A. Conway Morris

Evaluation of the effect of diagnostic methodology on the reported incidence of ventilator-associated pneumonia

Background: The optimal method for diagnosing ventilator-associated pneumonia (VAP) is controversial and its effect on reported incidence uncertain. This study aimed to model the impact of using either endotracheal aspirate or bronchoalveolar lavage on the reported incidence of pneumonia and then to test effects suggested from theoretical modelling in clinical practice. Methods: A three-part single-centre study was undertaken. First, diagnostic performance of aspirate and lavage were compared using paired samples from 53 patients with suspected VAP. Secondly, infection surveillance data were used to model the potential effect on pneumonia incidence and antibiotic use of using exclusively aspirate or lavage to investigate suspected pneumonia (643 patients; 110 clinically suspected pneumonia episodes). Thirdly, a practice change initiative was undertaken to increase lavage use; pneumonia incidence and antibiotic use were compared for the 12 months before and after the change. Results: Aspirate overdiagnosed VAP compared with lavage (89% vs 21% of clinically suspected cases, p<0.0001). Modelling suggested that changing from exclusive aspirate to lavage diagnosis would decrease reported pneumonia incidence by 76% (95% CI 67% to 87%) and antibiotic use by 30% (95% CI 20% to 42%). After the practice change initiative, lavage use increased from 37% to 58%. Although clinically suspected pneumonia incidence was unchanged, microbiologically confirmed VAP decreased from 18 to 9 cases per 1000 ventilator days (p = 0.001; relative risk reduction 0.61 (95% CI 0.46 to 0.82)), and mean antibiotic use fell from 9.1 to 7.2 antibiotic days (21% decrease, p = 0.08). Conclusions: Diagnostic technique impacts significantly on reported VAP incidence and potentially on antibiotic use.

Combined dysfunctions of immune cells predict nosocomial infection in critically ill patients

BACKGROUND Nosocomial infection occurs commonly in intensive care units (ICUs). Although critical illness is associated with immune activation, the prevalence of nosocomial infections suggests concomitant immune suppression. This study examined the temporal occurrence of immune dysfunction across three immune cell types, and their relationship with the development of nosocomial infection. METHODS A prospective observational cohort study was undertaken in a teaching hospital general ICU. Critically ill patients were recruited and underwent serial examination of immune status, namely percentage regulatory T-cells (Tregs), monocyte deactivation (by expression) and neutrophil dysfunction (by CD88 expression). The occurrence of nosocomial infection was determined using pre-defined, objective criteria. RESULTS Ninety-six patients were recruited, of whom 95 had data available for analysis. Relative to healthy controls, percentage Tregs were elevated 6-10 days after admission, while monocyte HLA-DR and neutrophil CD88 showed broader depression across time points measured. Thirty-three patients (35%) developed nosocomial infection, and patients developing nosocomial infection showed significantly greater immune dysfunction by the measures used. Tregs and neutrophil dysfunction remained significantly predictive of infection in a Cox hazards model correcting for time effects and clinical confounders {hazard ratio (HR) 2.4 [95% confidence interval (CI) 1.1-5.4] and 6.9 (95% CI 1.6-30), respectively, P=0.001}. Cumulative immune dysfunction resulted in a progressive risk of infection, rising from no cases in patients with no dysfunction to 75% of patients with dysfunction of all three cell types (P=0.0004). CONCLUSIONS Dysfunctions of T-cells, monocytes, and neutrophils predict acquisition of nosocomial infection, and combine additively to stratify risk of nosocomial infection in the critically ill.

Development of two real-time multiplex PCR assays for the detection and quantification of eight key bacterial pathogens in lower respiratory tract infections

The frequent lack of a positive and timely microbiological diagnosis in patients with lower respiratory tract infection (LRTI) is an important obstacle to antimicrobial stewardship. Patients are typically prescribed broad-spectrum empirical antibiotics while microbiology results are awaited, but, because these are often slow, negative, or inconclusive, de-escalation to narrow-spectrum agents rarely occurs in clinical practice. The aim of this study was to develop and evaluate two multiplex real-time PCR assays for the sensitive detection and accurate quantification of Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, Moraxella catarrhalis, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii. We found that all eight bacterial targets could be reliably quantified from sputum specimens down to a concentration of 100 CFUs/reaction (8333 CFUs/mL). Furthermore, all 249 positive control isolates were correctly detected with our assay, demonstrating effectiveness on both reference strains and local clinical isolates. The specificity was 98% on a panel of nearly 100 negative control isolates. Bacterial load was quantified accurately when three bacterial targets were present in mixtures of varying concentrations, mimicking likely clinical scenarios in LRTI. Concordance with culture was 100% for culture-positive sputum specimens, and 90% for bronchoalveolar lavage fluid specimens, and additional culture-negative bacterial infections were detected and quantified. In conclusion, a quantitative molecular test for eight key bacterial causes of LRTI has the potential to provide a more sensitive decision-making tool, closer to the time-point of patient admission than current standard methods. This should facilitate de-escalation from broad-spectrum to narrow-spectrum antibiotics, substantially improving patient management and supporting efforts to curtail inappropriate antibiotic use.

