Background

Alemtuzumab has proven to be an effective treatment for patients with highly active multiple sclerosis (MS). However, its use has been limited by adverse events (AEs) as secondary autoimmunity, being the most frequent those involving the thyroid gland, observed in around 40% of patients.

Objectives

To explore whether patient blood lymphocyte profile before alemtuzumab treatment initiation can identify patients with an increased risk of developing later autoimmunity.

Methods

A multicentre prospective longitudinal study was performed, including fifty‐four Relapsing-Remitting MS (RRMS) patients diagnosed in five Spanish hospitals. Patient blood samples were collected before initiating treatment with alemtuzumab. Autoimmune AEs were defined as the development, at any point within 2 years of follow-up, of any autoimmune thyroid-associated event, immune thrombocytopenia and/or autoimmune nephropathy. Differences were assessed using Man-Whitney U tests. Cut-off values were established using ROC curves to predict autoimmune AEs. Odds ratios were calculated by Fisher tests.

Results

Fifty‐four RRMS patients, 36 (66.7%) women, with a median (range) age of 28 (13–67) years and median (range) follow-up of 6 (0-20) years. Fourteen patients (25.9%) experienced autoimmune AEs, and all of them were dysthyroidism. No immune thrombocytopenia or nephropathies were observed. No statistical differences were found in clinical and demographic characteristics between patients who developed autoimmune AEs and those who did not. Patients who experienced autoimmune AEs before treatment onset had a higher percentage of blood CD19+ B cells (p=0.001), with a higher relative percentage of naïve B cells and plasmablasts. When explored total cell numbers, only plasmablast levels remained significant (p=0.02). A lower risk of autoimmune AEs after alemtuzumab was observed among patients with less than 7.6% of blood CD19+ B cells [odds ratio (OR) 16, confidence interval (CI) 3.86–58.95, p<0.0001] or less than 0.13% of plasmablast cells [OR 9.33, CI 2.17–42.65, p=0.002].

Conclusions

A low percentages of blood CD19+ B cells or plasmablasts before Alemtuzumab treatment predicted a lower risk of autoimmune AEs.

Background

Evobrutinib (EVO), a highly selective Bruton’s tyrosine kinase (BTK) inhibitor, has a dual mode of action on B cells and myeloid cells involved in multiple sclerosis (MS) pathogenesis. A Phase II randomized study (NCT02975349) investigated the effect of EVO on immune cells and immunoglobulins (Ig). After a 48-week randomized, double-blind period (DBP), relapsing MS (RMS) patients treated with EVO showed no evidence of B cell depletion or clinically relevant changes in memory or mature-naïve B cell subsets. IgG levels remained stable and slight elevations and reductions, respectively, in IgA and IgM levels were observed.

Objectives

To investigate the long-term effects of EVO on B cells (total, mature-naïve and memory subsets), T cells (total, helper and cytotoxic subsets), NK cells, and Ig levels after 48 additional weeks in the ongoing open-label extension (OLE).

Methods

Adults with RMS were randomized double-blind to EVO 25 mg QD, 75 mg QD, 75 mg BID, or placebo (PBO). PBO patients switched to EVO 25mg QD at Week 24. At Week 48, all patients were OLE-eligible, and received EVO 75 mg QD (median ≈48 weeks), then 75 mg BID. Safety of EVO, including assessment of total B cell counts and Ig levels, was a secondary endpoint; effects on B cell subsets, T cells, and NK cells were exploratory. Immune cell counts were assessed at OLE Week 48 relative to DBP baseline, and Ig levels at OLE Weeks 24 and 48.

Results

Of 213 patients receiving EVO during the DBP, 164 (77%) entered the OLE and 148 (90%) completed ≥60 additional treatment weeks. Investigation of total CD19+ B cells and B cell subsets revealed a decrease in CD19+ B cells and in mature-naïve B cells in all groups originally randomized to EVO. The decrease in mature-naïve B cells was consistent with that observed for CD19+ B cell counts, however no evidence of a change in the memory B cell levels was observed. No relevant changes in IgG levels relative to DBP baseline were observed. Mean IgA and IgM levels remained increased and decreased, respectively, but mean values were within normal ranges. Furthermore, there was no evidence of a change in T or NK cell parameters. Overall, EVO treatment was not associated with an increased risk of infections.

