PRophylaxis Against Early VENTilator-associated Infections in Acute Brain Injury

Overview

This research is about whether treatment with a commonly used antibiotic can prevent infections in airway and lungs and improves the chance of surviving, if it is given soon after patients commence mechanical ventilation when they have been admitted to hospital with an acute severe brain injury. An acute severe brain injury can occur as a result of a stroke, a traumatic injury or due to lack of oxygen to the brain that happens as a result of a cardiac arrest. Patients who are unconscious after an acute severe brain injury often need assistance to breath adequately, and this assistance is given by a breathing tube, connected to a mechanical ventilator. This treatment is an emergency medical treatment. The breathing tube is inserted into the patients' airway by either their mouth or neck. For patients who need assistance with their breathing from a mechanical ventilator, infections in the airways and lungs, known as pneumonia, are a common complication. Everyone naturally has bacteria in their mouth, esophagus and stomach. Clinicians think that during the process of inserting the breathing tube, small amounts of these bacteria can be introduced into the airways and lung when people are unconscious following an acute severe brain injury, or during the process of placing the breathing tube into the airways. These bacteria are now in a place they aren't meant to be and can cause an infections in the airways and lungs known as pneumonia. The purpose of this research is to see if giving one dose of a common antibiotic can prevent patients developing pneumonia, which is associated with having a breathing tube inserted and being on a ventilator, improving the chance of recovery following the acute severe brain injury and ultimately improving the chance of surviving. When patients have a known infection, current guidelines are to treat them with antibiotics. Antibiotics work to kill the bacteria causing the infection. When a patient has an infection in their lungs, they often need to stay on the mechanical ventilator for longer. While current practice is to give patients with a proven infection in their airways and lungs (pneumonia) antibiotics, it is unknown if giving an antibiotic to patients to prevent these infections before they show signs of pneumonia may lead to better outcomes.

Key Concepts * Lower respiratory tract infections are a common complication in invasively mechanically ventilated ICU patients. * Micro-aspiration before or during endotracheal intubation may contaminate the tracheobronchial tree and lungs with bacteria that can subsequently cause a lower respiratory tract infection. * Patients with a depressed level of consciousness are at particular risk of developing such infections. * In patients who have been hospitalised for fewer than 3-4 days, common community-acquired bacteria including S. aureus, and H. influenzae are the pathogens isolated most frequently from the lower respiratory tract. * Several small, single centre pilot RCTs, have suggested that prophylactic antibiotics might be effective at preventing lower respiratory tract infections in invasively mechanically ventilated ICU patients with acute neurological conditions. * In a multicentre placebo-controlled randomised clinical trial (RCT) prophylactic antibiotics significantly reduced early ventilator-associated pneumonia (VAP) compared to placebo in patients who were mechanically ventilated and treated with targeted temperature management after an out of hospital cardiac arrest. * In another recent RCT (PROPHY-VAP), trial participants allocated to receive a single dose of 2g of ceftriaxone within 12 hours of endotracheal intubation, who were comatose and expected to require at least 48 h of invasive mechanical ventilation in the ICU, had a significantly lower rate of VAP than trial participants allocated to placebo. * These RCTs were not adequately powered to determine the effect of antibiotic prophylaxis on patient-important outcomes. However, in the PROPHY-VAP trial a total of 32 of 161 participants (20%) allocated to prophylactic ceftriaxone had died and a total of 46 of 157 (30%) allocated to placebo had died (hazard ratio, 0.66 (95%CI, 0.42-1.04); P=0.074). * Lower respiratory tract infections that occur within the first week of hospitalisation potentially have major consequences for ICU patients with acute neurological conditions. * Such infections occur at a time when critically ill patients with acute neurological conditions are at risk of secondary brain injury. These infections can cause fevers, hypoxaemia, hypercapnia, and elevated