is a traditional Chi-

nese herbal medicine extracted from Panax ginseng(Panax ginseng

C. A. Mey, steamed and dry) and Ophiopogon japonicus (Ophiopo-

gon japonicus (L.f.) Ker-Gawl, root), with effects of tonifying Qi,

nourishing Yin and replenishing bodily uids (Chen et al., 2012;

Ma et al., 2010;Yu et al., 2014). Therefore, SMI has long been used

as a complement to the standardized treatments for CHF in China.

Although its efcacy against heart failure has been demonstrated

in several studies, the poor quality of these trials compromised the

reliability of the evidence (Chen et al., 2012). This clinical trial was

aimed to further evaluate whether CHF patients with CAD can

benet from the integrative treatments with standard medicines

plus SMI.

2. Material and methods

2.1. Study design

This study was designed as a multicenter, randomized, double-

blind, placebo-controlled study based on standard therapy and

parallel groups. The trial was conducted in eight A-level hospitals

and enrolled 240 patients with CHF and CAD based on the inclu-

sion and exclusion criteria. The study conforms to the Declaration

of Helsinki and was performed in accordance with the Good Clin-

ical Practice and national regulations. The trial protocol and its

informed consent form were approved by appropriate in-

dependent ethics committees. All participants submitted informed

consent before the start of the study. The clinical trial has been

registered at Chinese Clinical Trial Registry with the registration

number: ChiCTR-TRC-12003063.

Block randomization with stratication was used to assign

eligible patients to either SMI group or placebo group in the ratio

of 1:1. The corresponding randomization code was generated by

SAS software (SAS Institute, Cary, North Carolina) according to thenumber of enrolled patients in each center and was held centrally

by the rst afliated hospital of Guangzhou University of Chinese

Medicine. All patients and investigators were kept unaware of the

treatment allocation.

Prior to this trial, patients had received standard medicines for

at least 1 month with reference to the Chinese Society of Cardi-

ology guidelines for the treatment and diagnosis of CHF. Besides

standard medicines, eligible patients were randomized to receive

100 ml/day SMI or 100 ml/day 5% glucose injection (placebo) in-

travenously at the rate of 2040 drops per minute, for consecutive

7 days. The baseline of patients was tested before intervention.

After treatment for 4 and 7 days, patients were evaluated for the

effects from the intervention. Except for SMI, any other TCM that

could confuse the outcome of SMI was forbidden. In the case of

comorbidities, only the medicine that fullled the corresponding

guidelines was enabled.

2.2. Study products

SMI, prepared from Panax ginseng and Ophiopogon japonicus,

has been mass-produced as a patented drug based on the national

standards approved by CFDA (China Food and Drug Administra-

tion). Briey, Panax ginseng (100 g) and dried tuber roots ofOphiopogon japonicus (100 g) were respectively extracted by re-

uxing with 90% ethanol for 2 h each time (Panax ginseng, six

times; Ophiopogon japonicus, two times). The extract was con-

densed to the concentration of 0.30.4 g crude drug weight/ml

solution. The solution was then stirred with 1% acticarbon for 1 h,

followed by ltration and dilution with water for injection. This

process was repeated twice. Finally, polysorbate 80 was added and

the solution was diluted to 1000 ml with water for injection. SMI

(batch number: 1105079) used in our research was manufactured

in accordance with applicable GMP by Qingchunbao Pharmaceu-

tical Co., Ltd (Hangzhou, China). The batch sample analyses and

characteristic records of SMI were maintained. As reported pre-

viously (Wu et al., 2013), high-performance liquid chromatography

(HPLC) analysis was performed to identify the phytochemicalcharacteristics of SMI. The HPLC chromatogram of SMI including

ginsenosides as standard was showed in Fig. 1. The components

conformed to the national standards of Shenmai injection ap-

proved by CFDA.

The placebo (5% glucose injection) was proprietary prepara-

tions produced by Dongguan Lifeline Pharmaceutical Co., Ltd with

the batch number of 11070603. In order to avoid the exposure of

study drugs from the brown color of SMI, light-resistant containers

and infusion apparatus were used during treatment.