Pulmonary and systemic effects of mononuclear leukapheresis

Background and Objectives  There is increasing evidence that monocytes play a key role in the pathogenesis of acute lung inflammation. Mononuclear cell (MNC) leukapheresis can be used to remove large numbers of monocytes from circulating blood; however, the detailed characteristics of monocyte subpopulations removed by MNC leukapheresis, and the biological effects on the lung, remain incompletely defined.

Low-pathogenicity Mycoplasma spp. alter human monocyte and macrophage function and are highly prevalent among patients with ventilator-acquired pneumonia

Background Ventilator-acquired pneumonia (VAP) remains a significant problem within intensive care units (ICUs). There is a growing recognition of the impact of critical-illness-induced immunoparesis on the pathogenesis of VAP, but the mechanisms remain incompletely understood. We hypothesised that, because of limitations in their routine detection, Mycoplasmataceae are more prevalent among patients with VAP than previously recognised, and that these organisms potentially impair immune cell function. Methods and setting 159 patients were recruited from 12 UK ICUs. All patients had suspected VAP and underwent bronchoscopy and bronchoalveolar lavage (BAL). VAP was defined as growth of organisms at >104 colony forming units per ml of BAL fluid on conventional culture. Samples were tested for Mycoplasmataceae (Mycoplasma and Ureaplasma spp.) by PCR, and positive samples underwent sequencing for speciation. 36 healthy donors underwent BAL for comparison. Additionally, healthy donor monocytes and macrophages were exposed to Mycoplasma salivarium and their ability to respond to lipopolysaccharide and undertake phagocytosis was assessed. Results Mycoplasmataceae were found in 49% (95% CI 33% to 65%) of patients with VAP, compared with 14% (95% CI 9% to 25%) of patients without VAP. Patients with sterile BAL fluid had a similar prevalence to healthy donor BAL fluid (10% (95% CI 4% to 20%) vs 8% (95% CI 2% to 22%)). The most common organism identified was M. salivarium. Blood monocytes from healthy volunteers incubated with M. salivarium displayed an impaired TNF-α response to lipopolysaccharide (p=0.0003), as did monocyte-derived macrophages (MDMs) (p=0.024). MDM exposed to M. salivarium demonstrated impaired phagocytosis (p=0.005). Discussion and conclusions This study demonstrates a high prevalence of Mycoplasmataceae among patients with VAP, with a markedly lower prevalence among patients with suspected VAP in whom subsequent cultures refuted the diagnosis. The most common organism found, M. salivarium, is able to alter the functions of key immune cells. Mycoplasmataceae may contribute to VAP pathogenesis.

S79 Potential diagnostic significance of neutrophil proteases in ventilator-associated pneumonia

Introduction and Objectives The clinical diagnosis of ventilator-associated pneumonia (VAP) remains notoriously difficult, as several non-infective conditions mimic VAP. Microbiological confirmation of the diagnosis using conventional cultures typically takes 48–72 h. Identification of molecules measurable within a short time frame and closely associated with microbiologically confirmed VAP is therefore highly desirable. VAP is associated with significant influx of activated neutrophils into the alveolar space. We postulated that extracellular neutrophil proteases in bronchoalveolar lavage fluid (BALF) may reliably identify VAP in suspected cases. Methods Fifty-four intubated and mechanically ventilated patients in the intensive care unit developed clinically suspected VAP and were recruited. Bronchoalveolar lavage (BAL) was performed using a standardised protocol. An aliquot of BALF was sent to the diagnostic microbiology laboratory for quantitative culture, with confirmation of VAP defined as growth of a pathogen(s) at >104 colony forming units/ml. Remaining BALF was centrifuged. The following neutrophil-specific proteases were assayed in cell-free BALF supernatant—matrix metalloproteinase (MMP)-8 and MMP-9 by Luminex assay, and human neutrophil elastase (HNE) by enzyme-linked immunosorbent assay. Urea was simultaneously measured in serum and BALF, and used to correct for the dilution of epithelial lining induced by BAL. Receiver operating characteristic (ROC) curves were constructed and optimal specificity and sensitivity for each marker calculated. Results Eleven patients (20%) had confirmed VAP. For HNE (cut off 670ng/ml) the ROC area under curve (AUC) was 0.87 (p<0.0001), sensitivity 93%, specificity 79%. For MMP-8 (13 ng/ml), ROC AUC was 0.81 (p<0.005), sensitivity 91%, specificity 63%. For MMP-9 (22 ng/ml), ROC AUC was 0.79 (p<0.005), sensitivity 82%, specificity 63%. Conclusions Neutrophil proteases are strongly associated with confirmed infection in cases of suspected VAP. The values for HNE, in particular, compare extremely favourably with any previously published equivalent values. These data suggest that neutrophil protease concentrations in BALF deserve further attention as potentially diagnostic markers for VAP. They further suggest that neutrophil proteases, inappropriately released into the extracellular space, may contribute to the pathophysiology of VAP.

