Adjuvanted-influenza vaccination in patients infected with HIV: a systematic review and meta-analysis of immunogenicity and safety

ABSTRACT

Adjuvanted-influenza vaccination is an efficient method for enhancing the immunogenicity of influenza split-virus vaccines for preventing influenza. However, the medical community’s understanding of its performance in patients infected with HIV remains limited. To identify the advantages, we conducted a systematic review and meta-analysis with randomized controlled trials (RCTs) and cohort and case–control studies that have the immunogenicity and safety of influenza vaccines in patients infected with HIV as outcomes. We searched six different databases, and 1698 patients infected with HIV in 11 studies were included. Statistical analysis was performed to calculate the pooled standardized mean differences (SMD) or relative risk (RR) and 95% confidence interval (CI). Regarding immunogenicity, the pooled SMD of GMT (Geometric mean titer) for A/H1N1 was 0.61 (95%CI (0.40,0.82)), the pooled RR of seroconversion was 1.34 (95%CI (0.91,1.98)) for the H1N1 vaccine, 1.27(95%CI (0.64,2.52)) for the H3N2 vaccine, 1.19(95%CI (0.97,1.46)) for the B-type influenza vaccine. The pooled RR of seroprotection was 1.61 (95%CI (1.00,2.58)) for the H1N1 vaccine, 1.06 (95%CI(0.83,1.35)) for the H3N2 vaccine, and 1.13(95%CI(0.91,1.41)) for the B-type vaccine. Adjuvanted-influenza vaccination showed good general tolerability in patients infected with HIV, with the only significant increase being the rate of local pain at the injection site (RR = 2.03, 95%CI (1.06,3.86)). In conclusion, all studies evaluating injected adjuvanted influenza vaccination among patients infected with HIV showed acceptable levels of safety and immunogenicity.

1. Introduction

According to the antigenicity of nucleoprotein and matrix protein, the influenza virus is classified into three genera: influenza A, B, and C. Influenza A and B viruses usually cause seasonal outbreaks in crowds because of its antigenic variation, while influenza C viruses rarely causes illness.¹ Due to high morbidity and mortality, influenza pandemics continue to be a significant public health and economic burden. Based on the estimate of the World Health Organization (WHO), there are 3–5 million severe cases and 250,000 to 500,000 deaths worldwide.² Compared to antiviral drugs, the WHO suggests that influenza vaccination is the most effective method for preventing influenza infection and reducing the occurrence of severe infections. The most widely used influenza vaccine is the trivalent vaccine, which includes three different subtypes of the influenza A virus. However, with the change in epidemic strains of the influenza virus, the infection rate of the influenza B virus is increasing. For the populations immunized with the trivalent vaccine, the influenza vaccine does not provide sufficient protection. Therefore, another lineage of the influenza B virus is added to the trivalent vaccine to prepare a tetravalent vaccine including influenza A and B viruses.³

HIV (human immunodeficiency virus) infection is associated with increased severity of influenza and the risk of complications, resulting in excess hospitalization and mortality during influenza seasons. Therefore, influenza vaccination is recommended in many countries to these at-risk individuals.^4-8 Despite these recommendations, the rates of injection are low, mainly caused by negative attitudes to medication resulting in avoidance, fear of the side-effects, and not considering influenza a severe disease, which threatens more immune-deficient patients.^9,10 Existing systematic literature reviews and meta-analyses have revealed the function of trivalent inactive influenza vaccines among patients infected with HIV.^11,12 However, these reviews have limited recent evidence and do not report the adjuvanted influenza vaccines that might be more effective and safer for patients infected with HIV.

The majority of studies indicate that adjuvants play a critical role in regulating the immune system, antigen presentation, and immune elimination.^13-15 The presence of renewed adjuvants of human vaccines coming onto the market and the spreading of influenza vaccines compensates for the flaws of influenza vaccines, such as low immunogenicity.¹⁶At present, MF59 and AS03 are the only adjuvants that have been used in commercial influenza vaccines，and these are both oil-in-water adjuvants. One dose of AS03 (AS03A) adjuvanted vaccine contains 10.69 mg squalene, 11.86 mg a-tocopherol, 4.86 mg polysorbate 80, and 3.75 lg HA. One dose of MF59 (MF59(f)) adjuvanted vaccine contains 9.75 mg squalene, 1.175 mg polysorbate 80, 1.175 mg sorbitan trioleate, and 7.5 lg HA.^13,14 Their compositions are similar, and they efficiently reduce the dosage of antigens from influenza viruses in vaccines,^17,18 along with enlarging the productivity of the vaccine.¹⁹ To ensure optimal knowledge regarding vaccine-induced clinical protection and to optimize immunization policy decisions, we conducted a systematic review of the current medical literature to identify advantages in the data for the immunogenicity and safety of adjuvanted-influenza vaccination in patients infected with HIV.