2.3. Study participants

Inclusion criteria include male and female patients with 4080

years of age; history of CHF combined with CAD; New York Heart

Association (NYHA) functional classication ; diagnosis of Qi-

Yin deciency (Wang et al., 2013); subjects received standard

medicines for at least 1 month, whose condition remained un-

stable and required further treatment in hospital; submission of

informed consent. Exclusion criteria include the patients with

acute cardiac dysfunction; exacerbation of CHF induced by the

adverse effects of digitalis; comorbidities of cardiac shock, malig-

nant cardiac arrhythmia, degree type or severer atrioven-

tricular block, obstructive myocardiopathy, unrepaired valvular

heart disease, constrictive pericarditis, cardiac tamponade, pul-

monary embolism, acute myocardial infarction, or uncontrolled

Fig.1. The HPLC chromatogram of SMI (Shenmai injection). (A) The HPLC chro-

matogram of control solution that contains six main compounds of SMI (ginseno-

side Rg1, Re, Rf, Rb1, Rb2 and Rd); (B) The HPLC chromatogram of SMI. In both

(A) and (B), chromatographic peaks from 1 to 6 represented Rg1, Re, Rf, Rb1, Rb2

and Rd respectively.

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142 137

7/25/2019 Xian 2016

3/7

infection; subjects with serious liver, kidney, or hematopoietic

system disease; alanine transaminase or aspartate transaminase

levels480 U/L; mental disorder; diabetes patients with un-

controlled plasma glucose level (fasting plasma

glucose47.0 mmol/L, postprandial plasma glucose410.0 mmol/

L); pregnancy or breastfeeding; participation in other clinic trials

within 1 month; allergic to SMI.

2.4. Clinical endpoints

The primary endpoint was NYHA functional classication. The

secondary endpoints were 6-min walking distance (6MWD),

short-form 36 (SF-36) hearth survey score, TCM syndrome score

(see detailed evaluation method in Appendix A), left ventricular

ejection fractions (LVEF) and B-type natriuretic peptide (BNP) le-

vel. All these endpoints were tested at baseline. NYHA functional

classication, TCM syndrome score and BNP level were evaluated

following intervention for 4 and 7 days. 6MWD, SF-36 hearth

survey score and LVEF were performed after treatment for 7 days.

2.5. Laboratory tests and Safety monitoring

Routine examinations (blood and urine routine examination,

feces examination, blood biochemistry, liver and renal function,

and electrocardiogram) were carried out at baseline and after

treatment for 7 days. Safety was evaluated by vital signals (tem-

perature, blood pressure and heart rate), laboratory tests, and

adverse events (AEs) that were monitored throughout the study.

2.6. Statistical analysis

The sample size was estimated based on the expected im-

provement in the primary endpoint (NYHA functional classica-

tion) of SMI through the previous clinic trials, from which themean percentage of patients with improve NYHA function was

85.81% and 63.22% in SMI and placebo groups, respectively.

Therefore, assuming an additional 23% improvement in NYHA

function following treatment with SMI (SMI, 86%; placebo, 63%)

and given a power (1-) of 80% and two-tail typeerror of 5%, 68

patients would be required for each group to detect the superiority

of SMI as compared with placebo (superiority margin 0.05).Moreover, considering the requirements of the drug registration

regulation of CFDA and the dropout rate was approximately 20%, a

total of 240 patients (120 per treatment group) was needed to be

randomized to achieve the required number of patients for the

efcacy analysis.

Statistical analyses were performed using SAS. The analysis for

the primary endpoint was based on the full analysis set (FAS).

Continuous variables were presented as mean7SD, except for

BNP level, which is natural logarithm-transformed before analysis.

Generally, continuous variables were assessed by Student's t-test

between groups or by paired t-test within group, while discrete

variables were analyzed by chi-square test. For the comparison of

NYHA functional classication between groups, the Cochran

Mantel Haenszel (CMH) test was performed. The Fisher exact test

was used to analysis the ratio of adverse events between groups.

For NYHA functional classication and TCM syndrome score, the

missing data was estimated by the last observed carried forward

(LOCF) with at least one post-treatment evaluation (Shao and

Zhong, 2003). Signicance was attributed when a two-tail

po0.05.

3. Results

3.1. Patient characteristics

Patient enrolment was started in September 2011 and com-

pleted in August 2012. According to the inclusion and exclusion

criteria, 240 patients (SMI, n120; placebo, n120) were enrolled

and randomized from eight A-level hospitals in China. Patient

disposition is summarized inFig. 2. Overall, 240 patients receivedstudy drug were included in the safety evaluation. Among them,

228 patients (SMI, n114; placebo, n114) were eligible for the

primary analysis. Three patients (SMI, n2; placebo, n1) were

excluded from the primary analysis due to the violation of the

inclusion criteria. The other eight patients (SMI, n4; placebo,

n4) were excluded because of protocol deviations that could

affect efcacy assessments, the most common being use of other

TCMs during the treatment period (SMI, n4; placebo, n2). One

placebo recipient withdrew the informed consent form. The dis-

tribution of the demographic and baseline characteristics between

SMI and placebo groups were well balanced (Table 1).

