The Italian Multiple Sclerosis Register Experience With Cladribine: Impact on Relapses, PIRA, and Treatment Sequencing Strategies Evaluation.

Introduction

Immune reconstitution therapy (IRT), administered in short dose intervals to produce durable immunologic effects, has transformed multiple sclerosis (MS) management into the constantly evolving repertoire of disease-modifying treatments (DMTs), providing an additional treatment contributing to the prevention of the irreversible disability accrual related to MS. 1-3

Cladribine is an oral selective IRT, a synthetic analog of deoxyadenosine that causes a rapid and temporary depletion of lymphocytes followed by a process of immune reconstitution, with a modest impact on the innate immune system. 4,5 Cladribine tablets were approved by the European Medical Agency in August 2017 and are refundable in Italy for the treatment of adults with highly active relapsing multiple sclerosis (RMS), as defined by clinical and MRI features. The unique immune cell dynamics may explain the efficacy and safety profile of cladribine tablets: requiring only 2 treatment courses in 2 years has revolutionized the therapeutic paradigms and the life quality of patients with MS. 6 The CLARITY (CLAdRIbine Tablets treating multiple sclerosis orallY) and CLARITY Extension studies examined the safety and effectiveness profile of cladribine, showing significant improvements in relapse rate, disability progression, and MRI parameters of disease activity, in parallel with an excellent safety profile. 7-9 A growing amount of real-world evidence reinforced the strong anti-inflammatory effect of cladribine, highlighting the sustained efficacy related to the reconstitution of immune cells. 10-13

Formerly believed to be limited to the later stages of the disease, progression independent of relapse activity (PIRA) events have been amply identified even in the earliest stages, outlining a disease continuum of neuroinflammation and neurodegeneration processes. 14,15 Therefore, it is crucial to investigate how DMTs may affect PIRA events and, by extension, the underlying phenomena.

Over the past few years, an increasing number of studies derived from the real-world dataset of the Italian MS and Related Disorders Register (RISM) have demonstrated the ability to address and overcome the most debated topics in MS research arising from routine clinical practice. 16-18

In this study, we aimed to assess the effect of cladribine on the annualized relapse rate (ARR) and PIRA events in the Italian cladribine-treated cohort in the RISM. We also evaluated the strategies of DMT continuation after cladribine termination recorded in the RISM.

Methods

Standard Protocol Approvals, Registrations, and Patient Consents

The RISM was approved by the ethical committee at the “Azienda Ospedaliero—Universitaria—Policlinico of Bari” (Study REGISTRO SM001—approved on July 8, 2016) and by local ethics committees in all participating centers. Patients signed informed consent that allows collection and use of their clinical data for research purposes. According to the register rules, the Scientific Committee of the RISM granted the approval to conduct this project and extract and use the registry data.

Data Extraction

This was a study based on data extracted from the RISM. Data extraction was executed in March 2024. A minimum dataset of demographic and clinical characteristics from RISM was considered.

Study Population

Patients with RMS treated with at least one cycle of cladribine recorded in RISM after January 1, 2018, were retrospectively included in the analysis. Only patients in the RISM with consistent data were included, excluding patients with errors or incongruences in their clinical history.

Patients previously treated with other DMTs were stratified into moderately effective (ME) and highly effective (HE) DMTs. 19 The following treatments were classified as ME DMTs: azathioprine, dimethyl fumarate, glatiramer acetate, interferon-beta products, and teriflunomide. The following products were considered HE DMTs: alemtuzumab, fingolimod, mitoxantrone, natalizumab, ocrelizumab, ofatumumab, ozanimod, ponesimod, and rituximab.

Study Method and Statistical Analysis

The following variables were included in the dataset: age (in years), sex (male and female), type of onset (monofocal/multifocal/not known), IgG oligoclonal bands (OCBs) in the CSF (presence/absence/not known), number of EDSS evaluations with complete information regarding functional scores (FSs), start and end dates of all the administered DMTs, and number and dates of relapses.

In descriptive analyses, categorical data were expressed as frequency and proportion. Continuous data were expressed as mean and SD or as median and interquartile range (IQR).

