Hepatitis B virus vaccination in HIV-infected people: A review

ABSTRACT

HBV immunization is highly recommended in people infected with HIV. However, the classical schedule used in the general population has been shown to be insufficient in people living with HIV. This review summarizes the main studies dealing with HBV vaccination in people living with HIV, depending on their baseline status (in particular, never vaccinated, already vaccinated, or with an isolated anti-HBc serological profile). It shows that reinforced 40µg intramuscular HBV vaccination schedules are now frequently recommended, either initially in people never vaccinated, or in the lack of an anamnestic response in other situations. Adjuvants cannot be currently recommended. Anti-HBs titers have to be checked 1 to 2 months following the last vaccine dose, and annually thereafter a booster is necessary if antiHBs titers decrease below 10 mIU/mL. In patients with a CD4 cell count <200/µL, guidelines recommend starting the vaccination regimen as soon as possible after HAART has been started.

Introduction

Co-infection with HIV and HBV is common since both viruses share the same routes of transmission.^1-4 HBV infection in people living with HIV is less likely to spontaneously cure than is the case in HIV-negative persons,² and HBV infection is associated with increased rates of cirrhosis (10–20%), a higher risk of hepatocarcinoma,^2,3,5-7 and last a higher risk of liver-related death.⁷ In addition, the risk of hepatotoxic side effects of Highly Active Anti-Retroviral Therapy (HAART) is increased in patients with HBV coinfection.^15-18

Vaccination against HBV is thus highly recommended in the HIV-infected population.^8,9 Despite these recommendations, in 2015, only 2 third of persons living with HIV receive at least one dose of HBV vaccine 20 y after the beginning of the HAART era.^101,102 Response rate is usually defined by a seroconversion with anti-HBs antibodies > 10UI/mL. In addition, response rates to the “classic” schedule (20 µg of HBs antigen at months 0–1–6) vaccination against HBV are lower in people living with HIV, since only 20 to 70% of HIV-infected adults were seroprotected vs 90 to 95% in non HIV infected people ^11-14 The aim of this review is to summarize the currently available data regarding HBV vaccination in people living with HIV, according to their main characteristics and their vaccine and therapeutic background.

Reinforced vaccination strategies in people living with HIV and without anti-HBs antibodies

Why use reinforced strategies and what are the main predictive factors of response to HBV?

The response rates observed with classic HBV vaccination schedules in people living with HIV are lower than in the general population, and could be as low as 17.5%.²⁰ HIV RNA and CD4 cell counts are the main predictive factors of the response to HBV vaccination in people living with HIV. Numerous studies, from retrospective studies to randomized control trials (RCT), have found a correlation between undetectable HIV RNA, a high CD4 cell count and higher vaccination success.^{19,20,23,27,30,60,63,75,93-97,100} The HIV viral load even seems to be more important than the CD4 count [19, 21, 23,], even though a positive correlation between the CD4 cell count at the time of (re)vaccination and the response to vaccination was observed in many studies.^{10,11,20,22,23,95}

Other predictors of a poor response include HCV co-infection, poor general health status and occult hepatitis B ⁷ often observed in people living with HIV. In addition, while female sex ^20,28,62,65 and younger age ^24,60,95,99 are positive predictors of the response to vaccination, Afro-American ethnicity,⁹⁹ alcohol consumption ⁹⁹ and tobacco smoking ^24,99 are known to negatively influence the response to HBV vaccination.

Strategies in patients who have never been vaccinated against HBV (Table 1)

Table 1.

Vaccination against HBV in patients who had never been vaccinated.

