Changes in cerebral haemodynamics in cases of post-lumbar puncture headache: a prospective transcranial Doppler ultrasound study

Hartmut Göbel, Horst Klostermann, Volker Lindner, Sabine Schenkl

CEPHALALGIA Göbel H, Klostermann H, Lindner V, Schenkl S. Changes in cerebral haemodynamics in cases of post-lumbar puncture headache: a prospective transcranial Doppler ultrasound study. Cephalalgia 1990;10:117-22. Oslo. ISSN 0333-1024 We used transcranial Doppler ultrasonography in 45 patients to investigate if changes in haemodynamics in the major arteries of the brain base occurred after lumbar puncture and whether or not patients with or without post-lumbar puncture headache differ with respect to their cerebral haemodynamic parameters before and after lumbar puncture. Before lumbar puncture, patients with post-lumbar puncture headache differed from patients without post-lumbar puncture headache in that they showed significantly higher flow velocities and significant asymmetry of flow velocities with lateralization to the right (p £ 0.05). Patients without post-lumbar puncture headache, on the other hand, showed non-significant flow velocity lateralization to the left. Forty-eight hours after lumbar puncture, both groups demonstrated symmetrical flow velocities. In addition, only patients with post-lumbar puncture headache showed a significant reduction in the flow velocity of the right middle cerebral artery (p £ 0.05). These findings suggest that it is not only absolute flow velocity that plays a part in the event of headache, the interhemispheric relation of cerebral haemodynamics also plays a fundamental role. • Cerebral haemodynamics, flow velocity, interhemispheric relation, post-lumbar puncture headache, spinal tap, transcranial Doppler ultrasonography H Göbel, H Klostermann, V Lindner, S Schenkl, Department of Neurology, Christian Albrechts University, Niemannsweg 147, D-2300 Kiel 1, FRG; Correspondence to H Göbel; Accepted 9 March 1990 The pathogenesis of post-lumbar puncture syndrome, one of the most frequent types of iatrogenic pain, is a topic of controversial discussion (1-9). It is assumed that the dura defect produced by lumbar puncture is delayed in closing, and that continuously draining cerebrospinal fluid lost through this leak leads to a reduction of fluid pressure (puncture-hole-seepage theory). This leads to dilatation of intracranial vessels and also to pressure-induced shift of brain tissue causing traction on pain-sensitive structures (1-3, 5, 8, 9). Other theories favour vasomotor dysregulation (6), biochemical (7) or psychic variables (4) as the most important pathogenetic principle. The puncture-hole-seepage theory, however, fails to explain why only a proportion of patients develop typical position-dependent headache and other accompanying symptoms, despite the use of identical cannulas and under otherwise identical conditions. This leads to the question whether predictors for the genesis of post-lumbar puncture headache can be found. Since low pressure of cerebrospinal fluid and consequent dilatation of intracranial vessels has been postulated as being the critical pathogenetic principle, we investigated whether changes in the haemodynamics in the main branch of the middle cerebral and basilary arteries, which might possibly occur in vasodilatation due to low pressure, could be registered in cases of post-lumbar puncture headache using transcranial Doppler ultrasonography. A precise, non-invasive assessment of intracerebral vessel diameter is not possible. Using transcranial Doppler ultrasonography it is only possible to measure the flow velocity of blood in the main trunks of the cerebral vessels. Thus absolute statements about blood flow (ml/sec) or even vessel diameter are not obtainable. Under the premises of this limitation it was of particular interest to

see if patients with and without post-lumbar puncture headache differ in respect to the cerebral haemodynamic parameters measured by transcranial Doppler ultra-sonography before and after lumbar puncture. Method

