Immunogenicity of an investigational hepatitis B vaccine (hepatitis B surface antigen co-administered with an immunostimulatory phosphorothioate oligodeoxyribonucleotide) in nonresponders to licensed hepatitis B vaccine

Abstract

An additional one to three doses of hepatitis B vaccine are recommended for nonresponders to an initial standard three-dose series. We compared the safety and immunogenicity of an investigational hepatitis B surface antigen vaccine (HBsAg-1018) with a phosphorothioate oligodeoxyribonucleotide adjuvant that targets toll-like receptor-9 to a commercially available, alum-adjuvanted hepatitis B vaccine (HBsAg-Eng) in nonresponders to three previous doses (primary study) or to four to six previous doses (substudy) of HBsAg-Eng. Both vaccines were well tolerated, although HBsAg-1018 was associated with more injection-site tenderness (63.2% vs. 18.8%, p = 0.016 in the primary study and 81.8% vs. 15.4%, p = 0.003 in the substudy). No statistically significant differences in rates of seroprotection (anti-HBs concentration ≥ 10 mIU/mL) or geometric mean antibody concentrations were found in the primary study. In the substudy, a greater proportion of HBsAg-1018 recipients achieved an anti-HBs concentration ≥ 100 mIU/mL (54.5% vs. 8.3%, p = 0.027), and those responders had higher geometric mean antibody concentrations at 4 weeks (264 vs. 46.5 mIU/mL, p = 0.021) and 52 weeks (7.0 vs. 1.2 mIU/mL, p = 0.030) than HBsAg-Eng recipients. Although this study suggests that HBsAg-1018 may have improved immunogenicity in nonresponders to hepatitis B vaccine vaccination when compared with HBsAg-Eng, larger studies are required.

Keywords: :

• hepatitis B vaccine
• immunostimulatory sequences
• nonresponder
• randomized trial
• hepatitis B
• immunogenicity

Introduction

Control of hepatitis B is best achieved by universal immunization which is recommended by the World Health Organization.¹ Currently available hepatitis B vaccines comprise recombinant hepatitis B surface antigen (HBsAg) expressed in yeast cells adsorbed to an alum adjuvant.² After the recommended three-dose regimen with a minimum interval of one month between the first two doses and at least four months between the second and third doses,^2-4 a substantial proportion (5–10%) of healthy young adults fail to achieve a protective immune response (defined as an anti-HBs concentration ≥ 10 mIU/mL).^5-7 Risk factors associated with increased rates of nonresponse include obesity, smoking, increased age, immunodeficiency, and genetic factors.^8,9

Additional doses of vaccine (up to a second three-dose series) are recommended for nonresponders, and as many as 50% will respond to this second three-dose course.^6,10,11 Experimental and newly licensed vaccines have been compared with the standard, marketed, alum-adjuvanted vaccines with variable success.^7,12-18 Immunostimulatory sequences are bacterial DNA or synthetic oligonucleotides that stimulate the innate immune system through toll-like receptor-9 (TLR-9).^19,20 1018 is a 22-mer phosphorothioate oligodeoxyribonucleotide with immunostimulatory activity.²¹ An investigational hepatitis B vaccine, HBsAg adjuvanted with 1018 (HBsAg-1018), has been shown to provide a higher and more rapid antibody response after one and two doses than a marketed alum-adjuvanted hepatitis B vaccine in healthy adults.^22-25 This enhanced immunogenicity might be useful for inducing a seroprotective response in nonresponders to previous vaccination with a licensed hepatitis B vaccine. The purpose of this study was to compare the safety and immunogenicity of a single dose of HBsAg-1018 to a commercially available hepatitis B vaccine (HBsAg-Eng) in nonresponders to an initial vaccine series with a licensed vaccine. In a substudy, the safety and immunogenicity of HBsAg-1018 was studied in individuals who were nonresponders to up to two complete, three-dose courses of a licensed hepatitis B vaccine.

Results

Demographics and participant disposition

A total of 35 participants were enrolled in the primary study and 24 participants in the substudy (Table 1). Demographic characteristics of the participants randomized to receive HBsAg-1018 or HBsAg-Eng were similar for the primary study and substudy except for gender distribution and current smoking status. The mean age across groups ranged from 40.9 to 42.9 y; however, 89% of primary study participants were male whereas 63% of substudy participants were female. All 35 participants enrolled in the primary study completed the study; all but one of the participants enrolled in the substudy completed the study. One participant in the substudy who was randomized to receive HBsAg-Eng was excluded from the analysis because her post-primary vaccination series serology documenting that she was a nonresponder had been taken 13 mo after her last immunization rather than the protocol-specified upper limit of 6 mo.

