Multiplehttp://msj.sagepub.com/ Sclerosis Journal
Influence of treatments in multiple sclerosis disability: A cohort study Eleonora Cocco, Claudia Sardu, Gabriella Spinicci, Luigina Musu, Rita Massa, Jessica Frau, Lorena Lorefice, Giuseppe Fenu, Giancarlo Coghe, Serenella Massole, Maria Antonietta Maioli, Rachele Piras, Marta Melis, Gianluca Porcu, Elena Mamusa, Nicola Carboni, Paolo Contu and Maria Giovanna Marrosu Mult Scler published online 25 September 2014 DOI: 10.1177/1352458514546788 The online version of this article can be found at: http://msj.sagepub.com/content/early/2014/09/11/1352458514546788
Published by: http://www.sagepublications.com
On behalf of: European Committee for Treatment and Research in Multiple Sclerosis
Americas Committee for Treatment and Research in Multiple Sclerosis
Pan-Asian Committee for Treatment and Research in Multiple Sclerosis
Latin American Committee on Treatment and Research of Multiple Sclerosis
Additional services and information for Multiple Sclerosis Journal can be found at: Email Alerts: http://msj.sagepub.com/cgi/alerts Subscriptions: http://msj.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Sep 25, 2014 What is This?
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
546788
research-article2014
MSJ0010.1177/1352458514546788Multiple Sclerosis JournalE Cocco, C Sardu
MULTIPLE SCLEROSIS MSJ JOURNAL
Research Paper
Influence of treatments in multiple sclerosis disability: A cohort study Eleonora Cocco, Claudia Sardu, Gabriella Spinicci, Luigina Musu, Rita Massa, Jessica Frau, Lorena Lorefice, Giuseppe Fenu, Giancarlo Coghe, Serenella Massole, Maria Antonietta Maioli, Rachele Piras, Marta Melis, Gianluca Porcu, Elena Mamusa, Nicola Carboni, Paolo Contu and Maria Giovanna Marrosu
Multiple Sclerosis Journal 1–9 DOI: 10.1177/ 1352458514546788 © The Author(s), 2014. Reprints and permissions: http://www.sagepub.co.uk/ journalsPermissions.nav
Abstract Background and objective: A critical aspect of multiple sclerosis (MS) treatments is understanding the effect of disease-modifying drugs (DMDs) on the long-term risk of disability and whether the effect is related to disability at start of treatment. Methods: We performed an observational study on 3060 MS patients. The effect of therapy on progression to Expanded Disability Status Scale (EDSS) 3.0 and 6.0 from onset was analysed in treated vs untreated (UTP) patients using Cox regression analysis adjusted for propensity score and immortal time bias. Results: Compared to UTP, the risks of EDSS 3.0 were 94% and 73% lower in immunomodulant (IMTP-) and immunosuppressant (ISTP-) treated patients, respectively, while the risk of EDSS 6.0 was 86% lower in IMTP. The risk of EDSS 6.0 was, respectively, 91% and 75% lower in 1275 IMTP before and 114 after EDSS 3.0 than in 539 UTP; the risk was higher in IMTP starting therapy after EDSS 3.0 than before (HR = 4.42). Conclusions: DMDs delayed long-term disability in MS patients treated either in the early or, to a lesser extent, in the later phase of the disease. Thus, the window of therapeutic opportunity is relatively extended, assuming that early is better than late treatment, but late is better than never.
