Cross-clade immunogenicity and safety of an AS03A-adjuvanted prepandemic H5N1 influenza vaccine in Hong Kong.

Abstract

To present Hong Kong'Äìspecific data from a large Asian population (also involving Thailand, Singapore, and Taiwan) on safety and manufacturing consistency across four AS03(A)-adjuvanted H5N1 vaccine formulations in terms of immune response against the A/Vietnam/1194/2004 strain. Immunogenicity against the heterologous A/Indonesia/05/2005 strain was also assessed. NCT Number: 00449670.
Prospective, observer-blind study.
Out-patient clinic of a tertiary hospital in Hong Kong.
A total of 360 subjects aged 18 to 60 years were randomised into six groups to receive two doses (21 days apart) of the study vaccine.
One of the four adjuvanted formulations (3.75 microgram H5N1 haemagglutinin [HA]+AS03(A)) of the vaccine (H5N1-AS03(A)) or one of the two non-adjuvanted (3.75 microgram H5N1 [HA]) formulations of the vaccine (H5N1-DIL).
Blood samples collected before vaccination and 21 days after each vaccine dose were analysed using haemagglutination-inhibition and neutralisation assays. Solicited, unsolicited, and serious adverse events were recorded.
Manufacturing consistency across all four vaccine formulations was demonstrated. After two doses, the AS03(A)-adjuvanted prepandemic influenza vaccine demonstrated high seroprotection rates against the A/Vietnam/1194/2004 strain (95.8%) and good immunogenicity against the heterologous A/Indonesia/05/2005 strain (45.7%), as compared to the non-adjuvanted vaccine (4.6% and 1.5%, respectively). The seroconversion rates induced by the adjuvanted formulations in terms of viral neutralising antibodies against the two strains were much higher than those induced by the non-adjuvanted formulations. There were no safety concerns for any of the adjuvanted vaccine formulations.
The AS03(A)-adjuvanted H5N1 prepandemic influenza vaccine demonstrated good immunogenicity and an acceptable safety profile in Hong Kong.

Full text

Page 1

?Hong?Kong?Med?J??Vol?17?No?1?#?February?2011?#??www.hkmj.org?
39
Objective To present Hong Kong–specific data from a large Asian population
(also involving Thailand, Singapore, and Taiwan) on safety and
manufacturing consistency across four AS03A-adjuvanted H5N1
vaccine formulations in terms of immune response against
the A/Vietnam/1194/2004 strain. Immunogenicity against the
heterologous A/Indonesia/05/2005 strain was also assessed.
NCT Number: 00449670.
Design Prospective, observer-blind study.
Setting Out-patient clinic of a tertiary hospital in Hong Kong.
Participants A total of 360 subjects aged 18 to 60 years were randomised into six
groups to receive two doses (21 days apart) of the study vaccine.
Interventions One of the four adjuvanted formulations (3.75 µg H5N1
haemagglutinin [HA]+AS03A) of the vaccine (H5N1-AS03A) or one
of the two non-adjuvanted (3.75 µg H5N1 [HA]) formulations of
the vaccine (H5N1-DIL).
Main outcome measures Blood samples collected before vaccination and 21 days after
each vaccine dose were analysed using haemagglutination-
inhibition and neutralisation assays. Solicited, unsolicited, and
serious adverse events were recorded.
Results Manufacturing consistency across all four vaccine formulations
was demonstrated. After two doses, the AS03A-adjuvanted
prepandemic influenza
seroprotection rates against the A/Vietnam/1194/2004 strain
(95.8%) and good immunogenicity against the heterologous
A/Indonesia/05/2005 strain (45.7%), as compared to the
non-adjuvanted vaccine (4.6% and 1.5%, respectively). The
seroconversion rates induced by the adjuvanted formulations
in terms of viral neutralising antibodies against the two strains
were much higher than those induced by the non-adjuvanted
formulations. There were no safety concerns for any of the
adjuvanted vaccine formulations.
vaccine demonstrated high
Conclusions The AS03A-adjuvanted H5N1 prepandemic influenza vaccine
demonstrated good immunogenicity and an acceptable safety
profile in Hong Kong.
Cross-clade immunogenicity and safety of an AS03A-
adjuvanted prepandemic H5N1 influenza vaccine in
Hong Kong
O
A
R
R
IGIN
C
A
L
L
ETI
Key words
Adjuvants, immunologic; Disease
outbreaks; Influenza A virus, H5N1
subtype; Influenza vaccines; Influenza,
human
Hong Kong Med J 2011;17:39-46
Department of Medicine, Queen Mary
Hospital, 102 Pokfulam Road,
Hong Kong
DWS Chu, FHKCFP, FHKAM (Family Medicine)
Sai Ying Pun Jockey Club General
Outpatient Clinic, Hong Kong
ASK Kwong, FHKCFP, FHKAM (Family Medicine)
WWS Tsui, FHKCFP, FHKAM (Family Medicine)
JHL Wang, FHKCFP, FHKAM (Family Medicine)
CKH Ngai, FHKCFP, FHKAM (Family Medicine)
PKT Wan, FHKCFP, FHKAM (Family Medicine)
GlaxoSmithKline Biologicals, Wavre,
Belgium
G Ong, MB, BS, MMed (Public Health)
HW Tang, MD, PhD
F Roman, MD, PhD
M Dramé, MSc
HL Bock, MD
Correspondence to: Dr DWS Chu
Email: chuwsd@ha.org.hk
Daniel WS Chu
Alfred SK Kwong
Wendy WS Tsui
Jenny HL Wang
Charles KH Ngai
Peter KT Wan
Gary Ong
HW Tang
François Roman
Mamadou Dramé
Hans L Bock
朱偉星
鄺兆基
徐詠詩
王華力
魏家豪
溫敬東
王邦耀
唐海文
Introduction
Hong Kong reported the first-ever instance of human infection with H5N1 virus in the
world in 1997. This was followed by 17 other cases in the same year, which resulted in six
deaths. In 2003, two new H5N1 human cases were recorded. Since 2004, the H5N1 virus has
been isolated in wild birds, and in addition cross-species transmission was observed from
tigers and cats; however, no new H5N1 epidemic has ensued.1 As the H5N1 influenza virus
continues to circulate and cause sporadic human infections, there is a high probability of
Hong Kong being vulnerable to future influenza pandemics.1-4