Table 1. Participant characteristics

Characteristic	Primary study		Substudy
	HBsAg-1018 (n = 19)	HBsAg-Eng (n = 16)	HBsAg-1018 (n = 11)	HBsAg-Eng (n = 13)
Age in years; mean (range)	42.9 (27–60)	42.4 (28–59)	40.9 (18–57)	42.7 (19–58)
Gender female; n (%)	3 (15.8)	1 (6.2)	8 (72.7)	7 (53.8)
Male; n (%)	16 (84.2)	15 (93.8)	3 (27.3)	6 (46.2)
Race Caucasian; n (%)	19 (100)	16 (100)	10 (90.9)	12 (92.3)
Black; n (%)	0 (0)	0 (0)	1 (9.1)	1 (7.7)
Weight in kilograms (range)	86.7 (63.6–126)	90.1 (54.5–106.8)	81.3 (56.8–118.1)	85 (53.6–131.8)
Height in centimeters (range)	174.6 (152.4–188)	175.9 (168–187)	171.6 (162.5–189)	168.7 (156.2–182.8)
Current smoker; n (%)	1 (5.3)	5 (31.3)	5 (45.5)	3 (23.1)

Antibody response

Four weeks after the first dose of study vaccine was given, 10 (52.6%) HBsAg-1018 recipients and 6 (37.5%) HBsAg-Eng recipients had developed seroprotection (Table 2). Of these responders, seroprotective levels were still present 26 and 52 weeks postvaccination in 8 (88.9%) and 5 (50%) HBsAg-1018 recipients, respectively, and 4 (66.7%) and 1 (16.7%) HBsAg-Eng recipients; none of the differences were statistically significant. Similar trends were observed in the magnitude of the immune response as demonstrated by the proportion achieving anti-HBs ≥ 100 mIU/mL and the geometric mean antibody concentration (GMC).

Table 2. Antibody response following vaccination with HBsAg-1018 or HBsAg-Eng in nonresponders to HBV

Parameter	Time measured	Nonresponders to 3 doses (primary study)		Nonresponders to 4 to 6 Doses (substudy)
		HBsAg-1018	HBsAg-Eng	HBsAg-1018	HBsAg-Eng
% Anti-HBs ≥ 10 mIU/mL	Week 4	10/19 (52.6)	6/16 (37.5)	7/11 (63.6)	7/12 (58.3)
	Week 26^a	8/9 (88.9)	4/6 (66.7)	4/7 (57.1)	3/7 (42.9)
	Week 52^a	5/10 (50.0)	1/6 (16.7)	4/7 (57.1)	0/7 (0)
% Anti-HBs ≥ 100 mIU/mL	Week 4	4/19 (21.1)	2/16 (12.5)	6/11 (54.5)*	1/12 (8.3)
	Week 26^a	0/9 (0)	1/6 (16.7)	0/7 (0)	0/7 (0)
	Week 52^a	1/10 (10.0)	0/6 (0)	0/7 (0)	0/7 (0)
Geometric Mean Antibody Concentration (GMC)	Day 0	1.00 (n = 19)	1.05 (n = 16)	1.00 (n = 11)	1.00 (n = 12)
	Week 4 all	14.7 (n = 19)	5.33 (n = 16)	34.8 (n = 11)	11.2 (n = 12)
	Week 4 responders	90.0 (n = 10)	41.1 (n = 6)	264.0 (n = 7)**	46.5 (n = 7)
	Week 26 responders	20.3 (n = 9)	14.2 (n = 6)	14.8 (n = 7)	3.22 (n = 7)
	Week 52 responders	12.7 (n = 10)	1.96 (n = 6)	6.99 (n = 7)***	1.21 (n = 7)

^a Participants with anti-HBs levels ≥ 10mIU/mL at week 4; *p = 0.027; **p = 0.021; ***p = 0.030

In the substudy of participants who were nonresponders to four to six prior doses of licensed hepatitis B vaccine, a similar proportion of participants who received HBsAg-1018 (63.6%) or HBsAg-Eng (58.3%) achieved postvaccination levels ≥ 10 mIU/mL with similar rates of persistence at 26 weeks; a nonsignificant trend toward increased persistence at 52 weeks was observed in the HBsAg-1018 group. The magnitude of the immune response was significantly higher in HBsAg-1018 recipients as measured by the proportion achieving levels ≥ 100 mIU/mL (54.5% vs. 8.3%; p = 0.027). The geometric mean antibody level in responders was also higher in the HBsAg-1018 recipients and reached statistical significance at 4 and 52 weeks postvaccination.

