ANTIFIBROTIC COMPOSITION

Abstract

Methods of prophylactically treating fibrotic conditions using a synthetic lamellar body composition are provided, such as conditions of the lung, skin, gastrointestinal system, genitourinary system, heart, peritoneum, kidney, liver, and mucosa. In particular the present invention is concerned with lung injury which may be characterised by increased pulmonary vascular permeability. Suitably the lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions comprises phosphatidylcholine, cholesterol and optionally at least one phospholipid selected from phosphatidyl serine, phosphatidyl glycerol and phosphatidyl inositol to provide an anionic lamellar body.

Claims

1. A lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions, optionally wherein the conditions are selected from conditions of the lung, skin, gastrointestinal system, genitourinary system, heart, peritoneum, kidney, liver, and mucosa.

2. A lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions as claimed in claim 1 formulated for administration via the airway to the epithelium of the lower airways for the prevention and/or treatment of distal lung injury due to direct or indirect injury optionally wherein the lung injury is selected from sepsis, ventilator-induced lung injury, ischemia/reperfusion, hyperoxia, ALI, ARDS, or conditions caused by irradiation of the lower neck, thoracic structures or chest wall.

3. The lamellar body composition for use in the treatment of fibrotic or pro-fibrotic conditions formulated for administration via the airway to the epithelium of the lower airways for at least one of the prevention and treatment of distal lung injury due to direct or indirect injury as claimed in claim 2 wherein the lung injury is selected from conditions caused by irradiation of the lower neck, thoracic structures or chest wall.

4. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any one of the preceding claims wherein the lamellar bodies in the lamellar body composition are sized at less than 250 nm.

5. The lamellar body composition for use in the treatment of fibrotic or pro-fibrotic conditions of any one of the preceding claims wherein the lamellar bodies are provided as droplets with an average size about 1.5 microns.

6. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any one of the preceding claims comprising phosphatidylcholine, cholesterol and optionally at least one phospholipid selected from phosphatidyl serine, phosphatidyl glycerol and phosphatidyl inositol to provide an anionic lamellar body.

7. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any of the preceding claims comprising phosphatidylcholine, cholesterol and phosphatidyl serine.

8. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any of the preceding claims comprising about 44-70% phosphatidylcholine, about 2-18% phosphatidyl serine, and about 4-12% cholesterol by weight and optionally a further lipid.

9. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any of the preceding claims comprising about 44-70% phosphatidylcholine, about 15-23% sphingomyelin, about 6-10% phosphatidyl ethanolamine, about 2-15% phosphatidyl serine, about 2-4% phosphatidyl inositol and about 4-12% cholesterol by weight.

10. The lamellar body composition for use in the prophylactic treatment of fibrotic or pro-fibrotic conditions of any of the preceding claims comprising about 54% phosphatidylcholine, about 19% sphingomyelin, about 8% phosphatidyl ethanolamine, about 4% phosphatidyl serine, about 3% phosphatidyl inositol and about 10% cholesterol by weight, optionally comprising about 2% by weight lysophosphatidyl choline.

11. A method of supplementing lung surfactant in a mammalian subject with depleted lung surfactant, comprising administering to the subject a lamellar body composition comprising phosphatidylcholine, cholesterol and optionally at least one phospholipid selected from phosphatidyl serine, phosphatidyl glycerol and phosphatidyl inositol to provide an anionic lamellar body formulated for administration via the airway to the epithelium of the lower airways in a treatment effective amount.

12. A method of treating lung injury in a mammalian subject in need thereof, comprising administering to said subject a lamellar body composition comprising phosphatidylcholine, cholesterol and optionally at least one phospholipid selected from phosphatidyl serine, phosphatidyl glycerol and phosphatidyl inositol to provide an anionic lamellar body formulated for administration via the airway to the epithelium of the lower airways, optionally wherein the lung injury is selected from sepsis, ventilator-induced lung injury, ischemia/reperfusion, hyperoxia, ALI, ARDS, or radiation treatment

13. The method of treating a lung injury as claimed in claim 11 or 12, wherein said lung injury is a radiation treatment injury caused by alpha radiation, beta radiation, neutron radiation, or gamma radiation, optionally wherein said radiation injury is radiation pneumonitis or radiation-induced lung injury, or a fibrotic pulmonary condition caused by radiation.

14. The method of any one of claims 11 to 13, wherein said administering step is carried out as a single dose of said treatment effective amount, optionally wherein said administering step is carried out by inhalation administration.

15. The method of any one of claims 11 to 14, wherein said administering step is carried out within 1, 2 or 3 days of said ionizing radiation injury.

16. The method of any of claims 11 to 14 wherein said administering step is carried out 1, 2 or 3 days prior to treatment of the subject with ionizing radiation.

Description

DETAILED DESCRIPTION

[0096] In the present description three model systems are described: [0097] a Radiation-Induced Lung Injury (RILI) model; [0098] a TGF-β1 model system used to consider profibrotic mediators and methods of screening of candidates to counter such profibrotic mediators. As discussed further below, expression of TGF-β1 has been found to be associated with lung fibrosis six months after radiation exposure. Thus, TGF-β1 expression was considered to be an indicative measure of the efficacy of candidates to mediate and reduce fibrosis or profibrotic mediators. [0099] a cell entry model.

[0100] The RILI model was selected as a validated model of human pulmonary interstitial fibrotic response to damage. As would be known in the art, this model allows evaluation of the molecular mechanisms associated with radiation-induced lung injury and efficacy screening of candidate countermeasures.

[0101] Further this model allows the study of the development of fibrosis, and further recognised conditions of interstitial lung disease which cause effective disruption of the alveolar capillary membrane e.g. mild pneumonitis or pleural effusions.

RILI Model

[0102] To provide a model of lung injury, twelve commercially sourced adult Shetland sheep (bodyweight: 38.5 kg [33.0-43.0] median [range]; 6 female and 6 castrated male) were included in the described study. Identification of animals was by means of ear tags. Animals were housed for the duration of the study and otherwise maintained according to normal standards of farm animal husbandry. The sheep were treated with anthelminthic before the study began. The sheep were randomly allocated to one of two sex-matched treatment groups.

[0103] In order to confirm the absence of pre-existing pulmonary disease and collect baseline samples against which to judge change within animals, preliminary baseline examination (BBr1) involving bronchoscopic visualisation, bronchoalveolar lavage and bronchial brush biopsy under gaseous anaesthesia was conducted. Where bronchoalveolar lavage cytology failed to meet normal boundaries and was indicative of parasitism (% eosinophils>7.5%) the sheep were re-treated with anthelminthic and results confirmed within normal range prior to any further involvement in the experimental protocols. A further two baseline examinations (BBr2 & BBr3) involving bronchial brush biopsy sampling were thereafter conducted at fortnightly intervals. Measurements of bodyweight and rectal temperature were also made at these time points. At least two weeks after the last baseline assessment (BBr3) the sheep were re-anaesthetised and positioned in sternal recumbency in order to facilitate the acquisition of thoracic computed tomography images for subsequent radiation treatment planning. Following the last radiation treatment (t0) the sheep were closely monitored for any evidence of adverse effect. At t0+11d and t0+21d the sheep were re-anaesthetised and subject to bronchial brush biopsy in the same manner as during the preliminary baseline evaluation. At t0+23d the sheep were killed by overdose of anaesthetic and presented for necropsy examination.

[0104] During radiation treatment a total dose of 30 Gy was delivered in a fractionated regime. This involved the delivery of 6 Gy to a defined planning target volume (PTV) of the left caudal diaphragmatic lung lobe on each of five separate occasions over a period of two weeks (3-4 days intervals).

