Cerebrospinal fluid (CSF) analysis has perhaps the strongest diagnostic and prognostic potential for Alzheimer disease to date, yet some think the field is not harnessing its full power because patients, doctors, and even some investigators harbor misconceptions about spinal taps. In Sweden, lumbar punctures are as routine as drawing blood, but they are fairly shunned in many parts of the world, including most of the Americas, parts of Europe, and Asia.

For Alzheimer disease, amyloid imaging with PET ligands is increasingly seen as a better choice by some. Is the field missing out on a safe, fast, cheap, and radiation-free alternative to PET? Are physicians reluctant to perform lumbar punctures (LPs)? Are patients too squeamish of the infamous “LP headache” to consent to one? Studies show that complications from lumbar punctures are mild, and that the headache is extremely rare in older adults. Not convinced? Sweden’s Kaj Blennow, a world leader in neurological CSF analysis, has proposed a multinational study to track trends and outcomes in LP procedures, with a view to identifying the incidence and cause of complications. The answers could lead to greater acceptance of, and best-practice procedures for, spinal taps for AD around the world.

Blennow, together with panelists John Morris, Elaine Peskind, and Philip Scheltens, held a Webinar about LPs, a new research opportunity, and how CSF analysis fits into the bigger picture of AD research.

This video, courtesy of Washington University, St. Louis, School of Medicine, and hosted on their website, describes patients’ pre- and post-reactions to the procedure. It also shows a lumbar puncture being performed.

  • Click on the image to view the video.




  • Click on the image to view the Webinar.



View a larger version of the slides below.


Background Text
By Tom Fagan

Cerebrospinal fluid analysis could have immense diagnostic and prognostic potential for Alzheimer disease. The research field now shares a general consensus that reduced CSF Aβ42 indicates amyloid buildup in the brain, and together with increased CSF tau and phospho-tau, can predict progression from mild cognitive impairment to AD (see Hansson et al., 2006; Fagan et al., 2006; Shaw et al., 2009; and Mattsson et al., 2009). Data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) shows that CSF Aβ42 can predict brain atrophy in cognitively normal adults as well (see Fjell et al., 2010), confirming similar data by others (see Fagan et al., 2007). “CSF analysis continues to be an indispensable research tool and promises to be invaluable for diagnostic and prognostic purposes,” ADNI principal investigator Mike Weiner, University of California San Francisco, told ARF. “It is imperative that the Alzheimer’s community know that the procedures involved are extremely safe.”

But analyzing CSF requires a lumbar puncture (LP), a prospect that sends a shiver down the spines of many patients—and possibly their physicians as well, especially perhaps those who have not experienced the minor nature of epidural injection during the major throes of childbirth. Is fear of LP limiting CSF analysis? Are those fears unfounded? This is the opinion of Kaj Blennow of the University of Gothenborg, Sweden, who has been a leader analyzing CSF for AD prediction for many years.

In a handful of countries, performing an LP raises no more hackles than drawing blood, and studies suggest that complications, such as the infamous LP headache, are uncommon. A recently published analysis of nearly 1,100 patients at a memory clinic in Lund, Sweden, indicated that the incidence of all LP complications is low. Led by Blennow, the study found that only 2.6 percent of patients, most of them young, reported mild post-LP headache, and there were no other complications (see Zetterberg et al., 2010). But what about other countries? In ADNI roughly 50 percent of participants volunteered for LPs, but some researchers question whether that highly motivated group is representative of the general population (see Shaw et al., 2009). What kind of objective data would allay entrenched reservations among physicians and study participants?

Blennow has proposed a multinational, multicenter, prospective study to investigate the incidence of headache and other complications arising from lumbar punctures. The “Alzheimer’s Association Multi-Center Study on Lumbar Puncture Feasibility” will be sufficiently powered to identify factors—be they related to the procedure used, or the patient—that might increase the risk of such complications. Some evidence already exists that LP headache can be psychosomatic. An old study carried out in the U.S., for example, found that people who admitted being afraid of the process were more likely to develop a headache even after a sham LP (see Kaplan, 1967). Blennow’s proposed study will collect patient data, including age, cognitive status, and psychological attitude to the procedure. It will record details of the procedure, including type of needle, whether the fluid is collected passively (drip) or actively (by withdrawal), the posture of the patient during the procedure, and the experience of the physician or specialist involved. On all these points, leading research groups have developed procedural protocols to minimize the risk of side effects. A five-minute post-procedure survey will record the type and severity of complications. The full protocol is freely available from the Alzforum protocol database.

Blennow hopes to involve as many as 50 to 100 centers in the study, and plans to survey 5,000 to 10,000 LPs. He has not applied for specific funding for the study, believing that because it will take so little time on the part of participating centers, it could be built into existing clinical or research studies that ask patients to undergo LPs. The Alzheimer’s Association has given funding for one person to coordinate with interested researcher centers, curate samples, and oversee data collection.

On 5 August 2010, Blennow presented his study in the context of AD research and clinical care. Consider participating in the study. We were pleased to present a distinguished panel including Elaine Peskind, University of Washington, Seattle; Philip Scheltens, VU University Medical Center, Amsterdam, Holland; and John Morris, Washington University, St. Louis, Missouri, whose group has played a leading role in biomarker research using lumbar puncture.

imageBelow are questions that were not addressed during the Webinar. Answers are by Kaj Blennow.

