Monday, 1 September 2008

The Meningitis Quandry

Dear Bloggers

Today, I would like to discuss an area of medical uncertainty - the partially treated meningitis!

Sometimes a patient will present to the doctor with the classical features of meningitis and antibiotics are started before the lumbar puncture can be performed because for example, of the potential delay in obtaining a cranial CT scan or that the patient is too far from the nearest hospital and a delay in giving antibiotics might be deleterious. Note that GPs in the UK have traditionally administered i.m. or i.v. benzylpenicillin to patients with suspected meningitis before they are taken to hospital.

This of course is the correct management because a significant delay can result in adverse outcomes in meningitis.

However, partially treated bacterial meningitis can cause the conversion from a predominantly neutrophil predominant CSF cell count to a predominantly lymphocytic cell count. Therefore, on initial examination, it may look as if the patient has a viral meningitis!

How do we differentiate between viral and partially treated bacterial meningitis? Now that can be almost impossible ! However, there are some features of the CSF that can guide us although none are absolutes.

Bacterial meningitis has CSF that tends to appear turbid on gross visual examination and upon analysis, the glucose is usually <50%>

Conversely, viral meningitis fluid appears macroscopically clear, has lymphocyte predominant cells (early infection may show neutrophils!!), the CSF glucose >50% of the level of the serum glucose in most cases and the protein is usually raised only mildly (by <1.5mg/dl)

Protein levels are usually only mildly raised in viral infections (as above) whereas they can be >1.5mg/dl in bacterial infections.

Of course, gram stain can be negative in partially treated meningitis and so this cannot be relied up.

There is evidence to suggest that CSF can still remain positive for organisms up to approximately 4 hours after the administration of antibiotics, so this can be helpful to still look for the organisms within this time frame.

Blood cultures should always be taken in meningitis patients as bacteraemia can give positive results allowing isolation of a potential organism with a subsequent tailoring of therapy.

Remember that certain bacterial infections can present with a lymphocyte predominant cell count and they include listeria monocytogenes infection (consider in the over 50s age group) and tuberculosis. TB meningitis can produce a macroscopic fibrin web.

Fungal infection with cryptococcus neoformans can also produce lymphocyte predominant CSF and this should be considered in patients with severe liver disease and not just HIV patients. In HIV patients, the rest of the CSF can appear normal and hence, the CRAG testing and Indian Ink stain are the mainstay of confirming the diagnosis.

Hence, it is not as clear cut to make a diagnosis between viral and bacterial meningitis when antibiotic treatment has muddied the waters, as the vast majority of handbooks and textbooks out in the world arena would suggest-- I know as I have reviewed several recently and they all differ markedly in respect of the analysis of CSF. Most texts simply differentiate between how to make the diagnosis of viral versus bacterial meningitis and very few actually discuss partially treated meningitis.

One such book that does address this important point is the Oxford Textbook of Medicine and please consider reviewing this for more in-depth reading. Another good book has been written by Dr Makoto Aoki, Specialist in Infectious Disease medicine in Japan. This text is currently written in Japanese -- an English version is however a must please! :-)

So, when faced with the viral meningitis versus partially treated meningitis patient what do you do?

Firstly, still take the blood cultures even whilst on antibiotics - they may still prove to be positive.

Secondly, examine the CSF for bacteria, and if microscopically negative ( as in partial treatment meningitis ), send fluid for culture and rapid analysis e.g. meningococcus (PCR), pneumococcus (latex agglutination), HSV (PCR), TB (PCR) depending upon the clinical scenario and what you consider are the likely causes in your patient.
CSF can be stained for Acid Fast Bacilli and if in doubt, repeat CSF samples should be taken separately over several days (total of 3 samples) and fluid should also be sent for long term culturing. 90% of patients with TB meningitis also have a positive tuberculin skin test and it is still worthwhile performing. A negative skin test does not rule out TB.

Thirdly, continue the antibiotics until such time that it can be established (if possible) that it is or is not a bacterial infection e.g. PCR and cultures results (CSF and blood).

In the under 50s, adults should receive cetriaxone plus vancomycin [current UK guidelines 2005 suggest only using Cefotaxime] whereas over 50s require additional ampicillin therapy to cover listeria infection.

I would hope that future handbooks and textbooks fully address the problem of partial treatment meningitis.

It just goes to show that patients do not write the textbooks and that nothing is ever fixed in stone.

Have a great week....!

No comments: