Is Positioning During Lumbar Puncture Clinically Significant?

Diagnostic lumbar puncture (LP) is increasingly being performed by neuroradiologists, with variability in positioning during the measurement of opening pressure (OP). Two experts debate whether positioning during LP is clinically significant.

Point Counter-Point Section Editors: Andrew G. Lee, MD Gregory Van Stavern, MD Is Positioning During Lumbar Puncture Clinically Significant? Christian Joseph Lueck, PhD, FRCP, FRACP, Collin McClelland, MD Diagnostic lumbar puncture (LP) is increasingly being performed by neuroradiologists, with variability in positioning during the measurement of opening pressure (OP). Two experts debate whether positioning during LP is clinically significant. Pro: Prone Positioning During Lumbar Puncture Affects Opening Pressure Measurement in a Clinically Meaningful Way: Collin McClelland, MD Theory, eminence, and evidence support the century-old medical mantra that the lateral decubitus (LD) position during LP is the most accurate means of measuring OP and that prone positioning often leads to spuriously elevated OP measurements. The notion of accessing the cerebrospinal fluid (CSF) via the lumbar spinal canal dates back to 1891 when a German surgeon, Heinrich Quincke, first described the procedure as a therapy for hydrocephalus (1). By 1916, Hans Heinrich Georg Quenckenstedt, a German neurologist with an interest in CSF dynamics, described performing LPs in the LD position using a manometer to measure OP (2). He reported a test, later termed the Queckenstedt maneuver, in which manual jugular vein compression and Valsalva maneuver were performed to increase central venous pressure (CVP), followed 10-12 seconds later by increased intracranial pressure (ICP) as measured by lumbar manometer. Failure of the Queckenstedt maneuver to induce ICP elevation in the appropriate time frame was considered an indication of spinal canal obstruction. Since the conception of the LP, neurologists have favored the LD position because it facilitates both anatomical access to the spinal canal and reduces the risk of falsely elevated OP inherent with other positions. The bulk of OP normative data is based on LPs performed in the LD position (3-5). In adults, these studies collectively show a median OP of 17-19 cm H2O (4,5). Whiteley et al (5) found the 95% confidence intervals (CIs) for OP measured in the LD position among 242 adults to be 10-25 cm H2O. Avery et al (3) found the 90% CI to be 11.5-28 cm H2O in children. The most updated criteria for the diagnosis of definite pseudotumor cerebri syndrome (PTCS) require a "properly performed" LP with OP greater than 25 and 28 cm H2O in adults and children, respectively The Canberra Hospital (CJL), Garran, Australia; and Department of Ophthalmology and Visual Neurosciences (CM), University of Minnesota School of Medicine, Minneapolis, Minnesota. The authors report no conflicts of interest. Address correspondence to Collin McClelland, MD, Phillips Wangensteen Building-9th floor, 420 Delaware Street SE-MMC 493, University of Minnesota, Minneapolis, MN 55455; E-mail: [e-mail] 268 (6). These OP threshold criteria were based on the aforementioned studies of normative data derived from LPs performed in the LD position. In the discussion of proper LP technique, Friedman et al (6) clarify that the LD position is the most accurate means of measuring OP and that sedation and Valsalva should be avoided. The transition from LD positioning, historically favored among neurologists, to an increasing number of LPs performed in the prone position is the consequence of more LPs being performed by radiologists. Although fluoroscopic guidance certainly facilitates access to the spinal canal in patients with obesity, the greater variability in the LP technique may affect OP accuracy. In a survey of 577 neuroradiologists, 88% and 12% punctured the thecal sac in the prone and LD positions, respectively (7). Among those who puncture in the prone position, only 28% rotate the patient to the LD position. When OP is measured in the prone position, 21% of radiologists do not add the needle length to the measured manometer reading, potentially yielding falsely low results. Prone positioning introduces numerous theoretical and observed cardiopulmonary changes that may artificially increase ICP. These include hypoventilation-induced hypercapnia and increased CVP from elevated intrathoracic and intra-abdominal pressure (8-10). Transient hypoventilation and increased CVP from prone positioning may result in a Valsalva-like state, which can elevate OP values in a matter of seconds. A study of 15 patients undergoing LP in the LD position found that intentional Valsalva resulted in a mean pressure elevation of 17.7 cm H2O and a zenith OP of .25 cm H2O in all subjects (11). Of course, optimal LP technique in either the prone or LD position requires that patients be instructed to breathe normally and relax to minimize Valsalva (12-14). A