7/25/2019 ha2016

1/7

Prospective, randomized, double-blinded, double-dummy andmulticenter phase IV clinical study comparing the efcacy and safety ofPG201 (Layla) and SKI306X in patients with osteoarthritis

Chul-Won Ha a,b,1, Yong-BeomPark c,1, Byung-Woo Min d, Seung-Beom Han e,Jae Hyup Lee f, Ye-Yeon Won g, Ye-Soo Park h,n

a Department of Orthopedic Surgery, Stem Cell & Regenerative Medicine Institute, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81

Irwon-ro, Gangnam-gu, Seoul 06351, South Koreab Department of Health Sciences and Technology, SAIHST, Sungkyunkwan University, 81 Irwon-ro, Gangnam-gu, Seoul 06351, South Koreac Department of Orthopedic Surgery, Chung-Ang University Hospital, Chung-Ang University College of Medicine, 102 Heukseok-ro, Dongjak-gu, Seoul 06973,

South Koread Department of Orthopedic Surgery, Keimyung University Dongsan Medical Center, 56 Dalseong-ro, Jung-gu, Daegu 41931, South Koreae Department of Orthopedic Surgery, Korea University Medical Center Anam Hospital, 73 Inchon-ro, Seongbuk-gu, Seoul 02841, South KoreafDepartment of Orthopedic Surgery, Seoul Metropolitan Government Seoul National University Boramae Medical Center, 20 Boramae-ro 5-gil, Dongjak-gu,

Seoul 07061, South Koreag Department of Orthopedic Surgery, Ajou University Medical Center, 164 World Cup-ro, Yeongtong-gu, Suwon-si, Gyeonggi-do 16499, South Koreah Department of Orthopedic Surgery, Hanyang University Medical Center Guri Hospital, 153 Gyeongchun-ro, Guri-si, Gyeonggi-do 11923, South Korea

a r t i c l e i n f o

Article history:

Received 14 September 2015

Received in revised form

15 January 2016

Accepted 19 January 2016Available online 25 January 2016

Keywords:

Arthritis

Analgesic

Pain

Traditional medicine Asia & Oceania

Clinical trials

a b s t r a c t

Ethnopharmacological relevance: This prospective, randomized, double-blinded, double-dummy, multi-

center study compared the efcacy and safety of PG201 (Laylas), a new product from extracts of 12 plant

sources and SKI306X (Joinss) which have been well investigated and in relatively wide usage among

herbal medicine, for the treatment of patients with knee osteoarthritis.

Aim of the study: To compare the efcacy and safety of PG201 and SKI306X in patients with knee os-

teoarthritis.Materials and methods: A prospective, double-blinded multicenter study was conducted in 124 patients

with Kellgren and Lawrence grade 23 knee osteoarthritis. Patients were randomly assigned to receive

600 mg of PG201 (300 mg, twice daily) and 600 mg of SKI306X placebo (200 mg, thrice daily) or 60 0 mg

of SKI306X (200 mg, thrice daily) and PG201 placebo (300 mg, twice daily) for 12 weeks. The primary

outcome was the improvement of pain by week 8 as assessed by the 100-mm pain visual analog scale

(VAS). Secondary outcomes included pain VAS improvement level at week 12, pain VAS improvement

rate at weeks 8 and 12, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)

improvement level at weeks 8 and 12, the improvement of the quality of life (EQ-5D), overall symptom

self-assessment score, and rescue medication consumption.

Results: The pain VAS improvement at 8 weeks was 14.2716.2 in the experimental group and 11.9713.1

in control group (p0.557), conrming that the experimental group was not inferior to the control group

as lower limit (-8.38) of 95% CI of the difference of VAS improvement between two groups was well

above the allowed limit (-10 mm). There was no signicant difference in all secondary outcomes in-

cluding pain VAS, WOMAC, EQ-5D, overall symptom self-assessment score, and rescue medication

consumption. Adverse events were low and similar between the two groups.

Conclusions: The results of this study showed that PG201 signicantly reduced knee pain and improvedknee function and were comparable to SKI306X. PG201 can be suggested as an effective treatment of

knee osteoarthritis.

Trial registration ClinicalTrials.gov:NCT01768468

& 2016 Published by Elsevier Ireland Ltd.

1. Introduction

Non-steroidal anti-inammatory drugs (NSAIDs) and analgesics

are used for symptom relief in patients with osteoarthritis (OA).

However, the use of these drugs may not be safe and appropriate

Contents lists available at ScienceDirect

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

Journal of Ethnopharmacology

http://dx.doi.org/10.1016/j.jep.2016.01.029

0378-8741/&2016 Published by Elsevier Ireland Ltd.

n Corresponding author. Tel.: 82 31 560 2316; fax: 82 31 557 8781.

E-mail address: hyparkys@hanyang.ac.kr(Y.-S. Park).1 These authors contributed equally.

Journal of Ethnopharmacology 181 (2016) 17

http://www.sciencedirect.com/science/journal/03788741http://www.elsevier.com/locate/jephttp://dx.doi.org/10.1016/j.jep.2016.01.029mailto:hyparkys@hanyang.ac.krhttp://dx.doi.org/10.1016/j.jep.2016.01.029http://dx.doi.org/10.1016/j.jep.2016.01.029http://dx.doi.org/10.1016/j.jep.2016.01.029http://dx.doi.org/10.1016/j.jep.2016.01.029mailto:hyparkys@hanyang.ac.krhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.01.029&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.01.029&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.jep.2016.01.029&domain=pdfhttp://dx.doi.org/10.1016/j.jep.2016.01.029http://dx.doi.org/10.1016/j.jep.2016.01.029http://dx.doi.org/10.1016/j.jep.2016.01.029http://www.elsevier.com/locate/jephttp://www.sciencedirect.com/science/journal/03788741

7/25/2019 ha2016

2/7

symptom control may not be accomplished (Jordan et al., 2003;

Zhang et al., 2005; Zhang et al., 2008). Glucosamine and chon-

droitin sulfate are widely administered as supplementary treat-

ment drugs; however, their efcacy remains disputed (Hochberg,

2006;Wandel et al., 2010).