a pandemic outbreak is considered the most effective intervention to counter morbidity
and mortality caused by an influenza pandemic.5-7 The current strategy for prepandemic
influenza vaccine development is 2-fold: cross-clade protection and antigen-sparing.5,8
Vaccination with a prepandemic influenza vaccine just before or immediately after

Page 2

??#??Chu?et?al?#
40?
Hong?Kong?Med?J??Vol?17?No?1?#?February?2011?#??www.hkmj.org
Methods
Study design and subjects
This
conducted across Singapore, Thailand, Hong Kong,
and Taiwan between 24 March and 12 July 2007.
The study design was observer-blind due to the
difference in appearance of the AS03A-adjuvanted
and non-adjuvanted vaccine. Healthy subjects 18 to
60 years of age were randomised (allocation ratio
2:2:2:2:1:1) into six groups to receive two doses (21
days apart) of either one of the four formulations of
the AS03A-adjuvanted vaccine (group H5N1-AS03A) or
one of the two formulations of the non-adjuvanted
vaccine (group H5N1-DIL). The procedure used
to derive the adjuvanted and non-adjuvanted
formulations is described below under the sub-
section describing the vaccine. In Hong Kong, the
study was conducted at the general out-patient
clinic of Queen Mary Hospital according to Good
Clinical Practice guidelines. All necessary approvals
were obtained from the Institutional Review Board
of all participating institutions and the local health
authorities. Study procedures were performed after
obtaining written consent from the subjects.
multi-centre study (NCT00449670) was

