- Email: [email protected]
Intrathecal release of sICAM-1 into CSF in neuroborreliosis — increased brain-derived fraction
Journal of Neuroimmunology 103 Ž2000. 93–96 www.elsevier.comrlocaterjneuroim
Intrathecal release of sICAM-1 into CSF in neuroborreliosis — increased brain-derived fraction P. Lewczuk b
a,)
, H. Reiber a , G.C. Korenke b, E. Bollensen c , A.J. Dorta-Contreras
d
a Neurochemistry Laboratory, UniÕersity of Goettingen, Robert-Koch-Str. 40, 37075 Goettingen, Germany Department of Pediatric Neurology, UniÕersity of Goettingen, Robert-Koch-Str. 40, 37075 Goettingen, Germany c Department of Neurology, Hospital in Eschwege, Bahnhofstr. 24, 37269 Eschwege, Germany d Immunology Laboratory, Hospital Pediatrico SanMiguel, Apartado 10049 CP 11000 Ciudad HaÕana, Cuba
Received 30 July 1999; received in revised form 15 September 1999; accepted 29 September 1999
Abstract In the present study, we report sICAM-1 concentration in the cerebrospinal fluid ŽCSF. and serum of patients with neuroborreliosis ŽNB, n s 11., compared to the data from a control group of patients with corresponding bloodrCSF barrier dysfunction but without inflammation in the central nervous system Ždisc prolaps, DP, n s 11.. In NB, the sICAM-1 concentration in CSF was increased up to six-fold Žranges: 6.6–42.8 ngrml and 2.2–9.8 ngrml for NB and DP respectively. with no change in serum sICAM-1. The corresponding sICAM-1 CSFrserum concentration quotients Ž Q ICAM . were in the ranges: 22.5–171.3 = 10y3, and 8.8–27.8 = 10y3 for NB and DP respectively. This finding can be explained by increase of the brain-derived fraction of sICAM-1 in NB. In one case we observed increased Q ICAM on 6th day after admission to the hospital Ž171.3 = 10y3 at the time of the first lumbar puncture slightly increasing to 243.6 = 10y3 five days later., followed by normalization, in two remaining repunctured patients we observed decreasing Q ICAM with normalizing QAlb . q 2000 Elsevier Science B.V. All rights reserved. Keywords: Neuroborreliosis; Cerebrospinal fluid; Adhesion molecules; sICAM-1; Blood–CSF barrier dysfunction
1. Introduction Borreliosis is caused by tick-borne spirochete Borrelia burgdorferi. After a tick bite spirochetes may spread invading different organs including nervous system, heart, and joints. Involvement of the central nervous system ŽCNS. may result in neuroborreliosis ŽNB. manifesting with meningitis, cranial neuritis, radiculoneuritis, and other symptoms ŽPachner and Steere, 1985.. Basic cerebrospinal fluid ŽCSF. analysis in NB with disease-related ‘‘typical’’ immunoglobulin patterns has been described earlier ŽTumani et al., 1995.. Intercellular adhesion molecule-1 ŽICAM-1. is expressed on cells of CNS in normal conditions ŽEddleston and Mucke, 1993.. In inflammation it is up-regulated by )
Corresponding author. Max-Planck-Institute for Experimental Medicine, Hermann-Rein-Str. 3, 37075 Goettingen, Germany. Tel.: q495 5 1 - 3 8 - 9 9 - 6 2 5 ; f a x : q 4 9 - 5 5 1 - 3 8 - 9 9 - 6 7 0 ; e - m a i l: [email protected]
different factors Žfor review see Lee and Benveniste, 1999.. A soluble form of ICAM-1 ŽsICAM-1. has been reported in many body fluids ŽRothlein et al., 1991.. Recently we have found that 30% of sICAM-1 in normal CSF is brain-derived, and that this brain-derived fraction is increased during CNS inflammation ŽLewczuk et al., 1998.. Although the pshysiological function of soluble forms of adhesion molecules is not clear, their role in inhibiting immune response by competing cell–cell interaction is suggested ŽMartin et al., 1994; Lee and Benveniste, 1999.. As no data on sICAM-1 in CSF in NB are available so far, we studied sICAM-1 levels in CSF and serum of patients with different clinical picture of NB. These results were further compared to the results of patients with the corresponding influx of the blood proteins into the CSF but without CNS inflammation, and to the results of patients with other bacterial meningitis and viral encephalomeningitis. In three repunctured cases we provide also data on changes of sICAM-1 levels in the course of NB.
