Escolar Documentos
Profissional Documentos
Cultura Documentos
Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials
a r t i c l e i n f o a b s t r a c t
Article history: Alternatives to syringe-based administration are considered for vaccines. Intradermal vaccination with
Received 29 June 2017 dissolvable microneedle arrays (MNA) appears promising in this respect, as an easy-to-use and painless
Received in revised form method. In this work, we have developed an MNA patch (MNAP) made of hydroxyethyl starch (HES) and
16 August 2017
chondroitin sulphate (CS). In swines, hepatitis B surface antigen (HBsAg) formulated with the saponin
Accepted 25 August 2017
Available online 29 August 2017
QS-21 as adjuvant, both incorporated in HES-based MNAP, demonstrated the same level of immuno-
genicity as a commercially available aluminum-adjuvanted HBsAg vaccine, after two immunizations 28
days apart. MNAP application was associated with transient skin reactions (erythema, lump, scab),
Keywords:
Hydroxyethyl starch
particularly evident when the antigen was delivered with the adjuvant. The thermostability of the
Microneedle adjuvanted antigen when incorporated in the HES-based matrix was also assessed by storing MNAP at 37,
Hepatitis B antigen 45 or 50 C for up to 6 months. We could demonstrate that antigenicity was retained at 37 and 45 C and
QS-21 only a 10% loss was observed after 6 months at 50 C. Our results are supportive of MNAP as an attractive
Thermostability alternative to classical syringe-based vaccination.
Vaccine 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction that can be encountered, the costs associated with vaccine storage
and transportation concur for a signicant part of the global im-
Vaccination has greatly helped to reduce disease, disability and munization costs. So methods to produce thermostable vaccines
death worldwide. Nevertheless, access to vaccination may be are needed, as they would contribute to render vaccination more
limited for some population groups, as most vaccines need to be affordable for developing countries by decreasing these costs.
injected under the skin or deep into the muscle using needle-based In this context, microneedle array patch (MNAP) has been
syringes, which require skilled health care professionals for the gaining attention as an alternative vaccine delivery method [3e6].
administration, hence increasing the costs associated with the Visually MNAP is free from any needle and it is said to be painless as
immunization. In addition, needle fear is a reality for many people, it does not stimulate nerve ends that are associated with pain [7].
which may cause them to avoid vaccination [1,2]. New alternative Both elements might be an asset for patients who are reluctant to
vaccination strategies that are less invasive might allow over- injection. In addition, MNAP is user-friendly, as it is ready-to-use
coming the acceptance and access issues associated with injectable and does not require reconstitution, implying that it may be
vaccinations. administered by healthcare personnel with only minimal medical
Another issue limiting vaccine access is the fact that most vac- training. Lastly, the use of microneedles has been shown to allow
cines need to be kept and transported within a specic refrigeration antigen dose sparing while retaining equivalent degree of immune
temperature range (usually 2e8 C). Beyond the logistic problems responses as classical intramuscular injections in preclinical models
[8,9] and in humans [10].
* Corresponding author. GSK, Rue de lInstitut 89, 1330, Rixensart, Belgium. First generation MNAs were made of silicon or metal. The an-
E-mail address: JIM.J.JANIMAK@GSK.COM (J. Janimak). tigen in solution could be incorporated in hollow microneedles or
1
Current address: PAIRimmune, Quebec, Canada.
2
Current address: MSD Belgium, Brussels, Belgium.
be coated on them [3,5,6]. However, this type of MNA may break off
http://dx.doi.org/10.1016/j.biomaterials.2017.08.038
0142-9612/ 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
D. Poirier et al. / Biomaterials 145 (2017) 256e265 257
dissolvable MNA. These microneedles are made of hydroxyethyl Groups Prime Boost
starch, a polysaccharide already used in the pharmaceutical in- 1 Ctrl Ctrl
dustry [11]. Due to the solubility of their components, such needles 2 Ctrl QS-21-MNAP
are expected to resorb into the skin within minutes, releasing the 3 Ctrl Non-adjuvanted MNAP
vaccine components and bringing them in contact with the cuta- 4 QS-21-MNAP QS-21-MNAP
5 QS-21-MNAP None
neous immune system. In addition, antigens in solid MNA are
known to be more thermostable as compared with antigens in Ctrl, control vaccine, injected i.m; MNAP, microneedle array patch.
