Contents lists available at SciVerse ScienceDirect
Vaccine
jou rn al h om epa ge: www.elsev ier .com/ locate /vacc ine
eview
re immunosuppressive medications associated with decreased responses tooutine immunizations? A systematic review
ikhil Agarwala, Kevin Ollingtona, Marc Kaneshiroa, Robert Frenckb, Gil Y. Melmeda,∗
Division of Gastroenterology, Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, United StatesDepartment of Pediatrics, Cincinnati Children’s Hospital, Cincinnati, OH, United States
r t i c l e i n f o
rticle history:eceived 27 July 2011eceived in revised form6 November 2011ccepted 29 November 2011vailable online 21 December 2011
eywords:accines
nfluenza vaccinesneumococcal vaccinesmmunosuppressive medicationsumor necrosis alphaolid organ transplantheumatoid
nflammatory bowel diseaserohn’slcerative colitis
a b s t r a c t
Background: Long-term immunosuppressive medications are being used more commonly for a varietyof medical conditions, including immune-mediated diseases and organ transplantation. While thesemedications are often necessary, they are associated with an increased risk of serious infections. Vacci-nation may be a way to prevent a variety of infections but vaccine responses among patients receivingimmunosuppressive therapies have been variable.Purpose: To systematically review the literature describing immune responses among patients onimmunosuppressive therapies to vaccinations including influenza, pneumococcal, meningococcal, hep-atitis A and B, tetanus toxoid, pertussis, varicella, and zoster.Data sources: English language citations in the MEDLINE and EMBASE databases from 1985 to 2010.Study selection: Two reviewers independently screened titles and abstracts to identify prospective, con-trolled studies reporting pre- and post-vaccination titers of recommended vaccines in patients receivinglong-term immunosuppressive therapies for full-text review.Data extraction: Three reviewers independently assessed study characteristics including treatment regi-mens and pre- and post-vaccination titers.Data synthesis: Of the 972 identified titles, fifteen met inclusion criteria. Ten studies assessed the effects ofimmunosuppressive medications on responses to influenza vaccine, four studies investigated responsesfollowing pneumococcal vaccination, and one study assessed both influenza and pneumococcal vacci-nation. Five of the studies that evaluated influenza vaccination showed partially diminished responsesamong individuals receiving immunosuppressive therapies, while one of the pneumococcal vaccine stud-
ies showed significantly decreased responses following vaccination. Patients treated with more than oneimmunosuppressive medication were the least likely to respond to vaccination.Limitations: The heterogeneity of reported outcomes limits generalizeability.Conclusions: Immunosuppressive therapy, particularly combination regimens, may blunt response toinfluenza and pneumococcal vaccinations. To ensure the best chance of response, immunizations shouldbe administered prior to initiation of immunosuppressive medications whenever possible.
As part of the armamentarium against many disease states, moreeople are receiving an increasing array of immunosuppressivegents such as corticosteroids, 6-mercaptopurine, azathioprine,ethotrexate, cyclosporine, tacrolimus, mycophenolate mofetil
nd various monoclonal antibodies (anti-TNF�, anti-CD20 andthers). Immune suppression increases the risk of people devel-ping various vaccine-preventable infections, including influenza,neumococcal pneumonia, varicella, herpes zoster and heptati-is B [1–4]. Patients with inflammatory diseases (inflammatoryowel disease and rheumatological disorders) and organ trans-lant recipients commonly require long-term immunosuppressionhich may further increase their risk for developing serious infec-
ions. In an attempt to decrease this risk, the Centers for Diseaseontrol and Prevention (CDC) recommend that immunosuppressedatients be brought up to date against vaccine preventable infec-ions [5,6]. However, the response to immunizations in patientseceiving immunosuppressive therapy is unclear. It also is likelyhat the response to immunization varies depending on the specificmmunosuppressive drug or regimen administered. This has ledome groups to advocate checking post-vaccination titers to ensuren adequate immunological response following immunization [7].
To better delineate the response to vaccination among immuno-uppressed patients and if response varies by immunosuppressiveegimen, we performed a systematic review of prospective, con-rolled studies that assessed pre- and post-vaccination titers amonghildren and adults receiving various immunosuppressive thera-ies.
. Methods
.1. Search strategy
We conducted a systematic search of published, English-anguage studies from January 1, 1985 through January 2010 using
EDLINE and EMBASE, based on methods outlined in the Cochraneollaborative Working Group on Systematic Reviews [8]. A searchtrategy was created to capture commonly used immunosuppres-ive therapies, conditions in which such therapies might be used,nd routine vaccines (Table 1).
.2. Study selection and data collection process
Two independent reviewers (NA, KO) screened article titlesnd abstracts in duplicate to assess inter-observer reliability. After00 articles were screened in this manner, it was found that theappa for reliability was >0.87 and the remaining article titles
nd abstracts were divided between the screeners for review.rticles were excluded from full-text review for any of the fol-
owing: (i) not written in English, (ii) not a randomized controlledrial, controlled clinical trial or prospective cohort/case control
study, (iii) not involving humans, (iv) not involving immunosup-pressive medications, or (v) not dealing with immunizations forvaccine-preventable infections. Discrepancies between reviewerswere resolved by consensus (NA, KO, GYM, MK).
