The CLARION study design and status update: a long-term, registry-based study evaluating adverse events of special interest in patients with relapsing multiple sclerosis newly started on cladribine tablets

Abstract

Objective

To describe the design of the CLARION post-approval safety study (EU PAS Register number, EUPAS24484) and provide a status update, including characteristics of patients included up to 1 May 2021.

Methods

CLARION aims to further evaluate adverse events of special interest in patients who are newly initiating treatment with cladribine tablets for relapsing multiple sclerosis (MS). The study population consists of two cohorts: patients newly initiating cladribine tablets (cladribine cohort) and patients newly initiating oral fingolimod tablets (comparator fingolimod cohort), with an aim to include 8000 patients (4000 patients per cohort). The study relies on secondary use of data from pre-existing MS registries/data sources (except in Germany, where primary data collection is performed). The study is projected to last 15 years, with an anticipated 5-year inclusion period. Study outcomes are: malignancies; severe infections; tuberculosis; progressive multifocal leukoencephalopathy; other opportunistic infections; herpes zoster; severe lymphopenia (Grade ≥ 3); and treatment discontinuation.

Results

As of 1 May 2021, 2393 patients were included in CLARION from seven participating MS registries/data sources (cladribine cohort, n = 1266; fingolimod cohort, n = 1127). The majority of patients are female (cladribine cohort, 72.5%; fingolimod cohort, 68.0%), with mean age at onset of MS of 31.5 years for the cladribine cohort and 30.9 years for the fingolimod cohort. The majority of patients in both cohorts had relapsing MS (cladribine cohort, 92.1%; fingolimod cohort, 93.5%).

Conclusion

By providing further information on adverse events of special interest during long-term follow-up, CLARION will assist neurologists and patients regarding treatment decision-making for management of relapsing MS.

Keywords:

• Multiple sclerosis
• cladribine tablets
• study design
• safety

Introduction

Patients with multiple sclerosis (MS) require life-long treatment with disease-modifying therapies (DMTs); as such, it is important to understand the long-term safety of such therapies in routine clinical practice^1–3. This can be achieved through collection and assessment of high-quality drug utilization and safety data. The European Medicines Agency (EMA) Patient Registry Initiative provides guidance for a standardized approach to evaluating the benefit-risk of medicines in the European Economic Area using data from registries⁴. In response, MS registries have begun to collect safety data using a more systematic approach, including harmonization of data collection across different sources⁵.

In recent years, cladribine tablets 10 mg (3.5 mg/kg cumulative dose over 2 years) have been approved in many countries for the treatment of adult patients with relapsing MS. The mechanism of action of cladribine tablets in MS involves selective, transient reduction of lymphocytes; consequently, lymphopenia is an expected event during treatment^6–10. Long term safety data showed that, apart from herpes zoster, there was no increased risk of infections (including severe infections or opportunistic infections) in patients receiving cladribine tablets^9,10.

In placebo-controlled clinical trials of cladribine tablets, a numerical imbalance in the rate of malignancies was noted¹⁰. While this was not statistically significant, regulatory guidelines state that appropriate pharmacovigilance activities should be used to estimate the incidence rate of such rare events and generate long-term safety data¹¹. Regulatory authorities consequently requested a post-approval safety study to further characterize the safety profile of cladribine tablets in the real-world setting. The CLARION study (EU PAS Register number, EUPAS24484¹²) was therefore designed to compare adverse events of special interest (AESI) among patients with relapsing MS newly initiating either cladribine tablets or fingolimod tablets. AESI are defined as malignancies, severe infections, herpes zoster, tuberculosis, progressive multifocal leukoencephalopathy (PML), and other opportunistic infections. Fingolimod has been selected as the primary comparator for this study given that both cladribine tablets and fingolimod tablets are orally administered and have a similar indication, thereby minimizing possible indication and selection bias.

The objective of this manuscript is to describe the design of the CLARION study and provide a status update, including characteristics of included patients up to 1 May 2021. The main results of the first interim analysis are summarized elsewhere¹³.

Methods

Study design

CLARION is an observational, comparative cohort study involving multiple countries, mostly across Europe, where cladribine tablets have been launched. It is planned to obtain US data from the Department of Defense data source, pending results of an ongoing feasibility assessment. The study population consists of two cohorts: patients newly initiating cladribine tablets and patients newly initiating fingolimod tablets (hereafter referred to as the cladribine and fingolimod cohorts, respectively). In addition to the fingolimod control cohort, at completion of the study, an external comparator cohort of ocrelizumab-treated patients, obtained from the ongoing VERISMO study¹⁴, will be used to better contextualize the risk of malignancies.