liquid vaccines [3], hence MNAP represent a possibility to store and
transport vaccines without the constraints and the costs of main-
taining the cold chain. approximately 60 times using Pellicon XL 50 cm3 Biomax 50KD PES
The present work describes the production of dissolvable MNA kit (Merck Millipore). The term QS-21 used throughout this
containing hepatitis B surface antigen (HBsAg). Its capacity to manuscript refers to QS-21 formulated in liposomes.
induce immune responses, with or without adjuvant was evaluated
in swines, whose skin is similar to that of humans in terms of 2.2. Production of MNAP
thickness, structure and immune function [12e14]. In addition,
antigen stability in our MNAP was also investigated. MNAPs were produced in an aseptic isolator of Grade A in a GMP
pilot facility. First, hydroxyethyl starch 70000 (HES, Fresenius Kabi)
was dissolved in distilled water and mixed with HBsAg (4:1 w/w).
2. Materials and methods Alternatively, QS-21 was incorporated into the mixture as an
adjuvant. In the latter case, HES and sucrose (Wako Chemicals)
2.1. Antigen and adjuvant were dissolved in distilled water and mixed with HBsAg and QS-21
(HES:sucrose:QS-21:HBsAg 18:11:6:1 w/w). In both cases, the
HBsAg (GSK Vaccines) solution was concentrated using ultra- aqueous solution was sterilized through a 0.22-mm hydrophilic
ltration techniques involving Pellicon 3 with 88 cm3 Biomax 10KD polyethersulfone membrane (Millex SLGP033RS, Millipore) and
polyethersulfone (PES) kit (Merck Millipore) to reach a nal con- casted into thermosetting resin-based micromolds to shape the
centration of HBsAg 60 mg/ml, as determined by the Lowry microneedles. The same solution weight was distributed in each
method [15]. microneedle. The procedure yielded MNAP comprising 100
QS-21 (Quillaja saponaria Molina, fraction 21; licensed by GSK microneedles, each approximately 600 mm long. After drying at
from Antigenics LLC, a wholly owned subsidiary of Agenus Inc., a 23 C for approximately 3.5 h, a polyethylene terephthalate (PET)
Delaware, USA corporation) is an immuno-enhancer formulated in base lm coated with an aqueous solution containing chondroitin
1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) liposomes con- sulphate (CS, Maruha Nichiro), was attached to the micromolds and
taining cholesterol at 100 mg/ml. QS-21 was further concentrated dried at 35 C for 19e24 h. After drying, the PET base lm bound to
the MNAP was easily separated from the micromolds. The shape
Table 1 and length of the microneedles was conrmed under a digital
Immunization groups in the study using MNA containing non-adjuvanted HBsAg. microscope (VHX-5000, Keyence). MNAP was placed in a plastic
case, packed with desiccant into a sealed aluminium foil bag and
Groups Prime Boost
stored at 4 C until use. Final water content of MNAP was <5% (w/
1 Ctrl None w). The whole procedure is depicted in Fig. 1.
2 Ctrl Ctrl
3 Ctrl MNAP
4 MNAP MNAP 2.3. HBsAg distribution in MNAP
5 MNAP None
Ctrl, control vaccine, injected i.m; MNAP, microneedle array patch. HBsAg localization in the microneedles was evaluated after
258 D. Poirier et al. / Biomaterials 145 (2017) 256e265
Table 3
Scoring criteria for evaluation of skin reactions.
cutting at different levels by using a needle cutter (Fujilm pro- Assessment and Accreditation of Laboratory Animal Care (AAALAC)
prietary technology). The two different heights of cut, at 200 and recommendations and guidelines.
300 mm were dened after observation of the remaining micro- Twenty-eight-day-old female Yorkshire/Landrace F1(Y x L) x
needle length in pilot experiments. After cut, microneedle parts Duroc swines were obtained from a porcine reproductive and res-
were dissolved in phosphate-buffered saline (PBS) containing 0.1% piratory syndrome virus (PRRSV)-seronegative farm. The animals
polysorbate 20 (PBS-T). HBsAg content was quantied by a were pre-selected one week earlier, based on their health status
competitive enzyme-linked immunosorbent assay (ELISA) as and appearance to have homogeneity of the swine population. For
follows: all studies, animals were kept group-housed with 12 h light/dark
ELISA plates were rst coated with HBsAg overnight at 4 C and cycle and fed twice a day. They were four-month-old and weighed
then blocked by adding PBS containing 1% bovine serum albumin approximately 35 kg at the start of the study.
