Randomized Study of Early versus Late Immunization with Pneumococcal Conjugate Vaccine after...

1392 • CID 2009:48 (15 May) • Cordonnier et al.

M A J O R A R T I C L E

Randomized Study of Early versus LateImmunization with Pneumococcal ConjugateVaccine after Allogeneic Stem Cell Transplantation

Catherine Cordonnier,1 Myriam Labopin,2 Virginie Chesnel,2 Patricia Ribaud,3 Rafael De La Camara,5

Rodrigo Martino,6 Andrew J. Ullmann,7 Terttu Parkkali,10 Anna Locasciulli,11 Karima Yakouben,4 Karlis Pauksens,12

Hermann Einsele,8 Dietger Niederwieser,9 Jane Apperley,14 and Per Ljungman,13 for the Infectious Diseases WorkingParty of the European Group for Blood and Marrow Transplantation1Hematology Department, Henri Mondor Teaching Hospital, Assistance Publique–Hopitaux de Paris (AP-HP) and Paris 12 University, Creteil,and 2European Group for Blood and Marrow Transplantation Data and Study Office, Saint-Antoine Teaching Hospital, AP-HP, Universite Pierre et MarieCurie Paris 6 University, INSERM Unite Mixte de Recherche en Sante 893 group 14, 3Hematology and Bone Marrow Transplant Department, Saint-Louis Teaching Hospital, AP-HP and Paris 7 University, and 4Hematology Department, Robert Debre Teaching Hospital, Paris, France; 5HematologyDepartment, de la Princesa Teaching Hospital, Madrid, and 6Hematology Department, Sant Pau Teaching Hospital, Barcelona, Spain; 7JohannesGutenberg–Universitat Teaching Hospital, Mainz, 8Department of Internal Medicine, University Hospital, Wurzburg, and 9Division of Internal MedicineII, University Hospital, Leipzig, Germany; 10Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland; 11San Camillo TeachingHospital, Rome, Italy; 12Division of Infectious Diseases, University Hospital, Uppsala, and 13Department of Hematology, Karolinska University Hospitaland Karolinska Institute, Stockholm, Sweden; and 14Department of Haematology, Hammersmith Teaching Hospital, London, United Kingdom

Background. Invasive pneumococcal disease is a life-threatening complication after allogeneic stem cell trans-plantation, and at least 20% of cases occur within 1 year after transplantation. The 23-valent pneumococcalpolysaccharide vaccine (PPV23) has limited efficacy, especially during the first year after transplantation. Theimmune response to the conjugated vaccines is expected to be better than that to the polysaccharide vaccine, butthe optimal timing of vaccination is not defined. Our objective was to show that a 7-valent pneumococcal conjugatevaccine (PCV7; Prevnar) was not inferior when first given 3 months after transplantation, compared with whenfirst given 9 months after transplantation.

Methods. We performed a multicenter, randomized, noninferiority study involving 158 patients from 13 Eu-ropean Group for Blood and Marrow Transplantation centers who were randomly allocated at ∼100 days aftermyeloablative stem cell transplantation to receive a series of vaccinations (3 doses of PCV7 given 1 month apart)that was started immediately (i.e., 3 months after transplantation) or 6 months later (i.e., 9 months after trans-plantation). The primary evaluation criterion was the rate of response (antibody level, �0.15 mg/mL for each ofthe 7 serotypes) at 1 month after the third dose of PCV7. The noninferiority margin was 20%. All patients werefollowed up for 24 months after transplantation or until death, whichever occurred first.

Results. We found that the response rate was not lower after early vaccination (79% [45 of 57 patients]) than afterlate vaccination (82% [47 of 57 patients]) (difference, �3.5%; 90% confidence interval, �15.6 to 8.6; not significant).

Conclusions. We conclude that PCV7 vaccination at 3 months after stem cell transplantation is not inferiorto PCV7 vaccination at 9 months after transplantation. Because invasive pneumococcal disease can occur early,we recommend starting the PCV7 vaccination series at 3 months after transplantation to ensure earlier protectionagainst Streptococcus pneumoniae. However, the early vaccination may result in only short-lasting response andmay not prime for a 23-valent pneumococcal polysaccharide vaccine boost as efficiently as the late vaccination.

Allogeneic stem cell transplant (SCT) recipients are at

Received 16 September 2008; accepted 12 January 2009; electronicallypublished 14 April 2009.

Reprints or correspondence: Dr. Catherine Cordonnier, Hematology Dept., HenriMondor University Hospital, 94000 Creteil, France ([email protected]).

Clinical Infectious Diseases 2009; 48:1392–401� 2009 by the Infectious Diseases Society of America. All rights reserved.1058-4838/2009/4810-0011$15.00DOI: 10.1086/598324

increased risk for pneumococcal infections [1–3]. This

may be the result of functional asplenia or decreased

production of Streptococcus pneumoniae–specific op-

sonizing antibodies, perhaps because of B cell defi-

Presented in part: 47th Interscience Conference on Antimicrobial Agents andChemotherapy, Chicago, Illinois, September 2007, and 34th Annual Meeting ofthe European Group for Blood and Marrow Transplantation, Florence, Italy, March–April 2008.

by guest on May 17, 2011

cid.oxfordjournals.orgD

ownloaded from

http://cid.oxfordjournals.org/

Pneumococcal Vaccine after Stem Cell Transplantation • CID 2009:48 (15 May) • 1393

ciency, particularly in patients with graft-versus-host disease

(GVHD) [3–6].

Antibody response rates to the 23-valent pneumococcal poly-

saccharide vaccine (PPV23) after SCT are low; they are esti-

mated to be 25% among patients with chronic GVHD or who

receive steroid therapy and 50%–60% among other patients [5,

7, 8]. Neither repeated vaccination nor donor immunization

[6, 9] improved this response. Although PPV23 vaccination is

widely recommended [10, 11], a more effective method of pro-

tecting against pneumococcal disease is needed for SCT

recipients.

It is hypothesized that covalent conjugation of a polysac-

charide to a protein carrier converts most of the immune re-

sponse from T cell independent to T cell dependent, resulting

in a stronger and longer-lasting antibody response [12–14]. The

7-valent pneumococcal conjugate vaccine (PCV7; Prevnar; Wy-

eth Pharmaceuticals) is a solution of 7 capsular S. pneumoniae

antigens, each conjugated to the diphtheria CRM197 protein.

An antibody titer �0.15 mg/mL for all 7 serotypes was found

in 90% of infants given PCV7 at 2, 4, and 6 months of age

[15]. Two studies involving allogeneic SCT recipients showed

titers �0.5 mg/mL for all 7 serotypes in ∼70% of patients after

3 doses of vaccine [16, 17]. Severe invasive pneumococcal in-

fection can occur early after allogeneic SCT [2]. However, re-

sponses to vaccines may be weak immediately after SCT [8,

18], and the optimal timing of vaccination is yet not defined.

The aim of our study was to compare early vaccination,

started at the beginning of the high-risk period (on day 100),

with late vaccination, started 9 months after transplantation

(control condition). Our hypothesis was that early vaccination

was not inferior to late vaccination with respect to the rate of

response (antibody titers, �0.15 mg/mL to all 7 PCV7 serotypes)

at 1 month after the third PCV7 dose.

MATERIALS AND METHODS

Study design. This was a randomized noninferiority trial con-

ducted by the European Group for Blood and Marrow Trans-

plantation (EBMT) Infectious Diseases Working Party at 13

centers in 7 countries, from December 2001 through December

2004 (figures 1 and 2). The study was approved by the review

boards of all participating institutions. Informed consent was

obtained from the patients and/or their parents.

The primary end point was the percentage of patients with

antibody titers �0.15 mg/mL to all 7 PCV7 serotypes at 1 month

after the third dose of PCV7 (which was designated “S3” for

serum sample 3). This primary end point was compared be-

tween 2 groups of patients randomly allocated to receive early

vaccination (date of first dose, 100 days � 10 days after trans-

plantation) or late vaccination (date of first dose, 9 months �

10 days after transplantation). The secondary end points in-

cluded the geometric mean concentrations of specific antibod-

ies to the 7 PCV7 serotypes before vaccination (at baseline,

designated S1), 1 month after 3 doses of PCV7 (S3), imme-

diately before PPV23 vaccination (S4), 1 month after PPV23

injection (S5), and 24 months after transplantation (S6). Be-

cause 2 studies of PCV7 in SCT recipients, published while our

study was ongoing [16, 17], used 0.5 mg/mL as the cutoff for

defining an immune response, we conducted post hoc analyses

using this 0.5-mg/mL cutoff.

All patients were followed up for 24 months after trans-

plantation or until death, whichever occurred first. S. pneumo-

niae infections were recorded throughout the 21-month

follow-up period. Supportive care, including prophylactic an-

tibiotic treatment, was given at the discretion of the

investigators.

Eligibility. Inclusion criteria were age 5–65 years; first al-

logeneic SCT; transplant from an HLA-identical sibling, a fa-

milial 1-antigen mismatched donor, or a matched unrelated

donor; and myeloablative conditioning. Patients were included

in the study at 85–109 days after transplantation.

Exclusion criteria were reduced-intensity conditioning,

transplant from a haplo-identical donor, an intention to give

donor lymphocyte infusions in a planned protocol, platelet

count !50,000 platelets/mL, relapse of the underlying disease,

uncontrolled grade IV acute GVHD, known allergy to PCV7,

pneumococcal infection or vaccine since transplantation, im-

mune globulin received within the past 4 weeks, life expectancy

!2 weeks, and participation in another trial within the past 30

days.

Randomization. Randomization was performed by the

EBMT London office. On the basis of EBMT registry megafile

data, we estimated that ∼35% of patients would die between

3 months (date of eligibility) and 9 months (date of the first

vaccine in the late vaccination group) after transplantation. The

randomization procedure was designed to ensure that the num-

ber of vaccinated patients in the 2 groups would be similar

despite this expected mortality rate. Randomization was per-

formed by use of the minimization method, stratified by center,

donor relatedness (i.e., related vs. unrelated), and source of

cells (bone marrow or cord blood vs. peripheral stem cells).

Vaccine administration. PCV7 (0.5 mL) was injected in-

tramuscularly 3 times at intervals of 1 month. A dose of PPV23

was given at 12 months after transplantation in the early vac-

cination group and at 18 months after transplantation in the

late vaccination group (figure 2).

Definitions. Invasive pneumococcal disease was defined as

isolation of S. pneumoniae in blood or CSF samples, irrespective

of clinical symptoms, or in a bronchoscopy specimen from

patients with pneumonia. GVHD was defined according to

standard criteria [19].

Serum sample collection and antibody measurements.

The serum samples were labeled using code numbers, were



ownloaded from



Figure 1. Consort flow chart. PPV23, 23-valent pneumococcal polysaccharide vaccine; pt, patient; S3, 1 month after 3 doses of 7-valent pneumococcalconjugate vaccine; S6, 24 months after transplantation; SAE, serious adverse event.

immediately frozen at �20�C, and were shipped frozen by

BioInova to the US Wyeth laboratories. Serum IgG responses

to the 7 vaccine serotypes were determined using a double-

absorbent serotype-specific antipneumococcal capsular poly-

saccharide ELISA, as described elsewhere [20], and were ex-

pressed in units of mg/mL with use of the standard reference

serum 89-SF. The quantification limit was 0.01 mg/mL for all

serotypes. All serum samples were bar-coded and tested blindly

by qualified laboratory personnel.

