Keywords
hepatitis A - ulcerative colitis - Crohn's disease - inflammatory bowel disease -
immunization - vaccine
Introduction
The management of inflammatory bowel disease (IBD) has many complex issues which need
to be tackled by the clinicians. IBD not only involves the gastrointestinal system
but may have extraintestinal manifestations. Liver disease in patients with IBD may
be related to the extraintestinal manifestations, adverse effects of various therapies,
or as a consequence of underlying viral infections which may flare due to use of immunosuppression.[1] Around 5% of IBD patients may have underlying hepatobiliary disease while a third
could have derangements in liver function tests.[1]
[2] Therefore, any strategy which is effective in preventing hepatic morbidity in these
patients should be implemented. Vaccination is an effective strategy for prevention
of hepatitis A virus (HAV) and hepatitis B virus (HBV) and has been recommended for
IBD patients by various guideline panels.[3]
[4]
[5]
The need for vaccination with respect to HBV is unequivocal since the disease is associated
with chronic liver disease and may eventually lead to hepatocellular carcinoma. However,
there are several arguments for or against the strategy of routine vaccination for
hepatitis A in patients with IBD. HAV typically causes a self-limiting disease and
the occurrence of adverse outcomes like acute liver failure is relatively rare in
childhood.[6] Further, childhood infection is common in many regions of the world and this provides
an immunity against further infections during adulthood.[7] On the contrary, infection in adults may be associated with more risk of adverse
outcomes and the childhood infection rates may be much lower in certain countries
especially the Western world.[8] Further, patients with IBD may avoid eating out and be more likely to have safer
eating habits leading to lesser likelihood of exposure to HAV infection. Therefore,
the acquisition of infection in adulthood with possible underlying immunosuppression
(disease or therapy related) and the presence of underlying liver disease in a subset
may predispose to more adverse outcomes in patients with IBD.
Therefore, we conducted a systematic review to inform the clinical practices in relation
to HAV vaccination in patients with IBD. We assessed the overall seroprevalence of
HAV in patients with IBD, seroconversion after HAV vaccination in patients with IBD,
and the uptake of HAV vaccination in patients with IBD.
Methods
We followed the guidance provided in the Meta-analysis Of Observational Studies in
Epidemiology statement for the present systematic review.[9]
Database Search
We searched PubMed and Embase from inception till September 11, 2021. The keywords
used were hepatitis A, inflammatory bowel disease or ulcerative colitis (UC) or Crohn's
disease (CD), and vaccination combined with the operator AND. [Supplementary Table S1] (available in the online version) shows the detailed search strategy used for this
systematic review. We searched for additional papers from the references of these
studies. After combining the results, the duplicates were manually removed. The remaining
results underwent initial screening for titles and abstracts to identify papers eligible
for the full-text screening by two reviewers (V.S., P.B.). The differences were resolved
by discussion among the reviewers.
Study Selection
We included studies which provided information regarding baseline seroprevalence rates
of previous HAV infections with IBD, the vaccination rates for HAV in patients with
IBD, and the seroconversion rates after HAV vaccination (single and double dose) in
patients with HAV vaccination. These reports were included irrespective of language,
geographic, or age-related restrictions or the type of publications. For vaccination
rates we included all reports irrespective of the manner in which this was estimated
(records, questionnaires, or Internet-based surveys). However, physician surveys intended
to evaluate the practices regarding vaccination were not included. We excluded studies
where there was suspicion of duplication of data (same center with multiple reports
from overlapping time periods). For any analysis we excluded a study if the total
population reported was less than 10 patients. The studies were excluded if they did
not provide relevant information or if the data was not extractable.
Data Extraction
The data were extracted on a preformat and included the details regarding the study
duration, region, and study population (IBD, CD, and UC; adult or pediatric; any specific
population, e.g., pregnancy or postpartum). The details of study (retrospective, prospective,
trial) were also extracted.
Outcomes
The outcomes of interest relevant to the systematic review for which data were extracted
included:
-
Seroprevalence of HAV infection in IBD population. The data were also extracted for
UC and CD separately to compare the seroprevalence rates in these two groups. The
data from controls/healthy population was also extracted when available.