Conclusions

Immune cell numbers and Ig levels seen in patients receiving EVO for 48 weeks of the OLE were consistent with those in the DBP. The results suggest a gradual decline of B cells over time with consistent BTK inhibition, however the clinical meaningfulness of these changes remains to be determined. The observed changes in B cells, IgA and IgM levels were not associated with an enhanced risk of infections. These findings suggest that the continuous pharmacological inhibition of BTK over 96 weeks with EVO does not lead to substantial B cell reductions or changes in Ig levels.

Background

Serum neurofilament light chain (sNfL) levels are a promising biomarker for quantifying neuro-axonal injury in multiple sclerosis (MS). Recent studies showed sNfL levels relate to disease activity and treatment response.

Objectives

To analyze the predictive capacity of baseline (BL) sNfL levels for disease outcomes in the 15-year BENEFIT study long-term follow-up

Methods

Monofactorial regression analyses were conducted on outcomes at Year 15, including sNfL age-adjusted percentiles (80^th, 90^th, 95^th, 97.5^th and 99^th; Disanto et al., 2017).

Further BL covariates included sex; age; lowest EDSS up to Month 6; mono/multifocal onset; presence of optic nerve, brainstem, or spinal cord lesions; Paced Auditory Serial Addition Test 3 (PASAT-3), Timed 25 Foot Walk (T25W), and 9 Hole PEG test (9HPT) scores; number/volume of hypointense T1, gadolinium-enhancing T1, and T2 lesions; cerebral volume; and initial treatment assignment. On-study covariates at Years 1, 2, and 5 were annualized relapse rate; EDSS change; PASAT-3, T25W, and 9HPT scores; annualized rate of new lesions; lesion number/volume; brain volume change; and relative duration of treatment. Covariates with p≤0.1 (at ≥2 time points for on-study covariates) were entered into multifactorial models.

We assessed the predictive capacity of BL sNfL levels for the following outcomes: EDSS ≥4, clinically silent disease, T2 lesion volume, and cerebral volume.

Results

Patients with sNfL values at screening (N=258) above the 90^th (n=176), 95^th (n=157), 97.5^th (n=149), and 99^th (n=129) percentiles (adjusted for other relevant covariates) had a significantly higher risk of EDSS≥4 at Year 15 for BL model (odds ratio 2.45 [95% CI: 1.05,5.68] p=0.0376; 2.60 [1.18,5.75] p=0.0181; 2.61 [1.20,5.66] p=0.0154; 2.47 [1.19,5.13] p=0.0150, for the 90^th, 95^th, 97.5^th and 99^th percentiles, respectively), and for on-study models at Year 1 (3.86 [1.16,12.78] p=0.0272; 3.38 [1.13,10.14] p=0.0296; 2.92 [1.02,8.35] p=0.0461; 2.98 [1.11,8.00] p=0.0305, respectively) and Year 2 (5.36 [1.26,22.85] p=0.0231; 4.88 [1.34,17.77] p=0.0163; 3.86 [1.18,12.66] p=0.0259; 3.34 [1.17,9.48] p=0.0237, respectively). Covariates that remained statistically significant with EDSS≥4 as the endpoint in most percentile models were lowest EDSS value up to Month 6 for BL and Year 1, and change in brain volume at Year 1, 2, and 5.

Conclusions

Findings in this unique long-term BENEFIT study suggest that higher sNfL values in early MS disease stages are independent predictors of long-term disability.

Background

OPTIMUM was a multicenter, double-blind, active-comparator phase 3 superiority trial that assessed efficacy, safety, and tolerability of ponesimod 20 mg (PON) vs teriflunomide 14 mg (TER) in patients with relapsing multiple sclerosis (RMS). The primary endpoint of annualized relapse rate was met demonstrating PON’s superiority vs TER.