3.2. NYHA functional classication

At enrolment, the majority of patients were in NYHA to

(89.5%) and the remaining patients were in NYHA (10.5%).

Baseline NYHA functional classication was similar between the

two groups (Table 1). After treatment with either SMI or placebo

plus standard medicines, the frequency of NYHA to patients

gradually increased, whereas the frequency of NYHA to pa-

tients decreased continuously (Fig. 3). The frequency of NYHAand

NYHA within SMI group from baseline over 7 days presented an

increase of 22.8% and 12.3%, as opposed to 8.8% and 14.9% in

placebo group, while the frequency of NYHA and NYHA de-

creased by 28.1% and 7.0% in SMI group, in contrast to 19.3% and

4.4% in placebo group. SMI treatment resulted in superior im-

provements of NYHA functional classi

cation compared with theplacebo (p0.001,Fig. 3).

3.3. 6-min walking distance

As shown in Fig. 4, the two groups demonstrated similar

6MWD at baseline (p0.418). After treatment for 7 days, 6MWD

was signicantly improved within both groups (SMI, po0.001;

placebo, po0.001). On day 7, patients in the SMI group had a

mean 6MWD change from baseline of 112.21 m, whereas patients

in placebo group had a mean change of 63.03 m. Compared with

placebo, treatment with SMI resulted in a statistically signicant

treatment difference in the mean 6MWD change (po0.001,Fig. 4).

Fig. 2. Flow chart of patients throughout the study. (FAS, Full Analysis Set).

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142138

7/25/2019 Xian 2016

4/7

3.4. SF-36 health survey scores

Fig. 5 shows the mean change of SF-36 health survey scores

from baseline over 7 days. On day 7, patients receiving SMI pre-

sented higher SF-36 health survey scores compared with patients

with placebo (po0.001). Compared with baseline, the mean

change within the SMI group was an increase of 13.44 (po0.001),

and the change in the placebo group was an increase of 5.71

(po0.001). Moreover, the difference in the mean change of the SF-

36 health survey scores was also signicant between two groups

(po0.001).

3.5. TCM syndrome scores

A favorable effect of SMI was also observed on the TCM

syndrome scores. After treatment for 4 and 7 days, TCM syndrome

scores decreased signicantly within both SMI and placebo group

(Fig. 6A). The mean change in TCM syndrome score from baseline

over 4 and 7 days was higher in SMI group, compared with pla-

cebo group (Fig. 6B).

3.6. LVEF and BNP level

The analysis of LVEF was based on patients with LVEFr50%

(SMI, n46; placebo, n48; p0.70). At baseline, the LVEF level

in those patients did not differ between the SMI and placebo

Table 1

Baseline characteristics of patients enrolled in the study.

Char act erist ic s P lac eb o (n114) Shenmai injection

(n114)

p-Value

Sex 0.50

Men 66 (57.9%) 71 (62.3%)

Women 48 (42.1%) 43 (37.7%)

Age 68.1278.88 68.9579.91 0.51

Measurements

Weight (kg) 61.03711.05 62.06712.50 0.51

Height (cm) 161.8577.88 163.0078.13 0.28

BMI (kg/m2) 23.2573.75 23.2573.44 1.00

Heart rate (beats/min) 77.46711.28 78.85715.12 0.43

Systolic BP (mmHg) 128.96720.79 128.39719.43 0.83

Diastolic BP (mmHg) 75.72710.78 75.57712.09 0.92

NYHA class 0.62

0 (0.0%) 0 (0.0%)

56 (49.1%) 55 (48.2%)

43 (37.7%) 50 (43.9%)

15 (13.2%) 9 (7.9%)

LVEF (r50%) 48 (42.1%) 46 (40.4%) 0.70

LVEF (%) of Patients

with LVEFr50%

39.2577.59 38.8477.65 0.80

Medications

ACE inhibitor 101 (88.6%) 93 (81.6%) 0.14

ARB 33 (29.0%) 40 (35.1%) 0.32

Beta-blocker 70 (61.4%) 66 (57.9%) 0.59

Digoxin 36 (31.6%) 34 (29.8%) 0.77

Loop diuretic 44 (38.6%) 43 (37.7%) 0.89

Calcium channel

blockers

32 (28.1%) 37 (32.5%) 0.47

ADP receptor blocker 58 (50.9%) 51 (44.7%) 0.35

Cephalosporins 7 (6.1%) 12 (10.5%) 0.23

Statins 74 (64.9%) 64 (56.1%) 0.18

Laboratory

measurements

Sodium (mmol /l ) 13 8. 61713.01 140.0673.29 0.25Potassium (mmol/l) 5.32712.98 4.0570.47 0.30

