Id

ST

a

b

c

d

e

a

ARR1A

KPRI

1

r

I9

m

0d

Vaccine 30S (2012) A106– A113

Contents lists available at SciVerse ScienceDirect

Vaccine

j ourna l ho me pag e: www.elsev ier .com/ locate /vacc ine

mmunogenicity of the pentavalent rotavirus vaccine among infants in twoeveloping countries in Asia, Bangladesh and Vietnam

unheang Shina,∗, Dang Duc Anhb, K. Zamanc, M. Yunusc, Le Thi Phuong Maib, Vu Dinh Thiemb,asnim Azimc, John C. Victord, Michael J. Dallase, A. Duncan Steeled, Kathleen M. Neuzild, Max Ciarlete,1

International Vaccine Institute, Seoul, South KoreaNational Institute of Hygiene and Epidemiology, Hanoi, Viet NamInternational Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR, B), Mohakhali, Dhaka, BangladeshPATH Rotavirus Vaccine Program, Seattle, Washington, United StatesMerck Research Laboratories, North Wales, PA, United States

r t i c l e i n f o

rticle history:eceived 19 June 2011eceived in revised form4 November 2011ccepted 23 November 2011

eywords:entavalent rotavirus vaccineandomized controlled trial

mmunogenicity

a b s t r a c t

Background: We evaluated the immunogenicity of the pentavalent rotavirus vaccine (PRV) in two GAVI-eligible Asian countries, Bangladesh and Vietnam, nested in a larger randomized, double-blind, placebo-controlled efficacy trial conducted over a two-year period from 2007 through 2009.Methods: 2036 infants were randomly assigned, in a 1:1 ratio, to receive three oral doses of PRV orplacebo approximately at 6, 10, and 14 weeks of age. Concomitant use of EPI vaccines, including oralpoliovirus vaccine (OPV) and diphtheria-tetanus-whole cell pertussis (DTwP) vaccine, was encouragedin accordance to the local EPI schedule. A total of 303 infants were evaluated for immunogenicity andblood samples were collected before the first dose (pD1) and approximately 14 days following the thirddose (PD3). The seroresponse rates (≥3-fold rise from pD1 to PD3) and geometric mean titers (GMTs)were measured for anti-rotavirus immunoglobulin A (IgA) and serum neutralizing antibody (SNA) tohuman rotavirus serotypes G1, G2, G3, G4, and P1A[8], respectively.Results: Nearly 88% of the subjects showed a ≥3-fold increase in serum anti-rotavirus IgA response inthe analysis of the two countries combined. When analyzed separately, the IgA response was lower inBangladeshi children (78.1% [95% CI: 66.0, 87.5]) than in Vietnamese children (97.0% [95% CI: 89.6, 99.6]),with a PD3 GMT of 29.1 (units/mL) and 158.5 (units/mL), respectively. In the combined population, theSNA responses to the individual serotypes tested ranged from 10 (G3) to 50 (G1) percentage points lowerthan the responses shown in the developed countries. However, the SNA response to G3 in Vietnamese

subjects was 37.3% (95% CI: 25.8, 50.0), which was similar to the G3 response rate in developed countries.Conclusions: Three oral doses of PRV were immunogenic in two GAVI-eligible Asian countries: Bangladeshand Vietnam. The GMTs of both the serum anti-rotavirus IgA and SNA responses were generally higherin Vietnamese than in Bangladeshi children. The SNA responses varied by individual serotypes and werelower than the results from developed countries. The clinical significance of these observations is notunderstood because an immune correlate of protection has not been established.

. Introduction

Rotavirus continues to be the leading cause of severe diar-hoea in Asia among young children in both high- and low-income

∗ Corresponding author at: Translational Research Division, International Vaccinenstitute, SNU Research Park San 4-8, Nakseongdae-dong, Gwanak-gu, Seoul 151-19, South Korea. Tel.: +82 2 872 2801; fax: +82 2 872 2803.

E-mail addresses: sunheangshin@gmail.com, sshin@ivi.int (S. Shin).1 Present address: Novartis Vaccine & Diagnostics, Clinical Research & Develop-ent, Cambridge, MA, United States.

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2011.11.091

© 2011 Elsevier Ltd. All rights reserved.

countries [1]. In the region, approximately 45% of all diarrhoearelated hospitalizations among children less than 5 years of agehave been found to be attributable to rotavirus [2–9]. Vaccinationholds the best hope for the reduction of rotavirus-associated mor-tality and morbidity [3]. Given that rotavirus causes such a largeproportion (25–60%) of all hospitalizations for diarrhoea, it is pos-sible that a safe, effective and affordable rotavirus vaccine couldresult in a significant reduction in overall childhood mortality in the

region. Two rotavirus vaccines, the pentavalent rotavirus vaccine(PRV; RotaTeq®, Merck & Co. Inc., Whitehouse Station, NJ) and themonovalent rotavirus vaccine (MRV; Rotarix®, GlaxoSmithKlineBiologicals Inc., Rixensart, Belgium), have been licensed in many

e 30S

A[scft

(PereerssscASctwic

tAI

2

2

daAp2iBoP

ao(a(fsiaaigti8i

2

t

S. Shin et al. / Vaccin

sian countries and have obtained global WHO pre-qualification10]. Immunogenicity and efficacy of these rotavirus vaccines weretudied mostly among infants in high- and middle-income Asianountries, such as Taiwan, Korea and Singapore [11–13]. However,ew clinical trials had been performed in low-income Asian coun-ries with high childhood mortality for either vaccine.