Study (year)	Design	Schedule	% HAART Mean CD4 cell count (/µL)	Age (years) / %male	long-term response	Response rate	Predictors
Rey et al, 2000	Prospective	20µG M0–1–2	85%	30.5	Not assessed	55%	CD4 > 500
	N = 20	Boosters M3–4–5	470	70%
Fonseca et al, 2005	RCT	20µ vs 40µ	86%	37	NA	41% overall	CD4 > 350
		M 0–1–6	50% > 350 CD4	45 vs 50%		34% 20µ	HIV RNA < 10000 cp/mL
	N = 192					47% 40µ p = 0.07
Cornejo-Suarez et al, 2006	RCT	10 vs 40µg	65%	35.6 vs 34.1	NA	60.7% overall	CD4 > 200
	N = 79	M 0–1–6	245 vs 225	69.2 vs 72.5%		60% 10µg
						61.5% 40µ NS
Pasrischa et al, 2006	Prospective	20 vs 40µg	0	> 200: 30 YO, 63%	NA	> 200: 100%	Baseline CD4 > 200
	N = 40	M 0–1–6	< 200 / > 200	< 200: 32 YO, 46%		< 200: 47%
Veiga et al, 2006	Prospective	40µg M 0–1–6	100%	Responders:32 YO and 46%	NA	64%	Baseline CD4 and HIV VL
	N = 47		<200 or > 200	NR: 37 YO and 70%			CD4 memory T-lymphocytes
Cruciani et al, 2009	Prospective	40µ M0–1–2	80%	41	32.7% at 2 years	Overall = 89%	High CD4 cell count
		+/− 1–3 more	533	66%		M3 = 60%	Male, VL > 1000 cp/mL
	N = 92
Potsch et al, 2009	Prospective	40µg M0–1–2–6	79%	36	NA	89%	None
	N = 47		402	36%		79% > 100 UI/mL
Psevdos et al 2010	RCT	40 vs 20 µg after classic schedule	95 vs 86.9%	43 vs 45	NA	85 vs 59%	CD4 > 200
	N = 101		421 vs 413.	75 vs 63.9%			HAART
Launay et al, 2011	RCT	20µg M0–1–6	100%	43 / 72%	NA	65%	Female, Young, No smoking, high CD4,
	N = 437	40µg M0–1–2–6		42 / 66%		82% p<0.01
		4µg ID M0–1–2–6		43 / 64%		77% p = 0.02	HIV VL undetectable
De Vries et al, 2011	Non-inferiority RCT	10µg W0–1–3 vs	72.2 vs 70%	40 / 64.9%	NA	38.70%	NA
	N = 761	20µg M0–1–6	430 vs 440	40 / 69.7%		50%
						Non-inferiority for CD4 > 500
Postch et al, 2012	Prospective	40µg M0–1–2–6	80%	37	NA	91%	Undetectable HIV-1 VL
	N = 163		385	40%		80% > 100UI/mL
Mena et al, 2012	Retrospective	First schedule: 40µg	62.50%	36.9	NA	Global: 71%	HIV VL
	N = 474	W 0–4–24/48	488	71.30%		1st schedule: 60.3%	CD4 count > 350
		W 0–4–8				58.10%	Rapid schedule or < 3doses
		D 0–7–21				69.20%
		+/− M 0–1–6				50.00%
						2nd schedule: 58.6%
Ni et al, 2013	Meta-analysis of 5 RCT					Pooled OR for increased dose = 1.96 (95%CI 1.47–2.61)
	N = 833
Chaiklang et al, 2013	RCT	20µg M0–1–6	100%	41 / 18.9%	NA	88.60%	Delay diagnosis – treatment < 3M
	N = 132	20µg M0–1–2–6		42.2 / 20.4%		93.2% NS
		40µg M0–1–2–6		41 / 43.2%		95.4% NS
Fuster et al, 2016	Prospective	20µg M0–1–6	95%	42 / 68%	NA	62%	CD4 > 200
	N = 245						CD4 / CD8 > 0.4

HAART: highly active antiretroviral therapy

RCT: Randomized Controlled Trial

M: Month

W: week

D: Day

YO: Years Old

NR: Non Responders

VL: Virus Load

OR: Odd Ratio

NA: Not Assessed

NS: Not Significative

ID: IntraDermal

RCT: Randomized Controlled Trial

NR: Non Responders

VL: Viral Load]

M: Month

The first strategy assessed was to use higher doses of HBV vaccine. In a randomized control trial (RCT), Cornejo-Suarez et al. ¹⁹ found no difference in seroconversion rates between intramuscular injections of 10 and 40µg, but with a surprising response rate in the 10µg group (63.5%) versus 60% in the 40µg group. Another RCT compared a triple double-dose regimen (40µg) with the classic regimen (20µg) in 210 HIV infected patients, stratified on their CD4 cell count.²¹ Even though the authors found a trend toward a better response in the 40µg group (47% vs 34%, p = 0.07), the difference was only significant in the subgroup of patients with a CD4 cell count > 350/µL (64.3 vs 39.3%, p = 0.008). The double dose also improved HBV immunization compared with the standard dose in patients with an HIV viral load < 10000cp/mL (58.3% vs 37.3%, p = 0.01). Two prospective studies assessed the efficacy of a triple double-dose HBV vaccination schedule. In one small study of 40 HIV-infected patients, Pasricha et al. ¹⁰ found a 100% response rate in patients with a CD4 count > 200 cells/µL against 47% in those below this threshold. In the other prospective study, Veiga et al.²² observed a 64% response rate with Euvax B 40 µg delivered at months 0–1–6.