Design A prospective study was done on 45 neurological inpatients who required lumbar puncture. All suitable patients were included in the sample. The criteria for exclusion were emergency punctures, the taking of analgesics and immobilization. The first reading of cerebral flow velocities (taken transcranially using Doppler ultrasonography) was performed on the day before lumbar puncture. Subsequently the patient's personality profile was documented quantitatively using the Freiburg Personality Inventory (10). Lumbar puncture was performed by any of 10 attending physicians. Puncture was performed following a standard protocol. On the day of lumbar puncture and on seven consecutive days, the intensity of position-dependent headache was registered quantitatively using a verbal pain rating scale on an hourly basis (11). Forty-eight hours after lumbar puncture, the second transcranial Doppler ultrasound reading was taken. Neither patient nor attending physician knew the results of the ultrasound test. By the same token, the two physicians performing the ultrasound examination were not informed of the clinical algesimetric results (double-blind set-up). Assignment of patients to investigators was made in a randomized fashion. Patients The 45 patients entered the study in the order in which they were admitted to the hospital. There were 19 women and 26 men. The mean age was 35 ± 16 years; the ages ranged from 20 to 80 years. They were all right-handed. Their weight was 69 ± 14 kg; their height was 172 ± 11 cm. Each patient's treatment as well as other diagnostic measures was performed independently of this study. Eight of the 45 patients were diagnosed as having disseminated encephalomyelitis. The other patients suffered from diverse diseases: no one particular diagnosis showed a significant incidence. Apparatus and procedure Transcranial Doppler ultrasonography. The readings were taken using a TC-64 by Eden, Überlingen, FRG. The ultrasound frequency was 2 MHz, the power 10 mW/ cm2. The Doppler signal was evaluated using a frequency analysis based on the Fast Fourier Transformation principle. The maximum systolic velocity and the mean velocity were evaluated. During the reading, the patients lay supine on the examining cot. After application of contact gel, the transducer was placed between the lateral orbital margin and the ear, above the zygomatic bone on the temporal squama. Readings were taken on both sides. The readings were taken in the early afternoon between 1300 and 1500. Lumbar puncture. This was performed between the spinous processes of L3 and L4 with the patient in a sitting position. Disposable spinal cannulas cut according to Quincke (0.9 x 86 mm; 20G x 31/2"), Spinocan(r), by Braun, Melsungen, FRG, were used. The cerebrospinal fluid drawn amounted to 8-10 ml. Punctures were performed between 0900 and 1100. Analysis of data The t-test for independent and dependent samples was used to determine the significance of the differences in mean value between and within the subject groups. Correlations were described using the product-moment-correlation coefficient. Alpha was set at 0.05 for all statistical tests. Results

Most of the post-puncture complaints occurred on the second day following lumbar puncture. The relative frequencies of various

position-dependent headache intensities are shown in Fig. 1. About 31% of the patients were shown to suffer from strong or very strong headache. No headache was reported by 58% of the patients. There proved to be no significant difference in the expression of post-lumbar puncture headache between men and women. Since cerebral blood flow velocities vary with age, the age distribution pattern of patients with post-lumbar puncture headache was compared to that of patients without post-lumbar puncture headache. The group of patients with post-lumbar puncture headache had a mean age of 37 ± 14 years, those without post-lumbar puncture headache 32 ± 11 years. This difference proved to be non-significant. Nor were there any important differences in systolic and diastolic blood pressure values between both groups of patients. The degree of personality dimensions in the Freiburg Personality Inventory and the intensity of post-puncture headache were not shown to be significantly correlated. In 36 of the 45 patients the skull bone could be penetrated with ultrasound without difficulty. The average depth for measurements was 52.5 ± 4.3 mm. In order to determine the reliability of transcranial Doppler ultrasonography the correlation coefficients between the respective parameters of the first and second reading as well as between the right and left middle cerebral artery were determined (Table 1). With one exception the correlations are all significant, so the reliability can be considered to be sufficient. Fig. 2 shows maximum systolic velocities in relationship to the time of transcranial Doppler ultrasonography and to occurrence of post-lumbar puncture headache for the right middle cerebral artery. There proved to be a significant difference in the maximum systolic velocity in the right middle cerebral artery between patients with and without post-lumbar puncture headache on the day preceding lumbar puncture: those patients later developing post-lumbar puncture headache showed significantly higher flow velocities. The change in flow velocity following lumbar puncture is meaningful only in the group of patients suffering from post-lumbar puncture headache: they showed a significant drop in maximum systolic flow velocity. Patients not suffering from post-lumbar puncture headache showed no significant change in flow velocity. This is illustrated in Table 2 showing the number of patients demonstrating an increase or drop in flow velocities or the cumulative change in flow velocities with respect to post-lumbar puncture headache. The last parameter in particular shows a highly significant drop in flow velocity patients with post-lumbar puncture headache (p £ 0.0001). Concerning the haemodynamic parameters of the left middle cerebral artery, no significant effects between or within the groups could be found. The patient's age and maximal systolic flow velocity in the middle cerebral artery are correlated in an inverse fashion (r = -0.41; p £ 0.01), i.e. as age increases, flow velocities tend to decrease. Flow velocities and height as well as body weight also showed