Articles passing primary screening were selected for full textreview to determine if they met the following additional inclu-sion criteria: (i) specific immunosuppressive medications wereidentified, (ii) responses in patients on immunosuppressive med-ications were evaluated relative to a comparator group and (iii)quantitative baseline and post-vaccination antibody titers werereported.
2.3. Data items and analysis
Abstracted data items included geographic region, disease entityand duration of disease, study design, delineation of study and con-trol arms, mean age of patients/subjects within each group, specificimmunosuppressive medications administered, vaccine adminis-tered, timing of post-vaccination serologic assessment, delineationof vaccine serotypes assessed, definitions of vaccine response,and pre- and post-vaccination geometric mean titers (GMT) foreach serotype. Descriptive statistics were performed using Excel2003 (Microsoft Corp., WA) and JMP 8.0 (The SAS Institute,Carey, NC).
3. Results
3.1. Search results
The MEDLINE and EMBASE database searches were performedfor a 15 year period ending January 2010, and yielded 972titles (Fig. 1). Of these, 45 abstracts met inclusion criteria, andthese manuscripts were reviewed. Among these, 29 studies wereexcluded for not meeting inclusion criteria, resulting in 15 pub-lications included in the final review. [9–23] While all studieswere prospective, only 2 studies involved treatment randomization(immunosuppression or placebo) [9,18], one of which was blinded[9]. The remaining studies involved assessment of vaccine responsein prospectively stratified subject groups or post hoc analyses ofheterogenous populations receiving various immunosuppressiveregimens.
3.2. Vaccines
Our search yielded 11 publications that assessed influenza vac-cination (Table 2 ) and five studies that evaluated pneumococcalvaccination (Table 3). No studies that met inclusion criteria were
found evaluating tetanus, hepatitis B, hepatitis A, diphtheria, orhaemophilus influenza type B. Quantitative meta-analysis was notperformed due to heterogeneity of the patient populations, medi-cation exposures, and outcome measures in the identified studies.
N. Agarwal et al. / Vaccine 30 (2012) 1413– 1424 1415
Table 1Search strategy.
#1Crohn’s OR colitis OR arthritis OR sarcoid OR lupus OR SLE OR sjogren OR psoriasis OR ankylosing OR myositis OR sclerosis OR scleroderma OR polymyalgia OR amyloid
OR transplant OR cancer OR neoplasm OR malignancy OR arteritis OR autoimmune OR behcet OR demyelinating OR eczema OR glomerulonephritis OR gravis ORpemphigus OR reiter OR thrombocytopenia 19:18:55 3111039
#2azathioprine OR thiopurine OR mercaptopurine OR methotrexate OR prednisone OR tacrolimus OR sirolimus OR cyclosporin OR methylprednisolone OR hydrocortisone
OR infliximab OR adalimumab OR certolizumab OR natalizumab OR rituximab OR thalidomide OR mycophenolate OR ATG OR muromonab OR basiliximab ORthymocyte OR daclizumab OR etanercept
#3vaccine OR vaccination OR immunization OR inoculation OR pneumonia OR hepatitis OR varicella OR influenza OR measles OR mumps OR rubella OR pneumonia OR
haemophilus OR diphtheria OR pertussis OR tetanus OR zoster OR papillomavirus OR polio OR pertussis pneumovax OR zostavax OR varivax OR twinrix OR hib OR dtapOR dtp OR mmr OR opv OR ipv OR comvax OR pedvax OR flu OR polio
#4Search #1 AND #2 AND #3
#5Search (((letter[Publication Type] OR editorial[Publication Type]) OR review[Publication Type]) OR news[Publication Type])
#6Search #4 NOT #5
3
mp
#7Limits: Humans, English
.3. Influenza vaccination studies
Of the 11 studies describing the effects of immunosuppressiveedications on response to influenza vaccination, five involved
atients with rheumatic diseases, four assessed patients with
Fig. 1. Flow diagram of studie
solid organ transplants and two were performed in patients withinflammatory bowel disease (Table 2). These studies involved intra-
muscular administration of the trivalent influenza vaccine (TIV),with blood samples drawn 4–6 weeks later for serologic assess-ment.
s selected for inclusion.
1416N
. A
garwal
et al.
/ V
accine 30 (2012) 1413– 1424
Table 2Studies evaluating immune response to trivalent inactivated influenza vaccine.
Study Geographic region Study groups Disease Mean age(years)
Disease duration(years)
Subjects(n)
Kaine et al. [9] United States A: RA + adalimumab Rheumatoid arthritis 52.2 Not specified 99B: RA + placebo 51.1 109
Oren et al. [10] Israel A: RA − rituximab Rheumatoid arthritis 64 15.5 29B: RA + Rituximab 53 16.5 14C: Healthy controls 58 – 21
Gelinck et al. [11] The Netherlands A: Disease + anti-TNF Rheumatoid arthritis, otherchronic arthritis, Still’s disease,IBD
Rituximab group had similarresponse rates in 2/3 antigenstested. One of the antigens(California) was significantly lowerthan other RA pts and controls.
93 86 5067 – –
N.
Agarw
al et
al. /
Vaccine
30 (2012) 1413– 14241417
Gelinck et al. [11] 73 A/H3N2, A/H1N1,Influenza B
IFX/ETA/ADA 71 20 Azathioprine Group A < Groups B, Cin 2/3 of tested GMTs.