2. Methods

2.1. Search strategy

This review was conducted and analyzed based on the recommendations of the Cochrane handbook and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. According to the searching strategies formulated by the PICOS principles, articles published before January 2019 were searched in PubMed, the Cochrane Library, Embase, Web of Science, the China National Knowledge Infrastructure (CNKI), and the WANFANG Data databases. The following search terms were used: ‘influenza,’ ‘flu,’ ‘grippe,’ ‘vaccine,’ ‘vaccination,’ ‘adjuvanted,’ ‘human immunodeficiency virus,’ ‘AIDS,’ and ‘HIV.’ The reference lists of all included articles were searched for additional studies.

2.2. Inclusion criteria

(1) Types of studies: studies were limited to randomized controlled trials (RCTs), cohort studies, case-control studies, and self-control studies. Only studies published in English or Chinese were considered for inclusion in this research. (2) Types of participants: the population studied was patients infected with HIV from anywhere in the world. (3) Types of interventions: the experimental group was injected with inactivated adjuvanted influenza vaccine, and the control group was injected with inactivated non-adjuvanted influenza vaccine or were unvaccinated. (4) Types of outcomes: GMT (Geometric mean titer), the rate of seroconversion, the rate of seroprotection, and adverse events from the adjuvanted influenza vaccine. GMT, reflecting on body immunoreaction, is a critical metric to predict the persistence of protective antibodies in serum. Seroconversion and seroprotection rates were also considered as indexes of the outcome of serological immunoreactions. The definition of seroprotection is the titer of antibody >1:40 and the definition of seroconversion is a change from a baseline titer of <1:10 to a post-vaccination titer of ≥1:40, or a four-fold increase in titer in those with an initial HI titer of ≥1:10.

2.3. Study selection and data extraction

Two authors (YCC and JHZ) identified and screened potentially eligible studies by reviewing the title and abstract of each article. Relevant publications were then selected for full-text review. Studies whose relevance was not clear were discussed and a decision on whether to exclude it was made. In the case of discrepancies in an included study between investigators, a third investigator (JMT) made the definitive decision via discussion.

YCC and JHZ, respectively, assessed the relevance of search results and extracted the basic information of included studies in an Excel format. Discrepancies were resolved by a discussion that persisted until a consensus was achieved. Extracted information included: general study data (first author, year of publication, country, study year, age, study setting, sample size, duration of follow-up, etc.), quality of data (hidden, blinded, follow-up, etc.), interventions and outcomes (GMT, seroconversion, seroprotection and adverse events). These study selections and data extractions were performed according to the Cochrane Handbook.²⁰

2.4. Quality assessment

To assess the quality of RCTs, Cochrane’s Collaboration tool for assessing the risk of bias was used,²⁰ which includes selection, performance, detection, attrition, reporting, and other biases. The methodological quality of the included cohort, case–control and self-control studies was investigated according to the Newcastle-Ottawa Scale (NOS),²¹ using the following criteria: selection (0 − 4), comparability (0 − 2), and outcome/exposure (0 − 3). NOS scores for self-control studies were modified to exclude comparability and higher scores indicated improved quality.

2.5. Data analysis

Statistical analyses were performed using RevMan version 5.3 (provided by the Cochrane Collaboration) and STATA, version 12.0 (Stata Corp., College Station, TX, USA). If the sample standard deviation (SD) was not disclosed, SD was calculated from 95% CI according to the Cochrane Handbook. Statistical heterogeneity among included studies was determined by Q test and I² index, in which I² < 50% or p > .10 indicated that significant heterogeneity did not exist. If heterogeneity was not observed among the studies, the fixed-effects model was applied. Otherwise, the random-effects model was used. The analysis of subgroups and sensitivity analysis were adopted to identify and reduce sources of variability between studies. Publication bias was assessed through funnel plots and Egger’s linear regression analysis. Standardized mean differences (SMD) were determined to acquire the overall estimated average of GMTs between each group. The RR (Relative Risk) for seroconversion and seroprotection was used to determine the immunogenicity of the influenza vaccine. The RR and 95% CI for general and localized symptoms were used to determine the safety of the influenza vaccine. The interval is estimated at 95% Cl, and the test level is α = 0.05.