Relapses in the 2 years before and after cladribine initiation were assessed and compared using negative binomial regression models, and results were given as ARRs along with their 95% confidence intervals (95% CIs). 20 The analyses were performed in the whole population and stratified by age at cladribine start (<50 vs ≥ 50 years) and by previous DMT (ME vs HE DMTs).

To ensure reliability of results, PIRA was analyzed in a subgroup of patients with complete and carefully checked start and end dates of every week of cladribine treatment.

PIRA was considered as a recurrent event through the follow-up. PIRA was defined as a sustained (for the remaining recorded follow-up time) progression of disability, confirmed at 6 months, in the absence of a relapse since 90 days before the EDSS evaluation with documented increase. 21 Cladribine was considered as a time-dependent exposure. Time variable was defined as the time between the baseline visit and the end of follow-up. PIRA was, therefore, analyzed using the Ghosh-Lin Cox-type regression for the marginal mean. Risks were reported as hazard ratios (HRs) along with their 95% CIs and p values. Furthermore, an interaction term (cladribine treatment-by-covariate) was added into the model to explore potential differential efficacy between clinical and/or demographic subgroups.

The proportion of patients without PIRA events (PIRA-free probability) stratified based on the time-dependent exposure to cladribine over the entire follow-up period was shown using Kaplan-Meier curves.

A sensitivity analysis was also performed including only the subset of patients who received 2 cycles of cladribine.

A p value less than 0.05 was considered as statistically significant. All analyses were performed using R (version 4.2.0, R Foundation for Statistical Computing, Vienna, Austria).

Data Availability

Anonymized data will be shared on reasonable request from a qualified investigator.

Results

Clinical data of 87,782 patients were available in the IMSR at the time of data extraction. After applying the inclusion criteria, we retrieved a cohort of 2,329 patients treated with at least one cycle of cladribine. The median (IQR) age at treatment start was 36.5 (29.2–45.2) years, and the median follow-up time after cladribine start was 1.7 (0.9–2.8) years; 71% (n = 1,654) were female. Table 1 lists the detailed demographic and clinical characteristics of the cohort. 1,488 patients (63.9%) received 2 courses of cladribine.

Table 1 Baseline Demographic and Clinical Characteristics Variable N = 2,329 Female sex, n (%) 1,654 (71) Age at onset (mean ± SD), y 29.47 ± 9.59 Age at diagnosis (mean ± SD), y 31.42 ± 9.72 Time from onset to diagnosis (mean ± SD), y 1.97 ± 3.66 Age at first cladribine cycle start (mean ± SD; median [IQR]), y 37.45 ± 10.61; 36.5 (29.2–45.2) Time from onset to first visit (mean ± SD), y 2.96 ± 4.99 Presence of CSF IgG oligoclonal bands, n (% of the total available data, 1,410 cases) 1,156 (82) Number of cerebrospinal fluid IgG oligoclonal bands, median (IQR) 10 (1–33) Disease duration at the time of the analysis (mean ± SD), y 10.78 ± 8.0 Age at the time of analysis (mean ± SD), y 40.21 ± 10.64 Updated patients (RISM-App codification), n (%) 1,915 (82.2) Patients starting cladribine as first DMT, n (%) 679 (2,29) Patients starting cladribine after ME DMTs, n (%) 1,321 (56.7) Patients starting cladribine after HE DMTs, n (%) 329 (14.1) Time between the last cladribine cycle and the following first DMT start (mean ± SD; median [IQR]), y (n = 268) 1.56 ± 1.04; 1.37 (0.82–2.26) Follow-up after cladribine (from first start date to last update) in those patients not receiving a new DMT, y (n = 2,061) 1.77 ± 1.22; 1.7 (0.9–2.8) Abbreviations: DMT = disease-modifying treatment; HE DMT = highly effective disease-modifying treatment; ME DMT = moderately effective disease-modifying treatment; RISM = Italian Multiple Sclerosis Register.

Baseline Demographic and Clinical Characteristics

Abbreviations: DMT = disease-modifying treatment; HE DMT = highly effective disease-modifying treatment; ME DMT = moderately effective disease-modifying treatment; RISM = Italian Multiple Sclerosis Register.