Objectives

To assess treatment effect on disability progression using time to worsening of timed 25-foot walk (T25FW), 9-Hole Peg Test (9HPT), Symbol Digit Modalities Test (SDMT), Paced Auditory Serial Addition Test [PASAT-3] , Expanded Disability Status Scale (EDSS) and composites of these endpoints.

Methods

OPTIMUM enrolled patients with RMS (EDSS:0-5.5), randomized (1:1) to daily PON or TER for 108 weeks. Change in MS functional composite (MSFC) Z-score (mean of T25FW, 9HPT, PASAT-3 Z-scores), and SDMT from baseline to Week 108 was assessed using Mixed-Effect Model Repeated Measures. Analyses of time to first confirmed (12-week) disability event were conducted post-hoc; disability was defined as 4-point worsening in SDMT, 20% worsening in T25FW or 9HPT, in addition to EDSS (as defined for the secondary endpoint).

Results

Of 1133 patients (PON=567, TER=566), 86.9% completed study. Changes from baseline in overall MSFC Z‑score: 0.02 PON vs −0.039 TER (mean difference, 0.059; p=0.047); for the 3 individual components of MSFC Z-score, mean differences (p-values) were: T25FW, 0.20 sec (p=0.37); 9HPT, –0.93 sec (p<0.0001); PASAT-3, 0.55 number correct (p=0.16). The PON vs TER worsening events up to EOS were, respectively: confirmed 20% worsening in T25FW, 9.2% vs 13.1% (hazard ratio [HR]:0.70; p=0.045); 4-point worsening in SDMT, 22.1% vs 26.4% (HR:0.82; p=0.12)], composite EDSS/SDMT, 26.3% vs 32.7% (HR:0.80; p=0.045)]; composite EDSS/9HPT/T25FW, 18.2% vs 24.0% (HR:0.76; p=0.035); numerical differences in favour of PON in time to confirmed worsening in 9HPT and SDMT assessed individually were not statistically significant.

Conclusions

Measures of worsening of impairment and disability in this exploratory analysis indicated benefits for PON vs TER. Composite endpoints provide more power to statistical assessments owing to greater number of events analyzed, making them particularly useful in RMS trials.

Background

Ozanimod, an oral sphingosine 1-phosphate receptor 1 and 5 modulator, is approved in the US and EU for the treatment of relapsing forms of multiple sclerosis (RMS).

Objectives

To characterize the long-term safety and efficacy of ozanimod in participants with RMS in an ongoing open-label extension (OLE) trial.

Methods

Participants with RMS who completed a phase 1, 2, or 3 ozanimod clinical trial were eligible to enroll in DAYBREAK (NCT02576717), where they received ozanimod 0.92 mg/d (equivalent to ozanimod HCl 1 mg). The primary objective was to evaluate safety in the overall population; treatment-emergent adverse events (TEAE) were monitored. Efficacy was evaluated with annualized relapse rate (ARR), calculated via negative binomial regression and pooled for all parent-trial treatment groups. Number of new/enlarging T2 and gadolinium-enhancing (GdE) MRI brain lesions were reported for the subset of participants who entered the OLE from an active-controlled phase 3 trial.

Results

In total, 2639 participants completed the parent trials; this interim analysis (data cut 20 December 2019) included 2494 participants with mean (range) ozanimod exposure of 35.4 (0.03–50.2) months in the OLE. Adjusted ARR in the OLE was 0.112 (95% confidence interval, 0.093‒0.135). At months 24 and 36, 79% and 75% of participants, respectively, were relapse free in the OLE. Three- and 6-month confirmed disability progression was observed in 10.8% and 8.6% of participants in the OLE, respectively. Mean number of new/enlarging T2 lesions per scan at 24 months was similar, regardless of parent-trial treatment group (range, 1.57–1.90), as were mean number of GdE lesions at month 24 (range, 0.2 ‒0.4). In the OLE, 2039 participants (81.8%) had any TEAE, 236 (9.5%) had a serious TEAE (SAE), and 56 (2.2%) discontinued due to a TEAE. Similar rates of TEAEs and SAEs occurred when assessed by parent-trial treatment group. The most common TEAEs were nasopharyngitis (17.9%), headache (14%), upper respiratory tract infection (9.9%), and lymphopenia (9.6%). TEAEs were generally similar to parent trial observations. There were no serious opportunistic infections. Exposure-adjusted incidence rates of TEAEs and SAEs have decreased over time.