3. Results

3.1. Search results

In total, 184 studies were identified in the initial literature search. There were 93 duplicate studies that were removed using EndNote reference manager, and 21 reviews and 46 irrelevant studies (patients not infected with HIV or use of an unadjuvanted influenza vaccine) were excluded by reading the titles and abstracts. The study by Launay was excluded from the meta-analysis to avoid double-counting, as the year of study and date were largely repeated in another study by Durier.^22,23 These exclusions resulted in only 11 studies^23-33 being identified for this systematic review which involved a total of 1698 patients infected with HIV. A schematic illustration flowchart of the literature search and the study-selection criteria are presented in Figure 1. Of the included studies, four were RCTs,^23-26 two were cohort studies^29,31 and five were self-control studies.^{27,28,30,32,33}

Figure 1.

Flowchart showing the selection of study for the meta-analysis.

3.2. Study characteristics

3.2.1. Randomized controlled trials

Four RCTs^23-26 included in this meta-analysis provided direct evidence for the safety and immunogenicity of adjuvanted influenza vaccination in patients infected with HIV. Three RCTs directly compared the vaccine efficacy of the MF59-adjuvanted trivalent influenza vaccine and the unadjuvanted trivalent influenza vaccine and the follow-up periods in the studies were 5 months,²⁴ 6 months,²⁵ and 3 months.²⁶ It was reported that the adjuvanted vaccine yielded a higher short-term immune response than the nonadjuvanted vaccine without impacting the HIV infection.^24-26 In another study,²³ adults of the test group infected with HIV received two-doses of the AS03_A-adjuvanted A (H1N1) vaccine which were administered 21 days apart. After injecting the first dose, it was observed that the adjuvanted vaccine conferred a more intensive and higher-level immune response than the nonadjuvanted vaccine. However, all RCTs originated from European countries (Italy and France), so the studies have some limitations. The details of the RCTs identified in this study are shown in Table 1.

Table 1.

Description of characteristics of included studies.

							Sample size		Interventions
First author[ref]	Year of publication	Study year	Country	Study design	Participants	Age	T	C	T	C	Vaccine	Dose	Antibody measurement	NOS score
Iorio, A. M.1	2003	NR	Italy	RCT	HIV-seropositive patients	NR	44	40	MF59-adjuvanted trivalent influenza vaccine	Unadjuvanted trivalent influenza vaccine	A/Moscow/10/99-like (H3N2), A/New Caledonia/20/99-like (H1N1) and B/Beijing/184/93	1 dose	HAI	-
Gabutti, G.2	2005	2002-2003	Italy	RCT	HIV-1-seropositive patients	18-65 y	18	19	MF59-adjuvanted trivalent influenza vaccine	Unadjuvanted trivalent influenza vaccine	A/Moscow/10/99 (H3N2), A/New Caledonia/ 20/99 (H1N1) and B/Hong Kong/330/2001	1 dose	HAI	-
Durando, P.3	2008	2005-2006	Italy	RCT	HIV-1-seropositive patients	18-65 y	46	49	MF59-adjuvanted trivalent influenza vaccine	Unadjuvanted trivalent influenza vaccine	A/New Caledonia/ 20/99 (H1N1), A/California/7/2004 (H3N2), and B/Shanghai/361/2002	1 dose	HAI	-
Fabbiani, M.4	2011	NR	Italy	Self-control	HIV-infected adults	46(41–55) y*	41	-	MF59-adjuvanted H1N1 vaccine	-	A/California/7/2009 (H1N1)	1 dose	HAI	S2, CNR, O3, T5
Okike, I. O.5	2011	2009	United Kingdom	Self-control	HIV-infected children	NR	31	-	AS03B-adjuvanted H1N1 vaccine	-	A/California/7/2009（H1N1）	1 dose	HAI	S2, CNR, O3, T5
Tremblay, C. L.6	2011	2009	Canada	Self-control	HIV-positive adults	49(41.5–55.5) y*	89	-	AS03-adjuvanted H1N1 vaccine	-	A/California/7/2009（H1N1）	1 dose	HAI，MN	S2, CNR, O2, T4
Nielsen, A. B.6	2012	2010	Denmark	Cohort	HIV positive adults	25-80 y	159	60	AS03-adjuvanted H1N1 vaccine	Unvaccinated	A/California/7/2009（H1N1）	1 dose (31); 2 doses (128)	HAI	S3, C2, O3, T8
Durier, C.8	2013	2009	France	RCT	HIV-1infected adults	46.9 (40.0–53.8) y*	155	150	AS03A-adjuvanted H1N1 vaccine	Unadjuvanted H1N1v vaccine	A/California/7/2009（H1N1）	2 doses	HAI, MN	-
Pauksens, K.9	2013	NR	Sweden	Self-control	HIV-1infected adults	25-82	42	-	AS03-adjuvanted H1N1 vaccine	-	A/California/7/2009（H1N1）	2 doses	HAI	S2, CNR, O3, T5
Bickel, M.10	2014	NR	Germany	Cohort	HIV-1infected adults	NR	163	203	AS03-adjuvanted H1N1 vaccine	Unvaccinated	A/California/7/2009（H1N1）	1 dose (67); 2 doses (96)	HAI	S3, C2, O3, T8
Schwarze-Zander, C.11	2016	2009	Germany	Self-control	HIV positive patients	22-78	389	-	AS03-adjuvanted H1N1 vaccine	-	A/California/7/2009（H1N1）	1 dose	HAI	S2, CNR, O3, T5

