Cutaneous Delivery of a Live, Attenuated Chimeric Flavivirus Vaccine ...

[Human Vaccines 1:3, 106-111; May/June 2005]; ©2005 Landes Bioscience

Cutaneous Delivery of a Live, Attenuated Chimeric Flavivirus Vaccine Against Japanese Encephalitis (ChimeriVax™-JE) in Non-Human Primates Research Paper

Cheryl H. Dean1 Jason B. Alarcon1 Andrea M. Waterston1 Ken Draper2 Richard Early2 Farshad Guirakhoo3 Thomas P. Monath3 John A. Mikszta1,*

ABSTRACT

.

UT E

RIB

IST

*Correspondence to: John A. Mikszta; BD Technologies; 21 Davis Drive; Research Triangle Park; North Carolina USA; Tel.: 919.597.6158; Fax: 919.597.6402; Email: john—[email protected]

SC

KEY WORDS

BIO

microneedles, Japanese encephalitis, vaccine, flavivirus, intradermal, epidermal, cutaneous

ES

ABBREVIATIONS

20

05

LA

ND

Japanese encephalitis Langerhans cell dendritic cell antigen presenting cell peripheral blood mononuclear cell French neurotropic virus yellow fever microenhancer array subcutaneous intradermal intramuscular intracerebral trans-epidermal water loss draining lymph node plaque forming unit neutralizing antibody

© 106

Skin is an attractive target for vaccination due to the abundance of potent antigen presenting cells (APC) including epidermal Langerhans cells (LC) and other myeloid dendritic cell (DC) populations. Comprising only 1–3% of epidermal cells, LC form a mesh-like structure covering nearly 20% of skin surface area. This feature provides an efficient method for capture and presentation of antigens to the immune system following natural infection or vaccination.1,2 The stratum corneum is the outermost skin layer, composed primarily of lipids and dead cells. Only 10–20 µm thick in humans, this layer serves as an effective physical and chemical penetration barrier.1 Targeting skin APC for vaccination requires disrupting or bypassing the stratum corneum. We previously developed prototype microenhancer array (MEA) devices intended for mild abrasion of the stratum corneum, thus enabling topical vaccine application to the epidermis.3 These devices were shown to disrupt skin barrier function in humans and enable topical vaccination in mice and rabbits.3,4 In addition, we have developed stainless steel microneedles intended for shallow intradermal (ID) delivery and have demonstrated the utility of these devices in providing protective immunity against lethal inhalational anthrax in a rabbit model.4 These minimally-invasive delivery platforms are ideal candidates for administration of live, attenuated flavivirus vaccines and may offer substantial benefits, especially for developing world applications. The natural route of flaviviral infection is the skin with transmission occurring through the bite of an infected mosquito or tick.5,6 Immature DCs are permissive for infection by arthropod-borne viruses.7-10 Following infection, immature DCs become activated, rapidly migrate to draining lymph nodes (DLN) and change from an antigen-processing to an antigen-presenting mature DC phenotype.5,6,8,11 Interestingly, inactivated virus does not efficiently induce such changes. Recently, CD209 (DC-SIGN) has been proposed as a cell surface receptor for flavivirus infection of DCs.12,13 Arthropod-borne viruses do not replicate widely or cause local skin lesions or inflammation at the site of inoculation. Rather, infected and activated DCs migrate to DLN, where replication and antigen presentation occur. Although DCs have been implicated in natural flavivirus infection, there has been little effort in developing vaccination strategies targeting skin APC. All currently licensed flavivirus vaccines, as well as those in various stages of preclinical and clinical trials, bypass

IEN

Previously published online as a Human Vaccines E-publication: http://www.landesbioscience.com/journals/vaccines/abstract.php?id=1797

INTRODUCTION

CE

Received 01/20/05; Accepted 05/09/05

OT D

Sparks, Nevada USA

3Acambis; Cambridge, Massachusetts USA

ON

2Charles River Laboratories; Discovery and Development Services; Sierra Division;