any licensed inactivated vaccine within 2 weeks
or live-attenuated vaccine within 4 weeks prior to
enrolment in the study. They were also excluded
if they had previously received any prepandemic
candidate vaccine or any AS03A-adjuvanted vaccine,
or had been previously exposed to any H5N1 wild-
type virus. Subjects with a history of hypersensitivity
to vaccine components and chronic illnesses were
also not considered for the study. The study excluded
subjects presenting symptoms of acute disease
(defined as moderate or severe illness with or without
febrile illness [axillary temperature ≥37.5ºC]) at the
time of enrolment. If a subject presented with acute
disease (including febrile illness) at the time of the
second dose, that dose was not administered, but he/
she could continue with other study procedures (at
the discretion of the investigator) and was monitored
till resolution of the medical condition.
Subjects were excluded if they had received
Vaccines
The study vaccine and adjuvant were manufactured
by GSK Biologicals as described previously.9 The
adjuvanted vaccine (0.5 mL) [Prepandrix] contained
3.75 µg HA of the A/Vietnam/1194/2004-like NIBRG-
14 Clade 1 strain (National Institute for Biological
Standards and Control Potters Bar, UK) adjuvanted
with AS03A (a tocopherol [11.86 mg] oil-in-water
emulsion-based Adjuvant System), while the non-
adjuvanted vaccine had the same constituents except
AS03A. Two lots of HA antigen were combined with
two lots of AS03A Adjuvant System to obtain four
  目的  報告四種AS03A佐劑疫苗在亞洲人口（香港、泰國、
星加坡、台灣）中的安全性及製造穩定性。研究尤?
其針對香港人口對於流感病毒?A/Vietnam/1194/2004
品種的免疫反應，也探討對流感病毒? A/Indonesia/05/
2005異種的免疫原性（NCT編號：00449670）。
  設計  前瞻性觀察員單盲研究。
 安排  香港一所教學醫院的門診診所。
 參與者  共360名18至60歲的成年人共分成6組，分別在相隔
21天的時間接種兩個劑量的疫苗。
 介入治療  參與者接種四種佐劑疫苗（3.75?
+AS03A）的其中一種，或兩種非佐劑疫苗（3.75? µg?
H5N1血凝素）的其中一種。
µg?
H5N1血凝素
  主要結果測量  利用血凝抑制及中和試劑分析參與者在接種前及接種
後第21天的血液樣本。所有不良反應，不論是否預見
的及嚴重的，都會被記錄下來?。
  結果  四種佐劑疫苗都具製造穩定性。參與者在接種兩個劑
量的疫苗後，AS03A佐劑組對流感病毒? A/Vietnam/1194/
2004品種有高的血清保護率（95.8%），並對流感病
毒A/Indonesia/05/2005異種具高免疫原性（45.7%）
；相對地，非佐劑疫苗組只有分別4.6%的血清保護率
及1.5%免疫原性。病毒中和抗體方面，AS03A佐劑疫
苗組比非佐劑疫苗組產生相對較高的血清保護率。所
有佐劑疫苗均無安全問題。
 結論  流感大流行的前期疫苗AS03A佐劑對這項香港研究的
參與者有良好的免疫原性及可接受的安全度。
香港流感大流行的前期疫苗AS03A佐劑的
交叉分化枝抗原性及安全性
An AS03A-adjuvanted split-virion prepandemic H5N1
influenza vaccine (Prepandrix; 3.75 µg haemagglutinin
[HA] antigen; AS03A: a tocopherol [11.86 mg] oil-in-
water emulsion-based Adjuvant System) developed
by GlaxoSmithKline (GSK) Biologicals has been found
to be immunogenic with an acceptable safety profile
in European populations and has also demonstrated
cross-clade immunogenicity.9-11

Asian population in Hong Kong, Thailand, Singapore,
and Taiwan to demonstrate
consistency in terms of the vaccine-homologous
immune response [A/Vietnam/1194/2004 NIBRG-
14 (clade 1)] across the four adjuvanted vaccine
formulations as well as from the vaccine’s safety
profile. In addition, the cross-clade immunogenicity
of the vaccine against a heterologous strain [A/
Indonesia/05/2005 IBCDC-RG-2 (clade 2.1)] was also
assessed.
The present study was conducted in a large
manufacturing

presented previously.12 This paper presents Hong
Kong–specific primary vaccination data on the
immunogenicity of the AS03A-adjuvanted vaccine.
Overall data from this study have been