Herbal medicines have been commonly used to treat osteoar-

thritis in Asia for centuries (Cameron et al., 2009; Cameron and

Chrubasik, 2014; Chen et al., 2014). Although the mechanism of

action of oral medicinal plant products have not been fully eluci-dated, their interactions with the mediators of inammation and

cartilage destruction provide a rationale for their use in the

treatment of OA (Cameron et al., 2009).

PG201 (Laylas) is a medicine extracted from 12 herbs which

have been used to treat osteoarthritis in traditional medicine in

Eastern Asia (Kim, 2000;Neung, 1998; Park and Lee, 2000; Shin

et al., 2003). The herbs used for the formulation of PG201 are as

follows: Chaenomeles sinensis (Dum.Cours.) Koehne, fructus (sy-

nonym; Chaenomeles chinensis(Dum.Cours.) Koehne), Achyranthes

japonica (Miq.) Nakai, radix (synonym; Achyranthes bidentata

Blume),Acanthopanax sessiliorus (Rupr. & Maxim.) Seem, cortex

(synonym; Eleutherococcus sessiliorus (Rupr. & Maxim.) S.Y.Hu.),Cinnamomum cassia (L.) J.Presl, cortex (synonym; Cinnamomum

aromaticum Nees), Gentiana macrophylla Pall., radix, (synonym;

Tretorhiza macrophylla (Pall.) Sojk), Clematis chinensis Osbeck,

radix, (synonym; Clematis benthamiana Hemsl.), Angelica gigas

Nakai, radix, (No synonyms), Cnidium ofcinale Makino, rhizoma

(synonym; Ligusticum ofcinale (Makino) Kitag.), Gastrodia elata

Blume, rhizoma (synonym; Gastrodia elata f. pilifera Tuyama),

Carthamus tinctorius L., os (synonym; Carthamus glaberBurm.f.),Saposhnikovia divaricata (Turcz.) Schischk, radix (synonym; Lede-

bouriella seseloides (Hoffm.) H.Wolff), Dipsacus asperoides C.Y.

Cheng & T.M.Ai, radix (synonym; Dipsacus inermis Wall.). These

herbs have played an important role in the medical care of for

thousands of years in Korea (Heo, 1999;Lee, 1978). For example,

Chaenomeles sinensis, Achyranthes japonica, Carthamus tinctorius,

Cnidium ofcinale and Cinnamomum cassia have been known for

maintaining or assisting blood circulation. Angelica gigas and

Gastrodia elata have been used as a tonifying the blood and pro-moting blood circulation. Saposhnikovia divaricata has been used

for headache, vertigo and arthralgia. Gentiana macrophylla has

been frequently used to treat rheumatoid arthritis. Acanthopanax

sessiliorus has been used for arthralgia due to activity in

strengthening muscle and bone. Clematis chinensis and Dipsacus

asperoides have been used for protection against degeneration of

cartilage and regeneration of damaged tissue. These 12 medicinal

herbal extracts used PG201 has demonstrated anti-inammatory

and cartilage protective effects inin vitroand in vivo studies (Park

et al., 2005;Shin et al., 2003;Choi. et al., 2012;Choi et al., 2013).

These results indicate that the application of PG201 in the treat-

ment of patients with OA may be promising.

SKI306X (Joinss) is another drug prepared from the ethanol

extract of 3 medicinal plants, Clematis manshurica Rupr., radix,(synonym; Clematis terniora var. manshurica (Rupr.) Ohwi), Tri-

chosanthes kirilowiiMaxim, radix, (synonym;Trichosanthes kirilowii

var. japonica Kitan.) and Prunella vaulgaris L., spica, (synonym;

Prunella asiatica f. albiora (Nakai) Kitag.), which has been widely

and traditionally used to decrease pain and improve functional

capacity in patients with OA (Kim et al., 2005). SKI306X also

showed the effects of cartilage protection and anti-inammationin

vitro and in vivo (Choi et al., 2002; Hartog et al., 2008;Kim et al.,

2005). The clinical safety and efcacy of SKI306X has been well

investigated, as evidenced by several studies (Jung et al., 2001;Lung

et al., 2004; Song et al., 2007). SKI306X is currently the most widely

used herb-derived medicine for OA in Korea.

The objective of this study was to evaluate the efcacy and

safety of PG201 in knee OA, compared to the herb-derived medicine

that currently is widely used. We hypothesized that PG201 was not

inferior to SKI306X with respect to safety and efcacy for the

treatment of patients with OA.

2. Methods

2.1. Study design

This prospective, randomized, double-blinded, double dummy,

active drug-comparative, and multi-center phase IV study was

conducted at 6 University Hospitals from October 26, 2012 to June

27, 2013. Each participating hospital obtained the approval of its

respective Institutional Review Board and written informed con-

sent was obtained from all subjects.

2.2. Study population

Eligible patients were aged between 40 and 80 years, diag-

nosed as primary OA according to the American College of Rheu-

matology criteria (Altman et al., 1986), had radiographic evidence

of tibiofemoral OA (Kellgren and Lawrence grade 2 or 3), exhibited

similar OA symptoms for at least 3 months, had less than 80 mm

on a 100 mm pain visual analogue scale (VAS) during screening,

and had greater than 40 mm on the VAS at baseline. Patients were

excluded from the study if they had spinal or lower extremity joint

diseases that could signicantly affect the indicator knee, history

of upper gastrointestinal ulcers within 6 months before screening,

history of upper gastrointestinal hemorrhage (including hema-

temesis) within 12 months prior to the screening, or history of

lower gastrointestinal hemorrhage (excluding hemorrhoids)

within 12 months before screening. The full list of inclusion and

exclusion criteria is detailed in the supplemental Table 1.