Conclusions

In DAYBREAK, ozanimod was associated with low ARR and low new/enlarging T2 and GdE lesion counts over time. Most participants were relapse free and did not experience disability progression. Ozanimod was generally well tolerated and no new safety concerns emerged with long-term use.

Background

In a Phase II randomized study (NCT02975349) in patients with relapsing MS, evobrutinib (EVO) 75 mg twice-daily (BID) reduced total T1 Gd+ lesions (primary endpoint) and annualized relapse rate (ARR) over 24 weeks versus placebo, with efficacy maintained through Week 108. EVO was generally well tolerated.

Objectives

To describe the safety profile of EVO in the long-term treatment of MS by reporting detailed safety data from the study’s open-label extension (OLE) over 60 weeks.

Methods

In the 48-week double-blind period (DBP), patients received EVO 25 mg once-daily (QD) or 75 mg QD, 75 mg BID, or placebo for the first 24 weeks. All arms continued with the original treatment assignment until 48 weeks, except placebo patients who were switched to EVO 25 mg QD. At Week 48, all patients could enter the OLE, where treatment was initially EVO 75 mg QD (for a median of ~48 weeks) before switching to 75 mg BID. Safety was assessed throughout the OLE by the nature, severity, and occurrence of treatment emergent adverse events (TEAEs) by NCI-CTCAE v4.03 criteria, as well as vital signs, ECGs, and clinical laboratory safety parameters.

Results

Of 213 patients who received EVO during the double-blind period, 164 (77%) entered the OLE (safety analysis population) and 148 (90%) completed 60 weeks of treatment. Overall, 107 (65.2%) patients had a treatment emergent adverse event (TEAE), the majority of which were mild (47.6%) or moderate (36.0%), and none led to death. TEAEs were balanced across previous DBP treatment groups; the most frequent TEAEs over the OLE period, including the dose-switch, were nasopharyngitis (7.9%, Grade 2 or less), increased lipase (7.9%, Grade 3 or less), upper respiratory tract infection (6.1%, Grade 2 or less), and urinary tract infection (4.9%, Grade 2 or less); analysis of TEAEs by exposure-adjusted incidence rate showed no evidence of an increase after patients switched to 75 mg BID. Thirteen patients (7.9%) reported a serious TEAE, most frequently related to infections (6 patients, not treatment-related). Five patients (3.0%) had a TEAE during the OLE that led to treatment withdrawal, of which 3 were considered related to treatment (nausea, increased lipase, and increased lipase and amylase). The incidence of overall infections in the OLE was similar to that observed in the DBP. Transient elevated liver aminotransferases reported in the 48-week DBP were not observed in the OLE after prolonged treatment or after the switch to 75 mg BID. No adverse ECG findings were noted across all evobrutinib groups. There was also no apparent effect of EVO dose received in the DBP on safety parameters in the OLE.

Conclusions

In a 60 week OLE period of a Phase II study, the safety of EVO was similar to that seen in the DBP. Overall, long-term EVO treatment was generally well tolerated in patients with relapsing MS.

Background

The rapidly evolving therapeutic landscape of multiple sclerosis (MS) can make treatment decisions challenging. Novel tools using artificial intelligence (AI) can provide estimations of MS disease progression, which may aid MS therapeutic decisions. However, whether neurologists are willing to utilize information provided by AI-based models when making therapeutic decisions is unknown.

