Dire Straits for Biomarkers of Neurosurgery-Associated Meningitis.
H. Puttgen,J. Shah
DOI: https://doi.org/10.1097/CCM.0000000000001244
IF: 8.8
2015-11-01
Critical Care Medicine
Abstract:Because inaction often marks the surest path to disaster in the critical care unit, the intensivist must accept the uneasiness of making vital decisions based on limited information. As each intervention carries a risk of nosocomial harm, the intensivist must fully consider the safest course of action. A dilemma begins with a pair of conflicting premises such as these, but the term only gains meaning when circumstances demand action along one or the other. Dilemma denotes the forced choice between two predictable, adverse consequences. We find the classic example in Odysseus’s voyage through the Strait of Messina. Sail along the Italian shore and Scylla will surely kill some of the crew. Sail near Sicily, and Charybdis may devour the entire ship. He must decide, for the narrow passage allows no middle course. Rarely does one read on the subject of meningitis after neurosurgery without encountering the word dilemma, and the article by Liu et al (1) in this issue of Critical Care Medicine is no exception. As they note, rates of external ventricular drain (EVD)-associated infection range from 2% to 22% or about 8.80% per patient and 8.08% per EVD (2, 3). When caring for patients with severe acute brain injury necessitating the placement of an EVD or another neurosurgical intervention, intensivists often face the difficulty of interpreting an inflammatory cerebrospinal fluid (CSF) profile in the context of signs and symptoms of meningitis (3, 4). They must then choose between the risk of a prolonged course of empiric, broad-spectrum antibiotics, and the risk of incompletely treated bacterial meningitis. The difficulty lies not in finding a solution to a complicated problem but in committing to one unappealing course of action over another. Liu et al (1) hope to offer us a means of escape from this dilemma. In their article, they investigate a third option: a diagnostic approach that does not rely on CSF Gram stain, culture, cell counts, or glucose but seeks to characterize the source of inflammation according to the cytokine profile. They studied 32 patients over a period of 1 year who had severe brain injury (secondary to intracerebral hemorrhage, subarachnoid hemorrhage, or traumatic brain injury), had undergone neurosurgical interventions, and then developed postoperative fever more than 7 days after surgery. CSF was collected regularly via EVD from insertion until 4 days after fever onset. Patients with fever were divided in to three groups: nonmeningitis group (n = 11), aseptic meningitis group (n = 13), and bacterial meningitis group (n = 8). Levels of tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, IL-8, transforming growth factor (TGF)-β, and procalcitonin (PCT) were measured in CSF. The authors found that the levels of TNF-α, IL-1β, IL-6, and IL-8 were significantly elevated on the day of the fever and up to 4 days before and after fever onset in patients with bacterial meningitis when compared with those with aseptic or noninfectious meningitis. TGF-β and PCT did not follow the same trend. IL-1β and IL-8 were found to have high area under the curve values, sensitivity, and specificity at fever onset (IL-1β: 1.000, 100%, 100%; IL-8: 0.894, 87.5%, 92.3%, respectively). The authors concluded that proinflammatory cytokines such as of TNF-α, IL-1β, IL-6, and IL-8 were elevated in the CSF of patients with bacterial meningitis at 4 days prior to onset of fever. Measurement of cytokines could provide a rapid prediagnostic tool, allowing closer clinical observation and timely intervention in these patients. The above-mentioned biomarkers along with several other proinflammatory and anti-inflammatory biomarkers have been extensively studied to distinguish bacterial from aseptic meningitis, with most studies reporting elevated levels of the biomarker in the former group (5). CSF lactate has been most commonly studied in this regard and is shown to have sensitivity of 93% and specificity of 96% in a recent meta-analysis (6). Cutoffs of 3–4 mmol/L have been reported to be most likely to predict bacterial meningitis (7). Most of these biomarkers have been studied independently or in combination with other biomarkers or other clinical or routine CSF profile elements, such as CSF, WBC, protein, and glucose. The combination may be able to better predict bacterial meningitis than individual biomarkers (8). However, the methods used in this study may not be practically applicable. CSF is not routinely collected for cytokine assessment post-EVD placement. Hence, a rise in cytokines preceding the development of fever, although physiologically appealing, may not be practically feasible to demonstrate. In addition, patients who developed fever in the first week postEVD placement/craniotomy were excluded to avoid confounding from the elevation in cytokines due to initial injury or surgical manipulation. This can hinder the extrapolation Hans Adrian Puttgen, MD Jharna Nitin Shah, MD, MPH Division of Neurosciences Critical Care Department of Neurology Johns Hopkins University Baltimore, MD