Another way to increase response rates to hepatitis B vaccination is to increase both the dose and the number of injections. In the largest RCT published to date, Launay et al. ²³ compared 3 different regimens in a population with a CD4 count > 200 µL. One arm had 4 intramuscular injections of 40 µg at months 0–1–2–6, one had 4 intradermal injections of 4µg at the same time, and the control arm had the classic schedule. Compared with the 65% seroconversion rate obtained with the classic schedule (threshold 10 mIU/mL), the 4 intramuscular injection with 40 µg of HBV showed an increased proportion of seroconvertors (82%, p<0.01). They also found more high-responders (i.e. with anti-HBs titers above 100 mIU/mL) in the 4 injections of 40 µg group (74% vs 41% with the classic schedule, p<0.01). The response rate in the 4 injection of 4µg intradermal arm was 77%. A more recent RCT assessing 40 µg injections at 0–1–2 and 6 months found no significant difference between the groups in terms of response rates (95.4 vs 88.6% in the “classical” arm).²⁴ Despite a higher response rate with the classic schedule than usually reported in the literature (probably linked to a study population with favorable prognostic factors to vaccination, such as female gender, no co-infection with HCV, undetectable HIV viral load, and/or CD4 count above 200/mm³), the high-responder rate at month 12 was nevertheless significantly higher in the 4 injections of 40 µg group (72.7% vs 43.2%, p = 0.04). In the same way, in 2 prospective studies, Potsch et al. observed ^25,26 high seroconversion rates following a 40 µg schedule at months 0–1–2 and 6. The first found a response rate of 91% and 80% of high responders one month after the last injection, and the second reported 89% and 79% of responders and high responders, respectively.

On the other hand, schedules with a shorter duration were also assessed in people living with HIV. Vaccination with 40 µg at months 0–1–2 ²⁷ was assessed in one cohort. However, after receiving this 3-dose schedule, patients could have up to 3 supplementary injections if seroconversion was not reached after the first schedule, leading to a global response rate of 89%. Rapidly-accelerated schedules with HBV vaccinations at 0, 1 and 2–3 weeks were evaluated in one non-inferiority RCT testing a 10µg dose against classic schedule, and found 38.7% of seroconversion (vs 50%).²⁸ A retrospective study with 20µg of HBs antigen found a seroprotection rate of 50%.²⁹

Finally, a meta-analysis by Ni et al,³⁰ which included 4 RCTs and one prospective study (883 patients), confirmed the benefit of increased doses (pooled OR 1.82, 95% CI 1.35–2.47) to improve response rates to HBV vaccination. As a result, this reinforced schedule is recommended in most international guidelines for the vaccination of people living with HIV.

Strategies in patients who were vaccinated but without seroprotective antiHBs titer

Revaccination of HIV-infected people who did not respond to a first schedule remains challenging. Considering the low immunogenicity of hepatitis B vaccine in HIV-infected populations (24–58%),^1,21,48-50 there are no global recommendations about which schedule to use. US and French guidelines recommend a second series of vaccinations^51,52 while British recommendations prefer 3 double-doses of vaccine.⁵³

Several non-randomized studies assessed double-dose vaccination.^27,58-61 Chatkittikunwong et al. found higher response rates in their double-dose rescue vaccination group (M0–1–6) than in their standard-dose group (97 vs 70%, p = 0.001). These results were confirmed in sub-group analyses (< 350 and 350–500 CD4 cells/mm3).⁵⁷ Two retrospective studies found similar response rates of 67% ⁵⁹ and 64%²⁸ to double-dose rescue vaccinations. Another retrospective study ⁶⁰ found a response rate of 81.5% in subjects receiving an accelerated rescue schedule (40 µg at M0–1–2) after failure of a classic chedule. De Vries-Suij et al.⁶⁰ found a lower response rate (50%) in their prospective study while Psevdos et al ⁶¹ reported a significant difference between their double-dose and standard-dose rescue vaccination schedule (85 vs 59%, p = 0.006). A prospective cohort study¹¹ found a global response rate of 90% with a classic schedule rescued with 3 supplementary 20µg vaccinations. After the first regimen, the response rate was only 55%. Predictive factors identified were the CD4 cell count and HIV RNA level.