Page 3

#??Prepandemic?H5N1?influenza?vaccine?#
?Hong?Kong?Med?J??Vol?17?No?1?#?February?2011?#??www.hkmj.org?
41
adjuvanted vaccine formulations, while the same
two lots of HA antigen were combined with a diluent
to obtain the two controls. The vaccine doses were
administered as intramuscular injections in the
deltoid region of the non-dominant arm.
Laboratory assays
Haemagglutination inhibition (HI) antibody titres
against the homologous strain (A/Vietnam/1194/2004)
and a heterologous strain (A/Indonesia/05/2005) were
assessed using standard assay methods (cut-off for
HI=1:10) modified to utilise an equine erythrocyte
suspension instead of
suspension, as described previously.9
an avian erythrocyte

antibody titres were assessed in a small subset of
subjects, using a standard procedure described
previously, the lowest dilution tested being 1:28.12 The
neutralisation assay was expressed as the reciprocal
of the highest dilution that achieved at least 50%
neutralisation of virus growth.
As part of the overall study, viral neutralising
Assessment of immunogenicity
Blood samples were collected before vaccination
and 21 days after each of the two vaccine doses.
The seroprotection rates, seroconversion rates, and
seroconversion factors for HI antibodies for the
A/Vietnam/1194/2004 and A/Indonesia/05/2005 strains
were calculated with their respective 95% confidence
intervals (CIs).

of subjects the neutralising antibody titres for the
A/Vietnam/1194/2004 and A/Indonesia/05/2005 strains
were determined before vaccination and 21 days
after the second vaccine dose.
As part of the overall assessment, in a subset

protection against pandemic influenza strains, the
Committee for Medicinal Products for Human Use
(CHMP) of the European Medicines Agency (EMA)
currently recommends that prepandemic candidate
vaccines should meet all three of the current
immunological criteria for adults aged 18 to 60 years
set for licensure of seasonal influenza vaccines (ie
seroconversion rate >40%, geometric mean increase
>2.5, and seroprotection rate >70%).13
In the absence of immunological correlates of
Assessment of safety
Safety and reactogenicity were recorded using
diary cards. Local injection site symptoms (pain,
redness, swelling, induration, and ecchymosis) and
general symptoms (fatigue, fever, headache, myalgia,
shivering, sweating, and arthralgia) were recorded
for 7 days (day 0-6) after each dose. Intensity of all
symptoms was defined using a three-grade scale,
except fever, which was measured on a four-grade
scale. Injection site pain that hindered normal daily
activity was recorded as grade 3, while injection site
redness, swelling, ecchymosis and induration of >100
mm was recorded as grade 3. Fever of >39°C and up
to 40°C was recorded as grade 3 (>40°C, as grade 4),
while other general symptoms (headache, fatigue,
arthalgia, myalgia, shivering, and sweating) that
hindered normal daily activities were recorded to be
of grade 3 intensity.

recorded for 21 days after dose 1 and 30 days after
dose 2. Unsolicited events were classified according
to the Medical Dictionary for Regulatory Activities.
Serious adverse events were recorded during the
whole study period and events occurring until 30
days after dose 2 were logged.
All other
(unsolicited) symptoms were
Statistical analyses
The sample size of the overall study was calculated
to fulfil the primary objective of demonstrating
consistency in terms of immunogenicity across
the four adjuvanted vaccine formulations.12 All
paired vaccine formulations were considered
consistent in terms of immune response against the
A/Vietnam/1194/2004 strain, if the two-sided 95% CIs
for the HI antibody geometric mean titre (GMT) ratios
were within the pre-defined clinical limits (0.5, 2.0).

defined as the percentage of initially seronegative
subjects with a post-vaccination titre of ≥1:40 or
initially seropositive subjects with at least a 4-fold
increase in titre, after each dose. The seroconversion
rate for neutralising antibodies was defined as
at least a 4-fold increase in titre, after each dose.
The seroconversion factor was defined as the
fold increase in serum HI antibody GMTs post-
vaccination compared to that before vaccination. The
seroprotection rate was defined as the percentage of
subjects with a post-vaccination serum HI antibody
titre of ≥1:40.
The seroconversion rate for HI antibodies was

antibodies were calculated with 95% CIs by taking the
anti-log of the mean of the log titre transformations.
Antibody titres below the assay cut-off were given
an arbitrary value of half the cut-off for the purpose
of GMT calculation. Immunogenicity analysis was
performed on the per-protocol cohort, while the
safety and reactogenicity analysis was performed on
the total vaccinated cohort.
Geometric mean titres for HI and neutralising

the Statistical Analyses System (SAS) version 9.1,
while all 95% CIs were calculated using Proc StatXact
5.0.
All statistical analyses were performed using

