SYSTEMATIC REVIEW

Immune Response of Hepatitis B Vaccine Among Persons With Diabetes A systematic review of the literature

1 SARAH F. SCHILLIE, MD among persons with diabetes show some 2 PHILIP R. SPRADLING, MD 1 heterogeneity compared with adults without TRUDY V. MURPHY, MD diabetes. Differences in diabetes type, man- agement, and glycemic control, as well as vaccine, dosage, administration route, or schedule, may underlie this heterogeneity. n October 2011, the Advisory Com- and other comorbid conditions (6–12). The 2011 ACIP recommendation for hepa- Imittee on Immunization Practices Results of studies of the hepatitis B vac- titis B vaccination for adults with diabetes, an (ACIP) recommended hepatitis B vac- cine among persons with diabetes gener- increasing incidence of diabetes, and the cination for adults with diabetes in the ally follow these patterns (13,14). high prevalence of diabetes among certain United States (1). Serosurvey data from Primary hepatitis B vaccination usu- groups recommended for hepatitis B vacci- the 1999–2010 National Health and Nu- ally consists of 3 (or 4) doses of 10 or nation (e.g., persons with end-stage renal trition Examination Survey found that 20 mg of recombinant hepatitis B surface disease) suggests that a review of vaccine ef- noninstitutionalized adults aged $18 antigen (HBsAg) protein administered in- ficacy among persons with diabetes may be years with diabetes have an increased se- tramuscularly into the deltoid muscle timely. We therefore performed a systematic roprevalence of past or current hepatitis B on a 0-, 1-, and 6-month schedule review of the literatureandsummarizedthe virus (HBV) infection (1). Outbreaks of (Table 1). Alternative schedules are U.S.- evidence for seroprotection after hepatitis B hepatitis B in elderly persons with diabe- approved for routine vaccination for spe- vaccination among persons with diabetes. tes in long-term care facilities have been cific ages and vaccine formulations, and fi fi linked to diabetes care procedures (in- they elicit dose-speci cand nal rates of RESEARCH DESIGN AND cluding blood glucose monitoring), with seroprotection similar to those obtained METHODS likely vehicles for transmission including on a 0-, 1-, and 6-month schedule (4). fi spring-loaded nger-stick devices used Two single-antigen recombinant vac- Search strategy on multiple patients and blood glucose cines, Recombivax HB (Merck & Co, Electronic searches of MEDLINE (via testing meters that were not cleaned be- Inc., Whitehouse Station, NJ) and PubMed), EMBASE (via Ovid), Cochrane tween uses on different patients. Engerix-B (GlaxoSmithKline Biologicals, Library, and Web of Knowledge databases In the U.S., hepatitis B vaccination is Rixensart, Belgium), and one combi- were performed. The search terms con- routinely recommended for infants, chil- nation hepatitis A/hepatitis B vaccine, sisted of (hepatitis b vaccin* OR hbv vac- dren, and adolescents. It also is recom- Twinrix (GlaxoSmithKline Biologicals), cin* OR hepatitis b immuni* OR hbv mended for adults at increased risk of are approved for use in adults in the U.S. immuni*) AND (immunogeni* OR im- HBV infection, including persons with (15). Hepatitis B vaccines are safe for all mune response OR antibody) AND (dia- end-stage renal disease or chronic liver age groups. Administration of additional betes). The search terms also were used as disease, health care personnel, injection- vaccine doses for nonresponders is not Medical Subject Heading terms (MEDLINE) drug users, and men who have sex with associated with an increase in adverse – or key words (EMBASE), as applicable. men (2 4). A hepatitis B vaccination se- events (16). Vaccination is not contrain- Where possible, limits included publica- ries results in protection in a high propor- dicated in persons with autoimmune or . tion date from 1986 through 30 April tion ( 95%) of infants, children, and chronic diseases, or in those who are 2012, English language, and study of hu- young adults (5). As with other vaccines, pregnant (4). mans. The MEDLINE and EMBASE the efficacy of the hepatitis B vaccine pro- No review has been published of the fi searches were limited to items containing gressively declines with advancing age, ef cacy of the hepatitis B vaccine among abstracts. Studies conducted in the U.S. and as well as with the presence of obesity persons with diabetes; results of studies other countries were eligible for inclusion. ccccccccccccccccccccccccccccccccccccccccccccccccc Inclusion criteria From the 1Division of Viral Hepatitis, Vaccine Research and Policy Team; National Center for HIV/AIDS, Viral Peer-reviewed, published randomized Hepatitis, STD, and TB Prevention; Centers for Disease Control and Prevention, Atlanta, Georgia; and the 2 clinical trials or observational studies (all Division of Viral Hepatitis, Epidemiology and Surveillance Branch; National Center for HIV/AIDS, Viral types) assessing response to hepatitis B Hepatitis, STD, and TB Prevention; Centers for Disease Control and Prevention, Atlanta, Georgia. Corresponding author: Sarah Schillie, [email protected]. vaccine among persons with type 1 or Received 13 February 2012 and accepted 6 June 2012. type 2 diabetes were included. Studies DOI: 10.2337/dc12-0312 among both children and adults were in- The findings and conclusions in this report are those of the author(s) and do not necessarily represent the cluded to ascertain the effect of age upon official position of the Centers for Disease Control and Prevention. © 2012 by the American Diabetes Association. Readers may use this article as long as the work is properly vaccine response among persons with cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/ diabetes. Reports had to specify the pro- licenses/by-nc-nd/3.0/ for details. portion (numerator and denominator) of

2690 DIABETES CARE, VOLUME 35, DECEMBER 2012 care.diabetesjournals.org Schillie, Spradling, and Murphy