Only one RCT assessed the efficacy of a double-dose ⁵⁴ vs. a standard-dose schedule. It included 178 HIV infected persons randomized as HAART patients with CD4 counts > 200 cells/mm3 and non-responding to a first 20µg booster injection. Even though they found no significant difference for the proportion of responders 4 weeks after the last vaccination, with 74% vs. 67% (p = 0.334), the difference was significant at week 72 (54 vs 31%, p = 0.01). Moreover, there were more high responders in the double-dose group at week 28 (51 vs 26%, p = 0.002) and 72 (22 vs 7%, p = 0.0198). The geometrical mean titers 4 weeks after the last dose were also greater (2.5-fold) in the high-dose group from week 4 to week 72.

Considering these results and international recommendations, double-dose rescue vaccination with at least 3 injections seems to procure better immunization against hepatitis B virus in HIV-infected populations.^51-53,56 Annual supervision of antiHBs titers is recommended to determine when new boosters are needed.

Strategies in patients with an isolated anti-HBc serological profile (Table 2)

Table 2.

Vaccination in people living with HIV with isolated anti-HBc serological profile.

Study (year)	Design	Schedule	% HAART Median CD4 cell count (/µL)	HCV Co-infection	Age / %male	Anamnestic response	Response rate	Predictors
Piroth et al, 2016	Prospective	20µg W0 then 40µg W5–9–24	100%	30%	46 / 59%	46%	89%	NA
	N = 54		538
Kaech et al, 2012	Prospective	20µg M0 +/−	89%	51%	42 / 49%	22%	60%	Injected drug use
	N = 37	20µg M0–1–6	443
Chakvetadze et al, 2010	Prospective	20µg +/− 2–5 20µg W4–8–12–16–20	87.5%	12.5%	40 / 55%	32.5%	40.7% W10	None
	N = 40		463				64% W21
Gandhi et al, 2005	Prospective	20µg W 0–4–24	72%	33%	43 / 62%	16%	62%	Male sex HCV co-infection
	N = 69		518
Jongjirawisan et al, 2006	Prospective	20µg once	80%	12%	38.1 / 59.3%	7%	NA	Intravenous drug use and HCV co-infection
	N = 140		300

HCV: Hepatitis C Virus

M: Month

W: Week

NA: Not Assessed

An isolated anti-HBc (IAHBc) profile, i.e., positive anti-HBc antibodies with negative HBs antigen and negative anti-HBs antibodies, is commonly observed in the HIV-1 population: from 7 to 20%.^31-34 This rate rises to 45% in hepatitis C co-infected patients.^35,36 The meaning of this profile remains unclear. It can represent a “window” phase following acute hepatitis B virus infection (when HBs antigen has disappeared but anti-HBs are not yet produced), a resolved hepatitis B infection with loss of anti-HBs antibodies, an occult chronic HBV infection with detectable HBV DNA³⁷ or a false-positive test.^38,39 The presence of antibodies to hepatitis Be antigen (anti-HBe) suggests previous infection.⁴⁰ Since cases of reactivation of HBV infection have been reported,^31,35,41-43 it is legitimate to ask whether and how such patients should be vaccinated. An anamnestic response is defined by the production of antiHBs > 10 mIU/mL after a single injection of hepatitis B vaccine, and can rise from 7 to 32.5%.^36,44-46

No RCTs have been conducted in this population to evaluate the efficacy and safety of a double-dose hepatitis B vaccination schedule. Four prospective cohorts worked on a standard vaccination schedule in IAHBc.^44-47 The most recent one ⁴⁷ found 46% of anamnestic responders, among whom 50% still had antibodies at month 18. The study included 54 well-controlled HIV-infected patients receiving 3 doses of 40 µg. More importantly, they found a higher persistence of seroprotection at month 18 in the rescue group (81 vs. 50% in initial responders). Other prospective studies reported response rates ranging from 60 to 72% with 3^44,46 to 6⁴⁵ injections of hepatitis B recombinant vaccine. The anamnestic response rate ranged from 16%⁴⁴ to 32.5%⁴⁵ through 22%.⁴⁶ Positive anti-HBe seemed to be a predictor of the anamnestic response (43% vs 7% in antiHBe-negative patients, p = 0.035).⁴⁴ No strong evidence concerning the role of HCV co-infection has been found. A single injection of 20µg failed to achieve a satisfactory response rate.³⁶

Facing the absence of randomized controlled trials, no general recommendations can be made, even though it can be advocated that an anamnestic response has to be sought using a booster. The French guidelines recommend vacci-nating patients who do not initially respond, but without describing how.⁵²

Is there a place for adjuvants for HBV vaccination in people living with HIV?