Objectives

To assess whether neurologists rely on clinical judgment (CJ) or quantitative/ qualitative estimations of disease progression provided by hypothetical AI-based models (assuming these models can reliably identify patients at high vs. low risk of disease progression) in simulated MS case scenarios.

Methods

Overall, 231 neurologists with expertise in MS from 20 countries were randomized to receive qualitative (high/low) or quantitative (85-90% vs. 15-20%) information regarding the likelihood of disease progression. Participants were presented with simulated MS case scenarios, and initially made 7 treatment decisions based on the clinical information using CJ. After randomization, participants made 10 treatment decisions using CJ and estimations of disease progression provided by AI models. We evaluated concordance and discordance of therapeutic decisions based on CJ and AI. The primary outcome was the proportion of “optimal” treatment decisions defined as treatment escalation when there was evidence of disease progression or continuing the same treatment when clinically stable. Mixed models were used to determine the effect of randomization group, case risk level, and CJ/AI. Clinicaltrials.gov #NCT04035720

Results

Of 300 neurologists invited to participate, 231 (77.0%) completed the study. Study participants had a mean age (SD) of 44 (±10) years. Of 2310 responses, 1702 (73.7%) were classified as optimal. Optimal decisions were more common for the high-risk vs. low-risk CJ group (84.5% vs 57.6%; p<0.001). There were no differences in the estimated odds of optimal responses between the quantitative vs. qualitative groups (OR 1.09; 95%CI 0.86, 1.39) after adjustment for pre-intervention responses. The estimated odds of optimal decisions for the high-risk vs low-risk CJ group was 2.96 (95%CI: 2.47, 3.56 ) after adjusting for group, pre-intervention responses, and AI-based estimations. For low-risk CJ cases, additional input by AI-based estimations was associated with a lower likelihood of optimal responses; being worse for high-risk vs. low-risk AI estimations (OR 0.235; 95%CI: 0.16, 0.340) adjusting for covariables.

Conclusions

Neurologists were more likely to make optimal treatment choices for high-risk simulated scenarios. The addition of hypothetical information provided by AI-based models- did not improve treatment decisions for low-risk cases. These results provide a framework for understanding therapeutic decision-making in MS neurologists, who are more reliant on their own CJ over AI-based tools.

Background

In 2017, a revision of the 2010 McDonald criteria for multiple sclerosis (MS) diagnosis in clinically isolated syndrome (CIS) patients has been proposed. However, its validation in a large multicenter cohort of CIS patients is still needed.

Objectives

To compare the performance of 2017 and 2010 revisions of the McDonald criteria with respect to MS development in a large multicentric cohort of CIS suggestive of MS.

Methods

Brain and spinal cord magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) examination obtained ≤5 months from CIS onset and a follow-up brain MRI acquired ≤15 months from CIS onset were assessed in 626 CIS patients from 9 European MS centres. The occurrence of a second clinical attack (clinically definite [CD] MS) was recorded. Performances of the 2017 and 2010 revisions of McDonald criteria for dissemination in space (DIS), time (DIT) and DIS plus DIT, also including OCB assessment, were evaluated with a time-dependent receiver operating characteristic curve analysis. Median time to MS diagnosis for the different sets of criteria was estimated through Kaplan-Meier curves.

Results

At the last evaluation (median=61.9 months [IQR=39.1-102.5]), 319 (51%) of 626 patients had CDMS. At 36 months, for DIS, the 2017 MRI criteria had higher sensitivity (0.84 [95% CI=0.79-0.88] vs 0.77 [0.72-0.82]), lower specificity (0.33 [0.28-0.39] vs 0.40 [0.35-0.46]), and similar area under the curve values (AUC, 0.59 [0.55-0.62] for both). The 2017 DIS plus DIT MRI criteria had higher sensitivity (0.68 [0.63-0.74] vs 0.62 [0.56-0.68]), lower specificity (0.55 [0.49-0.61] vs 0.62 [0.56-0.68]), and similar AUC values (0.62 [0.58-0.66] for both). CSF-specific OCB assessment as part of the 2017 criteria revision, increased the sensitivity (0.81 [0.75-0.85]), decreased specificity (0.40 [0.34-0.46]) and preserved AUC values (0.60 [0.56-0.64]). Median time to MS diagnosis was earlier with the 2017 revision compared to the 2010 or CDMS criteria, especially with OCB assessment (2017 revision with OCBs=3.6 months [3.1-4.0], 2017 revision without OCB=11.6 months [7.8-13.5], 2010 revision=13.9 months [12.4-15.3], CDMS=56.3 months [43.8-76.0]).