Statistical analyses. The primary end point was designed

to see whether the immune response to early vaccination was

noninferior to the immune response to late vaccination in terms

of the percentage of patients with antibody titers �0.15 mg/mL

to all the 7 PCV7 serotypes at 1 month after the third dose of

vaccine. A 1-sided test was considered acceptable because this

was a noninferiority trial. The noninferiority margin was set at

20% fewer patients with response, given the probable benefits

associated with earlier protection.

The response rate to pneumococcal vaccines has been defined

in various ways, reflecting the lack of data about the antibody

titers that confer protection. However, the lowest published

response rate for PPV23 was 25% in patients with chronic

GVHD [5, 8, 13]. To define the response rate in our study, we

chose 0.15 mg/mL as the cutoff for all 7 antibody titers, in

accordance with the largest study performed involving infants

[15, 21]. In a comparison of polysaccharide and conjugate Hae-

mophilus influenzae vaccines, the conjugate vaccine produced

a 30% higher response rate [12]. Therefore, a 55%

( ) response rate might be expected for PCV7 in25% + 30%

patients with GVHD. Given that GVHD was not expected in

all patients, we estimated that 65% was a likely response rate.

The primary end point was tested using the 90% CI for dif-



ownloaded from



Figure 2. Diagram of the study design. Prevnar, 7-valent pneumococcal conjugate vaccine (Wyeth Pharmaceuticals); S1–S6, serum samples 1–6.

ference in response rate. If the lower 90% confidence limit was

within the noninferiority region, we considered noninferiority

to be established [22]. With , 142 patients (71 pera p 0.05

group) were required to obtain 80% power ( ) for es-b p 0.20

tablishing noninferiority. Taking into account that 35% of pa-

tients in the late vaccination group were expected to drop out

of the study before the first vaccine, we decided to include 181

patients in the study to expect 142 vaccinated patients.

For secondary end points, we used the x2 test to compare

the percentages of patients with titers �0.15 mg/mL between

the 2 groups and used the nonparametric Mann-Whitney rank-

sum test to compare geometric mean concentrations. All P

values were 2-sided with type I error fixed at 0.05. Because we

wanted to allow comparisons between our results and those of

the 2 studies published while our study was under way [16,

17], we defined a response as titers �0.5 mg/mL to all 7 antigens

for a post hoc analysis that repeated the primary end point and

for a correlation analysis between patient and donor charac-

teristics and the response rate. Factors yielding P values !.10

by univariate analysis were entered into a multivariable logistic

regression model. A stepwise backward procedure was used

with a cutoff significance level of .05 for deletion of factors

from the model. Statistical tests were done using SPSS, version

14.0 (SPSS).

Safety. All patients given at least 1 dose of PCV7 were

included in the safety analysis. Local side effects were recorded

by the patient or their parents on diary cards for 7 days after

each dose. Adverse events were reported according to European

regulations, until 24 months after transplantation.

Role of funding sources. The study was funded by the

EBMT (for data management and statistical analysis) and by

Wyeth Pharmaceuticals (for provision and delivery of the vac-

cines, serum sample collection, and antibody assays). The

EBMT sponsored the trial, organized data collection, owns the

study data, and decided to publish this article.

RESULTS

Patient characteristics. From December 2001 through De-

cember 2004, 158 patients were randomized to receive early

vaccination (75 patients) or late vaccination (83 patients). The

groups were well balanced (table 1). The overall Kaplan-Meier

survival curves, starting at randomization, look similar for both

groups, but the curves cross; therefore, any comparison by a

log-rank test would be irrelevant. In the late vaccination group,

17 patients did not receive the study vaccine because of events

that occurred between inclusion in the study and month 9

(figure 1).

Adherence to the protocol and study withdrawals. Five

patients (1 in the early vaccination group and 4 in the late

vaccination group) were lost to follow-up. In the early vacci-

nation group, 3 patients decided to leave the study before the

third dose of PCV7. At 24 months after transplantation, 44 of

75 patients in the early vaccination group and 42 of 83 patients



ownloaded from


Table 1. Patient characteristics at randomization.

CharacteristicEarly vaccination group

( )n p 75Late vaccination group

( )n p 83All patients( )n p 158

Age at randomization, mean years (range) 38 (7–59) 36 (10–58) 37 (7–59)Aged 116 years at SCT 67 (89) 76 (92) 143Sex

Male 38 (51) 53 (64) 91Female 37 (49) 30 (36) 67

Underlying diseaseAcute leukemia 44 (59) 44 (53) 88Chronic leukemia 15 (20) 21 (25) 36Lymphoma 3 (4) 5 (6) 8Plasma cell disorders 0 (0) 1 (1) 1Myelodysplastic or myeloproliferative syndrome 4 (5) 7 (8) 11Aplastic anemia 7 (9) 5 (6) 12Hemoglobinopathy 2 (3) 0 (0) 2

Status at SCTCR1 or CP1 36 (48) 49 (59) 85CR2 14 (19) 14 (17) 28Advanced disease 12 (16) 9 (11) 21Not available 13 (17) 11 (13) 24

Karnofsky score at randomization70–80 14 (19) 11 (13) 25190 58 (77) 51 (61) 109

Source of stem cellsBone marrow 40 (53) 43 (52) 83Peripheral blood 33 (44) 40 (48) 73Bone marrow and peripheral blood 1 (1) 0 (0) 1Cord blood 1 (1) 0 (0) 1