intracranial pressure, all of which may potentially contribute to secondary brain injury. * Patients with impaired airway reflexes due to neurological injuries may have difficulty managing respiratory secretions. The sputum burden associated with a lower respiratory tract infection may prolong the duration of invasive mechanical ventilation in brain-injured patients putting them at risk of unfavourable outcomes. Lower respiratory tract infections in this setting may also increase the risk of patients requiring a tracheostomy. * Overall, it is highly plausible that antibiotic prophylaxis around the time of intubation in ICU patients who have acute neurological injuries and conditions may prevent early lower respiratory tract infections and might, thereby, mitigate secondary brain injury, and reduce mortality. Background and rationale Respiratory tract infections are common in ventilated adults with acute neurological conditions Ventilator associated pneumonia or VAP, defined as an infection of the lung parenchyma occurring \>48 hours after endotracheal intubation, is common in patients in the ICU receiving invasive mechanical ventilation. A diagnosis of VAP requires a new lung infiltrate evident on chest radiograph plus clinical evidence that the infiltrate has an infectious origin. Such evidence can include new onset of fever, purulent sputum, and/or leukocytosis. Although pneumonia complicating intubation can occur before 48 hours have elapsed, chemical pneumonitis due to aspiration is a common cause of radiological pulmonary infiltrates in this period and does not necessarily progress to pneumonia. While micro-aspiration does not always result in infection, organisms that colonise the upper airway are frequently subsequently grown from tracheobronchial secretions. Moreover, micro-aspiration of mouth organisms around the time of intubation may be a common cause of lower respiratory tract infection in patients who are mechanically ventilated. When lower respiratory tract infections develop within the first 3-4 days of hospitalisation, common community-acquired pathogens including S. aureus, and H. influenzae are the most frequently isolated from tracheobronchial secretions. Patients with a diminished level of consciousness, including those with traumatic brain injury and other acute neurological conditions, appear to be at particular risk of developing VAP. In such patients, impairment of airway reflexes for a period of time prior to securing the airway may be responsible for the heightened risk of VAP compared with other ICU patients. The daily risk of developing VAP appears to be greatest in the early period after intubation. This comparatively higher daily risk of developing early, as opposed to late VAP, may reflect the particular role of the introduction of bacteria into the lower respiratory tract around the time of intubation. Even among patients who do not develop pulmonary infiltrates, and thus do not fulfil the definition of VAP, purulent deep respiratory secretions developing in association with fever are common in invasively mechanically ventilated patients. While the role antibiotic treatment in patients who do not fulfil the criteria for VAP is uncertain, patients with these clinical findings are often treated with antibiotics and, frequently, no non-respiratory source of infection is apparent. It is plausible that even in the absence of an infection fulfilling the formal definition of VAP, a lower respiratory tract infection occurring in a mechanically ventilated patient may be clinically consequential. Consequences for ventilated adults with neurological conditions and respiratory infections. It has been estimated that the attributable mortality of VAP is approximately 13%, reflecting both the natural history of the disease as well as the effectiveness of therapies including antimicrobials and advanced organ support. While it is rare to encounter patients in clinical practice where death clearly occurs as a direct consequence of an ICU-acquired lower respiratory tract infection, such infections appear to be associated with prolongation of invasive mechanical ventilation by an average of ≈7 days, ICU length of stay by ≈8 days, and hospitalisation by ≈12 days. In neurologically injured patients who are invasively mechanically ventilated in the ICU, early lower respiratory tract infections can have major specific clinical consequences. Neurologically injured patients with lower respiratory tract infections may develop sputum plugging which impairs ventilation leading to pathological elevations of intracranial pressure. These elevations in intracranial pressure, particularly when coupled with hypoxaemia, may