Anti-TNF group had significantlylower response rates in 2/3 oftested antigens; however,protective levels ofpost-vaccination titers were stillachieved.
69 – 75 23 Azathioprine78 – – – –
Mamula et al. [12] 41 A/H1N1, A/H3N2,B/Hong Kong
Infliximab 38 25 6-Mercaptopurine Group A < Groups B,C,Din 2/3 of tested GMTs.
Transplant recipients had lowerresponse rates to vaccine.Cyclosporine had lower antibodyresponse when compared totacrolimus.
57 –Fomin et al. [22] 77 A/Panama, A/New
Caledonia, B/HongKong
IFX/ETA in33% of pts
68 59 Sulfasalazine,Leflunomide
Group A < Group B in1/3 tested GMTs.
Humoral response preseved in RApatients on immunosuppressivemedications including TNFblockers.
70 – – – –Scharpe et al. [15] 36 A/H1N1, A/H3N2,
B/ShangdongNone None 58 Azathioprine, MMF,
Cyclosporine,tacrolimus, sirolimus
Group A = Group B Immunization was safe andefficacious in transplant recipientsdespite immunosuppressivetherapy. MMF was assoc. withlower response rates, but wasnonetheless seroprotective.
71 –Kapetanovic et al. [23] 76 A/H1N1, A/H3N2, B IFX/ETA – 50 Other DMARDs not
specifiedGroup C > Group A,B MTX alone had better response
rates than TNF blockers alone orTNF blockers + MTX, but TNFblocker group still seroprotective.
Group D had highpre-vaccination titersand were excluidedfrom statisticalanalysis.
1418 N. Agarwal et al. / Vaccine 3
Tabl
e
2
(Con
tinu
ed
)
Stu
dy
%
Fem
ale
Vac
cin
e
sero
typ
es
TNF-
inh
ibit
oru
sed
Met
hro
trex
ate
(%)
Ster
oid
s(%
)O
ther
imm
un
osu
pp
ress
ive
Med
s.
use
d
Vac
cin
e
resp
onse
aA
uth
or’s
con
clu
sion
s
70IF
X/E
TA10
052
–68
–
100
51
–74
––
––
Lu
et
al. [
17]
40
A/H
1N1,
A/H
3N2,
B/M
alay
sia
–
–
–
–
Gro
up
A
=
Gro
up
B
=
Gro
up
C.
Sero
pro
tect
ion
agai
nst
A
stra
ins
>
B
stra
in
IBD
pat
ien
ts
resp
ond
edap
pro
pri
atel
y
to
infl
uen
zava
ccin
atio
n
irre
spec
tive
of
wh
eth
erth
ey
wer
e
on
imm
un
osu
pp
ress
ive
ther
apy.
Pati
ents
on
TNF
inh
ibit
ors
had
low
er
resp
onse
rate
s
toin
flu
enza
stra
in
B.
40IF
X/A
DA
/Th
alid
omid
eN
ot
spec
ific
8
Tacr
olim
us,
Aza
thio
pri
ne,
6-M
erca
pto
pu
rin
eN
/A
–
–
–
–
GM
T,
geom
etri
c
mea
nt
tite
r;
IBD
, in
flam
mat
ory
bow
el
dis
ease
;
IM, i
mm
un
omod
ula
tor;
IFX
, in
flix
imab
;
ETA
, eta
ner
cep
t;
AD
A, a
dal
imu
mab
;
MM
F,
myc
oph
enol
ate
mof
etil
;
PCA
, pre
dn
isol
one,
cycl
osp
orin
e,
azat
hio
pri
ne;
PCM
,p
red
nis
olon
e,
cycl
osp
orin
e,
MM
F;
DM
AR
DS,
dis
ease
mod
ifyi
ng
and
ti-r
heu
mat
ic
dru
gs.
aV
acci
ne
resp
onse
as
defi
ned
by
the
auth
ors:
eith
er
fou
r-fo
ld
incr
ease
or
abso
lute
HI t
iter
>1:4
0.
0 (2012) 1413– 1424
3.3.1. Rheumatic diseaseFomin et al. reported the safety and immunogenicity of
influenza vaccine in 82 adult patients with rheumatoid arthritiscompared to 30 age- and gender- matched healthy con-trols. Subjects with rheumatoid arthritis were receiving variousimmunosuppressive regimens most were on methotrexate. Thepercentage of sero-responders for each of the influenza A strainsin the vaccine were equivalent between the RA group and the nor-mal controls. In contrast, a lower percentage of sero-responders tothe influenza B strain was found in the RA group relative to controls(67% vs 87%, respectively). The authors concluded that administra-tion of TIV to patients with RA receiving immunosuppression wassafe and immunogenic and should become part of the routine careof patients with RA [22].
Kapetanovic et al. assessed TIV responses in 149 patients withRA compared to 18 health volunteers. Subjects were either on anti-TNF agents alone, methotrexate (MTX) alone or on combinationimmunosuppression with MTX and anti-TNF agents. Subjects onmethotrexate alone had significantly higher post-vaccination titersthan subjects on anti-TNF agents alone or combined immunosup-pression with anti-TNF agents and MTX [16].