Keywords: Multiple sclerosis, disability, treatment, immunomodulant, immunosuppressant Date received: 1 May 2014; revised: 27 June 2014; accepted: 7 July 2014
Introduction Although the heterogeneity of the multiple sclerosis (MS) course is widely recognised, in the great majority the typical relapsing form of the disease evolves over time as a progressive deterioration of neurological functions. The basic neuropathological mechanisms of the disability and at what stage of the disease it starts are still unknown. While interconnection between focal inflammation and axonal degeneration seems to be established in relapsing MS,1 such a link is ambiguous in the progressive forms.1 The core of the question is whether the inflammation, either focal or diffuse, is the principal trigger of axonal injury and, thus, the primum movens of the subsequent accumulation of disability. Assuming so or not, this perspective has a crucial impact on the design of novel treatments and on the appropriate use of the available drugs. The suggestion that the focal inflammation
determines the disability progression is based on clinical trials on relapsing–remitting MS, which have clearly illustrated the impact of various agents, such as β-Interferon, fingolimod, natalizumab, and alemtuzumab;2 these agents either impact the focal inflammatory component of the disease, which is expressed as a reduction of relapses and MRI activity,2 or produce a meaningful slowing of short-term disability progression.2 More conflicting data result from trials in the secondary progressive form of the disease,3,4 which report that although the focal inflammation is still reduced by treatments, a defined effect on disability progression was not observed. The limit of clinical trials depends on the time of observation, which is often too short to determine whether the antiinflammatory effect slows down the disability progression. A model that combines the various components of the disease has been proposed;5 such a
Correspondence to: Maria Giovanna Marrosu MS Center, ASL 8, Cagliari, Italy; Department of Public Health, Clinical and Molecular Medicine, University of Cagliari/ Binaghi Hospital, Via Is Guadazzonis, 2, 09126 Cagliari, Italy. [email protected] Claudia Sardu Paolo Contu University of Cagliari, Italy Gabriella Spinicci Luigina Musu Rita Massa Serenella Massole Maria Antonietta Maioli Gianluca Porcu Elena Mamusa Nicola Carboni MS Center, ASL 8, Cagliari, Italy Eleonora Cocco Jessica Frau Lorena Lorefice Giuseppe Fenu Giancarlo Coghe Rachele Piras Marta Melis Maria Giovanna Marrosu Multiple Sclerosis Center, ASL 8, Cagliari/ University of Cagliari, Italy
http://msj.sagepub.com 1
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Multiple Sclerosis Journal model would consider MS as a two-stage disease, with the first being strictly dependent on focal inflammation and the second being dependent on neurodegeneration and diffuse inflammation, both driving the progression phase of the disease.5 What is still unclear is the relationship and the consequentiality, if it exists, between the first and the second stage of the disease and thus the possibility to moderate the second stage using appropriate therapies during the first stage. As suggested,5 the best approach for understanding the relationship between the focal activity and the progression of the process is the design of studies that aim to objectively demonstrate that early suppression of inflammation can prevent subsequent long-term disability. Given the current lack of these trials, epidemiological studies based on large and representative cohorts of patients observed over prolonged periods of time can be useful in understanding whether and at which MS stage therapeutic interventions are able to moderate the disease progression. Observational studies6,7 delineate MS as a pathological process in which the progression of disability is independent of the early history of the disease, the duration of the period of attacks and the number and extension of clinical and radiological focal inflammatory relapses. Two recent observational studies8,9 reinforced the finding that in relapsing MS, the initial course of secondary progression is unpredictable based on the early relapses, indicating that commonly used MS therapies are essentially clinically irrelevant in the disability outcome. In practical terms, the existence of a ‘window of therapeutic opportunity’, considered as the time during which therapy eliminates new episodes but also prevents disability progression, is still nebulous. In this article, we have analysed a large cohort of MS patients from the MS Clinic of Cagliari (Italy), with the principal objective of understanding whether commonly used MS treatments may delay disability and how narrow the therapeutic window may be. Patients and methods Patients and data collection Patients were identified through the MS Clinic of Cagliari database containing more than 3000 in- and outpatients from the year 1985. The clinic is the major Sardinian MS hospital and the team of neurologists has been fundamentally the same for many years. Demographic and clinical data (disease course, onset, symptoms at onset, irreversible disability level, therapies, death and its cause) are collected in the database and annually updated.