[0105] Bronchoalveolar Lavage Collection

[0106] The bronchoscope (Model FG-15W; Pentax UK Ltd.) was wedged in the segmental bronchus of the right apical lobe. Two 20 ml aliquots of PBS were used to collect bronchoalveolar lavage fluid (BALF) from this lung segment. BALF samples were placed into sterile tubes and kept on ice until subsequent analysis. Five millilitres of BALF was removed and centrifuged at 400 g for seven minutes to separate out the cellular fraction. The resultant pellet was re-suspended in sterile phosphate buffered saline (PBS) and the total cell number counted before subsequent preparation of cytospins for differential cytology. Cells were counted using a Neubauer haemocytometer and values expressed per millilitre BALF. Cyto-centrifuge slides were prepared and stained using Leishman stain for differential counts on 500 cells. Cells were classified as neutrophils, macrophages, eosinophils, lymphocytes or mast cells according to standard morphological criteria.

[0107] Bronchial Brush Biopsy

[0108] On each occasion of baseline assessment three bronchial brush biopsy samples were derived from each of three separate areas of the lung (n=9 total). Samples were derived from bronchi within the left caudal diaphragmatic lung lobe (LCD), the right caudal diaphragmatic lung lobe (RCD), and also from bronchi within areas of the anterior right lung. On each occasion considerable care was taken (through manual mapping and reference to video recordings) to avoid sampling any area of bronchial epithelium that had previously been subject to bronchial brush biopsy. At t0+11d and t0+21d the sheep were subject to bronchial brush biopsy in the same manner as during the preliminary baseline evaluations.

[0109] Necropsy

[0110] Following euthanasia by intravenous injection of barbiturate, the heart and lungs were carefully removed from the carcase following standard necropsy protocols. The pulmonary circulation was perfused via the pulmonary artery with 2-3 litres of PBS before the heart was dissected away. The lungs were then photographed before being presented for further processing.

[0111] Inflation Fixation

[0112] Lung tissue was fixed by airway instillation of 10% neutral buffered formalin. The trachea was connected to a reservoir of fixative and the fixative allowed to flow until the ‘natural contours’ of the lung were established. The lungs were then floated in a tank of the same fixative and inflation-fixed at a constant pressure of 3.0 kPa for a period of 7 days.

[0113] Gross Tissue Sampling

[0114] After fixation each lung was carefully sliced along the transverse plane, starting at the caudal pole of each diaphragmatic lobe, into fifteen 1 cm thick tissue slices. These slices were then arranged in consecutive order for photographing prior to a representative tissue block from each contiguous slice being selected and carefully dissected from surrounding lung tissue. A further photographic image of the slices with their selected blocks in situ was captured to document the spatial origin of each block. This latter step was a necessary prerequisite to registering the position of each block with respect to the radiation field through reference to CT images previously collected from the same animals. Tissue blocks were then submitted for standard histological processing and paraffin embedding.

[0115] Block Selection

[0116] A formalin-fixed paraffin embedded (FFPE) tissue block from the left caudal diaphragmatic lung that represented the isocentre of the planning target volume was identified and selected, as was the corresponding block from the right contralateral control lung. These blocks were thenceforth labelled LL_Post and RL_Post respectively (“LL” and “RL” for left and right lung respectively, and “Post” for posterior). A further FFPE tissue block was sourced anterior (14.5 mm [9.7-23.0]) to the cranial margin of the PTV (LL_Ant), as well as its corresponding block from the right contralateral control lung (RL_Ant).

[0117] Histochemical and Immunohistochemical Staining

[0118] Sections cut from the above blocks were stained with haematoxylin-eosin (H&E) and Picrosirius Red, as well as being immunostained with antibodies specific for the following antigens—ASMA, DC-LAMP, and Ki67 protein. All slides were stained using standard immunohistochemistry methods with endogenous peroxidase blocked using 3% H.sub.2O.sub.2 in methanol and heat-induced antigen retrieval performed using 10 mM citrate buffer pH6.0.

[0119] ASMA—Non-specific binding was blocked with 10% Normal Goat Serum (Sigma G9023) in PBS+0.5% Tween 80. Primary antibodies Monoclonal ASMA (Sigma A2547) and Normal Mouse IgG isotype control (Sigma M5284) were diluted to 1 μg/ml in blocking buffer and incubated for 30 minutes at room temperature. Detection using biotinylated goat anti mouse IgG (Vector BA-2001) and Streptavidin peroxidase polymer (Sigma S-2438) followed by DAB substrate (Vector SK-4100) with Haematoxylin counterstain.

[0120] Ki67—Non-specific binding was blocked with 3% BSA (Sigma A3733) in PBS+0.05% Tween 20. Primary antibodies were Monoclonal anti Ki67 clone MIB-1 (Dako M7240) and Normal Mouse IgG isotype control (Sigma M5284) were diluted to 1 μg/ml for 45 minutes at room temperature. Detection using biotinylated goat anti mouse IgG (Vector BA2001) and Streptavidin peroxidase polymer (Sigma S-2438) followed by DAB substrate (Vector SK-4100) and Haematoxylin counterstain.

[0121] DC-LAMP Non-specific binding was blocked with 4% normal rabbit serum (Sigma R9133) in PBS+0.2% Tween 80. Primary antibodies DC-LAMP/CD208 (2BScientific DDX0191P-50) and Normal Rat IgG isotype control (Serotec MCA1125R) were diluted to 2.5 μg/ml in blocking buffer and incubated overnight at 4° C. Detection using biotinylated goat anti rat IgG (Vector BA4001) and Streptavidin peroxidase polymer (Sigma S-2438) followed by DAB substrate (Vector SK-4100) and Haematoxylin counterstain.

[0122] Bronchial Brush Cytokine Expression

[0123] Bronchial brush biopsy specimens were collected using cytology Brushes (Conmed Endoscopic Technologies 152R) agitated into 1 ml of cold sterile PBS (Sigma D8537) through 200 μl wide orifice pipette tips (Star Lab E1011-8000) and centrifuged at 10,000 g for 5 minutes. Pellets were resuspended in RLT buffer (Qiagen 74106) containing 1% β mercaptoethanol and stored at −80° C. until extraction. All samples were run through Qiashredder columns (Qiagen 79656) and RNA extractions were done using RNeasy mini kit (Qiagen 74106) with DNase treatment using Rnase free DNase set (Qiagen 79254). RNA was quantified on Nanodrop and quality checked on Agilent Tapestation with RNA screentape (Agilent 5067-5576). cDNA was prepared from 400 ng RNA with Transcriptor First Strand cDNA Synthesis kit (Roche 04 896 866 001) using random hexamer primers. Quantitative Real Time PCR was performed using Lightcycler 480 with 2.5 μl cDNA in LightCycler 480 Sybr Green I Master (Roche 04 887 352 001) and specific primers. Advanced relative quantification was calculated using Lightcycler 480 SW1.5 programme. Standard curves for each gene generated from pooled ovine alveolar macrophage cDNA. Melt curve analysis showed single peak for all samples. PCR efficiency was in range of 1.8 to 2.1. qPCR conditions and referenced primer sets (12-15) are stipulated in Tables 1_**qPCR conditions** and 2_**qPCR primer sets**.