CLARION relies on secondary use of data from pre-existing MS registries/data sources except in Germany, where primary data collection is performed. Data collection began from the launch of cladribine tablets in each participating country. The study is projected to last 15 years, with an anticipated 5-year inclusion period. Patient follow-up starts on the date a patient newly initiates cladribine tablets or fingolimod tablets, and continues for a period of 10 years (or until the patient is lost to follow-up, withdraws consent, or dies). Study objectives and corresponding outcomes are shown in Table 1.

Table 1. Objectives and cohort level outcomes in CLARION.

Primary objective	Primary outcome
To further characterize and compare the risk, in terms of incidence of AESI (malignancies, severe infections, tuberculosis, PML, other opportunistic infections, and herpes zoster)	Number of AESI in the cladribine and fingolimod cohorts
Secondary objectives	Secondary outcomes
To assess the impact of prior use of immunomodulatory/immunosuppressive agents on the incidence of AESI	Number of AESI by prior use of immunomodulatory/ immunosuppressive agents in the cladribine and fingolimod cohorts
To characterize the incidence and duration of severe lymphopenia among users of cladribine tablets	Number and duration of severe lymphopenia events in the cladribine cohort
To describe DMTs prescribed subsequent to cladribine tablets when disease activity recurs	DMT (immunomodulatory or immunosuppressive agents) after cladribine tablets treatment in patients with relapsing MS newly initiating cladribine tablets who experience a recurrence of the disease activity
Exploratory objectives	Exploratory outcomes
To characterize treatment discontinuation and the reasons for discontinuation	Number of patients discontinuing treatment, overall and by reason(s) for discontinuation, in the cladribine and fingolimod cohorts
To further characterize the relationship between severe lymphopenia and AESI among users of cladribine tablets	Number and time points of severe lymphopenia and AESI in the cladribine cohort

Abbreviations. AESI, adverse events of special interest; DMT, disease-modifying therapy; PML, progressive multifocal leukoencephalopathy.

Study population and data sources

Patients eligible for inclusion are those who newly initiated treatment with cladribine tablets or fingolimod tablets in line with the local labels for the respective treatment, after the date of cladribine tablets launch in the relevant country. Patients are considered as newly initiating their treatment if they have not previously received a prescription for cladribine tablets or fingolimod tablets. In Germany (primary data collection), patients are included after written informed consent is obtained at the date of treatment initiation or, at the latest, 24 weeks after treatment initiation.

Patients who have received fingolimod tablets before newly initiating cladribine tablets, or cladribine tablets before newly initiating fingolimod tablets, are not eligible for the study. The population coverage and data elements planned for inclusion in the study are shown in Table 2.

Table 2. Planned data sources, population coverage, and data elements to be included in CLARION.