decrease in respiratory compliance in the prone position, however, could increase the risk of hypercapnia and elevated CVP even in properly instructed patients (10). This would be a particular concern in the overweight subset of patients being evaluated for PTCS. Lueck and McClelland: J Neuro-Ophthalmol 2019; 39: 268-272 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Point Counter-Point Two prospective studies compared OP variance between otherwise optimally performed LPs in the LD and prone positions (12,14). In both studies, OP measurements were taken sequentially during the same procedure, patients were advised to breath normally, the length of the needle was added to the prone OP manometry reading, and OP measurements were recorded after CSF levels had equilibrated and respiratory fluctuations had begun. Among 51 patients who underwent measurement in the prone position, followed by the LD position, Abel et al (12) found the mean OP in the prone (27.7 cm H2O) as compared to the LD (26.5 cm H2O) position to be 1.2 cm higher. This was reported as neither statistically nor clinically significant. Among 55 patients whose order of positioning was randomized, Schwartz et al (14) determined that the mean OP in the prone (15.3 cm H2O) as compared to the LD (12.6 cm H2O) position to be 2.7 cm H2O, which was statistically significant (P = 0.001). The authors believed that OP should be measured in the LD position "whenever possible" and that OP position differences are clinically relevant. Interestingly, both studies showed similar findings of a higher OP in the prone position, yet their conclusions are contradictory (12,14). On closer inspection, both studies provide evidence that position-related differences in OP may be clinically relevant for an individual patient. Although Abel et al (12) found a small, statistically insignificant, average difference in OP between positions, 8 of 51 patients (15.7%) had OP measurements that crossed from below the 25 cm H2O threshold in LD to above the threshold in the prone position. In 4/51 patients (7.8%), prone OP was $10 cm H2O higher than the LD position. Schwartz et al (14) found that 25/55 patients (45.5%) had a prone OP .5 cm H2O as compared to LD. Considering the data at a granular level, we can see how a falsely elevated OP on prone LP could easily lead to an erroneous diagnosis of PTCS by meeting the 25 cm H2O threshold. Although it should be acknowledged that a single OP in any position is a snapshot surrogate value of ICP subject to error and should not be overly relied on, clinicians less familiar with PTCS may not be aware of this potential pitfall. Even for neuro-ophthalmologists who often trust their pre-LP clinical suspicion for high ICP more than the actual OP results, there are instances when an OP is diagnostically important to rule in or rule out elevated ICP. Despite their study conclusion that prone positioning is not on average clinically significant, Abel et al acknowledge the importance of LP positioning in the consideration of individual patients: "When OP is needed to certify a clinical diagnosis as in IIH, the trend toward higher OP in the prone position that we identified should be considered, especially when the measurement is borderline. If a clinical diagnosis is questionable because of an OP near the 25 cm H2O threshold, we think it is reasonable to roll the patient to the LD position or repeat the LP with OP measurement in the LD position." Indeed, when making decisions regarding the diagnosis and management of PTCS, clinicians should bear in mind that prone OPs commonly (up to 45.5%) exceed LD OPs by .5 cm H2O (14). The notion that LP position does and should influence clinical interpretation of the OP results is sufficient proof that LP position is clinically meaningful in the care of the individual patient. Con: Prone Positioning During Lumbar Puncture Does Not Affect Opening Pressure Measurement in a Clinically Meaningful Way: Christian J. Lueck, MD Measurement of ICP plays an integral part in the diagnosis and management of many conditions, including acute cerebral trauma, normal pressure hydrocephalus, and spontaneous intracranial hypotension. In neuro-ophthalmic practice, measurement of ICP is most pertinent to the management of idiopathic intracranial hypertension (IIH) and related PTCSs such as cerebral venous sinus thrombosis (6). Clinically, ICP is usually measured as CSF-OP at the time of LP, and formal diagnostic criteria for IIH require that the OP is above 25 cm H2O in adults or 28 cm H2O in children (6). For a number of reasons, many diagnostic LPs are now performed under fluoroscopic guidance in radiology departments. Patients usually need to lie prone rather than in the left LD position, and it is important to remember that this necessitates rezeroing of the OP measurement (13). In addition, there is a concern that lying prone could result in an artefactually elevated OP because of compression of the Lueck and McClelland: J