At the advice of WHO’s Strategic Advisory Group of ExpertsSAGE) [14], a multi-country, placebo-controlled, double-blindhase III efficacy trial of PRV was conducted in two Asian countriesligible for GAVI Alliance co-financing, Bangladesh and Vietnam. Aseported by Zaman et al. [15], PRV was well tolerated, and over anfficacy follow-up period of nearly 2 years, the vaccine was 48.3%fficacious (95% confidence interval [CI]: 22.3–66.1) against severeotavirus gastroenteritis. For evaluation of a rotavirus vaccine, mea-urements of serum anti-rotavirus immunoglobulin (Ig)A and/orerum neutralizing antibody (SNA) responses are considered as thetandard for assessing immune responses following rotavirus vac-ination [16–20]. Thus, the Phase III efficacy trial of PRV in twosian countries also aimed to measure the anti-rotavirus IgA andNA responses to human rotavirus serotypes contained in the vac-ine (G1, G2, G3, G4, and P1A[8]) at approximately 14 days after thehird dose. The availability of such immunogenicity data, coupledith efficacy data from the same population, might contribute to

dentification of an immune correlate of protection or to design oflinical trials of additional rotavirus vaccine candidates.

Here we report the detailed findings of the immune responseso a 3-dose regimen of PRV among infants in the two GAVI-eligiblesian countries, Bangladesh and Vietnam, where the pivotal Phase

II efficacy trial of PRV was conducted.

. Materials and methods

.1. Study design

As previously reported [15], a placebo-controlled, randomized,ouble-blinded trial to evaluate the efficacy of three doses of PRVgainst severe RVGE among infants in low-income populations insia was conducted in rural Matlab, Bangladesh, and in urban anderi-urban Nha Trang, Vietnam from March 2007 through March009. The study was approved by the investigators’ corresponding

nstitutional review boards and the Western Institutional Reviewoard. The study was conducted in accordance with the principlesf the Declaration of Helsinki and in compliance with Good Clinicalractice guidelines.

After obtaining informed consent, infants were randomized in 1:1 ratio to receive three oral doses of PRV or placebo given withther routine pediatric vaccines, including oral poliovirus vaccineOPV) and diphtheria-tetanus-whole cell pertussis (DTwP) vaccine,ccording to local Expanded Program on Immunization schedulesapproximately 6-, 10-, and 14-weeks of age). Participants wereollowed from the moment they were enrolled until the end of thetudy. Trial enrollment in Bangladesh began in March 2007, whilen Vietnam the enrollment began in September 2007. To evalu-te immune responses to PRV, a subset of trial participants werelso enrolled into an immunogenicity cohort of approximately 150nfants per country. In Bangladesh, enrollment into this immuno-enicity cohort ran from July to August 2007, while in Vietnam, itook place in a single month at pre-selected sites. A total of 303nfants (149 [74 PRV: 75 placebo] in Bangladesh and 154 [74 PRV:0 placebo] in Vietnam) out of 2036 trial participants were enrolled

n the immunogenicity cohort.

.2. Immunogenicity assessment

Blood serum samples were collected from each infant beforehe first dose (pD1) and approximately 14 days following the third

(2012) A106– A113 A107

dose (PD3). The seroresponse rates and geometric mean titers(GMTs) were measured for anti-rotavirus IgA and SNA to humanrotavirus serotypes G1, G2, G3, G4, and P1A[8], respectively [21].Sero-response was defined as ≥3-fold rise from pD1 to PD3 asdescribed elsewhere [21–25]. Traditionally, a 4-fold rise criterionhas been used for doubling dilution assays. For the assays employedin this study, however, as well as throughout the clinical devel-opment of PRV, a 3-fold rise in titer has been used as validationexperiments showed that these assays were specific, reproducible,and sensitive enough to be able to detect a 3-fold difference with90% power at the 5% significance level. Serum samples were frozenand kept at −20 ◦C in laboratories at ICDDR, B in Matlab, and at Pas-teur Institute in Nha Trang until the samples were shipped to MerckResearch Laboratories. All immunologic assays were performed atChildren’s Hospital Medicine Center, Cincinnati, OH, USA.

2.3. Immunogenicity analyses

The immunogenicity analyses were based on the per-protocolpopulation (i.e., excluding protocol violators), subjects with validdata based on laboratory results from samples taken within theprotocol-specified day range, and subjects without intervening lab-oratory confirmed wild-type rotavirus disease. The proportion ofsubjects achieving a seroresponse, as measured by serum anti-rotavirus IgA responses and SNA responses to human rotavirusserotypes contained in PRV, was calculated for the two coun-tries combined, as well as for each country. The GMTs for serumanti-rotavirus IgA and SNA were summarized at pD1 and PD3.The associated 95% confidence intervals were calculated basedon binomial and normal distribution methodology, respectively.Immunogenicity analyses were also performed on sub-populationsof particular interest that were not specified in the protocol (posthoc analysis), including those subjects who received OPV concomi-tantly (on the same day) with each of the 3 doses of PRV or placebo,and those who did not receive OPV concomitantly with each of the3 doses of PRFV or placebo.