Three (75%) of 4 participants who did not respond to the study dose of HBsAg-1018 and 2 (28.6%) of 7 who did not respond to the study dose of HBsAg-Eng responded with protective antibody levels with two additional doses of HBsAg-Eng (Table 3).

Table 3. Immunogenicity and adverse events reported after immunization with second and third doses of HBsAg-Eng in nonresponders to HBsAg-ISS or HBsAg-Eng in the primary study

Measurement	Second HBsAg-Eng dose in nonresponders to		Third HBsAg-Eng dose in nonresponders to
	HBsAg-1018 (n = 5)	HBsAg-Eng (n = 8)	HBsAg-1018 (n = 4)	HBsAg-Eng (n = 7)
Immunogenicity
n (%) anti-HBs ≥ 10 mIU/mL	-	-	3 (75.0)	2 (28.6)
n (%) anti-HBs ≥ 100 mIU/mL	-	-	1 (25.0)	1 (14.3)
GMC	-	-	5.2 (0.5–56.6)	75.7 (0.7–8781)
Injection site
Redness	0 (0)	0 (0)	0 (0)	0 (0)
Swelling	0 (0)	0 (0)	0 (0)	0 (0)
Site warmth	1 (20)	0 (0)	0 (0)	0 (0)
Tenderness	1 (20)	2 (25)	3 (75)	1 (14.3)
Pain with arm movement	2 (40)	2 (25)	2 (50)	1 (14.3)
Systemic
Fever	0 (0)	1 (14.3)	0 (0)	0 (0)
Headache	1 (20)	2 (25)	0 (0)	2 (28.6)
Muscle ache	0 (0)	1 (12.5)	0 (0)	0 (0)
Fatigue	1 (20)	1 (12.5)	0 (0)	1 (14.3)
Nausea	0 (0)	1 (12.5)	0 (0)	1 (14.3)
Vomiting	0 (0)	0 (0)	0 (0)	0 (0)
Diarrhea	1 (20)	0 (0)	0 (0)	1 (14.3)

Adverse events

Both vaccines were well tolerated by participants in both the study and the substudy (Table 4). Injection-site reactions were the most commonly reported adverse event and comprised mostly tenderness at the site of injection and pain on movement of the arm. Most reported adverse events were mild or moderate and resolved within three days. Significantly more injection-site tenderness was reported in HBsAg-1018 recipients (63.2% of the primary study and 81.8% of the substudy) than in recipients of HBsAg-Eng (18.8% and 15.4%, respectively). No differences between the groups were observed in pain on arm movement. Headache, muscle ache, and fatigue were the most common systemic adverse events reported, with no significant differences between the groups.

Table 4. Adverse events reported after immunization with HBsAg-1018 or HBsAg-Eng

Adverse event	Nonresponders to 3 doses (primary study)		Nonresponders to 4 to 6 doses (substudy)
	HBsAg-1018 (n = 19)	HBsAg-Eng (n = 16)	HBsAg-1018 (n = 11)	HBsAg-Eng (n = 13)
Injection site
Redness	0 (0)	0 (0)	0 (0)	1 (7.7)
Swelling	0 (0)	0 (0)	0 (0)	0 (0)
Site warmth	2 (10.5)	1 (6.3)	1 (9.1)	2 (15.4)
Tenderness	12 (63.2)*	3 (18.8)	9 (81.8)**	2 (15.4)
Pain with arm movement	7 (36.8)	3 (18.8)	6 (54.5)	3 (23.1)
Systemic
Fever	0 (0)	0 (0)	0 (0)	1 (7.7)
Headache	1 (5.3)	4 (25)	3 (27.3)	7 (53.8)
Muscle ache	2 (10.5)	1 (6.3)	3 (27.3)	3 (23.1)
Fatigue	2 (10.5)	3 (18.8)	4 (36.4)	3 (23.1)
Nausea	0 (0)	2 (12.5)	1 (9.1)	0 (0)
Vomiting	0 (0)	0 (0)	0 (0)	0 (0)
Diarrhea	1 (5.3)	1 (6.3)	2 (18.2)	1 (7.7)