(Sigma-Aldrich) for 30 min at 37 C. Concomitantly, dissolved Before immunization, the animals (n 6/group) were pre-
MNAP containing HBsAg was mixed with a known amount of anti- anaesthetized by intramuscular (i.m.) injection of ketamine/xyla-
hepatitis B human antibody (Nabi-HB, Biotest Pharma) and incu- zine (20 mg/kg: 2 mg/kg) in the neck region, and maintained under
bated overnight at 25 C with shaking. Aggregated HBsAg/antibody isourane (1e3%). Atropine (0.5 mg/kg) was given in prevention for
complex was sedimented by centrifugation and the supernatant tachycardia. The ank was rst carefully shaved using electric
containing non-aggregated antibodies was transferred to the plates clippers, followed by an electric shaver and a blade shaver in order
and incubated for 2 h at 37 C. Thus, the more HBsAg in the dis- to get an injection site free of any hair residues. Thereafter, the
solved MNAP, the less anti-hepatitis B human antibody left in the injection site was carefully wiped with warm water to clean the
supernatant. Horseradish peroxidase (HRP)-conjugated goat anti- skin before MNAP application. MNAP was administrated on the
human IgG (Sigma-Aldrich) was added to the plate and incubated shaved skin with a kinetic energy of 0.5 J using handheld applicator
for 1 h at 37 C. Between each process, plates were washed 6 times (Fujilm proprietary technology) and held rmly in place for
with saline containing 0.1% polysorbate 20. Bound antibodies were 10 min with a force of 5 N. A second immunization was made four
revealed by the addition of the peroxidase substrate tetrame- weeks later. As control, Engerix-B (GSK Vaccines), which contains
thylbenzidine (TMB) for 20 min at room temperature in the dark. 20 mg HBsAg/0.5 mL and Al(OH)3 adjuvant, was administered by
The reaction was stopped by adding 1 N H2SO4, and the optical intramuscular injection in the gluteus supercialis muscle (this
density was measured at 450 nm and 620 nm. The intensity of the vaccine when administered intramuscularly will be called control
reaction was inversely proportional to the original amount of vaccine [Ctrl] throughout the text). Tables 1 and 2 summarize the
HBsAg in the MNAP. different groups used in the studies.
An experiment was also performed to determine the adequate
2.4. In vivo study dose of QS-21. In this experiment, 20 mg HBsAg in solution was
intradermally injected together with 1.5 mg, 5 mg, 15 mg or 50 mg QS-
2.4.1. Immunization 21 in 0.1 mL using a ne hypodermic needle with the bevel facing
All animal studies were approved by the Institut National de la upwards to mimic a Mantoux injection.
Recherche Scientique e Institut Armand Frappier (INRS-IAF) An-
imal Care and Use Committee and were compliant with the Cana- 2.4.2. Determination of humoral responses
dian Council on Animal Care (CCAC) and the Association for Blood samples were collected 28 days after the boost. Serum
Fig. 2. A. Microphotograph of a whole microneedle array. B. View of porcine skin after insertion and removal of microneedles, showing delivery of a blue dye. C. Details of a
microneedle.
D. Poirier et al. / Biomaterials 145 (2017) 256e265 259
Fig. 3. Distribution of HBsAg inside the microneedles. A. The needles were laser cut at different levels, dening different regions. The height of the cuts is shown with deviations. B.
The different regions were dissolved in phosphate-buffered saline and their HBsAg content was quantied by ELISA. ND not determined.