T cell depletion of the graft 4 (5) 1 (1) 5Donor age, mean years (range) 35.5 (8–61) 37 (4–63) 37 (4–63)Donor sex

Male 46/74 (62) 48/77 (62) 94Female 28/74 (38) 29/77 (38) 57

Donor typeMatched related 45 (60) 53 (64) 98Mismatched related 1 (1) 4 (5) 5Unrelated 29 (39) 26 (31) 55

Prior vaccination of donorPPV23 1/37 (3) 0/38 (0) …PCV7 0/39 (0) 0/39 (0) …

Acute GVHDa

Grade 1 11 (15) 9 (12) 20Grade 2 13 (17) 11 (14) 24Grade 3 4 (5) 3 (4) 7None 47 (63) 54 (70) 101

Chronic GVHDNone 43 (57) 46 (55) 89Limited 26 (35) 26 (31) 52Extensive 6 (8) 11 (13) 17

Duration of follow-up, mean years 1.8 (1–4) 2.1 (0.5–4) …Time from SCT to first dose of PCV7, median days (range) 101 (86–117) 272 (252–342) …

NOTE. Data are no. or proportion (%) of patients, unless otherwise specified. CP1, chronic phase 1; CR1, first complete remission; CR2,second complete remission; GVHD, graft-versus-host disease; PCV7, 7-valent pneumococcal conjugate vaccine; PPV23, 23-valent pneumococcalpolysaccharide vaccine; SCT, stem cell transplantation.

a Data on GVHD data were available for 77 patients in the late vaccination group (and thus for 152 total patients).



ownloaded from



in the late vaccination group were alive. The main reasons for

failure to complete the study were leukemia relapse, death, and

other serious adverse events, none of which was attributed by

the safety committee to the study vaccine.

Primary end point: response rate. One month after the

third PCV7 dose, the response rate was not lower in the early

vaccination group (45 [79%] of 57 patients) than in the late

vaccination group (47 [82%] of 57 patients); the difference was

�3.5% (90% CI, �15.6 to 8.6), which was well above the �20%

noninferiority margin (table 2 and figure 3).

Secondary end points. Before vaccination, the percentage

of patients with antibody titers �0.15 mg/mL was significantly

higher in the early vaccination group than in the late vacci-

nation group (tables 2 and 3). Geometric mean concentrations

at 1 month after the third PCV7 dose were not significantly

different between the early and late vaccination groups for any

of the serotypes. At 24 months after SCT, geometric mean

concentrations and response rates were both significantly lower

in the early vaccination group than in the late vaccination group

for all serotypes except for 2: pn9v, for which only a trend was

observed ( ), and pn23f.P p .06

Before vaccination, the percentage of patients with antibody

titers �0.5 mg/mL was low in both groups (8% of patients in

the early vaccination group and 2% of patients in the late

vaccination group). The percentage of patients with antibody

titers �0.5 mg/mL to all 7 vaccine antigens at 1 month after

the third dose of PCV7 (figure 3) was not lower in the early

vaccination group than in the late vaccination group: the dif-

ference was 1.7% (90% CI, �13.5 to 17), once again above the

�20% margin. At study completion, there was a trend toward

a higher response rate (with 0.5-mg/mL cutoff) in the late vac-

cination group.

Effect of the PPV23 vaccine. A total of 101 patients re-

ceived the PPV23 vaccine. Before PPV23 vaccination, the per-

centage of patients with response (titers, �0.15 mg/mL) was

comparable between the early vaccination group (33 [59%] of

56 patients) and the late vaccination group (35 [66%] of 53

patients) ( ). One month later, the percentage of patientsP p .45

with response was higher in the late vaccination group (42

[88%] of 48 patients) than in the early vaccination group (37

[69%] of 54 patients) ( ). Finally, of 36 patients whoP p .02

did not have a response before PPV23 vaccination, 21 (58%)

remained nonresponders and 15 (42%) had a response 1 month

after receiving PPV23.

Variables correlated with antibody titers �0.5 mg/mL at 1

month after the third PCV7 dose. Because the response rate

was not different between the 2 groups, correlations were stud-

ied in the overall population (table 4). In the univariate analysis,

titers �0.5 mg/mL correlated significantly with donor age, ab-

sence of acute and chronic GVHD, and serum gammaglobulin

level �4 g/L. By multivariate analysis, only chronic GVHD

( ) and donor age ( ) significantly impaired theP p .009 P p .03

response.

S. pneumoniae infections during the study. Three cases of

invasive pneumococcal disease occurred in 2 patients. A patient

in the early vaccination group who experienced bacteremia due

to an unidentified serotype at 1 month after PPV23 injection

survived without complications. A patient in the late vacci-

nation group had 2 episodes of bacteremia with pneumonia

caused by different serotypes (serotype 14, included in PCV7,

and serotype 15, not included in PCV7 but included in PPV23)

at 66 and 154 days after randomization (i.e., before vaccina-

tion), respectively. The patient survived the infections but died

13 days later from leukemia relapse. No patient died of pneu-

mococcal infection.

Safety. Of the 123 reported adverse events, 80 consisted of

injection-site pain and/or skin changes, and 8 were fever within

48 h after vaccination. Most of the 83 serious adverse events

were disease relapses or transplant-related events (i.e, GVHD

[10 patients] and relapse [25 patients]). Seven fatal events oc-

curred in the early vaccination group and 17 in the late vac-

cination group. The safety committee found no evidence that

any of the serious adverse events were due to vaccination.

