A) B)

C) D)

E)

Bergin P1, Langat R2, Omosa-Manyonyi G2, Farah B2, Bizimana J3, Umvirigihozo G3, Nyombayire J3, Ouattara G2, Ingabire R3, Park H4, Chetty P5, Verlinde C4, Laufer D4, Cox J4, Stevens G5, Fast P4, Gilmour J1, Karita E3, Anzala O2

1 International AIDS Vaccine Initiative, Human Immunology Laboratory, Imperial College London, UK, 2Kenya AIDS Vaccine Initiative, University of Nairobi, Nairobi, Kenya, 3Project San Francisco, Kigali, Rwanda, 4International AIDS Vaccine Initiative, New York, USA and 5International AIDS Vaccine Initiative, Johannesburg, South Africa

Selection and Optimisation of a Mucosal Sampling Method for Application in Phase 1

Clinical Trials

REFERENCES ACKNOWLEDGEMENTS

BACKGROUND In sub-Saharan Africa the main route of transmission of HIV-1 remains across the genital

mucosa during sexual intercourse. A successful HIV preventative vaccine will likely require

the induction of an anti-HIV immune response at the mucosal surface. There are several

approaches that may be used in order to induce a longer lasting immune response at the

mucosal surface. The International AIDS Vaccine Initiative (IAVI) has generated a panel of

potential vaccine candidates; replicating vectors containing a vaccine insert intended to

induce a stronger and longer lasting immune response. Administered nasally, these vectors

may target the vaccine induced immune response to the genital mucosa and/or gastric

associated lymphoid tissue. A pilot study was performed to determine if anti-HIV antibodies

could be detected in nasal secretions collected from the mid-turbinate region of the nasal

cavity (NC), naso-pharynx (NP) and oral secretions (Figure 1). To determine the optimal

location for sampling that would allow for the most compliancy, the collection

device/material that would give the highest yield of sample, and whether anti-HIV-1

antibodies could be detected in the nasal compartment. Volunteers were recruited at

Clinical Research Centers in Nairobi, Kenya (Kenya AIDS Vaccine Initiative; KAVI) and Kigali,

Rwanda (Project San Francisco; PSF) to prepare for future Phase I Clinical Studies.

METHODSEthical approval was granted by University of Nairobi Ethics Committee (KAVI), Rwanda

National Ethics Committee and Emory University IRB (PSF), and all volunteers provided

informed consent to take part. At KAVI 15 HIV-1 seropositive and 15 HIV seronegative

volunteers were recruited into the study. KAVI staff transferred the sampling and processing

methodologies to PSF, where 20 HIV-1 seropositive and 20 HIV-1 seronegative volunteers

were recruited. Nasal samples were collected using FloQ swabs (COPAN), either from the

mid-turbinate region (NC) or deeper in the naso-pharynx (NP) (Figure 1), saliva was

collected from parotid glands (Figure 1) using Salimetrics Oral Swabs and whole oral fluid

(transudate) was collected into Falcon tubes. In addition, volunteers at KAVI provided a mid-

turbinate secretion collected using a Synthetic Absorptive Matrix (SAM) strip. Eluted

samples were aliquoted and stored at -80C until they were tested for gp140 Env (Clade A

UG37) and Gag p24 (Clade B LAI) IgG and IgA antibodies using a Enzyme Linked

Immunosorbent Assay (ELISA) modified from previously described (Keefer et al., Cranage et

al.; Figure 2). All mucosal samples were diluted 1:20 prior to analysis, and plasma

samples were diluted 1:100. All samples from KAVI were assessed on site at KAVI, while PSF

samples were assessed at the Human Immunology Laboratory in London. A panel of test

plasma were used to ensure parity between the two testing laboratories.

RESULTSAll volunteers provided all nasal and salivary samples although volunteers indicated that the NC

collection was preferable to the deeper NP sample.

