Factors associated with the immune response to hepatitis A vaccination in HIV-infected patients in the era of highly active antiretroviral therapy

Vaccine 31 (2013) 3668–3674

Contents lists available at SciVerse ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Factors associated with the immune response to hepatitis A vaccination in HIV-infected patients in the era of highly active antiretroviral therapy Guillermo Mena a,b,∗ , Alberto L. García-Basteiro a,b , Anna Llupià a,b , Consolación Díez a,b , Josep Costa c , Josep-María Gatell d , Felipe García d , José-María Bayas a,b a

Preventive Medicine and Epidemiology Unit, Hospital Clínic, Barcelona, Spain Barcelona Centre for International Health Research (CRESIB, Hospital Clínic-Universitat de Barcelona), Barcelona, Spain c Clinical Microbiology and Parasitology Unit, Hospital Clinic, Barcelona, Spain d Infectious Diseases Unit, Hospital Clinic, Barcelona, Spain b

a r t i c l e

i n f o

Article history: Received 4 February 2013 Received in revised form 26 May 2013 Accepted 4 June 2013 Available online 15 June 2013 Keywords: Hepatitis A vaccination HIV Immune response Associated factors Rapidly accelerated schedule

a b s t r a c t Introduction: HIV seropositivity is considered a risk factor for complications in hepatitis A virus (HAV) infection. HAV vaccination schedules are widely implemented in HIV-infected patients, but the immune response remains impaired. Methods: We analysed the response to vaccination (antiHAV titres ≥20 IU/l) in 282 HIV-infected patients included in a standard (1440 Elisa Units (EU) at 0, 6 months) or rapidly accelerated schedule (720 EU at 0, 7, 21 days and 6 months) between 1997 and 2009. Factors associated with the response to vaccination were analysed using logistic regression. Results: The overall response rate was 73.4%. Male sex (OR: 0.16, 95% CI 0.05–0.51) and hepatitis C virus co-infection (OR: 0.30, 95% CI 0.14–0.74) were associated with a lower probability of response. Protective antibody response was associated with a higher CD4/CD8 ratio (OR: 3.69, 95% CI 1.3–10.5) and having received two doses of standard schedule (compared with patients receiving only one dose of the same schedule) (OR: 2.51, 95% CI 1.22–5.15). Three doses of the rapidly accelerated schedule were not more effective than a single dose of 1440 EU (OR: 1.32, 95% CI 0.48–3.63). Conclusion: The low responses observed in patients receiving a single dose suggest the need to emphasize adhesion to vaccination protocols to avoid failure. The CD4/CD8 ratio may be considered as an immune status marker which could help to better choose the moment of vaccination. Our findings underscore the importance of identifying strategies that optimize the timing and effectiveness of hepatitis A vaccination in HIV-infected patients and of the need for further studies on individual factors such as sex and hepatitis C co-infection that may affect the response to vaccination. Likewise, the sub-optimal effectiveness of three doses of 720 EU in the rapidly accelerated schedule, if confirmed in future studies, might lead to a revision of the current schedule recommended for HIV-infected travellers. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction Liver disease is one of the three most-common causes of death in HIV-infected persons, and acute hepatitis A virus (HAV) infection is one cause of liver damage in HIV [1]. Given that drug-induced liver

Abbreviations: AVC, Adult Vaccination Centre; HAART, highly-active antiretroviral therapy; HAV, hepatitis A virus; IDD, Infectious Diseases Department; IDU, injection drug users; MSM, men who have sex with men. ∗ Corresponding author at: Preventive Medicine and Epidemiology Unit, Hospital Clínic, c/Rosselló 138, Baixos Desp. 4, 08036, Barcelona, Spain. Tel.: +34 645968135; fax: +34 934510405. E-mail addresses: [email protected], [email protected] (G. Mena). 0264-410X/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.vaccine.2013.06.012

injury associated with antiretroviral treatment has had significant side effects since the beginning of the highly-active antiretroviral therapy (HAART) era [2] and that HIV-infected subjects may experience prolonged HAV viraemia [3], these patients are considered especially susceptible to severe complications when they become infected by HAV. All persons at risk of infection with HIV types 1 and 2, including men who have sex with men (MSM), persons with multiple heterosexual contacts, injection drug users (IDU), and persons frequently exposed to blood and blood products, are also at high risk of HAV infection, since infection may occur through shared transmission routes [4]. Although the CDC Advisory Committee on Immunization Practices (ACIP) has made no official recommendation regarding HAV

