Multiple Sklerose ohne oligoklonale Banden im Liquor: Prävalenz und klinischer Verlauf

Abstract

Ziel der vorliegenden Arbeit war zu prüfen, wie häufig Multiple Sklerose (MS) Patienten ohne oligoklonale Banden (OB) im Liquor in einer repräsentativen MS Population sind, und ob diese sich bezüglich des Behinderungsgrades, kernspintomographischer Befunde sowie der klinischen Verlausform von Patienten mit sicherer MS mit oligoklonalen Banden im Liquor unterscheiden. Als Datenbasis diente die MUSIS Datenbank des Jüdischen Krankenhauses Berlin, in der prospektiv standardisiert Daten erfasst werden. Aus dieser Datenbank wurden alle Patienten ausgewählt, bei denen keine OB im Liquor nachweisbar waren und die sicher an MS erkrankt waren. Von Januar 1995 bis zum Oktober 2003 wurden insgesamt 1827 MS Patienten auf das Vorhandensein von OB im Liquor untersucht. Siebzig der untersuchten 1827 MS Patienten (3,6 Prozent), deren Liquor auf das Vorhandensein oligoklonaler Banden untersucht wurde, wiesen keine OB im Liquor auf und waren sicher an MS erkrankt. Um mögliche Unterschiede im Behinderungsgrad (EDSS) und der Zahl Gadoliniumaufnehmender Läsionen in der Kernspintomographie zwischen OB negativen und OB positiven MS Patienten zu ermitteln, wurde zu jedem der 70 OB negativen MS Patienten ein OB positiver MS Patient, der eine größtmögliche Übereinstimmung hinsichtlich Alter, Geschlecht, Krankheitsdauer und Alter zu Krankheitsbeginn aufwies, nach einem Zufallsprinzip gefunden. Es konnte ein hochsignifikanter Unterschied im Behinderungsgrad zwischen OB negativen und OB positiven MS Patienten mit gleichem Manifestationsalter, Alter und gleicher Krankheitsdauer gefunden werden (EDSS 2,7 vs. 3,8; p< 0,00001). Einen benignen Krankheitsverlauf (EDSS ≤ 2 nach ≥ 10 Jahren Krankheitsdauer [Pittock et al. 2004]) wiesen 30,4 Prozent der OB negativen MS Patienten auf. In der Gesamtgruppe untersuchter MS Patienten lag dieser Anteil nur bei 19,7%. Zudem konnte gezeigt werden, dass mehr OB positive MS Patienten einen sekundär chronisch progredienten Verlauf aufwiesen. Der höhere EDSS der OB positiven MS Patienten könnte durch die höhere Wahrscheinlichkeit OB positiver MS Patienten erklärt werden, einen sekundär chronisch progredienten Krankheitsverlauf zu entwickeln. Zudem wurden alle verfügbaren Kernspintomographie-Befund Daten der 70 OB negativen und 70 OB positiven MS Patienten, insbesondere hinsichtlich einer Gadoliniumaufnahme, ausgewertet. Es wurde die mittlere Zahl der Gadoliniumaufnehmenden Läsionen pro MRT und Patient berechnet und verglichen. OB negative MS Patienten wiesen im Mittel signifikant weniger Gadoliniumaufnehmenden Läsionen im MRT, als OB positive MS Patienten auf (2,06 vs. 2,88; p= 0,03). Dieser Unterschied könnte eine weitere Ursache für den geringeren Behinderungsgrad der OB negativen MS Patienten sein. Die Zahl der Gadoliniumaufnehmenden Läsionen im MRT korrelierte weder in der Gruppe der OB negativen, noch in der Gruppe der OB positiven MS Patienten mit dem Behinderungsgrad oder der Krankheitsdauer. Es wurden Unterschiede im Alter zu Krankheitsbeginn, Krankheitsdauer und Grad der Behinderung zwischen MS Patienten ohne OB im Liquor und Patienten der Gesamtpopulation untersuchter MS Patienten gefunden. OB negative MS Patienten erkrankten durchschnittlich später als Patienten der untersuchten Gesamtpopulation, hatten eine kürzere Krankheitsdauer und zeigten einen niedrigeren Behinderungsgrad. Außerdem wurden mögliche Korrelationen zwischen einzelnen Liquorparametern bei einer Gruppe von 205 OB positiven MS Patienten, deren Liquor im Liquorlabor der Neurologischen Klinik der Medizinischen Hochschule Hannover analysiert wurde, untersucht. Die Zahl der oligoklonalen Banden korrelierte signifikant mit dem IgG-Index, der Liquorzellzahl, der absoluten intrathekalen IgG-Synthese nach Reiber/Felgenhauer, der prozentualen intrathekalen IgG-Synthese nach Reiber und dem IgG-Quotienten. Dies zeigt, dass die OB Zahl gut mit der intrathekalen Entzündungsaktivität korreliert und von der intrathekal sezerniertem IgG Menge abhängig ist. Eine signifikant negative Korrelation fand sich zwischen der Zahl oligoklonaler Banden im Liquor und dem Alter zum Zeitpunkt der Lumbalpunktion. Je älter die Patienten zum Zeitpunkt der Lumbalpunktion waren, desto weniger OB ließen sich im Liquor nachweisen. Es konnten keine signifikanten Unterschiede zwischen den weiblichen und männlichen MS Patienten bezüglich der Zahl der OB im Liquor, anderen Liquorparametern, dem EDSS und dem Alter zum Zeitpunkt der Lumbalpunktion gefunden werden. Einzig der Albuminquotient war bei den Patientinnen signifikant niedriger als bei den untersuchten Patienten. Die Arbeit bestätigt damit anhand eines großen Patientenkollektives Ergebnisse vorheriger kleinerer Studien [Stendahl-Brodin und Link, 1980; Zeman et al., 1996], wonach Patienten mit einer sicheren MS ohne nachweisbare oligoklonale Banden im Liquor einen relativ benignen Krankheitsverlauf aufweisen. Oligoklonale Banden im Liquor sind ein prognostischer Marker für das Fortschreiten der Erkrankung und das Ausmaß der Behinderung bei Patienten mit einer sicheren MS. Unterschiedliche neuropathologische Muster der MS wurden kürzlich beschrieben [Lucchinetti et al., 2000] und es bleibt abzuwarten, ob MS Patienten ohne oligoklonale Banden im Liquor einer bestimmten Subgruppe zugeordnet werden können.