* p = 0.016; **p = 0.003

In the nonresponders to the first study vaccination who received two additional doses of HBsAg-Eng, similar patterns of injection-site reactions and systemic adverse events were observed, regardless of whether the study vaccine had been HBsAg-1018 or HBsAg-Eng (Table 3), and were similar to those reported after the first dose in the HBsAg-Eng group.

Discussion

Currently available hepatitis B vaccines are highly immunogenic and lead to high rates of seroprotection following a two- or three-dose vaccination series, particularly in children and young adults.^2-4 With increased age, obesity, smoking, and some underlying conditions (e.g., renal failure receiving dialysis), nonresponse rates postvaccination rise substantially.^6-9 In this study, the safety and immunogenicity of an investigational hepatitis B vaccine (HBsAg-1018), comprised of recombinant HBsAg adjuvanted with a phosphorothioate oligodeoxynucleotide, was compared with a commercially available vaccine, HBsAg-Eng. While HBsAg-1018 demonstrated a trend toward increased rates of seroprotection and higher geometric mean antibody in nonresponders to three previous doses of a standard hepatitis B vaccine, these results did not reach statistical significance. In the substudy of nonresponders to four to six doses of licensed hepatitis B vaccine, although a similar proportion of participants receiving HBsAg-1018 and HBsAg-Eng achieved seroprotective levels ≥ 10 mIU/mL, significantly more participants who received HBsAg-1018 achieved seroprotective levels ≥ 100 mIU/mL 4 weeks postvaccination, and those who responded at 4 weeks achieved higher geometric mean antibody levels at 4 and 52 weeks postvaccination than recipients of the licensed HBsAg-Eng vaccine. HBsAg-1018 was well tolerated in this population of vaccine nonresponders, with comparable rates of adverse events except for significantly higher rates of injection-site tenderness.

There have been many attempts to develop strategies to address nonresponse to hepatitis B vaccine. Many nonresponders can achieve seroprotective levels with one to three additional doses of the licensed, alum-adjuvanted vaccine.²⁶ In a randomized, controlled trial, increasing the antigen dose from 20 µg to 40 µg does not improve the rate of seroprotection to an additional dose of vaccine¹⁴ although, in a retrospective analysis, higher antigen content is associated with higher rates of seroprotection in nonresponders.¹⁸ A high-antigen-content vaccine with a novel adjuvant (MPL plus alum; AS04) is also associated with higher seroprotective rates and geometric mean antibody levels in nonresponders.^12,18 Other novel vaccines, including a triple-antigen HBsAg vaccine (S, pre-S1, pre-S2)^13,17 and a DNA vaccine,¹⁶ have demonstrated improved efficacy in nonresponders. Multisite revaccination with HBsAg-Eng vaccine via the intradermal route can also induce seroconversion in some nonresponders.²⁷

Although HBsAg-1018 tends to increase the seroprotective rates and geometric mean antibody levels in nonresponders, a substantial proportion do not respond to a single dose of HBsAg-1018 or several additional doses of HBsAg-Eng. There are many factors that contribute to nonresponsiveness after hepatitis B vaccination. Both low levels of anti-HBs and memory B cells directed toward HBsAg capable of in vitro differentiation into anti-HBs secreting plasma cells poststimulation can be detected in the serum of nonresponders, indicating that the response to vaccination is suboptimal rather than truly absent in this subset of individuals.²⁸ Cytokines are important contributors to the differentiation and function of T and B cells; polymorphisms within genes of cytokines such as interleukin (IL)-4, IL-4RA, and IL-13, and toll-like receptor (TLR)-2 have been shown to correlate with hepatitis B vaccine nonresponsiveness.²⁹ The co-administration of granulocyte–macrophage colony stimulating factor (GM-CSF), a cytokine with a variety of immunostimulatory effects including increased production of IL-2, results in increased rates of seroprotection in hepatitis B vaccine nonresponders.¹⁵ Although decreased levels of IL-2 and interferon gamma have been demonstrated in nonresponders with renal failure,³⁰ co-administration of IL-2 with hepatitis B vaccine was not effective in improving the immune response in these patients.³¹ Immunostimulatory sequences act through TLR-9 and stimulate the innate immune system to produce interferon-gamma and foster a cytokine milieu that promotes a Th1 phenotype.^32-34 Other genetic factors, including HLA haplotype, have been associated with hepatitis B vaccine nonresponsiveness, suggesting that the reasons for nonresponse may be multifactorial and complex.^9,35-39