G. Mena et al. / Vaccine 31 (2013) 3668–3674

vaccination of HIV-infected individuals [5], HAV vaccination schedules are widely implemented for HIV-infected subjects, especially for patients with hepatitis C virus (HCV) co-infection or chronic liver disease, MSM and IDU [6–8]. Various formulations are available when HAV vaccination is recommended: single-antigen vaccine in a 2-dose schedule at either 0 and 6–12 months (Havrix® or Epaxal® ), or 0 and 6–18 months (Vaqta® ), or the combined HAV and hepatitis B vaccine (Twinrix® ) in a 3 or 4-dose schedule (0, 1, and 6 months/0, 1, 2 and 6 months) or the alternate schedule, licensed in 2007, administered at 0, 7, and 21–30 days, followed by a dose at 6–12 months [5]. Sub-optimal HAV seroconversion rates after administration of a two or three dose vaccination schedule have been reported in HIV infected subjects [9–13]. Although there are a limited amount of studies focussing on the response to HAV vaccination in HIVinfected subjects and related factors, especially when compared to studies of the response to hepatitis B vaccination (HBV) [14,15], some of the conditions that have repeatedly been studied were associated with higher HAV response rates, including a better immunological status (higher CD4 count and lower HIV-1 viral load) [11–13,16], and compliance with the vaccination schedule [10,17]. The aim of this study was to analyse the proportion of responders among HIV-infected patients receiving the primary HAV standard or rapidly accelerated vaccination schedule and identify the factors associated with seroconversion rates. 2. Patients and methods 2.1. Study characteristics This observational retrospective study analysed HIV-infected patients aged ≥18 years attending the Infectious Diseases Department (IDD) who were referred to the Adult Vaccination Centre (AVC) for HAV vaccination between January 1, 1997 and January 1, 2009 at the Hospital Clinic of Barcelona, Spain. HAV vaccinationnaive HIV-infected subjects who met the criteria for vaccination began the AVC vaccination protocol. Criteria for vaccination were: anti-HAV negative, verbal consent to vaccination and no hypersensitivity to vaccine components. Patients for whom HAV serological results four months after administration of the last dose of vaccine could not be obtained were excluded. 2.2. Hospital Clinic vaccination protocol New cases of HIV infection are detected and followed-up by the IDD. In addition, patients previously diagnosed and followedup in other centres who are attended by the IDD for various reasons are also invited to be followed at the Hospital Clinic. When patients are informed they are infected with HIV, serum determinations include: CD4 count, CD8 count, HIV 1-RNA viral load, anti-HBs, anti-HBc, anti-HCV and anti-HAV IgG titres. One to three months after the serum determinations, the first dose of HAV vaccine is administered by the AVC. If patients require HAV vaccination, a standard schedule is administered, consisting of two intramuscular doses of 1440 Elisa Units (EU) HAV-Ag ELISA (1 ml) (HAVRIX® , non-infectious vaccine manufactured by GlaxoSmithKline, Biologicals, Rixensart, Belgium), 6–12 months apart. When patients also require HBV vaccination, the combined intramuscular HAV + HBV TWINRIX® vaccine (non-infectious vaccine manufactured by GlaxoSmithKline, Biologicals, Rixensart, Belgium) is administered, consisting of four doses containing 720 EU HAVAg ELISA and 20 ␮g HBsAg (1 ml) [5]. Patients with imminent travel plans to regions where HAV vaccination is recommended are administered a rapidly accelerated schedule (720 U HAV-Ag ELISA (TWINRIX® ) 0, 7, 21 days and a booster dose at 6–12 months) [18].

3669

Anti-HAV titres are determined within four months after administration of both doses in standard schedules or the first three doses of rapidly accelerated schedules. There is no determination after the booster dose (6–12 months) in TWINRIX recipients. In this study, we analysed the immunological response to HAV vaccination in patients included in the standard or rapidly accelerated schedules. 2.3. Laboratory tests Anti-HAV titres were determined by Advia Centaur (Siemens) technology. Seroprotection was defined as antibody levels ≥20 IU/l. HIV 1-RNA viral load was calculated using the VERSANT® kPCR molecular diagnostic system. The detection limit of the system is 50 HIV copies 1-RNA/ml since September 2005 and 199 copies prior to this date. The CD4 count was determined by flow cytometry. All samples were analysed by the Microbiology Department, Hospital Clinic of Barcelona. 2.4. Collection of variables Clinical and demographic data collected at the beginning of the vaccination protocol included sex, date of birth, HIV risk group [heterosexual, MSM or IDU], and HCV co-infection. Information on the immunosuppressive status (CD4 count, CD8 count), HIV 1-RNA viral load and antiretroviral treatment was obtained from the last analytical tests performed, as long as this information was available within the last three months before initiation of the vaccination protocol. These data were collected from the computerized medical record and the paper clinical history. Data on HAV vaccination (serological data before and after vaccination, number of doses and date of administration) were obtained from AVC computerized records. All data collected were entered into a Microsoft Access® database. 2.5. Statistical analysis A descriptive analysis of patients at the beginning of the vaccination protocol was made. Comparisons between groups were made using the Chi-square test for categorical variables and the Mann–Whitney U test for continuous variables. Factors associated with response to vaccination were analysed using multivariate logistic regression. Variables with an association (P < 0.1) with the outcome variable (response/non-response) in the univariate model were entered into the multivariate model. The analysis was performed using the SPSS® v. 15 statistical package. 2.6. Ethical considerations The study complied with Spanish Law 15/1999 136 on the Protection of Personal Data. Patient confidentiality and non-disclosure of data collected conformed to the Declaration of Helsinki and was respected at all times. The study was approved by the Research Ethics Committee of the Hospital Clinic of Barcelona, Spain. 3. Results 3.1. Clinical and sociodemographic characteristics Of the 499 patients originally included in either the standard or rapidly accelerated vaccination schedule, serological samples could be collected within four months of administration of the last dose of vaccine in 282 patients, of whom 79.4% were male, with a median age of 36.3 years (IQR: 31.3–40.8). A total of 58.1% were classified as MSM, 27.4% as heterosexual and 14.4% as IDU. Forty-one (15.4%) patients had HCV co-infection. At the start of vaccination, the median CD4 count level was 531.0 cells/␮l (IQR:

3670

G. Mena et al. / Vaccine 31 (2013) 3668–3674

Table 1 Baseline clinical and sociodemographic characteristics of subjects vaccinated against hepatitis A between 1997 and 2009 according to whether the response was analysed or not in the final analysis. Characteristic

Subjects in whom the response was analysed

Subjects in whom the response was not analysed

Total (n)

Median (IQR)

Total (n)

Median (IQR)

Patientsb Age, yearsa

282 282

36.3 (31.3–40.8)

217 216

32.3 (28.3–36.2)

Sexb Male Female

282

HCV co-infectionb No Yes

266

217 224 (79.4) 58 (20.6)

269

b

<0.001 0.050

156 (71.9) 61 (28.1) 200

225 (84.6) 41(15.4)

CD4 count, cells/␮la

P-value

0.068 156 (78.0) 44 (22.0)

531.0 (405.0–765.5)

215

578.0 (385.0–798.0)

215

0.616

CD4 count (range) <350 ≥350

269

CD8 count, cells/␮la

240

966.5 (702.8–1210.5)

194

882.5 (605.3–1225.8)

0.194

CD4/CD8 ratio

236

0.6 (0.4–0.8)

181

0.7 (0.4–0.9)

0.308

HIV RNA load log 10, copies/ml

269

2.3 (1.7–4.1)

206

1.7 (1.7–2.3)

<0.001

46 (17.1) 223 (82.9)

a

b

HIV RNA load, copies/ml Undetectable Detectable

269

HAARTb No Yes

274

Scheduleb Incomplete standard Completed standard 3 dose of rapidly accelerated

282

Year of first dose

282

0.284 45(20.9) 170 (79.1)

216 156 (58.0) 113 (42.0)

<0.001 170 (78.7) 46 (21.3)

217 98 (35.8) 176 (64.2)

<0.001 23 (10.6) 194 (89.4)

217 80 (28.4) 161 (57.1) 41 (14.5) 2005 (2003–2006)

<0.001 98 (45.2) 106 (48.8) 13 (6.0)

217

2004 (2002–2006)

0.63

Data collected before administration of the first dose of vaccine. n: for each category, the number of patients is that for which information was available. List of abbreviations: HCV, hepatitis C virus; HAART, highly active antiretroviral therapy; standard schedule (0, 24–48 weeks); and rapidly accelerated schedule (0, 7, 21 days and 24–28 weeks). a Continuous variables are presented as median and interquartile range (IQR). Comparisons between two groups were made using the Mann–Whitney U test. b Categorical variables are presented as proportions. Comparisons between groups were made using the Chi-square test.

405.0–765.5) and the median CD8 count level was 966.5 cells/␮l (IQR: 702.8–1210.5), while 58.0% of patients had an undetectable HIV 1-RNA viral load. One hundred and seventy six patients (64.2%) were receiving antiretroviral therapy (HAART). The median time on HAART was 10.8 months (IQR: 4.7–26.4). Table 1 shows that the 282 subjects included in the analysis of response to the vaccine were older compared with patients not included [36.3 years (IQR: 31.3–40.8) vs. 32.3 years (IQR: 28.3–36.2)]. In addition, a lower proportion of patients included in the analysis had an undetectable viral load (58.0% vs. 78.7%) and were receiving antiretroviral treatment (64.2% vs. 89.2%).

3.2. Response to vaccine and associated factors An overall response rate of 73.4% (207/282) was found in patients included in the analysis. The response was 60.0% (48/80) in patients who received only the first dose of the standard schedule and 80.7% (130/161) in those completing the schedule. The proportion of patients who responded positively to the first three doses of the rapidly accelerated schedule was 70.7% (29/41). Patient characteristics according to response to vaccination are shown in Table 2. The 41 patients included in the rapidly accelerated schedule had a higher median age, a higher proportion of subjects with undetectable viral load and a higher percentage of individuals in HAART than the 241 patients receiving one or two doses of the standard schedule (Table 3). In the univariate analysis (Table 4), six factors were associated with the response after HAV vaccination: sex, risk group, HCV

co-infection, CD4 count (cut-off at 350 cells/␮l), CD4/CD8 ratio, and schedule. Male sex (OR: 0.26, 95% CI 0.11–0.63) and HCV coinfection (OR: 0.46, 95% CI 0.23–0.91) were associated with a lower probability of response. The categorical variable “risk group” was related to the response after vaccination (P = 0.022): MSM were not less likely to respond than heterosexual patients (OR: 0.66, 95% CI 0.33–1.31), but IDU had a worse outcome (OR: 0.30, 95% CI 0.13–0.71). The OR for patients with a CD4 count ≥350 cells/␮l was 2.31 times that of the OR for patients with CD4 < 350 cells/mm3 (95% CI 1.19–4.50). A higher OR was found for patients with a higher CD4/CD8 count (95% CI 4.07–10.65). In the univariate analysis, the categorical variable, vaccine schedule which included HIV patients, was associated with the response to vaccination (P = 0.003). Having completed the standard schedule was related to a higher probability of a response when compared with having received one of the two doses (OR: 2.78, 95% CI 1.54–5.07). The multivariate analysis (Table 4) showed that hepatitis C virus co-infection (OR: 0.30, 95% CI 0.14–0.74) and male sex (OR: 0.16, 95% CI 0.05–0.51) were associated with a lower probability of response. A protective antibody response was associated with a higher CD4/CD8 ratio (OR: 3.69, 95% CI 1.3–10.5) and completing the standard schedule (in comparison with patients receiving only dose) (OR: 2.51, 95% CI 1.22–5.15).