Q: Can artificial CSF be replaced to reduce the headache?

A: This has, as far as I know, not been examined. It might introduce a risk of infection.

Q: Will there be any recommendations in the protocol regarding how to perform the LP (sitting vs. lying, time resting lying down after LP, etc.)?

A: This is published in the Nature Rev Neurol paper (see Blennow et al., 2010).

Q: We are beginning a multicenter study to evaluate the utility of evaluating event-related potentials in diagnosing Alzheimer disease. We are doing LPs on 100 subjects and would be interested in participating in the study. We would like to add your LP data collection forms to our forms and then send them to the coordinating center. Is it possible to get access to the LP data forms?

A: The LP data forms and all information will be available on Alzforum.

Q: What would you consider to be adequate training for a physician to be able to independently do LPs? Or otherwise stated, how many LPs does it take for the physician to be considered acceptably competent to participate in the AA-multicenter study on LP feasibility?

A: This is a difficult question. Doing LPs is part of the training to become an M.D. in most countries, and is necessary to serve, for example, in emergency wards or departments where patients with CNS infections may come. So I would guess that most physicians have the necessary training.

Q: How do LP-inexperienced neurologists (in the U.S.) make diagnoses of MS? Does it mean that neurologists using LP for MS diagnostics will not have any hesitation to perform LP for AD?

A: LP is much more commonly used for the diagnosis of MS.

Q: Will pre-LP neuroimaging information be collected?

A: This would be interesting, but would also make the study too complicated.

Q: It seems like the greatest value of this study would be obtained from enrolling asymptomatic subjects, say age 40 and older, rather than those already diagnosed with AD or dementia. The types of concerns and incidence of AEs would be more useful and better applied to future screening for potentially lowered Aβ. For the already diagnosed AD patients, the Aβ would have already dropped and that group may be of less research value in evaluating the CSF biomarkers as screening tests. The study could also lead to an estimate of the prevalence of lowered Aβ.

A: These are good points, but the aim of this study is not to examine low Aβ.

Q: Do you perform LP in patients treated with one anti-platelet drug? With two anti-platelet treatments?

A: This is up to the clinician.

Q: Would you not also welcome participants who might be in presymptomatic stages of AD, i.e., "normal elderly"?

A: Yes.

Q: Will CSF biomarker data be published from this study?

A: The aim is not to publish biomarker data, but to examine the incidence of complications, and factors that may predict potential complications.


  1. Briefly, my view, having been doing LPs for many years, is that it is not the fear of complications or the procedure, but the lack of interest and perceived utility of LPs (for dementia diagnosis) that is the root of low LP rates in the U.S. In the U.S. ADNI, the LP rates were relatively high for the research centers, but again, in the community, the physician-driven attitude is "we don't need this," and I think this then pervades the general community attitude (other than for research) and explains why it is not easy to recruit other than for specialized research purposes.

    Keep in mind that probably up to half of all dementia cases in the community in the U.S. go undiagnosed. The geriatrics medical community would be thrilled to simply see community-based doctors more routinely performing the mental status exams needed to diagnose dementia. LPs are not on the radar as a practical, widely deployable procedure needed for dementia diagnosis at this point.

  2. Other possible reasons why many clinicians hesitate to use lumbar puncture (LP) as a diagnostic test for AD and certain dementias include patient misconception that LP is a "dangerous" procedure, the cost of LP and care of its potential complications, non-routine CSF tests, and lack of specificity of biomarkers we have currently identified. We should also be aware that physician performance, even in major U.S. academic institutions, is based largely on Relative Value Units, and LP is viewed as labor intensive and not cost effective. The trend on what workup physicians should pursue to diagnose AD and many dementias due to neurodegenerative disorders has moved toward evidence-based practice, and both the American Academies of Neurology and Family Medicine do not recommend LP.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Association between CSF biomarkers and incipient Alzheimer's disease in patients with mild cognitive impairment: a follow-up study. Lancet Neurol. 2006 Mar;5(3):228-34. PubMed.
  2. . Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol. 2006 Mar;59(3):512-9. PubMed.
  3. . CSF biomarkers and incipient Alzheimer disease in patients with mild cognitive impairment. JAMA. 2009 Jul 22;302(4):385-93. PubMed.
  4. . Brain atrophy in healthy aging is related to CSF levels of Aβ1-42. Cereb Cortex. 2010 Sep;20(9):2069-79. PubMed.
  5. . Cerebrospinal fluid tau/beta-amyloid(42) ratio as a prediction of cognitive decline in nondemented older adults. Arch Neurol. 2007 Mar;64(3):343-9. Epub 2007 Jan 8 PubMed.
  6. . Low incidence of post-lumbar puncture headache in 1,089 consecutive memory clinic patients. Eur Neurol. 2010;63(6):326-30. PubMed.
  7. . Cerebrospinal fluid biomarker signature in Alzheimer's disease neuroimaging initiative subjects. Ann Neurol. 2009 Apr;65(4):403-13. PubMed.
  8. . The psychogenic etiology of headache post lumbar puncture. Psychosom Med. 1967 Jun-Aug;29(4):376-9. PubMed.
  9. . Cerebrospinal fluid and plasma biomarkers in Alzheimer disease. Nat Rev Neurol. 2010 Mar;6(3):131-44. PubMed.

Other Citations

  1. Shaw et al., 2009

External Citations

  1. View a larger version of the slides below.

Further Reading