Treatment Patterns Before and After Cladribine

A total of 679 patients (29.1%) were treatment naïve, and 1,321 (56.7%) were previously treated with ME DMTs and 329 (14.1%) with HE DMTs. All previous DMTs administered before treatment with cladribine are presented in Table 2 . The median time between the date of the suspension of the last recorded DMT and the first cladribine cycle was 1.4 (0.8–2.2) months.

Table 2 Last Disease-Modifying Treatment (DMT) Before Cladribine Initiation in the Overall Cohort DMT before cladribine N = 2,329 Interferon-beta products 320 Teriflunomide 178 Azathioprine 9 Glatiramer acetate 231 Dimethyl fumarate 583 Fingolimod 209 Alemtuzumab 10 Rituximab 5 Mitoxantrone 5 Natalizumab 81 Ocrelizumab 13 Ponesimod 2 Ozanimod 4 Naïve 679

Last Disease-Modifying Treatment (DMT) Before Cladribine Initiation in the Overall Cohort

The number of patients starting a new treatment after the last cycle of cladribine was 268, and the average time between the last cycle and the beginning of the next DMT was 1.56 (1.04) years, with a median of 1.37 (0.82–2.26) years. After cladribine termination, we identified 8 patients re-treated with an additional cycle of cladribine, 148 with an anti-CD20 treatment (92 with ocrelizumab, 54 with ofatumumab, 2 with rituximab), 46 with natalizumab, 39 with a S1P receptor-modulator (3 with fingolimod, 17 with ozanimod, 3 with ponesimod, 16 with siponimod), and 5 with alemtuzumab. A total of 45 patients were treated with ME DMTs ( Table 3 ).

Table 3 Disease-Modifying Treatments (DMTs) After Cladribine Treatment DMT after cladribine N. first DMT after cladribine treatment N. second DMT after cladribine treatment Teriflunomide 16 1 Azathioprine 1 Glatiramer acetate 3 Fingolimod 3 Ofatumumab 51 3 Alemtuzumab 5 Rituximab 1 1 Cladribine 7 1 Siponimod 14 2 Ocrelizumab 84 8 Interferon-beta 7 1 Ponesimod 3 Dimethyl fumarate 15 1 Natalizumab 42 4 Ozanimod 15 2 No DMT 2,062 2,305

Disease-Modifying Treatments (DMTs) After Cladribine Treatment

Treatment pathways are visualized in the Sankey plot illustrated in Figure 1 .

Sankey Plot of Treatment Pathways

ARR Evaluation

In the overall cohort, ARR decreased ( p < 0.0001) from 0.96 (95% CI 0.91–1.02) in the 2 years preceding cladribine start to 0.09 (95% CI 0.08–0.11) during the 2 years after. All the analyses stratified by age and previous treatment confirmed the significant reduction of ARR ( p < 0.0001), regardless of age at treatment start and previous DMT efficacy profile. Considering the age cutoff of 50 years, pretreatment ARR of 0.61 (0.52–0.73) in patients who started treatment at least at 50 years of age converted to a post-treatment ARR of 0.04 (0.02–0.06); pretreatment ARR of 1.01 (0.95–1.08) in patients who began treatment at an age younger than 50 years changed to a post-treatment ARR of 0.1 (0.09–0.12). In naïve patients, the pretreatment ARR was 1.66 (1.52–1.81), but after cladribine treatment, it dropped to 0.09 (0.07–0.12). A similar trend is seen in patients who had ME DMTs before cladribine treatment, with a change in ARR from 0.68 (0.63–0.74) to 0.07 (0.06–0.09) after treatment, and in patients treated with HE DMTs before cladribine, with a change in ARR from 0.67 (0.57–0.69) to 0.18 (0.13–0.24) after treatment ( Figure 2 ).

ARR in the Entire Cohort and in the Subgroup Analysis

ARR = annualized relapse rate; DMT = disease-modifying treatment; HE = highly effective; ME = moderately effective.