DISCUSSION

We investigated whether earlier vaccination of allogeneic SCT

recipients with PCV7 was not inferior to later vaccination with

regard to achieving protective antibody titers. Although pneu-

mococcal infection is often viewed as a late complication of

SCT, cases may occur within the first 6 months after trans-

plantation [2], at a time when the patient cannot respond to

PPV23. In the EBMT study [2], 7 of 51 episodes of pneu-

mococcal infection occurred before day 100 after SCT, and 4

of 10 deaths occurred before 6 months after SCT. We found

that the response rate to PCV7 after early vaccination that was

started at 3 months after transplantation was not inferior to

the response rate after vaccination that was started at 9 months

after transplantation. Therefore, earlier vaccination should be

recommended, to ensure protection against early life-threat-

ening pneumococcal disease.

Three smaller studies of PCV7 in allogeneic SCT recipients

have been published [16, 17, 23]. Molrine et al. [16] random-

ized 96 patients, mostly adults, and their donors to receive

PCV7 or not to receive PCV7 before transplantation. All pa-

tients received PCV7 injections at 3, 6, and 12 months after

transplantation. In the 65 assessable patients, donor immuni-

zation improved the rate of response (defined as antibody titers

�0.5 mg/mL) after the first PCV7 dose (67% vs. 36%) but not

after the third dose (75% vs. 64%). Furthermore, donor im-

munization is not always feasible for unrelated donors. Meisel

et al. [17] immunized 53 children with 3 monthly doses of

PCV7 started 6–9 months after transplantation. A response



ownloaded from



Table 2. Patients with response to all 7 antigens included in the pneumococcal conjugate vaccine (PCV7) at differenttime points, with titer cutoffs of 0.15 mg/mL (for primary objective) and 0.5 mg/mL (for post hoc analysis).

Timepoint

Titer �0.15 mg/mL to all 7 antibodies

P

Titer �0.5 mg/mL to all 7 antibodies

PEarly vaccination group Late vaccination group Early vaccination group Late vaccination group

S1 45 (33/74) 9 (6/64) !.001 8 (6/74) 2 (1/64) .08S3 79 (45/57) 82 (47/57) .64 56 (32/57) 54 (31/57) .85a

S6 59 (26/44) 83 (35/42) .013 34 (15/44) 55 (23/42) .054

NOTE. Data are percentage of patients with response (no. with response/total no. of patients), unless otherwise indicated. S1, baseline (i.e.,before vaccination); S3, 1 month after the third dose of PCV7; S6, 24 months after transplantation.

a By x2 test. See figure 3 for equivalence test.

Figure 3. Difference in the response rate between the 2 groups (i.e., rate in the early vaccination group minus rate in the late vaccination group),with 2 different titer cutoffs to define response. PCV7, 7-valent pneumococcal conjugate vaccine.

(antibody titers �0.5 mg/mL) was noted in 74% of patients.

Kumar et al. [23] compared PPV23 with PCV7 given to donors

before transplantation and to transplant recipients at 6 months

after transplantation. Rates of response (antibody titers �0.35

mg/mL and 2-fold increase from baseline) at 12 months among

44 patients were low in both groups but were better with PCV7

vaccination.

Comparisons of studies are hindered by differences in the

definition of a response. We chose the classic cutoff of �0.15

mg/mL [15, 21] to assess our primary end point. After we started

our study, Molrine et al. [16] and Meisel et al. [17] chose titers

of �0.5 mg/mL to define response in SCT recipients; however,

clinical evidence to substantiate the use of this threshold is

lacking [15]. On the basis of a pooled analysis of healthy infants,

the World Health Organization suggested in 2003 that titers

�0.35 mg/mL may be protective [24]. However, no protective

cutoff has been determined for adults. We conducted a post

hoc analysis using 0.5 mg/mL as the cutoff for defining the

response. Whatever the cutoff—0.15 or 0.5 mg/mL—response

rates in our study were consistent with noninferiority of early

vaccination.

Our study also provides the first data, to our knowledge, on

the effects of PPV23 after PCV7 vaccination in SCT recipients.

Among patients who did not have a response after 3 doses of

PCV7, 41% had a response to the PCV7 antigens after receiving

a dose of PPV23. Thus, PPV23 not only may confer protection

to 16 additional serotypes but also may boost the response to

the PCV7 serotypes. This boost effect seems to be lower after

early vaccination than after late vaccination. Administration of

polysaccharide vaccine after conjugate vaccine is also supported

by data showing that a prior exposure to polysaccharide impairs

the response to subsequent conjugate antigen exposure [12,

25].

Importantly, patients in the early vaccination group had

higher prevaccination antibodies, compared with patients in

the late vaccination group. This difference is probably ascribable

to decreasing antibody production after transplantation.

Chronic GVHD and donor age were the only significant

predictors of the immune response. Children usually exhibit

excellent responses to conjugate vaccines, so the favorable effect

of younger donor age is not surprising. Chronic GVHD impairs

immune reconstitution and requires immunosuppressive ther-

apy, 2 factors that explain its negative correlation with an an-

tibody response. The fact that GVHD had no effect in earlier

studies may be ascribable to smaller patient series [16, 17] and

to pediatric cohorts [17] with a low risk of GVHD.



ownloaded from


Tabl

e3.

Conc

entr

atio

nsof

spec

ific

antib

odie

sto

the

7se

roty

pes

incl

uded

inth

e7-

vale

ntpn

eum

ococ

cal

conj

ugat

eva

ccin

e(P

CV7)

atdi

ffere

nttim

epo

ints

inth

e2

patie

ntgr

oups

.

Tim

epo

int,

vacc

inat

ion

grou

p

Geo

met

ricm

ean

conc

entr

atio

n(9

5%C

I),mg/

mL

pn4

pn6b

pn9v

pn14

pn18

pn19

fpn

23f

S1 E

arly

grou

p(

)n

p74

0.13

(0.0

1–1.

