Thank you, everyone, for being here at the 2023 SCCM Congress. I am here to talk to you about nosocomial meningitis and ventriculitis for the next 20 minutes and give you some updates on what's happening on this topic. My name is Sarah Norton. I am a nurse practitioner working in critical care medicine with a subspecialty in neurocritical care working in Southern California. So nosocomial meningitis and ventriculitis, why are we here? Well, neurocritically ill patients are complex with many reasons why fever and leukocytosis would be present. Now add in the inability for your patient to communicate effectively and the ability to suspect meningitis and ventriculitis based on symptom reporting alone is severely reduced. So in these 20 minutes, we're going to cover the need to know on the most recent evidence and recommendations surrounding the diagnosis, management, and treatment of nosocomial meningitis and ventriculitis and briefly highlight some areas requiring further research, of which there are many. So a few key points to account for before we get started. Yes, community-acquired meningitis and ventriculitis does occur, as we all know. However, for the purpose of this lecture, we're going to be discussing the types of patients you see listed here, pretty much anyone who's had a craniotomy or who presents with head trauma slash TBI, as sometimes their skull fractures can lead to infection. As we all recognize, nosocomial infections have significant impacts on prognosis, particularly within the neurocritical care population. CSF culture remains the gold standard of diagnosis. However, this takes time, and if treatment were delayed, that could have a negative impact on patient outcomes. Also, cultures are specific, but they sometimes lack sensitivity in those patients who have received prior doses of antibiotics, which, to be honest, is most of the ICU patient population for one reason or another. Also, standard diagnostic tools used for the diagnosis of community-acquired meningitis can be unreliable given the amount of blood contamination that can occur in the majority of neurocritically ill patients. The most recent IDSA guidelines for the treatment of nosocomial meningitis and ventriculitis are from 2017, but these recommendations were based mostly on expert opinion at that time given the paucity of rigorous clinical data. Unfortunately, the paucity of solid data surrounding nosocomial meningitis and ventriculitis remains to be the case, as on this most recent review. The newest research we'll briefly highlight today is heading towards how to supplement the standard exams to diagnose these patients more efficiently and accurately. A little discussion now on biomarkers and their role in diagnosis of infection in this patient population. At the University of Texas in Houston, Montes and his colleagues wanted to investigate if other bio or inflammatory markers in conjunction with the cell index would be useful to extrapolate in aiding the diagnosis of nosocomial meningitis and ventriculitis in the post-op ICH patient population. They went ahead and matched cases and controls based on the GCS, the Apache scores, age, and diagnosis of altermental status. In all cases in this study, they had positive CSF culture data, and cases were reported to be well matched by age in Apache 2. They did propose a new mode for diagnosis to include the analysis of an aneurysmal subarachnoid hemorrhage patient with CRP, CSF to blood glucose ratio, CSF leak, CSF granulocytes, and CSF lactate, and assigned points for each based on the levels returned. A score of 6 or more had a high probability of infection. They then calculated the cell index for these patients and then analyzed in three different correction factors, 500 to 1, 750 to 1, 1000 to 1. Interestingly in their study, only 41% of patients with culture-proven meningitis had a positive gram stain. ICH patients were noted to have higher cell counts in the CSF, which was attributed to the pleocytosis in addition to the infection. There was a majority of infected patients who did score higher than a 6 by their scale. However, the granulocytes in CSF elevated CRP or presence of aneurysmal subarachnoid did not render significantly different results when compared to those patients who were infected versus not. Their scoring model had a sensitivity of 32.4% and a specificity of 94.6%, though the area under the curve was only 65%, which they attributed due to inconsistent CSF lactate and CRP monitoring throughout the study. In Germany, Lenske and colleagues conducted a retrospective analysis of 40 TBI patients who had EVDs from 2013 to 2017 to evaluate if inflammatory markers like IL-6 or total leukocyte count and protein in the CSF in conjunction with WBC and the serum had any correlation in diagnosing ventriculitis. Again, culture remains the gold standard. However, for TBI patients in particular, a majority of the time they've already received empiric antibiotics for one reason or another, or if they've received antibiotics during EVD insertion, making the value of the CSF culture less indicative of active infection. Of the 40 patients that they looked at, five of which developed culture-positive ventriculitis. Important findings that they noted was the IL-6 in the CSF had a high diagnostic potential with values above 4,064 pg per ml and had an area under the curve of 93% at that level. The coolest finding was that these values resulted about 24 hours prior to typical exam findings would manifest in patients with EVD ventriculitis in their TBI population. They also noted that the IL-6 marker was not typical for the TBI population to be monitored frequently, for signs of ventriculitis, however, it had been successful in aneurysmal subarachnoid hemorrhage patients. They also noted that the IL-6 metrics and TBI were correlated with a higher morbidity when the levels were greater than 2,000 pg per ml, which is about half of the cutoff for the ventriculitis value diagnosis. Other markers like total leukocyte count in the CSF were significantly increased in patients with ventriculitis and the