About
CDC vaccine recommendations are developed using an explicit evidence-based method based on the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.
Evidence Retrieval
- For evidence retrieval, we conducted a systematic review of data on hepatitis A vaccine and persons with HIV (PWHIV), including searches of Medline, EMBASE, CINAHL, Cochrane Library, and ClinicalTrials.gov through January 17, 2019.
- Our search terms were as follows:
(((Hepatitis OR HAV OR hepatovirus) AND vaccin*) OR HepA OR VAQTA OR AVAXIM OR EPAXAL OR HAVPUR OR HAVRIX OR nothav) AND (HIV OR human immunodeficiency)
- We did not restrict articles based on language or country of origin. We excluded articles based on the following criteria:
- Articles focused solely on children or that did not have information on ages of included individuals
- Articles with no data on HAVRIX or VAQTA, which are the two single-antigen hepatitis A vaccines currently licensed in the United States
- Articles that did not provide new data, only included safety data among populations other than our target population of PWHIV, discussed vaccine introduction, made recommendations, or proposed guidelines
- Articles that could not be obtained full-text or in English
- Articles on animals other than humans
- Clinical trials with no results available
- Publication prior to 1996, when hepatitis A vaccine was introduced in the United States
- Articles focused solely on children or that did not have information on ages of included individuals
- We identified 927 unique abstracts; 584 abstracts met one or more of the exclusion criteria (above), leaving 343 articles for full-text review.
- Based on review of the full publications, we eliminated another 319 articles per exclusion criteria. We also excluded 2 studies with populations that were a subset of other included studies.
- We included a total of 22 studies in our GRADE analysis.
GRADE of evidence for hepatitis A vaccination among persons living with HIV: Benefits*
Outcome #1: Hepatitis A infection
| Study | Type | Site | Population N = total |
Age | Intervention | Comparison | CD4+ Count at Vaccination | HIV Viral Load at Vaccination | Immunogenicity* | Main Outcomes #1 |
|---|---|---|---|---|---|---|---|---|---|---|
| Lin, 2018 | Obs | Taiwan | N = 1533 | Median, vaccinated group: 35 | At least 1 dose of HAV vaccine | Median, cells/lL: 550 |
Weeks 28–36: 63.8% (ITT) and 93.7% (PPA) | Vaccine effectiveness: 96.3% | ||
| Cheng, 2017 | Obs | Taiwan | N = 365 | Mean: 30 |
HAVRIX, 2 doses at 0, 6 months; HAVRIX, 3 doses 0, 1, 6 months | Mean: 485 cells/mm3 |
Primary responders: 87.3% (2 dose) 88.9% (3 dose) |
GMCs‡ of anti-HAV immunoglobulin G (IgG): significantly higher for 3-dose versus 2-dose | ||
| Tsachouridou, 2017 | Obs | Greece | N = 1210 | Mean: 34.51 | HAVRIX, 2 doses at 0, 6–12 months; ENGERIX, 3 doses at 0, 4, and 24 weeks; PNEUMOVAX 23 |
Mean: 2.70 log10 | Mean, log10 copies/ml: 4.18 |
80.7% seroconversion within 3 months of HepA series completion | Seroprotection not affected by nadir and current CD4+ cell count and plasma viral load | |
| Jablonowska, 2014 | Obs | Poland | N = 234 | Mean age, vaccinated: 30.7 | HAVRIX, 2 doses, 6 months apart |
Median: 450 cells/mm3 |
79.5%, one month after second dose 75.5%, 5 years after vaccination |
Most HIV-infected adults with high CD4+ counts had a durable response up to 5 years post vaccination | ||
| Kourkounti, 2014 | Obs | Greece | N = 897 | Mean, vaccinated group: 40.2 | HAVRIX or VAQTA, 2 doses, 6-12 months apart | Response rate: 76% | GMT‡: 305 mIU/ml (95% CI 255-361 mIU/ml) | |||