[0124] Semiquantitative Histopathological Evaluation

[0125] One pathologist (SHS) examined a reduced subset (saline treated only) to obtain preliminary histopathology results to inform further analysis, whereas the other pathologist (JDP) was blinded to the results of the preliminary analysis and to the source of the slides. All H&E stained sections were scanned on a whole slide scanner (Nanozoomer, Hamamatsu, Japan) to acquire whole slide images (WSI) at ×40 magnification. These sections were then subject to detailed examination by a veterinary pathologist (Dr del-Pozo) blinded to their specific identity. Following an initial appraisal in which principal pathologic features were identified a semi-quantitative scoring system was developed to capture the incidence and extent of each feature amongst the different sections. Briefly, to score lesions for severity each section was allocated a score ranging from 0 (absent), 1 (mild), 2 (moderate), to 3 (severe). Fibrosis was scored by allocation of an estimated % surface involved (note that this score does not consider severity, which was mild in all cases in areas affected). Three variables, pneumocyte type II cell hyperplasia and atypia, and epithelial atypia, were scored qualitatively i.e. presence or absence.

[0126] Quantitative Histological Analyses

[0127] Areas of alveolar oedema could be clearly identified and annotated in Masson's trichrome stained sections using the Hamamatsu NDP.view2 viewer software. The area of each tissue section was outlined by using the freehand region tool, as was the area occupied by large (cartilaginous) airways and associated blood vessels. Finally, the area of each section occupied by alveolar oedema was also annotated. Measured annotations were saved to file, and the percentage of ‘parenchyma’ occupied by oedema (% Oedema Area.sub.[parenchyma]=(total area occupied by alveolar oedema/(whole section area minus large airways and blood vessels))*100) was calculated.

[0128] Within ImageJ, the NDPITools custom extract to TIFF/mosaic plugin was used to extract each ndpi image file to multiple TIFF images. WSI from H&E stained sections were extracted at ×20 resolution, and the remaining WSIs at ×40 resolution. Image fields containing parenchyma (including airways no larger than respiratory bronchioles) were then manually selected from a random selection of these extracted files. These files were then converted to OME-TIFF using an ImageJ recursiveTiffConvert macro to engage the Bio-Formats exporter function.

[0129] Fractal analysis was applied to H&E-stained sections in order to assess the morphometry of the distal lung. Where available, 100 H&E-stained images were randomly selected from each section (LL_Ant, LL_Post, RL_Ant, and RL_Post). In the five instances where less than 100 images were available, 92, 78, 49, 90 and 56 images were selected. Parenchymal images were then converted to OME-TIFF using a recursiveTiffConvert macro to engage the Bio-Formats exporter function. In a manner similar to that described by Andersen et al (2012) these converted files were then processed to binary images using imageJ functionality and each binary image was analysed using the ImageJ FracLac plugin which calculates the fractal box dimension (D.sub.B) (16).

[0130] The parenchymal ×40 OME-TIFF files were batch processed using macros employing the colour devolution plugin for detecting the area of red-stained collagen in Picrosirius Red-stained sections, the area of diaminobenzidine (DAB) stain in ASMA and DC-LAMP immunostained sections, and the number of DAB-stained particles (>200 pixels.sup.2) in Ki67 immunostained sections. Sample size was considered acceptable if the standard error for percentage area measurement fell below 5% of the mean value for that measurement. In the six sections where this condition was not met, the standard error ranged from 5.0 to 7.3% of the mean. As the scarcity of Ki67-stained cells in control lung sections meant that the 5% limit could not be achieved for the majority of sections a pragmatic decision on sampling was taken. Between 172 and 198 fields on each Ki67-stained section were examined.

[0131] Statistical Analyses

[0132] Data was initially assessed for normality of distribution using a Kolmogorov-Smirnoff test. Where necessary data transformation was applied to normalise the distribution, and where such transformation failed, a rank-order transformation was applied prior to subsequent evaluation. For repeated measures data a General Linear Model was fitted in which responses in radio-exposed and contralateral control lung segments were evaluated with respect to time, and the experimental treatment (lamellar body composition (LMS), SAL). Sheep identity, nested within treatment, was considered a random factor in the design.

[0133] For the analysis of histopathological and immunohistochemical data a General Linear Model two-way analysis of variance was conducted on the influence of the two independent variables (Lung, Treatment) on the variable in question. Lung included four levels (LL_Ant, LL_Post, RL_Ant, RL_Post), and Treatment two levels (Lamellar body composition, SAL).

[0134] Results

[0135] No adverse effects were noted either as a consequence of aerosol delivery, or in relation to exposure to radiation. At every time point sheep were weighed and rectal temperature recorded.

[0136] Bodyweight data was rank transformed and analyzed by two-way ANOVA with repeated measures in one factor. Treatment (LMS, SAL) was statistically significant at the 0.05 significance level. The main effect for Treatment yielded an F ratio of F(1, 82)=6.32, p=0.031. The main effect for Time yielded an F ratio of F(9, 82)=1.58, p=0.136, indicating no significant effect with respect to bodyweight. The interaction Treatment*Time was significant F(9,82)=25.10, p=0.001, highlighting the increase in weight of only the LMS-treated sheep from Rx1 onwards.

[0137] Temperature data was similarly rank transformed and analyzed by two-way ANOVA with repeated measures in one factor. Neither treatment (LMS, SAL) nor Time were statistically significant at the 0.05 significance level. The main effect for Treatment yielded an F ratio of F(1, 82)=2.09, p=0.178, and the main effect for Time yielded an F ratio of F(9, 82)=1.19, p=0.313—both indicating no significant effect with respect to body temperature. The interaction Treatment*Time was significant F(9,82)=3.31, p=0.002, highlighting the increase in temperature of only the saline-treated sheep from baseline 2 & 3.

[0138] Whilst body temperature remained within normal limits at all times during the experimental protocol, there was a small but significant increase seen at baseline 2 and 3. As this may have been indicative of a subclinical phenomenon data from these time points were discarded. Data obtained at the first baseline evaluation was used instead as the selected baseline time point.

[0139] From the Bronchial brush biopsy cytokine expression, gene expression levels of IL1 beta, TGF beta and IL8 relative to ATPase were log.sub.10-, rank-, and log.sub.10-transformed respectively to normalise data distribution prior to two-way ANOVA with repeated measures in one factor.

[0140] Neither Treatment nor Time had a significant effect on Log10 IL1beta or rank-transformed TGFbeta expression levels in samples derived from RCD or LCD and there was no significant interaction between these terms. Similarly, for samples derived from RCD neither Treatment nor Time had a significant effect on Log10 IL8 expression levels and there was no significant interaction between these terms. However, for samples derived from LCD, although Treatment had no significant effect, Time did have a significant effect on Log10 IL8 expression levels (p=0.030)—reflecting a decrease in expression at Time point 4. There was no significant interaction between these terms.

[0141] At necropsy, the pleural surface covering the planning target volume was easily identifiable as a consequence of dark red discolouration (FIG. 1). The underlying lung substance felt firmer on palpation, and when investigated in one instance, the affected lung volume failed to inflate properly when connected to a large volume-calibration syringe.

[0142] From the histopathological evaluation, the main parenchymal abnormalities noted in all radiation treated lungs were subpleural, periarteriolar and peribronchial intraalveolar oedema (FIG. 2) characterized by periarteriolar and intraalveolar accumulation of homogeneous, proteinaceous material, with occasional fibrillar material (fibrin), and aggregates of eosinophilic smudged material (fibrin). This change was associated with increase in the number of intraalveolar macrophages, which featured foamy cytoplasm in these areas. In addition, there was evidence of alveolar fibrosis characterized by mild thickening of alveolar walls by deposition of pale eosinophilic fibrillar material, interstitial pneumonia involving infiltration of alveolar walls with small numbers of lymphocytes and plasma cells, scattered pneumocyte type II hyperplasia and occasional atypia, with increased nuclear:cytoplasmic ratio, apical blebbing, mild pleomorphism, and nuclei with finely stippled chromatin and small nucleoli.