2.3. Preparation of PG201

PG201 was prepared using the formulation on the basis of the

known functions of the composing herbs described previously(Shin et al., 2003). The moisture content of each herb was less

than 10% by weight, and the herbs were air-dried. The herbs used

were as follows: Chaenomeles sinensis (Dum.Cours.) Koehne,

fructus (synonym; Chaenomeles chinensis (Dum.Cours.) Koehne),

Achyranthes japonica (Miq.) Nakai, radix (synonym; Achyranthes

bidentata Blume), Acanthopanax sessiliorus (Rupr. & Maxim.)

Seem, cortex (synonym; Eleutherococcus sessiliorus (Rupr. &

Maxim.) S.Y.Hu.), Cinnamomum cassia (L.) J.Presl, cortex (syno-

nym;Cinnamomum aromaticumNees),Gentiana macrophyllaPall.,

radix, (synonym; Tretorhiza macrophylla (Pall.) Sojk), Clematis

chinensis Osbeck, radix, (synonym;Clematis benthamianaHemsl.),

Angelica gigas Nakai, radix, (No synonyms) Cnidium ofcinale

Makino, rhizoma (synonym; Ligusticum ofcinale (Makino) Ki-

tag.),Gastrodia elataBlume, rhizoma (synonym; Gastrodia elataf.pilifera Tuyama), Carthamus tinctorius L., os (synonym; Cartha-

mus glaber Burm.f.), Saposhnikovia divaricata (Turcz.) Schischk,

radix (synonym; Ledebouriella seseloides (Hoffm.) H.Wolff), Dip-

sacus asperoides C.Y.Cheng & T.M.Ai, radix (synonym; Dipsacus

inermis Wall.), 8:8:8:8:5:5:5:5:5:5:4:4 (w/w), respectively. All

herbs were nely cut to 1015 mm in size, mixed, and extracted

with 25% (v/v) ethanol at room temperature for 72 hr. The su-

pernatant was ltered with 10mm cartridge lter and con-

centrated under reduced pressure below 60 C to remove etha-

nol. As a result of specic plant extract, PG201 was manufactured

by blending, mixing, compressing, and coating. Each tablet of

PG201 tablet is tasteless and odorless, yellow oval lm-coated

and contains 300 mg of PG201. The quality of tablets was con-

trolled in various methods such as purity evaluation, identity and

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 172

7/25/2019 ha2016

3/7

quantitative analysis and so on. For this study, we used tablets of

batch number 12001 (exp.07/10/2013).

2.4. Randomization

Patients were randomly assigned to the experimental or con-

trol group at a 1:1 ratio. The subjects in the experimental group

were administered 600 mg of PG201 (300 mg, twice daily) and

600 mg of SKI306X dummy placebo (200 mg, thrice daily) for a

total of 12 weeks. The control group was administered 600 mg of

the SKI306X (200 mg, thrice daily) and 600 mg of PG201 dummy

placebo (300 mg, twice daily) for a total of 12 weeks. The dose and

regimen of PG201 was determined based on a previous preclinical

study and phase 2 and 3 clinical studies (Shin et al., 2003; Yoo

et al., 2014). Both the investigators and patients were blinded to

the drugs provided. The outpatient follow-up observations were

conducted at week 4, 8, and 12. For the NSAID interregnum after

screening, patients were requested to comply with the washout

period according to the provided conversion table; this period did

not exceed a total of 4 weeks from the screening. During the entire

period of the clinical trial (including the washout period), patients

with a severe level of pain were allowed to take the rescue med-

icine (acetaminophen 650 mg) up to a maximum of 6 tablets in 24

hours. Compliance with prescribed treatment was assessed by

counting the numbers of unused tablets at each visit.

2.5. Outcome measures

All efcacy assessment variables were evaluated for each visit.

The primary assessment variable was the 100 mm pain VAS im-

provement level at week 8. The secondary efcacy assessment

variables included the 100 mm pain VAS improvement level at

Assessed for eligibility (n=133)

Excluded (n=9)

Meet exclusion criteria (n=1)

Declined to participate (n=7)

Lost to follow-up (n=1)

Completed (n=52)

Excluded from PP analysis (n=4)

(Out of window period visits(2),Low compliance rate(1),Concomitant medication(1*))

*One case overlap between deviation ofProtocol and excluded from PP

Drop-out due to Adverse event (n=3)

Withdrawal of Informed Consent (n=4)

Allocated to PG201 (n=63)

Received allocated intervention (n=63)

Did not receive allocated intervention (n=0)

Drop-out due to Adverse event (n=6)

Withdrawal of Informed Consent (n=5)

Lost to follow-up (n=1)

Dissatisfied with efficacy of drug (n=2)

Allocated SKI306X (n=61)

Received allocated intervention (n=61)

Did not receive allocated intervention (n=0)

Completed (n=44)

Excluded from PP analysis (n=3))

(Out of window period visits(3))

Allocation

Analysis

Follow-Up

Randomized (n=124)

Enrollment

ITT populationITT population

PP population PP population

Fig. 1. Flow diagram of study population

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 17 3

7/25/2019 ha2016

4/7

week 12, 100mm pain VAS improvement rate at weeks 8 and 12,

Western Ontario and McMaster Universities Osteoarthritis Index

(WOMAC) improvement level at weeks 8 and 12, level of im-

provement of the quality of life (EQ-5D) at weeks 8 and 12, im-

provement in the overall symptom self-assessment score at weeks8 and 12, and amount of rescue medicine usage at weeks 8 and 12.

The overall symptom self-assessment score was performed using

the 5-point Likert scale with the following questions: Please mark

(select) how you feel with your knee, considering every aspect

inuenced by your knee arthritis for the last 24 hours (1very

good, 5very bad).