0165-5728r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 5 7 2 8 Ž 9 9 . 0 0 2 2 1 - 0
94
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
2. Materials and methods 2.1. Patients The group of NB Ž n s 11. consisted of 8 males, and 3 females aged 8–78 years. In this group three male children were included, aged 8–10 years. A tick bite was recognized by 6 patients Ž55%., and erythema chronica migrans was reported by 4 subjects Ž36%. to appear 1–2 weeks after the tick bite. Patients who recognized tick bite reported period from this event to onset of neurological symptoms andror admission to the hospital of 10 days to 8 weeks Žmedian: 6 weeks.. Lumbar puncture ŽLP. was performed on the day of admission, and treatment with antibiotics was started immediately. All patients fullfilled at least one of the clinical case definitions proposed by Oschmann et al. Ž1998., and all cases were classified as stage II. Symptoms in this group included: meningoradiculitis cranialis Ž3 cases., meningoradiculitis cranialis et spinalis Ž6 cases., meningoradiculitis spinalis Ž1 case., and meningoencephaloradiculitis Ž1 case.. The control group Ž n s 11. consisted of patients with disc prolaps ŽDP., including 9 males and 2 females. The age of this group ranged between 32 and 77 years. The diagnosis was based on clinical examination, and it was supported by neuroradiological and neurochemical data. Subjects included in this group were selected to match QAlb on routine CSFrserum analysis. The group of normal control patients ŽNCP, n s 33. consisted of subjects without any organic brain disorders ŽLewczuk et al., 1998..
sensitivity of the ELISA was determined by serial dilution of the standards in the sample diluent Žboth included in the kit.. The lowest concentration distinguishable from the blank was 0.35 ngrml. Thus, the assay was found sensitive enough to determine sICAM-1 concentration in all samples. Additionally, the results obtained with two assays ŽR & D Systems Europe, UK, and Bender MedSystems, Austria. were compared and gave essentially equal results. A positive control serum is included into the assay, and it was measured in duplicate in each run. The day-by-day coefficient of variation Žinter-assay imprecision. was 5.5%. Undiluted CSF samples were used for the assay and serum was diluted 1:20. All steps of the procedure followed the manufacturer’s instruction. Absorbance was measured with an automatic ELISA reader ŽSLT Labinstruments, Germany. with an evaluation program ŽEasy-Fit, from the same manufacturer..
3. Results 3.1. Routine CSF r serum analysis All cases of NB showed typical findings on the routine CSFrserum analysis ŽTumani et al., 1995.. Cases of DP showed increased CSF total protein, and increased QAlb Žrange: 11.2–35.5 = 10y3 . accompanied by normal cell counts. No signs of intrathecal immunologic response were found in these patients.
2.2. Follow-up study Changes of sICAM-1 concentration in CSF and serum during the course of the disease was studied in three cases of NB: BV — 39-year-old man with meningoradiculitis cranialis, GHJ — 42-year-old man with meningoradiculitis cranialis et spinalis, and SMH — 49-year-old woman with meningoencephaloradiculitis. CSF samples were obtained twice from two patients ŽBV on days 1st and 11th, and GHJ on days 1st and 30th., and from one patient ŽSMH. four times on days 1st, 6th, 19th and 37th. 2.3. Routine CSF r serum analysis Routine CSFrserum analysis was performed in all cases according to the protocol described earlier ŽReiber, 1995.. CSF samples contaminated with the blood were not included into the study. For measurement of sICAM-1, aliquots of CSF and serum were frozen and kept in y208C for further analysis in groups of 10–12 samples. 2.4. Analysis of sICAM-1 in CSF and serum sICAM-1 in CSF and serum was analysed by a sandwich ELISA method ŽR & D Systems Europe, UK.. The
Fig. 1. Correlation between albumin and sICAM-1 CSFrserum concentration quotients Ž QAlb and Q ICAM , respectively. in subjects with neuroborreliosis ŽNB., disc prolaps ŽDP., and in normal control patients ŽNCP — reported previously; Lewczuk et al., 1998.. All Q ICAM of NB samples are above the upper discriminatory curve. In cases of NB that were punctured more than once results obtained on the first LP are taken into account.