Creatinine (mol/l) 99.81746.00 97.21732.93 0.63

Hemogl ob in ( g/l) 12 4. 86718.55 128.00717. 13 0.18

ALT (U/l) 21.45713.10 22.69715.12 0.51

AST(U/l) 24.40710.86 26.58711.34 0.14

Six-minute walk test

(m)

318.587128.67

(n113)

304.467133.60 0.42

BNP (pg/ml)a 4.9471.84 (n107) 5.0771.80 (n106) 0.61

Values are expressed as Mean7SD or n (%); BMI, Body Mass Index; BP, Blood

Pressure; NYHA, New York Heart Association; LVEF, Left Ventricular Ejection Frac-

tion; ACE, Angiotensin-Converting Enzyme; ARB, Angiotensin-Receptor Blockers;

ADP, Adenosine Diphosphate; ALT, Alanine Transaminase; AST, Aspartate Transa-

minase; BNP, B-type Natriuretic Peptide.a BNP level is natural logarithm-transformed before analysis.

Fig. 3. The effect of Shenmai injection on the results of CHF NYHA functional

classication.

Fig. 4. Change in 6-Min Walking Distance (6MWD) from Baseline over 7 days

treatment with Shenmai injection. ***po0.001 compared with the Placebo group.

Fig. 5. Change in Short-Form 36 (SF-36) Health Survey scores from Baseline over7 days treatment with Shenmai injection. ***po0.001 compared with the Placebo

group.

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142 139

7/25/2019 Xian 2016

5/7

groups. A trend towards higher LVEF level was observed on day

7 within each groups, however the difference in the mean LVEF

level change was not statistically signicant between SMI and

placebo (Table 2). Baselines of BNP levels were similar in the two

groups. After treatment for 7 days, the mean change of BNP levels

was also not signicantly different between SMI and placebo

groups (Table 2).

3.7. Safety evaluation

The safety evaluation included all patients who received at

least one dose of the study drug and had at least one post-baseline

safety assessment. Compared with placebo (data not included),

there was no signicant difference on index of kidney and liverfunction (Cr, ALT and AST levels) after treatment with SMI for

7 days between two groups (p40.05). Moreover, no signicant

difference was also found within the group respectively (p40.05).

Treatment with SMI within 1 week was well tolerated with no

apparent safety concerns (Table 3andAppendix B). Seven patients

experienced AEs during the study: four (3.33%) patients in the SMI

group and three (2.50%) patients in the placebo group. The ma-

jority of events were of mild-to-moderate severity (See detailed

AEs information inAppendix B). One AE in SMI group was serious

but caused by acute appendicitis. Two AEs in placebo group was

suspected to be related to guideline medicines which possibly

induced chills, fever and high plasma glucose level. A similar

proportion of patients in each treatment group had an AE that led

to discontinuation of study drug (SMI: n1; placebo: n1). The

analysis of drug induced AEs, serious AEs and total AEs revealed no

differences between the two groups (Table 3).

4. Discussion

TCMs have been used for thousands of years in China for pro-

moting the homeostatic balance of the body, which was believed

to be effective, safe and natural (Chen et al., 2012;Tang and Huang,

2013). Shenmai, a traditional Chinese medicine, is derived from

the ancient TCM literature medicine origin as a treatment for Qi-

Yin deciency that has been recognized corresponding to the

syndromes of CHF combined with CAD by modern medical experts

in China. Similar to other TCMs, the effects of Shenmai are too

obscure to be understand in terms of TCM theory, which drives the

demand to re-evaluation them by contemporary evidence-based

trials. Although most previous studies of SMI have reported

Fig. 6. TCM syndrome score (A) and Mean Change of TCM syndrome score (B) after treatment with Shenmai injection for 4 and 7days. Mean change of TCM syndrome

score(Baseline-4days) or (Baseline-7days). ***po0.001 compared with the baseline of the same group; *po0.05 compared with the Placebo group.

Table 2

The changes of LVEF and BNP from baseline over 7 days treatment with Shenmai injection.