In the subcohort of 1,488 patients who completed 2 cycles of cladribine, ARR decreased ( p < 0.0001) from 1.05 (95% CI 0.98–1.13) in the 2 years preceding cladribine start to 0.12 (95% CI 0.10–0.14) during the 2 years after. Regardless of age at treatment initiation or previous DMT efficacy profile, all analyses stratified by age and previous treatment confirmed the significant decrease in ARR ( p < 0.0001). Considering the age cutoff of 50 years, pretreatment ARR of 0.62 (0.50–0.77) in patients who started treatment at least at 50 years of age converted to a post-treatment ARR of 0.06 (0.04–0.11); pretreatment ARR of 1.12 (1.04–1.20) in patients who began treatment at an age younger than 50 years changed to a post-treatment ARR of 0.13 (0.11–0.15). In naïve patients, the pretreatment ARR was 1.92 (1.74–2.12) and it dropped to 0.13 (0.09–0.18) after the 2 cycles of cladribine. In patients treated with ME DMTs before cladribine, ARR decreased from 0.77 (0.70–0.85) to 0.09 (0.07–0.11) after the 2 cycles. After receiving full cladribine treatment, the ARR in patients treated with HE DMTs before cladribine went from 0.75 (0.64–0.90) to 0.20 (0.15–0.28).

PIRA Evaluation

PIRA events were calculated in a subcohort of 978 patients, all treated with 2 cycles of cladribine. The median follow-up was 6.71 years (0.32–28.29).

Considering the overall population, 133 PIRA events in 93 patients were reported during the noncladribine treatment period and 54 in 35 patients during cladribine therapy (HR 0.711, 95% CI 0.531–0.952, p = 0.0219) ( Figures 3 and 4 ).

Proportion of PIRA Events in the Entire Cohort and in the Subgroup Analysis

DMT = disease-modifying treatment; HE = highly effective; ME = moderately effective; PIRA = progression independent of relapse activity.

Kaplan-Meier Curves of the Proportion of Patients With Event-Free Survival Defined as PIRA-Free Probability

The figure shows the Kaplan-Meier curves reporting the proportion of patients without PIRA events (PIRA-free probability) stratified based on the time-dependent exposure to cladribine over the entire follow-up period. PIRA = progression independent of relapse activity.

In the group starting therapy at an age younger than 50 years (865 patients), 96 PIRA events were reported during the noncladribine treatment period and 42 during cladribine therapy (HR 0.679, 95% CI 0.489–0.943, p = 0.0207); in the group starting therapy after 50 years of age (113 patients), 37 PIRA events were reported during the noncladribine treatment period and 12 during cladribine therapy (HR 1.36, 95% CI 0.657–2.833, p = 0.405). The p value of cladribine treatment-by-age class interaction term corresponded to 0.0858. In addition, a significant p value of 0.000143 for the treatment-by-age interaction term indicated a more pronounced efficacy in reducing PIRA events in younger patients.

PIRA events were also evaluated by stratifying the cohort according to the preceding DMT. Therefore, the analysis considered naïve subjects (238 patients) and patients previously treated with ME or HE DMTs (560 and 180 patients, respectively). In naïve patients, 4 PIRA events were reported before the start of treatment and 11 during cladribine therapy (HR 0.61, 95% CI 0.191–1.949, p = 0.405). In patients first treated with ME DMTs, 79 PIRA events were reported before treatment initiation and 32 during cladribine therapy (HR 0.903, 95% CI 0.627–1.299, p = 0.582). In patients treated first with HE DMTs, 50 PIRA events were reported before treatment initiation and 11 during cladribine therapy (HR 0.465, 95% CI 0.240–0.901, p = 0.0233). The p value of cladribine treatment-by-previous treatment interaction corresponded to 0.612, thus indicating no evidence of a differential efficacy among groups with different previous treatments. Kaplan-Meier curves showing the proportion of patients with event-free survival defined as PIRA-free probability are illustrated in Figure 4 .

Classification of Evidence

This study provides Class IV evidence that for patients with RMS, cladribine treatment is associated with a reduction in ARR and PIRA events.

Discussion

In this multicenter, observational, retrospective cohort study based on RISM data, we assessed the effectiveness profile of cladribine, evaluating the effect on ARR and the impact on PIRA events. For patients with RMS, both naïve and switchers, as well as younger and older patients, the treatment with cladribine was successful.

