Association Between 2-Dose vs 3-Dose Hepatitis B Vaccine and Acute Myocardial Infarction

Abstract

Importance: The 2-dose hepatitis B vaccine with a cytosine phosphoguanine adjuvant (HepB-CpG vaccine; Heplisav-B) generated higher seroprotection in prelicensure trials than did a 3-dose hepatitis B vaccine with an aluminum hydroxide adjuvant (HepB-alum vaccine; Engerix-B). However, in 1 trial, a higher number of acute myocardial infarction (MI) events were observed among those who received the HepB-CpG vaccine than among those who received the HepB-alum vaccine, an outcome requiring further study.

Objective: To compare the rate of acute MI between recipients of HepB-CpG vaccine and HepB-alum vaccine.

Design, setting, and participants: This prospective cohort noninferiority study was conducted at Kaiser Permanente Southern California (KPSC), an integrated health care system with 15 medical centers and approximately 4.7 million members. The study included 69 625 adults not undergoing dialysis who received at least 1 dose of a hepatitis B vaccine in either family medicine or internal medicine departments at KPSC from August 7, 2018, to October 31, 2019 (November 30, 2020, final follow-up).

Exposures: Receipt of HepB-CpG vaccine vs HepB-alum vaccine. The first dose during the study period was the index dose.

Main outcomes and measures: Individuals were followed up for 13 months after the index dose for occurrence of type 1 acute MI. Potential events were identified using diagnosis codes and adjudicated by cardiologists. The adjusted hazard ratio (HR) of acute MI was estimated comparing recipients of HepB-CpG vaccine with recipients of HepB-alum vaccine, with inverse probability of treatment weighting (IPTW) to adjust for demographic and clinical characteristics. The upper limit of the 1-sided 97.5% CI was compared with a noninferiority margin of 2.

Results: Of the 31 183 recipients of HepB-CpG vaccine (median age, 49 years; IQR, 38-56 years), 51.2% (n = 15 965) were men, and 52.7% (n = 16 423) were Hispanic. Of the 38 442 recipients of HepB-alum (median age, 49 years; IQR, 39-56 years), 50.8% (19 533) were men, and 47.1% (n = 18 125) were Hispanic. Characteristics were well-balanced between vaccine groups after IPTW. Fifty-two type 1 acute MI events were confirmed among recipients of HepB-CpG vaccine for a rate of 1.67 per 1000-person-years, and 71 type 1 acute MI events were confirmed among recipients of HepB-alum vaccine for a rate of 1.86 per 1000 person-years (absolute rate difference, -0.19 [95% CI, -0.82 to 0.44]; adjusted HR, 0.92 [1-sided 97.5% CI, ∞ to 1.32], which was below the noninferiority margin; P < .001 for noninferiority).

Conclusions and relevance: In this cohort study, receipt of HepB-CpG vaccine compared with HepB-alum vaccine did not meet the statistical criterion for increased risk of acute myocardial infarction.

Figure 1.. Flow of Recipients of Hepatitis B Vaccines

^aMedian number of individuals per medical center, 4920 (IQR, 3574-5952). ^bMedian number of individuals per medical center, 4834 (IQR, 2988-5605). ^cMedian number of individuals per medical center, 5235.5 (IQR, 4070.5-10 317.5). ^dExclusions were not mutually exclusive; recipients could have multiple reasons for exclusion. ^eMedian number of individuals per medical center, 4755 (IQR, 2881-5572). ^fMedian number of individuals per medical center, 4169.5 (IQR, 1856-8125). HepB-alum indicates hepatitis B vaccine with an aluminum hydroxide adjuvant; HepB-CpG, hepatitis B vaccine with a cytosine phosphoguanine adjuvant; KPSC, Kaiser Permanente Southern California.

Figure 2.. Cumulative Incidence of Confirmed Type 1 Acute Myocardial Infarction Among Recipients of Hepatitis B Vaccines^a

^aThe mean (SD) observation time was 365 (86) days for hepatitis B vaccine with a cytosine phosphoguanine adjuvant (HepB-CpG) vaccine and 363 (88) days for hepatitis B vaccine with an aluminum hydroxide adjuvant (HepB-alum) vaccine; the median observation time was 365 days (IQR, 365-365 days). Potential acute myocardial infarction (MI) events identified through diagnosis codes were reviewed by cardiologists who determined occurrence of type 1 acute MI. Details of acute myocardial infarction adjudication results are presented in eTable 3 in Supplement 2.

Figure 3.. Subgroup Analyses of Confirmed Type 1 Acute Myocardial Infarction Comparing Recipients of HepB-CpG Vaccine With Recipients of HepB-Alum Vaccine^a,b

^aSubgroups were prespecified in the study protocol (age group at index dose, index dose as first or subsequent hepatitis B vaccination) or were requested by the study’s independent data monitoring committee (concomitant vaccine recipients, diabetes, or hypertension). ^bPotential acute myocardial infarction (MI) events identified through diagnosis codes were reviewed by cardiologists, who determined occurrence of type 1 acute MI. Details of acute MI adjudication results are presented in eTable 3 in Supplement 2. ^cAdjusted hazard ratios (HRs) were estimated using Cox proportional hazards models with inverse probability of treatment weighting, numbers and details of which are presented in eTables 5 through 11 in Supplement 2. HepB-alum indicates hepatitis B vaccine with an aluminum hydroxide adjuvant; HepB-CpG, hepatitis B vaccine with a cytosine phosphoguanine adjuvant.