The status of development of adverse events (AEs) was con-

rmed during each visit based on a physical examination and the

vital signs. Patients were asked to report all AEs by non-leading

questions and cases with signicant symptoms were assessed by

detailed screening. All AEs were recorded in detail based on the

Common Terminology Criteria for Adverse Event v4.03 (Health and

Services (2009)). The laboratory tests including blood tests were

conducted during the screening and the visit at week 12.

2.6. Sample size estimation

Sample size for the two groups was calculated based on the

hypothesis that the average 100 mm pain VAS improvement level

is 29.12 with a standard deviation of 19.96 according to the results

of a previous phase 3 trial of PG201 ( Yoo et al., 2014). As minimal

clinically important difference in the change of pain VAS was re-

ported as 10 mm (Ehrich et al., 2000), we dened the non-in-

feriority margin of the difference between PG201 and SK306X as

10 mm. The sample size needed for the non-inferiority was 49 per

group, which would provide a statistical power of 80% at a sig-

nicance level of 0.05. Factoring in the 20% dropout rate, 124

subjects were selected, and 62 were assigned to each group.

2.7. Statistical analysis

Patients who received at least one dose of study medication

were analyzed as Intention-To-Treat (ITT) population. Per Protocol

(PP) population was dened as the randomized patients who

completed the study without major protocol violation (Fig. 1). The

demographic and safety data were analyzed using the ITT popu-

lation. Efcacy analyses were performed using the ITT population

and PP population. Missing data were processed using the LastObservation Carried Forward Analysis. After testing the normal

distribution status of the data for the efcacy evaluation variables,

the two groups were compared using the independentt-test or the

Wilcoxons rank sum test, and were tested using the analysis of

covariance (ANCOVA), applying the rescue medicine usage status

as a covariate. The non-inferiority of the experimental group was

tested against the control group when the conrmation of two-

sided 95% condence interval(CI) showed that the lower limit of

the CI was higher than the permitted limit (10 mm). Categorical

variables were comparatively tested using the Chi-square test or

Fishers exact test.

3. Results

3.1. Patient characteristics

A total of 133 patients were screened, and 124 were included in

the ITT analysis patient group. Among the 124 patients, 22 patients

were eventually excluded: 9 who withdrew their consent after

registration, 9 who displayed adverse events, 1 who failed to fol-

low up, 1 who violated the protocol, and 2 others. Of the 102 pa-

tients who completed the clinical trial, 5 were outside the visit

window and 1 showed less than 70% drug adherence, which met

the PP exclusion criteria resulting in a total of 96 patients for the

PP analysis subject group. Fifty-two patients were included in the

experimental group and 44 in the control group (Fig. 1). Group

sample sizes of 52 and 44 achieved 72% power to detect non-in-

feriority. No signicant differences were observed between thetwo groups with regards to the age, sex, height, body weight, stage

of arthritis (K-L grade), pain VAS, and the WOMAC score at base-

line (Table 1). Compliance with prescribed treatment was good:

the proportion of patients who took over 70% of medication was

93.7% in experimental group and 91.8% in the control group

(p0.7415).

3.2. Efcacy

The 100 mm pain VAS showed an improvement of 14.2716.2

and 11.9713.1 mm in the experimental and control groups, re-

spectively, determined by PP analysis at week 8 compared to the

baseline (po0.001 in experimental and control group), with no

signicant statistical difference between the two groups(p0.557). In addition, the lower limit (8.38) of the 97.5% one-

sided CI (two-sided 95% CI) of the experimental group improve-

ment level-control group improvement levelwas greater than the

allowed limit (-10 mm), conrming that the experimental group

was not inferior to the control group (Table 2and Fig. 2). The ITT

analysis also revealed no signicant differences in the average

improvement level of 100 mm pain VAS at week 8 between the

two groups (p0.318), with the lower limit of the experimental

group improvement level-control group improvement levelbeing

8.71 mm, demonstrating non-inferiority (SupplementalTable 2).

The average improvement in 100 mm pain VAS at week 12

determined by PP analysis was 18.8717.61 and 17.1715.7 mm in

the experimental and control groups, respectively, showing no

signicant difference (p0.794, Table 2). In addition, the ITT

Table 1

Baseline characteristics.

Experimental group

PG201 (n63)

Control group

SKI306X (n61)

p-value

Age (years) 63.877.2 65.477.0 0.2021

Gender

Male 8(12.7) 7(11.5)

Female 55(87.3) 54(88.5) 0.8352

Height (cm) 155.575.8 156.476.7 0.4351

Body weight (kg) 61.177.6 62.378.8 0.3961

Location target knee

Left 40(63.5) 35(57.4)

Right 23(36.5) 26(42.6) 0.4862

Other medical

history

49(77.8) 47(77.1) 0.9232

Kellgren-Lawrence

scale

Grade 2 30(47.6) 23(37.7)

Grade 3 33(52.4) 38(62.3) 0.2403

100 mm pain VAS 56.479.8 56.7710.9 0.4714

WOMAC Index

To tal score 35.0714.8 36.7715.4 0.5381

Pain subscale

score

7.073.0 7.573.1 0.3254)

Stiffness subscale

score

2.971.8 3.171.7 0.6604

Function subscale

score

25.1711.5 26.2711.6 0.5511

Note: Data are mean7SD or n(%).1 Independent t-test2 Chi-square test3 Fishers exact test4 Wilcoxons rank sum test