3. Results

3.1. Serum rotavirus-specific IgA responses

Among the 303 infants enrolled in the immunogenicity cohort,263 had both pD1 and PD3 data on anti-rotavirus IgA responses.Approximately 88% of these infants exhibited a ≥3-fold risebetween pD1 and PD3 (Table 1). The anti-rotavirus IgA serore-sponse rate was lower in Bangladeshi infants (78.1% [95% CI:66.0–87.5] than in Vietnamese infants (97.0% [95% CI: 89.6–99.6])(Table 1) and was accompanied by substantially lower PD3 GMTlevels among Bangladeshi infants (29.1 units/mL) compared tothat among Vietnamese infants (158.5 units/mL) (Table 1). In theplacebo group, 24 out of 132 infants (18%) showed a ≥3-fold risein anti-rotavirus IgA titer between pD1 and PD3, with a PD3 GMTlevel of 2.9 units/mL, indicating natural rotavirus infection amongsome infants during the first 6 months of life. Among those infantsin Bangladesh and Vietnam who received placebo, the proportionwith a ≥3-fold rise in anti-rotavirus IgA titer between pD1 and PD3,or the PD3 GMT level, was comparable between countries.

3.2. Serum neutralizing antibody (SNA) responses to humanrotavirus serotypes contained in PRV (G1, G2, G3, G4, P1A[8])

The SNA responses were shown to vary by the individual

serotypes contained in PRV as shown in previous clinical trials ofPRV [12,13,18,21–24]. In the per-protocol analysis, the SNA sero-responses were highest to serotype G1, followed by G3, P1A[8], G4,and G2 in the combined population of two Asian infant subjects

A108 S. Shin et al. / Vaccine 30S (2012) A106– A113

Table 1Combined country versus country-specific analysis of serum anti-rotavirus IgA responses (per-protocol analysis).

Two countries combined population Bangladesh Vietnam

PRV Placebo PRV Placebo PRV Placebo

pD1No. subjects available for analysisa 148 155 74 75 74 80GMT (units/mL) and 95% CI 1.7 (1.5, 2.0) 1.6 (1.4, 1.7) 1.9 (1.5, 2.4) 2.0 (1.6, 2.4) 1.6 (1.3, 1.9) 1.3 (1.2–1.3)

PD3No. subjects available for analysisb 131 132 64 58 67 74GMT (units/mL) and 95% CI 69.3 (49.9, 96.2) 2.9 (2.2, 3.9) 29.1 (18.6–45.7) 3.4 (2.1, 5.3) 158.5 (107.0, 234.6) 2.6 (1.8–3.8)

≥Three-fold rise% of subjects with ≥3-fold rise inanti-rotavirus Ig A and 95% CI

87.8% (80.9, 92.9%) 18.2% (12.0, 25.8%) 78.1% (66.0, 87.5%) 17.2% (8.6, 29.4%) 97.0% (89.6, 99.6%) 18.9% (10.7, 29.7%)

Abbreviations: GMT, geometric mean titer; pD1, pre-dose 1; PD3, post-dose 3; CI, confidence interval; PRV, pentavalent rotavirus vaccine.a Excludes protocol violators and subjects with invalid data based on laboratory determinations.b Excludes protocol violators, subjects with invalid data based on laboratory determinations, subjects with wild type rotavirus-positive stool results, or subjects with

samples taken out of a specified day range.

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

G1 G2 G3 G4 P1A[8 ]

Vietnam

Bangladesh

% S

NA

ser

ores

pons

e ra

te

Human rota virus serotypes conta ined in PRV

Fig. 1. Comparison of the percent (%) of the SNA seroresponse rates (≥3-fold increase from pD1 to PD3, %) to the five human rotavirus serotypes (G1, G2, G3, G4, and P1A[8])contained in PRV among Vietnamese and Bangladeshi infants who received PRV. Abbreviations: SNA, serum neutralizing antibody; pD1, predose 1; PD3, postdose 3; PRV,pentavalent rotavirus vaccine.

123.1

22.6

41.6

60.9

106.3

73.2

23.4 22.5

43.0

57.7

0

20

40

60

80

100

120

140

G1 G2 G3 G4 P1A[8]

Vietnam

Bangladesh

PD3

SNA

GM

T le

vels

Human rota virus serotypes contained in PRV

Fig. 2. Comparison of the PD3 SNA GMT levels to the five human rotavirus serotypes (G1, G2, G3, G4, and P1A[8]) contained in PRV among Vietnamese and Bangladeshiinfants who received three doses of PRV. Abbreviations: SNA, serum neutralizing antibody; PD3, postdose 3; PRV, pentavalent rotavirus vaccine; GMT, geometric mean titre.

e 30S (2012) A106– A113 A109

(ttai

vhot(ta

hra(oGGs

3O

tsdsoPpaowstwocnjd1rddPttotwtssaIcotrc

lizi

ng

resp

onse

s

to

hu

man

rota

viru

s

sero

typ

es

(G1,

G2,

G3,

G4,

and

P1A

[8])

con

tain

ed

in

PRV

in

the

com

bin

ed

per

-pro

toco

l pop

ula

tion

.

PRV

Plac

ebo

G1

G2

G3

G4

P1A

[8]

G1

G2

G3

G4

P1A

[8]

bjec

ts

wit

hla

ble

for

148

148

148

148

148

155

155

155

155

155

uti

on

un

its)

CI

53.1

(44.

7,

63.1

)

33.3

(28.

6,

38.8

)

21.1

(17.

8,

25.0

)

49.9

(43.

0,

58.0

)

63.8

(54.

0,

75.4

)

50.4

(43.

9,

59.2

)

33.3

(28.

5,

38.9

)

20.1

(16.