4. Discussion This study found an overall rate of response to HAV vaccination in HIV-infected patients of 73.4%. Having received a complete

G. Mena et al. / Vaccine 31 (2013) 3668–3674

3671

Table 2 Baseline characteristics of patients by response to hepatitis A vaccination. Characteristic Patientsb Age, yearsa

N

Overall

Non-responders

Responders

282

282 36.3 (31.3–40.8)

75 36.2 (30.6–40.7)

207 36.3 (32.0–41.1)

224 (79.4) 58 (20.6)

69 (92.0) 6 (8.0)

155(74.9) 52 (25.1)

74 (27.4) 157 (58.1) 39 (14.4)

14 (19.4) 41 (56.9) 17 (23.6)

60 (30.3) 116 (58.6) 22 (11.1)

225 (84.6) 41(15.4)

55 (76.4) 17 (23.6)

170 (87.6) 24 (12.4)

531.0 (405.0–765.5)

483.0 (332.0–632.0)

550.5 (424.3–772.5)

46 (17.1) 223 (82.9)

19 (26.8) 52 (73.2)

27(13.6) 171 (86.4)

Sexb Male Female

282

Risk groupb Heterosexual MSM IDU

270

HCV co-infectionb No Yes

266

CD4 count, cells/␮la

269

CD4 count (range)b <350 ≥350

269

CD8 cell count, cells/␮la

240

966.5 (702.8–1210.5)

984.0 (746.5–1283.5)

952.0 (680.0–1175.0)

CD4/CD8 ratioa

236

0.6 (0.4–0.8)

0.5 (0.3–0.8)

0.6 (0.4–0.9)

HIV RNA load log 10, copies/ml

269

2.3 (1.7–4.1)

2.3 (1.7–4.2)

2.3(1.7–4.0)

156 (58.0) 113 (42.0)

41 (58.6) 29 (41.4)

115 (57.8) 84 (42.2)

98 (35.8) 176 (64.2)

22 (30.1) 51 (69.9)

76 (37.8) 125 (62.2)

10.2(4.7–26.4)

9.2(4.4–24.5)

11.7(4.8–30.4)

80 (28.4) 161 (57.1) 41 (14.5)

32 (42.7) 31 (41.3) 12 (16.0)

48 (23.2) 130 (62.8) 29 (14.0)

2005 (2003–2006)

2004 (2002–2006)

2005 (2003–2006)

b

HIV RNA load, copies/ml Undetectable Detectable

269

HAARTb No Yes

274

Time on HAART, monthsa

176

Scheduleb Incomplete standard Completed standard 3 dose of rapidly accelerated

282

Year of first dose

282

Data collected before administration of the first dose of vaccine. n: for each category, the number of patients is that for which information was available. List of abbreviations: MSM, men who have sex with men; IDU, injecting drug users; HCV, hepatitis C virus; HAART, highly active antiretroviral therapy; standard schedule (0, 24–48 weeks); and rapidly accelerated schedule (0, 7, 21 days and 24–28 weeks). a Continuous variables are presented as median and interquartile range (IQR). Comparisons between two groups were made using the Mann–Whitney U test. b Categorical variables are presented as proportions. Comparisons between groups were made using the Chi-square test.

standard schedule (two doses of 1440 EU HAV Ag) compared with receiving only one dose of 1440 EU, and Lower immunosuppression, and having received a complete standard schedule (two doses of 1440 EU HAV Ag) compared with receiving only one dose of 1440 EU, were associated with a response to vaccination, while male sex and HCV co-infection were more likely to be associated with nonresponse to HAV vaccination. Higher response rates for completed 1440 U HAV-Ag standard schedules in HIV-infected patients have been reported in two randomized clinical trials conducted in patients with a median CD4

count similar to that observed in our study. These trials found a response rate of 77.9% [17] and 93.9% [19], close to that obtained in HIV-free adults [20]. It is difficult to compare results between studies with different methodologies and patient characteristics [21]. In our study, patients completing the standard schedule had a response rate of 80.7%, higher than the rates found by most retrospective analyses in the last decade, which ranged between 48.5% and 89% (Table 5). Receiving a single dose of 1440 EU HAV Ag vaccine was associated with a deficient response compared to receiving both doses

Table 3 Baseline characteristics of patients by vaccination schedule. Characteristic

Standard

Rapidly accelerated

P-value

Patients, n Age, yearsa Sex, % maleb HCV co-infection, nb CD4 count, cells/␮la HIV RNA, % undetectableb Patients on HAART, %b Year of first dosea