S104 Monocyte influx accompanies the early neutrophilic inflammation seen in bronchoalveolar lavage fluid following lipopolysaccharide inhalation

Introduction Acute lung injury (ALI) has a mortality rate of over 30%, with no proven pharmacological treatment. Inhalation of lipopolysaccharide (LPS) in healthy volunteers induces transient inflammation resembling that found in patients with ALI. Inhaled LPS causes neutrophilia that is detectable in bronchoalveolar lavage fluid (BALF) and blood, but its effect on BALF and blood monocyte populations is not well established. Methods 12 healthy volunteers were recruited and randomly allocated to receive either 60 μg of inhaled LPS or saline (n=6 each arm). Clinical parameters, including temperature, and any reported symptoms were recorded. Full blood counts were taken at baseline and 2, 4, 6, 8 and 24 h post-inhalation. BAL was performed at 8 h. BALF cell populations were analysed morphologically using cytospins and cytometrically by flow cytometry after staining for cell surface markers (alveolar macrophages: CD163, CD206, CCR5; neutrophils/monocytes: HLA-DR, CD14, CD16). Results 4 LPS volunteers developed pyrexia, two reported cough and one myalgia. The mean maximal increment in temperature was significantly greater in the LPS arm (p=0.047). Compared to saline inhalation, LPS caused a peripheral blood neutrophilia (p=0.006) that was evident from 4 h and greatest at 8 h. There was no significant difference in peripheral blood monocyte counts between treatment arms at any point measured (p=0.87). Although mean total alveolar macrophage numbers were similar between the two groups, their relative proportion in the LPS volunteers was significantly reduced due to the expansion in neutrophil and monocyte populations. Flow cytometry revealed a 24-fold expansion of the neutrophil population following LPS (in parallel with morphological data). These neutrophils were distinguishable by HLA-DR-/CD14-/CD16+ staining. There was a concomitant similar rise in the population of HLA-DR+/CD14+/CD16- ‘classical’ monocytes. Further analysis of these monocytes revealed that macrophage cell surface marker expression was absent. Conclusion Morphological analysis of BAL fluid in previous LPS inhalation studies has consistently suggested that there is no change in the monocyte population. Using flow cytometry enables a more detailed analysis. This study is the first to clearly demonstrate that an early expansion in the monocyte population accompanies the neutrophil influx seen in BALF 8 h following inhalation of LPS.

S81 Primary type II alveolar epithelial cells respond differentially to bacterial virulence factors

Introduction and objectives The pathogens most commonly implicated in ventilator-associated pneumonia (VAP) are Pseudomonas aeruginosa and Staphylococcus aureus. Although a florid inflammatory response characteristically occurs in the alveolar space in VAP, the underlying mechanisms remain unclear, partly owing to a lack of adequate models of alveolar injury. We therefore sought to characterise the response of primary human type II alveolar epithelial (ATII) cells to virulence factors from these pathogens. Methods Primary ATII cells were derived from seven patients undergoing surgical resection for lung cancer. Lung tissue was refrigerated overnight; flushed with saline; incubated with trypsin; diced; incubated with DNAse I; and strained/filtered. Macrophages and fibroblasts were removed by adherence. The resulting cell population was centrifuged, washed, resuspended and plated onto tissue culture plates pre-coated with type I bovine collagen at 2×106 cells/ml. When cells achieved confluence medium was replenished and the following were added for 24 h: 100 ng/ml P aeruginosa lipopolysaccharide (LPS); 10 ug/ml S aureus lipoteichoic acid (LTA); 10 ug/ml S aureus peptidoglycan (PGN); 10 ng/ml human recombinant tumour necrosis factor alpha (TNFα); or control medium. Supernatant was aspirated at 24 h and cytokines were measured by cytometric bead array. Results Interleukin (IL)-1β, IL-6, IL-8 IL-10, IL-12p70 and TNFα were all detectable in control medium at 24 h. None of the measured cytokines were significantly altered by application of LPS or LTA. In contrast, PGN induced a significant rise in concentrations of IL-1β, IL-6, IL-8 IL-10 and TNFα. Addition of TNFa induced a significant increase in IL-6, IL-8 and IL-10. The only cytokine to be uniformly uninfluenced by stimulation was IL-12p70. Conclusions In our hands primary ATII cells appeared to be unresponsive to P aeruginosa LPS or to S aureus LTA. By contrast, S aureus PGN provoked a brisk and significant inflammatory response simultaneously affecting a range of cytokines. These data suggest that ATII cells have strikingly different responses to individual bacterial virulence factors. They further suggest that PGN (but not LTA) contributes, at least in part, to the florid inflammatory response seen in Staphylococcal pneumonia.