Neuro-Ophthalmol 2019; 39: 268-272 abdomen and consequent impairment of venous return from the head. This elevation might then lead to a diagnostic error. A sensible consideration, but is the OP elevated, and does this really matter in clinical practice? Similar to mean arterial pressure, normal ICP varies from moment to moment: short-term variations are produced by both arterial pulse and respiration. Pulse-related changes measured in the cerebral ventricles are of the order of 6 cm H2O, but the pulse waves are attenuated to about 3 cm H2O by the time they reach the lumbar thecal sac (15). Normal respiration results in changes of the order of 0.5-1 cm H2O, but a Valsalva maneuver can increase OP by more than 20 cm H2O (11). Longer term variations can occur in the form of "plateau waves" (6-12 cm H2O), particularly in the context of intracerebral pathology (16,17). Thus, a single measurement of OP with LP is only a point estimate, and this must be borne in mind when considering published 269 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Point Counter-Point normal ranges of OP (3-5,18) and the diagnostic criteria for IIH (6). What is the evidence that prone positioning makes any difference to CSF-OP at all? One study of 55 patients found that there was a significant increase in mean CSF-OP of 2.7 ± 0.7 cm H2O (P , 0.001) in the prone position (14). In these patients, the absolute value of OP in the left LD position was 12.6 ± 4.8 cm H2O, meaning that the effect of positioning was, in practice, about half the SD of the normal range. Furthermore, the individual differences in pressure between positions were highly variable, with some patients actually demonstrating lower pressures when lying prone. Another study looked at 51 patients undergoing LP for a variety of clinical reasons (including IIH) and found a nonsignificant mean increase of 1.2 cm H2O in the prone position on a background of a mean LD OP of 26.5 cm H2O (range of 11.0-48.0 cm H2O) (7). Again, the effect of prone positioning was small compared with the range, and many individual patients' pressures actually fell when lying prone. In summary, the absolute change in OP that results from prone positioning varies considerably between individuals. This variability, the fact that any difference is similar to the changes expected from normal physiology, and the fact that there is a large normal range mean that, overall, the effect of prone positioning is clinically uninterpretable and, therefore, meaningless. Could this matter when assessing an individual patient? In the IIH Treatment Trial, the mean OP was approximately 34.5 cm H2O with an SD of up to 9.4 cm H2O (19). The Bland-Altman plot from the study by Abel et al (7) showed that, in an individual, the effect of prone positioning could vary by as much as ±10 cm H2O. It is therefore theoretically possible that a measurement of OP could be moved from "normal" to "abnormal" or, conceivably, the other way round, and this might, in turn, result in a falsepositive or a false-negative diagnosis of IIH. Measurement of OP during LP, although important, should not be seen as an exact science: there are many factors that can significantly influence a point estimate of an individual patient's OP including pain, anxiety, and sedation (13,20). Patients can show an increase of OP of up to 30 cm H2O in the context of an attack of migraine (21), and values of up to 47 cm H2O have been recorded when performing a Valsalva maneuver (11). In this context, an increase of only 1.2-2.7 cm H2O must be seen as trivial. Ultimately, the diagnosis of IIH does not depend exclusively on CSF-OP. Indeed, although rare, there are cases with a good history and definite papilledema whose diagnosis seems to be IIH despite a normal CSF-OP (22). Conversely, a false-positive diagnosis of IIH is possible in an overweight female patient with migraine and optic disc drusen (misinterpreted as papilledema) whose otherwise borderline-normal OP was artificially elevated into the "abnormal" range by prone positioning. It is important that clinicians look at all aspects of the patient rather than relying solely on the OP when making a diagnosis. In conclusion, prone positioning might well result in a small change in OP pressure but the magnitude is unpredictable and is small when compared with the normal range and normal variation of OP. Any change is likely to be small when compared with the pressures generally seen in the context of IIH. As with so many things in medicine, test results must be interpreted in their clinical context rather than relying on a single value to rule in, or rule out, a diagnosis. In practice, therefore, I would submit that prone positioning during LP does not matter. Rebuttal: Dr. McClelland First, Dr. Lueck should be commended for his research, which has generated valuable OP normative data for LP in the LD position (5). He has done an excellent job presenting the argument for why the average differences between the prone