Fig. 4. Microphotographs of needles before and after administration in swine. In example 1, the microneedle was not entirely dissolved, as 200 mm remained, which means that 80%
of HBsAg content was administered. In example 2, the entire microneedle was dissolved, which means that 100% of the HBsAg content was administered.
antigen-specic IgG were quantied by ELISA. Briey, Maxisorp 96- protected from light. The colorimetric reaction was stopped with
well at-bottom plates (Nunc cat: 439454) were coated with 1 mg/ H2SO4 1N. Results were read using a microplate reader at 450 nm
ml HBsAg (GSK Vaccines) in carbonate-bicarbonate buffer 50 mM and concentrations were extrapolated from the standard curve
pH 9.6 for 2 h at 37 C, For the reference curve, goat anti-swine IgG following a 4 parameters equation with the following criteria: % CV
capture antibody (1/100; Bethyl cat: A100-104A) was used for below 20% between a minimum of 3 calculated concentrations.
coating. Plates were washed and aspecic sites were blocked with
Tris-buffered saline containing 0.05% polysorbate 20 and 1% bovine 2.4.3. Conrmation of MNA dissolution
serum albumin for 1 h at room temperature. Two-fold serial di- Microneedle resorption was conrmed by measuring needle
lutions of reference serum (Bethyl cat: RS10- 107) diluted 1/8250 or length on microscopic images taken by a digital microscope (VHX-
samples (pre-diluted if needed) were added for a 1 h incubation at 5000, Keyence) after administration. A total of 48 MNAPs were
room temperature. After washing, HRP-conjugated anti-swine IgG administered, 24 without adjuvant and 24 with adjuvant.
(Bethyl cat: A100-104P) diluted 1/75000 in blocking solution were The percentage of antigen delivered to an animal was estimated
dispensed for a 30 min incubation at 37 C. Bound antibodies were based on the remaining length of each microneedle and the known
revealed by addition of TMB for 30 min at room temperature HBsAg distribution in MNAP.
260 D. Poirier et al. / Biomaterials 145 (2017) 256e265
Table 4
Remaining needle length after administration (MNAP containing non-adjuvanted
HBsAg).
The aim of each analysis was to evaluate whether the new 3.1. Shape of MNAP and HBsAg distribution
Fig. 6. Skin reactions over time after administration of MNAP with non-adjuvanted HBsAg. A. Skin reactions after one MNAP immunization (priming). B. Skin reactions after the
second immunization done with MNAP. Pre indicates the reactions before immunization. Day 0 refers to the reactions measured 2 h after administration. Ctrl refers to the control
vaccine. The same color code as in Fig. 5 has been used for the groups. One dot represents one animal. (For interpretation of the references to colour in this gure legend, the reader
is referred to the web version of this article.)
every needle was measured, as shown in Fig. 4. started to be visible in the majority of animals 2 h later. One day
When non-adjuvanted, more than 15 mg of HBsAg was admin- after immunization with MNAP, all animals showed maximum
istered (except in 1 case out of 24 administrations) (Table 4). When grade 2 erythema. No lump and no scab were detected. After 3 days,
formulated with QS-21 more than 17 mg of HBsAg was administered no skin reactions remained visible.
in all cases (Table 5). Skin reactions were also evaluated after MNAP containing non-
adjuvanted HBsAg was used as second immunization (boost;
3.3. Immunization with MNAP Fig. 6B). When priming was done with MNAP, the strongest ery-
thema reactions after MNAP boost were grade 1 and it took more
3.3.1. MNAP containing non-adjuvanted HBsAg than 5 days for them to disappear, which was a bit longer than what
was observed after priming. When priming was done with Ctrl and
3.3.1.1. MNAP-induced humoral responses. We evaluated whether
booster with non-adjuvanted MNAP erythema grade looked higher
MNAP administration of the antigen could be efcient without
and lasted longer than after two administrations with MNAP. Some
adjuvant, due to its unique way of administration and direct contact
slight lump was observed in both groups (Fig. 6B). Scab was not
with the skin immune cells. To this aim, serum IgG level was
observed.
measured 28 days after boost immunization (Fig. 5).
We observed, by comparing group 5 with group 1, that priming
with MNAP-administered non-adjuvanted HBsAg was not as 3.3.2. MNAP containing QS-21-adjuvanted HBsAg
effective as priming with Ctrl. A boost with MNAP after priming The results described in the former section tended to indicate
with MNAP (Group 4) did not induce humoral immunity either. that MNAP administration could not replace the use of an adjuvant
Further, administering a non-adjuvanted MNAP boost after priming for priming. This prompted us to evaluate MNAP priming with
with Ctrl was not found non-inferior to the benchmark (twice Ctrl) adjuvanted antigen.