contribute to secondary brain injury. Such infections can also cause fevers, which may also contribute to secondary brain injury. For neurologically injured patients, it is plausible that lower respiratory tract infections developing when patients are at high risk of secondary brain injury, adversely affect survival and long-term functional recovery. The sputum burden associated with a lower respiratory tract infection may be particularly likely to prolong invasive mechanical ventilation in brain-injured patients because such patients often have impaired airway reflexes. Prolonged ventilation itself may put patients at risk of unfavourable outcomes and may also result in more patients requiring tracheostomy to facilitate respiratory weaning. Preliminary data suggest that prophylactic antibiotics may be effective Several RCTs have been conducted evaluating the role of prophylactic intravenous antibiotics in the prevention of VAP in patients with acute severe neurological injury. In a single centre open-label prospective RCT in mechanically ventilated patients with head injury, stroke, or following surgery for space-occupying lesions, two doses of 1.5g of cefuroxime administered 12 hours apart significantly reduced microbiologically confirmed pneumonia, 70% of which occurred within the first four days following intubation. A total of 12 of 50 (24%) participants and 25 of 50 (50%) allocated to antibiotic prophylaxis and control respectively developed pneumonia (P=0.007). In another single open-label RCT in comatose mechanically ventilated patients with traumatic, surgical or medical brain injury, 3g of intravenous ampicillin-sulbactam given every 6h for 3 days following intubation significantly reduced the occurrence of pneumonia in the first four days of mechanical ventilation compared to usual treatment. A total of 4 of 19 (21.0%) and 11 of 19 (57.9%) participants allocated to antibiotic prophylaxis and control respectively developed early VAP (Relative risk, 0.36 (95% CI, 0.14-0.94); P=0.02). Two RCTs have been conducted in comatose post-cardiac arrest patients. In the first, 60 patients were enrolled in a single centre. This trial focussed on the effect of 1.2g of intravenous amoxicillin + clavulanic acid given every 8 h on the systemic inflammatory response (estimated based on the white cell count, C reactive protein, procalcitonin, and CD 64 concentrations from day 1 to 7). While some minor differences in inflammatory response markers between groups were evident, the clinical importance of these was unclear. However, consistent with the earlier studies a mini-bronchoalveolar lavage specimen taken on day 3 was positive in 2 of 27 (7%) and 10 of 24 (42%) participants in the prophylactic antibiotic and control groups respectively (P\<0.01). The second RCT, is the Antibiotherapy during Therapeutic Hypothermia to prevent Infectious Complications (ANTHARTIC) trial. Published in the New England Journal of Medicine, this trial demonstrated that, among patients who were mechanically ventilated and treated with targeted temperature management following out of hospital cardiac arrest, 1.2g of amoxicillin + clavulanic acid 8 hourly for two days significantly reduced early VAP compared to placebo. A total of 19 of 99 (19%) and 32 of 95 (35%) of participants in the prophylactic antibiotic and placebo groups developed early VAP (HR 0.53; 95% CI, 0.31-0.92; P=0.03). Mortality rates were similar by treatment group and no deaths considered attributable to VAP occurred. While neurological outcomes at six months and one year were also similar by treatment group, the ANTHARTIC trial lacked power to confirm or refute an effect of antibiotic prophylaxis on such outcomes. Finally, in the Prophylactic Antibiotics for VAP trial (PROPHY-VAP), 2g of prophylactic intravenous ceftriaxone in patients with acute brain injuries requiring mechanical ventilation in the ICU significantly reduced early VAP compared to placebo. A total of 23 of 162 allocated to ceftriaxone (14%) and 51 or 157 allocated to placebo (32%) developed early ventilated pneumonia (hazard ratio 0·60; 95% CI 0·38-0·95; p=0·03). A total of 32 of 161 in the ceftriaxone (20%) and 46 of 157 in the placebo group (30%) died by day 60 (hazard ratio 0·66; 95% CI 0·42-1·04; p=0·07). The effect of antibiotics on prevention of lower respiratory tract infections that do not fulfil formal criteria for diagnosis of VAP has not been studied. While data from existing RCTs suggesting antibiotic prophylaxis reduces VAP are promising, further research focusing on patient-important outcomes is needed before this treatment strategy can be implemented in clinical practice. Evidence to