Ref: reference; T: the test group; C: the control group;*: median (interquartile range); y: year; d: day; RCT: randomized controlled trial; NR: not reported; CI: confidence interval; HAI: Hemagglutination inhibition assay; MN: Microneutralization assay; NOS: the Newcastle–Ottawa Scale for studies; S: the number of stars for selection section; C: the number of stars for comparability section; O: the number of stars for outcome section; T: the total number of stars.

3.2.2. Cohort studies

Only two cohort studies^29,31 were identified in the systematic review. The subjects were adults who were HIV positive at hospitals in Denmark and Germany. One study³¹ was retrospective, and several limitations are acknowledged. In two cohort studies, the dates for influenza vaccination history, demographics, and post-vaccination reactions were collected through face-to-face interviews and questionnaires of all participants. All patients were divided into three groups (unvaccinated, one dose, and two doses) depending on their own information of vaccination status. One dose and two-dose groups were combined to create a single pair-wise comparison to overcome a unit-of-analysis error. The details of the cohort studies are shown in Table 1.

3.2.3. Self-control studies

Five self-control studies^{27,28,30,32,33} included in this meta-analysis originated from Italy, United Kingdom, Canada, Sweden, and Germany. Only one study²⁷ drew a comparison between before and after MF59-adjuvanted H1N1 vaccination, whereas the other trials used AS03-adjuvanted H1N1 vaccination. Patients were injected with two-dose vaccine during 21 days in two studies.^28,30 In only one study by Okike,²⁸ all participants were children. The concrete information is shown in Table 1.

3.3. Quality assessment

The risk of bias graph (see Figure 1, Supplemental Digital Content 1) showed the low quality of all RCTs, because few RCTs adopted blinding,²³ all RCTs merely mentioned random and unreported random methods. The NOS score of Cohort studies was eight stars, indicating that the quality of Cohort studies was well. Simultaneously, the modified NOS score of the self-control studies ranged from four to five stars, manifesting that the quality of the self-control studies was general (Table 1).

3.4. Meta-analysis of immunogenicity and safety

Adjuvants provide evident immune-enhancing effects in vaccines.^13,14 At present, MF59 and AS03 are the only adjuvants that have been used in commercial influenza vaccines. To better understand the immunogenicity and safety of adjuvants，MF59 & AS03 trials were treated in an overall meta-analysis and subdivided into MF59-only and AS03-only trials.

3.4.1. Immunogenicity

Immunogenicity was measured by determining GMTs, seroconversion rates, and seroprotection rates after vaccination in serum samples. All serum samples were tested by hemagglutination inhibition (HAI) assay in all included studies, and the sera in two studies were further tested by microneutralization (MN) assay. The meta-analysis specifically integrated the HAI results presented in included studies to reduce heterogeneity and bias.

3.4.1.1. GMT

Regarding the meta-analysis of GMT, only seven studies were included, because one study³² did not report it and three studies^24-26 reported the incomplete data of GMT. Because the incomplete data of GMT were reported in three trials with trivalent reported， only the meta-analysis for the GMT for A/H1N1 responses was performed. A random-effects model was employed due to the significant heterogeneity among these studies (I² = 65.6%>50.0%, p = .050 < 0.10). GMT were determined with SMD and the pooled SMD was 0.61 (95%CI (0.40,0.82), p = .000 < 0.01 Figure 2). The pooled SMD of GMT for adjuvanted vs. unadjuvanted was 0.44 (95%CI (0.29,0.60), p = .000 < 0.01), showing that adjuvanted-influenza vaccination could improve the GMT antibody titers for the A/H1N1 in patients infected with HIV. Thereafter, a subgroup analysis was conducted for deeply researching the multifactor effect which may leaded to heterogeneity and bias, such as study design, types of adjuvanted, contrast and the year of the vaccine (Table 2). Concurrently, the following sensitivity analysis showed that individual studies had little effect on the stability of final results (see Figure 2a, Supplemental Digital Content 2). Results from funnel plots and Egger’s linear regression analysis showed no significant asymmetry or publication bias (p = .562 > 0.05).