9,

23.8

)

48.0

(41.

2,

56.0

)

58.0

(49.

1,

68.3

)

bjec

ts

wit

hla

ble

for

131

131

131

131

131

132

132

132

132

132

uti

on

un

its)

CI

95.5

(77.

6,

119.

0)

23.0

(19.

7,

26.8

)

30.8

(25.

2,

37.7

)

51.4

(43.

1,

61.2

)

78.9

(65.

1,

95.5

)

19.9

(17.

1,

23.2

)

12.5

(11.

0,

14.3

)

10.1

(8.7

, 11.

7)

15.7

(13.

7,

18.0

)

18

.0

(14.

8,

21.9

)

e bjec

ts

wit

hla

ble

for

131

131

131

131

131

132

132

132

132

132

ects

wit

his

e

in

SNA

95%

CI

32.1

%

(24.

2,

40.8

)

9.9%

(5.4

, 16.

4)

28.2

%

(20.

7,

36.8

)

18.3

%

(12.

1,

26.0

)

27.5

%

(20.

0,

36.0

)

2.3%

(0.5

, 6.5

)

0.8%

(0.0

, 4.1

)

3.0%

(0.8

, 7.6

)

0

(0, 2

.8)

5.3

(2.2

, 10.

6)

:

SNA

, ser

um

neu

tral

izin

g

anti

bod

y;

GM

T,

geom

etri

c m

ean

tite

r;

pD

1,

pre

-dos

e

1;

PD3,

pos

t-d

ose

3;

CI,

con

fid

ence

inte

rval

;

PRV

, pen

tava

len

t

rota

viru

s

vacc

ine.

pro

toco

l vio

lato

rs

and

subj

ects

wit

h

inva

lid

dat

a ba

sed

on

labo

rato

ry

det

erm

inat

ion

s.

pro

toco

l vio

lato

rs, s

ubj

ects

wit

h

inva

lid

dat

a

base

d

on

labo

rato

ry

det

erm

inat

ion

s,

subj

ects

wit

h

wil

d

typ

e

rota

viru

s-p

osit

ive

stoo

l res

ult

s,

or

subj

ects

wit

h

sam

ple

s

take

n

out

of

a

spec

ified

day

ran

ge.

S. Shin et al. / Vaccin

Table 2). The sero-response in SNA titers ranged from 11.9% (G2)o 41.8% (G1) in Vietnam, approximately 1.5- to 2.5-fold higher thanhose measured in Bangladesh (Fig. 1). The higher SNA responsesmong infants in Vietnam compared to Bangladesh were also notedn the comparison of PD3 SNA GMT levels (Fig. 2).

The baseline (pD1) GMT levels of the SNA to each of the indi-idual rotavirus serotypes contained in PRV were considerablyigher than those obtained in clinical trials conducted in devel-ped countries [12,13,18,21–24], ranging from 24.2 units/mL (G3)o 79.1 units/mL (P1A[8]) in Bangladesh and from 18.4 units/mLG3) to 51.5 units/mL (P1A[8]) in Vietnam (Fig. 3). In both coun-ries, the pD1 SNA GMT levels were highest to serotypes P1A[8]nd G1, followed by serotypes G4, G2, and G3 (Fig. 3).

In both the PRV and placebo groups, the pD1 SNA GMTs wereigher in Bangladesh than in Vietnam against all five humanotavirus serotypes, possibly indicating higher levels of maternalntibodies present in Bangladeshi infants than those in VietnamFigs. 3 and 4). By PD3 (measured approximately at 14–26 weeksf age), the SNA GMT titers declined substantially; the PD3 SNAMTs to all 5 human serotypes were 2- to 4-fold lower than thoseMTs at pD1 (approximately 4–12 weeks of age) among the placeboubjects, and were comparable between the two countries (Fig. 4).

.3. Potential interference of immunogenicity between PRV andPV

Although the trial was designed to administer PRV concomi-antly with routine EPI vaccines, including OPV and DTwP, not allubjects received each dose of PRV/placebo and OPV on the sameay (Fig. 5). However, 91–92% of the Bangladeshi and Vietnameseubjects, respectively, in the immunogenicity cohort received eachf the 3 doses of OPV on the same day as each of the 3 doses ofRV/placebo. A post hoc analysis of the immunogenicity of PRV waserformed, as measured by the serum anti-rotavirus IgA responsesnd the SNA responses, in those subjects who received doses of OPVn the same day as each of the 3 doses of PRV and those subjectsho did not receive doses OPV with each of the 3 doses of PRV on the

ame day. This post hoc analysis was weighted towards the popula-ion in Vietnam because there was only one subject in Bangladeshho did not receive the 3 doses of PRV on the same day as doses

f OPV. The remainder of the infants received some doses of OPVoncomitantly with some, but not all, doses of PRV/placebo (dataot shown). The immunogenicity of PRV in those Vietnamese sub-

ects who received concomitant doses of OPV and PRV on the sameay showed generally lower GMT anti-rotavirus IgA levels (GMT,43.2 dilution units/mL) compared with those subjects who did noteceive doses of OPV with each of the 3 doses of PRV on the sameay (GMT, 232.7 dilution units/mL) (Fig. 5A). The same pattern ofecreased PD3 SNA GMT level was noted among those who receivedRV and OPV concomitantly compared to those who did not receivehe vaccines together (Fig. 5B). However, it is important to highlighthat this study was not designed to evaluate the immunogenicityf PRV when administered concomitantly with OPV or to evaluatehe immunogenicity of PRV when not administered concomitantlyith OPV. These comparisons are purely observational because

hese two groups were not randomized accordingly; the group ofubjects who did not receive PRV concomitantly with OPV cannoterve as a true control group for those subjects who received PRVnd OPV concomitantly. The groups also differ considerably in size.t is important to note that the subjects who did not receive OPVoncomitantly (on the same day) may have actually received OPV

ne or two days before or after administration of PRV. Adminis-ration of OPV one or two days before the administration of theotavirus vaccine can potentially interfere more with the repli-ation of the rotavirus vaccine than when OPV and the rotavirus Ta

ble

2Se

rum

neu

tra

pD

1 No.

of

sud

ata

avai

anal

ysis

a

GM

T

(dil

and

95%

PD3 No.