At a time when almost all MS patients receive some kind of immunomodulatory therapy, possible misdiagnoses should be avoided at all means. Therefore the German treatment consensus group insists that the diagnosis of MS should include both MRI and testing for oligoclonal bands (OB) in the cerebrospinal fluid. With modern isoelectric focusing (IEF) techniques OB can be detected in 95 - 100 % of patients with clinically definite MS. The literature on MS cases with negative OB is scarce and inconclusive. Most investigations have been carried out with inadequate methods and/ or insufficient sensitivity. Our objective was to define the frequency of OB negative MS patients in larger case material and to further analyse the differences concerning age at onset, gender, disease course, progression and disability after follow-up between OB negative and OB positive MS patients. Methods: The MS data bank MUSIS of the Jewish hospital in Berlin was reviewed for MS patients, whose CSF and serum had been analysed for oligoclonal IgG bands by IEF between 1995 and 2003. 1827 patients with a diagnosis of MS were identified. CSF analysis of the 70 OB negative MS patients had been performed at N= 24 different laboratories. Gender, age at onset, duration of disease, course of disease and severity (expanded disability status scale, EDSS) were extracted from the data base. Patient follow-up was assessed up to October 2003 and ranged from 1 year to 51 years (mean 14.8, ± 8.4 SD). Further we compared two matched groups of patients - 70 OB negative MS patients versus 70 OB positive MS patients. Matching was performed in the following manner. For each OB negative case we identified a OB positive MS patient of the same gender, age at onset, duration of disease, and of the same diagnostic category. Matching of the RRMS (relapsing and remitting MS) patients to SPMS (secondary progressive MS) was permitted. The patient closest to the index case was selected. Results: In 151/1827 (8.3 %) patients with a tentative diagnosis of MS no OBs were found. Among these 62 had possible MS, six turned out to have an alternative diagnosis (cerebral vasculitis, adrenoleukodystrophy (ALD), Polyneuritis and Myelitis, hydrocephalus internus due to aqueduct stenosis, Morbus Wilson and autoimmun triggered polyneuropathy) and 13 patients had changed to OB positive on a repeat lumbar puncture, leaving 70 patients with definite MS and no OBs (3.8%). Differences between the 70 OB negative and 70 positive MS patients were found regarding EDSS (2.7 versus 3.8, p< 0.00001) and progression index (0.30 versus 0.43, p< 0.01). Discussion: We find that clinical definite MS without OB is rare (3.8%), and should be diagnosed with caution (only 46.4% of all OB negative MS patients were definite MS patients). Despite the problems of a retrospective data collection concerning selection bias and incomplete documentation in medical records, the matched group of positive MS patients represents a good selection of all OB positive MS patients. MS patients in the OB negative group were significantly less disabled than the matched OB positive MS patients (EDSS 2.7 vs. 3.8 p<0.00001). This retrospective data analysis along with previous studies confirms that the absence of OB in a case of definite MS, indicates a relatively benign prognosis. The difference in EDSS between OB negative and positive MS patients could be explained with a higher probability to reach SPMS. In other words, the hypothesis that the presence of OBs is associated with worse disability predicts that there would be a higher proportion of patients with secondary progressive disease in the OB positive group than among the OB negative group (Zeman et al., 1996). In fact we could show that 28.6% of the matched OB positive MS patients developed SPMS, compared to only 11.4% in the OB negative group. The higher EDSS in the OB positive group, could be explained with a higher probability to reach SPMS among OB positive MS patients. The demonstration of OBs depends heavily on the employed method and the experience of the performing laboratory. Only techniques with > 95 % positive OB results for definite MS patients are acceptable. According to this criterion all the data (except Bergamanschi et al., 2004 and Zeman et al., 1996) in table 1 have to be considered as inadequate. The considerable difficulties associated with the alleged absence of OB are vividly illustrated in the report of Zeman et al. Initially 34/290 (11.7 %) were found to be OB negative, but subsequent reinspection of the foils led to a revision from negative to positive in 14 patients. For eight other patients a new diagnosis was established and 3/6 patients with a repeat lumbar puncture showed OBs on the second occasion. Altogether only 9/290 (3.1 %) remained with trustful negative OB. OBs may sometimes be absent in a small percentage of patients at initial presentation such as optic neuritis (Söderström et al., 1998) or in children (Sindern et al., 1992) and become positive in the course of the