The database was locked on 31 December 2012. The MS diagnosis was established according to Poser’s classification10 until 2001, and after 2001, the McDonald criteria11 and its subsequent revisions12,13 were used. All patients diagnosed before 2001 underwent neuroimaging studies to confirm the diagnosis according to the Barkof criteria.14 Clinical onset was established as the first symptom unequivocally related to the disease. Disease phenotype at onset was considered according to the classification of Lublin and Reingold.15 However, for the purpose of the study, relapsing phenotype (R) was identified in patients who initially had a relapsing course that may eventually progress to the secondary progressive form of the disease; the primary progressive phenotype (PP) was identified in patients who had continuous deterioration of neurological functions from onset with or without superimposed attacks. Clinical assessment Disability was assessed using the Expanded Disability Status Scale (EDSS).16 As benchmarks of disability, EDSS 3.0 and 6.0 were considered as key steps and score assignment was qualified as irreversible when persistent for at least eight months. Defining treated and untreated patients Treatments were qualified as immunomodulatory (IM) (interferon beta 1a and 1b, glatiramer acetate, natalizumab) or immunosuppressive (IS) (azathioprine, methotrexate, mitoxantrone, cyclophosphamide). Before 1996 (when the first IM treatment was licensed in Italy), patients were not treated at all (excluding steroids or adrenocorticotropic hormone (ACTH)) or were treated with IS as azathioprine, cyclophosphamide, methotrexate. A group of untreated patients or patients who were treated with IS before 1996 afterward started IM if they were eligible for that according to Pharmacological Italian Agency indications. Other patients decided to continue or to start IS, in the majority of cases because of fear of collateral effects of IM. The untreated (UTP) group included either patients who were never treated or those who discontinued therapy after less than one year. Patients were not treated because: they did not satisfy the criteria for disease-modifying drug (DMD) prescription or for personal reasons due to fear of injections or collateral effects or because they were not convinced of benefits. Reasons for early discontinuation of therapy were: scarce compliance, fear of injection, presence of collateral effect and injection site reactions, desire for pregnancy and because not convinced of benefit. UTP belong either to the
2 http://msj.sagepub.com
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
E Cocco, C Sardu et al. Table 1. Summary of the demographics and clinical characteristics of treated and untreated patients.
Men Women Mean age at onset Mean EDSS at last visit Mean disease duration
Untreated
Treated
21.2% 21.4% 31.9 3.0 14.7
78.8% 78.6% 29.7 3.0 13.6
EDSS: Expanded Disability Status Scale.
historical (N = 238) or the contemporary cohort (N = 299). For the contemporary group, we considered patients who were entered in the database after the year 1996. However, these patients were regularly followed at the clinic and treated with corticosteroids when necessary. Short IM breaks (less than three months) or switches between different IM therapies were not considered interruptions. Data regarding the beginning and duration of each treatment were available. Statistical analysis A descriptive analysis was conducted, according to data distribution. The effect of therapy on reaching EDSS 3.0 and EDSS 6 was studied using Cox regression models with therapy as a time-dependent covariate adjusted for propensity score.17 The time to EDSS 3.0 and EDSS 6 from onset was compared in patients treated with IM and/or IS (TP) for at least one year vs UTP. Using therapy as a timedependent covariate, data regarding TP and UTP were adjusted for the ‘immortal time bias’.18 Propensity score, calculated based on gender, calendar year of onset, age at onset and duration of the disease at the first drug, is included in the model as a fixed covariate. Moreover, the Cox regression models with therapy as a time-dependent covariate were also calculated using the single variables (gender, calendar year of onset, and age at onset) as covariates instead of the propensity score. All analyses conducted to assess the effect of therapy were performed both for the global group of TP and separately for patients treated with IS (ISTP) and IM (IMTP). Then, the risk of EDSS 6 from onset, according to whether treatment started before or after EDSS 3.0, was analysed using Cox regression analysis with therapy as a timedependent covariate and propensity score as a fixed covariate. The analysis was not performed considering IS alone because of the low numbers.