241 35.8 (30.6–40.6) 193 (80.1) 33 (14.5) 531.0 (402.0–765.5) 126 (55.3) 143 (61.4) 2004 (2002–2006)

41 38.2 (33.8–43.4) 31 (75.6) 8 (21.1) 543.0 (411.2–787.8) 30 (73.2) 33 (80.5) 2006 (2006–2006)

– 0.030 0.510 0.300 0.620 0.030 0.020 <0.001

List of abbreviations: HCV, hepatitis C virus; HAART, highly active antiretroviral therapy; standard schedule (0, 24–48 weeks); and rapidly accelerated schedule (days 0, 7, 21 and 24–48 weeks). a Continuous variables are presented as median and interquartile range (IQR). Comparisons between two groups were made using the Mann–Whitney U test. b Categorical variables are presented as proportions. Comparisons between groups were made using the Chi-square test.

3672

G. Mena et al. / Vaccine 31 (2013) 3668–3674

Table 4 Univariate and multivariate analysis of factors associated with response to vaccination (positive anti-HAV). Characteristic

Univariate OR (95% CI)

P

Age, years Sex Female Male

1.00 (0.97–1.04)

0.636

Reference 0.26 (0.11–0.63)

0.003

Risk group Heterosexual MSM IDU

Reference 0.66 (0.33–1.31) 0.30 (0.13–0.71)

0.233 0.006

HCV co-infection No Yes

Reference 0.46 (0.23–0.91)

0.026

CD4 cell count, cells/␮la

1.00 (1.00–1.01)

0.164

CD4 cell count (range) <350 ≥350

Reference 2.31 (1.19–4.50)

0.013

CD8 cell count

1.00 (0.99–1.00)

0.222

Multivariate OR (95% CI)

P

Reference 0.16 (0.05–0.51)

<0.001

Reference 0.30 (0.12–0.74)

0.009

3.69 (1.30–10.50)

0.014

0.022

CD4/CD8 ratio

4.07 (1.55–10.65)

0.004

HIV RNA load log 10, copies/ml

1.01 (0.83–1.23)

0.947

HIV RNA Undetectable Detectable

Reference 1.03 (0.59–1.79)

0.909

HAART No Yes

Reference 0.71 (0.40–1.26)

0.242

Time on HAART, months

1.00 (0.99–1.03)

0.539

Schedule Incomplete standard Completed standard 3 dose of rapidly accelerated

Reference 2.78 (1.54–5.07) 1.61 (0.72–3.61)

0.001 0.247

Year of first dose

1.12 (1.00–1.25)

0.048

0.003

0.038 Reference 2.51 (1.22–5.15) 1.32 (0.48–3.63)

0.012 0.587

Data collected before administration of the first dose of vaccine. Univariate and multivariate logistic regression analysis. Variables considered of interest and variables with a value of P < 0.1 in the univariate model were entered into the model using the stepwise method. List of abbreviations: OR, odds ratio; CI, confidence interval; MSM, men who have sex with men; IDU, injecting drug users; HCV, hepatitis C virus; HAART, highly active antiretroviral therapy; standard schedule (0, 24–48 weeks); and rapidly accelerated schedule (days 0, 7, 21 and 24–48 weeks). a Continuous variables.

of the standard schedule. Similar data have been reported in two randomized double-blind clinical trials by Wallace et al. [19] and Launay et al. [10]. This suggests that adhesion to the protocol is essential to avoid non-responsiveness. The rapidly accelerated HAV vaccination schedule (720 EU at 0, 7, 21 days and 6–12 months), used in people unable to comply with regular schedules, has been shown to be effective in immunocompetent patients [22]. The response to a rapidly accelerated schedule has not been previously evaluated in patients with HIV infection. In our study, post-vaccination serological data were available for 41 patients who received the three first doses of this schedule, and showed that 70.7% responded. However, in the multivariate analysis, there was no significant difference with the response rate of 60% achieved by patients receiving only one dose of 1440 EU. As reported by studies in HIV-free subjects, responders were more likely to be female [23,24]. Three studies have reported an association between sex and response to HAV vaccination, but with contrasting results: Weissman et al. [9] found that female sex was associated with a good response, but Overton et al. [16] and the small study by Armstrong et al. [25] both found that male sex was significantly associated with a response to vaccination. In our study, the better response in females was obtained after adjustment of the CD4/CD8 ratio: however, we did not include the variable HIV-1 RNA viral load (detectable/undetectable) in the multivariable analysis as it was not significantly associated with the outcome variable in the univariate analysis. Nevertheless, a higher proportion of women

had an undetectable viral load, which has been associated with the response to the vaccine [16,26]. Therefore, this aspect should be further studied. To our knowledge, this is the first study to report a significant association between hepatitis C virus co-infection and a lack of immunological response to HAV vaccination in HIV-infected subjects. It is suggested that HCV co-infection results in dendritic cell dysfunction [27–29] that may impair antigen presentation and might affect the response to vaccination, as already shown in the case of hepatitis B vaccination in HIV-infected patients [30]. In addition, lower geometric mean levels of anti-HAV after vaccination of patients with chronic liver disease, including HCV-infected patients, has been reported [31]. Larger studies are required to confirm this association. As expected, a weakened immune status was associated with a poorer response to HAV vaccination. Kourkounti et al. [11], Rimland and Guest [12] and Jimenez et al. [13] found an association between higher CD4 levels and a more-successful response to vaccination. The study by Jimenez et al. [13] and other investigations found that a minimum or non-existent viral load could act as a predictor for response to HAV vaccination [16,23]. In our study, CD4 and CD8 counts, taken separately, were not significantly associated with the response to vaccination, but the CD4/CD8 ratio, which is usually maintained over 1.2 in immunocompetent patients, was associated with the response. This finding supports the use of the CD4/CD8 ratio as an immune status marker which can aid the best choice

