Принудительный Дулаглутид в реплапсинге-излучающем многократном склерозе: Randomized Open-Lable Treaty-Of-Cost-Tribune.

Key Summary Points

Why carry out this study? MS is a chronic immune-mediated inflammatory and neurodegenerative disease in which progressive axonal injury and brain atrophy continue to drive irreversible disability and cognitive decline, even in patients with well-controlled inflammatory activity on effective disease-modifying therapies. GLP-1RAs have demonstrated metabolic, anti-inflammatory, and neuroprotective effects in preclinical models of neurodegeneration, but their potential as adjunctive therapy in MS—particularly in patients with residual neurodegeneration despite stable immunotherapy—had not been clinically evaluated. This proof-of-concept randomized trial investigated whether 12 months of adjunctive treatment with the GLP-1RA dulaglutide could influence biomarkers of neurodegeneration (plasma NfL and MRI volumetry), metabolic parameters, and functional outcomes in patients with RRMS receiving stable natalizumab therapy. What was learned from the study? Adjunctive dulaglutide 0.75 mg weekly for 12 months significantly improved body weight, BMI, adiposity, and glucose tolerance, and was well tolerated with no adverse events, injection-site reactions, or treatment discontinuations. Dulaglutide did not significantly affect plasma NfL concentrations or MRI volumetric measures compared with controls, and no significant between-group difference in cognitive performance was observed. Despite the absence of a detectable neuroprotective signal on plasma NfL or MRI volumetry—likely reflecting underpowering and a floor effect created by well controlled disease on natalizumab—exploratory analyses suggested modest improvements in walking speed and manual dexterity, warranting further investigation in adequately powered trials. These findings demonstrate that GLP-1RA therapy can deliver meaningful metabolic benefits in people with MS and provide important information for the design of future trials, which should incorporate larger sample sizes, longer follow-up, higher-dose regimens, blinded assessors, and populations with greater residual neurodegeneration or metabolic burden.

Introduction

Multiple sclerosis (MS) is a chronic immune-mediated inflammatory and neurodegenerative disease affecting predominantly the white matter of the CNS [ 1 , 2 ]. MS involves inflammatory and neurodegenerative processes from the earliest stages: beyond demyelination, ongoing axonal injury and neuronal loss contribute to accelerated brain atrophy that exceeds normal aging. This neurodegenerative component is a major driver of irreversible disability and cognitive decline, including progression independent of obvious relapses. MS remains a leading cause of disability in working-age adults with substantial socioeconomic impact [ 3 ].

Current MS management includes corticosteroids for acute relapses and disease-modifying therapies (DMTs) that suppress or modulate immune response to reduce the inflammatory disease activity [ 4 ]. However, available therapies have a limited ability to prevent or reverse neurodegeneration, which underscores the need for neuroprotective strategies.

Glucagon-like peptide 1 receptor agonists (GLP-1RAs) are potential candidates to address this gap. Although developed for type 2 diabetes mellitus to improve glycemic control and promote weight loss [ 5 , 6 ], GLP-1RAs have pleiotropic properties relevant to neurodegenerative conditions. Preclinical and early clinical evidence suggests neuroprotective effects mediated by anti-inflammatory signaling, reduced oxidative stress, insulin re-sensitization, support of neuronal and oligodendroglial survival, and potential pro-remyelinating actions [ 7 ].

Accordingly, the primary objective of this proof-of-concept trial was to evaluate whether adjunctive GLP-1RA therapy (dulaglutide) in relapsing–remitting MS (RRMS) influences biomarkers and measures associated with neurodegeneration, including plasma neurofilament light chain (NfL), brain magnetic resonance imaging (MRI) volumetry, and cognitive performance [ 8 ].

Methods

Study Design

This exploratory, unblinded, randomized, single-center clinical trial evaluated 12 months of adjunctive dulaglutide therapy (Trulicity, 0.75 mg subcutaneously once weekly) in patients with confirmed RRMS, receiving stable natalizumab treatment.

