Frequent early relapses and long term disease evolution of multiple sclerosis Antonio Scalfari 1 , Anneke Neuhaus 2 , Martin Daumer 2 , Paolo. A. Muraro 1 , Paolo Giannetti 1 and George Ebers 3 1 Centre for Neuroscience, Division of Experimental Medicine, Imperial College, London (UK), 2 Sylvia Lawry Centre, Munich (GER), 3 Oxford University, Department of Neurology, Oxford (UK) METHODS: CONCLUSIONS: RESULTS: OBJECTIVE: ACKNOWLEDGMENTS: This study was supported by FISM (Italian MS Foundation), Porticus Foundation and Hertie Foundation BACKGROUND: Although total relapses in multiple sclerosis (MS) does not adversely influence the long term outcome, large number of attacks during the first two years (early relapses) associates with worse prognosis (Scalfari et al. 2010). The predictive effect of early relapses is driven by those patients with high ;≥ 3 attacks) frequency and it is exerted by increasing the probability of entering the secondary progressive (SP) phase and by shortening the latency to progression . 1) To assess the independent predictive effect of latency to progression. 2) To iŶǀestigate loŶg teƌŵ eǀolutioŶ of patieŶt ǁith high ;≥ ϯͿ Ŷuŵďeƌ of attaĐks during the first two years (early relapses). Among 730 relapsing-remitting (RR) onset patients from the London Ontario (LO) database, Logistic regression and Kaplan Meier analyses assessed the risk of attaining hard disability outcome (DSS 6-8) according to duration of RR phase. Kaplan Meier analysis was used to assess the long term evolution of frequent relapsers ;≥ 3 attacks in the first two years). Early relapses (Y1+Y2) 1-2 attacks ≥ ϯ attaĐks p values Num of patients (% within tot RR pop.) 572 (78.3%) 158 (21.7%) Num of males 180 (31.5%) 48 (30.4%) Num of females 392 (68.5%) 110 (69.6%) Sex ratio (F/M) 2.1 2.3 0.79 RR MS 166 (29.0%) 55 (34.8%) 0.16 SP MS 406 (71.0%) 103 (65.2%) Mean disease duration: yrs (± SD) 26.2 (10.3) 19.3 (7.1) < 0.001 Mean age at onset: yrs (± SD) 28.8 (9.2) 27.4 (8.2) 0.105 Mean age at onset of SP: yrs (±SD) 41.5 (9.9) 35.3 (9.5) < 0.001 Mean duration of RR phase: yrs (±SD) 11.7 (7.5) 6.9 (5.5) < 0.001 Mean 1 st inter-attack interval: yrs (±SD) 4.8 (5.1) 1.1 (1.1) < 0.001 Kaplan Meier estimated mean time (median) from disease onset to DSS levels DSS 3 13.5 (11) years 8.3 (4) years < 0.001 DSS 6 21.6 (19) years 14.6 (10) years < 0.001 DSS 8 32.4 (31) years 20.4 (21) years < 0.001 DSS 10 49.8 (63) years 29.2 (32) years < 0.001 Table 1: Clinical and demographic features of relapsing onset patients grouped by number of early relapses. The onset of progression is the key determinant of long term prognosis and its latency strongly predicts late disability accumulation, accounting for the variability of the disease course. The coexistence of two extremes of outcome within a population sharing adverse clinical features suggests a possible genetic control of axons vulnerability, promoting a more aggressive course by influencing the probability of becoming progressive. Mean years to DSS 8 from disease onset Duration of RR phase Relapses Y1-2= 1 Relapses Y1-2= 2 Relapses Y1-Ϯ ≥ ϯ 1-5 yrs 19.7 (p<0.01) 18.0 (p<0.01) 11.7 (p<0.01) 6-12 yrs 23.6 (p<0.01) 21.2 (p<0.01) 18.0 (p<0.01) ≥ ϭϯ yƌs 34.4 29.3 27.3 Table 2: Kaplan Meier analysis. Estimated mean times to DSS 8 from disease onset in patients grouped by number of early relapses and duration of RR phase . Among RR onset population (n=730), majority (78%) presented with 1-2 attack in the first two years; 1/5 (21.7%) had high ;≥ 3 attacks) frequency (frequent early relapsers). In both groups, females predominated and age at disease onset was similar. Frequent early relapsers were younger at onset of SP phase, secondary to shorter latency to progression (Table 1). Time to DSS 8 33.9 years 22.2 years 17.5 years p < 0.01 16.7 years 13.7 years 14.4 years From disease onset From onset of progression P = 0.06 Figure 3: Kaplan Meier analysis. Estimated mean times to DSS 8 from disease onset and