Adjuvants to enhance the immunization achieved with hepatitis B vaccination have been tried (Table 3). They increase the immune response by improving presentation of the antigen and enhancing both the inflammatory and innate immune response. One of the most interesting is CPG 7909: a synthetic oligodeoxynucleotid containing CpG motifs acting as agonists for Toll-like receptor 9 (TLR 9) and then stimulating a Th-1 directed response.^64,65 One RCT⁶⁶ assessed a vaccination schedule containing Engerix B 40 µg with or without CPG7909 and found higher GMT at all measured times (from Week 6 to 48) in the CPG7909 group than in the control group (p < 0.01). They had a faster (more high responders at week 8, p = 0.022) and longer (week 48, p = 0.049) response in the CPG7909 group. Their population was relatively young (< 40 y old) and efficiently treated (HIV viral load < 50cp/mL, CD4 count > 200 cell/mm3). It was well tolerated but with more injection-site pain in the CPG 7909 group than in the saline placebo group (p = 0.02). When looking at long-term (60 months) seroprotection, the rate of responders was significantly higher throughout the study in the CPG 7909 group (p < 0.05), as was the rate of high-responders (67 vs 17% at week 24, p = 0.02 and 43 vs 12% at month 54, p<0.05).⁶⁷

Table 3.

Anti-HBV vaccination studies using adjuvants.

Study (year)	Design	Schedule	% HAART Median CD4 cell count (/µL)	Tolerance	Age / %male	Response rate
Cooper et al, 2005,	RCT	Engerix B 40µg	100%	Injection site pain	43 vs 41 YO	100% B40 + GMT vs 89% B40 alone
		+/−1 mg CPG 7909	543 (B40µ alone) vs 647 (B40 + CPG7909)
		M0–1–2			18 vs 14%
		N = 19 each arm				Higher GMT in B40 + CPG7909
Cooper et al, 2008	RCT II	Follow-up previous study at 5 years	idem	ok	Idem	90 vs50% M60 p<0.04
						HR M54 43 vs 12% p < 0.05
Sasaki et al, 2003	RCT	M0: 40µg HBsAg + 20µg GM-CSF or Placebo	90%	25% mild-moderate local SE.	25.9 vs 28.1	62 vs 30% J60 p<0.0074
		M1–6: 40µg HBsAg.	487 vs 462
		N = 40 each arm			52 vs 50%	NSD at D210
Overton et al, 2010	RCT	40µg HBsAg +/− 250µg GM-CSF M0–1–3	77 vs 79%	Mild-moderate local SE, higher in GM-CSF p = 0.01	40 vs 41	52 vs 65%
		N = 24 each arm	445 vs 461		79 vs 75%	GMT 27vs 92 UI/mL
						NSD
Overton et al, 2011	RCT	Failed 3*20µg	70% each arm	Higher GM-CSF group (p < 0.001)	37.3 vs 38.5	35 vs 50%, NS
		B 40µg +/− 250µg GM-CSF
		N = 60			40 vs 43.3%

RCT: Randomized Controlled Trial

HAART: Highly Active Anti RetroViral Therapy

YO: Years Old

SE: Secondary Effect

NSD: Non Significant Difference

M: Month D: Day GMT: Geometric Mean Titer

Recombinant human GM-CSF (rhGM-CSF) has also been tested as an adjuvant for hepatitis B vaccination in healthy individuals ⁶⁸ and chronic renal failure patients.^69-71 The results in the HIV-1 population are contradictory. Sasaki et al⁷² found a higher response rate in their GM-CSF with 40 µg of HBsAg than in the 40 µg HBsAg vaccine alone (62 vs 30%, p < 0.0074). But on day 210, the seroconversion rate did not differ (72 vs 60%, p = 0.23). Two RCT by Overton et al,^73,100 however, did not confirm these results. In the first one, their patients received 250µg GM-CSF as an adjuvant to a 40 µg accelerated schedule at 0–1–3 months but the vaccine-alone group had better response rates (65.2 vs 52.4%, NSD).⁷³ In the latest one,¹⁰⁰ comparing the adjunction of 250µg of GM-CSF to a classic schedule, the authors found that only 35% of persons receiving GM-CSF developed protective sAb while 50% in vaccine only arm developed protection (P = 0.47). Overall, only 28% of subjects maintained protective Ab 1 y after vaccination.

