Respiratory complications of organophosphorus pesticide poisoning

Abstract

Of the 800,000 suicides recorded globally every year, over a third are due to pesticide ingestion, the majority of which occur in rural Asia with organophosphorus (OP) compounds. These anticholinesterase pesticides cause an acute cholinergic syndrome characterised by decreased consciousness, excessive airway secretions and respiratory failure. A combination of these clinical features is the most common cause of death. Up to 30% of OP pesticide poisoned patients are admitted to the Intensive Care Unit (ICU) for tracheal intubation and lung ventilation, but up to half die. It is not understood why the case fatality for intubated poisoned patients is so high, but one hypothesis is that the patients, when unconscious, aspirate their stomach contents (including the OP and the solvent present in its agricultural formulation) causing a toxic lung injury which contributes to the observed high mortality. In this PhD, I aimed to characterise the lung injury caused by OP pesticide self-poisoning through both indirect (ingestion) and direct (aspiration) means. To achieve this, I analysed data from previous toxicological minipig work and designed and conducted a specific minipig pulmonary aspiration study which was complemented by an experimental OP poisoning ex vivo lung perfusion model and human data from pesticide poisoned patients in Sri Lanka. I first investigated the pulmonary pathophysiology resulting from orogastric administration of OP pesticide without aspiration. Analysis of my group’s Gottingen minipig in vivo work demonstrated that orogastric placement of agricultural OP (dimethoate EC40) produced lung injury via exposure to blood-borne pesticide. Pathological lung changes consisted of alveolar and interstitial oedema, pulmonary haemorrhage and modest neutrophilia with increased concentrations of protein, IL-6 and IL-8 when compared with controls, but with low concentrations of TNF-α and IL-10 in bronchoalveolar lavage fluid (BALF). In a second study, OP poisoned minipigs had increased concentrations of BALF protein, neutrophils, IL-8 and CRP six hours after orogastric poisoning when compared with their baseline values. Electron microscopy images of both studies demonstrated damage to the alveolar capillary membrane secondary to systemic OP poisoning. Prior to conducting the main pulmonary aspiration study in minipigs, there was considerable refinement of the processes involved through use of: (i) pilot aspiration and dose ranging studies; (ii) the development of a specific pulmonary histopathological scoring system; and (iii) employment of modern human anaesthetic equipment and intensive care patient management protocols. After this period of model development, an in vivo 48 hour study using Gottingen minipigs (n=26) was conducted to investigate the pulmonary pathophysiology in animals given either sham bronchoscopy (sham control) or 0.5 mL/kg of: saline (saline control), porcine gastric juice [GJ], OP (dimethoate EC40) + GJ [OP+GJ], or solvent (cyclohexanone) + GJ [Solv+GJ] into the right lung under bronchoscopic guidance. The results showed that in a minipig model OP and GJ placed into one lung created a direct (right) and indirect (left) lung injury significantly different to controls, and in some respects worse than GJ alone 48 hours after poisoning. The direct lung injury caused by OP+GJ was characterised by significantly worse pathology (p=0.0003) in terms of: pulmonary neutrophilia, alveolar haemorrhage, necrosis, oedema and fibrin deposition, when compared with sham controls at 48 hours. Lungs injured directly with OP+GJ also had significantly higher concentrations of BALF neutrophils (p≤0.01), protein (p≤0.05), IL-6 (p≤0.01), IL-8 (p≤0.01) and CRP (p≤0.05) at 24 hours, and BALF protein (p≤ 0.01), and CRP (p≤ 0.05) when compared with sham controls at 48 hours. The BALF from OP+GJ minipigs at 48 hours also had higher numbers of aerobic bacteria than other groups, suggesting the development of pneumonia could be a source of additional lung injury. Lung damage might also have resulted from a reduction in the surfactant component responsible for the lowering of alveolar surface tension. Direct lung injury with OP+GJ caused a proportional reduction of beneficial pulmonary surfactant phosphatidylcholine (PC) species 16:0/16:0 [29(±4) % vs. 38(±4) %] when compared with sham controls at 48 hours. Unlike the other groups, OP+GJ (direct and indirectly-injured) lungs had type 2 alveolar cell ultrastructural morphological differences in the lamellar bodies that stored the surfactant. The lamellar bodies were more numerous and more dense