Primary endpoints were plasma NfL and brain MRI volumetry. Secondary and exploratory endpoints included anthropometric and metabolic measures, OGTT-derived insulin sensitivity indices, resting metabolic rate (RMR), physical activity energy expenditure (PAEE), Timed 25-Foot Walk (T25FW), 9-Hole Peg Test (9HPT), and Symbol Digit Modalities Test (SDMT).

Patients receiving natalizumab were specifically selected as the study population for several reasons: to ensure clinical homogeneity in background immunological status; to provide a stable and well-characterized platform of inflammatory suppression against which adjunctive metabolic and neuroprotective effects could be evaluated; and because natalizumab-treated patients continue to experience residual neurodegeneration despite effective suppression of inflammatory disease activity—representing a clinically important unmet need and a rational target for adjunctive neuroprotective strategies [ 9 ]. The regular monthly infusion schedule also facilitated study visit adherence and monitoring over the 12-month follow-up period.

The 0.75-mg weekly dose was selected for several reasons. First, this is the approved starting and maintenance dose for dulaglutide in type 2 diabetes in the European Union, with an established safety and tolerability profile. Second, at the time of study design (2018–2019), no data were available on GLP-1RA use in MS, and a conservative dose was chosen to prioritize participant safety in this novel indication. Third, the higher approved dose of 1.5 mg weekly was considered for future dose-escalation studies pending tolerability data from this proof-of-concept trial.

The trial was registered in the EU Clinical Trials Register (EudraCT 2019-003001-94) and approved by the Ethics Committee of the Old Town Hospital, University Hospital Bratislava, on 25 September 2019, under reference VV-2018-R-EK. The study was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent after receiving detailed study information and training in the administration of dulaglutide.

Participants

Eligible participants were adults aged 18–65 years with RRMS and high disease activity, disease duration ≥ 12 months, Expanded Disability Status Scale (EDSS) < 5, and stable natalizumab treatment for ≥ 3 months. Exclusion criteria included diabetes treated with insulin or oral antidiabetic drugs; serious chronic rheumatologic, metabolic, hematologic, pulmonary, hepatic, or cardiovascular disease; malignancy; alcohol or substance use disorder; metal implants contraindicating MRI; long-term glucocorticoid or antidepressant use; and known hypersensitivity to the medication investigated. Pregnancy and breastfeeding were exclusion criteria at enrollment and throughout the study.

Thirty patients were enrolled. After baseline assessment, participants were randomized to an active group receiving adjunctive dulaglutide for 12 months or to a control group without adjunctive therapy. Two participants in the control group were excluded after randomization as a result of newly developed conditions meeting exclusion criteria, yielding 28 participants for analysis (Fig.  1 , CONSORT flow diagram).

CONSORT flow diagram. Thirty patients were screened and enrolled. After baseline assessment, participants were randomised to the active group (adjunctive dulaglutide, n  = 15) or the control group (no adjunctive therapy, n  = 15). Two participants in the control group were excluded post-randomization as a result of newly developed conditions meeting exclusion criteria, yielding n  = 13 in the control group. All 28 remaining participants completed 12-month follow-up and were included in the analysis (active: n  = 15; control: n  = 13)

Study Protocol

Participants underwent assessments at baseline and at 12 months of follow-up in the 1st Department of Neurology, Faculty of Medicine, Comenius University, Bratislava, Slovakia. Participants fasted for 12 h prior to testing and avoided vigorous physical activity for 24 h. Anthropometrics (height, weight, waist, and hip circumference) were recorded. Body fat percentage, visceral fat, and muscle mass were assessed using bioelectrical impedance (Omron BF 511; Omron Healthcare, Kyoto, Japan). Blood pressure and heart rate were measured after ≥ 5 min of rest.

Functional outcomes included the Timed 25-Foot Walk (T25FW; average of two trials) [ 10 ] and the 9-Hole Peg Test (9HPT; two trials per hand) [ 11 ]. Cognitive performance was assessed with the Symbol Digit Modalities Test (SDMT) [ 12 ]. Functional assessments were performed by trained research staff in accordance with standardized protocols. However, full assessor blinding was not implemented because participants were aware of their treatment allocation, which could not be fully concealed from assessors in a clinical setting.