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
95
3.2. sICAM-1 in CSF and serum; sICAM-1 quotient (Q IC A M ); correlation between Q A l b and Q IC A M sICAM-1 concentration in serum did not differ between NB and DP Žmedian values — 250 ngrml and 251 ngrml, respectively.. sICAM-1 concentration in CSF in NB was up to six-fold higher than in DP Žranges: 6.6–42.8 ngrml, and 2.2–9.8 ngrml for NB and DP, respectively.. This resulted in corresponding increase of sICAM-1 CSFrserum concentration quotient Ž Q ICAM . Žranges: 22.5–171.3 = 10y3 , and 8.8–27.8 = 10y3 for NB and DP, respectively.. Fig. 1 presents correlation between Q ICAM and QAlb of all investigated patients. All cases of NB were above the discriminatory curve separating subjects with inflammatory and non-inflammatory diseases of CNS, as reported previously ŽLewczuk et al., 1998.. Apart of the fact that one case of meningoencephaloradiculitis was associated with the highest sICAM-1 concentration in CSF, and the highest Q ICAM Ž42.8 ngrml, and 171.3 = 10y3 , respectively on the first LP, followed by increase to 46.3 ngrml and 243.6 = 10y3 on the second LP five days later. no correlation between sICAM-1 concentration in CSF and clinical symptoms of NB was observed. Q ICAM in the three children were not different from these in adult patients. There was no significant correlation between CSF sICAM-1 concentration or Q ICAM and other parameters
Fig. 3. Comparison of Q ICAM and QAlb in cases of neuroborreliosis ŽNB. in this study with data derived from other bacterial meningitis ŽOBM., and viral encephalomeningitis ŽVEM. reported previously ŽLewczuk et al., 1998..
derived from the routine CSFrserum analysis including total CSF cell count, various types of CSF leucocytes, quotients of IgG, IgA, IgM, and B. burgdorferi IgG, and IgM indices Ždata not shown.. 3.3. Follow-up study of sICAM-1 in NB In the repunctured cases we did not observe any significant changes in the serum concentration of sICAM-1 Ždata not shown., however, changes in sICAM-1 in CSF were observed, resulting in changes in Q ICAM ŽFig. 2.. In all cases, Q ICAM seemed to decrease in the course of recovery, however, in one case ŽSMH. it peaked on the second LP 5 days after the first LP and was followed by a steady decrease.
4. Discussion
Fig. 2. Correlation between albumin and sICAM-1 CSFrserum concentration quotients Ž QAlb and Q ICAM , respectively. during the time-course of recovery in three cases of NB. ^-BV, 39-year-old man with clinical picture of meningoradiculitis cranialis, punctured on days 1st, and 11th; I-GHJ, 42-year-old man with clinical picture of meningoradiculitis cranialis et spinalis punctured on days 1st and 30th; O-SMH, 49-year-old woman with symptoms of meningoencephaloradiculitis, punctured on days 1st, 6th, 19th, and 37th. Open figures represent results of the first LP.
Here we report an increased CSF concentration of sICAM-1 and an increased sICAM-1 CSFrserum concentration quotient Ž Q ICAM . in neuroborreliosis when compared to disc prolapse — a disorder with similar influx of the blood proteins into the CSF Žcorresponding QAlb ., but without CNS inflammation. Moreover, three repunctured cases demonstrated that sICAM-1 levels in CSF decreased as a response to the treatment with antibiotics. As we did not find correlation between sICAM-1 level in CSF and CSF cell count, we think that our results can be explained by increased brain-derived fraction of sICAM-1 in NB. On the ground of the data from the literature two CNS cell types could be taken into account as a possible source of the brain-derived sICAM-1 in NB: Ža. activated brain endothelial cells, as reviewed by Sigal Ž1997., and Žb.