Item Time point Placebo Shenmai injection

Outcome Mean changefrom baseline Outcome Mean changefrom baseline

LVEF(%) of Patients with LVEF r50% Baseline 39.2577.59 (n48) 38.8477.65 (n46)

7 days 42.50710.13 (n44) 3.7179.27 (n44) 40.68710.28 (n41) 2.2278.26 (n41)

Baseline 4.9471.84 (n107) 5.0771.80 (n106)

BNP (pg/ml)a 4 days 4.5871.82 (n97) 0.2770.78 (n93) 4.9971.85 (n92) 0.0570.94 (n87)

7 days 4.6571.74 (n103) 0.1670.86 (n99) 5.0271.73 (n95) 0.1370.84 (n90)

Values are expressed as Mean7SD; LVEF, Left Ventricular Ejection Fraction; BNP, B-type Natriuretic Peptide.a BNP level is natural logarithm-transformed before analysis.

Table 3

Summary of adverse events.

Placebo (n120) Shenmai injection

(n120)

p-Value

Adverse events 3 (2.50%) 4 (3.33%) 1.000

Serious adverse event 0 (0.00%) 1 (0.83%) 1.000

Drugs induced AEs 2 (1.67%) 0 (0.00%) 0.245

The analysis included all patients who received at least one dose of the study

medication; AEs, Adverse Events.

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142140

7/25/2019 Xian 2016

6/7

modest improvements of CHF, few were designed and regulated as

rigorously as the clinical trials in contemporary pharmaceuticals

(Chen et al., 2012), with serious problems involving lack of de-

tailed process of randomization, absence of allocation conceal-

ment, unclear baseline characteristics, open-label design, and

small scale of subjects. Therefore, a well-designed study was per-

formed to explore the benets from SMI. Besides these validated

measures, a new index system (Mao et al., 2009) that has been

developed to specially quantify the clinical effects from TCMsbased on TCM syndromes, was used in this clinic trial as a com-

plementary endpoint.

SMI is widely used in China for patients hospitalized due to CHF

and approved by China Food and Drug Administration as a patent

drug. In clinical practice, SMI with a dose of 50100 ml per day is

applied as a complement to standard medicines and is often dis-

continued after the relief of symptoms or disease remission ( Chen

et al., 2012), which usually limited the treatment duration to only

1 week. Importantly, although advocated, this approach (SMI dose:

100 ml/day, treatment duration: 1 week) has never been pro-

spectively validated in clinical trials in CHF. Therefore, this study is

tried to evaluate whether patients with CHF and CAD can benet

from this already acceptable treatment in China.

240 patients with CHF combined with CAD were enrolled and

randomized in this study. The distribution of demographic and

baseline characteristics of included patients were homogeneous.

For patients receiving SMI, the improvement of the primary end-

point (the NYHA functional classication) from baseline over

7 days was superior to those in placebo group. This nding par-

alleled the improvement in 6MWD, SF-36 hearth survey score and

TCM syndrome scores. Subsequent safety evaluation from the vital

signals, laboratory tests, and adverse events did not show sig-

nicant difference between SMI and placebo. Together, our data

suggested that standard medicines plus SMI with a dose of 100 ml

per day could further ameliorate the condition of CHF patients

combined with CAD within the treatment duration of 1 week,

when SMI was well tolerated with no apparent safety concerns. It

is of note that despite demonstration of improvements with these

endpoints based on the current scheme, there is no evidence in

this study to support any improvements in long-term outcomes

and safety of SMI. More extensive research should be performed if

a long-term duration is expected.

CHF is a heterogeneous and complex clinical entity with nu-

merous underlying pathophysiological mechanisms (Francis,

2001; Kemp and Conte, 2012). The underlying mechanisms of

Shenmai may be related to its multiple actions targeting at CHF.

Shenmai could improve a series of hemodynamic parameters

(Chen et al., 2002;Zhang et al., 2008), involving enhancement of

coronary ow, augment of myocardial contractility, reduction of

heart rate and down-regulation of blood viscosity. The study by Liu

showed that Shenmai could inhibit secretion of inammatory

factors, such as TNF-, IL-6 and IL-8 (Liu et al., 2013). In addition,

Wang found that Shenmai could prevent cells from oxidative in-jury through modulating cellular antioxidants. Therefore, the

mechanisms of Shenmai on CHF may be due to its inhibition of

cardiomyocytes apoptosis by balance between pro-inammatory

and anti-inammatory and down-regulation oxygen free radicals.

Furthermore, Shenmai may possibly delay the cardiac remodeling.