[0143] Radiation-induced abnormalities associated with the airways included mild submucosal infiltration by lymphocytes and plasma cells, and bronchial and bronchiolar epithelial atypia similar to that described for pneumocyte type II cells. Other histopathological abnormalities noted in a small number of sections involved parasite granulomas which were interpreted as unrelated to the treatment.

[0144] The results of the histopathological assessment are depicted in FIG. 3. Statistical analysis of semiquantitative and qualitative aspects of the histopathological assessment involved ranking the ordinal data and subjecting the ranked data, categorised according to group (LMS, SAL), lung (LL, RL) and segment (Ant, Post), to one-way ANOVA. Tukey pairwise comparisons indicated that for sheep treated with saline significantly increased features in radio-exposed lung relative to unexposed control lung of the same sheep included the number of intra-alveolar macrophages, the extent of alveolar oedema, the extent of interstitial pneumonia and pneumocyte type II hyperplasia, as well as the extent of peribronchial and periarterial inflammation. Pre-treatment with the lamellar body composition of the invention significantly mitigated the extent of radiation-induced interstitial pneumonia.

[0145] Results of Quantitative Histochemistry and Immunohistochemistry

[0146] Evidence of histopathological abnormality including oedema in the distal lung parenchyma raised the possibility that local lung compliance and hence morphometry of the alveolar region would be affected. Fractal analysis was applied in this context. As indicated by Porzionato A, Guidolin D, Macchi V, Sarasin G, Grisafi D, Tortorella C, et al. Fractal analysis of alveolarization in hyperoxia-induced rat models of bronchopulmonary dysplasia. Am J Physiol Lung Cell Mol Physiol. 2016; 310(7):L680-8. the fractal dimension, which measures the rate of addition of structural detail with increasing magnification, scale, or resolution, can be used to characterize the spatial pattern formed by the alveolar walls.

[0147] H&E stained slides were scanned to digital images and the NDPITools custom extract to TIFF/mosaic plugin was used to extract each .ndpi image file to multiple ×20 TIFF images. Non-parenchymal images were manually deleted and a random selection of these files was made. Where available, 100 images were randomly selected—if less than 100 images were available, then all images were selected. Parenchymal images were then converted to OME-TIFF using a recursive TiffConvert macro to engage the Bio-Formats exporter function. These converted files were then processed to binary images using imageJ functionality. Finally each binary image was analysed using the ImageJ FracLac plugin which calculates the fractal box dimension (D.sub.B).

[0148] For each sheep, the mean value for each lung segment was determined. A two-way analysis of variance was conducted on the influence of the two independent variables (Lung, Treatment) on mean D.sub.B. Lung was statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=4.15, p<0.05, indicating a significant difference between LL_Ant (M=1.70, SE=0.01), LL_Post (M=1.74, SE=0.01), RL_Ant (M=1.70, SE=0.01) and RL_Post (M=1.70, SE=0.01). The main effect for treatment yielded an F ratio of F(1, 40)=0.02, p=0.884, indicating that the effect for treatment was not significant, LMS (M=1.711, SE=0.007) and SAL (M=1.712, SE=0.007). The interaction effect was not significant, F(3, 40)=0.76, p=0.520.

[0149] It was considered these results indicated that direct exposure to radiation was associated with a significant increase in D.sub.B. Treatment had no significant influence on D.sub.B. As the alveolar fractal box dimension has been shown to inversely correlate with mean linear intercept in other animal models (Andersen M P, Parham A R, Waldrep J C, McKenzie W N, Dhand R. Alveolar fractal box dimension inversely correlates with mean linear intercept in mice with elastase-induced emphysema. International Journal of Chronic Obstructive Pulmonary Disease. 2012;7:235-43) it was considered the increased D.sub.B seen in the radio-exposed lung of sheep treated with saline corresponded to a reduction in airspace size. 50 binary images were randomly sampled and subjected to both fractal analysis as stated above, as well as conventional assessment using STEPanizer software for stereological assessment of digital images (Tschanz S A, Burri P H, Weibel E R. A simple tool for stereological assessment of digital images: the STEPanizer. Journal of Microscopy. 2011;243(1):47-59). The data confirmed that DB was significantly negatively correlated with Lm in this randomly selected subset of images (data not shown).

[0150] The percentage of parenchyma occupied by alveolar oedema (% Area Oedema) in each Masson's trichrome-stained section was calculated. A two-way analysis of variance was conducted on the influence of two independent variables (Lung, Treatment) on Rank % Area Oedema data. Lung was statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=39.76, p=0.000, indicating a significant difference between LL_Ant (M=21.96, SE=1.82), LL_Post (M=41.42, SE=1.82), RL_Ant (M=18.13, SE=1.82) and RL_Post (M=16.50, SE=1.82). The main effect for treatment yielded an F ratio of F(1, 40)=0.57, p=0.456, indicating that the effect for treatment was not significant, LMS (M=25.19, SE=1.29) and SAL (M=23.81, SE=1.29). The interaction effect was not significant, F(3, 40)=0.80, p=0.503. Taken together, it was considered these results indicate that direct exposure to radiation was associated with a significant increase in % Area Oedema. These results are in agreement with those obtained using the blinded semi-quantitative scoring scheme.

[0151] Picrosirius red stain was used to enable quantification of the extent of alveolar fibrosis.

[0152] Picrosirius Red staining in lung that had not been previously exposed to radiation was evident throughout the alveolar septa (FIG. 4). In the alveolar walls the most intense staining took the form of wavy filiform fibre stretches of variable length and thickness. The septal crests and alveolar walls abutting alveolar ducts often featured more diffuse staining where individual fibres seemed teased apart into subunit fibrils. In lung that had been exposed to radiation, fibres present in the thickened alveolar septa more often appeared teased apart giving an overall subjective impression of more abundant staining. Within sections from radio-exposed lung there was often substantial variation in the extent of staining between fields, with some areas appearing identical to those from control lung sections.

[0153] Data considering the % Area of lung parenchyma occupied by collagen (red colour) in lung sections derived from the radio exposed area of the left lung (LL_Post), its contralateral control (RL_Post), and the non-radio exposed area of the left lung (LL_Ant), and its contralateral control (RL_Ant) were analysed. The highest values are found in the radio exposed lung of sheep pre-treated with saline. A two-way analysis of variance was conducted on the influence of two independent variables (Lung, Treatment) on Rank % Area Collagen (Picrosirius Red). Lung was statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=5.72, p=0.002, indicating a significant difference between LL_Ant (M=3.685, SE=0.750), LL_Post (M=7.787, SE=0.750), RL_Ant (M=5.562, SE=0.750) and RL_Post (M=6.751, SE=0.750). The main effect for treatment yielded an F ratio of F(1, 40)=1.51, p=0.227, indicating that the effect for treatment was not significant, LMS (M=5.509, SE=0.530) and SAL (M=6.429, SE=0.530). The interaction effect (disordinal) was significant, F(3, 40)=4.53, p=0.008, indicating that the impact of Lung depends on the Treatment. Examining the fold change in % Area of collagen found in radio exposed lung relative to its contralateral control for sheep pre-treated with saline (SAL_Rx) or LMS (LMS_Rx), and in non-radio exposed lung relative to its contralateral control for sheep pre-treated with saline (SAL_CON) or LMS (LMS_CON), determines that the greatest fold change is seen in the SAL_Rx group. One-way ANOVA of Fold change versus Group indicates that the fold change in the SAL_Rx group is significantly greater than the fold change in any other group (P=0.001). The fold changes in the other groups do not differ significantly from each other.