96
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
reactive glial cells which show up-regulation of adhesion molecules on the surfaces during CNS inflammation ŽEddleston and Mucke, 1993; Lee and Benveniste, 1999.. The recently published data on involvement of the vascular wall in NB ŽOksi et al., 1996., and demonstration in vitro that spirocheta cells themselves can up-regulate ICAM-1 expression on endothelial cells ŽBoggemeyer et al., 1994; Sellati et al., 1995; Sellati et al., 1996. indicate that the brain activated endothelium can be capable for sICAM-1 release into CSF in NB. CSF concentration of sICAM-1 in NB in this study overlap with the data derived from cases of other bacterial meningitis ŽOBM. and viral encephalomeningitis ŽVEM. reported previously ŽLewczuk et al., 1998. ŽFig. 3.. However, when the blood-derived fraction of sICAM-1 Žwhich is higher for higher QAlb . is subtracted from Q ICAM , it becomes evident that the highest values of intrathecally released sICAM-1 are derived from NB cases. The direct comparison of NB and VEM Ži.e., two inflammatory diseases of CNS with similar QAlb . shows this larger Q ICAM in NB than in VEM quite obvious. We did not observe correlation between Q ICAM and clinical symptoms of NB. However, in all our cases, meningitis due to B. burgdorferi was a part of the clinical picture, so we suggest that increased Q ICAM was due to borrelia-related meningitis independently of involvement of other nervous structures. The question whether Q ICAM is increased in rare cases of NB without meningitis Ži.e., mononeuritis and polyneuritis; Oschmann et al., 1998. remains unanswered. We observed a decrease in Q ICAM after initialization of treatment in 3 cases of NB. Although this result must be interpreted carefully due to small number of patients, it seems that normalizing level of sICAM-1 in CSF reflects the process of recovery. Thus, sICAM-1 concentration in the CSF seems to correlate more with the clinical course of the disease than other parameters, like anti-borrelia indices ŽTumani et al., 1995.. Concluding, sICAM-1, a factor suggested to play a role in regulation of immunological reactions in the brain ŽLee and Benveniste, 1999., was found to be released by the cells of the CNS in NB. This is in agreement with the results of the experiments in vitro Žreviewed by Sigal, 1997., and it represents an area of research that may help to understand pathology of NB. Further studies are necessary to clarify a possible relevance of our findings to the
clinical practice Žfor example in monitoring of patients’ recovery.. Acknowledgements The authors thank Mr. Peter Lange for his excellent technical help. References Boggemeyer, E., Stehle, T., Schaible, U.E., Hahne, M., Vestweber, D., Simon, M.M., 1994. Borrelia burgdorferi upregulates the adhesion molecules E-selectin, P- selectin, ICAM-1 and VCAM-1 on mouse endothelioma cells in vitro. Cell Adhes. Commun. 2, 145–157. Eddleston, M., Mucke, L., 1993. Molecular profile of reactive astrocytes — implications for their role in neurologic disease. Neuroscience 54, 15–36. Lee, S.J., Benveniste, E.N., 1999. Adhesion molecule expression and regulation on cells of the central nervous system. J. Neuroimmunol. 98, 77–88. Lewczuk, P., Reiber, H., Tumani, H., 1998. Intercellular adhesion molecule-1 in cerebrospinal fluid — evaluation of blood-derived and brain-derived fractions in neurological diseases. J. Neuroimmunol. 87, 156–161. Martin, S., Lampeter, E.F., Kolb, H., 1994. A physiological role for circulating adhesion molecules?. Immunol. Today 15, 141. Oksi, J., Kalimo, H., Marttila, R.J., Marjamaki, M., Sonninen, P., Nikoskelainen, J., Viljanen, M.K., 1996. Inflammatory brain changes in Lyme borreliosis. A report on three patients and review of literature. Brain 119, 2143–2154. Oschmann, P., Dorndorf, W., Hornig, C., Schafer, C., Wellensiek, H.J., Pflughaupt, K.W., 1998. Stages and syndromes of neuroborreliosis. J. Neurol. 245, 262–272. Pachner, A.R., Steere, A.C., 1985. The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 35, 47–53. Reiber, H., 1995. Clinical relevance of neuroimmunological reaction patterns in cerebrospinal fluid. Lab. Med. 19, 444–462. Rothlein, R., Mainolfi, E.A., Czajkowski, M., Marlin, S.D., 1991. A form of circulating ICAM-1 in human serum. J. Immunol. 147, 3788–3793. Sellati, T.J., Abrescia, L.D., Radolf, J.D., Furie, M.B., 1996. Outer surface lipoproteins of Borrelia burgdorferi activate vascular endothelium in vitro. Infect. Immun. 64, 3180–3187. Sellati, T.J., Burns, M.J., Ficazzola, M.A., Furie, M.B., 1995. Borrelia burgdorferi upregulates expression of adhesion molecules on endothelial cells and promotes transendothelial migration of neutrophils in vitro. Infect. Immun. 63, 4439–4447. Sigal, L.H., 1997. Immunologic mechanisms in Lyme neuroborreliosis: the potential role of autoimmunity and molecular mimicry. Semin. Neurol. 17, 63–68. Tumani, H., Nolker, G., Reiber, H., 1995. Relevance of cerebrospinal fluid variables for early diagnosis of neuroborreliosis. Neurology 45, 1663–1670.