Table 1dRecommended dosages of hepatitis B vaccine for adults aged >20 years* (4) 85 duplicates were identified. Abstracts from the remaining 140 studies (62.2%) Single-antigen vaccine Combination vaccine were reviewed, and 94 (67.1%) were deemed not relevant. Forty-six (32.9%) Recombivax HB Engerix-B Twinrix full-text studies were retrieved and re- Dosage Volume Dosage Volume Dosage Volume viewed, and one additional duplicate was (mg) (mL) (mg) (mL) (mg) (mL) identified (Fig. 1). Fifteen studies did not fulfill inclusion criteria, and 13 met at Adults 10 1 20 1 20 1 least one criterion for exclusion (Table 2 Adult hemodialysis patients and other and Fig. 1), leaving 17 studies. † ‡ immunocompromised adults 40 140 2NANA The 17 studies included 2 random- NA, not applicable. *Please refer to package insert. †Dialysis formulation administered on a 3-dose schedule ized clinical trials, 12 prospective, and 3 at 0, 1, and 6 months. ‡Two 1-mL doses administered in 1 or 2 injections on a 4-dose schedule at 0, 1, 2, and 6 retrospective studies published between months. 1989 and 2012 (Table 3). The 17 studies included 16,310 unique subjects, of which approximately 9,286 had diabetes. subjects seroprotected by diabetes status publication year, country); subject charac- The subject categories consisted of adults (or data available that allowed for calcula- teristics (e.g., age, diabetes type); vaccine for 4 studies (285 subjects, of whom 152 tion of the proportions), or available odds characteristics (e.g., vaccine, dosage, had diabetes); children for 4 studies (472 ratios from a multivariate analysis, which route, schedule); and the proportion of subjects, of whom 206 had diabetes); and included diabetes as a predictor variable. subjects attaining seroprotection, includ- hemodialysis/chronic kidney disease pa- Studies of specific subject populations of ing the time interval (in months) from the tients for 9 studies (15,553 subjects, of which persons with diabetes represented a last dose at which immune response was whom approximately 8,928 had diabe- subset were included. measured. When vaccine or route of ad- tes). Subjects in the studies among adults ministration differed by study arm, extrac- and children generally lacked comorbid- Exclusion criteria tion elements were recorded for each arm ities (including kidney disease) other than Studies were excluded when immune when possible. diabetes. One study among hemodialysis/ response using an antibody to hepatitis An anti-HBs threshold of 10 mIU/mL chronic kidney disease patients included B surface antigen (anti-HBs) threshold of after series completion (an accepted only subjects with chronic kidney disease 10 mIU/mL was not reported; immune marker of immune protection) defined (i.e., no hemodialysis patients) (32), and response was not measured between seroprotection (4). When multiple im- three studies included both hemodialysis 0 and 6 months after the last vaccine mune response measurements were re- and chronic kidney disease patients dose (because anti-HBs titers wane over ported reflecting seroprotection at different (36,42,44), one of which also included time); 10 or fewer subjects with diabetes intervals during the primary vaccine se- peritoneal dialysis patients (44). The re- were included; vaccine was administered ries, the result after the final dose was re- mainder were comprised entirely of he- intradermally; or subjects were not naïve corded. When immune response was modialysis patients. Seroprotection to the vaccine or had positive serology for measured more than once after vaccina- against hepatitis B virus infection was as- HBsAg, anti-HBs, or antibody to hepatitis tion, the result from the interval closest sessed in subjects by diabetes status in 16 B core antigen, indicating past or current to 1 to 2 months after the last dose (cor- studies (1,764 subjects, 633 with diabe- hepatitis B infection. When two studies responding to the recommended interval tes). Five studies included diabetes in a reported results for duplicate subjects, for postvaccination testing for serologic multivariate model (31,32,37,42,44). one study was excluded. Studies were response in the U.S. [4]) was recorded. No serious vaccine-related adverse event notexcludedwhensubjectsreceivedad- When immune response was measured was reported in any study. ditional vaccine dose(s) because they did after a booster dose(s) or novel adjuvant The average age of subjects by study not achieve seroprotective concentrations administration, results were recorded ranged from 8.4 to 79.5 years. Diabetes or when inclusion of a fraction of subjects separately and noted. type was type 1 in five studies, type 2 in meeting an exclusion criterion was Odds ratios for attainment of sero- one study, and type 1 and type 2 in two deemed unlikely to affect the findings. protection were recorded for studies re- studies. Diabetes type was not specified The determination as to inclusion or porting results from a multiple logistic in eight of the nine studies among exclusion was based on the above criteria regression model. When odds ratios and hemodialysis/chronic kidney disease pa- and was made irrespective of study 95% CIs for the diabetes predictor vari- tients; persons with type 2 diabetes constitute results. able were reported for failure to achieve more than half of persons with diabetes seroprotection, the probability of attain- starting dialysis in the U.S. (45). The vac- Data extraction and categorization ing seroprotection was calculated by tak- cine type was exclusively recombinant in Data were manually extracted and elec- ing the inverse of the reported values. 15 studies, one of which used a combina- tronically recorded. Studies were classi- tion hepatitis A/hepatitis B vaccine (43), fied into one of three subject categories: RESULTSdThe systematic electronic and was exclusively plasma derived in one adults (mean age .18 years), children search yielded 225 studies (53 from study; both recombinant and plasma- (mean age #18 years), and hemodialysis/ MEDLINE, 105 from EMBASE, 20 from derived vaccines were used in one study. chronic kidney disease patients. Elements Cochrane Library, and 47 from Web of A novel adjuvant was administered to sub- for extraction included study charac- Knowledge; a list of these studies is avail- jects in two studies (40,41). Vaccine dos- teristics (e.g., study design, sample size, able upon request). Of these 225 studies, age ranged from 3 to 40 mg. Route of care.diabetesjournals.org DIABETES CARE, VOLUME 35, DECEMBER 2012 2691 Hepatitis B vaccine response in diabetes