from onset of SP in patients grouped by duration of RR phase (blue = 1-5 yrs; green = 6-12 yrs; grey = ≥ 13 yrs). FƌeƋueŶt eaƌly ƌelapseƌs ;≥ ϯ attaĐks iŶ Yϭ-Y2) RR MS SP MS p Num of patients 55 (34.8%) 103 (65.2%) Num of females/males 45/10 65/38 Sex ratio (F/M) 4.5 1.7 0.015 Mean disease duration 17.2 yrs 20.3 yrs 0.003 Mean age at onset: 25.2 yrs 28.4 yrs 0.014 Type of initial presentation Mono/poly-symptomatic 62%/38% 70%/30% 0.30 Sensory 63.6% 54.4 0.26 Motor 10.9% 14.6% 0.52 Median time to SP // 5.0 yrs Kaplan Meier estimated mean time from disease onset to DSS 3 16.2 yrs 4.5 yrs < 0.01 DSS 6 27.3 yrs 9.1 yrs < 0.01 DSS 8 // 17.2 yrs < 0.01 DSS 10 // 28.2 yrs < 0.01 By the end of the observation period, within the group with low relapse frequency, 84% reached DSS 3, 70% reached DSS 6, 46% reached DSS 8 and 6% reached DSS 10 in 13.5, 21.6, 32.4 and 49.8 estimated mean years respectively. RRMS; n = 55 SPMS; n = 103 50 percentile Years from disease onset Figure 1: Logistic regression analysis. Risk (OR) of converting to SP MS according to disease duration. Figure 2 : Logistic regression analysis. Risk (OR) of attaining DSS 8 according to duration of RR phase. The risk of conversion to SP MS increases proportionally with disease duration (OR=1.07) (Figure 1). The latency to SP exerts a strong predictive effect. Being free from progression for 10, 15 or 20 years reduced the probability of becoming bed bound (DSS 8) by 42% (OR=0.58), 56% (OR=0.44) and 67% (OR=0.33) respectively (Figure 2). 1.07 1.41 1.99 2.81 3.97 5.61 Duration of disease OR Risk of becoming SP 0.97 0.76 0.58 0.44 0.33 0.25 Duration of RR phase OR Risk of attaining DSS 8 Groups with longer duration of the RR phase attained disability endpoints from disease onset in significantly longer times. However, the slope of the SP phase was not affected by the latency to progression. Times to DSS 8 from onset of SP were similar between groups (Figure 3). Interestingly, time to onset of SP exerted the same predictive effect even when patients were grouped by number of early relapses, (1-2-≥ 3)(Table 2). Taďle ϯ : ĐliŶiĐal aŶd deŵogƌaphiĐ featuƌes of RR aŶd “P patieŶts, aŵoŶg gƌoup ǁith ≥ 3 attacks in the first two years (frequent early relapsers). Figure 4: Kaplan Meier analysis. Percent of patients converting to SP M“ aŵoŶg those ǁith ≥ ϯ attaĐks iŶ the fiƌst tǁo yeaƌs ;fƌeƋueŶt early relapsers. The expected association high early relapse frequency/poor outcome was observed only in 2/3 (n=103) of frequent relapsers who rapidly (5 median years) converted to SPMS and accumulated severe disability (Table 3). Nevertheless, latency to SP exerted a predictive effect even within this minority (Table 2). The remaining patients (n=55) had a benign outcome. They escaped progression and less than half (43%) reached DSS 3 in estimated 16.2 mean years (Table 3), despite 70% had disease duration longer than 16 years. The two subgroups had similar clinical presentation. The larger % of females and the slightly younger age at onset of RR patients cannot sufficiently explain the benign disease course. Among the total population of frequent relapsers (n=158), 50% had entered the SP phase by 9 years from disease onset, increasing to 65% (N=103) by the end of the observation period (Figure 4). Rate of conversion to SP MS FƌeƋueŶt ƌelapseƌs ;≥ ϯ attaĐks iŶ Yϭ-Y2) 000 Posters. Copyright protecte protected. F1000 Posters. Copyright protected. F1000 Pos ers. Copyright protected. F1000 Posters. Copyright protected. F1000 Posters. Copyrigh osters. Copyright protected. F1000 Posters. Copyright protected. F1000 Posters. Copyright protected. F ight protected. F1000 Posters. Copyright protected. F1000 Posters. Copyright protected. F1000 Posters . F1000 Posters. Copyright protected. F1000 Posters. Copyright protected. F1000 Posters. Copyright p ers. Copyright protected. F1000 Posters. Copyright protected. F1000 Poster protected. F1000 Posters. Copyright protected. 1000 Posters. Copyrig