Page 4

??#??Chu?et?al?#
42?
Hong?Kong?Med?J??Vol?17?No?1?#?February?2011?#??www.hkmj.org
Results
Study population
The study enrolled 1206 subjects of which 360 subjects
were from Hong Kong. Of these, 287 received any
one of the four H5N1-AS03A formulations while
73 received any one of the two non-adjuvanted
formulations. The mean age of subjects enrolled
from Hong Kong was 32.8 years (standard deviation,
9.8 years); the subjects were predominantly male
(77%) and the majority (99%) were of Asian origin.
Immunogenicity
The consistency in the immunogenicity was
demonstrated by all
formulations in terms of HI antibody response against
the A/Vietnam/1194/2004 strain 21 days post-dose 2;
these results having been presented previously.12
Hence, the pooled results from the H5N1-AS03A and
H5N1-DIL groups are presented here.
four adjuvanted-H5N1
Immunogenicity against the homologous
A/Vietnam/1194/2004 (clade 1) strain
Prior to vaccination, only 3.3% (95% CI, 1.5-6.2) of
the subjects in the H5N1-AS03A groups and 1.5%
(95% CI, 0.0-8.3) of those in the H5N1-DIL groups
had detectable levels of HI antibodies (≥1:10) against
the A/Vietnam/1194/2004 strain. Following the first
vaccine dose, the immune response against the
A/Vietnam/1194/2004 strain was higher in the H5N1-
AS03A groups, compared to the H5N1-DIL groups
(Fig).

GMTs against the A/Vietnam/1194/2004 strain was
observed in the H5N1-AS03A groups after the second
dose (202.2 [95% CI, 179.1-228.3]). In addition, the
second dose also induced a high seroconversion
rate (95.5% [95% CI, 92.2-97.6]), seroprotection rate
(95.8% [95% CI, 92.7-97.9]), and seroconversion factor
(38.0 [95% CI, 33.5-43.2]). In contrast, in the H5N1-
DIL groups, the H5N1 HI antibody GMTs reached
6.5 (95% CI, 5.5-7.6), while the seroconversion and
seroprotection rates (4.6% [95% CI, 1.0-12.9] each)
and seroconversion factor (1.3 [95% CI, 1.1-1.5]) did
not demonstrate an appreciable increase after both
doses.
A marked increase in the H5N1 HI antibody

vaccine, all three CHMP criteria in terms of the HI
antibody response against the A/Vietnam/1194/2004
strain were met (Fig). In contrast, none of the CHMP
criteria were met after any dose of the H5N1-DIL
vaccine.
After the second dose of the H5N1-AS03A
Cross-clade immunogenicity against the heterologous
A/Indonesia/05/2005 (clade 2.1) strain
Before vaccination, less than 1.0% subjects in the H5N1-
AS03A and H5N1-DIL groups had detectable levels of
HI antibodies (≥1:10) against the A/Indonesia/05/2005
strain. Following the first vaccine dose, the immune
response against the heterologous (cross-clade)
A/Indonesia/05/2005 strain was comparable in the
H5N1-AS03A and H5N1-DIL groups (Fig).
However, following
dose, the immune response to the heterologous
(cross-clade) A/Indonesia/05/2005 strain was higher
in the H5N1-AS03A groups than in the H5N1-DIL
groups (Fig). In the H5N1-AS03A groups, the H5N1
HI antibody GMTs increased to 20.9 (95% CI, 17.7-
24.7), while the seroconversion and seroprotection
rates were 45.7% (95% CI, 39.6-51.9) each; the mean
seroconversion factor value was 4.2 (95% CI, 3.5-
4.9). In contrast, in the H5N1-DIL groups, the HI
antibody GMTs remained low (5.2 [95% CI, 4.8-5.5]),
while no appreciable increase in seroconversion and
seroprotection rates (1.5% [95% CI, 0.0-8.3] each)
and seroconversion factor (1.0 [95% CI, 1.0-1.0]) were
observed.
the second vaccine