Cox regression analysis results are expressed as hazard ratios (HR) and 95% confidence intervals (CI). P values less than 0.05 were considered significant. The analyses were performed using IBM SPSS Statistics version 21. Results Characteristics of the MS population The database contained 3160 patients, of whom 3060 (96.8%) were still alive on 31 December 2012. Based on recent studies, the estimated prevalence of MS in Sardinia was 210.4/100.000;19 thus considering that there are 1,600,000 Sardinia inhabitants, the database captures approximately 93% of the total patients who reside on the island. There were 2079 (68%) females and 981 (37%) males; the clinical phenotype was R in 2647 and PP in 253 of patients; 79 patients were classified as clinically isolated syndrome and in 81 the phenotype was not clearly identified. The median time of the followup was 12 years (fifth percentile, two years; 95th percentile, 34 years), accounting for 42,025 person-years. The mean age at onset was 31.0 years (95% CI 30.6– 31.4), and the median 29.0 years. The mean age of the cohort was 46.1 (95% CI 45.7–46.6) (Table 1). Complete data on therapy were available for 2516 R patients, 1429 (56.8%) of them were treated with IM, 165 (6.6%) with IS, 385 (15.3%) with both IS and IM and 537 (21.3%) were untreated. Among the patients who received IM or IM and IS therapy, 254 (14.0%) were treated with natalizumab after failure of previous therapy. The median duration of treatment was 6.0 years (5th percentile, one year; 95th percentile, 14 years). Figure 1 shows the flowchart describing the selection of patients included in the final analysis. Effect of treatment on disability in relapsing patients The risk of progression to EDSS 3.0 in 1735 TP (IMTP and/or ISTP) was 90% lower than in 781 UTP; it was 94% lower (HR = 0.06, CI 95% 0.04–0.09, p < 0.001) in 1306 IMTP and 73% lower (HR = 0.27, CI 95% 0.16–0.44, p < 0.001) in 98 ISTP than in UTP (Table 2). Cox analysis using single variables confirmed the lower risk of progression to EDSS 3.0 in TP (globally and separately for IMTP and ISTP) and showed the effect of age at onset. Data are reported in Table 3.
http://msj.sagepub.com 3
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Multiple Sclerosis Journal
Figure 1. Flowchart describing the selection of patients included in the final analysis. MS: multiple sclerosis; CIS: clinically isolated syndrome; IM: immunomodulatory; IS: immunosuppressive.
The risk of progression to EDSS 6.0 in 1904 TP (IMTP and/or ISTP) was 62% lower than in 539 UTP (HR = 0.38, CI 95% 0.28–0.52, p < 0.001); it was 86% lower in 1389 IMTP than in 539 UTP (HR = 0.14, CI 95% 0.08–0.25, p < 0.001), while it did not significantly differ in 140 ISTP with respect to UTP (HR = 0.67, CI 95% 0.42–1.07, p = 0.09). Data are reported in Table 2.
95% CI 0.04–0.19, p < 0.001). In ISTP the analysis was not performed because of the low number of patients (Table 4, model A).
Cox analysis using single variables confirmed the lower risk of progression to EDSS 6 and showed the effect of age at onset (Table 3).
The risk of progression to EDSS 6 in 202 patients who started IM or IS therapy after EDSS 3.0 was 58% lower than in 539 UTP (HR = 0.42, CI 95% 0.26–0.66, p < 0.001); risk was 75% lower in 114 patients who started IM after EDSS 3.0 than in UTP (HR = 0.25, 95% CI 0.12–0.51, p < 0.001) (Table 4, model B).
The risk of progression to EDSS 6.0 in 1702 patients who started IM or IS therapy before EDSS 3.0 was 66% lower than in 539 UTP (HR = 0.34, CI 95% 0.24–0.49, p < 0.001) and it was 91% lower in 1275 IMTP before EDSS 3.0 than in 539 UTP (HR = 0.09,
Cox analysis using single variables confirmed the results and the effects of age at onset are reported in Table 3.
Cox analysis using single variables confirmed the results and showed the effect of age at onset (Table 2).