Ni found that Shenmai could inhibit myocardial brosis in rat with

diabetic cardiomyopathy (Ni et al., 2011). Zhang investigated the

effect of SMI on heart failure rats and found SMI could reduce the

plasma concentrations of Angand endothelin (Zhang et al., 2001).

Although the mechanism of Shenmai has been gradually re-

cognized, it's still unclear which ingredient is really effective.

Therefore, further studies on the isolated active components of

SMI are warranted.

Given the characteristics of the study, several limitations are

apparent. Because this study was designed based on the widely

used therapy (SMI dose: 100 ml/day, treatment duration: 1 week)

from clinical practice, the duration was limited to 7 days. However,

it's difcult to obtain the re-admission rate and mortality, which

are important and useful outcomes for evaluation. Moreover, si-

milar to other patented TCMs, the effects of SMI may be ques-

tionable due to the variability in concentration levels of active

ingredients. Fortunately, SMI is usually mass-produced based on a

standardized formula. And recent study by Li (Li et al., 2013a) hasalso proved that a panel of key ingredients of SMI was relatively

consistent within a statistically acceptable range through si-

multaneously capturing the entire spectrum of ingredients, which

suggests that the effects from SMI are reliable.

In summary, the results of this study suggest that the in-

tegrative treatments with guideline medicines plus SMI could

further improve the condition of patients with CHF and CAD with

respect to NYHA functional classication, 6MWD, SF-36 health

survey score and TCM syndrome score. Therefore, SMI with a dose

of 100 ml per day and treatment duration of 1 week could be re-

commended as a complement to the standard medicines for pa-

tients with CHF combined with CAD.

Contributors

Dr. Shaoxiang Xian and Dr. Zhongqi Yang are the co-rst au-

thors and contributed equally to this work. Shaoxiang Xian and

Zhongqi Yang designed the study. Zhiping Jiang, Xiaohan Ye, Luyi

Luo and Lili Jin conducted the patient inclusion, reviewed all cases,

collected patient information and compiled the data les. Tianlun

Yang, Jun Lee and Suilin Ye collected, processed and compiled the

laboratory data. Jun Lee, Suilin Ye and Dongfeng Lu performed the

statistical analyses. Shaoxiang Xian, Zhongqi Yang, Luyi Luo and

Jun Lee drafted the paper. Shaoxiang Xian, Zhongqi Yang, Xiaohan

Ye and Luyi Luo contributed to critical revision for important in-

tellectual content. All authors approved the nal manuscript.

Funding

This work was supported by the Major Project for Major New

Drugs Innovation and Development from the Ministry of Science

and Technology of China (2008ZX09202-006) and Chiatai Qing-

chunbao Pharmaceutical Co., Ltd (ZYF-Y-2010-056).

Competing interests

Dr. Shaoxiang Xian received research grants from Chiatai

Qingchunbao Pharmaceutical Co., Ltd. All other authors have de-

clared that they have no relationships relevant to the contents ofthis paper to disclose.

Acknowledgements

We are grateful for the contributions of our study team, in-

cluding study coordinators: Mulan Wang, Yanxin Qian, Linjie Xu

(Chiatai Qingchunbao Pharmaceutical Co., Ltd.); statistical analy-

sis: Zhaosi Xu, Shengpeng Zhang (Guangzhou Yushi Medicinal

Technology Co., Ltd.), Yi Shen (School of Public Health, Zhejiang

University, Hangzhou, China). The funders had no role in the de-

sign and conduct of the study; collection, management, analysis,

and interpretation of the data; or preparation and review of the

manuscript.

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142 141

7/25/2019 Xian 2016

7/7

Appendix A. Supplementary material

Supplementary data associated with this article can be found in

the online version at doi:10.1016/j.jep.2016.03.066.

References

Chen, J., Yao, Y., Chen, H., Kwong, J.S., 2012. Shengmai (a traditional Chinese herbalmedicine) for heart failure. Cochrane Database of Systematic Reviews. 11,CD005052.

Chen, W., Meng, Q., Hong, S., 2002. Efcacy of Shengmai injection on haemody-namics in heart dysfunction dogs. Chin. J. Crit. Care Med. 22, 81 83.

Francis, G.S., 2001. Pathophysiology of chronic heart failure. Am. J. Med. 110 (Suppl7A), 37s46s.