.D

1BD Technologies; Research Triangle Park; North Carlina USA

JE LC DC APC PBMC FNV YF MEA SC ID IM IC TEWL DLN pfu Nab

Flaviviral diseases such as yellow fever, Japanese encephalitis (JE) and dengue hemorrhagic fever cause enormous morbidity and mortality worldwide. There is an urgent need for alternative technologies for mass vaccination against these and other diseases, particularly in the developing world. Here, we administered a live attenuated, chimeric JE vaccine (ChimeriVax™-JE) to nonhuman primates by skin microabrasion and intradermal delivery using microneedles. Both cutaneous delivery methods induced mild viremia similar in magnitude to that observed following subcutaneous (SC) injection. The duration of viremia induced by cutaneous delivery (5–7 days), however, was substantially longer than via SC (0–3 days). In addition, mean neutralizing antibody titers in cutaneous delivery groups were up to 7-fold greater than via SC injection. There were no safety issues identified and both cutaneous delivery methods appeared to be well tolerated. Thus, cutaneous delivery may represent a minimally-invasive alternative approach for flavivirus vaccines that more closely resembles the natural route of viral infection.

Human Vaccines

2005; Vol. 1 Issue 3

Cutaneous Delivery of a Live, Attenuated Chimeric Flavivirus Vaccine Against Japanese Encephalitis (ChimeriVaxTM-JE) in Non-Human Primates

skin and are delivered to deeper subcutaneous (SC) Table 1 Study design or intramuscular (IM) tisGroup # # of monkeys Delivery Method Vaccine Dose Level Dose Volume sues.14-16 Precedent exists, (Male/Female) (log 10 PFU a ) (mL) however, for cutaneous delivery of flavivirus vac1 2/1 SC ChimeriVaxTM-JE Vaccine 4.4 0.1 cines. In 1939, an attenu2 1/2 ID, Microneedle ated mouse brain yellow 3 2/1 MEA, Preabrasion fever (YF) vaccine virus 4 1/2 MEA, Abrasion through suspension applied by vaccine scarification to rhesus monkey skin produced viremia aPFU = plaque-forming units. with a resultant immunity to YF that developed in MEA devices used in this study were fabricated from a plastic material the same timeframe as a SC injection.17,18 Although the device and methodology were crude, this approach was successfully used to vac- and consisted of a 1cm2 microabrading surface (approximately 150 µm 4 cinate over 56 million people in Africa through the 1950s.14,18-20 microprojection height) mounted to a hand-held applicator. Approximately These initial campaigns were carried out using the French three weeks before the study began, the number of passes of the device Neurotropic YF Virus (FNV). Upon introduction of the safer YF required for significant skin barrier disruption on the deltoid/upper arm of the monkey was determined by trans-epidermal water loss (TEWL) mea17D strain, scarification was ultimately abandoned in favor of SC surements27 on three animals as described previously.3 For immunization, injection. two methods of MEA-based delivery were evaluated: (1) Preabrasion: shaved ChimeriVax™ is a live, attenuated genetically engineered virus skin was treated with six passes of the device across each of two premarked prepared by replacing the premembrane (prM) and envelope (E) 1 x 2 cm sites. Vaccine (0.05 ml per site) was applied topically across each genes of YF 17D with the corresponding genes of JE or other pretreated site and left undisturbed and uncovered for at least 30 min to air flaviviruses.21,22 Extensive preclinical studies have tested safety and dry. (2) Abrasion through the vaccine inoculum: Vaccine (0.05 ml per site x demonstrated the immunogenicity and protective efficacy of the two sites, as above) was first applied to shaved skin, then the device was vaccine.23 ChimeriVax™-JE was shown to be fully attenuated in placed in contact with the vaccine droplet and abraded through the inoculum mice inoculated intra-cerebrally (IC)21,24 and to be significantly less with six passes and left to air dry as above. Immunization sites from all groups were visually inspected immediately neurotropic than commercial YF 17D vaccine after IC inoculation in monkeys.24,25 The vaccine is currently in Phase II clinical trials, after treatment and at 30 min, 24 hr and 48 hr post-delivery. Blood was with results to date suggesting that the vaccine is well-tolerated and collected under ketamine/domitor anesthesia before immunization on day highly immunogenic in humans when inoculated by the SC route.26 (d) 1, on d 2–11 for determination of viremia and on d 31 and d 61 for evaluation of neutralizing antibodies. Food consumption, body weight, and Furthermore, extensive preclinical and clinical trials have shown that activity levels were monitored daily. preexisting immunity to the YF vector does not negatively influence Vaccine. ChimeriVax™-JE, a live attenuated JE vaccine, has been the antibody response to the vector-encoded JE antigens.23,25,26 described previously.21,23-26 In brief, the vaccine was constructed from a full Here, we investigated cutaneous delivery of ChimeriVax™-JE in length cDNA clone of YF 17D with genes encoding prM and E proteins of Cynomolgus monkeys (Macaca fascicularis) using MEA and YF 17D replaced with the corresponding genes of JE (vaccine strain microneedle-based delivery platforms. Both cutaneous delivery SA14-14-2) virus and was propagated in Vero cells.21,22 The vaccine was approaches induced low level viremia and neutralizing antibody produced and tested in accordance with current good manufacturing pracresponses comparable to or greater in magnitude than via SC injec- tices (GMP). The lot (Lot No. 00C02) used for this study was a frozen liquid tion using standard needles. Notably, the method employed for formulation, containing the vaccine virus in a suspension of Minimal MEA-based delivery (“preabrasion” vs. “abrasion through the vaccine Essential Medium (MEM; Gibco, Grand Island, NY), 2.5% lactose USP and 6 inoculum”) appeared to impact the extent of viral uptake and resulting 7.5% human serum albumin USP. The undiluted vaccine contained 1.3 x 10 pfu/ml. All delivery groups were immunized with 0.1 ml of vaccine adminimmune response.