support the choice of antibiotic agent The data presented above provide a compelling case to support further research to assess the potential for prophylactic antibiotics to prevent early ventilator associated infection and improve patient centred outcomes. It is important to note that guidelines for the prevention of ventilator associated pneumonia do not recommend the use of prophylactic antibiotics nor do guidelines for the management of patients with acute severe brain injury due to hypoxic ischaemic encephalopathy, traumatic brain injury, subarachnoid haemorrhage and ischaemic stroke. The most recent trial of prophylactic antibiotics to prevent ventilator associated pneumonia in patients with acute severe brain injury compared a single dose of 2g ceftriaxone to placebo and, as noted above found a reduction in ventilator associated pneumonia and mortality at day 60 in the treatment group of 32/161 (20%) compared to 46 /157 (30%), p=0.07. Ceftriaxone is an appropriate choice for antibiotic prophylaxis in this population, as it has a spectrum of antibacterial activity that includes the common pathogens found in patients with early ventilator associated pneumonia, including Haemophilus influenzae, Staphylococcus aureus, Escherichia coli and Streptococcus pneumoniae. The pharmacokinetic profile of ceftriaxone, specifically the prolonged half-life, allows a single dose of ceftriaxone to provide sufficient coverage for antibiotic prophylaxis. When diluted to a concentration of 1g per 100mL, the solution is visually indistinguishable from normal 0.9%sodium chloride. 3.3 Implications Implications of recent trials of selective decontamination of the digestive tract (SDD) SDD, which consists of an oral antibiotic paste and a gastric antibiotic suspension, with or without a short course of intravenous antibiotics has been extensively studied in critically ill patients. In the Selective Decontamination of the Digestive Tract in the Intensive Care Unit (SuDDICU) Australia trial, among critically ill patients receiving mechanical ventilation, SDD, compared with standard care without SDD, did not significantly reduce in-hospital mortality. However, when SuDDICU trial data were combined with other data in a systematic review and Bayesian meta-analysis of 32 randomized trials that included 24,389 participants, there was a 99.3% posterior probability that SDD was associated with reduced hospital mortality compared with standard care. The mortality reduction associated with SDD was only evident in regimens that included an intravenous antibiotic component. In a secondary analysis of the SuDDICU trial data, in patients with an ICU admission diagnosis of acute brain injury, SDD was associated with significantly improved survival compared to standard care without SDD as well as increased ventilator free day, ICU and hospital free days. These benefits were not evident in patients without acute brain injury. Clinicians' concerns about the development of antimicrobial resistant organisms in the ICU environment remain an important obstacle to implementation of SDD as a standard of care. There is an imperative to conduct an RCT to test the hypothesis that IV antibiotic prophylaxis alone reduces mortality in patients with acute severe brain injuries who require invasive mechanical ventilation in the ICU. Basic principles of antibiotic stewardship support the notion that prophylactic antibiotic use should be targeted to the patients most likely to benefit from such prophylaxis and that the narrowest spectrum effective agent should be used. Accordingly, the PREVENT-NEURO Trial will assess the effect of intravenous antibiotic prophylaxis alone, using the narrowest spectrum and duration that is likely to be effective in the group of patients who appear most likely to benefit (i.e. patients with acute severe brain injuries). Implications of the proposed trial for antibiotic stewardship and antibiotic resistance Available data are insufficient to exclude the possibility that prophylactic peri-intubation antibiotics in ICU patients with acute brain injuries and conditions affect antimicrobial resistance. However, in the ANTHARTIC trial, the median percentage of days with antibiotic use during the ICU stay, after completion of the trial treatment, tended to be lower in the antibiotic group than in the control group (23% \[interquartile range, 0 to 64\] vs. 50% \[0 to 70\]; median difference, 0 days; 95% CI, -15 to 0). An antibiotic prophylaxis strategy that reduces overall antibiotic use is unlikely provoke major concerns in relation to antibiotic stewardship and may not have significant implications for development of antibiotic resistance, but