Kaine et al. studied the response to TIV in 99 RA patientsrandomized to adalimumab and 109 RA patients random-ized to placebo. Both groups included subjects on concomitantimmunosuppression with 54.9% on MTX, 23.6% on anotherdisease-modifying anti-rheumatic drug (DMARD) and 45.7% onconcomitant steroids. Post-vaccination HAI titers were lower in theadalimumab group and a smaller percentage of subjects achievedsero-conversion in at least two of the three antigens, although thiswas not statistically significant. In addition, in sub-group analy-sis, combined immunosuppression with adalimumab and MTX wasassociated with a lower vaccine response [9].
Gelinck et al. looked at TIV response in three groups: 64 patientswith RA, Still’s disease, IBD and other rheumatic diseases on anti-TNF agents, 48 patients not on anti-TNF therapy, and 18 healthycontrols. Post-vaccination GMTs were significantly lower in sub-jects treated with anti-TNF agents relative to those not on anti-TNFtherapy and healthy controls in two of the three tested antigens[11].
Oren et al. assessed TIV response in 14 RA patients on rituximab,29 RA patients on various immunomodulators, and 21 healthy con-trols. Vaccine response to more than one strain was obtained in 14%of the rituximab patients, 48% of the non-rituximab patients and40% of health controls (p = 0.53). Fewer rituximab-treated patientshad a serological response to the A/California strain. (p ≤ 0.05) rel-ative to those treated with conventional DMARDs [10].
In summary, 3 of 5 studies of TIV administered to patientswith reheumatologic diseases showed that vaccine responses weregenerally robust across all immunosuppressive subgroups andappeared to confer at least some degree of immunity (definedas HI > 1:40). However in 2 studies, while subjects still developedseroprotection at equivalent rates as controls, those receiving ritux-imab or anti-TNF agents had a lower GMT as compared to controls[10,11].
3.3.2. Solid organ transplantationFour studies looked at TIV responses in recipients of solid organ
transplants, including kidney (2 studies), lung (1 study), and a com-bined group of kidney, lung, and heart transplant recipients (1study).
Scharpe et al. reported TIV response in renal transplant recip-ients. 165 renal transplant recipients and 41 healthy volunteers
received one intramuscular injection of TIV. Transplant patientswere on different combined immunosuppressive regimens includ-ing corticosteroids, azathioprine, mycophenalate mofetil (MMF),cyclosporine, tacrolimus and sirolimus. Post-vaccination GMTs
N.
Agarw
al et
al. /
Vaccine
30 (2012) 1413– 14241419
Table 3Studies evaluating immune response to 23-valent pneumococcal polysaccharide vaccine.
Study Geographicregion
Study groups Disease Mean age(years)
Disease duration(years)
Subjects(n)
Kaine et al. [9] United States A: RA + AdalimumabB: RA + placebo
Rheumatoid Arthritis 52.251.1
Not specified 99
109Mease et al. [18] United States A: PsA + anti-TNF
B: PsA + placeboPsoriatic arthtritis 47 (for all
pts withPsA)
9 (for all pts withPsA)
9490
Kapetanovic et al. [16] Sweden A: RA + anti-TNFB: RA + anti-TNF +MTXC: RA + MTXD: healthy controls
Rheumatoid arthritis 53.752.861.330.3
20.810.87.0–
62503718
Elkayam et al. [19] Israel A: arthritis + anti-TNFB: arthritis − anti-TNF
Rheumatoid arthritis,ankylosing spondylitis
41.048.3
6.510.4
1617
Melmed et al. [20] United States A: IBD + anti-TNF + IMB: IBD (ASA only)C: healthy controls
Adalimumab does not impairresponse to pneumococcalvaccination when compared toplacebo. Patients withcombined immunosuppressionwere less likely to respond tovaccination.
Mease et al. [18] 48 (for all ptswith PsA)
9V, 14, 18C,19F, 23F
Etanercept–
45 (for all pts withPsA)
12 (for all pts withPsA)
None specified Group A = Group B Etanercept does not impairpneumococcal vaccineresponse in pts with PsA whencompared to placebo. MTX wasassociated with decreasedvaccine response.
Kapetanovic et al. [16] 76706874
6B, 23F IFX/ETAIFX/ETA––
–100100–
505251–
Other DMARDs notspecified–––
GroupA > Group D > GroupB > Group C
Anti-TNF agents did not affectvaccine response in pts withRA. However, MTX wasassociated with lower vaccineresponse rates.
Elkayam et al. [19] 6276
2, 4, 7F, 8, 9N,14, 23F
IFX/ETA–
6562
19.055.3
None specified Group A < Group B(statistically significantin 2/7 tested GMTs)
Anti-TNF agents did notsignificantly reduce the meanantibody response topneumococcal vaccination;however, a substantialproportion of pts did notrespond adequately.
Melmed et al. [20] 254447
6B, 9V, 14, 19F,23F
IFX/ADA––
15––
None 6-Mercaptopurine Group A < GroupB = Group C(45% vs 80% vs 84%)
Pts with IBD on anti-TNFagents and combinedimmunomodulator therapyhave diminished vaccineresponse compared tonon-immunosuppressed IBDpts and healthy controls.
a Vaccine response as defined by the authors: >2-fold increase in post-vaccination GMT.