Data source, per country	Population coverage	Data elements
DENMARK
Danish MS Registry	100% of Danish MS patients with DMT (personal communication with the register holder (February 2017))	Treatment discontinuation and reasons Exposure to DMTs Covariates
Registry of Medicinal Product Statistics	100% of the population with prescribed drugs purchased in community pharmacies in Denmark	Covariates
Danish National Patient Register	100% of the population with somatic inpatient or special outpatient care in Denmark	Severe infections, TB, PML, other OI, and HZ Covariates
Danish Cancer Register	100% of the population with cancer in Denmark	Malignancies Covariates
Danish Register of Causes of Death	100% of the deceased population in Denmark	Covariates
Danish Civil Registration System	100% of the population in Denmark	Covariates
FINLAND
Finnish MS Registry	16/20 hospital regions in Finland (as of April 2021)	Severe infections,^a TB, PML, other OI, and HZ Malignancies Severe lymphopenia Treatment discontinuation and reasons Exposure to DMTs Covariates
Prescription registers	100% of the population with prescribed reimbursed drugs purchased in community pharmacies in Finland e-register: 100% of the population with prescribed drugs purchased in community pharmacies in Finland	Covariates
Care Register for Health Care (HILMO; secondary care)	100% of the population with inpatient or special outpatient care visits in Finland	Severe infections, TB, PML, other OI, and HZ Malignancies Covariates
Register of Primary Health Care Visits (AvoHILMO)	100% of the population, all healthcare visits in primary care since 2011 in Finland	Severe infections, TB, PML, other OI, and HZ Malignancies Covariates
Finnish Cancer Registry	100% of the population with cancer in Finland	Malignancies Covariates
National Infectious Diseases Register	100% of the population diagnosed with one of the 70 selected infectious diseases in Finland	Severe infections, TB, PML, other OI, and HZ Covariates
Cause of Death Register	100% of the deceased population in Finland	Covariates
Finnish Population Information System	100% of the population in Finland	Severe infections, TB, PML, other OI, and HZ
FRANCE
OFSEP	Approximately 50% of MS patients in France	All outcomes Exposure to DMTs Covariates
GERMANY
MSDS 3D +/− NTD	The database is used purely for specific studies and all patients will be included per study protocol	All outcomes Exposure to DMTs Covariates
ITALY
Italian MS Registry	57 sites across Italy, approximately 80% (20,000 patients) of the Italian MS patient population^15	All outcomes Exposure to DMTs Covariates
NORWAY
Norwegian MS Registry and Biobank	Approximately 55% (6000 patients) of the Norwegian MS patient population^16	Severe lymphopenia Treatment discontinuation and reasons Exposure to DMTs Covariates
Norwegian Prescription Database (Nor PD)	100% of the population with prescribed drugs dispensed in pharmacies in Norway	Covariates
Norwegian Patient Registry	100% of the population with inpatient or special outpatient care visits in Norway	Severe infections, TB, PML, other OI, and HZ Severe lymphopenia Covariates
Cancer Registry of Norway	100% of the population with cancer in Norway	Malignancies Covariates
Cause of Death Registry	100% of the deceased population in Norway	Covariates
National Registry	100% of the population in Norway	Covariates
SWEDEN
Swedish MS Registry	Approximately 80% (14,500 patients) of the Swedish MS patient population^17	Severe lymphopenia Treatment discontinuation and reasons Exposure to DMTs Covariates
Swedish Prescribed Drug Register	100% of the population with prescribed drugs purchased in community pharmacies in Sweden	Covariates
National Patient Register	100% of the population with inpatient or special outpatient care visits in Sweden	Severe infections, TB, PML, other OI, and HZ Covariates
Swedish Cancer Register	100% of the population with cancer in Sweden	Malignancies Covariates
Cause of Death Register	100% of the deceased population in Sweden	Covariates
Total Population Register	100% of the population in Sweden	Covariates
SWITZERLAND
Swiss MS Cohort-Study	8 MS centers, 1302 MS patients in Switzerland	All outcomes Exposure to DMTs Covariates
US
Department of Defense (under feasibility)	Between 2004–2016, 8695 patients with a new diagnosis of MS^18	All outcomes Exposure to DMTs Covariates
MULTI-COUNTRY
MSBase	Multi-country, current participating sites: Australia: n = 8; Turkey: n = 2	All outcomes Exposure to DMTs Covariates

Abbreviations. DMT, disease-modifying therapy; HZ, herpes zoster; MS, multiple sclerosis; MSBase, Multiple Sclerosis database; MSDS 3D, Multiple Sclerosis Management System 3 Dimension; NTD, NeuroTransData; OFSEP, Observatoire Français de la Sclérose en Plaques; OI, opportunistic infections; PML, progressive multifocal leukoencephalopathy; TB, tuberculosis.

^aSeverity via linkage with national registries using hospitalization and diagnosis as proxy.

Outcomes and covariables

Study outcomes are shown in Table 1. Study-specific data will be registered for each of the following events by international classification of disease (ICD)-10 or MedDRA codes: malignancies, grouped and by individual type; severe infections, tuberculosis, PML, other opportunistic infections, and herpes zoster (regardless of severity). Severe lymphopenia (Grade ≥ 3) will also be recorded, either on the basis of laboratory test results (lymphocyte cell count <0.5 × 10⁹/L) or in Germany recorded as an adverse event in the electronic case report form.

Treatment discontinuation (including date and reasons for discontinuation) will also be evaluated.

The following variables will be collected at inclusion:

• Demographics.

• MS disease history and medical history, including infection status (e.g. hepatitis B virus [HBV]).

• Concomitant medicines and vaccines (type and date), including immunosuppressive/immunomodulatory agents (other than DMTs) such as corticosteroid treatment.

• Any personal and family history of malignancy.

• Use of sex hormones prior to inclusion (among women).

• Radiation exposure (e.g. UV light), alcohol consumption, smoking, and obesity.

The following variables will be collected during the follow-up period:

• Concomitant medications, including immunosuppressive/immunomodulatory agents (other than DMTs) such as corticosteroid treatment.

• Other infections (e.g. hepatitis B and C status).

• Use of sex hormones (among women).

• Radiation exposure (e.g. UV light), alcohol consumption, smoking, and obesity.

• Date of death, if applicable.