Intrathecal release of sICAM-1 into CSF in neuroborreliosis — increased brain-derived fraction P. Lewczuk b
a,)
, H. Reiber a , G.C. Korenke b, E. Bollensen c , A.J. Dorta-Contreras
d
a Neurochemistry Laboratory, UniÕersity of Goettingen, Robert-Koch-Str. 40, 37075 Goettingen, Germany Department of Pediatric Neurology, UniÕersity of Goettingen, Robert-Koch-Str. 40, 37075 Goettingen, Germany c Department of Neurology, Hospital in Eschwege, Bahnhofstr. 24, 37269 Eschwege, Germany d Immunology Laboratory, Hospital Pediatrico SanMiguel, Apartado 10049 CP 11000 Ciudad HaÕana, Cuba
Received 30 July 1999; received in revised form 15 September 1999; accepted 29 September 1999
Abstract In the present study, we report sICAM-1 concentration in the cerebrospinal fluid ŽCSF. and serum of patients with neuroborreliosis ŽNB, n s 11., compared to the data from a control group of patients with corresponding bloodrCSF barrier dysfunction but without inflammation in the central nervous system Ždisc prolaps, DP, n s 11.. In NB, the sICAM-1 concentration in CSF was increased up to six-fold Žranges: 6.6–42.8 ngrml and 2.2–9.8 ngrml for NB and DP respectively. with no change in serum sICAM-1. The corresponding sICAM-1 CSFrserum concentration quotients Ž Q ICAM . were in the ranges: 22.5–171.3 = 10y3, and 8.8–27.8 = 10y3 for NB and DP respectively. This finding can be explained by increase of the brain-derived fraction of sICAM-1 in NB. In one case we observed increased Q ICAM on 6th day after admission to the hospital Ž171.3 = 10y3 at the time of the first lumbar puncture slightly increasing to 243.6 = 10y3 five days later., followed by normalization, in two remaining repunctured patients we observed decreasing Q ICAM with normalizing QAlb . q 2000 Elsevier Science B.V. All rights reserved. Keywords: Neuroborreliosis; Cerebrospinal fluid; Adhesion molecules; sICAM-1; Blood–CSF barrier dysfunction
1. Introduction Borreliosis is caused by tick-borne spirochete Borrelia burgdorferi. After a tick bite spirochetes may spread invading different organs including nervous system, heart, and joints. Involvement of the central nervous system ŽCNS. may result in neuroborreliosis ŽNB. manifesting with meningitis, cranial neuritis, radiculoneuritis, and other symptoms ŽPachner and Steere, 1985.. Basic cerebrospinal fluid ŽCSF. analysis in NB with disease-related ‘‘typical’’ immunoglobulin patterns has been described earlier ŽTumani et al., 1995.. Intercellular adhesion molecule-1 ŽICAM-1. is expressed on cells of CNS in normal conditions ŽEddleston and Mucke, 1993.. In inflammation it is up-regulated by )
Corresponding author. Max-Planck-Institute for Experimental Medicine, Hermann-Rein-Str. 3, 37075 Goettingen, Germany. Tel.: q495 5 1 - 3 8 - 9 9 - 6 2 5 ; f a x : q 4 9 - 5 5 1 - 3 8 - 9 9 - 6 7 0 ; e - m a i l: [email protected]
different factors Žfor review see Lee and Benveniste, 1999.. A soluble form of ICAM-1 ŽsICAM-1. has been reported in many body fluids ŽRothlein et al., 1991.. Recently we have found that 30% of sICAM-1 in normal CSF is brain-derived, and that this brain-derived fraction is increased during CNS inflammation ŽLewczuk et al., 1998.. Although the pshysiological function of soluble forms of adhesion molecules is not clear, their role in inhibiting immune response by competing cell–cell interaction is suggested ŽMartin et al., 1994; Lee and Benveniste, 1999.. As no data on sICAM-1 in CSF in NB are available so far, we studied sICAM-1 levels in CSF and serum of patients with different clinical picture of NB. These results were further compared to the results of patients with the corresponding influx of the blood proteins into the CSF but without CNS inflammation, and to the results of patients with other bacterial meningitis and viral encephalomeningitis. In three repunctured cases we provide also data on changes of sICAM-1 levels in the course of NB.