| Jimenez, 2013 | Obs | USA | N = 226 | Mean: 41.8 | At least 1 dose: a) HAVRIX b) TWINRIX (720 EU) |
Median: 410 cells/mm3 |
Median: 1287 copies/mL | 53.5% overall 54% (HAVRIX) 53% (TWINRIX) |
Patients with CD4+ counts >350 cell/mm3 (60%) were more likely to respond than those with CD4+ counts <200 cell/mm3 (35%) (P = 0.0498). Responders were also more likely to be virologically suppressed (48% versus 32%; P = 0.0024). |
|
| Kourkounti, 2013 | Obs | Greece | N = 113 | Median: 40 | HAVRIX or VAQTA, 2 doses, 6–12 months apart | Median, cells/mm3: 570 |
Median, copies/mL: <50 |
After the second dose: 77.0% | GMT‡: Highly active antiretroviral therapy (HAART) patients, 237 mIU/mL [95% CI, 201–321 mIU/mL]; no HAART, 158 mIU/mL [95% CI, 130–221 mIU/mL]), P = 0.068 | |
| Mena, 2013 | Obs | Spain | N = 499 | Median: 36.3 | (a) HAVRIX, 1 dose (b) HAVRIX, 2 doses, 6 months apart (c) TWINRIX (720 EIU), 3 doses at 0,7,14–21 days |
Median, cells/mm3: - 531, standard schedule - 543, rapidly accelerated |
Median, log10 copies/ml: 2.3 |
Overall rate: 73.4%
(a) 60.0% |
Protective antibody response to vaccination was associated with a higher CD4+/CD8 ratio Higher response was associated with reception of 2 doses of standard schedule (in comparison with those receiving only one of those of the same schedule) |
|
| Tseng, 2013 | Obs | Taiwan | N = 582 (365 HIV+) | age range: 18–40 | (a) HAVRIX, 2 doses at 6 months apart (b) HAVRIX, 3 doses at 0, 1, and 6 months |
(c) HAVRIX, 2 doses at 6 months apart (HIV- group) |
Mean, cells/mm3: (a) 538 (b) 452 |
(a) 2.5 log10 copies/mL (b) 3.0 log10 copies/mL |
Week 48 (ITT**): (a) 75.7% for 2-dose HIV+ (b) 77.8% for 3-dose HIV+ (c) 88.5% for 2-dose HIV- |
GMC‡ at week 48 (P <0.01): (a) 2-dose, 1.94 log10 mIU/mL (b) 3-dose, 2.29 log10 mIU/mLProtective antibody response associated with higher CD4+ counts and undetectable plasma HIV RNA load |
| Kourkounti, 2012 | Obs | Greece | N = 351 | Median: 40 (range 34–45) | HAVRIX or VAQTA, 2 doses, 6–12 months apart | Median: 564 cells/mm3 |
60% had <50 copies/mL at or prior to dose 1 HAV | 1 month after the second dose: 74.4% GMTs: 315, 203, 153, and 126 mIU/ml at months 1, 6, 12, and 18 |
A higher response rate and higher GMTs were observed in patients with CD4+ counts ≥500 cells/mm3 (76.6%) than in patients with CD4+ counts 200–499 cells/mm3.- Protective antibody response to vaccination was associated with higher baseline median CD4+ count at vaccination. |
|
| Weinberg, 2012 | Obs | USA | N = 373 | Mean: responders: 41.7 non-responders: 41.6 |
HepA (unspecified 2 dose vaccine 6 months apart or 3 dose vaccine every 2 months) | Mean, cells/μl: responders: 519 non-responders: 450 |
Plasma HIV RNA <400 copies/ml: responders: 46% non-responders: 35% |
52% in HAV-seronaïve | Plasma HIV RNA <400 copies/ml, higher CD4+ cells/μl, and baseline antibody titers <20 mIU/ml (HAV seronaïve) were significantly associated with an antibody response to the vaccine | |
| Crum-Cianflone, 2011 | Obs | USA | N = 130 | Median: 35 | VAQTA or HAVRIX, 2 doses, 6–18 months apart | Controls: HIV-negative, VAQTA, 2 doses | Median: 461 cells/mm3 |
Plasma HIV RNA level, <1000 copies/mL: 49% | 89% overall
78%, CD4+ <350 cells/mm3 |
GMCs‡ among HIV+ adults: 154, 111, and 64 mIU/mL at 1, 3, and 6–10 years
Higher GMCs over time among HIV-infected adults were associated with lower log10 HIV RNA levels (P = 0.04) |
| Armstrong, 2010 | Obs | USA | N = 451 | Mean: 40 |