Go, A.S., Mozaffarian, D., Roger, V.L., Benjamin, E.J., Berry, J.D., Blaha, M.J., Dai, S.,Ford, E.S., Fox, C.S., Franco, S., Fullerton, H.J., Gillespie, C., Hailpern, S.M., Heit, J.A., Howard, V.J., Huffman, M.D., Judd, S.E., Kissela, B.M., Kittner, S.J., Lackland, D.T., Lichtman, J.H., Lisabeth, L.D., Mackey, R.H., Magid, D.J., Marcus, G.M., Marelli,A., Matchar, D.B., McGuire, D.K., Mohler 3rd, E.R., Moy, C.S., Mussolino, M.E.,Neumar, R.W., Nichol, G., Pandey, D.K., Paynter, N.P., Reeves, M.J., Sorlie, P.D.,Stein, J., Towghi, A., Turan, T.N., Virani, S.S., Wong, N.D., Woo, D., Turner, M.B.,2014. Heart disease and stroke statistics 2014 update: a report from theAmerican Heart Association. Circulation 129, e28e292.

Kemp, C.D., Conte, J.V., 2012. The pathophysiology of heart failure. Cardiovasc. Pa-thol. 21, 365371.

Li, X., Chen, H., Jia, W., Xie, G., 2013a. A metabolomics-based strategy for the quality

control of traditional chinese medicine: shengmai injection as a case study.Evid. Based Complement. Alter. Med. 2013, 836179.

Li, X., Zhang, J., Huang, J., Ma, A., Yang, J., Li, W., Wu, Z., Yao, C., Zhang, Y., Yao, W.,Zhang, B., Gao, R., 2013b. A multicenter, randomized, double-blind, parallel-group, placebo-controlled study of the effects of qili qiangxin capsules in pa-tients with chronic heart failure. J. Am. Coll. Cardiol. 62, 10651072.

Liu, Z.M., Li, N., Yu, H., Chen, Y., 2013. Experimental study of the effect of Shenmaiinjection on post-cardiac arrest syndrome in rabbit. Chin. Crit. Care Med. 25,664668.

Ma, R.G., Wang, C.X., Shen, Y.H., Wang, Z.Q., Ma, J.H., Huang, L.S., 2010. Effect ofShenmai Injection on ventricular diastolic function in patients with chronicheart failure: an assessment by tissue Doppler imaging. Chin. J. Integr. Med. 16,173175.

Maciocia, G., 2005. The Foundations of Chinese Medicine, 2nd revised edition.Elsevier Health Sciences, London, United Kingdom.

Mao, J., Hou, Y., Shang, H., Wang, H., Wang, X., Zhao, Y., Niu, T., Cui, J., Li, G., Lin, Q.,Shi, L., Jia, X., Fan, R., Wang, B., Ruan, J., 2009. Study on the evaluation of theclinical effects of traditional chinese medicine in heart failure by complex in-tervention: protocol of SECETCM-HF. Trials 10, 122.

McMurray, J.J., Adamopoulos, S., Anker, S.D., Auricchio, A., Bohm, M., Dickstein, K.,Falk, V., Filippatos, G., Fonseca, C., Gomez-Sanchez, M.A., Jaarsma, T., Kober, L.,Lip, G.Y., Maggioni, A.P., Parkhomenko, A., Pieske, B.M., Popescu, B.A., Ronnevik,P.K., Rutten, F.H., Schwitter, J., Sef erovic, P., Stepinska, J., Trindade, P.T., Voors, A.A., Zannad, F., Zeiher, A., Bax, J.J., Baumgartner, H., Ceconi, C., Dean, V., Deaton,C., Fagard, R., Funck-Brentano, C., Hasdai, D., Hoes, A., Kirchhof, P., Knuuti, J.,

Kolh, P., McDonagh, T., Moulin, C., Popescu, B.A., Reiner, Z., Sechtem, U., Sirnes,P.A., Tendera, M., Torbicki, A., Vahanian, A., Windecker, S., McDonagh, T.,Sechtem, U., Bonet, L.A., Avraamides, P., Ben Lamin, H.A., Brignole, M., Coca, A.,Cowburn, P., Dargie, H., Elliott, P., Flachskampf, F.A., Guida, G.F., Hardman, S.,

Iung, B., Merkely, B., Mueller, C., Nanas, J.N., Nielsen, O.W., Orn, S., Parissis, J.T.,Ponikowski, P., 2012. ESC guidelines for the diagnosis and treatment of acuteand chronic heart failure 2012: the task force for the diagnosis and treatment ofacute and chronic heart failure 2012 of the European Society of Cardiology.

Developed in collaboration with the Heart Failure Association (HFA) of the ESC.Eur. J. Heart Fail. 14, 803869.