hemodialysis/chronic kidney disease pa- tients, ranging from 41.8–85.3% (me- dian, 60.1%) for those with diabetes and 61.8–87.5% (median, 75.1%) for those without diabetes. Seroprotection proportions among subjects with diabetes were generally lower when vaccine was administered on a 0-, 1-, and 2-month schedule (not approved in the U.S.) and greater when vaccine was administered on a 0-, 1-, 2-, and 12-month schedule (U.S.-approved alternative schedule) compared with the standard 0-, 1-, and 6-month schedule. Two studies using a 0-, 1-, and 2-month schedule reported relatively low seropro- tection proportions for persons with di- abetes in their corresponding subject categories (54.2% among children [35] and 45.8% among hemodialysis/chronic kidney disease patients [34]), and two studies using a 0-, 1-, 2-, and 12-month schedule reported relatively high seropro- tection proportions among subjects with diabetes in their corresponding subject categories (88.6% and 94.4% [13] among adults and 66.7% among hemodialysis/ chronic kidney disease patients [33]). When a fourth dose was administered to children with diabetes after a primary se- ries on a 0-, 1-, and 2-month schedule, seroprotection increased from 54.2 to 100.0% in one study (35). Another study, however, achieved an 88.2% seroprotec- tion proportion when vaccinating adults with diabetes subcutaneously using Figure 1dSearch results. *Articles that were not a peer-reviewed, published randomized clinical a 0-, 1-, and 2-month schedule trial or observational study were not double-counted (among articles that did not specify the with thymopentin as an adjuvant (41). proportion [numerator and denominator] of subjects seroprotected by diabetes status [or data Thymopentin is a synthetic pentapeptide available to allow the calculation of the proportions] or the odds ratio from a multivariate with activity characteristic of the thymic analysis that included diabetes as a predictor variable) so that the total would equal 15. †The hormone thymopoietin and is not ap- number of articles does not total 13 because one article met two inclusion criteria. proved for use in the U.S. It has been used to enhance interleukin 2 production administration was exclusively intramus- was assessed were equivalent for the dia- and function of macrophages (46). cular in 15 studies and exclusively subcu- betes and comparison groups except as Administration of additional hepatitis taneous in one study; one study used noted in Table 4. B vaccine doses may improve immune intramuscular or intradermal adminis- Seroprotection proportions ranged response among adults with diabetes. tration routes in separate study arms. from 31.3–100.0% (median, 73.4%) Douvin et al. (13) reported seroprotective The standard 0-, 1-, and 6-month sched- among persons with diabetes and anti-HBs levels in 91.5% of adult subjects ule was used exclusively or partially in 35.2–100.0% (median, 87.1%) for those with diabetes compared with 75.0% as 11 studies. Others were a 0-, 1-, 2-, and without diabetes. The proportion pro- reported by Bouter et al. (14). Subjects 12-month schedule and a 0-, 1-, 2-, and tected was generally greatest among in both studies received 20-mgdosages 6-month schedule (both U.S.-approved children, ranging from 54.2–100.0% of recombinant hepatitis B vaccine intra- alternatives), as well as a 0-, 1-, and (median, 93.9%) among children with muscularly. Subjects in the study by 2-month schedule (not approved in the diabetes and 98.0–100.0% (median Douvin et al. were, on average, older U.S.). The sex distribution was generally 100.0%) among those without diabetes. than those in the study by Bouter et al. equivalent for most studies, although Among adults, seroprotection propor- (mean age, 49.2 vs. 34.3 years). It is there- men predominated in the studies among tions ranged from 31.3–94.4% (median, fore reasonable that subjects from the hemodialysis/chronic kidney disease pa- 88.2%) for those with diabetes com- study by Douvin et al. would have re- tients. The type of vaccine, dosage, route pared with 35.2–96.9% (median, sponded more poorly than those from of administration, schedule, and the inter- 93.6%) for those without diabetes. Sero- the study by Bouter et al. because older val after the last dose when seroprotection protection proportions were lowest for age (6–9,31,47) is associated with a decline

2692 DIABETES CARE, VOLUME 35, DECEMBER 2012 care.diabetesjournals.org Schillie, Spradling, and Murphy

