CHAPTER4 Routine cerebrospinal fluid (CSF) analysis

1 Huge: ch04 2006/6/29 14:54 page 14 #1 CHAPTER 4 Routine cerebrospinal fluid (CSF) analysisF. Deisenhammer,aA. Bartos,bR. Egg,aN. E. Gilhus,cG. Giovannoni,dS. Rauer,eF. SellebjergfBackgroundA great variety of neurological diseasesrequire investigation of the cerebrospinal fluid(CSF) to prove the diagnosis or to rule out relevantdifferential evaluate the theoretical backgroundand provide guidelines for clinical use in rou-tine CSF analysis including total protein, albumin,immunoglobulins, glucose, lactate, cell count,cytological staining, and investigation of infec-tious Medline search for the abovementioned variables. Review of appropriate publi-cations by one or more of the task force of evidence and recommendations wasbased on consensus by all task force should be analysed immediately after collec-tion.

2 If storage is needed 12 ml of CSF shouldbe partitioned into three to four sterile of Neurology, Innsbruck MedicalUniversity, Austria;bDepartment of Neurology, CharlesUniversity, Prague, Czech Republic;cDepartment ofClinical Medicine, University of Bergen, Bergen,Norway, and Department of Neurology, HaukelandUniversity Hospital, Bergen, Norway;dDepartment ofNeuroinflammation, Institute of Neurology, UniversityCollege London, Queen Square, London, UK;eDepartment of Neurology and ClinicalNeurophysiology, Albert-Ludwigs University, Freiburg,Germany;fDepartment of Neurology, CopenhagenUniversity Hospital, CSF/serum ratio (Qalb)should be pre-ferred to total protein measurement and normalupper limits should be related to patients a non-specific finding but occursmainly in bacterial, cryptococcal, and tubercu-lous meningitis, leptomingeal metastases as wellas acute and chronic demyelinating decrease of the CSF/serum glu-cose ratio or an increase in lactate concentrationindicates bacterial or fungal meningitis or lep-tomeningeal immunoglobulin G synthesis is bestdemonstrated by isoelectric focusing followed byspecific morphology (cytological staining) shouldbe evaluated whenever pleocytosis is found orleptomeningeal metastases or pathological bleed-ing is suspected.

3 Computed tomography-negativeintrathecal bleeding should be investigated bybilirubin cerebrospinal fluid (CSF) is a dynamic,metabolically active substance that has many14 Huge: ch04 2006/6/29 14:54 page 15 #2 Section 2: Investigation15 Table constellation of CSF parameters in some neurological protein Glucose ratio LactateCell countTypical cytology(g/l)(mmol/l)(per l)Normal valuesa< > < <15 MNCD iseaseAcute bacterialmeningitis >1000 PNCV iral neuro-infections(meningo/encephalitis)=/ =/ =10 1000 PNC/MNCA utoimmunepolyneuropathy ===Infectiouspolyneuropathy == MNCS ubarachnoidalhaemorrhage == erythrocytes,macrophages,siderophagesMNC M ultiple sclerosis====/ MNCL eptomeningealmetastases =/ NA=/ malignant cells,mononuclearsCSF,cerebrospinalfluid ;MNC,mononuclearcells;PNC,polymorphonucl earcells.

4 / , increased/decreased;=, within normal limits; NA, evidence not values aregiven for lumbar CSF in functions. It is invaluable as a diag-nostic aid in the evaluation of inflammatory con-ditions, infectious or non-infectious, involvingthe brain, spinal cord, and meninges as well asin CT-negative subarachnoidal haemorrhage andin leptomeningeal metastases. CSF is obtainedwith relative ease by lumbar puncture (LP). Alter-ations in CSF constituents may be similar in dif-ferent pathologic processes and cause difficultiesin interpretation. Combining a set of CSF vari-ables referred to as Routine parameters ( deter-mination of protein, albumin, immunoglobulin,glucose, lactate, and cellular changes, as well asspecific antigen and antibody testing for infectiousagents) will increase the diagnostic sensitivity aim of this guideline paper was to pro-duce recommendations on how to use this set ofCSF parameters in different clinical settings andto show how different constellations of these vari-ables correlate with diseases of the nervous system(table ) (Braininet al.)

5 , 2004).Search strategyA Medline search using the search terms cere-brospinal fluid (CSF), immunoglobulin G (IgG)immunoglobulin M (IgM), immunoglobulin A(IgA), and albumin was conducted. Also, the keywords cerebrospinal fluid or CSF were cross-referenced with glucose , lactate , cytology , cell in title excluding child . Furthermore, asearch for cerebrospinal fluid and immunoglobu-lin and diagnosis and electrophoresis or isoelec-tric focusing was performed limited to the timebetween 1 January 1980 and 1 January 2005, andreturned only items with abstracts, and Englishlanguage (274 references). A search for cere-brospinal fluid AND infectious limited for time(1 January 1980 until now) returned 560 that primarily did not deal with diagnos-tic issues and infectious CSF ( non-infectiousinflammatory diseases, vaccination, general CSFparameters, pathophysiology, cytokines and ther-apy) were excluded resulting in 60 : ch04 2006/6/29 14:54 page 16 #316 Chapter 4.