RESULTSAnti-HIV antibodies were detected in nasal secretions of 100% of HIV-seropositive samples with

IgG expressed at a higher level than IgA (by optical density O.D.). Anti-gp140 IgG and anti-p24

IgG were detected in 100% and 93.4% of nasal samples, respectively. IgA anti-gp140 and anti-

p24 IgA were detected in 94% and 88.5% of nasal samples, respectively. No significant

differences were detected between NC and NP samples in magnitude or quantity. All nasal

samples from HIV-seronegative volunteers were negative for IgG and IgA anti-gp140/p24 except

for 4 Rwandan volunteers with low levels of anti-p24 antibodies in nasal secretions and/or

salivary samples. In all cases the antibodies were present in their plasma samples, suggesting

these volunteers may have been exposed to HIV or early in infection.

Anti-HIV antibodies were present in both whole oral fluid and parotid saliva from HIV seropositive

volunteers, but at much lower levels than that seen in the nasal secretions (data not shown).

Whole oral fluid contained much more mucus than parotid saliva and was more difficult to

process and assess on the ELISA.

Data from KAVI and PSF were comparable, although PSF samples consistently yielded higher

O.D.’s than the KAVI samples.

IgG antibodies were expressed at higher levels than IgA antibodies.

Midturbinate (NC) samples were similar in expression to naso-pharyngeal (NP). In all cases FloQ

swab samples gave a higher yield than the SAM strip.

DISCUSSIONNP sampling appeared to have little benefit over NC sampling, and HIV antibodies were detected

in all HIV-seropositive individuals. Volunteers indicated that the NC sampling was preferable to

the deeper NP collection, suggesting that NC sampling may result in greater compliance with

repeated sampling. NC sampling may provide a unique and tolerable method to collect antibodies

in an HIV vaccine trial following intranasal vaccination. The SAM strip has been used in studies

where there has been rapid, repeated, sampling (i.e. minutes between samples). In this case,

where a single sample was taken per visit, the FloQ swab appeared to yield a better sample,

likely due to collecting from the entire cavity.

CONCLUSIONSSamples collected from the turbinate nasal cavity demonstrate antibodies against HIV-1

Sample collected from multiple research centres were comparable

The methods described here are likely to be acceptable to volunteers in a Phase 1 Clinical Trial

1. Keefer et al. PLOS One 2012;7(8):e41936

2. Cranage et al. Vaccine. 2011 Feb 4;29(7):1421-30

UG37 Ag Clade A gp-140/ LAI p24

Clade B

Sample containing Ab

Detection antibody anti-human IgG/IgA

Substrate /colour development

Figure 2. ELISA method

Mid turbinate

Nasopharynx

Parotid papilla, marks the entry to the parotid duct

A) B)

Figure 1. A) Nasopharyngeal tract with midturbinate (NC) and naso-pharynx (NP) indicated and B) parotid gland

Figure 3. Anti-HIV-1 IgG and IgA against p24 (Clade B LAI) and envelope gp140 (Clade A UG37) were present in HIV-1 seropositive samples from KAVI and PSF.

KAVI samples from A) HIV seronegative volunteers and B) HIV seropositive volunteers and PSF samples from C) HIV-1 seronegative volunteers and D) HIV-1 seropositive volunteers. All seronegative samples were below the cut-off, except samples from 4 PSF seronegative volunteers with low levels of anti-p24 antibodies (Figure 3 C).

E) A comparison of the samples collected at KAVI and PSF. PSF had a trend towards a higher yield of IgG and IgA antibodies

Key:

Circles = Nasal FloQ swabSquares = Nasal SAM stripTriangles = Nasopharyngeal FloQ swabBlack = Env (gp140) IgABlue = Env (gp140) IgGRed = Gag (p24) IgAGreen = Gag (p24) IgG

Figure 4. Devices used for nasal sample collection. A) COPAN FloQ swab used to collect samples from the midturbinate region, or up to 10cm into the naso-pharynx. B) The Synthetic Absorptive Matrix (SAM) strip (Hunt Diagnostics UK) used to collect repeat sample from the midturbinate region of the nasal cavity.

A) B)