Regarding the use of DMT, the following data will be collected over the period prior to inclusion and during follow-up: type of DMT, as well as start and stop date. This will allow for presentation of data according to DMTs used and the two most recent DMTs used (including reasons for discontinuation). For each cohort, DMTs will be presented by drug type and grouped as immunosuppressive/immunomodulatory agents (according to influence of the agent on the immune system) and short-to-mid-term or long-term DMTs, according to the duration of the effect.

Definition of exposure

The primary definition of exposure for all outcomes (except for severe lymphopenia) is intention to treat (ITT), in which patients are classified according to treatment received at inclusion. Regarding severe lymphopenia and as part of sensitivity analysis for other outcomes, the definition of exposure is as-treated (AT) in which exposure is time-dependent and classified per on-going treatment at a given time: current exposure to cladribine tablets; current exposure to fingolimod tablets; or no current exposure to study treatment. At treatment discontinuation or switch from initial study treatment to another study treatment or to any other DMT, current exposure will be extended by a washout period of 6 months (or 12 months) for the discontinued treatment after which the exposure to the newly initiated treatment will start. Other definitions of exposure to cladribine tablets or fingolimod tablets will also be explored (Table 3).

Table 3. Additional exposure definitions.

For all primary outcomes	Description
On-treatment	ITT, with patient FU time censored 6 (or 12 months) after first discontinuation or switch (only the first treatment episode is considered)
Time-on-drug (VERISMO)^14	Equivalent to the time-on-drug exposure in the VERISMO study As-treated, but a patient would contribute to both cohorts in case of overlapping exposure. Two pairs of treatment-specific risk windows will be used
For malignancies	Description
ITT with an induction period of 6 (or 24) months	ITT, with events and FU time from 6 (or 24) months after cohort entry not counted
Time since last dose exposure	Categories: Current, recent (1–365 days after last dose), or past (>365 days after last dose) exposure to cladribine tablets or fingolimod tablets (6 categories)
Cumulative exposure	Categories: 0, 0–12, 12–24, 24–36, 36–48, and >48 months of cumulative exposure to cladribine tablets or fingolimod tablets (12 categories)
Ever-exposed (VERISMO)^14	Equivalent to the ever-exposed exposure in the VERISMO study ITT, but a patient can contribute to both cohorts

Abbreviations. FU, follow-up; ITT, intention to treat.

Data management and flow

Three types of data transfer to the clinical research organization (CRO; IQVIA) supporting Merck Healthcare KGaA for this non-interventional study will be used, depending on the participating MS registry/data source (Figure 1): Patient-level data transfer of primary data (Multiple Sclerosis Management System 3D and NeuroTransData), patient-level data transfer of secondary data (Finnish and Norwegian registries and Swiss MS Cohort-Study [SMSC]), and aggregate-level data transfer of secondary use of data (Danish, Italian, and Swedish registries, MSBase, and Observatoire Français de la Sclérose en Plaques [OFSEP]).

Figure 1. Overview of data flow from collection to the consolidated database in the CLARION study. CRO, clinical research organization; MSDS 3D, Multiple Sclerosis management system 3 Dimension; MSBase, Multiple Sclerosis database: NTD, NeuroTransData database; OFSEP, Observatoire Français de la Sclérose en Plaques; SMSC, Swish MS Cohort.

For Nordic countries (secondary data), data from MS registries are linked with national registers. In Finland and Norway, data are anonymized before each transfer. In Denmark and Sweden, data will be transferred already aggregated and linked.

For harmonization purposes, and to allow merging and analysis of data across the individual registries, a common data model (CDM) is used. Based on the list of outcomes and variables, study data elements are cleaned, mapped to common terminology, and transformed into a CDM by each registry providing aggregated-level data or by the CRO for those providing patient-level data. Registries providing aggregated-level data create aggregated datasets from the recorded patient-level data in a predefined standard format. This CDM defines the level of aggregation and minimum information required to enable the statistical analysis of study objectives. The CRO will combine the datasets from registries and data sources and perform the statistical analysis of the whole study.

Study size and data analyses

The study aims to include 8000 patients (4000 patients per cohort). Considering the primary study objective to further evaluate the risk of AESI, the study size estimation was based on simulated data and driven by malignancy as the AESI with the lowest incidence rate among the list of AESI being evaluated in CLARION. Four interim study analyses are planned at 3, 6, 9, and 12 years from the start of data collection, and a final analysis after study completion. Table 4 shows an overview of the planned analyses.

Table 4. Overview of planned analyses.