[0154] ASMA expression in lung that had not been previously exposed to radiation was evident at the tips of secondary septal crests abutting alveolar ducts, as well as within alveolar walls similarly adjacent to ducts (FIG. 5). The same general pattern of expression, but increased in area, was evident for radiation-exposed lung.

[0155] The average percentage area of microscopic fields in each section occupied by cells staining positively for ASMA was determined. A two-way analysis of variance was conducted on the influence of two independent variables (Lung, Treatment) on Log10 % Area ASMA. Lung was not statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=1.26, p=0.303, indicating no significant difference between LL_Ant (M=0.1805, SE=0.0717), LL_Post (M=0.3261, SE=0.0717), RL_Ant (M=0.3180, SE=0.0717) and RL_Post (M=0.3638, SE=0.0717). The main effect for treatment yielded an F ratio of F(1, 40)=28.45, p=0.000, indicating that the effect for treatment was significant, LMS (M=0.1060, SE=0.0507) and SAL (M=0.4882, SE=0.0507). The interaction effect did not reach significance, F(3, 40)=2.58, p=0.067.

[0156] The relationship between the % Area occupied by collagen (Picrosirius red) and that occupied by ASMA was explored for sheep pre-treated with SAL and those pre-treated with LMS. In fitting the regression model using % Area ASMA as the response variable, % Area Sirius Red as the continuous predictor, and Treatment (LMS or SAL) as the categorical predictor, the coefficient for Treatment, indicating that the vertical distance between the two regression lines, was highly significant (p=0.000). The observation that the slope of the relationship appeared similar for the two groups was examined by including the interaction term % Area Sirius Red*Treatment in the model. This confirmed that the interaction was not significant (p=0.833).

[0157] DC-LAMP expression in lung that had not been previously exposed to radiation was evident in large well-rounded cells most commonly positioned at the intersection of neighbouring alveolar walls (FIG. 6). Their appearance and position was consistent with their presumed identity as type II pneumocytes. These cells were regularly arrayed throughout the parenchyma. In contrast, in radiation-exposed lung DC-LAMP expressing cells were usually arranged in clusters. Whilst the areas between clusters were largely devoid of the regular array of expression seen in the control lung, when cells were identified at the intersection of neighbouring alveolar walls, they appeared much larger than seen in the control lung sections. Clusters of hyperplasia comprised contiguous, usually rounded but sometimes elongated or flattened, cells lining the alveolar walls. Whilst not an absolute finding, clusters were often positioned close to respiratory bronchioles and/or alveolar ducts.

[0158] ImageJ macro routines were designed to measure the percentage area (% Area) of each section occupied by DAB stain, the number of DAB particles of a given size (150 pixels2-infinity), and the average particle size.

[0159] A two-way analysis of variance was conducted on the influence of two independent variables (Lung, Treatment) on the rank-transformed averaged median DC-LAMP NND. Lung was statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=47.92, p=0.000, indicating a significant difference between LL_Ant (M=17.83, SE=1.95), LL_Post (M=8.67, SE=1.95), RL_Ant (M=36.25, SE=1.95) and RL_Post (M=35.25, SE=1.95). The main effect for treatment yielded an F ratio of F(1, 40)=0.36, p=0.554, indicating that the effect for treatment was not significant, LMS (M=25.08, SE=1.38) and SAL (M=23.92, SE=1.38). The interaction effect was significant, F(3, 40)=5.66, p=0.002 indicating that the lung effect depended on what treatment the sheep had received. Taken together, radiation exposure caused a decrease in DC-LAMP NND in directly exposed lung, and pre-treatment with LMS also was associated with a decrease in DC-LAMP NND in non-radioexposed left lung (LL-Ant).

[0160] Cells expressing Ki67 were only rarely observed in lung that had not been previously exposed to radiation and were variously found in the septal walls, or alveolar airspaces. Proliferating cells were more commonly identified in lung previously exposed to radiation. Occasionally these cells appeared to co-locate with cells expressing DC-LAMP (FIG. 6). Ki67-expressing cells could also be identified in perivascular fascia, and within the interstitium.

[0161] Two-way ANOVA (Table 1) indicated a significant (P=0.000) lung effect, a non-significant treatment effect (P=0.221) and a significant interaction effect (P=0.034) indicating that the lung effect depended on what treatment the sheep had received.

TABLE-US-00003 TABLE 1 Summary of two-way ANOVA statistics examining the influence of two factors, Lung (with four levels: LL_Ant, LL_Post, RL_Ant, RL_Post), Treatment (with two levels: LMS, SAL), and their interaction (Lung*Treatment)(total degrees of freedom = 40) on % Area collagen, % Area ASMA, % Area DC-LAMP stain, the number and size of DC-LAMP positive particles, the nearest neighbour distance (NND) between DC-LAMP positive particles, and the number of Ki67 positive particles. The F ratios and P values are depicted in the factor columns, and the fitted means (SE mean) for each factor level is given in the level columns. For clarity, the latter are only shown where the relevant factor effect is significant. Interaction Factor Levels Factor Levels Lung* Variable Lung LL_Ant LL_Post RL_Ant RL_Post Treatment LMS SAL Treatment % Area F 5.72 3.685 7.787 5.562 6.751 1.51 4.53 collagen ratio (0.750) (0.750) (0.750) (0.750) P 0.002 0.227 0.008 % Area F 1.26 28.45 0.1060 0.4882 2.58 ASMA ratio (0.0507) (0.0507) P 0.303 0.000 0.067 % Area F 0.31 3.33 0.39 DC- ratio LAMP P 0.820 0.076 0.759 stain Number F 0.17 4.19 61.73 51.43 1.35 of ratio (3.56) (3.56) DC- P 0.918 0.047 0.273 Lamp positive particles Size of F 10.76 2.8953 2.9649 2.8327 2.8599 0.00 1.75 DC- ratio (0.0174) (0.0174) (0.0174) (0.0174) LAMP P 0.000 0.966 0.170 positive particles DC- F 47.92 17.83 8.67 36.25 35.25 0.36 5.66 LAMP ratio (1.95) (1.95) (1.95) (1.95) NND P 0.000 0.554 0.002 Number F 7.75 1.1208 1.3704 0.9011 0.9589 1.55 3.19 of ratio (0.0755) (0.0755) (0.0755) (0.0755) Ki67 P 0.000 0.221 0.034 positive particles

[0162] The results of two-way analyses of variance conducted on the influence of the two independent variables (Lung, Treatment) on the percentage area (% Area) of DC-LAMP positive staining, the number of DC-LAMP positive particles of a given size (150 pixels.sup.2-infinity) is shown in Table 1. Two-way ANOVA (Table 1) indicated a significant lung effect (P=0.000), a non-significant treatment effect (P=0.554) and a significant interaction between these terms (P=0.002) indicating that the lung effect depended on what treatment the sheep had received. Radiation exposure caused a decrease in DC-LAMP NND in directly exposed lung, and pre-treatment with LMS also was associated with a decrease in DC-LAMP NND in non-radioexposed left lung (LL-Ant).

[0163] A two-way ANOVA (Table 1) indicated that there was a significant difference in the mean percentage collagen between the different lung segments (P=0.002). Whilst the effect for treatment was not significant (P=0.227) the interaction effect was significant (P=0.008), indicating that the impact of Lung depends on the Treatment. Examining the fold change in % Area of collagen found in radio-exposed lung relative to its contralateral control for sheep pre-treated with saline (SAL_Rx) or LMS (LMS_Rx), and in non-radio-exposed lung relative to its contralateral control for sheep pre-treated with saline (SAL_CON) or LMS (LMS_CON), determines that the fold change seen in the SAL_Rx group (FIG. 4b) significantly exceeds that seen in any other group (P=0.001).