4 http://msj.sagepub.com
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Influence of treatments in multiple sclerosis disability: A cohort study Eleonora Cocco, Claudia Sardu, Gabriella Spinicci, Luigina Musu, Rita Massa, Jessica Frau, Lorena Lorefice, Giuseppe Fenu, Giancarlo Coghe, Serenella Massole, Maria Antonietta Maioli, Rachele Piras, Marta Melis, Gianluca Porcu, Elena Mamusa, Nicola Carboni, Paolo Contu and Maria Giovanna Marrosu Mult Scler published online 25 September 2014 DOI: 10.1177/1352458514546788 The online version of this article can be found at: http://msj.sagepub.com/content/early/2014/09/11/1352458514546788
Published by: http://www.sagepublications.com
On behalf of: European Committee for Treatment and Research in Multiple Sclerosis
Americas Committee for Treatment and Research in Multiple Sclerosis
Pan-Asian Committee for Treatment and Research in Multiple Sclerosis
Latin American Committee on Treatment and Research of Multiple Sclerosis
Additional services and information for Multiple Sclerosis Journal can be found at: Email Alerts: http://msj.sagepub.com/cgi/alerts Subscriptions: http://msj.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Sep 25, 2014 What is This?
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
546788
research-article2014
MSJ0010.1177/1352458514546788Multiple Sclerosis JournalE Cocco, C Sardu
MULTIPLE SCLEROSIS MSJ JOURNAL
Research Paper
Influence of treatments in multiple sclerosis disability: A cohort study Eleonora Cocco, Claudia Sardu, Gabriella Spinicci, Luigina Musu, Rita Massa, Jessica Frau, Lorena Lorefice, Giuseppe Fenu, Giancarlo Coghe, Serenella Massole, Maria Antonietta Maioli, Rachele Piras, Marta Melis, Gianluca Porcu, Elena Mamusa, Nicola Carboni, Paolo Contu and Maria Giovanna Marrosu
Multiple Sclerosis Journal 1–9 DOI: 10.1177/ 1352458514546788 © The Author(s), 2014. Reprints and permissions: http://www.sagepub.co.uk/ journalsPermissions.nav
Abstract Background and objective: A critical aspect of multiple sclerosis (MS) treatments is understanding the effect of disease-modifying drugs (DMDs) on the long-term risk of disability and whether the effect is related to disability at start of treatment. Methods: We performed an observational study on 3060 MS patients. The effect of therapy on progression to Expanded Disability Status Scale (EDSS) 3.0 and 6.0 from onset was analysed in treated vs untreated (UTP) patients using Cox regression analysis adjusted for propensity score and immortal time bias. Results: Compared to UTP, the risks of EDSS 3.0 were 94% and 73% lower in immunomodulant (IMTP-) and immunosuppressant (ISTP-) treated patients, respectively, while the risk of EDSS 6.0 was 86% lower in IMTP. The risk of EDSS 6.0 was, respectively, 91% and 75% lower in 1275 IMTP before and 114 after EDSS 3.0 than in 539 UTP; the risk was higher in IMTP starting therapy after EDSS 3.0 than before (HR = 4.42). Conclusions: DMDs delayed long-term disability in MS patients treated either in the early or, to a lesser extent, in the later phase of the disease. Thus, the window of therapeutic opportunity is relatively extended, assuming that early is better than late treatment, but late is better than never.