1 ccine 3
ap3rtea
papvtMppt
rwcica
rpwtvp
piptcta
3
Icwioicp(
twp(wtp
3
oac
420 N. Agarwal et al. / Va
gainst the B strain were significantly higher in renal trans-lant recipients compared to healthy controls. (95.1 ± 140.1 vs9.1 ± 32.3, p = 0.01). Multivariate analysis showed that vaccineesponse was 2.6–5-fold lower in renal transplant recipientsaking MMF relative to other immunosuppressive agents. How-ver, despite this diminished response, patients taking MMF stillchieved protective titers 74–88% of the time [15].
Smith et al. determined TIV response in 25 renal trans-lant recipients treated with prednisolone, cyclosporine, andzathioprine (AZA), 13 renal transplant recipients treated withrednisolone, cyclosporine and MMF, and 20 normal controls. Post-accination GMTs for healthy controls were greater than bothransplant groups. Vaccine response was significantly lower in the
MF-based regimen compared to the AZA-based regimen trans-lant group and healthy controls (both p < 0.01). Additionally, theroportion of patients failing to mount a four-fold increase in titero any of the three antigens was highest in the MMF group [14].
Mazzone et al. compared TIV response in 43 lung transplantecipients relative to 21 healthy controls. Lung transplant recipientsere on combined immunosuppression regimens including corti-
osteroids, tacrolimus, cyclosporine, azathioprine and MMF. GMTsn the transplant group were significantly lower than in healthyontrols. Subgroup analysis revealed that cyclosporine was associ-ted with lower vaccine response rates than tacrolimus [21].
Blumberg et al. assessed for differences in response among 68ecipients of solid organ transplants (heart, liver, kidney, lung) com-ared to 29 healthy controls. Transplant recipients were treatedith azathioprine, cyclosporine and prednisone. The proportion of
ransplant recipients who achieved a four-fold increase in post-accination titers in all 3 antigens was 16–26%, compared to aroportion of 72–86% in the control population (p < 0.001) [13].
In summary, for organ transplant recipients on immunosup-ressive therapies, immune response to influenza vaccination
s significantly diminished in 3 of 4 studies, although manyatients were still able to develop some degree of seroprotec-ive post-vaccination titers. Certain regimens including MMF andyclosporine may have a greater impact on vaccine responseshan regimens without these medications, although heterogeneitycross these studies limits generalizeability.
.3.3. Inflammatory bowel diseaseTwo studies were identified that assessed TIV in children with
BD. Mamula et al. compared influenza vaccine response in 51hildren with IBD compared to 29 controls. Subjects with IBDere on immunomodulators alone or in combination with inflix-
mab, or on non-immunosuppresive therapy. For IBD patientsn combined immunosuppression with immunomodulators andnfliximab, post-vaccination GMTs were significantly lower thanontrols. Furthermore, response rates were overall lower in IBDatients and were specifically lower for strain B/Hong Kongp = 0.01) [12].
Lu et al. also assessed TIV response in 146 children with IBD;hese were compared to 76 historical healthy controls. IBD patientsere categorized as non-immunosuppressed or immunosup-ressed (either tacrolimus, anti-TNF agents, immunomodulatorsAZA or methotrexate) or corticosteroids). Although response ratesere similar between the three groups, subgroup analysis revealed
hat IBD patients on anti-TNF therapy had a significantly lowerrotection rate to strain B relative to controls (p = 0.03) [17].
.4. Pneumococcal vaccination studies
Table 3 summarizes the studies that investigate the effectsf immunosuppressive medications in response to intramusculardministration of the 23-valent pneumococcal polysaccharide vac-ine. This summary includes four studies in adults with rheumatic
0 (2012) 1413– 1424
diseases, and one study in adults with IBD. Unless otherwise spec-ified, all studies assessed serologic titers at baseline and 4 weekslater.
3.4.1. Rheumatic diseasesFour studies were performed in patients with rheumatic dis-
eases; all assessed for response to specific serotypes followingpneumococcal vaccination among patients treated with TNF block-ers, with and without concomitant immunomodulators.
Kaine et al. evaluated pneumococcal vaccine response in 99 RApatients randomized to receive adalimumab and 109 RA patientsrandomized to placebo; concomitant methotrexate was allowed inboth groups. Vaccine response was defined as a greater than two-fold increase in post-vaccination titers from baseline in 3 of the5 tested antigens (9V, 14, 18C, 19F and 23F). No differences werenoted between the two groups for any of the tested antigens, or foroverall vaccine response [9].
Mease et al. assessed pneumococcal vaccine response in 184patients with psoriatic arthritis. After stratifying by MTX use,patients were randomized to receive either etanercept or placebo,and received pneumococcal vaccination 1 month later. Vaccineresponse was assessed for 5 antigens (9V, 14, 18C, 19F, and 23F). Theproportion of responders in the etanercept group (35%–62%) wassimilar to the placebo group (40%–62%). Fewer patients receivingMTX achieved response when compared to patients not receivingMTX [18].
Kapetanovic et al. assessed pneumococcal vaccine response in149 RA patients compared to 47 controls. RA patients were catego-rized by treatment: (1) anti-TNF therapy alone, (2) anti-TNF therapywith MTX and (3) MTX alone. Immune response was assessed for2 antigens (6B and 23F). Post-vaccination titers were significantlylower in group 2 (anti-TNF agent and MTX) and group 3 (MTX) whencompared to group 1 (anti-TNF therapy alone) [16].