MATERIALS AND METHODS

Animals and immunizations. Cynomolgus monkeys (Macaca fascicularis) were housed at Sierra Division, Charles River Laboratories, Sparks, NV. Animal care was in accordance with the USDA Animal Welfare Act (9 CFR, parts 1,2 and 3) and followed the conditions specified in the Guide for the Care and Use of Laboratory Animals (ILAR publication, 1996, National Academy Press). A total of 12 flavivirus seronegative monkeys, six female and six male, ranging from 1.2 to 5.3 years of age and approximately 1.2 to 4.3 kg weight were randomized into 4 treatment groups, as outlined in Table 1. For vaccine administration, monkeys were anesthetized with ketamine/domitor and dosing sites were shaved with electric clippers immediately before immunization and marked. ChimeriVax™-JE (1.3 x 106 plaque forming units (pfu)/ml) was administered across two sites (0.05 ml each) on the deltoid/upper arm. SC dosing was performed using a 30 G needle and 1cc syringe (BD, Franklin Lakes, NJ). ID delivery was accomplished using a 1 mm long stainless steel microneedle4 fitted to a 1cc syringe and inserted perpendicularly to the skin surface to control delivery depth.

www.landesbioscience.com

istered to two separate sites of 0.05 ml each for a dose of 4.4 log10 pfu. Viremia determination. Virus titers in serum were determined by plaque assay on Vero cell monolayers.24,25 Undiluted serum samples were plated in duplicate on Vero cells growing in 6 well tissue culture plates (Falcon, BD Labware, Franklin Lakes, NJ). Following a 1 hr adsorption at 37˚C, cells were overlaid with a 0.4% nutrient agarose containing 5% fetal calf serum (Hyclone, Logan, UT), 1% HEPES (Sigma, St. Louis, MO) and 1% glutamine (Sigma) in MEM and incubated for 4 d at 37˚C. A second overlay containing 3% neutral red (Sigma) was added and plates were incubated for an additional 24 hr at 37˚C before plaques were counted. Viremia results between groups were compared by ANOVA assuming unequal variances among treatment groups with Satterthwaite’s approximation for degrees of freedom. Plaque neutralization assay. Neutralizing antibody titers were determined by the serum-dilution constant-virus plaque-reduction neutralization test in Vero cells.24,25 Serum samples were heat inactivated at 56˚C for 30 min. Serial 2-fold dilutions of sample were made in MEM containing 50% fetal calf serum. Diluted samples were mixed in 96-well plates with an equal volume of virus suspension and incubated overnight at 4˚C. Vero cell monolayers

