key: cord-0962272-6h6568n8 authors: Jackson, Mark R; Stevenson, Katrina; Chahal, Sandeep K; Curley, Emer; Finney, George E; Gutierrez-Quintana, Rodrigo; Onwubiko, Evarest; Rupp, Angelika F; Strathdee, Karen; MacLeod, Megan KL; McSharry, Charles; Chalmers, Anthony J title: Low-dose lung radiotherapy for COVID-19 lung disease: a preclinical efficacy study in a bleomycin model of pneumonitis date: 2021-03-03 journal: bioRxiv DOI: 10.1101/2021.03.03.433704 sha: cdabfb1388de26d3acd228767ef6be92885d0f78 doc_id: 962272 cord_uid: 6h6568n8 Purpose Low-dose whole lung radiotherapy (LDLR) has been proposed as a treatment for patients with acute respiratory distress syndrome associated with SARS-CoV-2 infection and clinical trials are underway. There is an urgent need for preclinical evidence to justify this approach and inform dose, scheduling and mechanisms of action. Materials and methods Female C57BL/6 mice were treated with intranasal bleomycin sulphate (7.5 or 11.25 units/kg, day 0), then exposed to whole lung radiation therapy (0.5, 1.0, 1.5 Gy or sham, day 3). Bodyweight was measured daily and lung tissue harvested for histology and flow cytometry on day 10. Computed tomography (CT) lung imaging was performed pre-radiation (day 3) and pre-endpoint (day 10). Results Bleomycin caused pneumonitis of variable severity which correlated with weight loss. LDLR at 1.0 Gy was associated with a significant increase in the proportion of mice recovering to 98% of initial bodyweight and a proportion of these mice exhibited less severe histopathological lung changes. Mice experiencing moderate initial weight loss were more likely to respond to LDLR than those experiencing severe initial weight loss. Additionally, LDLR (1.0 Gy) significantly reduced bleomycin-induced increases in interstitial macrophages, CD103+ dendritic cells and neutrophil-DC hybrids. Overall,bleomycin-treated mice exhibited significantly higher percentages of non-aerated lung in left than right lungs and LDLR (1.0 Gy) prevented further reductions in aerated lung volume in right but not left lungs. LDLR at 0.5 and 1.5 Gy did not modulate bodyweight or flow cytometric readouts of bleomycin-induced pneumonitis. Conclusions Our data support the concept that LDLR can ameliorate acute inflammatory lung injury, identify 1.0 Gy as the most effective dose and provide preliminary evidence that it is more effective in the context of moderate than severe pneumonitis. Mechanistically, LDLR at 1.0 Gy significantly suppressed bleomycin-induced accumulation of pulmonary interstitial macrophages, CD103+ dendritic cells and neutrophil-DC hybrids. To date (February 2021), infection with the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has been associated with more than 2,500,000 deaths globally (1) . Infection by SARS-CoV-2 can cause a clinical syndrome termed COVID-19 which varies widely in severity with a small proportion of patients developing severe pneumonia and life-threatening acute respiratory distress syndrome (ARDS) (2) . COVID-19 lung disease is characterised by pathological inflammation, the severity of which correlates with morbidity and mortality (3) . The characteristic lung pathology and associated systemic deterioration are consistent with ARDS and cytokine release syndrome (CRS) respectively (4). Key features include florid alveolar infiltration by neutrophils, macrophages and lymphocytes and markedly increased pro-inflammatory cytokines including interleukin-6 (IL-6), IL-1, tumour necrosis factor (TNF) and interferon gamma (IFN) (5,6). Treatment options for COVID-19 remain limited. The open-label RECOVERY trial (UK) reported that dexamethasone reduced 28-day mortality compared with standard of care among hospitalised patients who required ventilatory support (Relative risk [RR] 0.65) or oxygenation (RR 0.80) (7). More recently, the REMAP-CAP trial reported improvements in survival and time to recovery following dual therapy with tocilizumab and sarilumab (unpublished data), and the anti-inflammatory agent colchicine has been reported to reduce hospitalisation and mortality in patients with COVID-19 infection who had a least one risk factor for complications (8). While some of these studies await peer review, the early efficacy data support the concept of acute inflammation being the critical pathological process in COVID-19 lung disease, and indicate that broad spectrum immunosuppressive therapies may be of therapeutic value. Low-dose whole lung radiotherapy (LDLR; radiation doses of 1.5 Gy or less) was used extensively as a treatment for pneumonias of various aetiologies in the pre-antibiotic era (9). In various preclinical models of inflammation, LDLR has been shown to induce anti-inflammatory cytokine production, reduce leucocyte-endothelial adhesion, and repolarise myeloid and lymphoid cells towards immune-suppressive phenotypes (10) . 