Elkayam et al. studied the response to pneumococcal vaccina-tion in 16 RA and AS patients receiving anti-TNF therapy with eitherinfliximab or etanercept, compared to 17 age-matched RA controlsnot on anti-TNF therapy. Methotrexate and steroid use was simi-lar between the two groups. Vaccine response was assessed for 7antigens (14, 23F, 4, 8, 9N, 7F, and 2). GMTs were lower for thoseon anti-TNF therapy for serotypes 23F and 7F [19].
3.4.2. Inflammatory bowel diseaseIn adults with IBD, vaccine response was assessed in 45 patients
and 19 controls. Those with IBD were recruited in 2 groups: GroupA was comprised of those on an anti-TNF agent (infliximab oradalimumab) and an immunomodulator (6-mercaptopurine, aza-thioprine, or methotrexate). Those in Group B had IBD but werenot receiving immunosuppressive therapy. Five serotypes wereassessed (6B, 9V, 14F, 19F and 23F). Post-vaccination GMT werelower for Group A compared to Group B in 4 out of the 5 testedantigens, and lower for Group A compared to healthy controls forall antigens [20].
3.5. Effect of anti-TNF therapy on vaccine response
Table 4 summarizes the 9 studies that investigate the effectof anti-TNF agents on influenza vaccine (4 studies) [11,12,17,22],pneumococcal vaccine (3 studies) [18–20], or both (2 studies)[9,16]. While one study suggested that anti-TNF agents decreasedvaccine response [19], five studies did not demonstrate a lowerresponse among anti-TNF treated patients [9,11,16,18,19,22]. In
the remaining 3 three studies, response rates in anti-TNF-treatedpatients were confounded by concomitant immunomodulators andan effect attributable specifically to the anti-TNF drug could not bedetermined [12,17,20].
N.
Agarw
al et
al. /
Vaccine
30 (2012) 1413– 14241421
Table 4Vaccine responses among patients on anti-TNF therapy.
Study Anti-TNF agentused
Disease(s) Study groups Influenza vaccination pre/post geometric mean titersa
A/H1N1 Response (%) A/H3N2 Response (%) B Response (%)
23F Response (%) Does anti-TNF therapydiminish vaccine response?
Mease et al.[18]
––
––
N/A 5059
N/A 5962
N/A 6262
N/A 3540
N/A 5158
No (pneumococcalvaccination)
Elkayam et al.[19]
––
––
––
––
5.18/14.391.89/5.22
4453
––
––
––
––
2.45/3.511.98/5.80
1353
No. (pneumococcalvaccination) However, forserotype 23F there was astatistically significantdiminished response.
Melmed et al.[20]
0.73/3.963.83/11.351.12/7.04
657695
1.38/3.081.56/6.301.51/5.23
607668
2.16/16.7811.00/43.764.43/64.47
656884
–––
–––
0.86/3.614.73/21.072.77/9.69
457279
1.65/4.993.68/11.141.17/5.52
507679
Unclear (pneumococcalvaccination); however,combinedimmunosuppressivetherapy with an anti-TNFagent and animmunomodulatorresulted in decreasedvaccine response.
1422N
. A
garwal
et al.
/ V
accine 30 (2012) 1413– 1424
Table 4 (Continued )
Study Pneumococcal vaccination pre/post titersb
6B Response (%) 9V Response(%)
14 Response(%)
18C Response(%)
19F Response(%)
23F Response (%) Does anti-TNF therapydiminish vaccine response?
No(influenza/pneumococcalvaccination); however,combinedimmunosuppressivetherapy has dimishedvaccine response(pneumococcalvaccination). In addition,anti-TNF therapy hadlower response rates thanmethrotrexate therapy forinfluenza vaccination.
Gelinck et al.[11]
–––
–––
–––
–––
–––
–––
–––
–––
–––
–––
–––
–––
Yes for 2/3 tested GMTs.(influenza vaccination);although protective levelswere still achieved.
Mamula et al.[12]
––––
––––
––––
––––
––––
––––
––––
––––
––––
––––
––––
––––
Unclear (influenzavaccination); however,combinedimmunosuppression withanti-TNF therapy andimmunomodulatorsresulted in decreasedvaccine response.
Fomin et al.[22]
––
––
––
––
––
––
––
––
––
––
––
––
No (influenza vaccination)
Lu at al. [17] –––
–––
–––
–––
–––
–––
–––
–––
–––
–––
–––
–––
Unclear (influenzavaccination);subgroupanalysis revealed patientson anti-TNF agents haddecreased vaccineresponse against influenzastrain B. However, it isunclear if this effect wasdue to anti-TNF agentsalone or in combinationwith immunomodulators.
a Pre/post vaccination titers given in mcg/ml; vaccine response as defined by authors: either >4-fold increase in pre-vaccination GMT or post-vaccination titer >40.b Pre/post vaccination titers given in mcg/ml; vaccine response as defined by the authors: >2-fold increase in post-vaccination GMT.
ccine 3
4
iWa
ditpviqdacvi
cotdmdmv
aacpatdnw
httvmaa
wtistarmol
ssacsaal
[
[
[
[
N. Agarwal et al. / Va
. Discussion
We aimed to determine the impact of immunosuppressive med-cations on immune responses to vaccines in at-risk populations.
e identified 14 relevant studies assessing responses to influenzand pneumococcal vaccines.