Table 2dExclusion criteria as applicable for 13 studies patients, Lacson et al. (37) reported no decrease in seroprotection associated – Authors (reference) Publication year Exclusion criteria with diabetes (OR 1.02 [95% CI 0.93 1.12]), although a subanalysis of 814 Ahishali et al. (17) 2008 #10 subjects with diabetes pairs (including subjects with and with- Chow et al. (18) 2006 #10 subjects with diabetes out diabetes) matched for age, sex, race, Eldesoky et al. (19) 2009 Immune response measured .6 months dialysis vintage, and dialysis modality after last dose reported a decrease in seroprotection as- Elwell et al. (20) 2003 Immune response measured .6 months sociated with diabetes (OR 0.71 [95% CI after last dose 0.52–0.97]). Halotaetal.(21) 2002 Somesubjectsanti-HBcpositive A lower seroprotection proportion Kramer et al. (22) 2009 Immune response threshold other was achieved among children with di- than 10 mIU/mL abetes who received the vaccine using Marseglia et al. (23) 2000 Immune response measured .6 months intradermal administration, which is not after last dose, duplicate subjects from a U.S.-approved route of administration. Marseglia et al., 1996 Li Volti et al. (38) reported 77.8% sero- Morais et al. (24) 2007 Vaccine administered ID protectionwithintradermaladministra- Pascasio et al. (25) 2008 Some subjects anti-HBc positive tion compared with 100.0% seroprotection Pozzilli et al. (26) 1987 Immune response threshold other with intramuscular administration. Among than 10 mIU/mL children without diabetes, 100.0% in Somboonsilp et al. (27) 2003 Vaccine administered ID both intradermal and intramuscular Tsouchnikas et al. (28) 2007 #10 subjects with diabetes arms achieved seroprotection (38). Other Wismans et al. (29) 1991 Duplicate subjects from Bouter studies have found improved results with et al., 1992 intradermal administration of the hepatitis Anti-HBc, antibody to hepatitis B core antigen (total or not specified); ID, intradermally. B vaccine in dialysis patients (48). Thymopentin or priming with teta- in response. However, differences in the Five studies among hemodialysis/ nus toxoid (TT) was used in two studies number of doses administered likely con- chronic kidney disease patients included to enhance the immune response to the tributed to higher seroprotective propor- diabetes in a multivariate model hepatitis B vaccine. Thymopentin was tions in the Douvin et al. study, in which (31,32,37,42,44) (Table 5). All studies administered to adults with diabetes three 30 subjects (42.3%) receiving vaccine on a controlled for age, and three controlled times per week for a week before hepatitis 0-, 1-, 2-, and 12-month schedule also for obesity. Four studies, with a total of B vaccination and for 3 weeks afterward. received a booster dose at month 4 (for 587 subjects (186 with diabetes), reported The results were compared with hepatitis an anti-HBs titer ,10 mIU/mL, according that diabetes was independently associated B vaccination among normal control sub- to the study protocol). Therefore these sub- with failure to achieve seroprotection jects(41).TTwasadministered2days jects received five total doses, compared (odds ratio [OR] 0.23–0.50 [95% CI before an additional dose of hepatitis B with three total doses on a 0-, 1-, and 0.09–0.96]) (31,32,42,44). In their study vaccine to hemodialysis/chronic kidney 6-month schedule reported by Bouter et al. of 14,546 predominantly hemodialysis disease patients who were nonresponders after 6 or 7 doses of hepatitis B vaccine thathadbeenadministeredduringthe Table 3dCharacteristics of included studies previous 18 months (40). Although dif- ferences in route of administration and Author(s) (reference) Publication year Country Study design N number of booster doses administered prevent comparison with other studies, Arslanoglu et al. (30) 2002 Turkey Prospective 150 seroprotection was reported in 88.2% Bouter et al. (14) 1992 Netherlands Prospective 64 of adults with diabetes who received the Chin (31) 2003 U.S. Retrospective 97 thymopentin adjuvant and in 100.0% of DaRoza et al. (32) 2003 Canada Prospective 165 previously nonresponding hemodialysis/ Douvin et al. (13) 1997 France RCT 71 chronic kidney disease patients with Eardley et al. (33) 2002 U.K. Prospective 86 diabetes who had received TT before vac- Fabrizi et al. (34) 1996 Italy Prospective 188 cination. Subsequent meta-analysis of Fiçicioglu et al. (35) 1995 Turkey Prospective 41 studies using thymopentin adjuvant Hashemi et al. (36) 2011 Iran Prospective 167 with hepatitis B vaccine in end-stage renal Lacson et al. (37) 2005 U.S. Retrospective 14,546 disease patients (with and without dia- Li Volti et al. (38) 1998 Italy RCT 42 betes) found limited benefit from the ad- Marseglia et al. (39) 1996 Italy Prospective 239 dition of thymopentin to hepatitis B Ocak et al. (40) 2008 Turkey Prospective 49 schedules (46). Pagani et al. (41) 1989 U.K. Prospective 64 Taheri et al. (42) 2005 Iran Prospective 125 CONCLUSIONSdWhen hepatitis B Williams et al. (43) 2012 U.S. Prospective 86 vaccination is performed in accordance Zitt et al. (44) 2012 Austria Retrospective 200 with standard administration procedures, RCT, randomized clinical trial. children and young adults with diabetes care.diabetesjournals.org DIABETES CARE, VOLUME 35, DECEMBER 2012 2693 2694 diabetes in response vaccine B Hepatitis D IABETES

C Table 4dStudies included in review: demographic data, vaccination schedule, and seroprotection (SP) proportion (anti-HBs cutoff of 10 mIU/mL) ARE , VOLUME When Seroprotection proportion seroprotection 5 D 35, Authors Study Diabetes Schedule assessed Diabetes, No diabetes, (reference) population Age (years)* type Vaccine Dose (mg) Route (months) (months) %(n/N) %(n/N) ECEMBER Bouter et al. (14) Adults 34.3 1 Recombinant 20 IM 0, 1, and 6 1 75.0 (24/32) 96.9 (31/32) Douvin et al. (13) Adults 52.4 1 and 2 Engerix-B 20 IM 0, 1, 2, and 12† 1 94.4 (34/36) NR 02care.diabetesjournals.org 2012 46.0 GenHevac B 20 IM 0, 1, 2, and 12† 1 88.6 (31/35) NR Pagani et al. (41) Adults 50 1 and 2 Haevac B (+thymopentin if diabetes) 5 SC 0, 1, and 2 3 88.2 (15/17) 93.6 (44/47) Williams et al. (43) Adults 79.5* NR Twinrix 20 IM 0, 1, and 6 1–2 31.3 (10/32) 35.2 (19/54) Arslanoglu et al. (30) Children 10.8 1 Engerix-B‡ 10 IM 0, 1, and 6 1–6x 93.9 (93/99) 98.0 (50/51) Fiçicioglu et al. (35) Children 11.5 1 GenHevac B 20 IM 0, 1, and 2 6 5 2 54.2 (13/24)|| 100.0 (17/17) Li Volti et al. (38) Children 8.4 1 Engerix-B 10–20 IM 0, 1, and 6 1 100.0 (9/9) 100.0 (12/12) 9.8 3 ID 0, 0.5, 1, 1.5–2 1 77.8 (7/9) 100.0 (12/12) Marseglia et al. (39) Children 17.3 1 Recombivax HB 10{ IM 0, 1, and 6 2 95.4 (62/65) 98.3 (171/174) Chin (31) HD/CKD 53.7* NR Recombivax HB 40 IM 0, 1, and 