HepA (standard dose) or HepB (standard dose) or TWINRIX | 64%, CD4+ >400 cells/mm3 36%, CD4+ ≤400 cells/mm3 |
HepA: 60%, overall 62.5%, CD4+ >400 55.56%, CD4+ ≤400 |
Immune development to HepA increased as CD4+ counts increased | ||
| Horster, 2010 | Obs | Germany | N = 131 | Mean: 40 |
HAVRIX, 2 doses at months 1 and 6 or TWINRIX (720 EU), 3 doses at months 1, 3, 6; plus additional vaccines*** | Median: 423.0 CD4+/µl | Median: below limit of detection | 63.6% | Seroconversion was 63.6% among those receiving hepatitis A vaccine | |
| Launay, 2008 | RCT | France | N = 99 | Mean: 38.8 years |
HAVRIX, 2 doses, 24 weeks apart | HAVRIX, 3 doses at weeks 0, 4, 24 | Median, cells/mm3: 355 | Median, copies/mL (IQR): <50 (<50–1300) | Week 28, ITT**: 69.4%, 2-dose group 82.6%, 3-dose group (P = 0.13) |
GMT‡, mIU/mL: 138.2, 2-dose vs. 323.5, 3-dose group at 28 weeks |
| Overton, 2007 | Obs | USA | N = 906 | Mean, vaccinated group: 38.1 | HAVRIX, at least 1 dose | Mean, cells/mm3: 447 | 49.6% overall | Protective antibody response to vaccination with HIV viral RNA load <1000 copies/ml | ||
| Weissman, 2006 | Obs | USA | N = 503 | Mean: responders, 43.5 non-responders, 45.0 |
HAVRIX, 2 doses, 6–12 months apart |
Mean, cells/mm3: overall: 424 responder: 508.6 non-responder: 344.3 |
After the 2nd dose (mean of 187 days post series completion): 48.5% | Protective antibody response to vaccination was associated with higher CD4+ count | ||
| Rimland, 2005 | Obs | USA | N = 659 | Age not published | HAVRIX, 2 doses | After the 2nd dose: 60.7% | Protective antibody response to vaccination was associated with higher CD4+ count, especially if >200 cells/mm3 | |||
| Wallace, 2004 | RCT | USA | N = 180 (90 HIV+) | Mean: 32.6 years |
VAQTA, 2 doses, week 0 and week 24 | Placebo | Mean, cells/mm3: 457.5, VAQTA 493.6, placebo |
Mean, copies/mL: 0.33 x 105, VAQTA 0.16 x 105, placebo |
Week 28: 94% among HIV-infected subjects 87%, CD4+ <300 cells/mm3 100%, CD4+ ≥300 cells/mm3 |
GMT‡, mIU/mL: 517 subjects with CD4+ <300 cells/mm3; 1959 subjects with ≥300 cells/mm3 |
| Kemper, 2003 | RCT | USA | N = 133 | Mean: 38 years |
HAVRIX, 2 doses, 6 months apart | Placebo, 2 doses, 6 months apart | Mean, cells/mm3: 376, vaccine 327, placebo (P, not significant) |
Mean, log10 copies/mL: 3.2, vaccine 3.39, placebo |
Month 9: 68%, CD4+ ≥200 cells/mm3 9%, CD4+ <200 cells/mm3 (P = 0.004) |
Protective antibody response to vaccination was significantly associated with CD4+ cell counts ≥200 cells/mm3 |
| Lederman, 2003 | Obs | USA | N = 643 | Median: 40 | HAVRIX, 2 doses, weeks 16 and 40 + multiple antigens**** | Median, cells/mm3: 226 |
Median copies/mL: ≤500 | 8 weeks after second dose: 46% | 46% of subjects seroconverted after 2 doses of hepatitis A vaccine | |
| Valdez, 2003 | Obs | USA | N = 38 | Median: 38 | HAART and IL-2 vaccinated with: HAVRIX + tetanus toxoid + REMUNE + ENGERIX |
HAART-only vaccinated with: HAVRIX + tetanus toxoid + REMUNE + ENGERIX |
Median, cells/µL: HAART/IL-2: 865 HAART: 445 |
Median, log10 copies/mL (IQR): HAART/IL-2: 1.7 (1.7 - 2.6) HAART: 1.7 (1.7 - 1.7) |
88% of HAART-only recipients 36% of HAART/IL-2 recipients |
Seroconversion was 88% among HAART-only and 36% among HAART/IL-2 groups |
Table 1 Footnotes
‡GMT/ GMC: geometric mean titer/geometric mean concentration
* Seroconversion defined as anti-HAV antibody concentrations:
• ≥10 mIU/ mL: Horster; Wallace
• ≥10 mIU/ mL at 12 (±6) months after second dose: Crum-Cianflone
• ≥20 mIU/mL: Kourkounti, 2012; Kourkounti, 2013; Kourkounti, 2014; Mena; Tsachouriou; Weinberg; Tseng; Launay; Jablonowska