Table 1. Correlation matrix for assessing the reliability of flow velocities in the right or left middle cerebral artery determined by transcranial Doppler ultrasound before and 48 h following lumbar puncture. Before LP After LP SFL MFL SFR MFR S FL M FL S FR M FR Before SFL 1.0 LP MFL 0.76** 1.0 SFR 0.59** 0.49* 1.0 MFR 0.58** 0.60** 0.89** 1.0 After SFL 0.53* 0.55** 0.43* 0.36 1.0 LP MFL 0.58** 0.67** 0.45** 0.50* 0.88** 1.0 SFR 0.53* 0.50* 0.78** 0.68** 0.60** 0.57** 1.0 MFR 0.54** 0.53** 0.69** 0.63** 0.57** 0.63** 0.91** 1-tailed Significance: *: p £ 0.01; **: p £ 0.001. S = systolic; M = mean; F = flow velocity of the middle cerebral artery; L = left; R = right.

1.0

significant negative correlation coefficients. The relationship is most obvious between the flow velocity in the left middle cerebral artery and body weight (r = -0.52; p £ 0.001) as well as between the flow velocity in the left middle cerebral artery and height (r = -0.50; p £ 0.01); thus short and especially slender patients have greater flow velocities. Concerning flow velocities in both middle cerebral arteries, significant findings were possible for the right side only. For this reason the data were also analysed under the aspect of lateralization of cerebral haemo-dynamics. Within the groups, a comparison of the mean flow velocities between the right and left middle cerebral arteries shows significant lateralization in patients with post-lumbar puncture headache before lumbar puncture (Fig. 3): the mean flow velocity on the right is significantly higher than on the left. Patients who did not develop post-lumbar puncture headache, however, showed no significant lateralization. It is of particular interest to note that the relationship between right and left flow velocities is inverse in the group with post-lumbar puncture headache. The right/left comparison of mean flow velocity 48 h following diagnostic spinal tap showed no more significant differences in either group. The differences were compensated by the time this reading was taken. Discussion

The genesis of the post-puncture headache syndrome is unclear. The outcome of our study concerning this problem is threefold: 1. Patients who develop post-lumbar

Table 2. Number of patients and cumulative changes in flow velocity (difference between systolic flow velocity of the right middle cerebral artery before and 48 h following puncture) in relation to post-lumbar puncture headache and increase or drop in flow velocity. Decrease of Increase of flow velocity flow velocity Cumulative Cumulative No. of flow velocity No. of flow velocity patients difference patients difference With PLPH 14 +141 5 -32 Without PLPH 7 +40 10 -66 2 x -tests Count of patients: X2 = 3.901; p £ 0.05 Cumulative flow velocity difference: X2 = 55.252: p £ 0.0001

puncture headache, as compared to patients without subsequent post-lumbar puncture headache, show significantly greater systolic flow velocity in the right middle cerebral artery prior to lumbar puncture; there is no such difference on the left side. The results suggest that increased flow velocity is a predictor of post-lumbar puncture headache. Young, slender and short patients tend to show higher flow velocities. Clinical experience shows that the probability of occurrence of post-puncture complaints in these patients is particularly high. Recently we could show that the intensity of post-puncture headache varies depending on the experimentally established pain sensitivity (12). 2. Only those patients who develop post-lumbar puncture headache showed a significant decrease in flow velocity of the right middle cerebral artery. Because of the limitations mentioned initially the method does not explain whether changes in flow velocity depend on changes in vessel diameter of the small cerebral vessels, on flow volume or on a combination of the two. Nevertheless, empirical evidence of a relationship between a decrease in intracranial flow velocity and post-lumbar puncture headache could be demonstrated prospectively because only those patients who developed post-lumbar puncture headache showed a significant decrease in flow velocity. Reinecke and Langohr (13) showed that flow velocity is significantly lower in a standing as opposed to a sitting position. Thus not only headache intensity, but also flow velocity prove to be position-dependent. The effectiveness of horizontal positioning can be explained by a reduction of hydrostatic pressure in supra-aortal vessels with corresponding increase of flow velocity and subsequent reduction of headache intensity.