protein CSF, neutrophil percentage CSF, serum CRP, serum WBCs were non-diagnostic in the TBI or subarachnoid patient populations. All in all, this was a small sample size and needs further investigation on a larger scale, but it does look interesting. Here's some of their numbers they found. In China, there seems to be some promising potential for the assessment of heparin binding protein in the CSF for accurate diagnosis of nosocomial meningitis and ventriculitis compared to the CSF lactate and CSF ProCal. In the CSF, heparin binding protein is released by neutrophils early in the infection stage, yielding the possibility for earlier detection of infection. This study was a single center observational study with 281 patients who were evaluated post neurosurgical intervention. They did include patients who received empiric antibiotics as well as steroids. They differentiated between negative of infection, infection positive based on criteria of fever, white blood cells, and other signs, as well as culture positive groups. And they analyzed the standard CSF labs plus heparin binding protein, ProCalcitonin, lactate, and culture when they had patients with suspicion for nosocomial meningitis and ventriculitis. What they found was helpful in that heparin binding protein had an area under the curve of 99% with a 95% confidence interval, whereas with the same confidence interval, CSF lactate and CSF ProCalcitonin had areas under the curve of 98% and 69% respectively. Other findings they noted were CSF ProCalcitonin had a low sensitivity for diagnosis in the positive cases, but the CSF ProCalcitonin was significantly higher than those who did not receive empiric antibiotics in those that were infected. The lactate, CSF lactate, showed that there was a lower specificity but a higher positive predictive value compared to the other biomarkers. CSF heparin binding protein has a large potential for aiding in early diagnosis of infection given that it is produced by activated neutrophils. If the CSF heparin binding protein is elevated, yet all other tests run are negative, recommendations from the authors were to treat on the basis of a positive infection until final CSF culture results were obtained for these post-neurosurgery patients. The heparin binding protein all in all had a higher sensitivity and negative predictive value when compared to the other biomarkers and was more superior in diagnosing infection even in the early stages. So this is why we like this study. That's really great information to know. They also included patients who received early empiric antibiotics with pertinent findings. They had a large-ish sample size and they focused on the post-neurosurgical community rather than the community-acquired meningitis patient population. Some areas for improvement for this study would be that it seems that CSF heparin binding protein testing is a relatively new concept, which does not seem that it's widely available worldwide yet. The patient population in this study was primarily tumor-related and did not have strokes or intracranial hemorrhages or traumas, which would be interesting to see if the findings held true for those patients as well. So with nosocomial meningitis and ventriculitis, we do have a special mention. That's a Sinusobacter baumannii, who is quite the character. Relating to Sari and those others, those who are at risk for infections with this bacteria are those patients with prolonged antibiotic times, particularly with exposure to broad-spectrum antibiotics, EVD catheters greater than five days, as well as those with ventriculostomies or shunts. A. baumannii works really well under pressure. It can survive a lot and it's super great at adapting, which is increasingly difficult for us because it is becoming more and more drug-resistant. It has a multitude of virulence factors contributing to its successful attacks, including outer enzymes, proteins, mobility, biofilms, polysaccharide force fields, etc. It can cause infections in multiple places, including the brain, and causes about 70% mortality in the neurocritically ill patient populations, particularly those with device-related infections. These are part of the first-line therapy, as is colistin and polymyxin B, and usually combo therapies seem to be the most effective, but a clear consensus on the best approach is lacking, which makes you wonder if resistance is futile. Sharma and their colleagues in India ended up doing a retrospective analysis of patients in their neuro-ICU from 2014-2018 who were diagnosed with Acinetobacter baumannii and looking at 72 patient outcomes, which was the largest case series to date. 20% of the isolated cultures were considered extremely drug-resistant, though none were resistant to a colistin-tugacycline combo, and the mortality was 40.3%. Seven patients with the extremely drug-resistant pathogen survived out of a total of 15. So that is the big highlights in biomarkers, and now we're going to move on to recommendations. They recommend that MRI with contrast as well as DWI can help be an adjunct to the diagnosis of nosocomial meningitis and ventriculitis. CT abdomens can be helpful in diagnosing VP shunt loculations associated with infection. And in terms of antibiotic regimens, they recommend vancomycin plus a beta-lactam with a goal to cover for coagulase-negative Staphylococci, such as Staph aureus, as well as other skin flora. Typically, it's a good idea to cover for gram-negatives to include pseudomonas coverage, either with a cephalosporin or a carbapenem. Your vanc trough for the critically ill is a 15 to 20. If your patient has a beta-lactam allergy, they recommend meropenem or astreonam or Cipro. If the patient has another resistant organism, colonization, or history, they do also recommend to cover for that. Coverage of antibiotics is usually no less than 10 days and can go as much as 21 days for gram-negative bacilli, but there's weak evidence to support these recommendations. And this is amount of time for antibiotic coverage after the last positive culture date. So a long, long time for some of these