-
The seroconversion after HAV vaccination after one and two doses of vaccine was also
extracted.
-
The vaccination rates of IBD patients for HAV were also extracted and the method of
estimating vaccination rates was also extracted (survey or records).
Analysis
Apart from the base package of the R statistical software version 4.0.1, meta and
metafor packages were also utilized for the analysis.[10]
[11] We used the random effect model with inverse variance approach to calculate pooled
seroconversion rates after HAV vaccination. The individual seroconversion/seroprevalence/vaccination
rates were logit transformed prior to pooling. The seroprevalence rates between the
IBD population and control population were compared using pooled relative risk obtained
by the Mantel–Haenszel method. I
2 and p-values were used for the assessment of heterogeneity. Subgroup analysis, where feasible,
were performed to address and investigate any significant heterogeneity (I
2 > 50%). Baujat plot was also constructed to identify any specific studies contributing
to the heterogeneity.
Methodological Quality and Risk of Bias Assessment
The publication bias was assessed by the visual assessment of the funnel plot (if
studies > 10). Also, the Egger test was conducted to identify the publication bias.
Two investigators (A.J. and P.B.) separately and independently assessed the methodological
quality and risk of bias for each study using the Joanna Briggs Institute critical
appraisal tool for prevalence studies.[12] Any disagreement was resolved by mutual consensus after discussion with V.S.
Results
Study Selection
We identified 444 records after PubMed and Embase search. After removing duplicates,
349 records were eligible for title and abstract screening. Three hundred and two
titles were excluded for various reasons ([Fig. 1]). Of the 47 titles assessed for full-text screening, 10 were excluded for various
reasons ([Supplementary Table S2], available in the online version),[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22] and eventually 37 papers were included in the systematic review. Fourteen studies
were included in the analysis about baseline seroprevalence of hepatitis A in the
setting of IBD,[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36] 5 studies reported on the seroconversion after complete HAV vaccination,[26]
[37]
[38]
[39]
[40] while 18 studies reported about the HAV vaccination rates in IBD patients[30]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57] ([Tables 1]
[2]
[3]).
Fig. 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart
showing the selection of studies for the systematic review.
Table 1
The studies reporting the seroprevalence of HAV the details regarding patients, vaccinations,
and outcomes
|
Reference
|
Region
|
Study period
|
Study type
|
Patient population
|
Adult/Pediatric
|
Age
|
Male gender
|
Definition of seropositivity
|
1
|
Hong et al 2022
|
South Korea
|
January 2016 and December 2018
|
Cohort
|
263 IBD
|
Adult
|
30 years (IQR, 22 to 46)
|
167 [67.3%]
|
anti-HAV IgG ≥ 20 U/mL
|
2
|
Kalra et al 2021
|
India
|
August 2015 to October 2016
|
Case–Control
|
41 CD
|
Adult
|
33.5 (median)
|
27 [65.8%]
|
IgG anti- HAV antibody
|
3
|
Ruiz-Cuesta et al 2016
|
Spain
|
January and March 2012
|
Cohort
|
153 IBD, 77 UC, 76 CD
|
Adult
|
43.30 ± 14.19
|
69 [45.1%]
|
Prevaccination serological testing
|
4
|
Urganci and Kalyoncu 2013
|
Turkey
|
2000 and 2012
|
Case–Control
|
47 IBD, 25 UC, 14 CD, 8 IBDU
|