• Primary responders: ≥20 mIU/mL at month 12: Cheng
• ≥33 mIU/mL: Kemper
** ITT: Intention to treat analysis.
*** Additional vaccines administered: trivalent influenza split-vaccine (INFLUSPLIT), pneumococcal vaccine (PNEUMOVAX 23), hepatitis B (ENGERIX; administered at months 1, 3, if HAVRIX given for hepatitis A).
**** Antigens included Candida albicans, mumps skin test, and TT US Pharmacopeia fluid; tetanus toxoid vaccine was also administered unless previously received in past 12 months.
GRADE of evidence for hepatitis A vaccination among persons living with HIV: Harms
| Study | Type | Site | Population N = total |
Age, years | Intervention | Comparison | Main Outcomes #2 |
|---|---|---|---|---|---|---|---|
| Kemper, 2003 | RCT | USA | N = 133 | Mean: 38 | HAVRIX, 2 doses, 6 months apart | Placebo, 2 doses, 6 months apart | Minor injection site soreness: 35% of vaccine doses administered versus 8% of placebo doses (P <0.01).
Reported bacterial, viral, or fungal infections post-vaccination similar for patients receiving vaccine or placebo (24% vs. 26%, respectively; P >0.20). Within 4 days of vaccination, 1 subject (1.6%) in each group experienced severe headache; 1 subject (1.6%) in vaccine group experienced severe fatigue. This difference was non-significant. We consider these to be relatively mild adverse events. The authors concluded that the vaccine was well tolerated in this population. |
| Wallace, 2004 | RCT | USA | N = 180 (90 HIV+) | Mean: 32.6 | VAQTA, 2 doses, week 0 and week 24 | Placebo | Local reaction at injection site in 57% of VAQTA group and 60% of placebo group.
Systemic adverse events (predominantly self-limited headache and fever) were more common among PWHIV who received VAQTA (37%) than among PWHIV who received placebo (23%). Only 3 subjects experienced clinically significant adverse events within 2 weeks after receipt of either vaccine dose. Only 1 of these 3 events (a severe headache) was thought to be vaccine-associated. There were no significant changes in complete blood counts or the results of liver function tests in any group at any point in this study. |
| Tseng, 2013 | Obs | Taiwan | N = 582 (365 HIV+) | range: 18–40 | HAVRIX, 2 doses, 6 months apart | HAVRIX, 3 doses at 0,1 and 6 months | 51.6% of all subjects (HIV+ 51.7% vs HIV- 51.6%, P = 0.98) experienced mild tenderness at local injection site within 24 hours of vaccination. |
| Study | Type | Site | Population N = total |
Age | Intervention | Comparison | Main Outcomes #2 |
|---|---|---|---|---|---|---|---|
| Launay, 2008 | RCT | France | N = 99 | Mean: 38.8 |
HAVRIX, 2 doses, 24 weeks apart | HAVRIX, 3 doses at weeks 0, 4, 24 | There were no serious adverse events associated with the vaccine.
No significant changes in CD4+ T-cell counts or plasma HIV-1 RNA levels during 28-week follow-up. |
| Bodsworth, 1997 | Obs | Australia | N = 180 | Mean: - case: 33.2 - control: 36.6 |
HAVRIX, 2 doses at 1 or 6 months apart | No vaccine for controls | No significant differences (P >0.2) between case and control groups after 1 year for: - AIDS progression, 10.1% versus 10.7% - Death, 7.3% versus 7.6% - Mean CD4+ decline, 125 x106/l versus 123 x106/l No serious adverse events attributable to vaccination. |
| Wallace, 2004 | RCT | USA | N = 180 (90 HIV+) | Mean: 32.6 |
VAQTA, 2 doses, week 0 and week 24 | Placebo | No adverse effect on either HIV viral load or CD4+ cell count found. |
*RCT – randomized control trial
Obs – observational study
GMT – geometric mean titer
GMC – geometric mean concentration
ITT- intention to treat