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 174

7/25/2019 ha2016

5/7

analysis also showed no signicant difference between the two

groups (p0.601, SupplementalTable 2). The average rate of im-

provement of 100 mm pain VAS in the experimental and control

groups compared to the baseline determined by the PP analysis

was 25.5730.7 and 20.3722.5% at week 8 (p0.450, Table 2),

and 33.2732.3 and 28.9726.3% at week 12, respectively

(p0.617; Table 2), showing no signicant difference. The ITT

analysis also showed no signicant differences in the average rate

of improvement of 100 mm pain VAS at week 8 or 12 between the

experimental and control groups (p0.204 and 0.414,

respectively, SupplementalTable 2). The average improvement in

the WOMAC total scores in the experimental and control groups

compared to the baseline determined by the PP analysis were

8.9711.3 and 8.979.8 at week 8, and 10.9712.2 and 13.779.3 at

week 12, respectively, showing no statistically signicant differ-

ence (p0.991 and 0.190, respectively, Table 2). The ITT analysis

also showed no signicant difference in the average improvement

in WOMAC total score at week 8 or 12 between the two groups

(p0.665 and 0.542, respectively, Supplemental Table 2).The average improvement in quality of life (EQ-5D) in the ex-

perimental and control groups compared to the baseline de-

termined by the PP analysis was 0.0470.13 and 0.0470.09 at

week 8, and 0.0570.14 and 0.0670.10 at week 12, respectively,

showing no signicant difference (p0.780 and 0.528, respec-

tively, Table 2). The ITT analysis also showed no signicant dif-

ference in the average improvement of quality of life at week 8 or

12 between the two groups (p0.740 and 0.959, respectively,

SupplementalTable 2). The average improvement in the self-as-

sessed overall symptom score on the effects of knee arthritis on

overall symptoms over the last 24 hours in the experimental and

control groups compared to the baseline values determined by the

PP analysis was 0.5470.75 and 0.4870.76 at week 8 (p0.826)

and 0.6970.98 and 0.7070.82 at week 12 (p0.766), respec-

tively, showing no signicant difference (Table 2). The ITT analysis

also revealed no signicant differences in the average improve-

ment of the self-assessed overall symptom score at week 8 or 12

between the two groups (p0.258 and 0.422, respectively, Sup-

plementalTable 2). The dose of rescue medicine administered in

the experimental and control groups compared to the baseline

determined by the PP analysis was increased by 7.4733.5 and

8.6734.0 tablets at week 8, and 12 8.6730.5 and 9.9737.6 ta-

blets at week 12, respectively (p0.845 and 0.840), showing no

signicant difference (Table 2). The ITT analysis also showed no

statistically signicant differences in the dose of rescue medicine

between the two groups (p0.693 and 0.542, respectively, Sup-

plementalTable 2).

3.3. Safety

All the observed AEs occurred in 12 (19.1%) and 19 (31.2%)

subjects in the experimental and control groups, respectively in

the ITT analysis group; 47 AEs were recorded in 31 subjects (Ta-

ble 3). Serious AEs were observed in 2 subjects (2 cases) in the

control group, which were 1 gastrointestinal disorder and 1 mus-

culoskeletal and connective tissue disorders. Treatment-emergent

AEs (TEAEs) occurred in 6 subjects (7 cases) in the experimental

group and 9 subjects (12 cases) in the control group. No signicant

difference was observed in the AE incidence between the groups

(p0.120,Table 3). Among the categories of vital signs, the systolic

blood pressure increased in the experimental group but decreased

in the control group at week 8 (1.57710.51 vs. 2.59710.21,

p0.034) and 12 (1.48712.0 vs. 2.61711.49, p0.033). Thisdifference between the two groups was statistically signicant,

although it did not appear to be clinically signicant. No signicant

differences were observed in any of the other vital sign categories.

In addition, no signicant abnormal result was observed upon

physical examination and in laboratory test results including blood

tests.

4. Discussion

This prospective, randomized, double-blinded, double-dummy,

multicenter, and comparative study showed that the efcacy and

safety of PG201 were comparable to that of SKI306X in patients

with knee OA. During the 12-week follow-up period, the pain VAS,

Table 2

Primary and Secondary efcacy outcomes in per-protocol population.

PP group

Experimental

(n52)

Control(n44) p-value1

Primary efcacy outcome

100 mm Pain VAS

(mean7SD)Baseline 57.679.1 58.179.5 0.746

Weeks 8 43.4719.0 46.3715.0 0.557

Change 14.2716.2 11.9713.1

Difference between groups (95%

CI)

2.3 (8.38, 3.70)

Secondary efcacy outcomes

100 mm Pain VAS

(mean7SD)

Baseline 57.679.1 58.179.5 0.746

Weeks 12 38.9719.5 41.0716.2 0.794

Rate of 100 mm Pain VAS

change from baseline (%)

Weeks 8 25.5730.7 20.3722.5 0.450

Weeks 12 33.2732.3 28.9726.3 0.617

WOMAC Total score

(mean7SD)

Baseline 35.27

14.9 36.67

14.7 0.383Weeks 8 26.3716.9 27.7715.2 0.991

Weeks 12 24.4716.5 22.9714.0 0.190

Quality of life(EQ-5D)

(mean7SD)

Baseline 0.7370.11 0.7270.14 0.740

Weeks 8 0.7670.12 0.7670.10 0.780

Weeks 12 0.7770.12 0.7970.11 0.528

Patient assessed overall symp-

tom (mean7SD)

Baseline 3.5470.50 3.6470.61 0.344

Weeks 8 3.0070.84 3.1670.61 0.826

Weeks 12 2.8571.04 2.9370.70 0.766

Rescue medicine consumption

(mean7SD)

Baseline 13.7722.7 14.5724.1 0.852

Weeks 8 21.1739.7 23.1736.2 0.845

Weeks 12 22.3735.0 24.4734.9 0.840

1 ANCOVA

Fig. 2. Change from baseline to week 12 in 100 mm pain visual analogue scale.

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 17 5

7/25/2019 ha2016

6/7

WOMAC, quality of life (EQ-5D), and overall symptom assessment

scores signicantly improved in both groups compared to the

baseline values. However, there was no signicant difference be-

tween the two groups. In addition, there was no difference be-

tween the two groups in AEs and TEAEs in terms of the safety of

the drug.