Keywords: Multiple sclerosis, disability, treatment, immunomodulant, immunosuppressant Date received: 1 May 2014; revised: 27 June 2014; accepted: 7 July 2014
Introduction Although the heterogeneity of the multiple sclerosis (MS) course is widely recognised, in the great majority the typical relapsing form of the disease evolves over time as a progressive deterioration of neurological functions. The basic neuropathological mechanisms of the disability and at what stage of the disease it starts are still unknown. While interconnection between focal inflammation and axonal degeneration seems to be established in relapsing MS,1 such a link is ambiguous in the progressive forms.1 The core of the question is whether the inflammation, either focal or diffuse, is the principal trigger of axonal injury and, thus, the primum movens of the subsequent accumulation of disability. Assuming so or not, this perspective has a crucial impact on the design of novel treatments and on the appropriate use of the available drugs. The suggestion that the focal inflammation
determines the disability progression is based on clinical trials on relapsing–remitting MS, which have clearly illustrated the impact of various agents, such as β-Interferon, fingolimod, natalizumab, and alemtuzumab;2 these agents either impact the focal inflammatory component of the disease, which is expressed as a reduction of relapses and MRI activity,2 or produce a meaningful slowing of short-term disability progression.2 More conflicting data result from trials in the secondary progressive form of the disease,3,4 which report that although the focal inflammation is still reduced by treatments, a defined effect on disability progression was not observed. The limit of clinical trials depends on the time of observation, which is often too short to determine whether the antiinflammatory effect slows down the disability progression. A model that combines the various components of the disease has been proposed;5 such a
Correspondence to: Maria Giovanna Marrosu MS Center, ASL 8, Cagliari, Italy; Department of Public Health, Clinical and Molecular Medicine, University of Cagliari/ Binaghi Hospital, Via Is Guadazzonis, 2, 09126 Cagliari, Italy. [email protected] Claudia Sardu Paolo Contu University of Cagliari, Italy Gabriella Spinicci Luigina Musu Rita Massa Serenella Massole Maria Antonietta Maioli Gianluca Porcu Elena Mamusa Nicola Carboni MS Center, ASL 8, Cagliari, Italy Eleonora Cocco Jessica Frau Lorena Lorefice Giuseppe Fenu Giancarlo Coghe Rachele Piras Marta Melis Maria Giovanna Marrosu Multiple Sclerosis Center, ASL 8, Cagliari/ University of Cagliari, Italy
http://msj.sagepub.com 1
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Multiple Sclerosis Journal model would consider MS as a two-stage disease, with the first being strictly dependent on focal inflammation and the second being dependent on neurodegeneration and diffuse inflammation, both driving the progression phase of the disease.5 What is still unclear is the relationship and the consequentiality, if it exists, between the first and the second stage of the disease and thus the possibility to moderate the second stage using appropriate therapies during the first stage. As suggested,5 the best approach for understanding the relationship between the focal activity and the progression of the process is the design of studies that aim to objectively demonstrate that early suppression of inflammation can prevent subsequent long-term disability. Given the current lack of these trials, epidemiological studies based on large and representative cohorts of patients observed over prolonged periods of time can be useful in understanding whether and at which MS stage therapeutic interventions are able to moderate the disease progression. Observational studies6,7 delineate MS as a pathological process in which the progression of disability is independent of the early history of the disease, the duration of the period of attacks and the number and extension of clinical and radiological focal inflammatory relapses. Two recent observational studies8,9 reinforced the finding that in relapsing MS, the initial course of secondary progression is unpredictable based on the early relapses, indicating that commonly used MS therapies are essentially clinically irrelevant in the disability outcome. In practical terms, the existence of a ‘window of therapeutic opportunity’, considered as the time during which therapy eliminates new episodes but also prevents disability progression, is still nebulous. In this article, we have analysed a large cohort of MS patients from the MS Clinic of Cagliari (Italy), with the principal objective of understanding whether commonly used MS treatments may delay disability and how narrow the therapeutic window may be. Patients and methods Patients and data collection Patients were identified through the MS Clinic of Cagliari database containing more than 3000 in- and outpatients from the year 1985. The clinic is the major Sardinian MS hospital and the team of neurologists has been fundamentally the same for many years. Demographic and clinical data (disease course, onset, symptoms at onset, irreversible disability level, therapies, death and its cause) are collected in the database and annually updated.