Pediatric
|
11.06 ± 3.74
|
25 [53.2%]
|
Prevaccination serological testing
|
5
|
Hafner et al 2008
|
Germany
|
July 1999 and May 2003
|
Case–Control
|
121 IBD, 73 CD, 48 UC
|
Adult
|
36.8 ± 10.5
|
65 [53.7%]
|
HAV titers greater than 10 IU/L
|
6
|
Melmed et al 2006
|
United States
|
2004–2005
|
Cohort
|
16 of 169 IBD patients with serostatus for HAV
|
Adult
|
NA
|
NA
|
Detectable anti-HAV antibody
|
7
|
Feeney et al 2002
|
United Kingdom
|
|
Case–Control
|
137 UC, 139 CD
|
Adult
|
NA
|
NA
|
Hepatitis A IgG status
|
8
|
Moon and Moon 2021
|
United States
|
2009–2010
|
Retrospective analysis of NHANES
|
55 IBD patients and 4,443 non-IBD
|
Adult
|
NA
|
NA
|
Hepatitis A seropositivity
|
9
|
Shao et al 2015
|
Australia
|
March–May 2015
|
Retrospective audit and prospective survey
|
150 IBD patients screened for HAV
|
Adult
|
NA
|
NA
|
NA
|
10
|
Moreira Goncalves et al 2015
|
Portugal
|
NA
|
Prospective study
|
139 patients, 68 UC, 71 CD
|
Adult
|
NA
|
53 [38.1%]
|
?serologic status
|
11
|
Morace et al 2013
|
Italy
|
NA
|
Questionnaire and serology
|
192 IBD
|
Adult
|
51 years
|
94 [48.9%]
|
Serology
|
12
|
Riestra et al 2013
|
Spain
|
September 2011 to October 2012
|
Prospective study
|
787 IBD, 467 CD, 293 UC, 27 unclassified
|
Adult
|
44.3 years
|
357 [45.4%]
|
Serological status
|
13
|
De Bruyn et al 2012
|
Canada
|
September 2011–August 2012
|
Cross-sectional
|
156 IBD, 93 CD, 37 UC, 16 unclassified
|
Pediatric
|
NA
|
NA
|
Positive detection of HAV IgG
|
14
|
Alkhouri et al 2010
|
United States
|
NA
|
Cohort
|
100 IBD, 91 CD, 9 UC
|
Pediatric
|
NA
|
60 [60%]
|
Serology HAV IgG
|
Abbreviations: CD, Crohn's disease; HAV, hepatitis A virus; IBD, inflammatory bowel
disease; IBDU, inflammatory bowel disease unclassified; IgG, immunoglobulin G; IQR,
interquartile range; NA, not available; NHANES, National Health and Nutrition Examination
Survey; UC, ulcerative colitis.
Table 2
Studies reporting seroconversion after HAV vaccination
|
Reference
|
Region
|
Study period
|
Study type
|
Patient population
|
Adult/Pediatric
|
Age
|
Male gender
|
Dosage
|
Response
|
1
|
Urganci and Kalyoncu 2013
|
Turkey
|
2000–2012
|
Prospective
|
47 IBD, 25 UC, 14 CD, 8 indeterminate
23 received vaccine
|
Pediatric
|
11.06 ± 3.74
|
25 [53.2]
|
720 milliELISA units of inactivated hepatitis A virus vaccine (HAV) (0 and 6 months)
|
anti-HAV IgG
|
2
|
Moses et al 2011
|
United States
|
NA
|
Prospective
|
12 IBD patients on infliximab
|
Pediatric
|
18.3 years
|
8 [66.6%]
|
710 milliELISA units at 0 and 6 months or double (if age > 19 years)
|
anti-HAV IgG
|
3
|
Radzikowski et al 2011
|
Poland
|
October 2006 and August 2009
|
Prospective
|
66 adolescents and children with IBD
|
Pediatric
|
13.61 ± 0.4
|
33 [50%]
|
Two doses of HAV in a 6- to 12-month interval
|
Seroconversion at 12 weeks after 2nd dose (> 20 IU/mL)
|
4
|
Dimas et al 2019
|
Greece
|
NA
|
Prospective
|
356 IBD patients
90 got two doses
21 got single dose
|
Adult
|
50.0 [33.5–63.1]
|
NA
|
1 mL, two doses, one at baseline and the second 6–12 months after the first dose
|
Positive anti-HAV IgG > 3 months after 2nd dose
|
5
|
Karmiris et al 2014
|
Greece
|
NA
|
Prospective
|
167 IBD patients, only 43 were eligible for vaccination
|
Adult
|
45.4 [30.3–60.3]
|
103 [61.7%]
|
Havrix, 1 mL, two doses, 0 and 6–12 months
|
anti-HAV IgG > 20 mIU/mL
|
Abbreviations: CD, Crohn's disease; HAV, hepatitis A virus; IBD, inflammatory bowel
disease; IgG, immunoglobulin G; NA, not available; UC, ulcerative colitis.