In summary, there are no sufficient data to legitimate the use of GM-CSF as adjuvant for hepatitis B vaccination in HIV-infected people, which thus does not appear in current recommendations. Regarding CPG 7909, further studies in the whole HIV population are needed.

Are there any specific regimens for HBV vaccination in children living with HIV? (Table 4)

Table 4.

Vaccination against HBV in children living with HIV.

Study (year)	Design	Schedule	% HAART Median CD4 cell count (/µL)	Mean Age (years old)	Response rate	Predictors
Diamant et al, 1994	Prospective	M0–1–6	88%	47 months	25%	% normal CD4 count (25 vs 52%, p = 0.014)
	N = 24	Recombivax 2.5µ / Engerix B10 < 11YO	460/mm3			Age (25 vs 55months, p = 0.016)
		Recombivax 5µ / Engerix 20 > 11YO.
Rutstein et al, 1994	Retrospective	W 0–4/8–6/9	None	3.1 months	35%	NA
	N = 355	Recombivax 2.5µg if mother HBsAg neg	1393 / µL	Weight = 2.71kg
	HIV = 17	Recombivax 5µ if mother HBsAg +
Scolfaro et al, 1996	Prospective	10µg M0–1–6	None	38 months	45%	NA
	N = 20	+/− 20µg	NA		Total: 85%
Zuin et al, 1992	Prospective	20µg M0–1–6	None?	659 days	78%	NA
	N = 18	+/− 20µg
Choudhury et al, 1995	Prospective	Booster doses		21–30 months	14%
	N = 14	< 11YO: Engerix B-20
		> 11YO: Engerix B 40
Bunupuradah et al, 2011	RCT	M0–2–6:	91.30%	12.1 YO	90,2% ID	Weight < 35kg
	N = 80	2µg ID/ 10 µg IM		50% ♂	92.3% IM
Flynn et al, 2011	RCT	M0–1–6	44%	20.44 months	60%	CD4 cell count OR1.32, p = 0.03
	N = 336	Engerix B 20µg
		Engerix B 40µg		39% ♂	73.2% p = 0.04
		Twinrix			75.45% p = 0.02
Lao-araya et al, 2010	Prospective	M0–2–6	100%	10.1YO	At 3 years: 71%	CD4 cell count > 500/µL
	N = 63	Engerix B 10µg		32% ♂
		Booster for non-responders			83% after booster
Bose et al, 2015	Prospective	Engerix B20µg M0–1–2–6	82%	8 YO	94% High responders 90%	NA
	N = 33			76% ♂
Abzug et al, 2009	Prospective	Fail classic schedule	100%	9.1 YO	46% W8	NA
	N = 204	Booster 5µ Recombivax + Pc + MMR		66% > 7 YO	38% W96
					Booster: 37%
				46% ♂

RCT: Randomized Controlled Trial

M: Month

ID: Intra Dermal

IM: Intra Muscular

OR: Odds Ratio

Pc: Pneumococcus

MMR: Measles Mumps Rubella

HAART: Highly Active Anti Retroviral Therapy

Vaccination against HBV is highly recommended in healthy infants and is included in the vaccine calendar.⁹⁰ It remains an easy, safe and effective way of protection. Healthy children and infants have demonstrated a good response to a 3-dose hepatitis B vaccination schedule with response rates around 90%.⁷⁴ By contrast, the response rate falls to only 28 to 78% in HIV-infected children,^75-80 and only 1% of HIV-infected children with immune recovery after HAART and prior HBV immunization still have protective antibodies at 5 y old or older.⁸¹ Double-dose regimens can raise seroconversion to 90%.^82,11,55