	1st interim analysis (Q3 2021)	2nd interim analysis (Q3 2024)	3rd interim analysis (Q3 2027)	4th interim analysis (Q3 2030)	Final analysis (Q4 2034)
Summary of disposition	x	x	x	x	x
Patient characteristics at inclusion	x	x	x	x	x
IR (and time) to malignancy (primary objective)	–	–	–	–	x
IR for first/recurrent event (primary objective)	x^a	x	x	x	x
CAIRR, first/recurrent event (primary objective)	–	x	x	x	x
IR difference for first/recurrent event (primary objective)	x^a	x	x	x	x
Analyses for secondary objective	x^b	x	x	x	x
Analyses for exploratory objectives	x^c	x	x	x	x

Abbreviations. IR, incidence rate; CAIRR, covariate adjusted incidence rate ratio.

Note. Sensitivity and other analyses to support the primary analysis will be conducted following the plan for the primary analyses.

^aThis interim analysis will not be conducted for malignancies; ^bThis interim analysis will not be conducted for the secondary objective “to assess the impact of the prior use of immunomodulatory/immunosuppressive agents on the incidence of AESI in patients with relapsing MS newly initiating cladribine tablets or fingolimod tablets”; ^cThis interim analysis will not be conducted for the exploratory objective “to further characterize the relationship between severe lymphopenia and AESI in patients with relapsing MS newly initiating cladribine tablets”.

The primary analysis will be the crude and adjusted incidence rates of AESI along with 95% confidence intervals (CI) in both cohorts. Incidence rates will be calculated as the number of events divided by the risk time in person-years. To compare the incidence of AESI between cohorts, the incidence rate ratios of AESI (together with the 95% CIs) will be estimated by Poisson regression models adjusted for key prognostic factors. For malignancies, time-to-event analysis using a weighted Cox proportional hazard model with inverse probability of treatment weights will be used to obtain a hazard ratio (HR) estimate and its corresponding 95% CIs using the ITT exposure definition. Specifically, time-to-malignancy will be analyzed using a meta-analytical approach in three steps: Step 1: estimation of propensity score (PS) weights within each data source. Step 2: estimation of the data source specific confounder-adjusted HRs from the time-to-event analysis using Cox proportional hazards model, incorporating the PS weights estimated in step 1. Step 3: estimation of the overall pooled confounder-adjusted HR from a meta-analysis of data source specific confounder-adjusted HRs.

The primary analysis described above will be repeated for the subset of patients with relapsing remitting MS only.

Secondary analyses will include the incidence rate and corresponding 95% CI of severe lymphopenia in patients treated with cladribine tablets; its duration will also be described. The effect of prior use of immunomodulatory/immunosuppressive agents on the incidence of AESIs in the two cohorts will be assessed by subgroup analysis. Also included are the number and percentage of patients that switch in the cladribine cohort as well as the type of DMT given after cladribine tablets.

Data quality

The validity of the primary outcome across all participating MS registries and data sources is acceptable, although some variation across registries may exist. For example, in the Nordic countries (Denmark, Finland, Norway, and Sweden), the ascertainment of malignancies is established through linkage with national cancer registries for the purpose of the study. In these countries, reporting a cancer event to the cancer registries is mandatory and the registries are considered as achieving high-quality standards in terms of completeness and accuracy of the registered data¹⁹. Close to 100% completeness of incidence of solid malignancies have been reported in each of the cancer registries^20,21. When compared to morphologically verified cases (i.e. biopsy or autopsy), the validity of the cancer registries ranges from 93% in Finland²⁰ to 98% in Sweden¹⁹.

In Germany, where primary data collection is used, as well as in the OFSEP (France) and SMSC (Switzerland) data sources, the completeness and validity of malignancies can be ascertained through data source verification at site level and feedback to centers with queries in case of missing data. This source verification process is not possible in the remaining data sources (MSBase [multi-country] and Italian MS Registry [Italy]).

The quality of data is assessed yearly using pre-specified data quality indicators (DQIs) and summarized in data quality assessment reports. Each participating data source is responsible for the quality and integrity of collected study data according to existing standard operating procedures or internal guidelines. A set of DQIs has been developed to monitor the quality of study data in terms of consistency, accuracy, completeness, and representativeness. These DQIs are based on the European Medicines Agency’s Patient Registries Initiative’s report on multiple sclerosis workshop held on 7 July 2017¹. This approach helps to identify potential data inconsistencies and methodological differences across registries early, including outcome definitions such as treatment discontinuation. To date, the application of data quality indicators in the CLARION study has improved data quality over time in terms of completeness and accuracy²². Detailed data quality results will be published in a separate manuscript.