6 Routine cerebrospinal Fluid (CSF) AnalysisSearching the items cerebrospinal fluid AND serology limited for time (1 January 1980 untilnow) and excluding abstracts not directly relatedto the topic returned 35 abstracts and a search for cerebrospinal fluid AND bacterial culture lim-ited for time (1 Jan 1980 until now) resulted in28 abstracts were selected by the author whowas in charge of the respective addition, text books and articles identified inreference lists of individual papers were selected ifconsidered are no guidelines for CSF analysispublished by the American Academy of Neurol-ogy (AAN). Individual task force members pre-pared draft statements for various parts of themanuscript. Evidence was classified as Class I IVand recommendations as Level A C according tothe scheme agreed for EFNS guidelines (Braininet al.)

7 , 2004). When only Class IV evidence wasavailable but consensus could be reached, the TaskForce has offered advice as Good Practice Points(Braininet al., 2004). The statements were revisedand adapted into a single document that was thenrevised until consensus was analysis of totalprotein and albuminThe blood CSF barrier is a physical barrier, con-sisting of different anatomical structures, for thediffusion and filtration of macromolecules fromblood to CSF. The integrity of these barriers andCSF bulk flow determine the protein content ofthe CSF (Thompson, 1988; Reiber, 1994). In new-borns, CSF protein concentrations are high, butdecrease gradually during the first year of life,and are maintained at low levels in adults, CSF protein concentrations increasewith age (Eeg-Olofsonet al.

8 , 1981; Statz andFelgenhauer, 1983) (Class I). The CSF to serumalbumin concentration quotient (Qalb)can alsobe used to evaluate blood CSF barrier integrity(Anderssonet al., 1994). TheQalbis not influencedby intrathecal protein synthesis, is corrected for theplasma concentration of albumin, and is an inte-gral part of intrathecal immunoglobulin synthesisformulae. TheQalbis a method-independent mea-sure, allowing the use of the same reference val-ues in different laboratories (Blennowet al., 1993;Reiber, 1995). However, there are no conclusivedata on how theQalbperforms compared to totalprotein as a measure of blood CSF barrier functionin large cohorts of unselected is a concentration gradient for total pro-tein and theQalbalong the neuraxis with thelowest concentrations in the ventricular fluid andthe highest concentrations in the lumbar sac(Thompson, 1988; Fishman, 1992).

9 A significantdecrease of theQalbwas observed from the first 0 4 ml of CSF to the last 21 24 ml of CSF obtained byLP (Blennowet al., 1993) (Class I). TheQalbis alsoinfluenced by body weight, sex, degenerative lowerback disease, hypothyroidism, alcohol consump-tion (Class II) and smoking (Class III) (Kornhuberet al., 1987; Skouenet al., 1994; Nystr met al.,1997; Seyfertet al., 2002). Posture and physicalactivity may influence the CSF protein concentra-tion, resulting in higher CSF protein concentra-tions in inactive, bed-ridden patients (Seyfertet al.,2002) (Class III). Elevated CSF protein concentra-tions can be found in the majority of patients withbacterial ( g/l), cryptococcal ( g/l),tuberculous ( g/l) meningitis and neurobor-reliosis (Stockstill and Kauffman, 1983; Sabeta,1985; Kaiser, 1998; Negriniet al.)

10 , 2000) (Class II).A concentration of> g/l is specific (99%), butinsensitive (55%) for bacterial meningitis as com-pared to a variety of other inflammatory diseases(Lindquistet al., 1988) (Class I).In viral neuroinfections CSF protein concentra-tions are raised to a lesser degree (usually< g/l)(Negriniet al., 2000) (Class II). The concentrationin herpes simplex virus encephalitis is normal inhalf of the patients during the first week of illness(Koskiniemiet al., 1984) (Class IV).Non-infectious causes for an increased CSF pro-tein and sometimes with an increased cell countinclude subarachnoidal haemorrhage, central ner-vous system (CNS) vasculitis, and CNS neoplasm(Jerrardet al., 2001) (Class IV). Elevated total pro-tein concentration with normal CSF cell count(albuminocytologic dissociation) is a hallmarkin acute and chronic inflammatory demyelinat-ing polyneuropathies but protein levels may beHuge: ch04 2006/6/29 14:54 page 17 #4 Section 2: Investigation17normal during the first week (Seguradoet al.