PG201 is a herb-derived medicine extracted from 12 herbs, and

has been developed using the information on the known function

of each herb and the traditional herbal medicine theories re-

searched from the literature of traditional Chinese and Korean

medicine (Shin et al., 2003). Recently, the components of PG201

have gained scientic evidences in the treatment of osteoarthritis.For example, Gentiana Macrophylla and Acanthopanax sessiliorus

were reported to assist in the functions of both muscles and bones

and Clematis chinensis has been used for protection against de-

generation of cartilage and regeneration of damaged tissue (Park

et al., 2005;Shin et al., 2003). PG201, containing the 12 medicinal

herbal extracts, exhibited anti-inammatory activity by inhibiting

the production of pro-inammatory cytokines such as tumor ne-

crosis factor (TNF-) and interleukin (IL-1), and by upregulatingthe anti-inammatory cytokine IL-4 in in vivo studies (Park et al.,

2005;Shin et al., 2003). In addition, PG201 inhibited the expres-

sion of matrix degradation enzymes including matrix metallo-

proteinase (MMP) (Park et al., 2005). PG201 exhibited anti-in-

ammatory activity by decreasing nitrite production via upregu-

lation of heme oxygenase-1 by regulating phosphatidylinositol3-kinase and nuclear factor, erythroid 2 (NF-E2)-related factor 2

(Choi et al., 2013). In a recent in vitro study, 14 compounds (cou-

marin, demethylsuberosin, xanthotoxin, psoralen, decursinol, de-

cursin, decursinol angelate, chikusetsusaponin IVa, chikusetsusa-

ponin IVa methyl ester, ethyl caffeate, syringaresinol, cnidilide,

farnesol, and linoleic acid) were isolated from phytopharmaceu-

tical PG201 by activity-guided fractionation using inhibitory ac-

tivity on nitric oxide production. All compounds except linoleic

acid showed suppressive effects on nitric oxide and prostaglandin

E2 production (Kim et al., 2015). This study was performed based

on the encouraging results obtained in previous studies showing

the activities and benecial effects of PG201 on symptomatic OA.

In this study, PG201 signicantly improved pain and function in

patients with OA. For the primary efcacy parameter, both groups

showed signicant improvement in pain VAS without signicant

difference between the two groups. The secondary efcacy para-

meters also showed signicant improvement without signicant

difference between the two groups. In a study comparing SKI306X

activity to that of diclofenac, the daily administration of 600 mg of

SKI306X over 4 weeks signicantly improved the pain symptoms;

and the effect of SKI306X was not different from that of diclofenac

(Lung et al., 2004). An 8-week study comparing PG201 to celecoxib

revealed that treatment with 600 mg of PG201 signicantly im-proved pain symptoms; and it was not different from the effect of

celecoxib (Yoo et al., 2014). In this study, the efcacy of PG201 was

compared with that of SKI306X, which is known to have clinical

effects and in relatively wide usage among herb-derived medicine

in patients with knee OA (Jung et al., 2001;Lung et al., 2004). The

results of this study showed no signicant differences between

PG201 and SKI306X. Therefore, PG201 could be an effective agent

in the treatment of patients with symptomatic knee OA.

In the safety analysis, the incidence of AEs and TEAEs in the

PG201 group was lower than that observed in the SKI306X group

(19.1 vs. 31.2%, 9.5 vs. 14.8%, respectively). However, this difference

was not statistically signicant (p0.120; p0.372). During this

clinical trial, 2 cases of serious AEs were observed in 2 subjects in

the SKI306X group while no serious AEs were observed in thePG201 group. The previous study comparing SKI306X and diclo-

fenac showed a lower incidence of AEs in SKI306X group (29.6 vs.

34.7%, p0.390), which was not statistically signicant, but the

incidence of TEAEs was signicantly low in SKI306X group (17.6 vs.

29.0%, p0.033) (Lung et al., 2004). The previous study comparing

celecoxib and PG201 showed higher incidence of AEs and adverse

drug reactions(ADRs) in patients treated with PG201 than in those

treated with celecoxib but this difference was not statistically

signicant (28.6 vs. 26.5%, p0.677 in AEs and 14.9 vs 13.6%,

p0.727 in ADRs) (Yoo et al., 2014). Although a direct comparison

is difcult owing to differences in the study designs, these results

suggest that PG201 can be safely used in the treatment of patients

with knee osteoarthritis in terms of gastrointestinal adverse

reactions.

There were some limitations to this study. Firstly, this study

was designed to demonstrate the non-inferiority of PG201 to

SKI306X in patients with OA. Although the efcacy of SKI306X was

assessed by randomized controlled trials (Jung et al., 2001;Lung

et al., 2004), further high-quality, fully powered studies must be

conducted to conrm the efcacy of herbal medicinal plant pro-

ducts (Cameron and Chrubasik, 2014). This study was a well-de-

signed, double-blinded, double dummy, randomized, and con-

trolled study with sufcient number of patients in the experi-

mental and control groups, according to sample size calculation.

Therefore, the results of this study could be considered reliable.

Secondly, no objective efcacy parameter such as X-ray or mag-

netic resonance imaging was used in this study. However, the goal

of the administration of the oral medication was to provide pain

relief and effect functional improvement in patients with OA.

Therefore, we believe the composite of several efcacy parameters

in this study should be enough to assess the comparative efcacies

of PG201 and SKI306X.

5. Conclusions

This study showed that PG201 has effects on reducing pain and

improving function of the knee in patients with OA. The results

were comparable to SKI306X, a widely used herb-derived medi-

cine in Korea. PG201 can be suggested as an effective treatment for

knee OA.

Table 3

Summary of safety including severity and incidence of adverse events.