Figure 2.

Forest plot of GMT for the A/H1N1 vaccine included in seven studies. (GMT, Geometric mean titer; SMD, standardized mean differences; CI, confidence interval. The diamond represents the pooled estimate with a 95%CI).

Table 2.

GMT for A/H1N1 responses by subgroup analysis.

Subgroup	Number of studies	SMD (95%CI)
Overall	7	0.61(0.40,0.82)
Study design
RCT	1	0.45(0.21,0.69)
Cohort	2	0.89(0.71,1.06)
Self-control	4	0.45(0.22,0.68)
Types of adjuvanted
MF59	1	0.25(−0.19,0.68)
AS03	6	0.65(0.44,0.87)
Age
≥18	6	0.57(0.35,0.80)
< 18	1	0.89(0.37,1.42)
The year of the vaccine
2009	7	0.61(0.40,0.82)
Before 2009	0	-
Contrast
Adjuvanted vs. Unadjuvanted	5	0.44(0.29,0.60)
Adjuvanted vs. Unvaccinated	2	0.89(0.71,1.06)

3.4.1.2. Seroconversion

The outcomes of seroconversion rates were constituted by four RCTs, one cohort studies and five self-control studies. However, a retrospective cohort study by Bickel³¹ was excluded because of lacking seroconversion rate. For each virus strain of influenza vaccine, the combined seroconversion rate of injecting adjuvanted-influenza vaccine was approximately 60% in patients infected with HIV (A/H1N1:57.74% (481/833); A/H3N3:61.97%(88/142); B:61.27%(87/142)). The pooled results showed (Figure 3a, 3b, 3c) that the adjuvanted vaccination could increase the seroconversion rate in patients infected with HIV, but not significantly. The pooled RR for the H1N1 vaccine was 1.34 (95%CI (0.91,1.98)). For the H3N2 vaccine, the pooled RR was 1.27(95%CI (0.64,2.52)). For the B-type influenza vaccine, the pooled RR was 1.19(95%CI (0.97,1.46)). We conducted subgroup analysis of potential factors leading to heterogeneity (Table 3), types of adjuvanted, contrast and the year of the vaccine were sources of high heterogeneity for seroconversion. For A/H1N1 of influenza vaccine, the pooled RR (RR = 1.97, 95%CI (1.14,3.41), p = .015 < 0.05) of AS03 was in the direction of favoring vaccination, but MF59 (RR = 0.95, 95%CI (1.14,3.41), p = .632 > 0.05) did not clearly show its effect. And the difference between adjuvanted and unadjuvanted vaccination on seroconversion was not significant, but adjuvanted was generally improved (RR = 1.15, 95%CI (0.80,1.63), p = .452 > 0.05). Finally, the sensitivity analysis manifested the results were credible (see Figure 2b, Supplemental Digital Content 2). Because of the small number of studies included in this meta-analysis of seroconversion, it was difficult to visually determine the absence or presence of asymmetry by examining this plot. However, the Egger’s regression tests (p = .912 > 0.05) did not show any evidence of publication bias.

Figure 3.

Forest plot for seroconversion in RCTs and cohort studies. (a) Forest plot of seroconversion for the A/H1N1 vaccine. (b) Forest plot of seroconversion for the A/H3N3 vaccine. (c) Forest plot of seroconversion for the B-type influenza vaccine. (RR, relative risk; CI, confidence interval. The diamond represents the pooled estimate with a 95%CI.).

Table 3.

Seroconversion and seroprotection by subgroup analysis.