Human Vaccines

107


was 3.3 ± 0.6 g/m2/hr (Fig. 1). The low baseline TEWL indicated that shaving alone did not disrupt the skin barrier to a significant extent. There was a substantial rise in TEWL that increased with number of passes across the site. After six passes, mean TEWL was 21.7 ± 5.9g/m2/hr, representing a 4- to 12-fold increase over pretreatment levels across individual animals (Fig. 1). These results are similar to those observed previously in humans.3 The treatment process was associated with a very mild and transient erythema that completely resolved within 24 hr. To examine the feasibility of cutaneous delivery of ChimeriVax™-JE, monkeys were dosed with 4.4 log10 pfu of vaccine by MEA, microneedle, or SC injection using a standard needle. As detailed in the Materials and Methods, two protocols for MEA-based delivery were investigated: (1) a “preabrasion” protocol whereby the skin was treated with the device before topical application of the vaccine and (2) “abrasion through the vaccine inoculum” whereby the vaccine was first applied to skin, followed by abrasion. Based on the TEWL results (Fig. 1), 6 passes of the MEA were employed. Photographs of skin dosing sites are displayed in Figure 2. Microneedlebased ID delivery resulted in a raised wheal, typical of a shallow ID injection, but was not associated with erythema or edema (Fig. 2A). Skin microabrasion by MEA was also well-tolerated. The only visible sign of treatment was slight Figure 1. Average TEWL as a function of the number of passes of the MEA erythema that resolved within 24 hr (Fig. 2B and C). SC injection resulted device across the shaved skin of the deltoid/upper arm region of three in an occasional drop of blood at the injection site (Fig. 2D), but otherwise monkeys. Data represent mean TEWL value (g/m2/hr) ± SEM. was well-tolerated. There were no differences in daily food consumption, activity levels or bodyweight between treatment groups during the 30 d post-dose observation period (data not shown). SC injection induced a lowlevel, transient viremia in one of three monkeys (Table 2). The peak magnitude in this animal was 1.8 log10 pfu/ml with a duration of 3 d. In contrast, MEAbased delivery by abrading through the vaccine inoculum resulted in viremia in all three monkeys. While the peak level in these animals (ranging from 1.8–2.3 log10 pfu/ml) was similar to the viremic monkey in the SC group, the duration was significantly longer (5–7 d; p = 0.02) and the onset delayed by 2–3 d. Notably, only one of three monkeys in the preabrasion group was viremic, with peak magnitude (1.7 log10 pfu/ml) similar to SC, but longer duration (7 d) and slightly delayed onset. ID injection by microneedle also induced viremia in three of Figure 2. Photographs of skin dosing sites. (A) ID injection with microneedle. (B) Delivery by MEA abrasion through three animals (peak ranging from the vaccine inoculum. (C) Delivery by MEA preabrasion. (D) SC injection. Photographs were taken within 15 minutes 1.7–2.3 log10 pfu/ml), with duration generally longer than via SC after dosing. inoculation (Table 2). The onset time was similar to SC in two of were inoculated in duplicate with 0.1 ml of the serum/virus suspension, and three ID-injected animals, while the third monkey from this group displayed processed as above. The neutralizing antibody titer was defined as the highest a delayed onset of viremia similar to that observed in the MEA group. dilution giving a 50% reduction in plaques relative to the plaque titer Neutralizing antibody (Nab) titers against the vaccine virus strain were obtained from the same monkey prior to immunization. measured on d 31 and d 61 post-immunization. By d 31, all but one monkey across the groups had sero-converted (Table 3). Nab titers for monkeys in the SC group ranged from 640 to 1,280 consistent with previous studies.24,25 RESULTS Individual animals immunized ID using microneedles responded with We previously showed that MEA devices provide significant skin barrier similar or greater titers, ranging from 1,280 to 10,240. Likewise, MEAdisruption in humans, as determined by TEWL measurements.3 We used based delivery by abrading through the inoculum induced elevated Nab this same approach to examine the effects of the device on monkey skin. The titers in three of three animals with titers ranging from 160 to 10,240. The mean baseline TEWL on the deltoid/upper arm region following shaving Nab response was more variable in the MEA preabrasion group, with one

108

Human Vaccines

2005; Vol. 1 Issue 3


Table 2

Viremia results

Monkey

Delivery Method

F21555M

SC

2**

3

Viremia (log 4 5

10

pfu/ml) by day post immunization 6 7 8 9

10

11