Murdoch, D.R., Love, M.P., Robb, S.D., McDonagh, T.A., Davie, A.P., Ford, I., Capewell,S., Morrison, C.E., McMurray, J.J., 1998. Importance of heart failure as a cause ofdeath. Changing contribution to overall mortality and coronary heart disease

mortality in Scotland 19791992. Eur. Heart J. 19, 18291835.Ni, Q., Wang, J., Li, E.Q., Zhao, A.B., Yu, B., Wang, M., Huang, C.R., 2011. Study on the

protective effect of shengmai san (see text) on the myocardium in the type

2 diabetic cardiomyopathy model rat. J. Tradit. Chin. Med. 31, 209 219.Owan, T.E., Hodge, D.O., Herges, R.M., Jacobsen, S.J., Roger, V.L., Redeld, M.M.,

2006. Trends in prevalence and outcome of heart failure with preserved ejec-tion fraction. New Engl. J. Med. 355, 251259.

Roger, V.L., 2013. Epidemiology of heart failure. Circ. Res. 113, 646659.Roger, V.L., Weston, S.A., Redeld, M.M., Hellermann-Homan, J.P., Killian, J., Yawn, B.

P., Jacobsen, S.J., 2004. Trends in heart failure incidence and survival in acommunity-based population. JAMA 292, 344350.

Shao, J., Zhong, B., 2003. Last observation carry-forward and last observation ana-

lysis. Stat. Med. 22, 24292441.Tang, W.H., Huang, Y., 2013. Cardiotonic modulation in heart failure: insights from

traditional Chinese medicine. J. Am. Coll. Cardiol. 62, 10731074.Wang, J., Yang, X., Chu, F., Chen, J., He, Q., Yao, K., Teng, F., Gao, Y., Xing, Y., Wu, A.,

Xing, Y., 2013. The effects of xuefu zhuyu and shengmai on the evolution ofsyndromes and inammatory markers in patients with unstable angina pec-toris after percutaneous coronary intervention: a randomised controlled clin-

ical trial. Evid. Based Complement. Alter. Med. 2013, 896467 .Wu, L., Wang, Y., Nie, J., Fan, X., Cheng, Y., 2013. A network pharmacology approach

to evaluating the efcacy of chinese medicine using genome-wide transcrip-tional expression data. Evid. Based Complement. Alter. Med. 2013, 915343.

Yancy, C.W., Jessup, M., Bozkurt, B., Butler, J., Casey Jr., D.E., Drazner, M.H., Fonarow,G.C., Geraci, S.A., Horwich, T., Januzzi, J.L., Johnson, M.R., Kasper, E.K., Levy, W.C.,Masoudi, F.A., McBride, P.E., McMurray, J.J., Mitchell, J.E., Peterson, P.N., Riegel,B., Sam, F., Stevenson, L.W., Tang, W.H., Tsai, E.J., Wilkoff, B.L., 2013. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American

College of Cardiology Foundation/American Heart Association Task Force onPractice Guidelines. J. Am. Coll. Cardiol. 62, e147e239.

Yu, J., Xin, Y.F., Gu, L.Q., Gao, H.Y., Xia, L.J., You, Z.Q., Xie, F., Ma, Z.F., Wang, Z., Xuan,Y.X., 2014. One-month toxicokinetic study of SHENMAI injection in rats. J.

Ethnopharmacol. 154, 391399.Zhang, X., Wang, Q., LM, C., GL, C., 2001. Effects of Shenmai injection on the plasma

angiotensin,endothelin, atrial natriuretic polypeptide and the myocardial

contractility and relaxation of left ventricular in experimental rats with heartfailure. Chin. J. Integr. Tradit. West. Med. Intensive Crit. Care 8, 3637.

Zhang, Y.C., Lu, B.J., Zhao, M.H., Rong, Y.Z., Chen, R.M., 2008. Effect of Shengmaiinjection on vascular endothelial and heart functions in patients with coronary

heart disease complicated with diabetes mellitus. Chin. J. Integr. Med. 14,281285.

S. Xian et al. / Journal of Ethnopharmacology 186 (2016) 136142142

http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref9http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref9http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref9http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref24http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref23http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref22http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref21http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref20http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref19http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref18http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref17http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref16http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref15http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref14http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref13http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref12http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref11http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref10http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref9http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref9http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref8http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref7http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref6http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref5http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref4http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref3http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref2http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1http://refhub.elsevier.com/S0378-8741(16)30189-1/sbref1