position and LD position LP are relatively small and not meaningful. He has emphasized that both the inherent inaccuracy of OP and the wide variability in OP values between the LD and prone positions for any individual patient negates the clinical significance of average higher OP values in the prone position. In the study by Schwartz et al (14) comparing OP in LD and prone positions, however, 37/55 patients (67%) had an OP that was .2 cm H2O higher in the prone position, whereas only 4/ 55 (7%) had an OP that was higher in the LD position (14). Although a difference of 3 cm H2O could be regarded as clinically trivial, 25/55 patients (45.5%) had a prone OP .5 cm H2O higher than that in the LD position. Although 270 OP in the prone position may occasionally be lower than that in the LD position, this should not diminish the statistical and clinical significance of a strong trend for higher measurements in the prone position. Stepping back from our assigned hard line stances in this discussion, Dr. Lueck and I both agree that OP is a highly variable indicator of actual ICP and must be interpreted cautiously in the context of the entire clinical picture. An argument could be made that OP in any position is not clinically meaningful to the astute neuro-ophthalmologist. We can all likely recollect cases when we have mentally discarded an erroneous "normal" OP in a patient with typical clinical features of PTCS or a "high" OP in a patient without examination or radiographic evidence of increased ICP. For clinicians to ignore OP guidelines, however, there must exist guidelines that are generally accurate. The reality is that many providers interpreting LP results are not PTCS Lueck and McClelland: J Neuro-Ophthalmol 2019; 39: 268-272 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Point Counter-Point experts, and their diagnosis often rests on OP values alone. The diagnostic criteria for PTCS are based on normative OP values measured in the LD position. Theory and evidence support that prone positioning is subject to falsely elevated OP. On average, the difference in measured OP between positions is relatively small, although in individual patients, false elevation can easily reach clinical significance. Indeed, positioning of the LP should be kept in mind when considering OP results in any individual, de facto demonstrating that position is clinically meaningful. Rebuttal: Dr. Lueck Dr. McClelland's statement is thoughtful and well considered. However, I would suggest that most his arguments relate to inaccurate performance of the test (i.e., failing to rezero when the patient is in the prone position) or inaccurate interpretation of the test in its clinical context. Both of these are likely to relate to suboptimal education of the clinicians involved and not to whether measurement of OP in the prone position itself matters. Provided that the radiologist and treating clinician know what they are doing, these arguments become insignificant. As I have discussed above, measurement of OP is not a precise science: any alteration in OP generated by lying the patient prone falls within the range of "experimental error." Therefore, I cannot agree that this practice should make a difference to an astute physician. One further aspect that must be considered is the patient's view. In practice, many "unscreened" LPs in the LD position fail (often after multiple attempted penetrations), so patients are then referred for a repeat LP in the radiology department. These patients are understandably anxious at having to have the test repeated, and the process of rolling over from prone to LD with a needle inserted into their back is likely to generate a significant increase in anxiety, with subsequent distortion of the OP readings and the potential to make interpretation even more difficult. As long as the radiologist knows what to do to obtain an accurate reading (i.e., add the length of the LP needle to the pressure measured from the manometer), working with the patient lying prone is the fastest, and least traumatic, way of obtaining a measurement of OP. This, in turn, means that the reading is more likely to be representative of the true value (bearing in mind the comments above). In practice, the result of a CSF-OP measurement is usually unequivocal because the pressure will be either normal or clearly elevated in the vast majority of patients. My experience is that only a small number of patients fall into the "borderline" range. When they do, it is clinical judgment, not the measurement of OP itself, that will yield the final diagnosis and guide management, and this applies whatever the patient's position when the OP is measured. While I try to obtain the OP in the LD position wherever possible, in practice, prone positioning turns out to be the preferred (or only) option for many of my patients. The bottom line is that any clinician using any test must be aware of both the indications and the limitations of the test if it is to