Conclusions

The 2017 revision of the McDonald criteria showed overall similar accuracy to the 2010 McDonald criteria in predicting CDMS development. The suggested modifications are expected to simplify the clinical use of MRI criteria without reducing accuracy and allow an earlier diagnosis of MS.

Background

Therapeutic inertia (TI) is a worldwide phenomenon affecting physicians who manage patients with chronic conditions. Previous studies in Multiple Sclerosis (MS) showed TI affects 60 to 90% of neurologists and up to 25% of daily treatment decisions.

Objectives

To determine the most important factors and levels of attributes associated with treatment escalation in an international sample of neurologists with expertise in the management of patients with MS.

Methods

We conducted an international study comprised of 300 neurologists with expertise in MS from 20 countries (Europe: 59.4%, Asia/Australia: 18.3%, America: 22.3%). Participants were presented with 12 pairs of simulated MS patient profiles reflective of case scenarios encountered in clinical practice. Patient profiles included information on age, sex, previous MS history of relapses, MRI findings, desire for pregnancy, and other relevant details. We used disaggregated discrete choice experiments (a conjoint analysis), which is a standard technique used in economic research to estimate the weight of factors and attributes (e.g. categories) affecting physicians’ decisions when considering treatment selection by asking respondents to choose between pairs of options. In our study, participants were asked to select the ideal candidate (Patient A, B or neither) for treatment escalation (from first-line to second-line therapies- eg. Fingolimod, Cladribine, Monoclonal antibodies).

Results

Of 300 neurologists invited to participate, 229 (76.3%) completed the study. The mean age (SD) of study participants was 44 (±10) years. The mean (SD) number of MS patients seen per week by each neurologist was 18 (±16).

The top 3 factors (relative importance) associated with treatment escalation were: previous relapses (20%), EDSS (18%), and MRI activity (13%). Patient demographics and desire for pregnancy had a modest influence (<3%) in treatment escalation.

Participants were 13% less likely to escalate treatment for patients with EDSS >7.0 (compared to EDSS <6.0), whereas symptom severity during most recent relapse and higher number of MRI lesions at 1 year were each associated with 6% higher likelihood of treatment escalation.

We observed differences in the weight of factors associated with treatment escalation between MS specialists and non-specialists and participants practicing in European vs. non-European countries.

Conclusions

This is the first study applying a conjoint design to assess factors associated with treatment escalation and therapeutic inertia in neurologists caring for people living with MS. Our results provide critical information on factors influencing neurologists’ treatment decisions and should be applied to continuing medical education strategies.

Background

Comorbidities and adverse health behaviors are associated with worse outcomes in multiple sclerosis (MS) and increased use of healthcare resources. Some comorbid conditions are more frequent in MS compared to the general population. The frequency and impact of comorbidities in our region is unknown.

Objectives

Describe the frequency of common comorbidities, adverse health behaviors, and healthcare resource usage in MS patients compared to the general population in Catalonia.

Methods

This is a population-based case-control study of the primary healthcare information system that covers 80% of the population of Catalonia. Cases were identified using the CIE-10 MS code (G35), and age/sex- matched controls (ratio 1:5) were randomly chosen, if they had at least one visit since 2006 and did not have active CIE-10 coding for any demyelinating disease. We obtained information on demographics (age, sex, socioeconomic status), comorbidities (by count and type), adverse health behaviors (smoking, alcohol), annual visits (primary care and specialists), sick leave days and medication dispensing.

Bivariate analysis and adjusted logistic regressions were done and odds ratios (OR) with 95% confidence interval (CI) were calculated.