Experimental group

PG201 (n63)

Control group

SKI306X

(n61)

p-value1

Patient of AEs 12 (19.1) 19 (31.2) 0.120

Adverse events (cases) 19 (30.2) 28 (45.9)

Grade 1(mild) 17 (89.5) 19 (67.9)

Grade 2(moderate) 2 (10.5) 4 (14.3)Grade 3(severe) 0 (0.0) 5 ( 17.9)

Grade 4 0 (0.0) 0 (0.0)

Grade 5 0 (0.0) 0 (0.0)

Serious AEs (case) 0 (0.0) 2 (7.1)

Patient of TEAEs 6 (9.5) 9 (14.8) 0.372

TEAEs (case) 7 (11.1) 12 (19.7)

Constipation 0 (0.0) 2 (3.3)

Diarrhea 2 (3.2) 0 (0.0)

Dyspepsia 1 (1.6) 5 (8.2)

Dyspnoea 0 (0.0) 2 (3.3)

Face oedema 1 (1.6) 0 (0.0)

Headache 0 (0.0) 1 (1.6)

Insomnia 1 (1.6) 0 (0.0)

Nausea 0 (0.0) 1 (1.6)

Tongue paralysis 0 (0.0) 1 (1.6)

Urticaria 2 (3.2) 0 (0.0)

Data are n(%).1 Chi-square test

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 176

7/25/2019 ha2016

7/7

Role of the funding source

This study was funded by PMG Pharm Co., Ltd, South Korea. The

sponsor did not play any role in the collection, analysis, and in-

terpretation of data, or in the writing of the report and decision to

submit the paper for publication.

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

YBP and CWH analyzed the data and wrote the manuscript.

CWH, BWM, SBH, JYL, YYW, and YSP obtained funding and de-

signed the study. CWH, BWM, SBH, JYL, YYW, and YSP recruited

the participants and conducted the trial. All authors read and

approved the manuscript.

Acknowledgements

All authors thank the study coordinators at the participating

centers for their efforts in collecting clinical data and ensuring the

accuracy and completeness of the data.

Appendix A. Supplementary material

Supplementary data associated with this article can be found in

the online version athttp://dx.doi.org/10.1016/j.jep.2016.01.029.

References

Altman, R., Asch, E., Bloch, D., Bole, G., Borenstein, D., Brandt, K., Christy, W., Cooke,T.D., Greenwald, R., Hochberg, M., et al., 1986. Development of criteria for theclassication and reporting of osteoarthritis. Classication of osteoarthritis ofthe knee. Diagnostic and Therapeutic Criteria Committee of the AmericanRheumatism Association. Arthritis Rheum 29, 10391049.

Cameron, M., Chrubasik, S., 2014. Oral herbal therapies for treating osteoarthritis.Cochrane Database Syst Rev 5, CD002947.

Cameron, M., Gagnier, J.J., Little, C.V., Parsons, T.J., Blumle, A., Chrubasik, S., 2009.Evidence of effectiveness of herbal medicinal products in the treatment of ar-thritis. Part I: Osteoarthritis. Phytother Res 23, 14971515.

Chen, F.P., Chang, C.M., Hwang, S.J., Chen, Y.C., Chen, F.J., 2014. Chinese herbalprescriptions for osteoarthritis in Taiwan: analysis of National Health Insurancedataset. BMC Complement Altern Med 14, 91.

Choi, J.H., Choi, J.H., Kim, D.Y., Yoon, J.H., Youn, H.Y., Yi, J.B., Rhee, H.I., Ryu, K.H.,Jung, K., Han, C.K., Kwak, W.J., Cho, Y.B., 2002. Eff ects of SKI 306X, a new herbalagent, on proteoglycan degradation in cartilage explant culture and col-lagenase-induced rabbit osteoarthritis model. Osteoarthritis Cartilage 10,471478.

Choi., J., Kim., S.H., Kim., S., 2012. Suppressive effects of PG201, an antiarthriticbotanical formulation, on lipopolysaccharide-induced inammatory mediatorsin Raw264.7 cells. Exp Biol Med (Maywood) 237 (5), 499508.

Choi, J., Lee, J., Lee, J., Kim, S.H., Kim, J., Kim, S., 2013. PG201 downregulates theproduction of nitrite by upregulating heme oxygenase-1 expression throughthe control of phosphatidylinositol 3-kinase and NF-E2-related factor 2. NitricOxide 33, 4255.

Ehrich, E.W., Davies, G.M., Watson, D.J., Bolognese, J.A., Seidenberg, B.C., Bellamy, N.,2000. Minimal perceptible clinical improvement with the Western Ontario and

McMaster Universities osteoarthritis index questionnaire and global assess-ments in patients with osteoarthritis. J Rheumatol 27, 26352641.

Hartog, A., Hougee, S., Faber, J., Sanders, A., Zuurman, C., Smit, H.F., van der Kraan, P.M., Hoijer, M.A., Garssen, J., 2008. The multicomponent phytopharmaceutical

SKI306X inhibits in vitro cartilage degradation and the production of in-

ammatory mediators. Phytomedicine 15, 313320.Health, U.D.o., Services, H. National Cancer Institute. Common Terminology Criteria

for Adverse Events (CTCAE). 28 May 2009 (v4. 03: 14 June 2010). http.evs.nci.nih.gov/ftp1/CTCAE/About.html (October 2010, date last accessed).

Heo, J., (1613 A.D.). 1999. Dong Ui Bo Gam [The precious mirror of oriental medi-cine]. Namsandang. Seoul.

Hochberg, M.C., 2006. Nutritional supplements for knee osteoarthritisstill no re-

solution. N Engl J Med 354, 858860.Jordan, K.M., Arden, N.K., Doherty, M., Bannwarth, B., Bijlsma, J.W., Dieppe, P.,

Gunther, K., Hauselmann, H., Herrero-Beaumont, G., Kaklamanis, P., Lohman-

der, S., Leeb, B., Lequesne, M., Mazieres, B., Martin-Mola, E., Pavelka, K., Pen-dleton, A., Punzi, L., Serni, U., Swoboda, B., Verbruggen, G., Zimmerman-Gorska,I., Dougados, M., 2003. EULAR Recommendations 2003: an evidence based

approach to the management of knee osteoarthritis: Report of a Task Force ofthe Standing Committee for International Clinical Studies Including TherapeuticTrials (ESCISIT). Ann Rheum Dis 62, 11451155.

