This question arises with predictable infrequency, roughly once or twice a year. ER physicians may desire some initial disposition/determination to be in process as the fluid specimens is fairly readily clinically available. By the time of a neurosurgical consultation and dye testing, it may be that the question can be presumptively answered. And, particularly in the patient with allergic rhinitis, we must remember that there could be a mixture of watery nasal secretions and CSF.

Dr. Carter encountered an interesting (and his first CSF-positive case by these studies) on the weekend of June 5, 2004: A 37-year-old lady with no history of trauma, tumor, or infection presented to the ER with clear watery fluid "running out of her nose". Three biochemical tests are recommended for this differential diagnosis:

1. Glucose: Nasal fluid has very little glucose -- < 10 mg/ml in fluids that we’ve tested. Normal CSF glucose level is 40-70 mg/dl.
2. Chloride: Nasal fluid chloride levels approximate that of serum; CSF chloride levels are significantly higher, in the range of 122-128 meq/L.
3. Transferrin immunofixation electrophoresis (IFE): Nasal fluid shows a transferrin pattern similar to that of serum, namely a single band in the beta globulin region. CSF transferrin presents as two β-region transferrin bands, a biclonal pattern with the 2^nd (β-2) band of somewhat weaker intensity.
4. Brown paper bag test: "Snot" dries shiny; CSF dries out like water  (we heard mention many years ago of the "brown paper bag test" in which the dripping nasal fluid is allowed to run down this type of heavy paper...mucoid nasal secretion dries shiny [somewhat like a snail track] & CSF dries out like water, leaving a dull surface).

The above patient had a nasal fluid glucose level of 48 mg/dl, chloride of 128 mmol/L, and with a biclonal transferrin band, all findings very consistent with CSF leakage. On the basis of the laboratory results, a skeptical neurosurgeon agreed to further investigate with dye studies.

Dr. Carter’s most recent similar case of several months ago showed a glucose < 10, chloride was not done, and the specimen had a monoclonal beta transferrin band (the case was signed out as "c/w nasal fluid").

Reference textbooks claim that glucose and chloride levels are too unpredictable to be useful. In a short personal experience, I (JBC) cannot agree: The glucose and chloride levels are readily and rapidly available at all hours. Then for confirmation LML is perhaps one of few hospital laboratories that can do the transferrin IFE, but on a next weekday basis.

Seroma vs. pseudomeningocele:

Especially post perispinal surgery, this DDX may come into play & be unclear if there is only rare or occassional headache. The serous fluid in a seroma has more like a normal serum protein level, whereas CSF protein is very low.

References:

1. Carter JB, In-house memo 18 June 2004