Table 3
Studies reporting the vaccination rates for HAV in patients with inflammatory bowel
disease
1
|
Pittet et al 2021
|
Switzerland
|
|
Cross-sectional study
|
306 IBD
|
Adults
|
Median 42.7
|
|
Using questionnaire, immunization records review
|
2
|
Chiarella-Redfern et al 2022
|
Canada
|
September 2012–December 2018
|
Retrospective cohort study
|
303 pregnant IBD
|
Pregnancy
|
31.24
|
0
|
Vaccine records
|
3
|
García-Serrano et al 2020
|
Spain
|
Records until December 2016
|
Cross-sectional, retrospective study
|
1,722 IBD
|
Adults
|
NA
|
866 [50.3%]
|
Online records
|
4
|
Mao et al 2019
|
USA
|
2007 onwards
|
Prospective national registry
|
628 IBD pregnant
|
Adults
|
|
|
Questionnaire
|
5
|
Pham et al 2018
|
USA
|
2010 and 2013
|
Retrospective cohort study
|
1,401 IBD on anti-TNF therapy
|
Adults
|
40.6 ± 16.8
|
668 [48.4%]
|
Electronic database
|
6
|
Malhi et al 2015
|
Canada
|
September 1, 2013 and January 31, 2014
|
Patient survey
|
300 IBD
|
Adults
|
35.4 ± 13.2
|
146 [48.7%]
|
Patient Survey
|
7
|
Yun et al 2013
|
South Korea
|
November 2011 to February 2012
|
Questionnaire survey
|
192 IBD
|
Adults
|
39.7
|
121 [63.0%]
|
Patient survey
|
8
|
Moon and Moon 2021
|
USA
|
2009–2010
|
Retrospective analysis of NHANES
|
48 IBD
|
Adults
|
NA
|
NA
|
NHANES data
|
9
|
Hussain et al 2020
|
USA
|
March 1, 2019 to December 31, 2019
|
Retrospective study
|
356 IBD
|
Adults
|
NA
|
172 [48.3]
|
Two IBD clinics electronic checklist
|
10
|
Strasse et al 2019
|
Brasil
|
July 2015 to June 2016
|
Questionnaire survey
|
239 IBD
|
Adults
|
46.4 ± 12.7
|
87 [36.4]
|
Patient survey
|
11
|
Kowalska-Duplaga et al 2019
|
Poland
|
September 2015 and September 2016
|
Prospective study
|
214 children IBD
|
Children
|
14.5 (11.5–16.0)
|
117 [54.7%]
|
Review of immunization cards
|
12
|
Kakati et al 2018
|
USA
|
2016
|
Retrospective charts
|
210 IBD
|
Adults
|
NA
|
NA
|
Review of charts
|
13
|
Waszczuk et al 2018
|
Poland
|
NA
|
Self-completed patient survey
|
195 IBD
|
Adults
|
37 ± 15
|
98 [50.2%]
|
Patient survey
|
14
|
Vellanki et al 2018
|
USA
|
January 1, 2007 to December 31, 2016
|
Retrospective data review
|
1396 IBD
|
Adults
|
NA
|
NA
|
Database review
|
15
|
Ryu et al 2010
|
South Korea
|
November 2013 and February 2015
|
Questionnaire survey
|
287 IBD
|
Adults
|
29 years (range, 16–69)
|
188 [67.6%]
|
Questionnaire
|
16
|
Love et al 2012
|
USA
|
44-month period
|
Retrospective
|
24 IBD on anti-TNF
|
Adults
|
NA
|
NA
|
Electronic records
|
17
|
Xu et al 2019
|
USA
|
2015–2016
|
National Health Interview Survey
|
951 IBD
|
Adults
|
NA
|
360 [37.9]
|
Patient survey
|
18
|
Wasan et al 2012
|
USA
|
NA
|
Prospective Internet patient survey
|
958 IBD
|
Adults
|
45 (IQR 31–57)
|
260 [27.2%]
|
Internet-based survey
|
Abbreviations: HAV, hepatitis A virus; IBD, inflammatory bowel disease; IQR, interquartile
range; NA, not available; NHANES, National Health and Nutrition Examination Survey,
TNF, tumor necrosis factor.
Seroprevalence of Hepatitis A in Inflammatory Bowel Disease
The pooled seroprevalence of HAV antibodies in patients with IBD on the basis of 14
included studies (2,370 patients) was 0.36 (95% confidence interval [CI]: 0.22–0.53,
I
2 = 95%) ([Fig. 2]). To explain the high degree of heterogeneity we performed a subgroup analysis on
the basis of factors which are likely to impact the underlying seroprevalence, that
is, age of included population and the region of the studies. As expected, the seroprevalence
was higher in adults (0.42, 95% CI: 0.27–0.59, I
2 = 95%) as compared with the pediatric studies (0.18, 95% CI: 0.04–0.55, I
2 = 94%), however, the heterogeneity was high ([Supplementary Fig. S1], available in the online version). Regional differences were apparent with studies
from Asia reporting highest seroprevalence rates and those from North America reporting
the lowest seroprevalence rates but the heterogeneity even among various region-based
groups was high ([Supplementary Fig. S2], available in the online version). Baujat plot could identify two studies which
contributed the maximum to the heterogeneity ([Supplementary Fig. S3], available in the online version). The sensitivity analysis by excluding the two
studies which did not report the method to determine seroprevalence did not change
the pooled seroprevalence rates (0.32, 0.19–0.49, I
2 = 95%) ([Supplementary Fig. S4], available in the online version). On comparing the seroprevalence of HAV in IBD
as compared with controls, the pooled relative risk (4 studies) was not different
between the two groups (0.94, 95% CI: 0.66–1.34, I
2 = 76%) ([Fig. 3]).
Fig. 2 Forest plot depicting the pooled seroprevalence rates for hepatitis A virus (HAV)
in patients with inflammatory bowel disease.
Fig. 3 Pooled relative risk of seropositivity of hepatitis A virus (HAV) in patients with
inflammatory bowel disease (IBD) as compared top controls.
Seroconversion after HAV Vaccination
There were five studies (221 patients, two in the pediatric age group) reporting response
to two doses of hepatitis A vaccination in patients with IBD. The pooled seroconversion
rate to two doses of HAV vaccination was 0.93 (95% CI: 0.88–0.96, I
2 = 0%) ([Fig. 4]). Three studies (104 patients) reported on the seroconversion after a single dose
of HAV vaccination and the pooled seroconversion rate was 0.47 (95% CI: 0.35–0.59,
I
2 = 20%) ([Supplementary Fig. S5], available in the online version).
Fig. 4 Pooled seroconversion rates after two doses of hepatitis A virus (HAV) vaccination
in patients with inflammatory bowel disease (IBD).
Vaccination Rates for Hepatitis A in Patients with IBD
The pooled vaccination rate for hepatitis A among patients with IBD (18 studies, 9,521
patients) was 0.21 (95% CI: 0.14–0.30, I
2 = 99%) ([Fig. 5]). To explain the heterogeneity, we performed subgroup analysis based on the region
(continents) and the population (adults, pediatric, and pregnant). Interestingly,
the HAV vaccination rates were highest among the pregnant population (0.56, 0.45–0.66,
I
2 = 90%) as compared with adults (0.18, 012–0.27, I
2 = 98%) ([Supplementary Fig. S6], available in the online version). The vaccination rates were higher in the North
American population (0.29, 0.19–0.41, I
2 = 99%) as compared with the other three continents ([Supplementary Fig. S7], available in the online version). A subgroup analysis on the basis of study type
indicated that the vaccination rates in prospective studies was 0.37 (0.15–0.66, I
2 = 98%), in retrospective studies was 0.18 (0.09–0.33, I
2 = 99%), and in surveys was 0.19 (0.11–0.32, I
2 = 97%) ([Supplementary Fig. S8], available in the online version). Baujat plot identified three studies which had
maximum contribution to heterogeneity ([Supplementary Fig. S9], available in the online version).