For influenza vaccination, 5 of 11 studies showed significantlyecreased responses for at least one antigen in patients receiving
mmunosuppressive therapy [2–6]. Patients in these studies werereated for a variety of conditions with a variety of immunosup-ressive medications. Four of the studies that reported ‘adequate’accine response showed decreased post-vaccination titers inmmunosuppressed patients compared to controls, raising theuestion as to whether these studies were sufficiently powered toetect differences in response rates for individual antigens. Takens a whole, these studies suggest that immunosuppressive medi-ations may partially dampen the immunologic response influenzaaccination, particularly when multiple immunosuppressive med-cations are used.
Among the 5 studies that assessed response to pneumo-occal vaccination, the majority showed similar proportionsf response among those treated with anti-TNF therapy rela-ive to those that were not. However, several of these studiesemonstrated diminished responses among those treated withethotrexate as monotherapy or in combination with anti-TNF
rugs, suggesting that methotrexate, not the anti-TNF drug,ay be primarily responsible for inhibiting immune response to
accination.We attempted to assess various classes of immune suppressant
gents. We found that in most studies, anti-TNF agents alone did notdversely affect immune response to both influenza and pneumo-occal vaccination. However, the majority of these studies were notowered to investigate the effects of monotherapy with anti-TNFgents relative to the effects of concomitant therapy. We concludehat in most cases, anti-TNF therapy alone is not associated withecreased vaccine response. However, anti-TNF therapy in combi-ation with other immunosuppressive medications is associatedith decreased vaccine responses.
Our systematic review included patients with a variety ofealth conditions, including rheumatologic conditions, solid organransplant recipients, and inflammatory bowel disease. In post-ransplant populations, the majority of studies show diminishedaccine responses across a variety of immunosuppressive regi-ens. Due to heterogeneity of treatment regimens and outcome
ssessments from study- to study, we cannot draw conclusionsbout specific drugs or drug regimens.
Interestingly, some general differences in vaccine responsesere noted between rheumatologic conditions and IBD, which
end to utilize similar therapeutic classes (anti-TNF agents,mmunomodulators). Among those with rheumatic diseases, mosttudies showed a relatively normal vaccine response among thosereated with immunosuppressive therapies. In contrast, studiesmong subjects with IBD generally suggested impaired vaccinationesponses among those treated with immunosuppressive regi-ens. This discrepancy might be explained by the higher doses
f immunosuppressive therapies used in IBD relative to rheumato-ogic diseases [5,7,8].
Our systematic review is limited by the paucity of publishedtudies assessing vaccine response in patients receiving immuno-uppressive medications; we therefore included only studiesssessing vaccine responses to the influenza and pneumococ-al vaccines. Even within these studies, differences in specific
erotypes assessed, and definitions of response, limit generalize-bility. Meta-analysis was not performed due to this heterogeneitycross studies, specifically with regard to post-vaccination sero-ogic assessment.
[
0 (2012) 1413– 1424 1423
In summary, we have demonstrated that immunosuppressivemedications may adversely affect pneumococcal and influenza vac-cine responses in certain populations receiving combinations ofimmunosuppressive therapies. Given the decreased responses tovaccination in these settings, it is reasonable to recommend thatcertain populations should receive influenza and pneumococcalvaccinations prior to the initiation of immunosuppressive therapy.This approach has long been recommended for patients being con-sidered for organ transplantation [24]. Similar recommendationsto vaccinate newly diagnosed patients with immune-mediatedconditions may improve the likelihood of responses to vac-cination, and optimize protection against vaccine-preventableinfections. Finally, a strategy for surveillance of titers over timeamong those vaccinated while immunosuppressed should bedeveloped to ensure adequate protection among these at-riskindividuals.
Acknowledgements
Gil Y. Melmed had full access to all of the data in the studyand takes responsibility for the integrity of the data and theaccuracy of the data analysis. Funding. None. Financial disclosures.GYM: Jannsen (consultant), Abbott Labs (non-CME speaker),Amgen (consultant), Celgene (consultant). Contributors. Study con-cept and design was done by Gil Melmed. Acquisition of data wasperformed by all authors. Analysis and interpretation of data wasdone by Nikhil Agarwal, Robert Frenck, and Gil Melmed. Draft-ing of the manuscript by Nikhil Agarwal and Gil Melmed. Criticalrevision of the manuscript for important intellectual content wasdone by all authors. Statistical analysis by Nikhil Agarwal andGil Melmed.
References
[1] Vonkeman H, ten Napel C, Rasker H, van de Laar M. Disseminated primary vari-cella infection during infliximab treatment. J Rheumatol 2004;31(12):2517–8.
[2] Esteve M, Saro C, Gonzalez-Huix F, Suarez F, Forne M, Viver JM. Chronic hepatitisB reactivation following infliximab therapy in Crohn’s disease patients: needfor primary prophylaxis. Gut 2004;53(9):1363–5.
[3] Baghai M, Osmon DR, Wolk DM, Wold LE, Haidukewych GJ, Matteson EL. Fatalsepsis in a patient with rheumatoid arthritis treated with etanercept. Mayo ClinProc 2001;76(6):653–6.
[4] Vergara M, Brullet E, Campo R, Calvet X, Blanch L. [Fulminant infection causedby varicella herpes zoster in patient with Crohn disease undergoing treatmentwith azathioprine]. Gastroenterol Hepatol 2001;24(1):47.