of

sud

ata

avai

anal

ysis

b

GM

T

(dil

and

95%

≥3-F

old

ris

No.

of

sud

ata

avai

anal

ysis

b

%

of

subj

≥3-f

old

rti

ter

and

Abb

revi

atio

nsa

Excl

ud

esb

Excl

ud

es

A110 S. Shin et al. / Vaccine 30S (2012) A106– A113

45.7

32.6

18.4

41.0

51.5

61.7

34.1

24.2

60.9

79.1

0

10

20

30

40

50

60

70

80

90

G1 G2 G3 G4 P1A[8]

Vietnam

Bangladesh

pD1

SNA

GM

T le

vels

Human rotavir us serotypes contained in PRV

F n rotaB ibody

vt

4

to[PtaatBetsos

Fn

ig. 3. Comparison of the baseline (pD1) SNA GMT levels to each of the five humaangladeshi subjects who received PRV. Abbreviations: SNA, serum neutralizing ant

accine are given on the same day, due to the active replication ofhe poliovirus vaccine strains.

. Discussion

The clinical trial of PRV conducted in Bangladesh and Vietnam ishe only Phase III study evaluating the efficacy and immunogenicityf a rotavirus vaccine performed in GAVI-eligible countries in Asia14]. Our study allowed the evaluation of the immunogenicity ofRV, an oral vaccine, in infants in two lower socio-economic coun-ries in Asia. In the present study, nearly 88% of the infants showed

≥3-fold rise in serum anti-rotavirus IgA response. However, thenti-rotavirus IgA seroresponse rates appeared different betweenhe two countries: the rate was approximately 78% and 97% inangladesh and Vietnam, respectively, likely reflecting the differ-nt socio-economic conditions between the subjects from each of

hese two GAVI-eligible countries. Although the anti-rotavirus IgAeroresponse rate in Vietnam was high and consistent with thosef the US, EU, Latin America, and Asian countries with a higherocioeconomic level (Taiwan and Korea) [12,13], the anti-rotavirus

A: Bang ladesh

65.4

36.3

24.2

61.0

81.1

20.7

11.7 10.0

17.9 20.320

30

40

50

60

70

80

90

G1 G2 G3 G4 P1A[ 8]

Pre-dose Post-dose 3

SNA

GM

T le

vels

Human rotavirus serotypes c

ig. 4. Comparison of the pD1 and PD3 SNA GMT levels in the placebo group in Bangladeeutralizing antibody; pD1, predose 1; PD3, postdose 3, PRV, pentavalent rotavirus vaccin

virus serotypes (G1, G2, G3, G4, and P1A[8]) contained in PRV in Vietnamese and; pD1, predose 1; PRV, pentavalent rotavirus vaccine; GMT, geometric mean titre.

IgA seroresponse rate in Bangladesh was approximately 10–15percentage points lower than those observed in other regionsof the developed world, and more on par with those observedin Sub-Saharan African countries [25,26]. Given that PRV is anoral vaccine, these results likely reflect that, in developing coun-tries, oral vaccines have a history of being less immunogenic thanin the developed world. These differences of oral vaccines havebeen postulated due to differences in the level of transplacentallyacquired maternal antibody, immune and non-immune compo-nents of breast milk, the amount of gastric acid in the digestive tract,micronutrient malnutrition, interfering gut flora, and diarrheal andimmune system disease [15,27–29]. In the case of Bangladeshversus Vietnam, the reasons for the decreased immunogenicity ofPRV in Bangladeshi infants may be due to a combination of the dif-ferences in host populations and their associated health conditions,which include malnutrition and concomitant infections of the gut

with several enteropathogens.

In addition, the PD3 anti-rotavirus IgA GMT levels were alsoreduced in Asian subjects when compared to those of sub-jects in developed world countries [12,13,18,21–24]. The GMT

B: Vietnam

39.4

30.8

16.8

38.442.3

19.413.3

10.114.1 16.4

0

10

20

30

40

50

60

70

80

90

G1 G2 G3 G4 P1A[ 8]

Pre-do se Post-do se 3

ontained in PRV

shi infants (A) and Vietnamese (B) infants, respectively. Abbreviations: SNA, serume; GMT, geometric mean titre.