[0164] ASMA expression: Two-way ANOVA (Table 1) indicated no significant lung effect (P=0.303), a highly significant treatment effect (P=0.000), and a non-significant interaction effect (P=0.067). The relationship between the % Area occupied by collagen (Picrosirius red) and that occupied by ASMA was explored and a significant association could be demonstrated for both groups (P=0.029 for SAL, and P=0.009 for LMS).

[0165] A two-way analysis of variance was conducted on the influence of two independent variables (Lung, Treatment) on Log10% Ki67 count. Lung was statistically significant at the 0.05 significance level. The main effect for Lung yielded an F ratio of F(3, 40)=7.75, p=0.000, indicating a significant difference between LL_Ant (M=1.1208, SE=0.0755), LL_Post (M=1.3704, SE=0.0755), RL_Ant (M=0.9011, SE=0.0755) and RL_Post (M=0.9589, SE=0.0755). The main effect for treatment yielded an F ratio of F(1, 40)=1.55, p=0.221, indicating that the effect for treatment was not significant, LMS (M=1.1347, SE=0.0534) and SAL (M=1.0409, SE=0.0534). The interaction effect was significant, F(3, 40)=3.19, p=0.034 indicating that the lung effect depended on what treatment the sheep had received.

[0166] A model system in sheep was used in this study. It is proposed this model demonstrates similar findings as would be expected to be observed in humans. The consistent histopathological features associated with lung irradiation in this study, which developed within 37 days of the first exposure to radiation (within 23 days of the last exposure), were intraalveolar oedema, alveolar fibrosis, interstitial pneumonia, and pneumocyte type II hyperplasia. Whilst observations of the earliest effects of radiation to human lungs are unavailable Gross (Gross N J. Pulmonary effects of radiation therapy. Ann Intern Med. 1977; 86(1):81-92) in reviewing autopsy studies of humans dying of pneumonitis 4-12 weeks after radiotherapy also found alveolar septa thickened with oedema, cell infiltrates and the laying down of connective tissue, together with atypia, hyperplasia and desquamation of alveolar epithelial cells and the presence of hyaline membranes The early appearance of alveolar septal fibrosis is not an isolated finding. Indeed, Jennings & Arden (Jennings FL, Arden A. Development of radiation pneumonitis. Time and dose factors. Arch Pathol. 1962; 74:351-60) found that alveolar septal fibrosis could be seen in some instances less than 30 days after radiation exposure. Bennett et al (Bennett DE, Million RR, Ackerman LV. Bilateral radiation pneumonitis, a complication of the radiotherapy of bronchogenic carcinoma. (Report and analysis of seven cases with autopsy). Cancer. 1969; 23(5):1001-18), in analysing seven autopsies in which bilateral radiation pneumonitis following radiotherapy was the primary or major contributory cause of death, found alveolar septal fibrosis to be a prominent feature in five of the cases, with these patients dying between 40 and 95 days after completion of radiotherapy.

[0167] Despite sharing aspects of pathology with patients dying of radiation pneumonitis the sheep in the present study demonstrated no clinically overt adverse effect as a consequence of radiation exposure. It is considered this is a function of the fractionated dose regime and volume targeted in the present study. Indeed it has previously been shown that sheep will indeed develop radiation pneumonitis typical of that seen in humans given sufficient dose and lung volume targeted (Ohkuda K, Abe Y, Ohnuki T, Koike K, Watanabe S, Nitta S, et al. Effects of irradiation on the pulmonary vascular fluid and protein exchange. The Tohoku journal of experimental medicine. 1982; 138(3):309-12; Perkett E A, Brigham K L, Meyrick B. Increased vasoreactivity and chronic pulmonary hypertension following thoracic irradiation in sheep. J Appl Physiol. 1986; 61(5):1875-81; Loyd J E, Bolds J M, Sheller J R, Duke S S, Gillette A W, Malcolm A W, et al. Acute effects of thoracic irradiation on lung function and structure in awake sheep. J Appl Physiol. 1987; 62(1):208-18; Guerry-Force M L, Perkett E A, Brigham K L, Meyrick B. Early structural changes in sheep lung following thoracic irradiation. Radiat Res. 1988; 114(1):138-53; and Tillman B F, Loyd J E, Malcolm A W, Holm B A, Brigham K L. Unilateral radiation pneumonitis in sheep: physiological changes and bronchoalveolar lavage. J Appl Physiol. 1989; 66(3):1273-9).

[0168] Unilateral single fraction irradiation (30 Gy) of the sheep thorax produced radiation pneumonitis typical of the syndrome in humans at 4 weeks after irradiation, and whole lung irradiation (15 Gy) of sheep resulted in the development of dyspnoea three weeks after exposure that continued to progress until the animals were killed at 4 weeks. The inventors consider that sheep replicate many aspects of the early human response to lung irradiation, and that substantial pathology including alveolar fibrosis develops in sheep subjected to a radiotherapy regime (30 Gy/5 F/2 wk), which bears resemblance to palliative regimens routinely applied to patients with metastatic lung cancer, and some patients with locally advanced disease.

[0169] Whilst peribronchial and peribronchiolar inflammatory cells, comprising mostly plasma cells and lymphocytes, were relatively frequently identified in irradiated lung, the inventors found no evidence of a radiation influence on bronchial epithelial cytokine expression. Previous clinical studies which have assessed changes in plasma cytokine concentration during radiotherapy for lung cancer have demonstrated increased circulating TGF-β1, IL-6 and IL-10, and MCP-3, δMIP-1a, and IP-10. The specific cellular source of these cytokines has not been definitively ascertained.

[0170] Radiation was not considered to influence bronchial epithelial cytokine expression in this model, and is therefore unlikely to prove a useful indicator of the effect of radiation exposure in this context.

[0171] A significant increase in the percentage of parenchyma affected by oedema in radiation-exposed lung was determined by the inventors. In the method used by the inventors—manually annotating low power whole slide images of Masson Trichrome-stained sections—it is only sufficiently sensitive to identify lung areas where there is complete alveolar flooding with oedema fluid that takes up the stain. Hence the method likely underestimates the true extent of oedema and may be susceptible to variable stain uptake as a consequence of variation in the composition of the oedema fluid.

[0172] In radiation-exposed lung of sheep treated with saline, the inventors demonstrated a median fold change in the percentage area of parenchyma occupied by Sirius red-stained collagen (relative to the contralateral control lung) of 1.68 (range: 1.25-4.01).

[0173] The fibroblast populations in the alveolar mesenchyme are responsible for producing tropocollagen, the molecular component of collagen fibres, and the ground substance that fills the spaces between the cells and various fibres in the interstitial space. A particular population of differentiated fibroblasts comprise the myofibroblasts which are characterised by their expression of ASMA, as well as their ability to contract in a smooth muscle cell-like manner. Myofibroblasts play a fundamental role in alveologenesis.

[0174] In healthy lung sections ASMA expression is recognised at the tips of secondary septal crests, representing the cross-sectioned ridges running between the alveoli surrounding the alveolar ducts.