Table 5 Studies assessing the immunological response after HAV vaccination in the HAART era (n = 10).a Author/year

Study design

Vaccine schedule

Baseline immunological status of participants

Response rate (%)b

Main findings

Neilsen et al. [17] 1997

RCT

HAVRIX1440 EU, Two doses, either 1 or 6 months apart

Kemper et al. [32] 2003 Wallace et al. [19] 2004

RDBCT

HAVRIX 1440 EU, Two doses, 6 months apart VAQTA (50 U), Two doses, 6 months apart

CD4 count mean = 515 cells/mm3 at baseline CD4 count mean = 376 cells/mm3 CD4 count mean = 458 cells/mm3 76% on HAART

Vaccine schedule did not affect response (0.1 vs. 0.6 months) – GMT Anti-HAV greater in subjects with a CD4 count ≥200 cells/mm3 Higher response in subjects with a CD4 count ≥200 cells/mm3

Weissman et al. [9] 2004 Rimland and Guest [12] 2005 Overton et al. [16] 2006

Retrospective Observational Retrospective Observational Retrospective Observational

Launay et al. [10] 2008

After the 1st dose: 77.9% (67/86) After the 2nd dose (either 1 or 6 months apart): 88.1% (67/76) After the 1st dose: 11.1% (5/45) After the 2nd dose: 52.1% (25/48) One month after the 1st dose: 61.0% (33/54) One month after the 2nd dose: 93.9% (46/49) After the 2nd dose: 48.5% (67/138) After the 2nd dose: 60.7% (130/214) After the 1st or the 2nd dose: 49.6% (133/238) After the 2nd dose: 52.2% (94/180) A: After the 1st dose: 37.9% After the 2nd dose: 69.4%(34/49) B: After the 3rd dose: 82.6% (38/46) After the 2nd dose: 89% (116/130)

RDBCT

85% on HAART

HAVRIX 1440 EU, Two doses

CD4 count mean = 447 cells/mm3 68% on HAART

RCT

HAVRIX 1440 EU, A. Two doses (6 months apart) vs. B. Three doses (0.1 and 6 months)

CD4 count median = 355 cells/mm3 80% on HAART

Crum-Cianflone et al. [26] 2011

Retrospective Observational

Kourkounti et al. [11] 2012

Retrospective Observational

CD4 count median = 461 cells/mm3 62% on HAART CD4 count median = 564 cells/mm3

Jimenez et al. [13] 2013

Retrospective Observational

Mena et al., 2013

Retrospective Observational

VAQTA (50 U) or HAVRIX 1440 EU, Two doses (6–18 months apart) HAVRIX 1440 EU or VAQTA (50 U), Two doses (6 months apart) At least one dose of A.TWINRIX (0.1 and 6 months) or B. HAVRIX (6–12 months apart) HAVRIX 1440 EU, A.One dose, B. Two doses (6 months apart) or C.TWINRIX (720 EU) 0, 7, 21 days

No data

CD4 count median = 410 cells/mm3 70.4% on HAART CD4 count median = 631 461 cells/mm3 64.2% on HAART

After the 2nd dose: 74.4% (260/351) A. At least one dose of TWINRIX: 53.5% (54/101). B. At least one dose of HAVRIX: 53.6%(67/125) A. After 1 dose of HAVRIX: 60.0% (48/80) B. After 2 doses of HAVRIX: 80.7% (130/161) C. After 3 doses of rapidly accelerated schedule of TWINRIX: 70.7% (29/41)

Responders more likely to be female and have a higher CD4 count Responders more likely to have a higher CD4 count, especially if exceeds 200 cells/mm3 Responders more likely to be men and have a baseline HIV viral RNA load <1000 copies/ml

The 3-dose group induced a significantly higher antibody titre – absence of tobacco smoking was a predictor of response

Younger age associated with a higher initial GMC, with a trend for lower log 10 plasma HIV RNA levels with better response Responders more likely to have a higher CD4 count, A higher response rate and higher GMTs were observed in patients with CD4 counts ≥ 500 cells/mm3 Responders more likely to have higher baseline median CD4 count and lower median viral RNA load. Worse response in TWINRIX recipients when virologically not suppressed and when vaccination series not completed. Response associated with female sex, absence of HCV co-infection, a higher CD4/CD8 ratio and a standard schedule (in comparison with those receiving only one of those of the same scheme).