6 protein S is protective of bleomycin pneumonitis (24) and has been proposed as a potential treatment for . To test the hypothesis that LDLR would reduce the severity of bleomycin-induced acute lung injury by exerting suppressive effects on cellular and molecular components of the inflammatory response, we measured the effects of 0.5, 1.0 and 1.5 Gy whole lung irradiation on bodyweight (primary endpoint); pulmonary cytology and histology, and lung computed tomography (CT) appearances (secondary endpoints). Our data show that 1.0 Gy LDLR enhances recovery in a proportion of bleomycin-treated mice, with corresponding improvements in lung histopathology and imaging parameters and modulation of specific immune cell populations. All reagents were purchased from Biolegend unless otherwise stated. Bleomycin sulphate was obtained from European Pharmacopeia EDQM, Council of Europe, France. Intranasal bleomycin dose was 11.25 units/kg except in the initial pilot study, when 7.5 units/kg was also used. Bleomycin generates a well-established murine model of pneumonitis (20) with a dynamic pathology similar to that of COVID-19. Female, 11-13 week old C57BL/6 mice (Charles River Laboratories) were administered one intranasal 40 L dose of bleomycin sulphate (7.5 or 11.25 units/kg) or phosphate buffered saline (PBS) vehicle control under light isoflurane anaesthesia. Mice were maintained in a pathogen-free facility, provided with additional high calorie, soft diet to ease feeding, and monitored daily for wellbeing and change in bodyweight. Those demonstrating signs of illness such as lethargy, isolation, reduced mobility, altered respiration, or ≥25% weight-loss were humanely culled. The experimental design optimised mice numbers to comply with the principles of Replacement, Reduction and Refinement for humane animal research. Procedures were governed by the Animals Scientific Procedures Act 1986 and approved by Home Office licence PP6245051. Bleomycin-treated mice exhibiting a day 3 relative bodyweight area-under-the-curve (AUC) 2.92 were randomised to receive LDLR or sham irradiation. Anaesthetised mice were irradiated with low-dose whole lung radiotherapy (0.5, 1.0, 1.5 Gray or sham) on day 3, using the Small Animal Radiation Research Platform (SARRP) developed by XStrahl. A 220 kVp, 13 mA X-ray beam was used with a dose rate of approximately 280 cGymin -1 at the chosen aperture size. Treatment was delivered with anterior and posterior parallel opposed fields. The broad focal spot (5.5 mm) was used and the SARRP's motorised variable collimator set to an aperture size of 20 x 20 mm to ensure full coverage of both lungs. Mice were culled on day 10 and lung tissue harvested for experimental endpoint analysis as described below. Lung changes were measured using the SARRP's in-built cone-beam CT (CBCT) function to image anaesthetised mice on days 3 (pre-irradiation) and 10 (experimental endpoint). Images were reconstructed using the FDK (Feldkamp, Davis and Kress) CBCT reconstruction algorithm from 1440 projections taken at 60 kVp and 0.8 mA using the fine focal spot (1 mm). For quantification of aerated lung volumes, Hounsfield unit (HU) clinical ranges were used: poorly aerated lung defined as -500 to -100 HU and normo-aerated defined as -900 to -500 HU. Images were analysed using the Lung CT analyzer module from the 3D Slicer software extension SlicerCIP (26,27). Mice were culled by terminal intraperitoneal injection of 100 L sodium pentobarbital (200 mg/ml) and cardiac exsanguination. The trachea was exposed, a small transverse opening cut between cartilage rings and a ligature tied loosely distal to the cut. The protruding 0.5 cm tip of a cannula sheath around a 23G syringe needle was inserted into the opening and the ligature tightened. The lungs were lavaged twice with 0.8 mL PBS and then perfused via the right ventricle with cold PBS until they blanched, after which lungs and heart were removed en bloc. The left lobe of the lung was excised, submerged in 4% neutral buffered formalin fixative for 24hr and processed for histology while the right lung lobes were processed for cytology. Serial 4 μm sections