6** 4–6 52.1†† (25/48) 79.6†† (39/49) DaRoza et al. (32)# HD/CKD 59.8* NR Recombinant or 0, 1, and 6 OR 3 71.7 (33/46) 86.6 (103/119) plasma 40 IM 0, 1, 2, and 6 Eardley et al. (33) HD/CKD 61* NR HB Vax II 40 IM 0, 1, 2, and 12 1.5–2 66.7 (8/12) 74.3 (55/74) Fabrizi et al. (34)xx HD/CKD 63.4* NR Engerix-B 40 IM 0, 1, and 2 1 45.8 (11/24) 72.3 (68/94) Hashemi et al. (36) HD/CKD 57.4* NR Engerix-B 40 IM 0, 1, and 6 2–3 85.3 (29/34) 75.9 (101/133) Lacson et al. (37) HD/CKD 59* NR Recombivax HB 40 IM 0, 1, and 6 6 NR NR Engerix-B 40 IM 0, 1, 2, and 6 6 NR NR Ocak et al. (40) HD/CKD 61.6* 2 Euvax B 6 tetanus 40 IM 0, 1, 2, and 6 6 2 2 57.9|||| (11/19) 70.0|||| (21/30) toxoid OR 3 boosters Taheri et al. (42) HD/CKD 50* NR Heberbiovac HB 40 IM 0, 1, and 6 1–6 62.2 (23/37) 87.5 (77/88) Zitt et al. (44) HD/CKD 64* NR HBVaxPro or 40 IM 0, 1, and 6 2 41.8 (23/55) 61.8 (89/144) Engerix-B HD/CKD, hemodialysis/chronic kidney disease patients; ID, intradermally; IM, intramuscularly; NR, not reported; SC, subcutaneously. *Mean age for subjects with and without diabetes. †Thirty subjects received one booster at month 4 because of anti-HBs ,10 mIU/mL. ‡GenHevac B used in 23 of 51 controls. xTime in months after last dose when measured range of anti-HBs among persons with diabetes was 0.1–23.4; range among controls 1–6 months. ||Response 100.0% after fourth dose. {One subject aged 4.5 years received a 5-mg dose. #DaRoza et al. (32) included only included patients before dialysis initiation. **Second series given if no response. ††Response assessed on two occasions; subjects responding on only the first occasion were not classified as responder or nonresponder. xxSome subjects were antibody to hepatitis B core antigen positive. ||||Responses of 78.9% and 96.6% after additional doses in diabetes and comparison groups, respectively; response 100.0% in each group after administration of tetanus toxoid and vaccine. Schillie, Spradling, and Murphy

Table 5dMultiple logistic regression model results for attainment of seroprotective response

Authors (references) Predictor variable Other variables in model Matching Odds ratio 95% CI Chin (31) Diabetes Age, weight, initial nPCR – 0.29*† 0.11–0.77* DaRoza et al. (32) Diabetes Age, sex, use of erythropoietin, – 0.27 0.10–0.71 albumin level, hemoglobin level, urea level, weight, GFR Lacson et al. (37)‡ Diabetes Age, sex, race, vaccine type, albumin, Age, sex, race, 1.02x 0.93–1.12 BSA, hemoglobin, vintage vintage, modalityx (square root), modality, hepatitis C Taheri et al. (42) Diabetes Age, sex, dialysis treatment – 0.23 0.09–0.59* Zitt et al. (44) Diabetes Age, sex, modality, 25(OH)D, – 0.50 0.26–0.96 VDRA treatment 25(OH)D, 25-hydroxyvitamin D; BSA, body surface area; GFR, glomerular filtration rate; nPCR, normalized protein catabolic rate; VDRA, vitamin D receptor ac- tivator. *Inverse of values reported by author because reported values were for failure to achieve seroprotection response. †Diabetes was no longer significant when the parameter for baseline serum albumin was added into the model. ‡Seroprotection response threshold of 10 mIU/mL was confirmed by personal communication with author. xAuthors performed a matched analysis on a subset of 814 pairs (for this subset analysis, matched variables were excluded from multivariate model); odds ratio and 95% CI from matched analysis were 0.71 and 0.52–0.97, respectively. generally have responses similar to per- tissue may delay antigen presentation to effective (Centers for Disease Control sons of comparable age without diabetes. the B and T cells responsible for the im- and Prevention, 2011, unpublished Older adults have a reduced response, mune response (52). However, the needle data) and not recommended, using and older adults with diabetes seem to should not be so long that it involves the schedules requiring four rather than three have further impairment in vaccine re- underlying bone (4). doses or administering additional doses sponse, particularly those with coexisting In addition to obesity (52), older age of hepatitis B vaccine remains an option. kidney conditions. (6–9,47), comorbid conditions (11), and A longer interval between the final There are several hypotheses about medication use (12) have been associated two doses of the primary hepatitis B the biological basis for potentially im- with impaired vaccine response and may vaccine series has been associated with paired response to vaccination among confound the relationship between diabe- higher final anti-HBs levels in general (4). persons with diabetes. Although persons tes and immune response. Four multivari- Findings from this review suggest that a with diabetes have appropriate humoral ate analyses among hemodialysis/chronic longer interval between the final two immune responses to vaccination (49), kidney disease patients included in this doses also is associated with increased se- impaired cellular response may account review, all of which controlled for age roprotection proportions among persons for less robust antibody production after and three of which controlled for obesity, with diabetes. Excessively long intervals hepatitis B vaccination (14). Proposed ex- found diabetes to be an independent fac- have the drawback of increasing the risk planations include a reduction in the tor for impaired vaccine response for acquisition of HBV infection among number of circulating helper T cells, the (31,32,42,44). Lacson et al. (37), however, those with an incomplete series (4). CD4-to-CD8 lymphocyte ratio, and lym- reported a null association with diabetes Other strategies that may augment the phocyte blastogenesis (23) and defects and seroprotection among hemodialysis/ response to hepatitis B vaccine in persons with antigen presentation (39). Impaired chronic kidney disease patients (N = with diabetes include the use of novel vaccine response also has been linked to 14,546; OR 1.02 [95% CI 0.93–1.12]) in