Management and reporting of adverse events

The management and reporting of adverse events (AEs) vary according to the study model in the countries involved following EU GVP module VI revision 3. Non-interventional studies based on secondary use of data do not require reporting of suspected adverse reactions in the form of Individual Case Safety Reports, where the reporting obligation rests with the investigator. In Germany where primary data are collected, the reporting of AEs is based on additional forms as appropriate, including: (serious) AE, pregnancy, parent-child/fetus AE, and AE of special interest report forms.

Ethical considerations

The study is conducted according to Guidelines for Good Pharmacoepidemiology Practice (GPP) and the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) Code of Conduct, which ensures compliance with General Data Protection Regulation (GDPR). The study protocol was approved in Germany by the Ethikkommission of the Technische Universität Dresden (reference number: EK 338092018) and in Switzerland by the Ethikkommission Nordwest- und Zentralschweiz (EKNZ) (reference number: 2019-01949). Data permits are granted by approval bodies in each Nordic country.

Study status

As of May 2021, 2393 patients were included in the CLARION study (Figure 2); registry participation in the study, over time, is summarized in Table 5. Of the 2393 patients included, 1266 patients were in the cladribine cohort and 1127 in the fingolimod cohort; their characteristics are summarized in Table 6. The majority of patients are female (cladribine cohort, 72.5%; fingolimod cohort, 68.0%), with mean age at onset of MS of 31.5 years for the cladribine cohort and 30.9 years for the fingolimod cohort. The majority of patients included in each cohort had relapsing MS (cladribine cohort, 92.1%; fingolimod cohort, 93.5%). Approximately 75% of patients have at least one annual visit with their treating physician, on average, over the follow-up period.

Figure 2. Inclusion status in the CLARION study over time. The figure shows the number of patients included in the study at different data cut-offs, and this is regardless of the initial treatment time (some patients added in 2020 were already treated in 2019).

Table 5. Registry participation in the CLARION study, over time.

19.Nov.2018	01.May.2019	01.Nov.2019	01.Feb.2020	10.May.2020	01.Aug.2020	01.Nov.2020	01.Feb.2021	01.May.2021
MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD	MSDS 3D / NTD
	Finnish MS	Finnish MS	Finnish MS	Finnish MS	Finnish MS	Finnish MS	Finnish MS	Finnish MS
	Nor MS	Nor MS	Nor MS	Nor MS	Nor MS	Nor MS	Nor MS	Nor MS
	Swedish MS	Swedish MS	Swedish MS	Swedish MS	Swedish MS	Swedish MS	Swedish MS	Swedish MS
		Swiss MS	Swiss MS	Swiss MS	Swiss MS	Swiss MS	Swiss MS	Swiss MS
		Danish MS	Danish MS	Danish MS	Danish MS	Danish MS	Danish MS	Danish MS
			MSBase	MSBase	MSBase	MSBase	MSBase	MSBase

Abbreviations. Danish MS, Danish Multiple Sclerosis Registry; Finnish MS, Finnish Multiple Sclerosis Registry; MSBase, Multiple Sclerosis database; MSDS 3D, Multiple Sclerosis Management System 3 Dimension; Nor MS, Norwegian Multiple Sclerosis Registry and Biobank; NTD, NeuroTransData; Swedish MS, Swedish Multiple Sclerosis Registry; Swiss MS, Swiss Multiple Sclerosis Cohort-Study.

Table 6. Demographics and disease characteristics of patients in CLARION (cut-off date: 1 May 2021).

	Cladribine cohort n = 1266	Fingolimod cohort n = 1127
Age at inclusion, years
Mean (SD)	31.5 (9.5)	30.9 (9.8)
Range (min, max)	6.4–68	6–65
Female sex, n (%)	918 (72.5)	766 (68.0)
MS registry/data source (country), n (%)
Danish MS Registry (Denmark)	180 (14.2)	458 (40.6)
Finnish MS Registry (Finland)	128 (10.1)	159 (14.1)
MSBase (Australia, Turkey)	154 (12.2)	68 (6.0)
MSDS 3D (and NTD) (Germany)	294 (23.2)	192 (17.0)
Norwegian MS Registry and Biobank (Norway)	360 (28.4)	159 (14.1)
Swedish MS Registry (Sweden)	137 (10.8)	64 (5.7)
Swiss MS Cohort-Study (Switzerland)	13 (1.0)	27 (2.4)
MS clinical course at inclusion, n (%)
Relapsing remitting MS	1166 (92.1)	1054 (93.5)
Secondary progressive MS	41 (4.5)	46 (4.8)
Primary progressive MS	16 (1.3)	1 (0.1)
Primary relapsing MS	2 (0.2)	0 (0)
Other	1 (0.1)	6 (0.6)
Unknown	40 (3.2)	19 (1.7)
History of DMT use for MS prior to inclusion, n (%)
None	456 (36.0)	267 (23.7)
One DMT	372 (29.4)	409 (36.3)
Two DMT	211 (16.7)	242 (21.5)
≥ 3 DMT	227 (17.9)	209 (18.5)

Note. The Norwegian MS Registry and Biobank categorization does not include secondary progressive MS or primary relapsing MS; instead, they report data on relapsing remitting MS, primary progressive MS, and unknown. Denmark, Finland, and MSBase are able to report patients with “Other” MS course. Percentages are calculated out of 906 for the cladribine cohort and out of 968 for fingolimod.