of the left lobe were cut and stained with haematoxylin and eosin (H&E) and Masson's trichrome and evaluated independently by a veterinary pathologist and a pulmonary immunologist both of whom were blinded to the experimental treatment. In brief, semiquantitative scoring (described in detail in Supplementary data) examined the extent of interstitial mononuclear cell infiltrates, specifically interstitial (to intra-alveolar) macrophage infiltrates and perivascular/peribronchiolar lymphocyte aggregates. Statistical analyses were performed in R 3.6.3 (29) using the packages "MESS" (30), "survival" (31) and "survminer" (32). Box and whiskers were plotted according to the Initially, pilot studies were conducted to characterise the bleomycin-induced pneumonitis model and establish optimum dosing and scheduling parameters. Using mouse bodyweight as a marker of systemic response, we observed variable responses to intranasal bleomycin (Fig. 1A) , as reported in other studies. Area-underthe-curve (AUC) analysis revealed that, despite visible effects at day 3 (Supp. Fig. S2 ), by day 10 the bodyweight of mice treated with 7.5 units/kg bleomycin was not significantly different to controls (Fig. 1B) . In contrast, administration of 11.25 units/kg induced progressive weight loss in the majority of mice, with 25% exhibiting a severe reduction that triggered humane endpoint sacrifice but 25% failing to show a demonstrable response (Fig. 1A,B) . Histological assessment on day 3 revealed multifocal, small, interstitial to intra-alveolar macrophage infiltrates and appreciable but small, perivascular and peribronchiolar lymphocyte aggregates in the majority of mice (Fig. 1C ). These were accompanied by a robust reduction in alveolar macrophages and an increase in interstitial macrophages measured by flow cytometry in dispersed lung tissue (Fig. 1D ). To mirror the clinical scenario, in which LDLR would only be considered in patients exhibiting moderate to severe COVID-19 lung disease, we opted to deliver LDLR three days after bleomycin treatment. Mice showing minimal weight loss at day 3 were excluded and those exhibiting a sustained drop in bodyweight (defined by day 3 AUC 2.92) were randomly allocated to receive LDLR or sham irradiation (Fig. 1E ). Treatment of vehicle-only control mice with LDLR (1.0 Gy) was well tolerated with no effect on bodyweight and no detectable deviation from normal behaviour (Fig. 1F ). Despite the variability inherent to the bleomycin model, treatment with 1.0 Gy was associated with a modest increase in mean bodyweight in irradiated versus sham- compared with sham-irradiated mice (3.3%, n=30; p=0.0265), with recovery also occurring earlier (Fig. 2C ). This definition of recovery (regaining 98% of initial weight) was used as a reference in subsequent analyses. Recovery was also significantly increased in irradiated mice if a recovery threshold of 100% was imposed (p=0.0230) and a strong trend was observed at 96% (p=0.0776) ( Table 1) S4 ). Early fibrotic changes were observed (Supp. Fig. S5 ) but were deemed not substantial enough to be quantified with existing scoring systems for fibrosis, which have been created and validated for later timepoints than those under investigation in this study. Immunocytological assessment of mouse lungs on day 10 showed that while the bleomycin-induced increase in interstitial macrophages was significantly blunted by lung irradiation (Fig. 4A ), the associated reduction in alveolar macrophages was not affected (Fig. 4B ). Bleomycin associated increases in CD103+ dendritic cells and neutrophil-DC hybrids (33) were also significantly attenuated in mice exposed to 1.0 Gy LDLR (Fig. 4C,D) . Representative FACS plots are shown in Supp. Fig. S6 . In addition to changes in cell number, bleomycin inhalation was associated with increased expression of the co-stimulatory molecule CD86 on alveolar macrophages and on neutrophil-DC hybrids, but reduced expression on CD103+ dendritic cells and interstitial macrophages (Supp. Fig. S7 ). Importantly, the reduction in expression of CD86 induced by bleomycin in interstitial macrophages was significantly attenuated by LDLR. To enable longitudinal assessment of lung infiltration, mice underwent CT imaging of the thorax on day 3 (pre-irradiation) and day 10. As previously reported, bleomycin related changes were significantly more pronounced in the left lung (Fig. 5A ,B); this is thought to be due to morphological differences between left and right main bronchi (34). Left