S. Shin et al. / Vaccine 30S (2012) A106– A113 A111

112.6

172.5

22.7 22.4

35.4

76.2

55.4

87.0

99.4

137.1

0

20

40

60

80

100

120

140

160

180

200

PRV and OPV concomitantly administered at all 3 doses PRV and OPV not administered concomitantly at all 3 doses

G1 G2 G3 G4 P1A[8 ]

PD3

SNA

GM

T le

vels

1.4 2.8

143.2

232.7

0

50

250

PRV and OPV concomitantly administered at al l 3 doses (N at

A

B

pD1=59, N at PD3 =53 )PRV and OPV not adm inistered

concomitantly at all 3 doses (N at pD 1 =15, N at PD3 =14 )

Serum IgA at pD 1

Serum IgA at PD3Seru

m a

n�-r

otav

iurs

IgA

GM

T le

vels

Fig. 5. (A) pD1 and PD3 GMT levels of serum anti-rotavirus IgA in Vietnamese infants who received OPV concomitantly with PRV at all three doses (left) compared to thosewho did not receive 3 doses of PRV and OPV concomitantly (right). (B) PD3 SNA GMT levels in Vietnamese subjects who received OPV concomitantly with PRV concomitantly,s y (righ3

(AsHcau1w(mAnb

ame day, (left) and those who did not receive 3 doses of PRV and OPV concomitantl; PRV, pentavalent rotavirus vaccine; GMT, geometric mean titre.

69.3 dilution units/mL) of the serum anti-rotavirus IgA at PD3 ofsian subjects was approximately 2-fold lower than those mea-ured 14 or 42 days after Dose 3 in subjects in developed countries.owever, once again, the pattern was not the same when the twoountries were evaluated separately. The GMT level of the serumnti-rotavirus IgA at PD3 of Bangladeshi subjects was 29.1 dilutionnits/mL, approximately 5- to 10-fold lower than those measured4 or 42 days after Dose 3 in subjects in developed countries,hile the PD3 GMT level of the serum IgA in Vietnamese subjects

158.5 dilution units/mL) was approximately the same as those

easured 14 or 42 days after Dose 3 in subjects in the EU and Latinmerica [21,24]. The clinical significance of these observations isot understood because an immune correlate of protection has noteen established.

t). Abbreviations: SNA, serum neutralizing antibody; pD1, predose 1; PD3, postdose

SNA responses to each of the five human serotypes, G1, G2,G3, G4, and P1A[8], contained in PRV were also evaluated at pD1and PD3 in Asian subjects. The results showed a ≥3-fold rise inSNA responses to rotavirus serotypes G1, G2, G3, G4 and P1A[8]in varying percentages in the Asian subjects. A consistent andsimilar pattern was observed when the data from Bangladeshand Vietnam were compared to those of the African subjects[25,26]. For serotypes G1, G2, G3, G4, and P1A[8], the ≥3-foldSNA response rates in Bangladeshi subjects were approximately50, 30, 10, 35, and 40 percentage points, respectively, lower than

those exhibited by subjects in the US, EU, Taiwan, Korea, and LatinAmerica [12,13,18,21–24]. In Vietnamese subjects, the ≥3-fold SNAresponse rates were approximately 30, 25, 25, and 15 percentagepoints, respectively, lower for serotypes G1, G2, G4, and P1A[8]

A e 30S

teccibGcoh

stcmitipaimtfid

2icUGwps(asiT

∼InaamiinFBhm

shbaeaoisa

P

[

[

[

112 S. Shin et al. / Vaccin

han those exhibited by subjects in the developed countries. How-ver, it is important to point out that the pD1 SNA GMT levels wereonsiderably higher in these populations than those in developedountries. Therefore, achievement of a seroresponse, which by def-nition, requires a ≥3-fold increase from pD1 to PD3, might haveeen more difficult in these populations because of the higher pD1MT levels, which is likely a reflection of SNA acquired transpla-entally or via breastmilk. The lower immunogenicity and efficacyf PRV in poor developing countries could be explained, in part, byigher titers of SNA in breast milk at the time of immunization [30].

For serotype G3, the ≥3-fold SNA response rates in Vietnameseubjects were approximately 10 percentage points higher thanhose exhibited by subjects in the developed world settings. Coin-identally, rotavirus strains belonging to the G3 genotype were theost prevalent during the duration of the study [15], also suggest-

ng the possibility that natural exposure might have contributed tohe appearance of a relatively enhanced G3 specific SNA responsen Vietnam. Looking at the baseline SNA responses (Fig. 3), theD1 SNA titers to serotype G3 were high not only in Vietnam butlso in Bangladesh: 24.2 and 18.4 dilution units/mL of pD1 GMTn Bangladesh and Vietnam, respectively. This may indicate com-

on circulation of G3 strains in both countries before and/or duringhe clinical trial. Nevertheless, G3 rotavirus strains were not identi-ed in Bangladesh among the rotavirus cases detected and enrolleduring the clinical trial.

In terms of the GMT levels at PD3, there was a decrease of about.5-fold in the GMTs corresponding to the G1 and P1A[8] serotypes

n the Bangladeshi subjects who received PRV in this study whenompared to the GMT levels shown in studies conducted in theS, EU, Taiwan, Korea, and Latin America [12,13,18,21–24]. TheMTs for serotypes G2, G3, and G4 among Bangladeshi subjectsho received PRV were generally similar when they were com-ared to GMTs for the corresponding rotavirus serotypes amongubjects who received PRV in the other studies. There was little1.5-fold) to no decrease in the GMTs to serotypes G1, G2, G3, G4,nd P1A[8] in the Vietnamese subjects who received PRV in thistudy when compared to the GMTs to the same rotavirus serotypesn subjects who received PRV in studies conducted in these US, EU,aiwan, Korea, and Latin America [12,13,18,21–24].