The database was locked on 31 December 2012. The MS diagnosis was established according to Poser’s classification10 until 2001, and after 2001, the McDonald criteria11 and its subsequent revisions12,13 were used. All patients diagnosed before 2001 underwent neuroimaging studies to confirm the diagnosis according to the Barkof criteria.14 Clinical onset was established as the first symptom unequivocally related to the disease. Disease phenotype at onset was considered according to the classification of Lublin and Reingold.15 However, for the purpose of the study, relapsing phenotype (R) was identified in patients who initially had a relapsing course that may eventually progress to the secondary progressive form of the disease; the primary progressive phenotype (PP) was identified in patients who had continuous deterioration of neurological functions from onset with or without superimposed attacks. Clinical assessment Disability was assessed using the Expanded Disability Status Scale (EDSS).16 As benchmarks of disability, EDSS 3.0 and 6.0 were considered as key steps and score assignment was qualified as irreversible when persistent for at least eight months. Defining treated and untreated patients Treatments were qualified as immunomodulatory (IM) (interferon beta 1a and 1b, glatiramer acetate, natalizumab) or immunosuppressive (IS) (azathioprine, methotrexate, mitoxantrone, cyclophosphamide). Before 1996 (when the first IM treatment was licensed in Italy), patients were not treated at all (excluding steroids or adrenocorticotropic hormone (ACTH)) or were treated with IS as azathioprine, cyclophosphamide, methotrexate. A group of untreated patients or patients who were treated with IS before 1996 afterward started IM if they were eligible for that according to Pharmacological Italian Agency indications. Other patients decided to continue or to start IS, in the majority of cases because of fear of collateral effects of IM. The untreated (UTP) group included either patients who were never treated or those who discontinued therapy after less than one year. Patients were not treated because: they did not satisfy the criteria for disease-modifying drug (DMD) prescription or for personal reasons due to fear of injections or collateral effects or because they were not convinced of benefits. Reasons for early discontinuation of therapy were: scarce compliance, fear of injection, presence of collateral effect and injection site reactions, desire for pregnancy and because not convinced of benefit. UTP belong either to the
2 http://msj.sagepub.com
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
E Cocco, C Sardu et al. Table 1. Summary of the demographics and clinical characteristics of treated and untreated patients.
Men Women Mean age at onset Mean EDSS at last visit Mean disease duration
Untreated
Treated
21.2% 21.4% 31.9 3.0 14.7
78.8% 78.6% 29.7 3.0 13.6
EDSS: Expanded Disability Status Scale.
historical (N = 238) or the contemporary cohort (N = 299). For the contemporary group, we considered patients who were entered in the database after the year 1996. However, these patients were regularly followed at the clinic and treated with corticosteroids when necessary. Short IM breaks (less than three months) or switches between different IM therapies were not considered interruptions. Data regarding the beginning and duration of each treatment were available. Statistical analysis A descriptive analysis was conducted, according to data distribution. The effect of therapy on reaching EDSS 3.0 and EDSS 6 was studied using Cox regression models with therapy as a time-dependent covariate adjusted for propensity score.17 The time to EDSS 3.0 and EDSS 6 from onset was compared in patients treated with IM and/or IS (TP) for at least one year vs UTP. Using therapy as a timedependent covariate, data regarding TP and UTP were adjusted for the ‘immortal time bias’.18 Propensity score, calculated based on gender, calendar year of onset, age at onset and duration of the disease at the first drug, is included in the model as a fixed covariate. Moreover, the Cox regression models with therapy as a time-dependent covariate were also calculated using the single variables (gender, calendar year of onset, and age at onset) as covariates instead of the propensity score. All analyses conducted to assess the effect of therapy were performed both for the global group of TP and separately for patients treated with IS (ISTP) and IM (IMTP). Then, the risk of EDSS 6 from onset, according to whether treatment started before or after EDSS 3.0, was analysed using Cox regression analysis with therapy as a timedependent covariate and propensity score as a fixed covariate. The analysis was not performed considering IS alone because of the low numbers.