Current guidelines⁸³ recommend a 3-dose hepatitis B vaccine regimen for all children and adolescents aged <19 y old who were not previously vaccinated. However, antibody responses to hepatitis B vaccination may be diminished in HIV-infected children, especially in older children or those with CD4 T lymphocyte (CD4 cell) counts <200 cells/mm³.^79,81

One RCT ⁸⁵ compared 3 vaccination regimens: Engerix B* 20µg, 40 µg and Twinrix* 20µg / 720ELU; administered at weeks 0, 4 and 24 in children and adults aged 12–25 y old. Better response rates at week 28 were observed in the 40 µg (73.2%, p = 0.04) and the Twinrix* groups (75.45%, p = 0.02) than in the standard group, with no safety issues for any of the regimens. The GMT was also higher in the Twinrix* group than in the 20µg group (97.7 vs 52 mIU/mL, p = 0.03). However, no significant difference in the persistence of the response to vaccine at week 72 was observed between groups. The only independent predictor of a positive response was the baseline CD4 T cell count.

The only RCT ⁸⁴ comparing intradermal and intramuscular routes found a lower response with the intradermal route (GMT 157 vs 459 when administered intramuscularly, p < 0.0001), and with fewer high responders (56.1 vs 82%, p = 0.01).

Regarding re-vaccination, Lao-araya et al. ⁸⁶ observed satisfactory response rates following a 3-10µg-dose HBV vaccine schedule in a relatively young population with controlled HIV: up to 92.1% at 7 months with 76.2% of high responders and a GMT of 4076 mIU/mL. Children whose HIV was controlled (HIV viral load below the limit of detection at the time of re-vaccination) were more likely to respond to HBV re-vaccination (p<0.001). Scolfaro et al. achieved better seroconversion results by doubling the dose of intramuscular HBV vaccine given at 0, 1 and 6 months.⁸⁷ Another prospective cohort study observed high rates of responders and of high-responders 6 weeks after the last injection with a 4-double-dose schedule at months 0–1–2–6 (94 and 90%, respectively).⁸⁸ Their young (median age 8 y old) and well controlled (median CD4 count cell = 738/mm³) population with almost all subjects under HAART can explain these good results. Other authors⁸⁹ observed a response rate of 46% in a population of 2–18 year-olds with perinatal infections, who were vaccinated with 5µg of HBs antigen combined with pneumococcal and measles-mumps and rubella. A response to the booster vaccination occurred in 37%. GMT after boosting remains low (190 mIU/mL) compared with mean concentrations of 7096–15263 MIU/mL among HIV-uninfected children 1 month after boosting 5 y after the primary vaccination.^91,92

Limited data suggest that modified hepatitis B vaccine dosing regimens, including a doubling of the standard antigen dose and the use of combined hepatitis A and B (HAV/HBV) vaccine, can increase response rates in HIV-uninfected non-responders and in HIV-infected adults and adolescents. Therefore, the use of double-dose HBV vaccines or combination HAV/HBV vaccines may be considered for HBV vaccination in HIV-infected adolescents.⁸³ Monitoring the antiHBs titer remains necessary, as it has been reported that only 42% of HIV-infected responders kept their protective titers 13–18 months after their last dose of vaccine.⁸⁷

Conclusion

In conclusion, several clear recommendations regarding HBV vaccination in HIV-infected populations now exist.^{8,18,51-53,83} Reinforced HBV vaccination, using the schedule of 40 µg IM at months 0–1–2–6, is now recommended in HIV-infected patients with negative hepatitis B serology. The anti-HBs titer has to be systematically checked 1 to 2 months following the last dose of HBV vaccine, so as not run the risk of missing a genuine response to vaccination, which may be difficult to establish later considering the decrease in anti-HBs antibodies over time. Annual antiHBs control is also recommended so that a booster can be given if the antiHBs titer falls below 10 mIU/mL. In patients with a low CD4 cell count (< 200/µL), current guidelines recommend starting the vaccination regimen as soon as possible after HAART has been started, without waiting for immune restoration.

On the other hand, some uncertainties remain.^8,18,53 For IAHBc, a first booster injection is required to seek an anamnestic response. The need to fully vaccinate those who do not respond to the booster injection is not yet established, but it can be advocated that such a vaccination could be of interest in patients with practices that carry a high-risk of contamination. In addition, in the light of current data, the use of adjuvants does not appear to be recommended for HBV vaccination in people living with HIV. New studies are needed to better define the best strategies in such populations.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.