disease. The conversion from negative to positive OB was also observed by us for 13 definite MS patients who underwent a second lumbar puncture. However, in most of the cases the change in the OB rating may be explained by a better performance of the second laboratory (11/11 reexamined at the neurochemical laboratory of the medical school Hannover were found to be positive), rather than by the development of an intrathecal IgG response. The minimum number of OB chosen as cut-off may also influence the percentage of MS patients with negative OB. Most authors demand 2-3 OB for a positive rating. The disappointingly low frequency of only 76 % OB positive MS patients ( Pirttilla et al., 1995) may be explained by the unusually high cut-off of 5 OB. Conversely patients with only 1 OB are negative by definition, but show an increased propensity to develop a complete oligoclonal pattern on a repeat lumbar puncture (Davies et al., 2003). Interestingly 4 of the13 MS patients (30.8 %) without OB who were investigated at the Medical School Hannover displayed one OB. Information on the remaining 57 OB negative MS patients was not available from the other laboratories. Despite the drawback that the results for OB had to be assembled from N= 24 different laboratories, the proportion of definite MS patients with negative OB was remarkably low with only 70/1827 (3.8 %). Thus OB negative CDMS is rare and should be diagnosed with caution. In six patients the absence of OB prompted reconsideration of the initial diagnosis of MS which was substituted by cerebral vasculitis, adrenoleukodystrophy (ALD), polyneuritis and myelitis, hydrocephalus internus due to aqueduct stenosis, Morbus Wilson and autoimmun triggered polyneuropathy. If OBs are lacking it is recommended to repeat the laboratory investigations because (1) a false negative result may probably have been reported by a laboratory with suboptimal performance, or (2) an alternative diagnosis should be considered, or (3) OB may develop with time in patients with a short course of suspected MS. Our study supports the conclusion of previous studies [Stendahl-Brodin et al., 1980; Zeman et al., 1996] that the prognosis of true OB negative MS patients is relatively benign. Different neuropathological patterns have recently be defined in MS (Lucchinetti et al., 2000) and it remains to be seen, if MS patients without OBs can be allocated to a certain subgroup. Subanalysis: CSF analysis in 205 patients with multiple sclerosis. Relevance of the number of oligoclonal bands, interrelation of CSF variables and correlation with clinical data Background: Currently investigations of the CSF of patients with MS are gaining renewed interest. While modern methods of CSF analysis are indispensable for making an early and secure diagnosis, they may also turn out to be of value in predicting clinical disease course and response towards different therapies. However, it is rarely recognized that common CSF parameters may show a considerable degree of interrelation. The recently suggested parameter "number of oligoclonal bands" was tested for such dependencies and the correlation of common CSF variables with clinical data was investigated. Methods: In a retrospective analysis the clinical data for 205 patients with definite MS and positive oligoclonal bands (OB) were extracted from the MUSIS data bank at the Jewish Hospital Berlin. Identification of the oligoclonal band pattern was carried out by isoelectric focusing on polyacrylamide gels with silverstaining in one laboratory. Counting of bands was always performed by the same person (U.W.). Results: Mean age of the 144 women and 61 men at the time of lumbar puncture was 41.2 (SD: 11.6) years. Mean cell count amounted to 4.9/µl with 65/174 (37.4 %) showing more than 4 cells/µl. EDSS score at follow-up correlated negatively with the number of cells (r = - 0.292), but not with the number of OB. The CSF/serum quotient for albumin (blood-CSF-barrier) showed a significant correlation with EDSS (r = 0.234, p = 000.1) and was higer in men than in women (p = 0.017). The mean number of CSF restricted OB was 15.2 (SD: 7.8), with a minimum of 2 (set cut-off) and a maximum of 34 OB. Only 7/205 (3.4%) had less than 4 OB. Additional identical OB in serum (oligoclonal pattern type 3) were encountered in 69/205 (33.7 %), but only 25/205 (12.2 %) displayed at least 4 or more identical OB. The mean IgG index was 0.943 and in 119/205 (58%) it was higher than the usual limit of 0.700. The number of OB correlated closely with the IgG index (r = 0.671, p< 0.0001) and more weakly with the cell count (r = 0.292). It showed a slight decrease with age (r = - 0.192). Conclusion: Comparisons between different methods must consider that the number of OB depends (i) on the resolution of the gels and (ii) on the amount of intrathecally synthesized IgG. CSF analysis has the potential to assist in prognosis and therapy, but a multifactorial approach must be used, including parameters for demyelination, neuronal and axonal damage.