adjuvants, higher dosages, or combination the presence of DR3, DR7, and DQ2 hu- their principal multivariate analysis. vaccines. No hepatitis B vaccine with a man leukocyte antigen alleles among per- Administration of additional doses of novel adjuvant is currently licensed for use sons with diabetes (13). However, no hepatitis B vaccine improves the propor- in the U.S. Recent prelicensure trials using a association between seroprotection and tion of persons responding to it. Addi- two-dose schedule with a hepatitis B vac- glycemic control (13,14,30,39), duration tional doses have not caused unusual cine containing a toll-like receptor 9 of diabetes (13,30,39), insulin require- adverse reactions (16). It is recommended (HBsAg-1018 ISS) agonist as the adjuvant ment (30,39), or microangiopathic com- that health care personnel at continuing seems promising (53). Other strategies that plications (39) was demonstrated in risk of hepatitis B exposure and infants could be explored in clinical trials include studies included in this review. with perinatal hepatitis B exposure the use of the combined hepatitis A virus Obesity is a major risk factor for type repeat a primary series of hepatitis B vac- and HBV vaccine Twinrix (GlaxoSmithK- 2 diabetes; 53% of adults with diabetes cine if they are found to be nonresponsive line Biologicals) (54,55) or hemodialysis are obese (50). Studies have demon- to the initial series (4,5). Among health dosage of the hepatitis B vaccine (56). Be- strated an association between obesity care personnel who did not respond cause response declines with advancing age and a reduced response to hepatitis B vac- after a primary series, 67.6% mounted a and comorbidities (11), vaccination soon after cine in adults (51). A needle length that is protective response up to three additional diabetes diagnosis likely would constitute the inadequate to penetrate the deltoid fat doses (11). Data exclusively on revaccina- most practical way to optimize seroprotection. pad and reach the muscle mass may ac- tion of nonresponding persons with This review has several limitations. It count for the reduced immune response diabetes are sparse. Although postvacci- is possible that all relevant studies may among persons with obesity (4,52). The nation serology to determine responses not have been identified in the literature less abundant blood supply in adipose among adults with diabetes is not cost search, especially those studies in which care.diabetesjournals.org DIABETES CARE, VOLUME 35, DECEMBER 2012 2695 Hepatitis B vaccine response in diabetes persons with diabetes represented a sub- the final dose, four vs. three doses, or 6. Halsey NA, Moulton LH, O’Donovan JC, set of the population. There is the poten- additional vaccine doses may achieve et al. Hepatitis B vaccine administered to tial for publication bias. Studies with seroprotection in a greater proportion of children and adolescents at yearly inter- – significant results are more likely to be adults with diabetes, including those who vals. Pediatrics 1999;103:1243 1247 published than those with null results, do not respond to an initial series. Other 7. Clemens R, Sänger R, Kruppenbacher J, et al. Booster immunization of low- and and this review consisted only of pub- strategies to increase response among non-responders after a standard three lished studies. A null result for some stud- older adults with diabetes, including those dose hepatitis B vaccine scheduledresults ies in this review would consist of an with other comorbid conditions, should of a post-marketing surveillance. Vaccine equivalent vaccine response in persons be explored. 1997;15:349–352 with and without diabetes. It is therefore 8. Hussain Z, Ali SS, Husain SA, Raish M, unlikely that publication bias would lead Sharma DR, Kar P. Evaluation of immu- to a reported overestimate of vaccine re- AcknowledgmentsdNo potential conflicts of nogenicity and reactogenicity of recombi- sponse among persons with diabetes. interest relevant to this article were reported. nant DNA hepatitis B vaccine produced fi S.F.S. performed the literature review, ab- in India. World J Gastroenterol 2005;11: Only 17 of the identi ed studies were – included in this review (and only 5 mul- stracted data, and wrote the manuscript. P.R.S. 7165 7168 9. Honorati MC, Mariani E, Dolzani P, tivariate analyses), and data pertaining critically reviewed and edited the manuscript. T.V.M. suggested the topic and critically re- Facchini A. Biological parameters influencing to adults with diabetes but without viewed and edited the manuscript. S.F.S. is the the immunological response to plasma de- hemodialysis/chronic kidney disease were guarantor of this work and, as such, had full rived and recombinant hepatitis B vaccines. sparse. Observational studies were included, access to all the data in the study and takes Ann Ist Super Sanita 1996;32:369–374 and the results of these studies may not be responsibility for the integrity of the data and 10. Fabrizi F, Martin P, Dixit V, Bunnapradist as robust as randomized clinical trials. The the accuracy of the data analysis. S, Dulai G. Meta-analysis: the effect of age matched comparison groups and ob- A portion of the data from five included on immunological response to hepatitis B jective outcome measure (i.e., anti-HBs studies was presented at a Meeting of the Ad- vaccine in end-stage renal disease. Aliment visory Committee on Immunization Practices, Pharmacol Ther 2004;20:1053–1062 levels) strengthen the data quality. – Outcomes consisted of serological corre- Atlanta, Georgia, 25 26 October 2011. 