	A/H1N1		A/H3N3		B
Subgroup	n	RR (95%CI)	n	RR (95%CI)	n	RR (95%CI)
Seroconversion
Overall	5	1.34(0.91,1.98)	3	1.27(0.64,2.52)	3	1.19(0.97,1.46)
Study design
RCT	4	1.15(0.80,1.63)	3	1.27(0.64,2.52)	3	1.19(0.97,1.46)
Cohort	1	2.62(1.30,3.83)	0	-	0	-
Types of adjuvanted
MF59	3	0.95(0.77,1.17)	3	1.27(0.64,2.52)	3	1.19(0.97,1.46)
AS03	2	1.97(1.14,3.41)	0	-	0	-
Age
≥18	5	1.34(0.91,1.98)	3	1.27(0.64,2.52)	3	1.19(0.97,1.46)
< 18	0	-	0	-	0	-
The year of the vaccine
2009	3	0.95(0.77,1.17)	0	-	0	-
Before 2009	2	1.97(1.14,3.41)	3	1.27(0.64,2.52)	3	1.27(0.64,2.52)
Contrast
Adjuvanted vs Unadjuvanted	4	1.15(0.80,1.63)	3	1.27(0.64,2.52)	3	1.19(0.97,1.46)
Adjuvanted vs Unvaccinated	1	2.62(1.30,3.83)	0	-	0	-
Seroprotection
Overall	6	1.61(1.00,2.58)	3	1.06(0.83,1.35)	3	1.13(0.91,1.41)
Study design
RCT	4	1.12(0.87,1.45)	3	1.06(0.83,1.35)	3	1.06(0.83,1.35)
Cohort	2	1.61(1.00,2.58)	0	-	0	-
Types of adjuvanted
MF59	3	1.02(0.82,1.27)	3	1.06(0.83,1.35)	3	1.06(0.83,1.35)
AS03	3	2.52(1.18,5.37)	0	-	0	-
Age
≥18	6	1.61(1.00,2.58)	3	1.06(0.83,1.35)	3	1.06(0.83,1.35)
< 18	0	-	0	-	0	-
The year of the vaccine
2009	3	1.02(0.82,1.27)	0	-	0	-
Before 2009	3	2.52(1.18,5.37)	3	1.06(0.83,1.35)	3	1.06(0.83,1.35)
Contrast
Adjuvanted vs. Unadjuvanted	4	1.12(0.87,1.45)	3	1.06(0.83,1.35)	3	1.06(0.83,1.35)
Adjuvanted vs. Unvaccinated	2	1.61(1.00,2.58)	0	-	0	-

3.4.1.3. Seroprotection

There were eleven studies providing dates on seroprotection rates of adjuvanted influenza vaccination in patients infected with HIV, and the total rate over 70% (A/H1N1:73.04% (867/1187), A/H3N3:85.21% (121/142), B:91.55% (130/142)). The combined RR of RCTs and cohort studies in seroprotection rates was 1.61 (95%CI (1.00,2.58)) for the H1N1 vaccine, 1.06 (95%CI(0.83,1.35)) for the H3N2 vaccine, 1.13(95%CI(0.91,1.41)) for the B-type vaccine (Figure 4a, 4b, 4c). For the H1N1 vaccine, the types of adjuvanted and the year of the vaccine were sources of high heterogeneity by subgroup analysis (Table 3), and individual studies had no effect on the stability of results by sensitivity analysis (see Figure 2c, Supplemental Digital Content 2). In addition, the Egger’s regression tests (p = .145 > 0.05) for seroprotection rates did not show significant publication bias.

Figure 4.

Forest plot for seroprotection in RCTs and cohort studies. (a) Forest plot of seroprotection for the A/H1N1 vaccine. (b) Forest plot of seroprotection for the A/H3N3 vaccine. (c) Forest plot of seroconversion for the B-type influenza vaccine. (RR, relative risk; CI, confidence interval. The diamond represents the pooled estimate with a 95%CI).

Generally, the results of GMT, seroconversion and seroprotection rates indicated that adjuvanted influenza vaccination in patients infected with HIV could improve the immunogenicity.

3.4.2. Safety

Eight studies^{23,25,26,28,29,31-33} reported on the tolerability after injecting the adjuvanted-influenza vaccine, and none of these reported any fatal adverse reactions. The adjuvanted influenza vaccination in patients infected with HIV was generally well tolerated. However, local side effects (overall; 39.83% (231/580); MF59-trivalent: 29.68% (19/64); MF59 H1N1 (monovalent): not reported; and ASO3 H1N1 (monovalent): 41.09% (212/516)) were common in patients. The rates of swelling and redness were 11.2% (MF59-trivalent: 4.69% (3/64); and ASO3 H1N1 (monovalent): 12.01% (62/516)) and 6.90% (MF59-trivalent: 6.25% (4/64); and ASO3 H1N1 (monovalent): 6.98% (36/516)) respectively. With regards to the overall systemic reaction, the rates of fatigue, headache, fever, myalgia, and arthralgia were 26.55% (154/580), 19.48% (113/580), 6.55% (38/580), 8.10% (47/580), and 5.69% (33/580), respectively. However, the majority of the side effects were mild to moderate. Compared to the unadjuvanted-vaccine, the adjuvanted-vaccine was well tolerated but also associated with the increasing occurrence of local pain at the injection site (RR = 2.03, 95%CI (1.06,3.86), p < .05; MF59-trivalent: 25.00% (16/64); and ASO3 H1N1 (monovalent): 39.34% (203/516)) in patients infected with HIV.