vaccine, two out of the three CHMP criteria for
licensure of seasonal influenza vaccine, in terms of
HI antibody response against the A/Indonesia/05/2005
strain were met (Fig). In contrast, none of the CHMP
criteria were met after any dose of the H5N1-DIL
vaccine.
Following the second dose of the H5N1-AS03A
Safety and reactogenicity
There was a consistent trend towards higher
frequency of solicited and unsolicited symptoms
(local and general) reported during a 7-day period
after each vaccination in the H5N1-AS03A groups as
compared to the H5N1-DIL groups. Overall, injection
site pain was the most frequently recorded solicited
local symptom reported by 96.1% (95% CI, 93.1-98.0)
of subjects in the H5N1-AS03A groups and 26.0% (95%
CI, 16.5-37.6) of subjects in the H5N1-DIL groups
(Table). Injection site pain of grade 3 intensity was
reported by 9.6% (95% CI, 6.4-13.7) of subjects in the
H5N1-AS03A groups. None of the subjects in the H5N1-
DIL groups reported any solicited local symptom
of grade 3 intensity. Frequencies of reporting of all
solicited local symptoms are presented in the Table.

symptom in the H5N1-AS03A groups was myalgia
61.6% (95% CI, 55.6-67.3), while in the H5N1-DIL
groups it was fatigue (42.5% [95% CI, 31.0-54.6]).
The most frequently reported solicited general

by 43.2% (95% CI, 37.4-49.2) of subjects in the H5N1-
AS03A groups and 42.5% (95% CI, 31.0-54.6) of subjects
in the H5N1-DIL groups. Influenza-like illness was
the most commonly reported unsolicited symptom
in the H5N1-AS03A groups (8.4% [95% CI, 5.4-12.2]),
while it was headache in the H5N1-DIL groups (8.2%
At least one unsolicited symptom was reported

Page 5

#??Prepandemic?H5N1?influenza?vaccine?#
?Hong?Kong?Med?J??Vol?17?No?1?#?February?2011?#??www.hkmj.org?
43
[95% CI, 3.1-17.0]). Unsolicited symptoms of grade 3
intensity were reported in 4.2% (95% CI, 2.2-7.2) of
subjects in the H5N1-AS03A groups and 1.4% (95% CI,
0.0-7.4) in the pooled H5N1-DIL group, influenza-like
illness (1.4% [95% CI, 0.4-3.5]) and irritation of the
larynx (1.4% [95% CI: 0.0-7.4]) being most frequently
reported in the respective groups.

both in the H5N1-AS03A group. One subject was
diagnosed with acute appendicitis after the second
dose, which resolved in 3 days, while another subject
Two serious adverse events were reported,
died on duty in a fire accident. None of these were
assessed to be causally related to vaccination.
Discussion
The first prepandemic vaccine from GSK Biologicals,
Prepandrix, was approved by the EMA in May 2008
(subsequently the licensure was extended to GSK
Biologicals’ pandemic vaccine, Pandemrix). The data
from this study conducted in Hong Kong, Thailand,
Singapore, and Taiwan were an essential component
FIG. Immunogenicity of the prepandemic influenza vaccine (in terms of seroprotection rate, seroconversion rate, and
seroconversion factor) against the A/Vietnam/1194/2004 (clade 1) and A/Indonesia/05/2005 (clade 2.1) strains, 21 days after each
dose (per-protocol cohort for immunogenicity)
CHMP denotes Committee for Medicinal Products for Human Use
100
100
50
60
60
30
20
20
10
80
80
40
40
40
20
0
0
0
H5N1-AS03A
H5N1-DIL
CHMP criterion >70%
Post-dose 1
A/Vietnam/1194/2004
Post-dose 1
A/Vietnam/1194/2004
Post-dose 1
A/Vietnam/1194/2004
Post-dose 1
A/Indonesia/05/2005
Post-dose 1
A/Indonesia/05/2005
Post-dose 1
A/Indonesia/05/2005
Post-dose 2
Post-dose 2
Post-dose 2
Post-dose 2
Post-dose 2
Post-dose 2
CHMP criterion >40%
CHMP criterion >2.5
Seroconversion factor
Seroprotection rate (%)
Seroconversion rate (%)
36.2
35.8
95.8
95.5
1.1
1.1
45.7
45.7
1.5
1.5
4.6
4.6
0.0
0.0
1.5
1.5
38.0
4.23.3
1.1
1.0 1.1
1.0
1.3