Resting metabolic rate (RMR; kcal/24 h) and respiratory quotient (RQ) were measured by indirect calorimetry (Cosmed Quark RMR with OMNIA software; COSMED, Rome, Italy) using a canopy hood under standardized conditions. Participants rested supine for ≥ 10 min before testing. After flowmeter and gas sensor calibration, oxygen consumption (VO 2 ) and carbon dioxide production (VCO 2 ) were recorded for 20 min. RMR, RQ, and substrate utilization were calculated automatically. Steady state was defined as consecutive measurements with < 10% variation in both VO 2 and VCO 2 .

Daily physical activity energy expenditure (PAEE; kcal/24 h) was estimated using the Slovak version of the Lagerros questionnaire based on self-reported frequency, duration, and intensity of activities [ 13 ].

A cubital vein cannula was placed for blood sampling. After a baseline sample (0 min), participants underwent a 75-g oral glucose tolerance test (OGTT; 250 mL water consumed within 3 min). Samples were collected at 15, 30, 45, 60, 90, and 120 min. Serum and EDTA plasma aliquots were stored at − 70 °C until analysis.

Biochemical Analyses and Calculations

Plasma glucose was measured in a certified hospital laboratory using an autoanalyzer (Siemens Healthcare Diagnostics, Tarrytown, NY, USA) with enzymatic kits (Roche Diagnostics, Lewes, UK). Plasma insulin was measured by ELISA (ALPCO Diagnostics, Salem, NH, USA).

Insulin resistance/sensitivity indices were calculated using fasting values (HOMA-IR; [ 14 ]) and OGTT-derived glucose and insulin (Cederholm index [ 15 ]; Matsuda index [ 16 ]).

Plasma NfL was measured using the SIMOA® platform according to the manufacturer’s protocol (Quanterix, Lexington, MA, USA) [ 17 , 18 ].

Brain MRI was performed on a 3-T scanner (Philips Ingenia 3.0 T; Philips North America) using a standardized MS protocol as previously described [ 19 ], within 4 weeks of baseline and follow-up assessments. Volumetry was analyzed using Icobrain, including whole brain volume, gray matter volume, FLAIR lesion volume, T1 hypointense lesion (“black hole”) volume, and gadolinium-enhancing lesion volume, as previously described as described extensively in our previous study [ 8 ].

Statistical Analysis

Analyses were performed with IBM SPSS Statistics v25 (IBM, Chicago, IL, USA). Normality was assessed using the Shapiro–Wilk test. Between-group comparisons at a single time point used Student’s t  test or Mann–Whitney U  test, as appropriate. Within-group changes used paired t  tests or Wilcoxon signed-rank tests. Group, time, and group × time effects across 12 months were evaluated using a repeated-measures general linear mixed model. Data are presented as mean ± SD unless stated otherwise. Two-sided  p  < 0.05 was considered statistically significant.

Effect sizes were calculated as Cohen’s d for parametric between-group comparisons (T25FW and non-dominant 9HPT measures). For the primary neurodegeneration endpoints (plasma NfL and brain volume), a post hoc power analysis was performed using G*Power 3.1 to estimate the power in a current sample.

Results

Twenty-eight participants (14 women/14 men; age 39.4 ± 8.7 years) were randomized to the active (adjunctive dulaglutide) or control group. Baseline characteristics are shown in Table  1 . There were no statistically significant baseline differences in age, sex, EDSS, or body mass index (BMI), indicating broadly comparable groups. Changes over 12 months are summarized in Table  2 . No relapses were observed in both groups. At baseline, the metabolic profile of the active group was characterized by a mean BMI of 26.4 ± 3.4 kg/m 2 , placing the group on average in the overweight range, with 2 participants (13.3%) meeting criteria for obesity (BMI ≥ 30 kg/m 2 ) and 7 (46.7%) classified as overweight (BMI 25–29.9 kg/m 2 ). Mean fasting glucose was 5.1 ± 0.4 mmol/L and mean HOMA-IR was 1.43 ± 0.64, both within the normal range. One participant (6.7%) had HOMA-IR > 2.5, exceeding the commonly used threshold of HOMA-IR for insulin resistance. No participant met formal diagnostic criteria for metabolic syndrome (International Diabetes Federation 2006 criteria) or type 2 diabetes at enrollment. In the active group, BMI decreased significantly ( p  = 0.003), driven by weight loss (80.3 ± 11.4 vs 78.0 ± 11.4 kg;  p  = 0.002).