87)

0.73

(0.0

9–6.

27)

0.58

(0.0

8–4.

29)

1.9

(0.1

6–23

.04)

0.46

(0.0

8–2.

58)

0.72

(0.0

6–8.

13)

0.55

(0.0

5–5.

76)

Late

grou

p(

)n

p64

0.04

(0.0

1–0.

41)

0.33

(0.0

2–4.

6)0.

23(0

.03–

2.01

)0.

8(0

.05–

12.6

4)0.

23(0

.02–

2.2)

0.31

(0.0

3–3.

51)

0.19

(0.0

2–1.

7)S

2 Ear

lygr

oup

()

np

630.

63(0

.01–

32.8

9)1.

52(0

.05–

49.4

8)1.

05(0

.07–

15.6

1)3.

67(0

.11–

121.

24)

1.05

(0.0

5–20

.85)

1.09

(0.0

5–25

.45)

1.52

(0.0

3–80

.06)

Late

grou

p(

)n

p57

1.01

(0.0

3–38

.96)

1.53

(0.0

4–63

.85)

1.34

(0.0

5–35

.15)

4.04

(0.1

2–13

4.12

)2.

01(0

.03–

126.

88)

1.58

(0.0

6–40

.26)

1.52

(0.0

3–12

3.75

)S

3a Ear

lygr

oup

()

np

570.

98(0

.01–

93)

2.69

(0.0

5–15

.4)

1.57

(0.0

7–34

.93)

4.65

(0.1

1–20

3.15

)1.

9(0

.06–

61.0

9)2.

11(0

.05–

97.6

8)2.

81(0

.03–

254.

14)

Late

grou

p(

)n

p57

2.03

(0.0

4–98

.74)

3.08

(0.0

5–17

9.27

)2.

03(0

.06–

68.2

8)5.

83(0

.24–

140.

91)

3.26

(0.0

6–17

7.33

)3.

24(0

.09–

114.

44)

3.34

(0.0

3–33

4.06

)S

4 Ear

lygr

oup

()

np

560.

39(0

.01–

22.1

8)1.

10(0

.02–

55.0

2)0.

79(0

.03–

20.8

8)2.

85(0

.05–

177.

04)

0.96

(0.0

4–24

.62)

1.12

(0.0

3–41

.48)

1.36

(0.0

2–10

2.47

)La

tegr

oup

()

np

530.

72(0

.02–

21.9

6)1.

71(0

.03–

84.4

2)1.

05(0

.05–

21.8

9)3.

23(0

.13–

81.1

9)1.

67(0

.04–

69.8

5)1.

45(0

.05–

43.0

3)1.

56(0

.02–

122.

43)

S5 E

arly

grou

p(

)n

p54

1.18

(0.0

1–11

2.65

)1.

71(0

.02–

174.

5)1.

61(0

.04–

64.9

8)5.

44(0

.09–

317.

73)

1.55

(0.0

4–65

.76)

3(0

.03–

328.

76)

1.85

(0.0

2–13

9.6)

Late

grou

p(

)n

p48

4.48

(0.1

5–13

4.62

)3.

72(0

.08–

162.

91)

3.52

(0.1

8–69

.14)

7.71

(0.3

6–16

6.81

)4.

33(0

.17–

110.

43)

11.4

9(0

.29–

455.

22)

4.08

(0.0

9–17

6.08

)S

6b Ear

lygr

oup

()

np

440.

39(0

.01–

23.7

5)0.

77(0

.01–

55.0

5)0.

84(0

.03–

20.9

4)2.

49(0

.03–

218.

55)

0.70

(0.0

2–24

.15)

0.91

(0.0

1–58

.04)

1.06

(0.0

2–58

.46)

Late

grou

p(

)n

p42

1.65

(0.0

6–42

.2)

2.03

(0.0

4–93

.37)

1.75

(0.1

–30.

62)

5.09

(0.2

6–97

.65)

2.24

(0.1

–52.

39)

4.89

(0.1

4–17

6.14

)2.

22(0

.04–

118.

1)

NO

TE

.S

1,ba

selin

e(i.

e.,

befo

reva

ccin

atio

n);

S2,

afte

rth

ese

cond

dose

ofP

CV

7;S

3,af

ter

the

third

dose

ofP

CV

7;S

4,be

fore

the

23-v

alen

tpn

eum

ococ

calp

olys

acch

arid

eva

ccin

e(P

PV

23);

S5,

1m

onth

afte

rP

PV

23;

S6,

24m

onth

saf

ter

tran

spla

ntat

ion.

aP

valu

esfo

rco

mpa

rison

ofth

eea

rlyva

ccin

atio

ngr

oup

with

the

late

vacc

inat

ion

grou

pfo

rsp

ecifi

can

tibod

ies

are

asfo

llow

s:fo

rpn

4,;

for

pn6b

,;

for

pn9v

,;

for

pn14

,;f

orP

p.0

9P

p.6

2P

p.4

0P

p.4

6pn

18,

;fo

rpn

19f,

;an

dfo

rpn

23f,

.P

p.1

7P

p.3

4P

p.7

9b

Pva

lues

for

com

paris

onof

the

early

vacc

inat

ion

grou

pw

ithth

ela

teva

ccin

atio

ngr

oup

for

spec

ific

antib

odie

sar

eas

follo

ws:

for

pn4,

;for

pn6b

,;f

orpn

9v,

;for

pn14

,;f

orP

p.0

01P

p.0

25P

p.0

6P

p.0

4pn

18,

;fo

rpn

19f,

;an

dfo

rpn

23f,

.P

p.0

04P

!.0

01P

p.1

2



ownloaded from



Table 4. Impact of pretransplantation, transplant-related, and posttransplantation factors on the response rate at 1 monthafter the third dose of 7-valent pneumococcal conjugate vaccine (PCV7), by univariate analysis.