The major limitation of this study was the difficulty in achieving the intended sample size, resulting in insufficient power to assess the primary outcome adequately. Given the small sample size, the contribution of participant characteristics, particularly gender and smoking status, could not be assessed. An additional limitation was that, as per protocol, nonresponders to HBsAg-Eng or HBsAg-1018 in this study were all given additional doses of HBsAg-Eng. HBsAg-1018 is more immunogenic than licensed, alum-adjuvanted hepatitis B vaccine for the primary immunization of healthy adults.^22-25 Additional doses of HBsAg-1018 may have improved the rates of seroconversion in study participants. While the results from this study are suggestive of increased immunogenicity of HBsAg-1018 among nonresponders to three to six prior doses of alum-adjuvanted recombinant hepatitis B vaccine, the benefits of the use of HBsAg-1018 in nonresponders must await the results of further studies.

Materials and Methods

Vaccines

The investigational vaccine (HBsAg-1018) comprised 3,000 µg 1018 (sequence 5′-TGACTGTGAACGTTCGAGATGA-3′) in combination with 20 µg yeast-derived recombinant HBsAg (manufactured for Dynavax Technologies by Rhein Americana SA, Buenos Aires, Argentina) per 0.5 mL dose. The reference vaccine was 20 µg commercially available, yeast-derived, recombinant HBsAg adsorbed onto 0.5 mg aluminum hydroxide (Engerix^®-B [HBsAg-Eng], GlaxoSmithKline Biologicals, Rixensart, Belgium). A single lot of each vaccine was used for the study.

Study population

Healthy adults 18 to 65 y of age in Halifax, Montreal, and Quebec City, Canada, were eligible to participate in the primary study if they had completed an initial three-dose series of hepatitis B vaccine and had a documented anti-HBs concentration < 10 mIU/mL measured sometime within six months after the third injection. Healthy adults 18 to 65 y of age who had completed 4 to 6 doses of licensed vaccine and had an anti-HBs concentration < 10 mIU/mL measured within six months of the last vaccination were eligible for the substudy. There was no restriction related to the interval since the previous hepatitis B series and enrollment into the study. Exclusions to participation included having a history of hepatitis B infection; being seropositive for HBsAg or antibody to hepatitis B core antigen (anti-HBc); being pregnant or breast-feeding; being unwilling to use effective contraception during the study; having an immunodeficiency, receiving immunosuppressive medication, or having an autoimmune disorder; having received blood products or immunoglobulins in the previous three months; having a sensitivity to any component of any of the vaccines; having ever received an injection of DNA plasmids or oligonucleotides; having any clinically significant acute or debilitating chronic illness; or having received any vaccines in the previous month.

Study design and procedures

The study was a phase 2, observer-blinded, randomized, controlled, multicenter study. The study was approved by the Research Ethics Board of each of the participating centers; participants provided written informed consent prior to any study procedure. Participants were randomly allocated in a 1:1 ratio using a computer-generated randomization list to receive HBsAg-1018 or HBsAg-Eng. Randomization was stratified according to anti-HBs levels at baseline (undetectable [nonresponders] vs. ≥ 2.1 to < 10 mIU/mL [hyporesponders]); separate randomization lists were generated for the study and the substudy.

All participants were given a single dose of vaccine at the first visit (day 0); vaccines were given as a 1.0 mL intramuscular injection into the deltoid muscle with a 1–1½ inch (2 inch in obese participants), 25-gauge needle by a nurse who was not involved in any other aspects of the study. Participants were monitored by study staff for 30 min after the injection for immediate reactions. Adverse events were recorded on a diary card by participants until four weeks after their injection. Subjective events were categorized as mild (aware but easily tolerated), moderate (some interference with activity), and severe (incapacitating, preventing daily activity). Erythema and swelling were defined as mild (≥ 10 mm but < 20 mm), moderate (≥ 20 mm but < 50 mm), and severe (≥ 50 mm). Fever was defined as mild (≥ 38^◦C and ≤ 39.0^◦C), moderate (> 39.0^◦C and ≤ 40.0^◦C), and severe (> 40^◦C).