further research is required to evaluate this issue. Problems with recording lower respiratory tract infections or VAP as outcomes The investigators have carefully considered the possibility of recording lower respiratory tract infections or ventilator associated pneumonia as trial outcomes. However, such outcomes are extremely labour intensive to adjudicate, are subjective, have poor sensitivity and specificity. Moreover, because antibiotics suppress growth of bacteria in the laboratory, the investigators intervention may uncouple the diagnosis of lower respiratory tract infections from the patient-important consequences of such infections. Accordingly, the investigators will focus on patient-important outcome measures like mortality and functional recovery and will record the physiological and process of care measures such as duration of mechanical ventilation and requirement for reintubation and requirement for a tracheostomy that may be impacted by acquisition of a ventilator associated lower respiratory tract infection. The epidemiology of acute brain injuries in Australian and New Zealand ICUs Patients with acute brain injuries account for around 13% of all unplanned ICU admissions in Australia and New Zealand. However, patients with such brain injuries account for a disproportionally high number of the in-hospital deaths among patients who require unplanned ICU care. Data from the Australian and New Zealand Intensive Care Society Centre for Outcome and Resource Evaluation Adult Patient Database (ANZICS-CORE APD) indicate that from 2013 to 2022, a total of 24,540 or 93,664 (26.2%) of patients admitted to the ICU with acute brain injuries died in hospital; the mortality rate for patients with other reasons for unplanned ICU admissions for the same period was 63,510 of 601,433 (10.5%). Linkage to death registry data showed that by day 90 the mortality rate for all brain injured patients, including those not requiring invasive mechanical ventilation, rose to 33% and for other patients was 11%. Health-economic implications of antibiotic prophylaxis In a health economic analysis of the ANTHARTIC trial, over a 3-month time horizon, hospital costs were significantly lower in patients allocated to prophylactic antibiotics with a median cost reduction of €7,193 per patient. Patients with acute neurological conditions who survive with dependency require ongoing care. For example, approximately 1 in 20 patients who are admitted to the ICU following a cardiac arrest will have moderately severe (mRS4) or severe disability (mRS5) at six months corresponding to being "unable to walk and attend to bodily needs without assistance" and being "bedridden, incontinent and requiring constant nursing care and attention" respectively. Such patients typically require hospital-level nursing home care, which costs around $1300 per week ($68k p.a.). Approximately another 1 in 20 patients have moderate disability (mRS3), corresponding to "requiring some help, but able to walk without assistance". While it is difficult to accurately estimate ongoing costs for this "moderate-disability" group, it is clear that if prevention of early lower respiratory tract infections can alter the extent of secondary brain injury and prevent such disability, the health-economic implications of this strategy would be substantial. 3.4 Clinical significance There are more than 20,000 acute brain injury admissions to ICUs in Australia and New Zealand each year. Survivors are commonly left with lifelong severe physical and mental health disability. Providing care for survivors of severe brain injury requires prolonged hospitalisation and extensive rehabilitation that is costly to the health system and community alike. There has been little improvement in patient-centred and functional outcomes in this population over recent years. Short-term, targeted, prophylactic antibiotics to prevent ventilator-associated lower respiratory tract infections in brain-injured patients has the potential to deliver a substantive health impact by reducing mortality and improving functional neurological outcomes. Confirmation of the benefits of this strategy, suggested by the SuDDICU-neuro analysis, would mean a number needed to treat to prevent one death of 17: a clinically important benefit from a simple, cost-effective and readily implementable intervention, translating to more than 1200 lives saved annually. For patients who survive a severe acute brain injury, mitigating secondary brain damage through infection prevention may prevent the need for decades of care, improve the quality of life of survivors, allow people to return to the workforce, and reduce healthcare costs.