FactorTotal no.

of patientsNo. of patients

without responseNo. (%) of patients

with response P

Patient age at randomization, years .55!37 55 23 32 (58)�37 59 28 31 (53)

Patient sex .84Male 66 29 37 (56)Female 48 22 26 (54)

Donor age, years .003!37 59 18 41 (69)�37 51 30 21 (41)

Donor sex .14Male 74 29 45 (61)Female 39 21 18 (46)

Diagnosis .73Acute leukemia 65 30 35 (57)Other 49 21 28 (53)

Complete remission at transplantation .97No 20 9 11 (55)Yes 88 40 48 (55)

Source of stem cells .64Bone marrowa 61 26 35 (57)Peripheral blood 51 24 27 (53)

Blood lymphocyte count, cells/mm3 .89!200 90 39 51 (57)�200 20 9 11 (55)

Serum gammaglobulin level, g/L .044!4 21 13 8 (38)�4 75 28 47 (63)

Acute GVHD .03No 75 28 47 (63)Yes 39 23 16 (41)

Chronic GVHD before the primary end pointNo 58 16 42 (72) .002Yes 41 24 17 (41)

Any immunosuppressive drug at time of first PCV7 dose .25Yes 89 42 47 (53)No 21 7 14 (67)

NOTE. Response was defined as antibody titers �0.5 mg/mL to all 7 serotypes in PCV7. The primary end point was 1 month after the thirddose of PCV7. GVHD, graft-versus-host disease.

a Only 1 patient in the early vaccination group received a transplant of cord blood; this patient was analyzed in the bone marrow transplantgroup.

Our study has several limitations. First, antibody titers to 6

of the 7 PCV7 serotypes decreased faster in the early vaccination

group than in the late vaccination group. Although this finding

does not argue against the benefit of early vaccination, it sug-

gests a need for a PCV7 or PPV23 boost during the second

year. Indeed, we compared 2 different timings of vaccination,

and the times from PPV23 vaccination to study completion

were different (12 vs. 6 months). Second, the benefits of PCV7

may vary across regions [26, 27] and according to the vaccine

policy in the community [28–30]. Third, we did not use the

0.35-mg/mL cutoff proposed by the World Health Organization

[24] for infants because we had, at that time, already defined

our primary end point. We are aware that the minimum an-

tibody titer associated with long-term clinical protection may

be within the 0.15–0.5 mg/mL range in healthy infants [21],

but it is unknown whether this applies to immunocompro-

mised adults.

On the basis of our results, we recommend early adminis-



ownloaded from



tration of 3 doses of PCV7 at monthly intervals, starting 3

months after transplantation, followed by PPV23. Whether a

booster injection during the second posttransplantation year

prevents the antibody decrease seen after early vaccination at

24 months cannot be determined from our study but should

be a logical conclusion until further data are obtained.

Acknowledgments

We are indebted to the members of the Safety Committee: Dan Engel-hard, MD, Hadassah University Hospital, Israel; Pierre Reusser, MD, Por-rentruy, Switzerland; and Philippe Reinert. Creteil, France. We are gratefulto Eric Bonnet, MD, Clinical Trial Coordinator of the study at WyethPharmaceuticals, and to all the physicians, pharmacists, data managers, andnurses of the study transplant centers, for their participation in the study.We thank Branda Hu, at the Wyeth US laboratory, for performing theELISA tests, as well as Kim Champion and the staff of the European Groupfor Blood and Marrow Transplantation London Office for randomizing thepatients.

Financial support. European Group for Blood and Marrow Trans-plantation (grants for data management and statistical analysis) and WyethPharmaceuticals (provision and delivery of the vaccines, serum samplecollection, and antibody assays).

Potential conflicts of interest. All authors: no conflicts.

References

1. Kulkarni S, Powles R, Treleaven J, et al. Chronic graft versus host diseaseis associated with long term risk for pneumococcal infections in recipientsof bone marrow transplants. Blood 2000; 95:3683–6.

2. Engelhard D, Cordonnier C, Shaw PJ, et al. Early and late invasive pneu-mococcal infection following stem cell transplantation: a European BoneMarrow Transplantation survey. Br J Haematol 2002; 117:444–50.

3. Winston DJ, Schiffman G, Wang DC, et al. Pneumococcal infections afterhuman bone marrow transplantation. Ann Intern Med 1979; 91:835–41.

4. Giebink G, Warkentin P, Ralsay N, et al. Titers of antibody to pneu-mococci in allogeneic bone marrow transplant recipients. J Infect Dis1986; 154:590–6.

5. Hammarstrom V, Pauksen K, Azinge J, Oberg G, Ljungman P. Pneu-mococcal immunity and response to immunization with pneumococcalvaccine in bone marrow transplant patients: the influence of graft versushost reaction. Support Care Cancer 1993; 1:195–9.

6. Lortan JE, Vellodi A, Jurges ES, Hugh-Jones K. Class- and subclass-specificpneumococcal antibody levels and response to immunization after bonemarrow transplantation. Clin Exp Immunol 1992; 88:512–9.

7. Winston DJ, Ho WG, Schiffman G, Champlin RE, Feig SA, Gale RP.Pneumococcal vaccination of recipients of bone marrow transplants. ArchIntern Med 1983; 143:1735–7.

8. Avanzini MA, Carra AM, Maccario R, et al. Antibody response to pneu-mococcal vaccine in children receiving bone marrow transplantation. JClin Immunol 1995; 15:137–44.