G. Mena et al. / Vaccine 31 (2013) 3668–3674

HAVRIX 1440 EU, Two doses, 6–12 months apart HAVRIX 1440 EU, Two doses

Higher response in subjects with a CD4 count ≥300 cells/mm3

RCT, randomized controlled trial; RDBCT, randomized double-blind controlled trial; GMT, geometric mean titre; GMC, geometric mean concentration; EU, ELISA units; anti-HAV, antibody against hepatitis A virus. a Inclusion criteria: N ≥ 40, immune response reported between 1 month and 1 year after last dose of vaccine. b Protective titre: anti-HAV ≥10 IU/l: Wallace et al., Crum-Cianflone et al.; ≥18 IU/l: Weissman et al.; ≥20 IU/l: Neilsen et al., Rimland and Guest, Launay et al., Mena et al.; ≥33 IU/l: Kemper et al.; not reported. Overton et al.

3673

3674

G. Mena et al. / Vaccine 31 (2013) 3668–3674

of the time to vaccinate, i.e. before the CD4 count starts falling or delaying vaccination until the ratio is higher after the initiation of HAART. The strengths of this study include the comprehensive clinical data, collected through the well-established and coordinated healthcare system of three hospital departments during twelve years. However, the study had some limitations. Some subjects simultaneously received the hepatitis B vaccine, as well as other vaccines, thus complicating the analysis of vaccine coadministration as a factor in the response. This factor has not yet been studied. Secondly, only AVC patients with post-vaccination serology were included. Our data show that, although patients excluded from the HAV response analysis had worse compliance, with a high percentage of patients receiving only one dose, the response to vaccination would probably be higher in the whole population of patients because they were younger and had lower immunosuppression, with a higher proportion of subjects having an undetectable viral load and being on HAART. Thirdly, as we performed a naturalistic study that analysed responses obtained from patients attending our vaccination unit, we could not obtain further information on serologies. In the case of TWINRIX recipients, these were obtained only after the third dose, and there was a lack of information on the response after the booster dose at 6–12 months. Therefore, caution is required in drawing general conclusions. The low responses observed in HIV-infected patients receiving less than two standard doses suggest the need to emphasize adherence to vaccination schedules to avoid failure. Likewise, three doses of 720 EU in the rapidly accelerated schedule were not shown to be more effective than one dose of 1440 EU. If confirmed in future studies, these results might lead to a revision of the current schedule recommended for HIV-infected travellers. The CD4/CD8 ratio may be considered as an immune status marker which could aid a better choice of the time of vaccination: either before the CD4 count starts falling or by delaying vaccination until the ratio is higher after the initiation of HAART. Our findings underscore the importance of identifying strategies that optimize the timing and effectiveness of hepatitis A vaccination in HIV-infected patients and of the need for further studies on individual factors such as sex and hepatitis C coinfection that may affect the response to vaccination. Conflicts of interest: JMB has participated in vaccine trials by GlaxoSmithKline and Sanofi Pasteur MSD. The remaining authors declare no conflicts of interest. Financial support: This study was supported by the Hospital Clínic of Barcelona. Acknowledgments We thank David Buss for technical assistance. References [1] Puoti M, Moioli MC, Travi G, Rossotti R. The burden of liver disease in human immunodeficiency virus-infected patients. Semin Liver Dis 2012;32:103–13. ˜ [2] Jones M, Núnez M. Liver toxicity of antiretroviral drugs. Semin Liver Dis 2012;32:167–76. [3] Ida S, Tachikawa N, Nakajima A, Daikoku M, Yano M, Kikuchi Y, et al. Influence of human immunodeficiency virus type 1 infection on acute hepatitis A virus infection. Clin Infect Dis 2002;34:379–85. [4] Laurence JC. Hepatitis A and B immunizations of individuals infected with human immunodeficiency virus. Am J Med 2005;118(Suppl. 10A):75S–83S. [5] Centers for Disease Control and Prevention. Morbidity and Mortality Weekly Report (MMWR). Recommended Adult Immunization Schedule—United States, 2012. February 3, 2012/61(04); 1–7. Available at: http://www.cdc.gov/mmwr/ preview/mmwrhtml/mm6104a9.htm