11. Averhoff F, Mahoney F, Coleman P, Schatz lates of protection (anti-HBs levels), The authors thank Jack (Rick) Colbert, MLIS, G, Hurwitz E, Margolis H. Immunogenicity Reference Librarian, CDC Public Health Library of hepatitis B Vaccines. Implications for which may not correlate with maintaining and Information Center, Atlanta Georgia, for his persons at occupational risk of hepatitis B full vaccine effectiveness over time. Im- assistance with the literature search. virus infection. Am J Prev Med 1998;15:1–8 munocompetent persons found to have 12. de Rave S, Heijtink RA, Bakker-Bendik M, anti-HBs levels of $10 IU/L after the pri- Boot J, Schalm SW. Immunogenicity of mary series, consistent with the threshold References standard and low dose vaccination using used by studies included in this review, 1. Centers for Disease Control and Prevention yeast-derived recombinant hepatitis B sur- are considered to be protected (4,5,57). (CDC). Use of hepatitis B vaccination for face antigen in elderly volunteers. Vaccine Duration of protection was not assessed adults with diabetes mellitus: recommen- 1994;12:532–534 in this review, although other evidence dations of the Advisory Committee on Im- 13. Douvin C, Simon D, Charles MA, et al. suggests protection lasts for two decades munization Practices (ACIP). MMWR Morb Hepatitis B vaccination in diabetic patients. – Randomized trial comparing recombinant in healthy primary vaccine responders. Mortal Wkly Rep 2011;60:1709 1711 2. André FE. Summary of safety and efficacy vaccines containing and not containing pre- However, data on the duration of immune data on a yeast-derived hepatitis B vac- S2 antigen. Diabetes Care 1997;20:148–151 memory in immunocompromised per- cine. Am J Med 1989;87(3A):14S–20S 14. Bouter KP, Diepersloot RJ, Wismans PJ, sons are limited (4). Although general 3. Zajac BA, West DJ, McAleer WJ, Scolnick et al. Humoral immune response to a yeast- conclusions were drawn from a synthesis EM. Overview of clinical studies with derived hepatitis B vaccine in patients with of the results, no attempt was made to hepatitis B vaccine made by recombinant type 1 diabetes mellitus. Diabet Med 1992; test for statistically significant differences. DNA. J Infect 1986;13(Suppl A):39–45 9:66–69 A meta-regression was not conducted 4. Mast, EE, Weinbaum CM, Fiore AE, et al.; 15. Historic dates and events related to vac- because colinearity likely exists among Advisory Committee on Immunization cines and immunization. Available from: sources of heterogeneity (e.g., children Practices (ACIP) Centers for Disease Control http://www.immunize.org/timeline/. Ac- most likely have type 1 diabetes and are and Prevention (CDC). A comprehensive cessed 5 January 2012 immunization strategy to eliminate trans- 16. Hadler SC, Margolis HS. Hepatitis B im- treated with insulin). Heterogeneity be- mission of hepatitis B virus infection in the munization: vaccine types, efficacy, and tween studies by diabetes type, manage- United States: recommendations of the indications for immunization. Curr Clin ment, glycemic control, and vaccine type, Advisory Committee on Immunization Top Infect Dis 1992;12:282–308 dosage, administration route, and sched- Practices (ACIP) Part II: immunization of 17. Ahishali E, Boztas G, Akyuz F, et al. Re- ule limit synthesis of results. adults. MMWR Recomm Rep 2006;55(RR- sponse to hepatitis B vaccination in pa- High levels of seroprotection from 16):1–33; quiz CE1–4. tients with celiac disease. Dig Dis Sci recombinant hepatitis B vaccine are 5. Mast EE, Margolis HS, Fiore AE, et al.; 2008;53:2156–2159 achieved safely in children with diabetes. Advisory Committee on Immunization Prac- 18. Chow KM, Law MC, Leung CB, Szeto CC, Data from studies reviewed here suggest tices (ACIP). A comprehensive immunization Li PK. Antibody response to hepatitis B that adults with diabetes have a reduced strategy to eliminate transmission of hep- vaccine in end-stage renal disease patients. atitis B virus infection in the United States: Nephron Clin Pract 2006;103:c89–c93 response to vaccination. Seroprotection recommendations of the Advisory Com- 19. Eldesoky A, et al. Protective immunity is decreased among older adults and per- mittee on Immunization Practices (ACIP) after hepatitis B vaccination. Arab J Gas- sons on hemodialysis in general, including Part 1: immunization of infants, children, troenterol 2009;10:68–71 those with diabetes. Administration of and adolescents. MMWR Recomm Rep 20. Elwell RJ, Neumann M, Bailie GR. Factors schedules using an extended interval to 2005;54(RR-16):1–31 associated with long-term antibody

2696 DIABETES CARE, VOLUME 35, DECEMBER 2012 care.diabetesjournals.org Schillie, Spradling, and Murphy

production induced by hepatitis B vaccine characteristics. Hemodial Int 2003;7: 44. Zitt E, Sprenger-Mähr H, Knoll F, Neyer U, in patients undergoing hemodialysis: 296–303 Lhotta K. Vitamin D deficiency is associated a retrospective cohort study. Pharmaco- 32. DaRoza G, Loewen A, Djurdjev O, et al. with poor response to active hepatitis B therapy 2003;23:1558–1563 Stage of chronic kidney disease predicts immunisation in patients with chronic 21. Halota W, Muszynska M, Paw1owska M. seroconversion after hepatitis B immuni- kidney disease. Vaccine 2012;30:931–935 Hepatitis B virus serologic markers and zation: earlier is better. Am J Kidney Dis 45. Molitch ME, DeFronzo RA, Franz MJ, anti-hepatitis B vaccination in patients 2003;42:1184–1192 et al.; American Diabetes Association. with diabetes. Med Sci Monit 2002;8: 33. Eardley KS, Jones HE, Osman H, Smith Nephropathy in diabetes. Diabetes Care CR516–CR519 SA. Efficacy of the accelerated hepatitis B 2004;27(Suppl 1):S79–S83 22. Kramer ES, Hofmann C, Smith PG, vaccination schedule used in haemodial- 46. Fabrizi F, Dixit V, Martin P. Meta-analysis: Shiffman ML, Sterling RK. Response to ysis patients post-exposure to virus: the adjuvant role of thymopentin on im- hepatitis A and B vaccine alone or in a single-centre experience. Nephrol Dial munological response to hepatitis B virus combination in patients with chronic Transplant 2002;17:1982–1987 vaccine in end-stage renal disease. Aliment hepatitis C virus and advanced fibrosis. 