GRADE Hepatitis A vaccine for post-exposure prophylaxis in adults >40 years of age
Post-exposure prophylaxis (PEP) with HepA vaccine or immune globulin (IG) effectively prevents hepatitis A infection when administered within 2 weeks of exposure. Previous ACIP recommendations for PEP included HepA vaccine for persons aged 1–40 years and IG for persons outside of this age range; if IG was not available, HepA vaccine could be administered for persons older than 40 years. During November 2016 to February 2018, the ACIP Hepatitis Work Group held conference calls to review and discuss relevant scientific evidence, including the use of HepA vaccine versus IG for PEP. The ACIP Hepatitis Work Group evaluated the quality of evidence related to the benefits and harms of administering a dose of HepA vaccine for PEP for persons aged >40 years using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework.
A randomized, double-blind clinical trial among 1,090 persons aged 2–40 years who were contacts of hepatitis A cases and susceptible to hepatitis A virus infection suggested that performance of HepA vaccine, when administered <14 days after exposure, approaches that of IG in healthy children and adults aged <40 years. Only limited data are available comparing HepA vaccine and IG in healthy adults >40 years of age, showing reduced response to vaccine in older age groups compared to younger adults. The body of evidence assessing a dose of HepA vaccine for PEP to prevent HAV infection in adults aged >40 years was determined to be GRADE evidence type 4 for benefits and 3 for harms.
Policy Question: Should hepatitis A vaccine be recommended instead of immune globulin (IG) as post-exposure prophylaxis for prevention of hepatitis A disease in adults >40 years of age?
| Population | Healthy adults >40 years of age |
|---|---|
| Intervention | Hepatitis A vaccine administered within 14 days of exposure |
| Control | Immune globulin administered within 14 days of exposure |
| Outcomes |
|
Outcome measures included in evidence profile
| Outcomes | Importance |
|---|---|
| Benefit1. Immunogenicity
2. Hepatitis A infection 3. Deaths 4. Hospitalizations |
Critical Critical Critical Critical |
| Harms1. Adverse events |
Critical |
Evidence types
| Initial Evidence Type | Study Design |
|---|---|
| 1 | Randomized controlled trials (RCTs) or overwhelming evidence from observational studies |
| 2 | RCTs with important limitations, or exceptionally strong evidence from observational studies |
| 3 | Observational studies, or RCTs with notable limitations |
| 4 | Clinical experience and observations, observational studies with important limitations, or RCTs with several major limitations |
GRADE of Evidence for Hepatitis A vaccines in persons aged >40 years: Benefits
Outcome #1-4
| Author, year | Study Design | No. of subjects | Population | Outcome* |
|---|---|---|---|---|
| Briem et al., 1994 | 3 | 60 | 40 to 62 years, Iceland | Seroconversion (≥20 mIU/mL anti-HAV) at 15 days: 77%
GMT at 15 days: 262 mIU/mL (range: 65-995 mIU/mL) |
| Reuman et al., 1997a | 3 | 23 | ≥ 40 years, U.S. | Seroconversion (≥10 mIU/mL anti-HAV) at 2 weeks: 31%
GMT at 2 weeks: 6.1 mIU/mL |
| Van Der Meeren et al., 2015 | 3 | 80 | ≥ 40 years, Belgium, Finland, Iceland | Seroconversion (≥20 mIU/mL anti-HAV) at 15 days: 79.7% (68.8-88.2%)
GMT at 15 days: 126.5 mIU/mL (88.6-180.7 mIU/mL) |
| Goilav et al., 1995 | 3 | 306 | Mean age 24 years (range 18-63),UK, Belgium, Germany, France | Seroconversion (≥20 mIU/mL anti-HAV) at 2 weeks: 76.1% (95% CI: 71.2-81.1%)
GMT at 2 weeks: 30.8 mIU/mL (95% CI: 28.1-33.7) |
| Bertino et al., 1998b | 3 | 153 | ≥ 30 years (median age 42 years), U.S. | Seroconversion (≥10 mIU/mL anti-HAV) at 2 weeks: 46%
GMT at 2 weeks: 10.0 mIU/mL |
| Williams et al., 2000c | 1 | 149 | Mean age 42,Alaska Natives | See results for Nelson et al., 2014a |
| Nelson et al., 2014c | 3 | 172 | 40-49 years; 50-59 years; ≥ 60 years, Alaska Natives | Seroconversion (≥20 mIU/mL anti-HAV) at 15 days: 40-49 years: 74% (n=125); 50-59 years: 54% (n=37); ≥ 60 years: 30% (n=10)
GMTs at 15 days: 40-49 years: 26.09 mIU/mL (n=125); 50-59 years: 12.80 mIU/mL (n=37); ≥ 60 years: 1.62 mIU/mL (n=10) |
| Parrón et al., 2017 | 3 | 80 | >40 years, Catalonia | Hepatitis A infection, deaths, hospitalizations |
| Freeman et al. 2014 | 3 | 40(90) | ≥40, Australia | Hepatitis A infection, deaths, hospitalizations |
| Whelan et al., 2013 | 3 | 10 | ≥41 years, Netherlands | Hepatitis A infection, deaths, hospitalizations |
*No immune globulin comparator studies were identified in the systematic review
aVAQTA 25U; bVAQTA 50U results shown; cNelson et al. 2014 is a reanalysis of Williams et al., 2000
Three studies included in the final review had direct comparisons between adults ≤40 years of age and adults >40 years of age: 1) Briem (retrospective cohort, conducted in Iceland), 2) Reuman (retrospective cohort conducted in the US) and 3) Van der Meeren (post-hoc retrospective pooled analysis from 3 randomized clinical trials conducted in Belgium, Finland and Iceland). In the Briem study, seroconversion at 15 days occurred in 77% of persons aged 40-62 years. The Reuman study used half the current recommended adult dose of VAQTA (25 U). Thirty-one percent of persons aged ≥40 years seroconverted at 2 weeks. In the Van Der Meeren study, seroconversion occurred in ~80% of persons aged ≥40 years. Two studies included adults aged >40 years of age in the population, but did not present results differentiated by age: 1) Goilav (multicentre, controlled, randomised, open, comparative study performed in the UK, Belgium, Germany and France). Seroconversion occurred in about 76% of adults aged 18-63 years. 2) Bertino et al, an observational study conducted in the US, 46% of persons aged ≥30 years seroconverted. The only study to stratify by age groups >40 years was Nelson et al., a reanalysis of Williams et al., which included adults with a mean age of 41 years; seroconversion occurred in 74% of adults age 40-49 years; 54% of adults age 50-59 years and 30% of adults aged ≥60 years.