Abbreviations. DMT, disease-modifying therapy; MS, multiple sclerosis; MSBase, Multiple Sclerosis database; MSDS 3D, Multiple Sclerosis Management System 3 Dimension; NTD, NeuroTransData; SD, standard deviation.

Discussion

Data generated with cladribine tablets in the clinical development program have demonstrated the efficacy and safety profile in the treatment of patients with relapsing MS^9,10,23–27, findings that have been confirmed in follow-up reports²⁸ and real-life studies of the safety and effectiveness of such treatment²⁹. As a large, non-interventional study reflecting real-world treatment and management of such patients¹², CLARION will provide additional evidence relevant to clinical practice given that it will involve a larger number of patients and a longer observation period than the clinical development program and the aforementioned studies. Furthermore, it’s design – using mainly pre-existing data – makes it an invaluable resource to potentially address new safety issues in patients treated with cladribine tablets.

Over time, national and international MS registries have improved data coverage and established collaborations with regulatory authorities and the pharmaceutical industry in the conduct of post-approval safety studies such as CLARION. The existing framework to combine data from multiple registries is expected to contribute to the advancement of knowledge and management of patients with MS⁵. Patient counts by cohort are obtained every 4 months from each participating MS registry/data source. This allows tracking of the number of patients included and implementation of mitigation plans as needed. For example, the feasibility of adding further MS registries to the CLARION study has been examined, so that the range of data sources can be broadened to increase the rate of patient inclusion at a later date if needed.

Limitations of the study and mitigation plans

Selection bias

To minimize selection bias, the inclusion criteria for the study were selected to be as broad as possible for this study population.

Generalizability of results

Since the study is based on pre-existing MS registries and conducted mainly via secondary use of data, some countries where no registry or data source was identified were not eligible for this study. However, countries with the highest MS incidence in Europe, such as Finland, Denmark, France, and Norway, are included or planned³⁰. Per protocol, those patients receiving fingolimod tablets before initiation of cladribine tablets were also excluded. The proportion of excluded patients is 19.5% overall, and varies across MS registries/data sources and countries.

Information bias

This is a distortion in the estimate of association between risk factor and disease that is due to systematic measurement error or misclassification of patients on one or more variables, either risk factor or outcome. To minimize this type of bias in the current study, instructions on variables and outcome definitions are provided to all neurologists/physicians in the centers participating in the study with primary data collection. In the Nordic national registries, the study outcomes and covariables were defined using validated algorithms when available.

Channeling bias

Cladribine tablets was a new drug on the market at the start of this study while fingolimod was well established. Thus, patients newly initiating cladribine tablets might be different from patients treated with fingolimod tablets, despite similar therapeutic indications, especially at the very beginning of the study. To address this potential channeling bias³¹, baseline characteristics in the two cohorts will be checked over time to identify any differences between the two cohorts. If such differences are identified, sub-group analysis will be performed.

Confounding by indication

This bias can arise because of the presence of an underlying diagnosis or other clinical features that trigger preferred use of either cladribine tablets or fingolimod and are related to patient outcomes. For example, patients on annual skin screening due to risk of basal cell carcinoma might be prescribed cladribine tablets more often than fingolimod, and this may impact study outcomes.

Limitations related to differences across data sources

In all Nordic countries, the national healthcare registers have a lag time of 3–12 months. Different categories may also be used by the data sources for recording the type of MS disease. For example, Norway’s NMSRB does not include primary and secondary progressive MS in their categorization. Different data sources may have different missing variables, such as date of diagnosis. The magnitude of this limitation, and impact on study results, will be assessed using analyses for Germany in which fewer missing variables are expected as it is primary data collection versus other data sources. Due to data protection restrictions in Denmark, a “tiny cell rule” applies for database cells with fewer than five observations (e.g. rare events). If fewer observations are contained in the aggregated data, a further aggregation might be applied. Alternatively, cells with fewer than five observations are clouded before the aggregated data is transferred. In CLARION, the inclusion period and number of patients included also varies across registries/data sources. This might introduce differences in patient characteristics that may be associated with, or reflect, the effect of underlying determinants and bring potential bias and/or confounding. To account for this potential impact, descriptive analysis of patient characteristics by cohort, country, and calendar time will be performed. Further analysis of study outcomes will be performed if the description suggests that imbalances or differences in patient characteristics have the potential to introduce bias and/or confounding.