At four weeks after the first dose of study vaccine was given, anti-HBs concentrations were reviewed in a blinded fashion. Participants who had not attained a seroprotective anti-HBs concentration ≥ 10 mIU/mL were offered two additional injections of HBsAg-Eng. Participants considered seroprotected were given no further injections and were followed for antibody persistence by repeat serological testing at 52 weeks postvaccination. Participants who did not respond to the first study dose and who received additional booster doses were monitored for adverse events and had blood drawn for anti-HBs four weeks after the second and third injections.

Blood was collected by venipuncture at study entry and at four weeks postvaccination. For participants with anti-HBs ≥ 10 mIU/mL at week 4, blood samples were collected at weeks 26 and 52. For participants with anti-HBs < 10 mIU/mL at week 4, blood samples were collected 28 d after the first HBsAg-Eng booster injection, 28 d after the second HBsAg-Eng booster injection, and at week 52. Laboratory testing of sera collected for anti-HBs concentrations at specified time points was conducted at Dynavax Technologies using a multiparticle enzyme immunoassay (AxSYM AUSAB^®, Abbott Laboratories, Abbott Park, IL, USA) and reported as mIU/mL. In addition to anti-HBs levels, hematological testing (complete blood count), biochemical testing (electrolytes, blood urea nitrogen and creatinine, and hepatic enzymes), and urinalysis were tested at baseline and 4 weeks postvaccination, and antinuclear antibodies and anti-single-stranded and anti-double-stranded DNA antibodies were tested at baseline, and at 4, 26, and 52 weeks after the first study vaccination.

Data analysis and statistical considerations

The primary outcome for the study was the proportion of participants in each group exhibiting seroprotection four weeks after a single injection of HBsAg-1018 or HBsAg-Eng (seroprotection rate [SPR] defined as the proportion of participants with an anti-HBs concentration ≥ 10 mIU/mL). For participants achieving seroprotection four weeks after a single injection, the secondary outcomes included the proportion of participants with seroprotection at 26 and 52 weeks, the proportion of participants with anti-HBs ≥ 100 mIU/mL at 4, 26, and 52 weeks, and the geometric mean antibody concentration at 4, 26, and 52 weeks postvaccination. The safety outcomes were evaluated by the proportion of participants with adverse events and laboratory abnormalities postvaccination.

The analysis population for immunogenicity and safety was the intention-to-treat population defined as participants who received at least one study injection. The within-center analysis of proportions was accomplished by construction of binomial estimates and exact binomial confidence intervals for each group at each visit and differences between groups were assessed using Fisher’s exact tests. The multicenter analysis consisted of Mantel-Haenszel point and interval estimates of relative risk and test of treatment differences using the Mantel-Haenszel test of unit relative risk. The main effect of center and interaction of center and treatment were assessed using logistic regression. Anti-HBs concentrations were log transformed prior to analysis to better approximate a normal distribution. For within-center analysis, geometric means and 95% confidence intervals were estimated for each treatment group at each visit. A multicenter analysis was performed using an analysis of variance (ANOVA), assessing the difference between treatment groups, centers, and the interaction of center and treatment group. Statistical tests were performed at the 0.05 significance level; no adjustments were made for multiple comparisons. All data analyses were performed using Statistical Analysis Systems (SAS Version 8.2 or later, SAS Institute, Cary, NC, USA). A sample size of 50 per group in the primary study was calculated with the assumption of an 80% SPR in the HBsAg-1018 group and 50% SPR in the HBsAg-Eng group, a power of 90%, and a dropout of 15%.

Disclosure of Potential Conflicts of Interest

J.T.M. and W.L.H. are employees of Dynavax Technologies Corporation. All other authors have no financial interest in the vaccine or its manufacturer but received research funding to undertake the study.

The study was funded by Dynavax Technologies Corporation.

Acknowledgments

We thank the participants of the study and the research staff at the three enrollment sites for their careful attention to detail. This study was funded by Dynavax Technologies Corporation.