There was no significant relationship between personality traits and post-lumbar puncture headache. In view of the changes in cerebral haemodynamics it then appears improbable that post-lumbar puncture headache is induced by purely psychic mechanisms. In retrospect it would be of interest to apply these results to the pressure of cerebrospinal fluid, as such investigations have not yet been done. Earlier investigations indicate that a significant relationship between the pressure of cerebrospinal fluid measured during puncture and post-lumbar puncture headache does not exist (14). For this reason we consciously did not determine cerebrospinal fluid pressure in order to keep the standardized puncture procedure as simple as possible. 3. The relationship of the flow velocity between the right and left hemispheres, however, appears to be just as significant as the absolute value of the haemodynamic parameters. Patients who develop post-lumbar puncture headache are characterized before lumbar puncture by significant lateralization of the mean flow velocity of the middle cerebral artery. The flow velocity in the right artery predominates significantly. Patients who do not develop post-lumbar puncture headache differ in two ways: first, they show no significant lateralization and, secondly, the flow velocity in the left middle cerebral artery outweighs the flow velocity in the right. Forty-eight hours after lumbar puncture both groups show flow velocity symmetry. This prospective observation may suggest that this effect is brought about by the breakdown of significant lateral asymmetry of flow velocity in patients who show marked increase in flow velocity in the right middle cerebral artery prior to lumbar puncture. The observations described here give reason for further investigations. Seemingly critical is the inference that a disturbed imbalance between hemispheres is significant for the genesis of headache. References

1.

Diener HC, Bendig M, Hempel V. Der post-punktionelle Kopfschmerz. Forschr Neurol Psychiatr 1985;53:344

2.

Dieterich M. Postpunktionelles Liquorunterdruck-syndrom. In: Diener HC. Dichgans J, Brandt T eds Therapie und Verlauf neurologischer Erkrankungen. Stuttgart: Kohlhammer 1987:49-51

3.

Hilton-Jones D. What is postlumbar-puncture headache and is it avoidable? In: Warlow C, Garfield J eds Dilemmas in the management of the neurological patient. Edinburgh, London. Melbourne, New York: Churchill Livingstone 1984;144-57

4.

Kaplan G. The psychogenic etiology of headache post lumbar puncture. Psychosom Med 1967;29:376

5.

Olsen KS. Epidural blood patch in the treatment of post-lumbar puncture headache. Pain 1987;30:293

6.

Säker G. Stichlochdrainage, postpunktionelle Kopfschmerzen und Liquorresorptionsprüfungen. Nervenarzt 1953;24:237

7.

Sicuteri F. Headache and meningism evoked by lumbar puncture and subarachnoid hemorrhage: A biochemical interpretation. Headache 1966:5:108

8.

Tourtellotte WW, Haerer AF, Heller GL, Somers JE. Post-lumbar puncture headaches. Springfield: CC Thomas 1964

9.

Wolff HG. Headache and other head pain. New York: Oxford University Press 1972, 3rd ed

10.

Fahrenberg J, Selg H, Hampel R. Freiburger Persönlichkeits-inventar. Göttingen: Hogrefe 1978

11.

Göbel H, Heller O, Nowak T, Wesphal W. Zur Korrespondenz yon Schmerzreiz and Schmerz-erleben. Der Schmerz 1988;2:205

12.

Göbel H, Schenkl S. Post-lumbar puncture headache: the relation between experimental suprathreshold pain sensitivity and a quasi-experimental clinical pain syndrome. Pain 1990;40:267

13.

Reinecke M, Langohr HD. Darstellung vegetativer Regulationen (Valsalva. Aufstehen. Atemanhalten) über der A. cerebri media mit der transcraniellen Doppler-Sonographie. In: Widder Bed Transcranielle Doppler-Sonographie bei zerebro-vaskulären Erkrankungen. Berlin, Heidelberg. New York. London. Paris. Tokyo: Springer 1987: 149-54

14.

Marschall J. Lumbar-puncture headache. J Neurol Neurosurg Psychiatry 1950:13:71