Cox regression analysis results are expressed as hazard ratios (HR) and 95% confidence intervals (CI). P values less than 0.05 were considered significant. The analyses were performed using IBM SPSS Statistics version 21. Results Characteristics of the MS population The database contained 3160 patients, of whom 3060 (96.8%) were still alive on 31 December 2012. Based on recent studies, the estimated prevalence of MS in Sardinia was 210.4/100.000;19 thus considering that there are 1,600,000 Sardinia inhabitants, the database captures approximately 93% of the total patients who reside on the island. There were 2079 (68%) females and 981 (37%) males; the clinical phenotype was R in 2647 and PP in 253 of patients; 79 patients were classified as clinically isolated syndrome and in 81 the phenotype was not clearly identified. The median time of the followup was 12 years (fifth percentile, two years; 95th percentile, 34 years), accounting for 42,025 person-years. The mean age at onset was 31.0 years (95% CI 30.6– 31.4), and the median 29.0 years. The mean age of the cohort was 46.1 (95% CI 45.7–46.6) (Table 1). Complete data on therapy were available for 2516 R patients, 1429 (56.8%) of them were treated with IM, 165 (6.6%) with IS, 385 (15.3%) with both IS and IM and 537 (21.3%) were untreated. Among the patients who received IM or IM and IS therapy, 254 (14.0%) were treated with natalizumab after failure of previous therapy. The median duration of treatment was 6.0 years (5th percentile, one year; 95th percentile, 14 years). Figure 1 shows the flowchart describing the selection of patients included in the final analysis. Effect of treatment on disability in relapsing patients The risk of progression to EDSS 3.0 in 1735 TP (IMTP and/or ISTP) was 90% lower than in 781 UTP; it was 94% lower (HR = 0.06, CI 95% 0.04–0.09, p < 0.001) in 1306 IMTP and 73% lower (HR = 0.27, CI 95% 0.16–0.44, p < 0.001) in 98 ISTP than in UTP (Table 2). Cox analysis using single variables confirmed the lower risk of progression to EDSS 3.0 in TP (globally and separately for IMTP and ISTP) and showed the effect of age at onset. Data are reported in Table 3.
http://msj.sagepub.com 3
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
Multiple Sclerosis Journal
Figure 1. Flowchart describing the selection of patients included in the final analysis. MS: multiple sclerosis; CIS: clinically isolated syndrome; IM: immunomodulatory; IS: immunosuppressive.
The risk of progression to EDSS 6.0 in 1904 TP (IMTP and/or ISTP) was 62% lower than in 539 UTP (HR = 0.38, CI 95% 0.28–0.52, p < 0.001); it was 86% lower in 1389 IMTP than in 539 UTP (HR = 0.14, CI 95% 0.08–0.25, p < 0.001), while it did not significantly differ in 140 ISTP with respect to UTP (HR = 0.67, CI 95% 0.42–1.07, p = 0.09). Data are reported in Table 2.
95% CI 0.04–0.19, p < 0.001). In ISTP the analysis was not performed because of the low number of patients (Table 4, model A).
Cox analysis using single variables confirmed the lower risk of progression to EDSS 6 and showed the effect of age at onset (Table 3).
The risk of progression to EDSS 6 in 202 patients who started IM or IS therapy after EDSS 3.0 was 58% lower than in 539 UTP (HR = 0.42, CI 95% 0.26–0.66, p < 0.001); risk was 75% lower in 114 patients who started IM after EDSS 3.0 than in UTP (HR = 0.25, 95% CI 0.12–0.51, p < 0.001) (Table 4, model B).
The risk of progression to EDSS 6.0 in 1702 patients who started IM or IS therapy before EDSS 3.0 was 66% lower than in 539 UTP (HR = 0.34, CI 95% 0.24–0.49, p < 0.001) and it was 91% lower in 1275 IMTP before EDSS 3.0 than in 539 UTP (HR = 0.09,
Cox analysis using single variables confirmed the results and the effects of age at onset are reported in Table 3.
Cox analysis using single variables confirmed the results and showed the effect of age at onset (Table 2).
4 http://msj.sagepub.com
Downloaded from msj.sagepub.com at University of North Carolina at Chapel Hill on November 10, 2014
![[Oral treatments in multiple sclerosis].](/assets/img/hold.png)
