Monte Minute‎ > ‎Monte Minute‎ > ‎

Bacterial Meningitis - 3/19/2013

posted Mar 19, 2013, 8:07 AM by Rohit Das   [ updated Mar 19, 2013, 1:39 PM by Purnema Madahar ]

At report yesterday, we discussed a case of a woman in her 40s who presented with a few days of headache, nausea/vomiting and fevers who ultimately received the diagnosis of bacterial meningitis based on her CSF study profile. The etiology wasn’t completely clear, as the gram stain and culture results were confounded by antibiotic therapy prior to lumbar puncture. Lots to talk about:

·         What is the epidemiology of meningitis? Breaking it down etiologically, what are the most common causes of “aseptic” versus bacterial meningeal disease?

·         What are the clinical signs and symptoms of meningitis, and how helpful are they from a diagnostic accuracy standpoint?

·         What CSF findings are consistent with bacterial meningitis?

·         How should one approach treatment of bacterial meningitis?

·         What is the prognosis of bacterial meningitis and potential neurologic complications?

What is the epidemiology of acute meningitis? Breaking it down etiologically, what are the most common causes of “aseptic” versus bacterial acute meningeal disease?

·         Very difficult to get an accurate incidence for meningitis, whether aseptic or bacterial. For some of the more common causes, the CDC estimates that 10-15 million symptomatic enteroviral infections occur in the US annually, of which 30,000 to 75,000 are cases of aseptic meningitis. The incidence of bacterial meningitis has dropped DRAMATICALLY since the advent of the H. influenza type B vaccine, which decreased incidence of bacterial meningitis from around 3 cases per 100,000 population/year to 0.2 cases per 100,000 population/year…

·         Aseptic Meningitis – there are COUNTLESS causes of non-bacterial meningitis, and I shall focus on the most common ones:

o   Enteroviruses – BY FAR the most common cause of aseptic meningitis, representing nearly 90% of cases. In temperate climates (like ours) they follow a typical seasonality, with incidence mainly being in the summer and fall; in tropical climates, the epidemiology is year round. The main culprits are Coxsackie and Echovirus. In this context, it’s important to look for other manifestations of enteroviral disease – exanthems, myopericarditis, conjunctivitis…or specific enteroviral syndromes like herpangina or hand-foot-mouth syndrome.

o   Mumps – not a very common etiology in these parts thanks to vaccination, but in a nonimmunized population, a very common cause. It’s estimated that 10-30% of patients with symptomatic mumps have meningitis, and it can occur without evidence of parotitis.

o   Herpesvirusesthis includes HSV, CMV, EBV, VZV…HSV, though, carries the most clinical significance, representing around 3% of all cases of aseptic meningitis. It is most commonly associated with HSV-2 primary genital infection, and rarely occurs with recurrences of genital herpes.

o   Other – this includes quite a long list…HIV, influenza/parainfluenza, LCV (lymphocytic choriomeningitis virus), arboviruses, spirochetes (Lyme Disease, syphilis)…etc. A VERY comprehensive list is provided on Table 1 of the attached review.

·         Bacterial Meningitis – Pre 1990, the predominant cause across all comers with bacterial meningitis was H. influenzae. That spectrum has shifted however with advent of the Hib vaccine. In adults, though, the epidemiology has not changed as dramatically, with the most common causes being…

o   Streptococcus pneumoniae – this destructive bug is the most common cause of bacterial meningitis, representing around 60% of diagnosed cases. Unfortunately, it also has the worst prognosis…bad combination.

o   Neisseria meningitidis - the second most common cause, representing around 10-15% of cases. Serogroups B, C and Y represent the majority of cases. A distinguishing aspect of meningococcal meningitis is the development of a petechial à purpuric rash of the extremities, which develops in up 60-80% of cases.

o   Listeria monocytogenes – probably the third most common cause in adults (around 8-10%), this is mainly a disease of patients older than 50 years old…risk factors include alcoholism, immunosuppression, DM, chronic liver/kidney disease, or conditions of iron overload (the invasive capability of Listeria is enhanced with increased iron availability). Interestingly, patients with Listeria meningitis have an increased tendency to present with seizures and/or focal deficits as compared to other pathogens…

o   Other – H. influenza (still happens, around 4-8%), Group B Strep (majority of the time in infants, but does occur in adults), Staphylococcus (usually trauma related, or in context of neurosurgical procedures), Gram negative bacilli (usually nosocomial, similar to the epidemiology of Staphylococcal meningitis, and usually in older adults in association with gram-negative sepsis) amongst several others.

What are the clinical signs and symptoms of meningitis, and how helpful are they from a diagnostic accuracy standpoint?