Hepatitis A infection was described in 3 studies: Parron et al (included adults aged >40 years in Catalonia), Freeman et al (included adults aged ≥40 years in Australia) and Whelan et al (included adults aged ≥41 years in the Netherlands). Deaths and hospitalization outcomes were evaluated in these studies.
GRADE of Evidence for Hepatitis A vaccines in persons aged >40 years: Harms
Outcome #1
| Author, year | Study Design | No. of subjects | Population | Outcome |
|---|---|---|---|---|
| Briem et al., 1994 | 3 | 60 | 40 to 62 years, Iceland | Adverse events |
| Reuman et al., 1997 | 3 | 23 | ≥ 40 years, U.S. | Adverse events |
| Van Der Meeren et al., 2015 | 3 | 80 | ≥ 40 years, Belgium, Finland, Iceland | Adverse events |
| Goilav et al., 1995 | 3 | 306 | Mean age 24 years (range 18-63), UK, Belgium, Germany, France | Adverse events |
| Bertino et al., 1998 | 3 | 153 | ≥ 30 years (median age 42 years) | Adverse events |
| Nelson et al., 2014 (Williams et al., 2000) | 3 | 172 | 40-49 years; 50-59 years; ≥ 60 years, Alaska Native | Adverse events |
| Parrón et al., 2017 | 3 | 80 | >40 years, Catalonia | Adverse events |
| Freeman et al. 2014 | 3 | 40 (90) | ≥40, Australia | Adverse events |
| Whelan et al., 2013 | 3 | 10 | ≥41 years, Netherlands | Adverse events |
Incidence of outcomes
| Outcome | No. of subjects (# studies) | Incidence in comparison group | Incidence in vaccinated | Vaccine efficacy | Absolute risk | Number needed to vaccinate |
|---|---|---|---|---|---|---|
| Hepatitis A infection | 130(3 observational) | N/A | 4 cases (3.1%) | N/A | N/A | N/A |
| Deaths | 130(3 observational) | N/A | None reported | N/A | N/A | N/A |
| Hospitalizations | 130(3 observational) | N/A | None reported | N/A | N/A | N/A |
| Serious adverse events | 974(9 observational) | N/A | None reported | N/A | N/A | N/A |
Four cases among adults aged >40 years were identified in the 3 studies with information on hepatitis A infection (one case occurred in a 43 year old man; no other information was provided). Incidence in comparison group could not be assessed because no studies were identified that compared IG and vaccine in adults aged >40 years. Vaccine efficacy, absolute risk and number needed to vaccinate also could not be evaluated.
GRADE SUMMARY
| Outcome | Design (# studies) | Risk of bias | Inconsistency | Indirectness | Imprecision | Evidence type | Overall evidence type |
|---|---|---|---|---|---|---|---|
| BENEFIT | |||||||
| Immunogenicity | 6 observational | No serious | Serious (-1)b | Serious (-1)c | Serious (-1)d | 4 | 4 |
| Hepatitis A incidence | 3 observational | Serious (-1)a | No serious | No serious | No serious | 4 | 4 |
| Deaths | 3 observational | Serious (-1)a | No serious | No serious | No serious | 4 | 4 |
| Hospitalizations | 3 observational | Serious (-1)a | No serious | No serious | No serious | 4 | 4 |
| HARMS | |||||||
| Serious adverse events | 9 observational | No serious | No serious | No serious | No serious | 3 | 3 |
aTwo of three studies were surveillance, with little information provided about the potential for missed follow-up.
bImmunogenicity results in older adults showed high variability by study, with orders of magnitude differences in GMT results.
cFour of six studies did not provide results broken down by age group, included younger individuals, or did not provide details on age ranges.
dStudies included very few adults in the oldest age groups. Only one study provided results for adults >60 years and this included only 10 subjects.
The benefit outcome immunogenicity was downgraded for inconsistency (Immunogenicity results in older adults showed high variability by study, with orders of magnitude differences in GMT results); downgraded for indirectness (Four of six studies did not provide results broken down by age group, included younger individuals, or did not provide details on age ranges) and downgraded for imprecision (Studies included very few adults in the oldest age groups. Only one study provided results for adults >60 years and this included only 10 subjects.) Hepatitis A incidence, death and hospitalizations was downgraded for risk of bias (Two of three studies were surveillance, with little information provided about the potential for missed follow-up). The overall evidence was type 4 for benefits and type 3 for harms.