(Group 3 versus Group 2). It was not found inferior either, probably
due to a lack of power of the analysis. The ratio between these two 3.3.2.1. Selection of QS-21 dose. QS-21 was selected as adjuvant and
groups was 2.58 (95% CI: 0.42e15.85). the rst step was to determine the optimal dose of QS-21 able to
elicit similar anti-HBsAg antibody levels as those induced by Ctrl
3.3.1.2. MNAP-induced reactogenicity in skin. After a single MNAP which was needed to further compare MNA-induced with Ctrl-
application of non-adjuvanted HBsAg (priming; Fig. 6A), erythema induced antibody levels. For that, HBsAg in solution was
262 D. Poirier et al. / Biomaterials 145 (2017) 256e265
Fig. 7. Selection of QS-21 dose. A. IgG levels on Day 56 after intradermal immunization with HBsAg in solution, alone or adjuvanted with escalating doses of QS-21 (from 1.5 mg to
50 mg, as indicated under the graph). One dot represents one animal, and geomeans with 95% CI are shown. Statistical analyses evaluated the non-inferiority of the groups compared
with group 6 (benchmark: Control vaccine [Ctrl] prime and boost). An equal sign indicates a group that is non-inferior to the benchmark. B. Skin reactions over time after the second
intradermal immunization with HBsAg adjuvanted with escalating doses of QS-21. Pre indicates the reactions before immunization. Day 0 refers to the reactions measured 2 h
after immunization. In A and B the same color code has been used for the groups. (For interpretation of the references to colour in this gure legend, the reader is referred to the
web version of this article.)
intradermally injected in the presence of different doses of QS-21. equivalent humoral responses, which prompted us to choose 15 mg
Antibody levels and skin reactogenicity were evaluated to deter- QS-21 for further experiments.
mine the adequate dose of QS-21 to be used in MNA (Fig. 7).
It was observed that 1.5 and 5 mg QS-21 were not sufcient to
3.3.2.2. MNAP-induced humoral responses. Serum IgG level was
obtain anti-HBsAg antibody levels similar to those induced by Ctrl
measured 28 days after boost immunization (Fig. 8). We observed
(Fig. 7A). Moreover, the response was highly variable. With both
that single MNAP priming with QS-21-adjuvanted HBsAg was able
15 mg and 50 mg QS-21, anti-HBsAg antibody levels were not inferior
to elicit an anti-HBsAg antibody response (Group 5), but it was not
to those induced by Ctrl. When considering skin reactivity (Fig. 7B),
non-inferior to the benchmark (priming and boost with Ctrl; Group
it appeared that immunizing with 50 mg QS-21 was more reacto-
1). Boosting with QS-21-adjuvanted HBsAg led to antibody levels
genic than with 15 mg. Scab was particularly important when
not inferior to those obtained in the benchmark (Group 4 versus
immunizing with 50 mg QS-21, as compared with 15 mg, for almost
Group 1). Likewise, priming with Ctrl and MNA boosting with QS-
D. Poirier et al. / Biomaterials 145 (2017) 256e265 263
4. Discussion
Fig. 9. Skin reactions over time after administration of MNAP with QS-21-adjuvanted HBsAg. A. Skin reactions after one MNAP immunization (priming). B. Skin reactions after the
second immunization done with MNAP. Pre indicates the reactions before immunization. Day 0 refers to the reactions measured 2 h after immunization. Ctrl refers to the control
vaccine. The same color code as in Fig. 8 has been used for the groups. One symbol represents one animal. (For interpretation of the references to colour in this gure legend, the
reader is referred to the web version of this article.)
formulated in liposomes in a single device is unique. Further, the intramuscular injections with less dependence to cold chain re-
antigen was found very stable when incorporated in the HES-based quirements. These quality attributes will need to be further
matrix that constitutes the microneedles. Therefore, dissolvable conrmed in subsequent human studies.
MNAP could represent a suitable alternative to conventional
D. Poirier et al. / Biomaterials 145 (2017) 256e265 265
Science) for editorial input and Ulrike Krause (GSK Vaccines) for the
coordination of manuscript development.
References