[0175] The structure of the alveolar interstitial matrix is significantly compromised as a consequence of radiation exposure. Central among the growth factors that co-ordinate matrix tissue re-modelling is transforming growth factor-beta. This growth factor, which is ubiquitously expressed by all cells and tissues within the body, promotes extracellular matrix (ECM) deposition by stimulating different collagen, elastin, fibronectin and proteoglycan genes to produce ECM components. During synthesis two TGF-β precursor proteins form a dimer which is then cleaved by furin into two products, the first being a latency-associated peptide (LAP) and the other being mature TGF-β. These products thereafter maintain a non-covalent association forming a complex referred to as the small latent complex, which in turn covalently links to latent TGF-β binding protein (LTBP) to form the large latent complex which is then secreted and incorporated into the extracellular matrix as an inactive molecule. In addition to physical influences such as acidification or temperature changes TGF-β can be activated by proteases, by reactive oxygen species, or by interacting with thrombospondin or the αv-containing integrins (αvβ5, αvβ6, and αvβ8). Integrin αvβ5 is expressed by airway epithelial cells, endothelial cells, fibroblasts and monocytes in the lung, integrin αvβ8 by airway epithelial basal cells, and αvβ6 by airway epithelial cells. Activated TGF-β can then interact with its receptors leading to phosphorylation of transcription factors Smad2 and/or Smad3 which in turn associate and form a complex with Smad4 before translocating to the nucleus to influence the transcription of target genes and the production of ECM components. In a rat model of radiation-induced lung injury protein expression of integrin αvβ6, TGF-β1, TβRII, Smad3, and p-Smad2/3 was undetectable in the normal alveolar epithelium but increased in association with lung fibrosis six months after radiation exposure.

[0176] The present result in significant positive association between ASMA expression and collagen deposition in lung sections from sheep exposed to radiation.

[0177] Analysis by the inventors of expression of DC-LAMP found that radiation exposure was associated with clustering and an increase in size of DC-LAMP-positive cells. Type II pneumocytes are considered to proliferate in response to injury and serve as progenitors for replacing lost or damaged type I pneumocytes lining the alveolar surface.

[0178] Type II cells are well-recognised to be early susceptible targets of radiation effects and the inventors found that their reduced presence in their normal niches at alveolar corners was associated with an increase in size of the remaining cells found in these sites.

[0179] Pre-treating sheep with nebulised lamellar body compositions of the invention prior to each radiation exposure abrogated the increase in collagen seen in the PTV of sheep pre-treated with saline.

[0180] Lamellar body composition pre-treatment significantly increased the number of DC-LAMP positive cells in the lung relative to sheep pre-treated with saline, which contributed to a non-significant trend (P=0.067) towards an increase in the DC-LAMP area percentage. The lamellar body composition, in influencing the ability of type II cells to manage the interstitium, is proposed to change the proportion of myofibroblasts in this compartment in health, which would explain the significant treatment effect on ASMA.

[0181] Pre-treatment with the lamellar body composition discussed herein was associated with clustering of DC-LAMP positive cells and an increase in Ki67 count in non-radioexposed left lung (giving rise to significant interaction effects). The inventors considered the proximity of the anterior blocks to the cranial margin of the PTV might be a factor in dictating this difference between the treatment groups. Anterior blocks were selected as follows: The block containing the cranial margin of the PTV was identified. Progressing cranially, its immediate neighbour was disregarded and the next block along selected as LL (or RL)_Ant. As this procedure was consistent it was assumed that any variation in the spatial relationship between the selected blocks and the cranial margin of the PTV would be randomly spread between the SAL and lamellar body groups. However, when the inventors retrospectively examined these distances, they determined that there was a significant difference between the groups in terms of the distance of the “Ant” section from the cranial margin of the PTV—the lamellar body sections were approximately 7 mm closer (data not shown). Further, when the relationship between LL_Ant and RL_Ant DC-LAMP NND and Ki67 cell counts was expressed as absolute difference (LL-RL), and fold-change (1+((LL-RL)/RL) respectively, and compared to the distance of the section from the cranial edge of the PTV, significant correlations were found which would appear to explain the observation that pre-treatment with lamellar body composition was associated with a decrease in DC-LAMP NND and an increase in Ki67 count in non-radioexposed left lung (data not shown). This suggests a ramped decline in the biological effect of radiation (at least in terms of DC-LAMP cell clustering and cell proliferation) that extends beyond the margins of the PTV in this model.

[0182] TGF-β1 Model

[0183] An established in vitro model of myofibroblast activation using primary fibroblasts derived from the lungs of healthy donors and patients with idiopathic pulmonary fibrosis (IPF) was selected to investigate the anti-fibrotic effect of prepared lamellar body compositions outlined in the table below. The model involves a 96-well assay in which cells are treated with TGF-β1 in order to stimulate the differentiation of fibroblasts to myofibroblasts allowing for high-throughput screening of anti-fibrotic compounds using cells from multiple donors.

TABLE-US-00004 Formulation/ No Lipid Composition Mass Ratio LMS-611 DOPC/ESM/DOPE/DOPS/HSPI/Chol 55.5/19.4/8.2/4.1/3.1/10.1 (LMS-611) 1 DOPC/ESM/DOPE/DOPS/HSPI/Chol 55.1/19.4/8.2/4.1/3.1/10.1 (LMS-611 downsized to ca 125 nm) 2 DOPC/ESM/DOPE/DOPG/Chol 55.3/19.4/8.2/6.8/10.1 3 DOPC/ESM/DOPE/Chol 62.1/19.5/8.2/10.1 4 DOPC/ESM/DOPS/Chol 61.7/19.4/8.9/10.0 8 DOPC/ESM/DOPE/DOPS/HSPI/Chol/LysoPC 54.5/19.2/8.1/4.0/3.1/10.0/1.1

[0184] For the purposes of this study, fibroblasts from healthy donors were used at Passage 4. On Day 0, cells were seeded in 96-well plates and incubated at 37° C. for 48 hours. On Day 2, the cell culture medium was refreshed. On Day 5, the cells were treated with an 8-point concentration curve of each lamellar body formulation. The concentration curve was generated by two-fold serial dilutions of a top total lipid concentration of 1.5 mg/ml diluted with 0.9% saline solution. One hour post-treatment, the cells were stimulated with 1.25 ng/ml TGF-β1. The cells were incubated for a further 72 hours.

[0185] To confirm the assay was functional, 1nM SB-525334 was used a positive control for anti-fibrotic activity. SB525334 is an inhibitor of ALK5 (TGF-β Receptor 2) and inhibits TGF-β1 signalling. Cells were treated with SB-525334 for one hour in parallel to the lamellar body treatments. Cells treated with 0.1% DMSO or 0.9% saline (3% of the final volume per well) served as vehicle controls for the SB-525334 and lamellar body formulations, respectively. Nintedanib, a drug clinically approved for treatment of IPF, was also used as a reference compound against which to compare the efficacy of lamellar body formulations. An 8-point concentration curve was used, with a top concentration of 10 μM.

[0186] On Day 8, 72 hours post-stimulation with TGF-β1, cells were fixed with 4% formaldehyde.

[0187] Analysis of myofibroblast differentiation in response to TGF-β1 was carried out by means of confocal microscopy imaging of fluorescent staining of alpha smooth muscle actin (αSMA) and DAPI staining of the cell nuclei. This method measures the density times the area (D×A) of αSMA staining. The number of nuclei staining positive for DAPI serves as an indicator of possible compound cytotoxicity.

[0188] The assay was deemed to be suitable for the analysis of potential anti-fibrotic effects of LMS-611 composition and other novel formulations of lamellar bodies. Determining formulations (1, 2 and 8) induced a partial inhibition of TGF-β1-induced αSMA, the DOPS-enriched Formulation 4 caused a full dose-dependent inhibition in upregulation of αSMA in response to TGF-β1 stimulation.