Fig. 5 Pooled rates of vaccination for hepatitis A virus (HAV) in patients with inflammatory
bowel disease (IBD).
Risk of Bias
For seroprevalence of HAV, both the funnel plot and the Egger test suggest absence
of any publication bias ([Supplementary Fig. S10], available in the online version). The risk of bias analysis of studies included
in the analysis of seroprevalence, seroconversion rates, and vaccination rates are
shown in [Supplementary Tables S3]–[S5] (available in the online version), respectively. For pooled HAV vaccination rates
in IBD of HAV, the visual analysis of the funnel plot and the Egger test suggest that
publication bias is unlikely ([Supplementary Fig. S11], available in the online version). However, majority of studies were of low risk.
Some of the studies available as abstracts could not be completely assessed for all
domains.
Discussion
Hepatitis A vaccination is recommended for patients with IBD. However, it is unclear
if the vaccination should be done routinely or after checking the presence of previous
exposure. Also, in regions with high baseline seroprevalence, the utility of vaccination
at the time of diagnosis is uncertain as many individuals may have been exposed to
HAV in the past. The present systematic review provides evidence that the seroprevalence
of hepatitis A infection is variable across the globe and the highest rates are found
in Asia while North America may have the lowest rates. Importantly, the seroprevalence
rates in the IBD population are similar to the controls, suggesting that the population
level data can be used as a proxy indicator of seroprevalence in IBD patients and
guide the testing/vaccination strategy. Further, the results suggest that the vaccination
rates against hepatitis A are low across the globe. This is even true in North America
where the baseline seroprevalence is low and the vaccination rates are higher as compared
with other regions of the globe.
Usually, there is a concern about response to vaccination (like HBV and coronavirus
disease 2019) in patients who are on certain drug therapies like immunomodulators
and biologicals.[58]
[59] Our analysis demonstrates that this may not be a significant concern with complete
HAV vaccination although the data are limited. Four of the five included studies provided
information about the underlying drugs used. While seroconversion was good with most
drugs, one report suggested that seroresponse may be attenuated in those on anti-tumor
necrosis factor.[39] Another study, however, suggested good response in the patients who were on infliximab.[37] This suggests that post-vaccination testing of seroconversion may not be required
even in patients with underlying exposure to immunomodulators. The response to a single
dose of vaccination is also fair and therefore, vaccination can be administered at
the time of starting any immunosuppressive therapy.
Hepatitis A is an important cause of acute viral hepatitis in the world. It is a positive
single-stranded ribonucleic acid virus belonging to the Picornaviridae family and
primarily undergoes replication in hepatocytes. The risk factors of HAV infection
include intravenous drug abuse, chronic liver disease, travel to endemic areas, occupational
exposure, prisoners, older persons, and immunocompromised. The transmission is usually
by feco-oral route while the virus is excreted in human bile and stool. With prolonged
immunosuppression in patients of IBD, hepatitis A infection is a concern. Infection
with hepatitis A can lead to acute viral hepatitis and acute liver failure.[6] Rarely hepatitis A infection can cause relapsing hepatitis and prolonged cholestasis.
With vaccination available, hepatitis A-related liver disease is preventable in patients
with IBD. Although this is the first systematic review on hepatitis A vaccination
in IBD, there are certain limitations. We could not separately assess the effect of
underlying disease (UC or CD), activity, and drug classes because of the limited number
of studies. Further, the studies for a particular region regarding seroprevalence
were limited and may not be directly applicable to the entire region.
In regions with low seroprevalence, a strategy to vaccinate without testing of prior
exposure may be appropriate. However, in regions with high seroprevalence (100% in
one study from India), the need for routine vaccination of all IBD patients is questionable.[24] However, with improvements in sanitation there may be decline in childhood exposure
to HAV and changes in such policies may be needed with changing seroprevalence data.
In this regard, the seroprevalence rates in the general population can guide the strategy
because these are proxy indicators of seroprevalence of HAV in the IBD population.
Further, increasing immigration rates bring their own challenges in making blanket
policies for a particular region. The systematic review suggests that a single global
policy for HAV vaccination may not be appropriate and the policy should be based on
regional seroprevalence rates, cost of testing, and the cost of vaccination.