be used correctly. This applies to LP just like anything else. The clinician should be aware of a possible, but unpredictable, small effect of measuring OP with the patient lying in the prone position. Provided that this is taken into account, and assuming the radiologist knows, what he/she is doing and despite the arguments of my colleague, I remain of the opinion that prone positioning does not matter in clinical practice. Conclusion: Drs. Lee and Van Stavern As the performance of LP has moved primarily into the hands of the neuroradiologists, it is important that neuroophthalmologists understand the technical issues related to, and limitations of, a single measurement of OP. Clinical context is critical in all cases, but particularly in patients for REFERENCES 1. Aschoff A, Kremer P, Hashemi B, Kunze S. The scientific history of hydrocephalus and its treatment. Neurosurg Rev. 1999;22:67-93; discussion 94-165. 2. Zambito Marsala S, Gioulis M, Pistacchi M. Cerebrospinal fluid and lumbar puncture: the story of a necessary procedure in the history of medicine. Neurol Sci. 2015;36:1011-1015. 3. Avery RA, Shah SS, Licht DJ, Seiden JA, Huh JW, Boswinkel J, Ruppe MD, Mistry RD, Liu GT. Reference range for Lueck and McClelland: J Neuro-Ophthalmol 2019; 39: 268-272 whom the OP measurement is equivocal. In these cases, careful review of the procedure note and discussion with the radiologist often is necessary. Expertise in the diagnosis and management of PTCS may be a critical element in this process. cerebrospinal fluid opening pressure in children. N Engl J Med. 2010;363:891-893. 4. Bo SH, Davidsen EM, Benth JS, Gulbrandsen P, Dietrichs E. Cerebrospinal fluid opening pressure measurements in acute headache patients and in patients with either chronic or no pain. Acta Neurol Scand Suppl. 2010;190:6-11. 5. Whiteley W, Al-Shahi R, Warlow CP, Zeidler M, Lueck CJ. CSF opening pressure: reference interval and the effect of body mass index. Neurology. 2006;67:1690-1691. 271 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Point Counter-Point 6. Friedman DI, Liu GT, Digre KB. Revised diagnostic criteria for the pseudotumor cerebri syndrome in adults and children. Neurology. 2013;81:1159-1165. 7. Abel AS, Brace JR, McKinney AM, Harrison AR, Lee MS. Practice patterns and opening pressure measurements using fluoroscopically guided lumbar puncture. AJNR Am J Neuroradiol. 2012;33:823-825. 8. Dharmavaram S, Jellish WS, Nockels RP, et al. Effect of prone positioning systems on hemodynamic and cardiac function during lumbar spine surgery: an echocardiographic study. Spine (Phila Pa 1976). 2006;31:1388-1393; discussion 1394. 9. Kwee MM, Ho YH, Rozen WM. The prone position during surgery and its complications: a systematic review and evidence-based guidelines. Int Surg. 2015;100:292-303. 10. Mobley SR, Miller BT, Astor FC, Fine B, Halliday NJ. Prone positioning for head and neck reconstructive surgery. Head Neck. 2007;29:1041-1045. 11. Neville L, Egan RA. Frequency and amplitude of elevation of cerebrospinal fluid resting pressure by the Valsalva maneuver. Can J Ophthalmol. 2005;40:775-777. 12. Abel AS, Brace JR, McKinney AM, Friedman DI, Smith SD, Westesson PL, Nascene D, Ott F, Lee MS. Effect of patient positioning on cerebrospinal fluid opening pressure. J Neuroophthalmol. 2014;34:218-222. 13. Lee SC, Lueck CJ. Cerebrospinal fluid pressure in adults. J Neuroophthalmol. 2014;34:278-283. 14. Schwartz KM, Luetmer PH, Hunt CH, Kotsenas AL, Diehn FE, Eckel LJ, Black DF, Lehman VT, Lidell EP. Position-related 272 15. 16. 17. 18. 19. 20. 21. 22. variability of CSF opening pressure measurements. AJNR Am J Neuroradiol. 2013;34:904-907. Bering EA. Choroid plexus and arterial pulsation of cerebrospinal fluid. AMA Arch Neurol Psychiatry. 1955;73:165-172. Gomes JA, Bhardwaj A. Normal intracranial pressure physiology. In: Irani DN, ed. Cerebrospinal Fluid in Clinical Practice. Philadelphia, PA: Saunders, 2009:19-25. Kasprowicz M, Lalou DA, Czosnyka M, Garnett M, Czosnyka Z. Intracranial pressure, its components and cerebrospinal fluid pressure-volume compensation. Acta Neurol Scand. 2016;134:168-180. Fleischman D, Berdahl JP, Zaydlarova J, Stinnett S, Fautsch MP, Allingham RR. Cerebrospinal fluid pressure decreases with older age. PLoS One. 2012;7:e52664. The NORDIC Idiopathic Intracranial Hypertension Study Group Writing Committee. Effect of acetazolamide on visual function in patients with idiopathic intracranial hypertension and mild visual loss. JAMA. 2014;311:1641-1651. Friedman DI. The pseudotumor cerebri syndrome. Neurol Clin. 2014;32:363-396. Van Oosterhout WPJ, Gagaouzova BS, Terwindt GM, Schoonman GG, Vein AA, Ferrari MD. Transient intracranial pressure changes during a migraine attack. Cephalalgia. 2013;33:956. Green J, Newman NJ, Stowe Z, Nemeroff C. "Normal pressure" pseudotumor cerebri. J Neuroophthalmol. 1996;16:241-246. Lueck and McClelland: J Neuro-Ophthalmol 2019; 39: 268-272 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.