[0189] The following table summarises the formulations tested and whether or not they were able to demonstrate an antifibrotic effect in this FMT assay model.

TABLE-US-00005 Formulation Antifibrotic No Lipid Composition Mass Ratio Effect * LMS-611 DOPC/ESM/DOPE/DOPS/HSPI/Chol (LMS-611) 55.5/19.4/8.2/4.1/3.1/10.1 Inconclusive** 1 DOPC/ESM/DOPE/DOPS/HSPI/Chol (LMS-611 55.1/19.4/8.2/4.1/3.1/10.1 Y downsized to ca 125 nm) 2 DOPC/ESM/DOPE/DOPG/Chol 55.3/19.4/8.2/6.8/10.1 Y 3 DOPC/ESM/DOPE/Chol 62.1/19.5/8.2/10.1 N 4 DOPC/ESM/DOPS/Chol 61.7/19.4/8.9/10.0 Y 8 DOPC/ESM/DOPE/DOPS/HSPI/Chol/LysoPC 54.5/19.2/8.1/4.0/3.1/10.0/1.1 Y * based on results from 2 human donors **results from one donor showed inhibition however data from a second donor did not.

[0190] Cell Entry Model

[0191] To elucidate cellular interaction, all prepared formulations, of lamellar body formulations, were labelled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (Dil), a lipophilic non exchangeable fluorescent lipid label. Cellular interactions were measured by flow cytometry using trypan blue, a quencher of extracellular fluorescence, to discriminate between cell association and internalisation of the vesicles. This is a standard technique as described earlier (Sahlin et al., 1983; Feldmann et al., 2017). All variants were tested in HeLa cells with some also tested in A549 cells. The selected cell lines were taken as model cell lines to illustrate that the formulations are taken up by cells.

[0192] Preparation Methods

[0193] Dual centrifugation was used to effectively homogenise a lipid/water blend to form a vesicular phospholipid gel (VPG) and after subsequent dilution of the VPG to prepare lamellar body formulations. This process method is described in (Massing, U., Ingebrigtsen, S. G., S̆kalko-Basnet, N., Holster, A. M., 2017. Dual Centrifugation—A Novel “in-vial” Liposome Processing Technique, in: Catala, A. (Ed.), Liposomes. InTech. https://doi.org/10.5772/intechopen.68523). Other methods such as extrusion and microfluidisation can be used to prepare lamellar body aqueous dispersions as well.

[0194] Lipid mixtures were prepared by mixing of dissolved lipids with 0.5 mol % Dil with respect to total lipid amount in a suitable organic solvent or solvent mixture followed by removal of the solvent by drying under vacuum. Aqueous dispersions of the lamellar body formulations were prepared by hydration of the dry lipid film in 250 mM sucrose and 25 mM sodium chloride and processed in a dual centrifuge (ZentriMix 380R, Andreas Hettich GmbH, Germany) at 1200 rpm, 15° C. for 20 min. The resulting vesicular phospholipid gel (VPG) was diluted with aqueous medium and processed again at 1200 rpm, 15° C. for 5 min. Ceramic beads were used as mixing aids in the vials. Finally, the formulations were further diluted to the required concentration.

[0195] Characterisation of Preparations

[0196] Phospholipid concentrations after extrusion were determined using the Bartlett assay (Bartlett, 1959 Bartlett, G. R., 1959. Phosphorus assay in column chromatography. J. Biol. Chem. 234, 466-468.). For all formulations, vesicle size and size range were measured. The composition, size range and zeta potential of the tested lamellar body formulations are provided in the following Table (Characteristics of Different Formulations—Note: All formulations were labelled with 0.7 weight % of Dil.).

[0197] Characteristics of Different Formulations

TABLE-US-00006 Zeta Identifica- Size Potential Lipids Ratio w/w tion Label (nm) (mV) DOPC/ESM/DOPE/DOPS/ 54.8/19.3/8.1/4/3.1/10 LMS-611 123 −24 HSPI/Chol DOPC/Chol 89.5/9.8 Not Applicable, 124 0 used as Neutral Control DOPC/ESM/DOPE/DOPG// 54.9/19.4/8.2/6.8/10 LMS_DOPG 125 −21 Chol DOPC/ESM/DOPE/DOPS/ 54.9/19.4/8.2/4/2.9/10 LMS_DOPS/DSPG 126 −24 DSPG//Chol DOPC/ESM/DOPE/DSPG// 54.9/19.4/8.2/6.9/10 LMS_DSPG 127 −24 Chol DOPC/ESM/DOPS/Chol 69.5/14.9/9.9/5 009 113 −25 DPPC/DOPE/DOPS/Chol 26.5/32.3/29.3/11.2 010 131 −58 DOPC/DOPE/DOPS/Chol 27.9/31.7/28.8/11 011 108 −55 DOPC/DOPS/Chol 60.3/28.2/10.8 012 110 −53

[0198] In vitro Uptake Experiments

[0199] For cellular uptake experiments 6.5.Math.10.sup.4 cells (A549 and HeLa) were seeded into the wells of a 24 well plate. After 24 hours the medium was changed and cells were incubated with the lamellar body formulations (0.15 mM) for 2 hours at 37° C. Cells were then analysed using a flow cytometer (BD LSRFortessa™ with BD FACSDiva software 8.01, Becton Dickinson, Germany) to assess the Dil fluorescence (excitation 561 nm, emission 585/15 nm).

[0200] A 0.08% aqueous solution of trypan blue, a quencher of the fluorescence of Dil, which cannot permeate cells, was used to discriminate between association and uptake of the lamellar bodies as described previously (Sahlin, S., Hed, J., Rundquist, I., 1983. Differentiation between attached and ingested immune complexes by a fluorescence quenching cytofluorometric assay. J. Immunol. Methods 60, 115-124.; Feldmann, D. P., Xie, Y., Jones, S. K., Yu, D., Moszczynska, A., Merkel, O. M., 2017. The impact of microfluidic mixing of triblock micelleplexes on in vitro/in vivo gene silencing and intracellular trafficking. Nanotechnology 28, 224001. https://doi.org/10.1088/1361-6528/aa6d15). The correction for spectral overlap of Dil and trypan blue was carried out with BD FACSDiva software.

[0201] Uptake experiments were analysed by normalising the Dil fluorescence intensities of the tested formulations to the Dil fluorescence of neutral control liposomes (DOPC/Chol).

[0202] Results

[0203] Lamellar body vesicle formulation LMS-611 was able to enter both cell lines. Fluorescence intensities were 40-fold (A549 and HeLa) higher compared to DOPC/Chol control liposomes. Additionally, in both cell lines LMS-611 lamellar body formulations were internalised to a high degree (93% of fluorescence was not quenched by trypan blue). The lamellar body formulations all demonstrated significantly greater cell entry than the DOPC/Chol control liposomes.

[0204] FIG. 9 depicts cellular association (red bars) and internalisation (blue bars) of formulations in A549 cells (A) and HeLa cells (B) after a 2 hour incubation period. The fold change of Dil fluorescence is normalised to neutral DOPC/Chol liposomes.

[0205] Cellular entry of lamellar body formulations was observed with formulations prepared from six, five or four or three lipids, containing negatively charged lipids. The phospholipids in the formulations comprise esterified saturated and unsaturated fatty acids. The example illustrates that lamellar body formulations are suitable to be taken up in general by cells. Thus, it is considered these compositions can act at the site of pathology within the cell to minimise or prophylactically treat fibrotic conditions.

[0206] Although the invention has been particularly shown and described with reference to particular examples, it will be understood by those skilled in the art that various changes in the form and details may be made therein without departing from the scope of the present invention.