34. Fabrizi F, Di Filippo S, Marcelli D, et al. Pharmacol Ther 2006;23:1559–1566 Dig Dis Sci 2009;54:2016–2025 Recombinant hepatitis B vaccine use in 47. Wolters B, Junge U, Dziuba S, Roggendorf 23. Marseglia G, Alibrandi A, d’Annunzio G, chronic hemodialysis patients. Long-term M. Immunogenicity of combined hepati- et al. Long term persistence of anti-HBs evaluation and cost-effectiveness analysis. tis A and B vaccine in elderly persons. – protective levels in young patients with Nephron 1996;72:536–543 Vaccine 2003;21:3623 3628 type 1 diabetes after recombinant hepa- 35. Fiçicioglu C, Mikla S, Midilli K, Aydin A, 48. Fabrizi F, Dixit V, Messa P, Martin P. titis B vaccine. Vaccine 2000;19:680– CamH,Ergin S. Reduced immune re- Intradermal vs intramuscular vaccine 683 sponse to hepatitis B vaccine in children against hepatitis B infection in dialysis 24. Morais EO, Resende MR, Oliveira AM, et al. patients: a meta-analysis of randomized with insulin dependent diabetes. Acta – Intradermal hepatitis B vaccination in pa- Paediatr Jpn 1995;37:687–690 trials. J Viral Hepat 2011;18:730 737 49. Smith SA, Poland GA; American Diabetes tients with advanced chronic renal failure: 36. Hashemi B, Mahdavi-Mazdeh M, Abbasi fl immunogenicity and follow-up. Aliment M, Hosseini-Moghaddam SM, Zinat NH, Association. In uenza and pneumococcal – fi immunization in diabetes. Diabetes Care Pharmacol Ther 2007;25:849 855 Ahmadi F. Ef cacy of HBV vaccination in – 25. Pascasio JM, Aoufi S, Gash A, et al. Re- various stages of chronic kidney disease: is 2004;27(Suppl 1):S111 S113 – 50. Age-adjusted percentage of obesity for sponse to a vaccination schedule with 4 earlier better? Hepat Mon 2011;11:816 – doses of 40 microg against hepatitis B vi- 820 adults with diabetes, United States, 1994 2007. Available from: http://www.cdc.gov/ rus in cirrhotic patients evaluated for liver 37. Lacson E, Teng M, Ong J, Vienneau L, diabetes/statistics/comp/fig7_obesity.htm. transplantation. Transplant Proc 2008; Ofsthun N, Lazarus JM. Antibody re- Accessed 6 February 2012 40:2943–2945 sponse to Engerix-B and Recombivax-HB 51. Roome AJ, Walsh SJ, Cartter ML, Hadler 26. Pozzilli P, Arduini P, Visalli N, et al. Re- hepatitis B vaccination in end-stage renal JL. Hepatitis B vaccine responsiveness in duced protection against hepatitis B virus disease. Hemodial Int 2005;9:367–375 Connecticut public safety personnel. following vaccination in patients with 38. Li Volti S, Caruso-Nicoletti M, Biazzo F, JAMA 1993;270:2931–2934 type 1 (insulin-dependent) diabetes. Dia- et al. Hyporesponsiveness to intradermal – 52. Middleman AB, Anding R, Tung C. Effect betologia 1987;30:817 819 administration of hepatitis B vaccine in of needle length when immunizing obese 27. Somboonsilp W, Eiam-Ong S, Tungsanga insulin dependent diabetes mellitus. Arch adolescents with hepatitis B vaccine. Pe- K, Tirawatanapong T. Immune response Dis Child 1998;78:54–57 – ’ diatrics 2010;125:e508 e512 of intradermal hepatitis B vaccination at 39. Marseglia GL, Scaramuzza A, d Annunzio 53. Barry M, Cooper C. Review of hepatitis B lower dose versus intramuscular vaccina- G, Comolli G, Gatti M, Lorini R. Suc- surface antigen-1018 ISS adjuvant-containing tion at double standard dose in predialytic cessful immune response to a recombi- vaccine safety and efficacy. Expert Opin Biol chronic renal failure patients. J Med Assoc nant hepatitis B vaccine in young patients Ther 2007;7:1731–1737 – Thai 2003;86:1122 1127 with insulin-dependent diabetes mellitus. 54. Nyström J, Cardell K, Björnsdottir TB, – 28. Tsouchnikas I, Dounousi E, Xanthopoulou Diabet Med 1996;13:630 633 Fryden A, Hultgren C, Sällberg M. Im- K, Papakonstantinou S, Thomoglou V, 40. Ocak S, Eskiocak AF. The evaluation of proved cell mediated immune responses Tsakiris D. Loss of hepatitis B immunity in immune responses to hepatitis B vaccination after successful re-vaccination of non- hemodialysis patients acquired either nat- in diabetic and non-diabetic haemodialysis responders to the hepatitis B virus surface urally or after vaccination. Clin Nephrol patients and the use of tetanus toxoid. Ne- antigen (HBsAg) vaccine using the com- – – 2007;68:228 234 phrology (Carlton) 2008;13:487 491 bined hepatitis A and B vaccine. Vaccine 29. Wismans PJ, van Hattum J, de Gast GC, 41. Pagani S, Cruciani L, Chianelli M, Procaccini 2008;26:5967–5972 et al. A prospective study of in vitro E, Pozzilli P. Thymopentin administration 55. Murdoch DL, Goa K, Figgitt DP. Com- anti-HBs producing B cells (spot-ELISA) and increase of sero-conversion after bined hepatitis A and B vaccines: a review following primary and supplementary B-hepatitis vaccine in diabetic patients. of their immunogenicity and tolerability. vaccination with a recombinant hepatitis Diabetes Res 1989;12:199–201 Drugs 2003;63:2625–2649 B vaccine in insulin dependent diabetic 42. Taheri Sh, et al. Response rate to hepatitis 56. Greub G, Genton B, Safary A, Thoelen S, patients and matched controls. J Med B vaccination in patients with chronic Frei PC. Comparison of the reactogenicity Virol 1991;35:216–222 renal failure and end-stage-renal-disease: and immunogenicity of a two injection 30. Arslanoglu I, Cetin B, Isgüven P, Karavus influence of diabetes mellitus. J Res Med combined high-dose hepatitis A and hepa- M. Anti-HBs response to standard hepatitis Sci 2005;10:384–390 titis B vaccine to those of Twinrix. Vaccine B vaccination in children and adolescents 43. Williams RE, Sena AC, Moorman AC, 2000;19:1113–1117 with diabetes mellitus. J Pediatr Endocrinol et al. Hepatitis B vaccination of suscep- 57. Jack AD, Hall AJ, Maine N, Mendy M, Metab 2002;15:389–395 tible elderly residents of long term care Whittle HC. What level of hepatitis B an- 31. Chin AI. Hepatitis B virus vaccine re- facilities during a hepatitis B outbreak. tibody is protective? J Infect Dis 1999; sponse in hemodialysis: baseline patient Vaccine 2012;30:3147–3150 179:489–492

care.diabetesjournals.org DIABETES CARE, VOLUME 35, DECEMBER 2012 2697