Table 1 Baseline characteristics of participants randomized to adjunctive dulaglutide vs control Active GLP-1RA group Control group p n 15 13 Age (years) 41.7 ± 10.0 36.8 ± 6.3 0.140 Sex 7F/8 M (46.7/53.3%) 7F/6 M (53.8/46.2%) 0.708 EDSS 4.0 (1.0–4.5) 2.0 (1.0–4.5) 0.130 Disease duration (months) 166.0 ± 82.8 157.4 ± 64.7 0.761 BMI (kg/m 2 ) 26.4 ± 3.4 23.5 ± 3.9 0.053 Body fat (%) 29.4 ± 9.7 27.5 ± 10.8 0.514 Muscle mass (%) 32.4 ± 5.8 32.3 ± 6.9 0.980 PAEE (kcal/24 h) 3647.2 ± 1175.7 3153.2 ± 750.7 0.205 Systolic BP (mmHg) 122.9 ± 14.8 113.9 ± 10.9 0.082 Diastolic BP (mmHg) 79.7 ± 11.1 76.3 ± 6.9 0.344 HR (beats/min) 75.3 ± 12.0 80.0 ± 12.3 0.320 Fasting glucose (mmol/L) 5.1 ± 0.4 5.0 ± 0.4 0.425 Fasting insulin (mU/L) 6.2 ± 2.6 4.7 ± 3.6 0.214 Data are presented as mean ± SD for normally distributed variables, median (min–max) for non-normally distributed variables, and numbers and percentages (%) for categorical variables EDSS Expanded Disability Status Scale, BMI body mass index, PAEE physical activity energy expenditure, BP blood pressure, HR heart rate, GLP-1RA glucagon-like peptide 1 receptor agonist

Baseline characteristics of participants randomized to adjunctive dulaglutide vs control

Data are presented as mean ± SD for normally distributed variables, median (min–max) for non-normally distributed variables, and numbers and percentages (%) for categorical variables

EDSS Expanded Disability Status Scale, BMI body mass index, PAEE physical activity energy expenditure, BP blood pressure, HR heart rate, GLP-1RA glucagon-like peptide 1 receptor agonist