9. Molrine DC, Guinan EC, Antin JH, et al. Donor immunization withHaemophilus influenzae type b (HIB)-conjugate vaccine in allogeneicbone marrow transplantation. Blood 1996; 87:3012–18.

10. Ljungman P, Engelhard D, Da La Camara R, et al. Vaccination of stemcell transplant recipients: recommendations of the Infectious DiseasesWorking Party of the EBMT. Bone Marrow Transplant 2005; 35:737–46.

11. Centers for Disease Control and Prevention. Guidelines for preventingopportunistic among hematopoietic stem cell transplant recipients: rec-ommendations of the CDC, the Infectious Diseases Society of America,and the American Society of Blood and Marrow Transplantation. MMWRMorb Mortal Wkly Rep 2000; 49(RR-10):1–125.

12. Barra A, Cordonnier C, Precziosi MP, et al. Immunogenicity of Hae-mophilus influenzae type b conjugate vaccine in allogeneic bone marrowrecipients. J Infect Dis 1992; 166:1021–8.

13. Guinan EC, Molrine DC, Antin JH, et al. Polysaccharide conjugate vac-cine responses in bone marrow transplant patients. Transplantation1994; 57:677–84.

14. Parkkali T, Kayhty H, Ruutu T, Volin L, Eskola J, Ruutu P. A comparisonof early and late vaccination with Haemophilus influenzae type b andpneumococcal polysaccharide vaccines after allogeneic bone marrowtransplantation. Bone Marrow Transplant 1996; 18:961–7.

15. Shinefield HR, Black S, Ray P, et al. Safety and immunogenicity of hep-tavalent pneumococcal CRM197 conjugate vaccine in infants and tod-dlers. Pediatr Infect Dis J 1999; 18:757–63.

16. Molrine D, Antin JH, Guinan EC, et al. Donor immunization with pneu-mococal conjugate vaccine and early protective antibody responses fol-lowing allogeneic hematopoietic stem cell transplantation. Blood 2003;101:831–6.

17. Meisel R, Kuypers L, Dirksen U, et al. Pneumococcal conjugate vaccineprovides early protective antibody responses in children after related andunrelated allogeneic stem cell transplantation. Blood 2007; 109:2322–6.

18. Storek J, Mendelman PM, Witherspoon RP, et al. IgG response to pneu-mococcal polysaccharide-protein conjugate appears similar to IgG re-sponse to polysaccharide in bone marrow transpant recipients and healthyadults. Clin Infect Dis 1997; 25:1253–5.

19. Armitage JO. Bone marrow transplantation. N Engl J Med 1994; 330:827–38.

20. Conception NF, Frash CE. Pneumococcal type 22F polysaccharide ab-sorption improves the specificity of a pneumococcal polysaccharide en-zyme-linked immunosorbent assay. Clin Diagn Lab Immunol 2001; 8:266–72.

21. Black S, Shinefield H, Fireman B, et al. Efficacy, safety and immuno-genicity of heptavalent pneumococcal conjugate vaccine in children.Northern California Kaiser Permanente Vaccine Study Center Group.Pediatr Infect Dis J 2000; 19:187–95.

22. Pocock SJ. The pros and cons of non-inferiority trials. Fundam ClinPharmacol 2003; 17:483–90.

23. Kumar D, Chen MH, Welsh B, et al. A randomized, double blind trialof pneumococcal vaccination in adult allogeneic stem cell transplantdonors and recipients. Clin Infect Dis 2007; 45:1576–82.

24. World Health Organization Expert Committee on Biological Stan-dardization. Pneumococcal conjugate vaccine: recommendations forproduction and quality control of pneumococcal conjugate vaccines.2003. Available at: http://www.who.int/biologicals/pneumococcal_conjugate_Vaccines_Recomm_Nov_2003. Accessed 1 September2008.

25. De Roux A, Schmoele-Thoma B, SIber GR, et al. Comparison of pneu-mococcal conjugate polysaccharide and free polysaccharide vaccines inelderly adults: conjugate vaccine elicits improved antibacterial immuneresponses and immunological memory. Clin Infect Dis 2008; 46:1015–23.

26. Debbache K, Varon E, Hicheri Y, et al. Epidemiology of S. pneumoniaeinfection in hematology patients in Ile de France: serotypes covered bythe polysaccharide 23-valent (PS23) and the heptavalent conjugate vac-cine [abstract]. Blood 2007; 110:4963.

27. Hausdorff WP, Hajieh R, Al-Mazrou A, Shibl A, Soriano-Gabarro M.The epidemiology of pneumococcal, meningoccal, and Haemophilus dis-ease in the Middle East and North Africa (MENA) region—current statusand needs. Vaccine 2007; 25:1935–44.

28. Kyaw MH, Lynfield R, Schaffner W, et al. Effect of introduction of thepneumococcal conjugate vaccine on drug-resistant Streptococcus pneumo-niae. N Engl J Med 2006; 354:1455–63.

29. Albrich WC, Baughman W, Schmotzer B, Farley MM. Changing char-acteristics of invasive pneumococcal disease in metropolitan Atlanta,Georgia, after introduction of a 7-valent pneumococcal vaccine. ClinInfect Dis 2007; 44:1569–76.

30. Farrell DJ, Klugman KP, Pichichero M. Increased antimicrobial resistanceamong nonvaccine serotypes of Streptococcus pneumoniae in the pediatricpopulation after the introduction of 7-valent pneumococcal vaccine inthe United States. Pediatr Infect Dis J 2007; 26:123–8.



ownloaded from


Randomized Study of Early versus Late Immunization with Pneumococcal Conjugate Vaccine after...

Documents

Transcript of Randomized Study of Early versus Late Immunization with Pneumococcal Conjugate Vaccine after...