4. Discussion

The risk of influenza-related complications and a more severe course of disease are higher for individuals infected with HIV. Therefore, annual influenza vaccination is recommended for those infected with HIV as it can efficiently reduce the occurrence of fatal diseases related to influenza infection.^5,7,34 Adjuvants, in the context of vaccines, are defined as components capable of enhancing or shaping antigen-specific immune responses.¹⁴ Adjuvanted-influenza vaccination in patients infected with HIV has certain advantages. However, individuals were still concerned about its effectiveness and safety, causing the rates of injection to be low.^10,18 Based on this occurrence, we conducted this systematic review to further explore the immunogenicity and safety of adjuvanted-influenza vaccination in patients infected with HIV. In addition, this is the first systematic review and meta-analysis to directly evaluate the immunogenicity and safety of adjuvanted influenza vaccination in patients infected with HIV. These results can provide guidance for scheduling adjuvanted influenza vaccination policies in patients infected with HIV.

By reviewing the published literature results, a lack of direct comparison between individuals who are infected with HIV and those who are healthy, with regards to the immunogenicity and safety of the influenza vaccine, was identified. By comparing the results of the published studies, it was found that the incidence of serious adverse events due to vaccination, such as cardiovascular, was very low in both the general population and the patients infected with HIV.³⁵ In addition, Trotta F et al. indicated that the incidence of adverse events, whether injected with MF59 adjuvant influenza vaccine or not, was similar in pregnant women.³⁶ However, there were still confounding factors that implied the increased incidence of congenital malformations. Therefore, further analysis is needed. Notably, O’Brien WA et al. found that the peak level of HIV viral RNA in patients infected with HIV’s peripheral blood mononuclear cells (PBMCs) increased significantly after influenza vaccination. This result might indicate that the influenza vaccine will promote HIV infection progression.³⁷ In immunogenicity, individuals infected with HIV have a weaker immune response to the influenza vaccine than individual not infected by HIV. However, by adjusting the dose and increasing the adjuvant, the efficiency of the influenza vaccine in patients infected with HIV can be increased. The reason that individuals infected with HIV have lower response rates to the influenza vaccines than the individuals not infected by HIV are mainly due to the destruction of the immune system by HIV. Pallikkuth S et al. injected 16 individuals infected with HIV and eight healthy individuals with H1N1/09 influenza vaccine, and eight infected individuals did not have a significant immune response compared with the healthy individuals. It was found that this result was associated with the impaired function of the peripheral blood T-follicular helper cell peripheral in patients infected with HIV.³⁸ Moreover, the failure of the influenza vaccine in inducing antibody response in individuals infected with HIV is also associated with impaired B cell function. Pallikkuth S et al. also found defects in BAFF (B cell-activating factor) and APRIL (a proliferation-inducing ligand) and their receptors on memory B cells in individuals infected with HIV, which is associated with the failure of the antibody response in individuals infected with HIV.³⁹

Our findings suggest that adjuvanted influenza vaccination is highly immunogenic in patients infected with HIV when compared with unadjuvanted influenza vaccination or no vaccinated. Based on these studies, GMT, seroconversion rates, and seroprotection rates were evaluated. The outcome of GMT for A/H1N1 responses showed high levels in the titer of antibodies after the adjuvanted-influenza vaccination in patients infected with HIV, compared with unadjuvanted-influenza vaccination and no vaccinated. In addition, adjuvanted influenza vaccination resulted in high seroconversion rates (A/H1N1:57.74%; A/H3N3:61.97%; B:61.27%) and seroprotection rates (A/H1N1:73.04%; A/H3N3:85.21%; B:91.55%). The seroconversion results were not statistically significant, with the 95% CI broadly crossing 1, yet the adjuvanted influenza vaccination for A/H1N1 responses was improved compared to no vaccination. And the reason that the difference of seroconversion and seroprotection between adjuvanted and unadjuvanted is not clear, it may be related to the dose of vaccination. All subjects of the studies by Durier and Pauksens and some of the study subjects for Nielsen and Bickel were vaccinated with two doses, while others were vaccinated with one dose. A few studies evaluating immune response after one or two doses of adjuvanted influenza vaccine in patients infected with HIV, it was found that one dose of adjuvanted influenza vaccine was quite immunogenic in HIV patients, while a second dose significantly increased protection and offered long-term antibody response, supported the recommendation of two double doses of adjuvanted influenza vaccine for individuals infected by HIV.^29,31,40-43 Overall, adjuvanted influenza vaccine could improve the immunogenicity in patients infected with HIV. However, the confirmation of these results still need further studies to monitor the immunogenicity of adjuvanted vaccination. Based on the results of this study, it was found that the AS03 adjuvant caused more adverse reactions than the MF59 adjuvant in patients infected with HIV. The composition of AS03 and MF59 was found to be different. AS03 contains the immunostimulant α-tocopherol, while MF59 is only an emulsion containing squalene. Related studies have confirmed that α-tocopherol can induce muscle and lymph node-specific immune activation, while MF59 only specific immune activation of the injected muscle is activated.^44,45 In addition, Morel et al. demonstrated that α-tocopherol in AS03 promotes cytokine production and enhances the immune response.⁴⁶