Table 2 Changes from baseline to 12 month follow-up in the GLP-1RA group and controls GLP-1RA subjects ( n  = 15) a Controls ( n  = 13) b Baseline Follow-up P Baseline Follow-up P BMI (kg/m 2 ) 26.4 ± 3.4 25.6 ± 3.5 0.003** 23.5 ± 3.9 24.0 ± 4.1 0.058 Systolic BP (mmHg) 122.9 ± 14.8 119.4 ± 13.3 NS 113.9 ± 10.9 117.2 ± 12.2 NS Diastolic BP (mmHg) 79.7 ± 11.1 81.3 ± 9.7 NS 76.3 ± 6.9 77.0 ± 6.7 NS HR(/min) 75.3 ± 12.0 80.5 ± 9.1 0.052 80.0 ± 12.3 81.5 ± 16.8 NS Waist circumference (cm) 92.4 ± 11.9 92.5 ± 11.0 NS 83.2 ± 10.6 84.2 ± 11.0 NS Hip circumference (cm) 105.5 ± 7.3 105.5 ± 7.3 NS 100.0 ± 8.8 101.7 ± 8.5 NS Body fat (%) 29.4 ± 9.7 27.8 ± 9.4 0.065 27.5 ± 10.8 28.5 ± 10.5 0.081 Body muscle mass (%) 32.4 ± 5.8 32.8 ± 5.8 NS 32.3 ± 6.8 31.8 ± 6.5 NS Visceral fat 8.2 ± 3.5 7.3 ± 3.8 0.012* 5.7 ± 2.3 5.9 ± 2.7 NS RMR—measured (kcal/24 h) 1854.3 ± 282.2 1680.8 ± 262.3 0.013* 1583.1 ± 285.5 1516.8 ± 257.3 NS PAEE (kcal/24 h) 3647 ± 1176 3535 ± 1281 NS 3153 ± 751 3079 ± 654 NS 9-HPT dominant hand (average/s) 25.5 (19.5–43.1) 24.7 (18.6–68.7) NS 21.0 (17.6–48.8) 19.3 (15.9–42.6) NS 9-HPT non-dominant hand (average) (s) 25.9 (18.9–44.5) 25.4 (18.5–38.1) 0.077 20.7 (17.2–44.6) 20.2 (16.9–51.1) NS T25FW (average) 5.3 (3.6–11.1) 5.0 (3.5–11.0) 0.065 4.7 ± 0.6 4.9 ± 1.1 NS SDMT 44.0 ± 17.5 46.9 ± 18.4 0.027* 47.0 ± 10.5 49.9 ± 10.9 0.002** NfL (pg/mL) 8.1 ± 3.7 8.2 ± 3.2 NS 7.6 ± 2.9 7.5 ± 4.2 NS Brain volume (percentile) 5.7 (1.0–79.8) 4.8 (1.0–45.7) 0.069 9.1 (1.0–58.0) 4.9 (1.0–58.2) NS Volume gray matter (percentile) 34.9 ± 28.4 29.5 ± 21.4 NS 38.0 ± 22.3 38.0 ± 25.5 NS FLAIR lesions volume (mL) 11.9 ± 8.0 12.3 ± 8.1 NS 10.7 ± 8.6 11.6 ± 10.0 NS T1 black holes volume (mL) 8.6 ± 6.3 9.3 ± 6.8 0.008** 8.0 ± 6.9 8.7 ± 7.9 0.094 Gd enhancing lesion volume (mL) 0 0 NS 0. 0 (0–0.01) 0 NS AUC glucose 872 ± 162 744 ± 166 < 0.001*** 864 ± 117 815 ± 171 NS AUC insulin 7426 ± 2554 7278 ± 1448 NS 7278 ± 5017 6217 ± 4129 NS IR HOMA 1.43 ± 0.64 1.26 ± 0.67 NS 1.11 ± 0.84 0.99 ± 0.50 NS ISI Matsuda 5.70 ± 2.31 7.02 ± 3.97 0.099 7.76 ± 3.91 8.58 ± 4.47 NS ISI Cederholm 51.8 ± 13.1 62.0 ± 16.6 0.003** 52.9 ± 10.4 59.3 ± 14.0 0.035* Data are presented as mean ± SD for normally distributed variables and median (min–max) for non-normally distributed variables.  p  values reflect within-group change from baseline to follow-up BMI body mass index, BP blood pressure, HR heart rate, RMR resting metabolic rate, PAEE physical activity energy expenditure, 9HPT 9-Hole Peg Test, T25FW Timed 25-Foot Walk, SDMT Symbol Digit Modalities Test, NfL neurofilament light chain, FLAIR fluid-attenuated inversion recovery, Gd gadolinium, GLP-1RA glucagon-like peptide 1 receptor agonist a 7 women, 8 men; age 41.7 ± 10.0 years b 7 women, 6 men; age 36.8 ± 6.3 years Statistical significance: * p  < 0.05, ** p  < 0.01, *** p  < 0.001

Changes from baseline to 12 month follow-up in the GLP-1RA group and controls

Data are presented as mean ± SD for normally distributed variables and median (min–max) for non-normally distributed variables.  p  values reflect within-group change from baseline to follow-up