and right lung datasets were therefore analysed separately. Consistent with evolving acute lung injury, aerated lung volume decreased between days 3 and 10 in sham irradiated mice (both lungs) and in the left lungs of irradiated mice (Fig. 5C ). In contrast, no statistically significant deterioration was observed in the right lungs of irradiated subjects (Fig. 5C, right panel) . Furthermore the mean decrease in right lung aerated volume was significantly less in irradiated mice than controls (-3.8% and -11.9% respectively, Fig. 5D ). Indeed 36% (n=22) of irradiated mice showed an improvement (change >0%) in right lung aeration at day 10, compared to only 5% (n=19) of controls. No effect of irradiation was observed in the left lungs. These observations led us to question whether LDLR might be more effective in alleviating moderate (right lungs) than more severe lung pathology (left lungs). To interrogate this further we looked for correlations between the severity of the bleomycin response prior to irradiation (day 3), as indicated by bodyweight AUC, and the likelihood of response to LDLR. Of the mice receiving 1.0 Gy, those that went on to recover had significantly higher AUC values at day 3 than those that did not recover ( Fig. 5E ). This analysis also confirmed that there was no significant difference in mean pre-LDLR AUC between irradiated and sham irradiated mice, and that mice experiencing a severe initial response to bleomycin (low day 3 AUC) were more likely to go on to experience severe weight loss (humane endpoint), regardless of further treatment. Finally we evaluated two additional LDLR doses (0.5 and 1.5 Gy) that are also being tested in clinical trials. Neither dose was associated with an improvement in outcome compared to sham irradiation, either in terms of mean bodyweight or likelihood of recovery (Fig. 6A ). In keeping with this, no effect of these doses was observed on the immune cell subsets previously shown to respond to 1.0 Gy (Fig. 6B ). It should be noted, however, that in this experiment the bleomycin effects were less pronounced than in previous experiments (Fig. 6B, Supp. Fig S8) . (Fig. 4) Chemokines such as MIP-2 and CXCL5 are released in the first few days after acute lung injury and, together with other factors including extracellular ATP, may play a role in initiating and sustaining accumulation of neutrophils within the lungs following bleomycin inhalation (45). While we saw no increase in classical neutrophils in bleomycin exposed lungs, we did observe accumulation of a hybrid population that expressed markers of both neutrophils (Ly6G) and dendritic cells (CD11c, MHCII). These hybrid cells are thought to differentiate from neutrophil precursors, retaining their phagocytic function while gaining the ability to present antigen to CD4 T cells (33). Inflammation induced by thioglycolate, bacterial or fungal infection leads to an increase in this hybrid population in mouse models of tissue inflammation and these cells have also been found in human tumours (46-48). Our data extend these observations to show that bleomycin also drives accumulation and differentiation of these cells, an effect that we showed to be significantly blunted by LDLR. Our data provide preclinical evidence of efficacy of LDLR in a subset of mice with Relative bodyweight Mice whose bodyweight returned to 98% of baseline were classified as recovered. Mice sacrificed early due to an excessive reduction in bodyweight were classified as having experienced severe weight loss. * P<0.05, ** P<0.01, *** P<0.001. Number of interstitial macrophages 14 14 14 14 14 14 14 12 11 11 9 13 13 13 13 13 13 11 9 9 9 9 12 12 12 12 12 12 10 Johns Hopkins Coronavirus Resource Center Clinical Characteristics of Coronavirus Disease 2019 in China The trinity of COVID-19: immunity, inflammation and intervention Radiobiological mechanisms in inflammatory diseases of low-dose radiation therapy Low doses of radiation increase the immunosuppressive profile of lung macrophages during viral infection and pneumonia Low-dose whole-lung radiation for COVID-19 pneumonia: Planned day 7 interim analysis of a registered clinical trial Lack of supporting data make the risks of a clinical trial of radiation therapy as a treatment for COVID-19 pneumonia unacceptable Low dose radiation therapy for COVID-19: Effective dose and estimation of cancer risk Unethical not to Investigate Radiotherapy for COVID-19 Clinician Attitudes to Using Low-Dose Radiation Therapy to Treat COVID-19 Lung Disease