key: cord-0268949-i572ljzo
authors: Martini, Veronica; Hinchcliffe, Michael; Blackshaw, Elaine; Joyce, Mary; McNee, Adam; Beverley, Peter; Townsend, Alain; MacLoughlin, Ronan; Tchilian, Elma
title: Distribution of droplets and immune responses after aerosol and intra-nasal delivery of influenza virus to the respiratory tract of pigs
date: 2020-06-05
journal: bioRxiv
DOI: 10.1101/2020.06.04.134098
sha: c0d0446a42763d52bd3c2cfea97f303a456bb1a4
doc_id: 268949
cord_uid: i572ljzo

Recent evidence indicates that local immune responses and tissue resident memory T cells (TRM) are critical for protection against respiratory infections but there is little information on the contributions of upper and lower respiratory tract (URT and LRT) immunity. To provide a rational basis for designing methods for optimal delivery of vaccines to the respiratory tract in a large animal model, we investigated the distribution of droplets generated by a mucosal atomization device (MAD) and two vibrating mesh nebulizers (VMNs) and the immune responses induced by delivery of influenza virus by MAD in pigs. We showed that droplets containing the drug albuterol, a radiolabel (99mTc-DTPA) or a model influenza virus vaccine (S-FLU) have similar aerosol characteristics. 99mTc-DTPA scintigraphy showed that VMNs deliver droplets with uniform distribution throughout the lungs as well as the URT. Surprisingly MAD administration (1ml/nostril) also delivered a high proportion of the dose to the lungs, albeit concentrated in a small area. After MAD administration of influenza virus, antigen specific T cells were found at high frequency in nasal turbinates, trachea, broncho-alveolar lavage, lungs, tracheobronchial nodes and blood. We conclude that the pig is useful for investigating optimal targeting of vaccines to the respiratory tract.

gamma camera fitted with a Low Energy General Purpose collimator (Mediso Medical Imaging 135 Systems, Hungary), and anterior, posterior and lateral images were recorded. A pilot study 136 ( Fig. 2A) was conducted to ensure optimal dosage and gamma camera settings. Radiolabelled to cover the whole lung area (Fig. 2B) . Radioactive material deposited in the stomach was as 153 a rectangle below the lung area. Counts in the individual ROI were then adjusted for 154 background and decay relative to first image taken for each animal. Deposition in each specific 155 ROI was calculated relative to total counts as follows: standards. Analysis indicated that one VMN, termed "small VMN", generated droplets with a 202 volume mean diameter (VMD) of 3.5 µm and mass median aerodynamic diameter (MMAD) of 203 2.5 µm and the other, termed "medium VMN", generated droplets with VMD of 4.5 µm and 204 MMAD 3.5 µm (Fig. 1A) .The small VMN generated significantly more droplets of less than 5 205 µm diameter, the fine particle fraction, than the medium VMN (Fig. 1B) . These devices were 206 selected with the expectation that the aerosols generated, would preferentially deposit in the lung, 207 and minimise deposition within the URT.

In order to quantify the drug dose delivered to the LRT, we developed an in vitro model 209 based on the breathing pattern and anatomical features of pig nasal cavities and trachea. A 3D 210 printed pig head was attached to a breathing simulator with a collection filter, representing lung 211 deposition, and albuterol was delivered using a custom-made mask. Because mammalian lung 212 volume is influenced by bodyweight we modelled the breath of 15 kg (tidal volume of 115 ml) 213 and 20 kg pigs (tidal volume of 150 ml), assuming 7.7 ml/kg bodyweight tidal volume (33).

These weights are in the range of the pigs used in the scintigraphy experiments described below.

The small VMN deposited more albuterol on the filter (18.1% for 15 kg and 22.3% for 20 kg 216 pigs) than the medium VMN (15.9% and 20.5% respectively) (Fig. 1C) . These data showed that 217 small and medium VMNs generated droplets with different size distributions which could 218 potentially influence total lung deposition in vivo. (Fig. 1D) . 231 We also characterized the droplets generated using the MAD with 99m Tc-DTPA and S- (Fig. 1E) . 236 The results indicate that in vivo deposition using 99m Tc-DTPA would be an accurate vivo, 99m Tc-DTPA was administered to sedated pigs (n=3) using small and medium VMNs or 243 i.n. with the MAD according to a randomised crossover experimental design ( Fig. 2A) . 