Our results are in line with the growing amount of evidence about cladribine effectiveness published in the past few years. A recent monocentric real-world study on 114 patients with relapsing remitting MS treated with cladribine proved a reduction in relapses and MRI activity, along with a stabilization of disability, over the long term, with 74.9% of patients achieving no-evidence-of-disease-activity-3 (NEDA-3) status at 24 months. 22 In the MSBase Registry, clinical outcomes of dimethyl fumarate, teriflunomide, cladribine, and fingolimod were compared, proving that cladribine cohorts had significantly longer time to treatment discontinuation, longer time to first relapse, and lower ARR. 23 In the CLARITY study, cladribine tablets significantly decreased the probability of 3-month and 6-month confirmed disability worsening and the ARR by 58% when compared with placebo ( p < 0.001). 7 Another study demonstrated that, compared with the ARR of 0.67 (95% CI [0.56–0.79]) in the year before initiation of cladribine, ARR was reduced to 0.11 (95% CI [0.08–0.15]) in years 0–2 after 3-month re-baseline with cladribine (84.8% reduction). The analysis stratified by previous therapy confirmed a significant ARR reduction in all subgroups, although less pronounced in cladribine-treated patients switching from HE DMTs. 24 After a median follow-up period of 19 months, the mean ARR in a Finnish nationwide register study including 179 patients was found to be 0.1, as opposed to 1.0 before the start of cladribine therapy. 25 A reduction in the ARR at the 12-month follow-up after cladribine completion compared with the year before starting therapy (0.07 ± 0.25 vs 0.82 ± 0.80, p < 0.001) was also observed in a retrospective, multicenter Italian study. 26 Similar trends were observed in our cohort, regardless of age at treatment start and previous DMT efficacy profile.

PIRA is now considered the main driver of disability accumulation in MS. 14,15 Therefore, it is crucial to collect real-world evidence to understand whether the pathogenetic mechanisms sustaining PIRA can be modified by currently approved DMTs. Our study provides novel insights into the effect of cladribine on PIRA events, stratifying patients in different cohorts based on age and treatment history. We found a reduction in PIRA events, comparing the observation period before and after cladribine therapy, in all subcohorts of patients. The exception is the group of patients naïve to treatment. In this specific subgroup, it was not possible to see an effect of cladribine on the PIRA risk just because the model we used compared the periods before and after the treatment start. Because cladribine in naïve patients is used with a short delay after the diagnosis, there were more chances to detect a disability worsening independent of relapse activity after the treatment start than before. A longer follow-up is required to detect an effect on PIRA risk in naïve patients.

IRTs may be most beneficial when used early in the disease course. Zanetta et al. reported that cladribine was more effective in first-line therapy switchers or naïve patients. 22 Another study analyzing data from 243 patients with MS followed at 8 Italian MS centers showed that patients with a higher number of previous treatments had a lower probability of retaining NEDA-3 status and an increased risk of clinical relapses. 27 On the contrary, a recent review considering the changing demographic background of RMS and the challenges associated with aging people with RMS pointed out that long-term follow-up of RMS populations in randomized trials alongside real-world evidence suggests that the effectiveness of cladribine in suppressing MS disease activity is similar in older and younger age groups, with a solid safety profile. 28 In our cohort, cladribine resulted effective in both younger and older patients.

Research questions increase in tandem with real-world challenges: there have been an increasing number of questions in recent years concerning the continuation of therapy after the completion of 2 cladribine courses. Our results are consistent with other reports based on real-life data, confirming that ocrelizumab and natalizumab are the primary choices after cladribine therapy. 26,29 A recent expert opinion identified 5 patient categories classified by clinical and MRI activity and biomarker evaluation that could benefit from treatment with cladribine tablets beyond year 4, underlining that patients who do not achieve NEDA-3 status should be candidates for re-treatment with cladribine tablets. 30 In our register, we identified 8 patients treated with an additional cycle. This number may still be minimal because of the rules of prescription and refundability of the drug currently present in Italy.