0165-5728r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 5 7 2 8 Ž 9 9 . 0 0 2 2 1 - 0
94
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
2. Materials and methods 2.1. Patients The group of NB Ž n s 11. consisted of 8 males, and 3 females aged 8–78 years. In this group three male children were included, aged 8–10 years. A tick bite was recognized by 6 patients Ž55%., and erythema chronica migrans was reported by 4 subjects Ž36%. to appear 1–2 weeks after the tick bite. Patients who recognized tick bite reported period from this event to onset of neurological symptoms andror admission to the hospital of 10 days to 8 weeks Žmedian: 6 weeks.. Lumbar puncture ŽLP. was performed on the day of admission, and treatment with antibiotics was started immediately. All patients fullfilled at least one of the clinical case definitions proposed by Oschmann et al. Ž1998., and all cases were classified as stage II. Symptoms in this group included: meningoradiculitis cranialis Ž3 cases., meningoradiculitis cranialis et spinalis Ž6 cases., meningoradiculitis spinalis Ž1 case., and meningoencephaloradiculitis Ž1 case.. The control group Ž n s 11. consisted of patients with disc prolaps ŽDP., including 9 males and 2 females. The age of this group ranged between 32 and 77 years. The diagnosis was based on clinical examination, and it was supported by neuroradiological and neurochemical data. Subjects included in this group were selected to match QAlb on routine CSFrserum analysis. The group of normal control patients ŽNCP, n s 33. consisted of subjects without any organic brain disorders ŽLewczuk et al., 1998..
sensitivity of the ELISA was determined by serial dilution of the standards in the sample diluent Žboth included in the kit.. The lowest concentration distinguishable from the blank was 0.35 ngrml. Thus, the assay was found sensitive enough to determine sICAM-1 concentration in all samples. Additionally, the results obtained with two assays ŽR & D Systems Europe, UK, and Bender MedSystems, Austria. were compared and gave essentially equal results. A positive control serum is included into the assay, and it was measured in duplicate in each run. The day-by-day coefficient of variation Žinter-assay imprecision. was 5.5%. Undiluted CSF samples were used for the assay and serum was diluted 1:20. All steps of the procedure followed the manufacturer’s instruction. Absorbance was measured with an automatic ELISA reader ŽSLT Labinstruments, Germany. with an evaluation program ŽEasy-Fit, from the same manufacturer..
3. Results 3.1. Routine CSF r serum analysis All cases of NB showed typical findings on the routine CSFrserum analysis ŽTumani et al., 1995.. Cases of DP showed increased CSF total protein, and increased QAlb Žrange: 11.2–35.5 = 10y3 . accompanied by normal cell counts. No signs of intrathecal immunologic response were found in these patients.
2.2. Follow-up study Changes of sICAM-1 concentration in CSF and serum during the course of the disease was studied in three cases of NB: BV — 39-year-old man with meningoradiculitis cranialis, GHJ — 42-year-old man with meningoradiculitis cranialis et spinalis, and SMH — 49-year-old woman with meningoencephaloradiculitis. CSF samples were obtained twice from two patients ŽBV on days 1st and 11th, and GHJ on days 1st and 30th., and from one patient ŽSMH. four times on days 1st, 6th, 19th and 37th. 2.3. Routine CSF r serum analysis Routine CSFrserum analysis was performed in all cases according to the protocol described earlier ŽReiber, 1995.. CSF samples contaminated with the blood were not included into the study. For measurement of sICAM-1, aliquots of CSF and serum were frozen and kept in y208C for further analysis in groups of 10–12 samples. 2.4. Analysis of sICAM-1 in CSF and serum sICAM-1 in CSF and serum was analysed by a sandwich ELISA method ŽR & D Systems Europe, UK.. The
Fig. 1. Correlation between albumin and sICAM-1 CSFrserum concentration quotients Ž QAlb and Q ICAM , respectively. in subjects with neuroborreliosis ŽNB., disc prolaps ŽDP., and in normal control patients ŽNCP — reported previously; Lewczuk et al., 1998.. All Q ICAM of NB samples are above the upper discriminatory curve. In cases of NB that were punctured more than once results obtained on the first LP are taken into account.