Results

5548 MS cases and 27710 controls were included. 70% were female and mean age was 48.3 years. A total of 3334 (60.1%) of cases vs. 15756 (56.8%) of controls had at least one comorbidity (p<0.001). A higher frequency of comorbidities was found in MS in the 20-39 (OR: 1.377; 95%CI: 1.229-1.542) and 40-59 (OR: 1.231; 95%CI: 1.135-1.336) age ranges, whereas it was lower in 60-79 (OR: 0.648; 95%CI: 0.551-0.763) and >80 years (OR: 0.268; 95%CI: 0.115-0.625) age ranges. Socioeconomic deprivation was associated with a higher presence of comorbidities in MS cases (OR: 1.456; 95%CI: 1.161-1.824). Stroke (OR: 1.513; 95%CI: 1.173-1.952) and epilepsy (OR 2.566; 95%CI: 2.034-3.237), as well as any psychiatric disorder (OR 1.425; 95%CI: 1.377-1.519), bipolar disorder (OR: 1.882 95%CI: 1.335-2.654), or major depression (OR: 1.791: 95%CI 1.660-1.932) were more frequent in MS. Cardiovascular diseases were more frequent in males, whereas psychiatric diseases were more frequent in females. MS cases had higher annual sick leave days (11 vs. 6.7; p<0.001) and nurse (3 vs. 1.7; p<0.001), primary care (5 vs. 3.8; p<0.001) and specialist visits (11.9 vs. 0.5; p<0.001), as well as yearly medication dispensing. MS patients were more prone to smoking but not to alcohol consumption, especially among males.

Conclusions

MS patients have higher risk of psychiatric comorbidities, stroke and epilepsy, as well as adverse health behaviors and higher healthcare resource usage than the general population. The profile of comorbidities differs between women and men. Comorbidities are more frequent in the mos deprived socioeconomic strata.

Background

New T2 lesions count is routinely used for assessing disease activity in multiple sclerosis (MS), although their visual detection is challenging (low sensitivity, high variability).

Objectives

We assessed two different automatic decision-support systems to detect new T2 lesions in longitudinal brain MRIs of patients with MS.

Methods

The study included 100 MS patients with two MRI exams (median interval 12 months [range 3-27 months]; relapse free 85%). MRI scans were acquired on a 3T magnet following a standardized protocol (3D-FLAIR, 3D- MPRAGE, and 2D dual-echo T2-weighted sequences).

Two different automated methods were used: M1, based on an unsupervised approach that used intensity-derived features from the subtraction images together with deformation fields information obtained from the non-rigid registration between the two scans; and M2, a supervised approach based on the application of convolutional neural networks (CNN) trained to detect the presence of new T2 lesions in the follow-up scan. The outcomes of these automated tools were compared to the results of two operator-related methods based on visual analysis: the standard radiological report (O1); and revision of the MRIs by an expert observer non-blinded to the radiological report (O2). A “Gold Standard Outcome” (GOS) was created by consensus of two expert observers based on combined visual assessment of all the MRI images, the radiological reports, and the outcomes of the automated methods.

Results

GOS identified 104 new T2 lesions in 38 patients. Automated tools doubled the number of new T2 lesions (125 for M1; and 119 for M2) compared to operator-related methods (59 for O1 and 73 for O2). Specificity for detecting patients with at least one new T2 lesion was 100% for operator related methods while for automatic tools was 83% for M1 and 87% for M2. Sensitivity was higher with both automated tools (92.1% for M1; 97.4% for M2) compared to operator related methods (76.3% for O1, and 89.5% for O2).

Conclusions

The CNN model was more sensitive for detecting new T2 lesions and active patients, compared to standard and expert visual analysis, and to an unsupervised automated tool. However, visual supervision of the CNN model outcomes is still required due to its suboptimal specificity. Automatic tools, based on the application of CNN models are promising for detecting MRI disease activity, and shows potential to be used as an aid to the neuroradiological visual assessment in clinical practice.