Jung, Y.B., Roh, K.J., Jung, J.A., Jung, K., Yoo, H., Cho, Y.B., Kwak, W.J., Kim, D.K., Kim, K.H., Han, C.K., 2001. Effect of SKI 306X, a new herbal anti-arthritic agent, inpatients with osteoarthritis of the knee: a double-blind placebo controlled

study. Am J Chin Med 29, 485491.Kim, H.C., 2000. Herbalogy. Young Lim Sa, Seoul .Kim, H.J., Kim, H.M., Ryu, B., Lee, W.S., Shin, J.S., Lee, K.T., Jang, D.S., 2015. Con-

stituents of PG201 (Layla), a multi-component phytopharmaceutical, with in-hibitory activity on LPS-induced nitric oxide and prostaglandin E productions in

macrophages. Arch Pharm Res.Kim, J.H., Ryu, K.H., Jung, K.W., Han, C.K., Kwak, W.J., Cho, Y.B., 2005. SKI306X

suppresses cartilage destruction and inhibits the production of matrix me-

talloproteinase in rabbit joint cartilage explant culture. J Pharmacol Sci 98,298306.

Lee, S., (1590 A.D.). 1978. Bon Cho Gang Mok [Compendium of material medica].ChungKuK ChungEuiHak. Peking.

Lung, Y.B., Seong, S.C., Lee, M.C., Shin, Y.U., Kim, D.H., Kim, J.M., Jung, Y.K., Ahn, J.H.,Seo, J.G., Park, Y.S., Lee, C.S., Roh, K.J., Han, C.K., Cho, Y.B., Chang, D.Y., Kwak, W.J.,

Jung, K.O., Park, B.J., 2004. A four-week, randomized, double-blind trial of theefcacy and safety of SKI306X: a herbal anti-arthritic agent versus diclofenac in

osteoarthritis of the knee. Am J Chin Med 32, 291 301.Neung, I.G., 1998. Herbalogy.. Sanitation Press, SeoulPark, J.H., Lee, C.H., 2000. The encyclopedia of medicinal plants. SinIl Sangsa, Seoul.Park, K.C., Park, E.J., Kim, E.R., Kim, Y., Chung, S.H., Cho, B.W., Kim, S., Jin, M., 2005.

Therapeutic effects of PG201, an ethanol extract from herbs, through cartilageprotection on collagenase-induced arthritis in rabbits. Biochem Biophys Res

Commun 331, 14691477.

Shin, S.S., Jin, M., Jung, H.J., Kim, B., Jeon, H., Choi, J.J., Kim, J.M., Cho, B.W., Chung, S.H., Lee, Y.W., Song, Y.W., Kim, S., 2003. Suppressive effects of PG201, an ethanol

extract from herbs, on collagen-induced arthritis in mice. Rheumatology (Ox-ford) 42, 665672.

Wandel, S., Juni, P., Tendal, B., Nuesch, E., Villiger, P.M., Welton, N.J., Reichenbach, S.,Trelle, S., 2010. Effects of glucosamine, chondroitin, or placebo in patients withosteoarthritis of hip or knee: network meta-analysis. BMJ 341, c4675.

Yoo, W.H., Yoo, H.G., Park, S.H., Baek, H.J., Lee, Y.J., Shim, S.C., Kang, S.W., Kim, H.A.,

Song, J.S., Suh, C.H., Choi, S.J., Yoon, B.Y., Tae, D.N., Ko, H.S., Song, Y.W., 2014.Efcacy and safety of PG201 (Layla) and celecoxib in the treatment of symp-

tomatic knee osteoarthritis: a double-blinded, randomized, multi-center, activedrug comparative, parallel-group, non-inferiority, phase III study. RheumatolInt.

Zhang, W., Doherty, M., Arden, N., Bannwarth, B., Bijlsma, J., Gunther, K.P., Hau-selmann, H.J., Herrero-Beaumont, G., Jordan, K., Kaklamanis, P., Leeb, B., Le-quesne, M., Lohmander, S., Mazieres, B., Martin-Mola, E., Pavelka, K., Pendleton,

A., Punzi, L., Swoboda, B., Varatojo, R., Verbruggen, G., Zimmermann-Gorska, I.,Dougados, M., 2005. EULAR evidence based recommendations for the man-

agement of hip osteoarthritis: report of a task force of the EULAR StandingCommittee for International Clinical Studies Including Therapeutics (ESCISIT).Ann Rheum Dis 64, 669681.

Zhang, W., Moskowitz, R.W., Nuki, G., Abramson, S., Altman, R.D., Arden, N., Bierma-Zeinstra, S., Brandt, K.D., Croft, P., Doherty, M., Dougados, M., Hochberg, M.,

Hunter, D.J., Kwoh, K., Lohmander, L.S., Tugwell, P., 2008. OARSI re-commendations for the management of hip and knee osteoarthritis, Part II:

OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage 16,137162.

C.-W. Ha et al. / Journal of Ethnopharmacology 181 (2016) 17 7

http://dx.doi.org/10.1016/j.jep.2016.01.029http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref2http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref2http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref2http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref13http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref13http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref17http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref18http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref18http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref24http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref23http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref22http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref21http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref20http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref19http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref18http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref17http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref16http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref15http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref14http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref13http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref12http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref11http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref10http://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://http.evs.nci.nih.gov/ftp1/CTCAE/About.htmlhttp://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref9http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref8http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref7http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref6http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref5http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref4http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref3http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref2http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref2http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://refhub.elsevier.com/S0378-8741(16)30029-0/sbref1http://dx.doi.org/10.1016/j.jep.2016.01.029