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
95
3.2. sICAM-1 in CSF and serum; sICAM-1 quotient (Q IC A M ); correlation between Q A l b and Q IC A M sICAM-1 concentration in serum did not differ between NB and DP Žmedian values — 250 ngrml and 251 ngrml, respectively.. sICAM-1 concentration in CSF in NB was up to six-fold higher than in DP Žranges: 6.6–42.8 ngrml, and 2.2–9.8 ngrml for NB and DP, respectively.. This resulted in corresponding increase of sICAM-1 CSFrserum concentration quotient Ž Q ICAM . Žranges: 22.5–171.3 = 10y3 , and 8.8–27.8 = 10y3 for NB and DP, respectively.. Fig. 1 presents correlation between Q ICAM and QAlb of all investigated patients. All cases of NB were above the discriminatory curve separating subjects with inflammatory and non-inflammatory diseases of CNS, as reported previously ŽLewczuk et al., 1998.. Apart of the fact that one case of meningoencephaloradiculitis was associated with the highest sICAM-1 concentration in CSF, and the highest Q ICAM Ž42.8 ngrml, and 171.3 = 10y3 , respectively on the first LP, followed by increase to 46.3 ngrml and 243.6 = 10y3 on the second LP five days later. no correlation between sICAM-1 concentration in CSF and clinical symptoms of NB was observed. Q ICAM in the three children were not different from these in adult patients. There was no significant correlation between CSF sICAM-1 concentration or Q ICAM and other parameters
Fig. 3. Comparison of Q ICAM and QAlb in cases of neuroborreliosis ŽNB. in this study with data derived from other bacterial meningitis ŽOBM., and viral encephalomeningitis ŽVEM. reported previously ŽLewczuk et al., 1998..
derived from the routine CSFrserum analysis including total CSF cell count, various types of CSF leucocytes, quotients of IgG, IgA, IgM, and B. burgdorferi IgG, and IgM indices Ždata not shown.. 3.3. Follow-up study of sICAM-1 in NB In the repunctured cases we did not observe any significant changes in the serum concentration of sICAM-1 Ždata not shown., however, changes in sICAM-1 in CSF were observed, resulting in changes in Q ICAM ŽFig. 2.. In all cases, Q ICAM seemed to decrease in the course of recovery, however, in one case ŽSMH. it peaked on the second LP 5 days after the first LP and was followed by a steady decrease.
4. Discussion
Fig. 2. Correlation between albumin and sICAM-1 CSFrserum concentration quotients Ž QAlb and Q ICAM , respectively. during the time-course of recovery in three cases of NB. ^-BV, 39-year-old man with clinical picture of meningoradiculitis cranialis, punctured on days 1st, and 11th; I-GHJ, 42-year-old man with clinical picture of meningoradiculitis cranialis et spinalis punctured on days 1st and 30th; O-SMH, 49-year-old woman with symptoms of meningoencephaloradiculitis, punctured on days 1st, 6th, 19th, and 37th. Open figures represent results of the first LP.
Here we report an increased CSF concentration of sICAM-1 and an increased sICAM-1 CSFrserum concentration quotient Ž Q ICAM . in neuroborreliosis when compared to disc prolapse — a disorder with similar influx of the blood proteins into the CSF Žcorresponding QAlb ., but without CNS inflammation. Moreover, three repunctured cases demonstrated that sICAM-1 levels in CSF decreased as a response to the treatment with antibiotics. As we did not find correlation between sICAM-1 level in CSF and CSF cell count, we think that our results can be explained by increased brain-derived fraction of sICAM-1 in NB. On the ground of the data from the literature two CNS cell types could be taken into account as a possible source of the brain-derived sICAM-1 in NB: Ža. activated brain endothelial cells, as reviewed by Sigal Ž1997., and Žb.