Interestingly, approximately 18% (∼17% in Bangladesh and19% in Vietnam) of the subjects who received placebo had an

gA seroresponse. This would suggest that the rate of exposure toaturally occurring rotavirus is very high in these Asian countries,nd the relatively rapid acquisition of infection by 5–6 months ofge (i.e. 14 days PD3). Thus, it is important to note that enroll-ent patterns and rotavirus circulation patterns may influence the

nterpretation of background rates of antibody. Although rotaviruss known to circulate throughout the year in Bangladesh and Viet-am, rotavirus activity is highest during certain months of the year.or the subjects who participated in the immunogenicity cohort,angladesh enrolled some of the subjects during the months ofighest rotavirus activity, while Vietnam enrolled them in a singleonth during the high rotavirus season.Another important observation is that at the time these Asian

ubjects received Dose 1, at approximately 4–10 weeks of age, theyave little to no pre-existing serum anti-rotavirus IgA as evidencedy the low GMT levels. However, at the time of the first dose, nearlyll subjects, whether they received PRV or placebo, had high lev-ls of SNA against all the rotavirus serotypes tested, suggestingcquisition of SNA transplacentally or via breastmilk (the isotypef the prevalent neutralizing antibody responsible for the neutral-zation activity in the SNA assay is not known). This observation

upports the suggestion that pre-existing maternal antibody playsn important role in vaccine take of live oral rotavirus vaccines [27].

Our clinical trial demonstrated that the immunogenicity ofRV was consistently higher in Vietnamese than in Bangladeshi

[

(2012) A106– A113

subjects in all immunogenicity assays performed. In addition,higher immune response levels translated into higher efficacy levelas demonstrated in the same trial (Vietnam, 68.1% [95% CI: 7.6,90.9%]; Bangladesh, 42.7% [95% CI: 10.4, 63.9%]) [15]. The differ-ences in efficacy between the two countries may be the result of thedifferent intensity of the immune responses in these populationstogether with heterogeneous socio-epidemiological circumstancesof the study populations. However, it is important to note that thehigher point estimate of efficacy in Vietnam than in Bangladeshmay be attributable to a low degree of precision in this study, as thestudy was not designed to make statistical comparisons betweenthe countries.

In brief, three oral doses of PRV were immunogenic in twoGAVI-eligible Asian countries, Bangladesh and Vietnam, althoughdifferences were noted between these two countries. Both theserum anti-rotavirus IgA response and SNA GMT levels followingthe third dose of PRV were lower among infants in Bangladeshthat in Vietnam. While the immune responses measured in Viet-namese children were comparable to those among children in LatinAmerica and Europe [21,24], the immune responses measured inBangladeshi children were more comparable to those shown inimpoverished populations in Africa [25]. Understanding differencesbetween these two populations might help elucidate the well-recognized difficulties of live oral vaccines in developing countries.

Acknowledgements

We thank all the local collaborators and project staff workingin Khanh Hoa Province in Vietnam and in Matlab, Bangladesh. Thistrial (Merck protocol V260-015) was funded by PATH’s RotavirusVaccine Program under a grant from the GAVI Alliance and the trialwas co-sponsored by Merck & Co. Inc.

Conflict of interest statement: MC and MJD were employees ofMerck when the study was conducted and owned equity in thecompany. No other conflicts of interest are declared.

References

[1] Bresee J, Fang ZY, Wang B, Nelson EA, Tam J, Soenarto Y, et al. First report fromthe Asian rotavirus surveillance network. Emerg Infect Dis 2004;10(6):988–95.

[2] Bresee JS, Glass RI, Ivanoff B, Gentsch J. Current status and future priorities forrotavirus vaccine development, evaluation and implementation in developingcountries. Vaccine 1999;17:2207–22.

[3] Glass RI, Breese JS, Turcios R, Fischer TK, Parashar UD, Steele AD. Rotavirus vac-cine: targeting the developing world. J Infect Dis 2005;192(Suppl. 1):S160–6.

[4] Santosham M, Nelson EA, Bresee JS. Implementing rotavirus vaccination in Asia.Vaccine 2007;25:7711–6.

[5] Nguyen VM, Luan LT, Dang DT, Nguyen TH, Tu PV, Long NT, et al. Epidemiologicalprofile and burden of rotavirus diarrhea in Vietnam: 5 years of Sentinel hospitalsurveillance, 1998–2003. J Infect Dis 2005;192(Suppl. 1):S127–32.

[6] Parashar UD, Hummelman EG, Bresee JS, Miller MA, Glass RI. Global ill-ness and deaths caused by rotavirus disease in Children. Emerg Infect Dis2003;9(5):565–72.

[7] Nguyen VM, Nguyen VT, Huynh PL, Dang DT, Nguyen TH, Phan VT, et al. Theepidemiology and disease burden of rotavirus in Vietnam: sentinel surveillanceat 6 hospitals. J Infect Dis 2001;183:1707–12.

[8] Jiraphongsa C, Bresee JS, Pongsuwanna Y, Kluabwang P, Poonawagul U, Arporn-tip P, et al. Epidemiology and burden of rotavirus diarrhea in Thailand: resultsof sentinel surveillance. J Infect Dis 2005;192(Suppl. 1):S87–93.

[9] Parashar UD, Gibson CJ, Bresse JS, Glass RI. Rotavirus and severe childhooddiarrhea. Emerg Infect Dis 2006;12:304–6.

10] World Health Organization. Meeting of the immunization Strategic AdvisoryGroup of Experts, April 2009 – conclusions and recommendations. Wkly Epi-demiol Rec 2009;84:220–36.