A few limitations of our study should be taken into account. As with previous retrospective observational studies from the IMSR, incomplete or inaccurate data entered in the register must be acknowledged. In defining PIRA events, we considered solely the EDSS score because we could not incorporate MRI data because of the lack of a systematic acquisition of neuroimaging data in the RISM database. Because MRI biomarkers are a key element in determining the disease burden and constitute a critical prognostic factor, especially when assessing disease activity on therapy, this consideration must be emphasized. A longer follow-up will provide a deeper understanding of the sustained effectiveness of cladribine, especially on PIRA events. However, our study used the large and validated database of RISM to assess the effectiveness profile of cladribine and therapeutic choices after treatment completion.

In conclusion, results from our cohort extracted from the large and validated RISM database highlighted the effectiveness profile of cladribine in suppressing relapse activity and in reducing the occurrence of PIRA events, thus addressing neuroinflammation and neurodegeneration, the 2 faces of MS disability accumulation. A longer follow-up will be essential to validate these findings and to better characterize treatment patterns after the completion of both cladribine cycles.

Acknowledgment

The authors thank the Research Assistants of RISM (Italian MS Foundation - FISM): Beatrice Biolzi, Camilla Borgo, Antonella Carta, Paola Crida, Daniele Dell’Anna, Marco Delogu, Daniele Di Giulio Cesare, Sonia di Lemme, Chiara Di Tillio, Teresa Fonsdituri, Ilaria Maietta, Agata Marchese, Martina Marciano, Silvia Marinetto, Federica Martini, Chiara Monetti, Cristiana Morano, Ornella Moreggia, Silvia Perugini, Giovanna Ramona Piredda, Chiara Raimondi, Antonino Rallo, Monica Romoli, Ilaria Rossi, Valentina Tallarico, Stefania Treccarichi, Eliana Zaccone. The Research Assistants of RISM contributed to data collection.

Glossary

annualized relapse rate

disease-modifying treatment

functional scores

highly effective

interquartile range

immune reconstitution therapy

moderately effective

multiple sclerosis

oligoclonal band

progression independent of relapse activity

Italian Multiple Sclerosis Register

relapsing multiple sclerosis

Author Contributions

T. Guerra: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. M. Copetti: drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data. C. Zanetta: drafting/revision of the manuscript for content, including medical writing for content. F. Patti: drafting/revision of the manuscript for content, including medical writing for content. C.G. Chisari: drafting/revision of the manuscript for content, including medical writing for content. E. Barbuti: drafting/revision of the manuscript for content, including medical writing for content. E. Portaccio: drafting/revision of the manuscript for content, including medical writing for content. M. Foschi: drafting/revision of the manuscript for content, including medical writing for content. A. Conte: drafting/revision of the manuscript for content, including medical writing for content. D. Ferraro: drafting/revision of the manuscript for content, including medical writing for content. E.E. Cocco: drafting/revision of the manuscript for content, including medical writing for content. R. Fantozzi: drafting/revision of the manuscript for content, including medical writing for content. G.T. Maniscalco: drafting/revision of the manuscript for content, including medical writing for content. G. Salemi: drafting/revision of the manuscript for content, including medical writing for content. C. Tortorella: drafting/revision of the manuscript for content, including medical writing for content. D. Paolicelli: drafting/revision of the manuscript for content, including medical writing for content. M. Filippi: drafting/revision of the manuscript for content, including medical writing for content. M.P. Amato: drafting/revision of the manuscript for content, including medical writing for content. M. Trojano: drafting/revision of the manuscript for content, including medical writing for content. P. Iaffaldano: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no conflicts of interest with respect to the contents of the current study, but note that the patients in the study were treated with a number of disease-modifying drugs and that authors have received advisory board, membership, speakers honoraria, travel support, research grants, consulting fees, or clinical trial support from the manufacturers of those drugs, including Actelion, Allergan, Almirall, Alexion, Bayer Schering, Biogen, Celgene, Excemed, Genzyme, Forward Pharma, Ipsen, Medday, Merck, Mylan, Novartis, Sanofi, Roche, Teva, and their local affiliates. Go to Neurology.org/NN for full disclosures.

References

Associated Data

Data Availability Statement

Anonymized data will be shared on reasonable request from a qualified investigator.