·         The Rational Clinical Exam series, which I hope everyone is a big fan of, published a nice article in 1999 highlighting the value of the symptoms and certain signs in diagnosing meningitis. Of the 10 cases they used, all but 1 were retrospective chart reviews (so only sensitivity was able to be assessed), around 850 patient-episodes were compiled, and meningitis was confirmed by either lumbar puncture or autopsy. The majority of the studies looked mainly at bacterial meningitis. Some highlights from this paper (which is of course attached):

o   The pooled sensitivity was 50% for headache, 30% for nausea/vomiting, and 28% for neck pain (evaluated in only one of the studies). In the one prospective study (34 patients) the sensitivity for headache was 50%, and 60% for nausea/vomiting. Based on these numbers, the authors appropriately conclude that clinical symptoms alone are not sufficient for establishing a diagnosis of meningitis, probably related to the altered cognition that often accompanies this illness…on to the physical exam…

o   Physical exam is more helpful in the evaluation of potential meningitis. The main things to assess for are the individual parts of the classic triad – fevers, nuchal rigidity, and altered mental status – along with some adjunctive maneuvers – Kernig and Brudzinski signs, and jolt accentuation of headache. Highlights from the paper:

§  The Classic Triad – not so classic unsurprisingly, as the individual parts occur together in 30-40% of patients. Individual sensitivities – fever (~90%), nuchal rigidity (70%) and altered mental status (~70%). More importantly then these individual accuracies is their value taken together – one study reported that 95% of patients had 2+  parts of the triad, and two studies reported that nearly 100% had at least 1 of these clinical findings. HOWEVER, a recent prospective study in 2002 containing 297 patients (217 of which had meningitis) showed that the negative LRs for both the absence of fever and neck stiffness approached 1…goes to show how discrepant analyses can be in prospective versus retrospective studies…

§  Kernig and Brudzinski – hard to comment on the accuracy of these tests as they haven’t been extensively studied. In the largest prospective study looking at these signs, the sensitivity for both were very low (5%) but the specificity was very high (95%).

§  Jolt accentuation of Headache ­ this is done by asking the patient to simply rotate their head horizontally at a frequency of 2-3 rotations per second – worsening of the headache is considered a positive sign. In a prospective study of 34 patients, 30 of which had meningitis (mostly aseptic), the overall sensitivity was 97%, specificity was 60%, and the +LR and –LR were 2.4 and 0.05 respectively. So, the absence of jolt accentuation of headache is helpful in lowering your probability of meningitis.

·         Ok…so after all that, some conclusions. Certainly, clinical symptoms are not enough to rule out meningitis, and physical exam is much more helpful in that regard…but how helpful? Though past literature seems to favor that the absence of all three aspects of the triad is a safe way to exclude meningitis, recent prospective literature sways away from that conclusion. The Kernig and Brudizinski signs are definitely helpful if present, but not at all helpful if absent. Finally, jolt accentuation of headache, a fairly easy maneuver, is probably the most sensitive sign one can elicit on exam.

What CSF findings are consistent with bacterial meningitis?

·         Most should know this well, but worth reviewing. Typical findings include elevated CSF pressure (usually 200-500 mmHg) significant pleocytosis with neutrophil predominance (1000-5000, >70-80% neutrophils), low glucose (<40 mg/dL) and high protein (100-500 mg/dL). In one study, the combination of >1180 neutrophils, CSF glucose <34, CSF/serum glucose ratio of <0.23 and CSF protein >220 was able to differentiate bacterial versus viral meningitis with 99% certainty.

·         Obviously, Gram Stain and Culture are key. The sensitivity of Gram stain ranges from 60-90%, and the specificity is 100%. The sensitivity also differs for different pathogens – highest for pneumococcus and hemophilus, and on the lower side for meningococcus (75%) and gram-negative bacilli (50%). Additionally, the sensitivity is far lower (40-60%) in patients who have received antibiotic therapy prior to LP – studies looking at pneumococcal meningitis specifically have shown that complete CSF sterilization can occur as early as 4 hours after initiation of appropriate antimicrobials.

·         In cases where Gram Stain and Culture are negative, there are other adjunctive tests that may lead to a diagnosis. These include agglutination techniques for detecting bacterial antigens, and of course, PCR. The latter is particularly useful, as it detects low levels of bacteria and recent assays have been developed to implement PCR as a way to assess for antimicrobial sensitivity, particularly for meningococcus.

How should one approach treatment of bacterial meningitis?

Several important principles to review here…

·         When to get a CT Head – in a large prospective study of 300 adults with suspected meningitis, 235 patients underwent CT prior to lumbar puncture to assess for the risk of herniation. Of those patients, only 5% had evidence of mass effect. Furthermore, the risk of having an abnormal CT of the head was accurately predicted by the following factors: immunosuppression, history of CNS disease, new-onset seizure, papilledema, altered mental status and/or focal neurologic deficit. So, in the absence of those risk factors, a CT Head prior to LP is not indicated by the most recent IDSA guidelines.