Although the latest guidelines have recommended annual adjuvanted-influenza vaccination for individuals infected by HIV, the rates of injection were still low in each region. The main reason was that individuals are concerning about vaccine’s safety, therefore evaluating the safety of adjuvanted-influenza vaccine is a vital step to popularize adjuvanted-influenza vaccination.^9,10,47 This study showed that local pain was common, and no evidence proving that the adjuvanted-influenza vaccination attributed to a severe adverse reaction. According to the published article,⁴⁴ we found that adjuvanted-influenza vaccination could upregulate the proinflammatory genes which prompting leukocyte migration to produce more inflammatory factor and raise more inflammatory cells in the vaccination site, which might be a strong evidence for the specified occurrence of local pain. Some studies^26,48-50 showed no significant changes either in the occurrence of viremia or the declining of CD4⁺ cells count, which certified that the safety of adjuvanted-influenza vaccination was excellent at any time point. Esposito and colleagues, in their studies, found that the adjuvanted influenza vaccine is safe in patients infected with HIV as well as the individuals not infected with HIV.⁴⁸ Overall, no new unsafety information was identified in this meta-analysis and the results support an acceptable safety profile of adjuvanted influenza vaccination.

This study has several limitations that must be acknowledged. First, only six databases were searched and some unpublished studies or publications in other databases may not have been identified. Second, this meta-analysis included a limited number of studies. The self-control studies were included in the analysis due to a lack of RCTs. Although self-control study evidence is generally weaker than RCTs, these studies do provide reference values. Third, a significant amount of heterogeneity was observed with respect to the study design. The sources of heterogeneity were attempted, but some studies did not report complete data on risk factors. Nevertheless, the results of sensitivity analyses indicated that the conclusions are reliable. Furthermore, although seroconversion and seroprotection are regarded as reliable parameters for the evaluation of the immunogenicity of influenza vaccination, the pre-vaccination baseline antibody titers were not similar in the included studies.

5. Conclusion

Available evidence suggests that adjuvanted-influenza vaccinations provide superior protection to reduce the incidence of influenza in individuals infected with HIV. With regard to side effects, local pain at the injection site has a higher incidence rate after adjuvanted vaccination compared with unadjuvanted vaccination, but other side effects were not significantly different, and no serious adverse events occurred. Taking these factors into consideration, adjuvanted influenza vaccination should be considered for use in patients infected with HIV.

Funding Statement

This work was supported by the China Postdoctoral Science Foundation [2017M622587];National Natural Science Foundation of China [81703919];Hunan Provincial Higher Educational Institutions Research Team “Traditional Chinese Medicine prevention and treatment research on infectious diseases” Funding program [15];Hunan Provincial Traditional Chinese Medicine Key Research Project [201701];Hunan Province Teaching and Science “Thirteenth Five-Year Plan” Project [XJK17BGD057];Hunan Provincial Natural Science Foundation [2017JJ3232];

Acknowledgments

We thank for the funding support from the National Natural Science Foundation of China (No:81703919 and No:81774126), China Postdoctoral Science Foundation (No:2017M622587), Hunan Provincial Natural Science Foundation (No:2017JJ3232), Hunan Provincial Traditional Chinese Medicine Key Research Project (No:201701), Hunan Education Department Scientific Research Project (17C1238), Hunan Provincial Higher Educational Institutions Research Team “Traditional Chinese Medicine prevention and treatment research on infectious diseases” Funding program (No. 15), Hunan Province Teaching and Science “Thirteenth Five-Year Plan” Project (No. XJK17BGD057) and the Basic Medicine of Hunan University of Chinese medicine.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.