[6] Rockstroh JK, Bhagani S, Benhamou Y, Bruno R, Mauss S, Peters L, et al. European AIDS Clinical Society (EACS) guidelines for the clinical management and treatment of chronic hepatitis B and C coinfection in HIV-infected adults. HIV Med 2008;9:82–8. [7] Kahn J. Preventing hepatitis A and hepatitis virus infections among men who have sex with men. Clin Infect Dis 2002;35:1382–7. [8] Villano SA, Nelson KE, Vlahov D, Purcell RH, Saah AJ, Thomas DL. Hepatitis A among homosexual men and injection drug users: more evidence for vaccination. Clin Infect Dis 1997;25:726–8. [9] Weissman S, Feucht C, Moore BA. Response to hepatitis A vaccine in HIVpositive patients. J Viral Hepat 2006;13:81–6. [10] Launay O, Grabar S, Gordien E, Desaint C, Jegou D, Abad S, et al. Immunological efficacy of a three-dose schedule of hepatitis A vaccine in HIV-infected adults: HEPAVAC study. J Acquir Immune Defic Syndr 2008;49:272–5. [11] Kourkounti S, Mavrianou N, Paparizos VA, Kyriakis K, Hatzivassiliou M, Kordosis T, et al. Immune response to hepatitis A vaccination in HIV-infected men in Greece. Int J STD AIDS 2012;23:464–7. [12] Rimland D, Guest JL. Response to hepatitis A vaccine in HIV patients in the HAART era. AIDS 2005;19:1702–4. [13] Jimenez HR, Hallit RR, DeBari VA, Slim J. Vaccine 2013;31:1328–33. [14] van den Berg R, van Hoogstraten I, van Agtmael M. Non-responsiveness to hepatitis B vaccination in HIV seropositive patients; possible causes and solutions. AIDS Rev 2009;11:157–64. [15] Kim HN, Harrington RD, Crane HM, Dhanireddy S, Dellit TH, Hepatitis Spach DH. B vaccination in HIV-infected adults: current evidence, recommendations and practical considerations. Int J STD AIDS 2009;20:595–600. [16] Overton ET, Nurutdinova D, Sungkanuparph S, Seyfried W, Groger RK, Powderly WG. Predictors of immunity after hepatitis A vaccination in HIV-infected persons. J Viral Hepat 2007;14:189–93. [17] Neilsen GA, Bodsworth NJ, Watts N. Response to hepatitis A vaccination in human immunodeficiency virus-infected and -uninfected homosexual men. J Infect Dis 1997;176:1064–7. [18] Centers for Disease Control and Prevention. Morbidity and Mortality Weekly Report (MMWR). October 12, 2007, vol 56, 40. FDA Approval of an Alternate Dosing Schedule for a Combined Hepatitis A and B Vaccine (Twinrix® ). Available at: http://www.cdc.gov/mmwr/PDF/wk/mm5640.pdf [19] Wallace MR, Brandt CJ, Earhart KC, Kuter BJ, Grosso AD, Lakkis H, et al. Safety and immunogenicity of an inactivated hepatitis A vaccine among HIV-infected subjects. Clin Infect Dis 2004;39:1207–13. [20] Clemens R, Safary A, Hepbrun A, Roche C, Stanbury WJ, Andre FE. Clinical experience with an inactivated hepatitis A vaccine. J Infect Dis 1995;171(Suppl. 1):S44–8. [21] Shire NJ, Welge JA, Sherman KE. Efficacy of inactivated hepatitis A vaccine in HIV-infected patients: a hierarchical Bayesian meta-analysis. Vaccine 2006;24:272–9. [22] Kallinowski B, Jilg W, Buchholz L, Stremmel W, Engler S. Immunogenicity of an accelerated vaccination regime with a combined hepatitis a/b vaccine in patients with chronic hepatitis C. Z Gastroenterol 2003;41:983–90. [23] Chlíbek R, Smetana J, Sindelár R, Cecetková B, Prymula R, Kohl I. [Immunogenicity of vaccines against viral hepatitis A and B in the population above 40 years of age – impact of risk factors]. Epidemiol Mikrobiol Imunol 2007;56: 119–28. [24] Van der Wielen M, Van Damme P, Chlibek R, Smetana J, von Sonnenburg F. Hepatitis A/B vaccination of adults over 40 years old: comparison of three vaccine regimens and effect of influencing factors. Vaccine 2006;29:5509–15. [25] Armstrong KE, Bush HM, Collins JD, Feola DJ, Caldwell GC, Thornton AC. Role of CD4 count in immunity development after hepatitis A and B vaccination among HIV-infected patients: Kentucky, 2002–2007. J Int Assoc Physicians AIDS Care (Chic) 2010;9:179–86. [26] Crum-Cianflone NF, Wilkins K, Lee AW, Grosso A, Landrum ML, Weintrob A, et al. Long-term durability of immune responses after hepatitis A vaccination among HIV-infected adults. J Infect Dis 2011;203:1815–23. [27] Anthony DD, Yonkers NL, Post AB, Asaad R, Heinzel FP, Lederman MM, et al. Selective impairments in dendritic cell-associated function distinguish hepatitis C virus and HIV infection. J Immunol 2004;172:4907–16. [28] Sarobe P, Lasarte JJ, Casares N, López-Díaz de Cerio A, Baixeras E, Labarga P, et al. Abnormal priming of CD4(+) T cells by dendritic cells expressing hepatitis C virus core and E1 proteins. J Virol 2002;76:5062–70. [29] Auffermann-Gretzinger S, Keeffe EB, Levy S. Impaired dendritic cell maturation in patients with chronic, but not resolved, hepatitis C virus infection. Blood 2001;97:3171–6. [30] Gandhi RT, Wurcel A, Lee H, McGovern B, Shopis J, Geary M, et al. Response to hepatitis B vaccine in HIV-1-positive subjects who test positive for isolated antibody to hepatitis B core antigen: implications for hepatitis B vaccine strategies. J Infect Dis 2005;191:1435–41. [31] Keeffe EB, Iwarson S, McMahon BJ, Lindsay KL, Koff RS, Manns M, et al. Safety and immunogenicity of hepatitis A vaccine in patients with chronic liver disease. Hepatology 1998;27:881–6. [32] Kemper CA, Haubrich R, Frank I, Dubin G, Buscarino C, McCutchan JA, et al. Safety and immunogenicity of hepatitis A vaccine in human immunodeficiency virus-infected patients: a double-blind, randomized, placebo-controlled trial. J Infect Dis 2003;187:1327–33.