Limitations related to the aggregated data model

Compared with patient-level data, the use of aggregate-level data will potentially limit the performance of additional analyses for the overall population.

Limitation related to missing variables

As an example, this may be an issue when measuring the duration of lymphopenia. Given that no regular blood tests are requested per protocol to confirm recovery of lymphocyte counts, there is a risk to overestimate the duration of lymphopenia. However, for patients with severe lymphopenia, which is the outcome of interest in CLARION, this risk will likely be minimized as the physician would be expected to perform a closer follow-up with more frequent blood testing.

Limitations relating to the comparison with an external cohort, regarding malignancy

Results of the comparison with the external cohort of ocrelizumab-treated patients need to be interpreted with caution due to potential selection bias and other differences in patient characteristics and their management. Inclusion and exclusion criteria for the ocrelizumab study may differ from the CLARION study population, and ocrelizumab administration is via intravenous infusion while cladribine and fingolimod are both oral. However, ocrelizumab is approved for a similar indication to cladribine tablets and its approval date will allow comparison of incidence of malignancies for patients included during the same time period as CLARION; moreover, both cohorts are expected to have a similar size.

Data representativeness and generalization of results

From the data collected currently in CLARION, the percentage of women and mean age of onset of MS is representative of the general MS population. The majority of included patients had previously received MS treatment, with a similar proportion of treatment-naïve patients across the two cohorts. However, patients included in the study are new users for cladribine tablets and fingolimod tablets with no prior fingolimod use. Potential exclusion of a number of patients with prior fingolimod use may limit the generalization of results to patients treated with cladribine tablets. Early findings from CLARION indicate that 18% of patients were excluded from the cladribine cohort because of prior fingolimod use¹³.

Conclusions

By providing further information on AESIs during long-term follow-up, CLARION will assist neurologists and patients regarding treatment decision-making for management of relapsing MS.

Transparency

Declaration of funding

The CLARION study is sponsored by Merck [CrossRef Funder ID: 10.13039/100009945].

Declaration of financial/other relationships

HB’s institution (Monash University) received compensation for consulting, talks, and advisory/steering board activities from Alfred Health, Biogen, Merck, Novartis, and Sanofi; research support from Biogen, Merck, MS Research Australia, National Health and Medical Research (Australia), Novartis, the Oxford Health Policy Forum, and Roche. He has received personal compensation for steering group activities from Oxford Health Policy Forum and Merck. NM has consulted with Merck. AA is an employee of EMD Serono Research & Development Institute, Inc. Billerica, MA, USA (an affiliate of Merck KGaA). JS was an employee of IQVIA at the time of this study, and is currently an employee of CATO SMS, Durham, NC, USA. IB is an employee of IQVIA, a contract research organization that performs commissioned pharmacoepidemiological studies for several pharmaceutical companies. PK was an employee of IQVIA at the time of this study, and is currently an employee of REWODAT Oy, Lohja, Finland. MS is an employee of Merck Healthcare KGaA, Darmstadt, Germany. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

Nicholas Moore, Aida Aydemir, Jaak Sõnajalg, Pasi Korhonen and Meritxell Sabidó were involved in the conception and design of the study whilst all authors were responsible for the analysis/interpretation of data, drafting and revision of the intellectual content and final approval of the paper.

Data availability statement

Any requests for data by qualified scientific and medical researchers for legitimate research purposes will be subject to the Data Sharing Policy of Merck. All requests should be submitted in writing to the data sharing portal of Merck https://www.merckgroup.com/en/research/our-approach-to-research-and-development/healthcare/clinical-trials/commitment-responsible-data-sharing.html. When Merck has a co-research, co-development, or co-marketing or co-promotion agreement, or when the product has been out-licensed, the responsibility for disclosure might be dependent on the agreement between parties. Under these circumstances, Merck will endeavour to gain agreement to share data in response to requests.

Acknowledgements

Medical writing assistance (in the form of writing assistance, including preparation of the draft manuscript under the direction and guidance of the authors, collating and incorporating authors’ comments for each draft, assembling tables and figures, grammatical editing and referencing) was provided by Mark O’Connor and Joseph Ward of inScience Communications, Springer Healthcare Ltd, Chester, UK, and funded by Merck Healthcare KGaA, Darmstadt, Germany.