·         DO NOT delay Antibiotic Therapy – though the appropriate timing of antibiotics has never been assessed in a prospective study, retrospective data has shown that a delay of antibiotic therapy of greater than 6 hours after presentation to the ED leads to a 8-9-fold greater risk for adverse outcome (though the extent of that risk varies significantly by the etiologic pathogen). So, if you do plan on getting a CT, make sure you check blood cultures and give antibiotics beforehand.

·         Steroids – the rationale behind this is mainly from animal studies, which showed that steroids significantly ameliorated CSF inflammation, cytokine production, and cerebral edema associated with pneumococcal meningitis in rats treated with PCN. In a randomized, double-blinded study of 301 adults with bacterial meningitis, patients received dexamethasone 10-15 min before, or at the time, of antibiotic administration. In those given steroids, both the mortality (7% vs 15%) and neurologic disability (15% vs 25%) at 8 weeks were significantly reduced. When analyzed based on culture results, these differences were only seen in cases due to Pneumococcal meningitis. However, since we don’t usually know the cause at the time of initiating treatment, dexamethasone is recommended as part of the initial treatment for all adults with bacterial meningitis, and only continued if gram stain and culture are consistent with pneumococcal disease.

·         What treatment to give? – Initial empiric therapy is guided by the patient’s epidemiology. Community acquired meningitis should cover what I mentioned above, and thus includes Vancomycin (due to a world-wide emergence of PCN-resistant pneumococci), Ceftraixone and Ampicillin (for adults > 50 – Listeria coverage). In patients who are immunocompromized, have experienced penetrating trauma or are post-neurosurgery, consideration for coverage against Pseudomonas also needs to be considered – in such scenarios, a fourth-generation cephalosporin (Cefepime or Ceftazidime) or carbapenem (Imipenem or Meropenem) is thus recommended.

o   Very importantly, steroids decrease the penetrance of Vancomycin into the CSF – worth considering for patients with PCN-resistant pneumococcus who are on steroids and aren’t getting better with appropriate therapy.

What is the prognosis of bacterial meningitis and potential neurologic complications?

·         Importantly, bacterial meningitis leads to a host of neurologic complications, most of which are increased with (surprise) pneumococcal disease. Generally, three factors – hypotension, altered mental status and seizures – correlate best with incidence of adverse outcomes (defined as in-hospital death or neurologic deficit at discharge). 

o   Altered Mental Status, Increased ICP – these are things usually present upon diagnosis, but the latter can also worsen with treatment due to subarachnoid inflammation resulting from bacterial death (thus the logic behind the use of dexamethasone).

o   Seizures – occur in 15-30% of cases, with higher rates in pneumococcal meningitis. As mentioned, the presence of seizures during an episode of meningitis is a poor prognostic sign. Not surprisingly, patients with meningitis who have seizures also have a higher rate (around 8%) of recurrent seizures in 5-year follow-up.

o   Focal Deficits – didn’t go into this too much, but meningitis has an interesting pathophysiology which includes marked changes in cerebral vasculature that can lead to cerebral infarction; focal deficits may also occur as a result of cerebral edema. Types of deficits include CN palsy, varying extents of paresis, aphasia, ataxia…etc. The incidence ranges from 20-50%, and yes…higher with pneumococcus.

o   Sensorineural Hearing Loss – ranges between 12-20%, either being transient (usually due to conductive defect from increased ICP) or permanent (from direct bacterial invasion/CSF inflammation and consequent damage to important hearing structures). Yup, higher with pneumococcus.

o   Neuropsychological Impairment – one meta-analysis showed that 30% of survivors of bacterial meningitis had some degree of cognitive impairment in long-term follow up…that study actually DIDN’T show a difference in incidence with pneumococcus versus meningococcus…surprising…

·         Despite our best efforts, the mortality of bacterial meningitis is very high, with an average of around 25% for pneumococcal disease, 20-25% for Listeria and around 10% for meningococcus. This number varies with age however - death rates are around 2-3-fold higher in patients >65 years old. Death rates are even higher for gram-negative infections (30-40%).

 I didn’t focus too much on the management and prognosis of aseptic meningitis…BUT, I did attach a review article worth reviewing for that purpose. Also attached is the RCE Book Chapter on Meningitis, as well as the most recent IDSA Guidelines on Diagnosis and Management of Bacterial Meningitis…Enjoy


Aseptic Meningitis
Lee et. al., Curr Opin Infect Dis 2007, Volume 20: 272-77

Hatala et. al., JAMA Evidence - Rational Clinical Exam, Adult Meningitis (its the book chapter, with updates from the 1999 Paper)

Tunkel et. al., Clin Inf Dis 2004, Volume 39: 1267-84


Comments