in the OP+GJ lungs compared with other groups and could signify a failure of surfactant release or some other pathology pertinent to OP aspiration lung injury. Computed tomography analysis showed that direct lung injury with OP+GJ caused significantly more lung tissue to be poorly or non-aerated [77 (±13) % ; p≤0.0001 when compared with sham] as opposed to 62 (±27) % in GJ, 53(±13)% in sham and 47(±0.2)% in saline control animals by 47.5 hours and was mainly due to pulmonary haemorrhage and oedema fluid. The key differences between aspiration of OP+GJ versus GJ alone was that the majority of inflammatory markers (e.g. BALF protein, IL-6 and CRP) appeared to increase from 24-48 hours in OP+GJ treated animals, but decreased in GJ pigs, possibly signifying resolution. Treatment with GJ alone produced less severe histopathological damage, bacterial BALF numbers and percentage of poorly and non-aerated lung tissue. Importantly, there was less evidence of indirect lung injury within the GJ pigs when compared with animals treated with OP+GJ. Solvent placed into the lung seemed to offer some form of protection from the effects of GJ aspiration. This was dramatically demonstrated by the histopathology scores, proportional percentage of beneficial phosphatidylcholine (PC) species 16:0/16:0 and the percentage of poorly and non-aerated lung tissue all approaching control animal levels by 48 hours in minipigs that had Solv+GJ placed in the directly-injured (right) lung. Further evidence of benefit was provided by statistically significant reductions (p≤ 0.05) in BALF concentrations of IL-8, IL-6 and CRP in minipigs which had aspirated Solv+GJ when compared with OP+GJ and/or GJ minipig groups at 24 hours. The pathophysiology of aspirated OP+GJ was also investigated in a pilot ovine ex vivo lung perfusion (EVLP) model (n=4). Lungs directly-injured with OP+GJ had higher concentrations of total protein (4300 mg/L vs. 350 mg/L) with a proportional reduction of beneficial pulmonary surfactant phosphatidylcholine species 16:0/16:0 (27% vs.34%) when compared with control lungs. Analysis of toll-like receptor (TLR) lung tissue expression in the OP+GJ directly and indirectly-injured lungs indicated that inflammatory mechanisms might also involve upregulation of TLR 3 and 5, unlike other lung injuries e.g. those induced with lipopolysaccharide, which typically upregulates TLR 2 and 4. To compare OP-induced lung injury in humans and the minipigs, a small feasibility study was conducted in the ICUs of the University of Peradeniya hospital, Sri Lanka. Unfortunately, ethics review and recruitment proved more difficult than expected and we failed to recruit to target. We did however find raised BALF concentrations of IL-6, IL-8 and CRP and low concentrations of TNF, IL-1β, IL-10 in intubated OP poisoned patients at 24 hours when compared with controls. We also found that two plasma micro-RNA biomarkers thought to be involved in inflammation and lung injury, MiR-21 and MiR-146a, had significantly reduced expression in OP-poisoned patients with aspiration compared to non-intubated control patients from the UK (p=0.008 and p=0.0083 respectively). The work from this thesis has allowed the characterisation of both indirect and direct lung injuries caused by OP pesticide ingestion and aspiration. The minipig model showed that at 48 hours the lung injury created by aspiration of OP+GJ appeared more severe than GJ alone, but the addition of the solvent cyclohexanone seemed protective and even beneficial in the context of GJ aspiration. The cytokine expression profiles from both the human and minipig work, combined with the preliminary TLR lung tissue analysis from the EVLP model, suggest that OP+GJ aspiration is unlike normal GJ aspiration and classic ARDS. Increased concentrations of aerobic bacteria in the minipig OP+GJ lungs at 48 hours and evidence of suppression of plasma miR-21 and miR-146a in OP poisoned patients could be linked, and may involve cholinergic immune system modulation. These molecular mechanisms need to be investigated further in both in vitro and in vivo models. These discoveries indicate the complex nature of the pulmonary injury that occurs after OP pesticide poisoning, and suggests that damage is not caused by gastric contents alone. Preliminary findings indicate that aspiration of OP+GJ could create favourable conditions for the development of aspiration or ventilator-associated pneumonia but this would need confirmation in larger clinical studies. The potential roles of micro RNA as a biomarker of OP poisoning and lung injury, and solvent as a therapy for aspiration should be explored in further pre-clinical studies.