96
P. Lewczuk et al.r Journal of Neuroimmunology 103 (2000) 93–96
reactive glial cells which show up-regulation of adhesion molecules on the surfaces during CNS inflammation ŽEddleston and Mucke, 1993; Lee and Benveniste, 1999.. The recently published data on involvement of the vascular wall in NB ŽOksi et al., 1996., and demonstration in vitro that spirocheta cells themselves can up-regulate ICAM-1 expression on endothelial cells ŽBoggemeyer et al., 1994; Sellati et al., 1995; Sellati et al., 1996. indicate that the brain activated endothelium can be capable for sICAM-1 release into CSF in NB. CSF concentration of sICAM-1 in NB in this study overlap with the data derived from cases of other bacterial meningitis ŽOBM. and viral encephalomeningitis ŽVEM. reported previously ŽLewczuk et al., 1998. ŽFig. 3.. However, when the blood-derived fraction of sICAM-1 Žwhich is higher for higher QAlb . is subtracted from Q ICAM , it becomes evident that the highest values of intrathecally released sICAM-1 are derived from NB cases. The direct comparison of NB and VEM Ži.e., two inflammatory diseases of CNS with similar QAlb . shows this larger Q ICAM in NB than in VEM quite obvious. We did not observe correlation between Q ICAM and clinical symptoms of NB. However, in all our cases, meningitis due to B. burgdorferi was a part of the clinical picture, so we suggest that increased Q ICAM was due to borrelia-related meningitis independently of involvement of other nervous structures. The question whether Q ICAM is increased in rare cases of NB without meningitis Ži.e., mononeuritis and polyneuritis; Oschmann et al., 1998. remains unanswered. We observed a decrease in Q ICAM after initialization of treatment in 3 cases of NB. Although this result must be interpreted carefully due to small number of patients, it seems that normalizing level of sICAM-1 in CSF reflects the process of recovery. Thus, sICAM-1 concentration in the CSF seems to correlate more with the clinical course of the disease than other parameters, like anti-borrelia indices ŽTumani et al., 1995.. Concluding, sICAM-1, a factor suggested to play a role in regulation of immunological reactions in the brain ŽLee and Benveniste, 1999., was found to be released by the cells of the CNS in NB. This is in agreement with the results of the experiments in vitro Žreviewed by Sigal, 1997., and it represents an area of research that may help to understand pathology of NB. Further studies are necessary to clarify a possible relevance of our findings to the
clinical practice Žfor example in monitoring of patients’ recovery.. Acknowledgements The authors thank Mr. Peter Lange for his excellent technical help. References Boggemeyer, E., Stehle, T., Schaible, U.E., Hahne, M., Vestweber, D., Simon, M.M., 1994. Borrelia burgdorferi upregulates the adhesion molecules E-selectin, P- selectin, ICAM-1 and VCAM-1 on mouse endothelioma cells in vitro. Cell Adhes. Commun. 2, 145–157. Eddleston, M., Mucke, L., 1993. Molecular profile of reactive astrocytes — implications for their role in neurologic disease. Neuroscience 54, 15–36. Lee, S.J., Benveniste, E.N., 1999. Adhesion molecule expression and regulation on cells of the central nervous system. J. Neuroimmunol. 98, 77–88. Lewczuk, P., Reiber, H., Tumani, H., 1998. Intercellular adhesion molecule-1 in cerebrospinal fluid — evaluation of blood-derived and brain-derived fractions in neurological diseases. J. Neuroimmunol. 87, 156–161. Martin, S., Lampeter, E.F., Kolb, H., 1994. A physiological role for circulating adhesion molecules?. Immunol. Today 15, 141. Oksi, J., Kalimo, H., Marttila, R.J., Marjamaki, M., Sonninen, P., Nikoskelainen, J., Viljanen, M.K., 1996. Inflammatory brain changes in Lyme borreliosis. A report on three patients and review of literature. Brain 119, 2143–2154. Oschmann, P., Dorndorf, W., Hornig, C., Schafer, C., Wellensiek, H.J., Pflughaupt, K.W., 1998. Stages and syndromes of neuroborreliosis. J. Neurol. 245, 262–272. Pachner, A.R., Steere, A.C., 1985. The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 35, 47–53. Reiber, H., 1995. Clinical relevance of neuroimmunological reaction patterns in cerebrospinal fluid. Lab. Med. 19, 444–462. Rothlein, R., Mainolfi, E.A., Czajkowski, M., Marlin, S.D., 1991. A form of circulating ICAM-1 in human serum. J. Immunol. 147, 3788–3793. Sellati, T.J., Abrescia, L.D., Radolf, J.D., Furie, M.B., 1996. Outer surface lipoproteins of Borrelia burgdorferi activate vascular endothelium in vitro. Infect. Immun. 64, 3180–3187. Sellati, T.J., Burns, M.J., Ficazzola, M.A., Furie, M.B., 1995. Borrelia burgdorferi upregulates expression of adhesion molecules on endothelial cells and promotes transendothelial migration of neutrophils in vitro. Infect. Immun. 63, 4439–4447. Sigal, L.H., 1997. Immunologic mechanisms in Lyme neuroborreliosis: the potential role of autoimmunity and molecular mimicry. Semin. Neurol. 17, 63–68. Tumani, H., Nolker, G., Reiber, H., 1995. Relevance of cerebrospinal fluid variables for early diagnosis of neuroborreliosis. Neurology 45, 1663–1670.