11] Phua KB, Quak SH, Lee BW, Emmanuel SC, Goh P, Han HH, et al. Evaluation ofRIX 4414, a live attenuated rotaviurs vaccine, in a randomized, double-blind,placebo-controlled phase 2 trial involving 2464 Singapore inifants. J Infect Dis2005;192(Suppl. 1):S6–16.

12] Chang CC, Chang MH, Lin TY, Lee HC, Hsieh WS, Lee PI. Experience of pentavalent

human-bovine reassortant rotavirus vaccine among healthy infants in Taiwan.J Formos Med Assoc 2009;108(4):280–5.

13] Kim DS, Lee TJ, Kang JH, Kim JH, Lee JH, Ma SH, et al. Immunogenicity and safetyof a pentavalent human-bovine (WC3) reassortant rotavirus vaccine in healthyinfants in Korea. Pediatr Infect Dis J 2008;27(2):177–8.

e 30S

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

S. Shin et al. / Vaccin

14] World Health Organization. Conclusions and recommendations from theimmunization Strategic Advisory Group. Wkly Epidemiol Rec 2006;32:2–11.

15] Zaman K, Anh DD, Victor JC, Shin S, Yunus M, Dallas MJ, et al. Efficacy of thepentavalent rotavirus vaccine against severe rotavirus gastroenteritis amonginfants in developing countries in Asia: a randomized, double-blind, placebo-controlled trial. Lancet 2010;376(9741):615–23.

16] Ward RL. Mechanisms of protection against rotavirus in humans and mice. JInfect Dis 1996;174(Suppl. 1):S51–8.

17] Ward RL, Knowlton DR, Zito ET, Davidson BL, Rappaport R, Mack ME. Serologiccorrelates of immunity in a tetravalent reassortant rotaviurs vaccine trial. JInfect Dis 1997;176:570–7.

18] Ciarlet M, Schödel F. Development of a rotavirus vaccine: clinical safety,immunogenicity, and efficacy of the pentavalent rotavirus vaccine, RotaTeq® .Vaccine 2009;27(Suppl. 5):G72–81.

19] Heaton PM, Ciarlet M. The pentavalent rotavirus vaccine: discovery to licensureand beyond. Clin Infect Dis 2007;45:1618–24.

20] Velazquez FR, Matson DO, Guerrero ML, Shults J, Calva JJ, Morrow AL. Serumantibody as a marker of protection against natural rotavirus infection and dis-ease. J Infect Dis 2000;182:1602–9.

21] Ciarlet M, Sani-Grosso R, Yuan G, Liu GF, Heaton PM, Gottesdiener KM, et al.Concomitant use of the oral pentavalent human-bovine reassortant rotavirus

vaccine and oral poliovirus vaccine. Pediatr Infect Dis J 2008;27:874–80.

22] Rodriguez ZM, Goveia MG, Stek JE, Dalls MJ, Boslego JW, DiNubile MJ, et al. Con-comitant use of an oral live pentavalent human-bovine reassortant rotavirusvaccine with licensed parenteral vaccines in the United States. Pediatr InfectDis J 2007;26(3):221–7.

[

[

(2012) A106– A113 A113

23] Vesikari T, Matson DO, Dennehy P, Van Damme P, Santosham M, RodriguezZ, et al. Safety and efficacy of pentavalent human-bovine (WC3) reassortantrotavirus vaccine in preventing rotavirus gastroenteritis and reducing associ-ated healthcare resource utilization. N Engl J Med 2006;354(1):23–33.

24] Ciarlet M, He S, Lai S, Petrecz M, Yuan G, Liu GF, et al. Concomitant use ofthe 3-dose oral pentavalent rotavirus vaccine with a 3-dose primary vaccina-tion course of a diphtheria-tetanus-acellular pertussis-hepatitis B-inactivatedpolio-Haemophilus influenzae type b vaccine: immunogenicity and reacto-genicity. Pediatr Infect Dis J 2009;28(3):177–81.

25] Armah G, Samba SS, Breiman RF, Dallas MJ, Tapia MD, Feikin DR, et al. Efficacy ofthe pentavalent human-bovine reassortant rotavirus vaccine against rotavirusgastroenteritis in sub-Saharan Africa. Lancet 2010;376(9741):606–14.

26] Armah G, Breiman RF, Tapia MD, Sow S, Dallas MJ, Binka F, et al. Immuno-genicity of the pentavalent rotavirus vaccine in African infants. Vaccine 2012;30(Suppl. 1):A86–93.

27] Patel M, Shane A, Parashar U. Oral rotavirus vaccines: how well will they workwhere they are needed most. J Infect Dis 2009;200:S39–48.

28] Zaman K, Sack DA, Yunus M, Arifeen SE, Podder G, Azim T, et al. Success-ful co-administration of a human rotavirus and oral poliovirus vaccines inBangladeshi infants in a 2-dose schedule at 12 and 16 weeks of age. Vaccine2009;27:1333–9.

29] Rennels MB. Influence of breastfeeding and OPV on the immunogenicity andefficacy of rotavirus vaccine. J Infect Dis 1996;174(Suppl. 1):S107–11.

30] Moon S, Wang Y, Shane AL, Nguyen T, Ray P, Dennehy P, et al. Inhibitory effectof breast milk on infectivity of live oral rotavirus vaccines. Pediat Infect Dis J2010;29:919–23.