[5] Recommendations of the Advisory Committee on Immunization Practices(ACIP): use of vaccines and immune globulins for persons with altered immuno-competence. MMWR Recomm Rep 1993;42(RR-4):1–18.
[6] Kroger AT, Atkinson WL, Marcuse EK, Pickering LK. General recommendationson immunization: recommendations of the Advisory Committee on Immuniza-tion Practices (ACIP). MMWR Recomm Rep 2006;55(RR-15):1–48.
[7] Sands BE, Cuffari C, Katz J, Kugathasan S, Onken J, Vitek C, et al. Guidelines forimmunizations in patients with inflammatory bowel disease. Inflamm BowelDis 2004;10(5):677–92.
[8] Higgins JPTG. S. Cochrane Handbook for Systematic Reviews of Interventions4.2.6 [updated September 2006]. Chichester, UK: n Wiley & Sons, Ltd.; 2006.
[9] Kaine JL, Kivitz AJ, Birbara C, Luo AY. Immune responses following adminis-tration of influenza and pneumococcal vaccines to patients with rheumatoidarthritis receiving adalimumab. J Rheumatol 2007;34(2):272–9.
10] Oren S, Mandelboim M, Braun-Moscovici Y, Paran D, Ablin J, Litinsky I, et al.Vaccination against influenza in patients with rheumatoid arthritis: the effectof rituximab on the humoral response. Ann Rheum Dis 2008;67(7):937–41.
11] Gelinck LB, van der Bijl AE, Beyer WE, Visser LG, Huizinga TW, van Hogezand RA,et al. The effect of anti-tumour necrosis factor alpha treatment on the antibodyresponse to influenza vaccination. Ann Rheum Dis 2008;67(5):713–6.
12] Mamula P, Markowitz JE, Piccoli DA, Klimov A, Cohen L, Baldassano RN. Immuneresponse to influenza vaccine in pediatric patients with inflammatory boweldisease. Clin Gastroenterol Hepatol 2007;5(7):851–6.
13] Blumberg EA, Albano C, Pruett T, Isaacs R, Fitzpatrick J, Bergin J, et al. The
immunogenicity of influenza virus vaccine in solid organ transplant recipients.Clin Infect Dis 1996;22(2):295–302.
14] Smith KG, Isbel NM, Catton MG, Leydon JA, Becker GJ, Walker RG. Suppres-sion of the humoral immune response by mycophenolate mofetil. Nephrol DialTransplant 1998;13(1):160–4.
1 ccine 3
[
[
[
[
[
[
[
[
[
424 N. Agarwal et al. / Va
15] Scharpe J, Evenepoel P, Maes B, Bammens B, Claes K, Osterhaus AD, et al.Influenza vaccination is efficacious and safe in renal transplant recipients. AmJ Transplant 2008;8(2):332–7.
16] Kapetanovic MC, Saxne T, Sjoholm A, Truedsson L, Jonsson G, Geborek P. Influ-ence of methotrexate. TNF blockers and prednisolone on antibody responsesto pneumococcal polysaccharide vaccine in patients with rheumatoid arthritisRheumatology (Oxford) 2006;45(1):106–11.
17] Lu Y, Jacobson DL, Ashworth LA, Grand RJ, Meyer AL, McNeal MM, et al. Immuneresponse to influenza vaccine in children with inflammatory bowel disease. AmJ Gastroenterol 2009;104(2):444–53.
18] Mease PJ, Ritchlin CT, Martin RW, Gottlieb AB, Baumgartner SW, Burge DJ, et al.Pneumococcal vaccine response in psoriatic arthritis patients during treatment
with etanercept. J Rheumatol 2004;31(7):1356–61.
19] Elkayam O, Caspi D, Reitblatt T, Charboneau D, Rubins JB. The effect of tumornecrosis factor blockade on the response to pneumococcal vaccination inpatients with rheumatoid arthritis and ankylosing spondylitis. Semin ArthritisRheum 2004;33(4):283–8.
[
0 (2012) 1413– 1424
20] Melmed GY, Agarwal N, Frenck RW, Ippoliti AF, Ibanez P, Papadakis KA, et al.Immunosuppression impairs response to pneumococcal polysaccharide vac-cination in patients with inflammatory bowel disease. Am J Gastroenterol2010;105(1):148–54.
21] Mazzone PJ, Mossad SB, Mawhorter SD, Mehta AC, Schilz RJ, Maurer JR. Thehumoral immune response to influenza vaccination in lung transplant patients.Eur Respir J 2001;18(6):971–6.
22] Fomin I, Caspi D, Levy V, Varsano N, Shalev Y, Paran D, et al. Vaccination againstinfluenza in rheumatoid arthritis: the effect of disease modifying drugs, includ-ing TNF alpha blockers. Ann Rheum Dis 2006;65(2):191–4.
23] Kapetanovic MC, Saxne T, Nilsson JA, Geborek P. Influenza vaccination asmodel for testing immune modulation induced by anti-TNF and methotrexate
therapy in rheumatoid arthritis patients Rheumatology (Oxford) 2007;46(4):608–11.
24] Duchini A, Goss JA, Karpen S, Pockros PJ. Vaccinations for adult solid-organtransplant recipients: current recommendations and protocols. Clin MicrobiolRev 2003;16(3):357–64.