Images taken of each delivery system before and after administration to pigs revealed 254 that the residual dose of 99m Tc-DTPA in the MAD was only 11% compared to approximately 255 50% with nebuliser plus mask; the latter was predominantly associated with the mask filter 256 which is designed to capture exhaled air ( Table I and Fig. 2C ).The regional deposition of

Tc-DTPA in the pig was subsequently estimated as a percentage of the total amount detected 258 (i.e. sum of counts associated with ROIs), as is standard practice. The 'face' counts include those associated with deposition of 99m Tc-DTPA in the nasal 280 cavities as well as those due to contamination of the external facial skin. We attempted to 281 estimate the amount of facial skin contamination by removing external 99m Tc-DTPA with a 282 skin wipe and imaging; revealing that more than 20% of the counts delivered by VMN to the 283 face area were associated with the skin compared to 7% with the MAD (Fig. 2E ). This fraction VMNs were used. In contrast, the majority of face counts appeared to be associated with the 288 nasal cavities with the MAD (Fig. 3A) . 289 Lung distribution patterns differed dramatically between VMNs and MAD. The MAD 290 delivered a high dose of 99m Tc-DTPA to a relatively small area of the lung. In contrast, both 291 nebulizers delivered droplets more uniformly throughout both lungs with small VMN resulting 292 in higher and more consistent total lung deposition (mean ± SD 32.9% ± 1.6%) compared to 293 the medium VMN (25.4% ± 6.8%) ( Fig. 2D & 3A) . 294 To establish the reproducibility of the scintigraphy technique, we compared the results 295 between the pilot and second study legs, where the same devices were used in each animal.

This comparison, although limited to 1 animal per device type, highlighted a potential for 297 increased variability of i.n. MAD instillation compared to aerosol delivery by VMNs in all 298 regions of the RT (Fig. 3B) . 299 Together these data suggest that targeting to the pig lung is best achieved with use of a 300 VMN. Although the MAD delivers a higher but variable proportion to the lung as well as the 301 URT and the stomach, delivery via a nebulizer gives a much more even targeting of both lungs 302 with deposition in the URT as well. IFNγ (49.3%) (Fig. 4B) . (1:320 50% inhibition titre at day 21) and a much lower titre in BAL (1:15) ( Figs. 5A and B) . 349 In the nose, IgA dominated over IgG and reached a plateau by day 14 (Fig. 5C) . Taken together 

Human respiratory tract model for radiological protection. A report of a Task 603

Group of the International Commission on Radiological Protection

Mechanisms of pharmaceutical aerosol deposition in the 606 respiratory tract

Scaling of respiratory variables in mammals

EANM 611 guideline for ventilation/perfusion single-photon emission computed tomography 612 (SPECT) for diagnosis of pulmonary embolism and beyond

Challenges in assessing regional distribution of 615 inhaled drug in the human lungs

Comparison of 617 aerosol delivery across combinations of drug delivery interfaces with and without 618 concurrent high-flow nasal therapy

SPECT-CT Comparison of Lung Deposition 621 using a System combining a Vibrating-mesh Nebulizer with a Valved Holding 622 Chamber and a Conventional Jet Nebulizer: a Randomized Cross-over Study

SARS-CoV-2 infection protects against rechallenge in rhesus macaques

Rapid development of an inactivated vaccine candidate for SARS-CoV-2

Generation and persistence of human tissue-resident memory T cells 642 in lung transplantation

Resident memory CD8(+) T cells in the upper 645 respiratory tract prevent pulmonary influenza virus infection

Rapid turnover of effector-memory CD4(+) T 648 cells in healthy humans

Quantifying Memory CD8 T Cells Reveals 651 Regionalization of Immunosurveillance

Selective accumulation of differentiated CD8+ T cells specific for 654 respiratory viruses in the human lung

CD8(+) T cells with an intraepithelial 658 phenotype upregulate cytotoxic function upon influenza infection in human lung

Prevention of 664 tuberculosis in macaques after intravenous BCG immunization

Relative immunogenicity of the cold-666 adapted influenza virus A/Ann Arbor/6/60 (A/AA/6/60-ca

A/AA/6/60-ca, and parental strains with similar surface antigens

Small-particle aerosolization of live 670 attenuated influenza vaccine virus

Nasal deposition and clearance in man: 672 comparison of a bidirectional powder device and a traditional liquid spray pump