people. Infected hardware needs to be removed in order to gain source control, and intraventricular antibiotics may be appropriate if the response to systemic antibiotics is low. They will need dose adjustments, thank you to our pharmacy friends, and the goal is 10 to 20 of the MIC of the organism that's causing the problem. Periop antibiotics is recommended to reduce the incidence of incision-related infections, and patients with basilar skull fractures with a CSF leak should always get antibiotics, and you should also consider pneumococcal vaccinations. So intraventricular and intrathecal routes should be considered for all patients who are not responding to traditional IV antibiotics. Given the increasing resistance of organisms, there's an increasing need to administer antibiotics in new ways, such as intrathecal or intraventricularly. Again, our common culprits are Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa. By using this different route, it allows for different combos of antibiotics that previously were not available given poor CSF or blood-brain barrier penetration. The pros of this is that there's better penetration and less systemic toxicity risk related to high-dose IV antibiotics. However, the cons of this route of administration is that it's not FDA-approved and would be considered off-label use, as well as there not really being any well-designed studies currently available to review. The Neurocritical Care Society does support the use of intraventricular and intrathecal antibiotics when a multidrug-resistant or extremely drug-resistant organism is suspected, or if the patient fails IP therapy. So Alnami and Almari did a systematic review for intraventricular and intrathecal colistin administration for multidrug-resistant meningitis, ventriculitis, and kiddos. They looked at 20 articles retrospectively and saw who administered colistin via either of these routes. They found 23 patients had recovered from their infection, 3 died, and 5 had permanent disability. They did conclude, however, that it is a safe route to administer antibiotics when IV options have failed for children with multidrug-resistant nosocomial meningitis and ventriculitis. Abdallah and Asale wanted to see the safety and effectiveness of this route with ticocycline, specifically for Acinetobacter bimani meningitis and ventriculitis. Usually, ticocycline has a poor CSF penetration when it's administered IV and generally requires higher doses. It is considered salvage therapy in most cases, particularly with Acinetobacter bimani. They did a search to see who received this treatment and saw that there was a paucity of studies and mostly had case reports available to review. They did note that the intraventricular ticocycline was always paired with another antibiotic, either through the IV or paired with intraventricular polymyxin. And due to the pharmacokinetics and pharmacodynamics, the preservative-free state and overall general safety, they recommended that a dose of 4 mg a day with a max of 8 mg a day should be safe for administration via the intraventricular or intrathecal route as a salvage maneuver when other options have failed. Using optimization of antibiotics, don't forget about the MIC. The goal is to get the antibiotics optimized in the serum with the maximum amount of time without causing toxicity. Remember your pharmacokinetics, pharmacodynamics, and the blood-brain barrier is a great thing, but sometimes it'll take longer for antibiotics to cross. Phone a friend in your pharmacy department with extra questions. So the Neurocritical Care Society ended up putting out a EVD insertion and management recommendations. They did note that there was insufficient evidence surrounding specific antibiotics for perioprophylaxis, and they recommend following your local antibiogram data during insertion. And recommend at least one dose of antibiotics given prior to EVD insertion. This is a conditional recommendation with low-quality evidence. They do say, however, not to administer antibiotics prophylactically for the entirety of the EVD placement. The use of an impregnated catheter, if they're available, is recommended, and there was no comparison available between either antibiotic-impregnated or silver-impregnated catheters. Avoid routine sampling of the CSF as to not introduce contaminants unnecessarily. Don't change the catheter without cause, and don't change the tubing without cause. Wean the EVD as soon as possible, and get it out as quickly as you can. Bundling. It's super great for auto-insurance as well as EVD management. The Neurocritical Care Society recommends creating bundles at your system level that includes aseptic insertion, limited manipulation or breaking of the closed system, as well as standardized dressings and weaning plans that get the EVD ASAP out of the patient's head. In a meta-analysis done in the UK, they found that clorhexidine dressings may reduce incidence of infection, but they do say that a randomized controlled trial to determine safety and efficiency as well as efficacy should be conducted. This leads us to our areas to explore. Bundling protocols to include standardization of insertion, including antibiotics timing, placement, where you place it, etc., frequency of the drain manipulation, and even possibly guidance surrounding timeframe of insertion length can all help reduce incidence of nosocomial meningitis and ventriculitis. Further research into heparin-binding protein's role in the diagnosis of infection seems super promising and would appear to be exceedingly helpful if we were able to see it approved and extrapolated worldwide. And research into safety and efficacy of a standardized random controlled trial would be ideal to assess the safety of intraventricular and intrathecal antibiotic administration, though admittedly it would likely be difficult to design. Thank you all for being here today. Your attendance is very much appreciated and please feel free to reach out with any questions.

nosocomial meningitis

ventriculitis

diagnosis

treatment

antibiotic administration

research