BMI body mass index, BP blood pressure, HR heart rate, RMR resting metabolic rate, PAEE physical activity energy expenditure, 9HPT 9-Hole Peg Test, T25FW Timed 25-Foot Walk, SDMT Symbol Digit Modalities Test, NfL neurofilament light chain, FLAIR fluid-attenuated inversion recovery, Gd gadolinium, GLP-1RA glucagon-like peptide 1 receptor agonist

a 7 women, 8 men; age 41.7 ± 10.0 years

b 7 women, 6 men; age 36.8 ± 6.3 years

Statistical significance: * p  < 0.05, ** p  < 0.01, *** p  < 0.001

No adverse events, injection-site reactions, or treatment-related complications were recorded in either group during the 12-month follow-up. No participant discontinued the study as a result of adverse effects. No serious adverse events occurred. All 15 participants in the active group completed the 12-month treatment course. The absence of observed adverse events may reflect the use of a low, well-tolerated dose (0.75 mg weekly) and the inclusion of patients with prior experience of regular injectable therapy (natalizumab infusions), which may have facilitated adherence and tolerability.

No changes in waist and hip circumference and muscle mass were observed in both groups. Repeated-measures analyses demonstrated significant group × time interactions for BMI ( F 1,26  = 16.3, p  < 0.001), body fat percentage ( F 1,26  = 7.3, p  = 0.012), and visceral fat ( F 1,26  = 8.0, p  = 0.009), indicating improvement in the active group. At baseline, RMR was higher in the active group than in controls ( p  = 0.029); there was no between-group difference at 12 months. Within the active group, RMR decreased over 12 months ( p  = 0.013). In a repeated-measures analysis, there was a significant main effect of time ( F 1,26  = 10.5, p  = 0.004), and a borderline main effect of group ( F 1,26  = 4.3, p  = 0.05), with no significant effect of group × time interaction. Questionnaire-based physical activity did not change significantly.

Glucose AUC decreased significantly over 12 months in the active group ( p  < 0.001) but not in controls. Repeated-measures analysis showed a significant effect of time and group × time interactions ( F 1,26  = 5.1, p  = 0.032), consistent with a larger reduction in the active group. Insulin AUC and HOMA-IR did not change significantly. The Matsuda index showed a non-significant trend toward improvement in the active group ( p  = 0.099), whereas the Cederholm index increased over time in both groups (control, p  = 0.035; active, p  = 0.003), without a group × time interaction. Clinically, patients in both groups remained stable: no relapses occurred, and EDSS did not change significantly during follow-up.

Functional and cognitive measures are shown in Table  2 . For the 9HPT, no significant changes were observed for the dominant hand. For the non-dominant hand, the active group showed a trend toward improvement p  = 0.077), and repeated-measures analysis indicated a significant effect of group × time interaction ( F 1,26  = 4.7, p  = 0.039). For the T25FW, the active group showed a trend toward faster completion times ( P  = 0.065), with significant time and group × time effects in repeated-measures analysis ( F 1,26  = 9.4, p  = 0.005). The effect sizes for T25FW and non-dominant 9HPT were in the small-to-moderate range ( d  = 0.44–0.52). For context, the minimal clinically important difference (MCID) for T25FW is generally considered to be approximately 20% change from baseline [ 20 ]; the median baseline T25FW in the active group was 5.3 s, and the observed improvement of approximately 0.35 s represents approximately 6.6%, which is below the MCID threshold. Similarly, the median baseline 9HPT on non-dominant hand was 25.9 s, and the observed improvement of approximately 0.82 s represents approximately 3.2%. SDMT improved minimally in both groups; repeated-measures analysis indicated a small overall effect of time ( F 1,26  = 5.5, p  = 0.027), with no significant group × time interaction.

Plasma NfL did not change significantly over the 12-month follow-up, and no group × time interaction was observed. MRI measures showed increases in lesion-related volumes over 12 months (FLAIR lesion volume and T1 hypointense lesion volume), with significant time effects (FLAIR ( F 1,26  = 4.7, p  = 0.041); T1 black hole volume ( F 1,26  = 11.4, p  = 0.002)), but no group or group × time effects. We performed a post hoc power analysis for the primary neurodegeneration-related outcomes. For plasma NfL, the observed between-group difference in change from baseline was negligible (dulaglutide, + 0.1 pg/mL; control, − 0.1 pg/mL), with pooled SD of approximately 3.32 pg/mL, yielding an effect size (Cohen’s d ) of approximately 0.06. Achieving only 5.3% power. For brain volume percentile, the observed effect size was similarly small ( d ≈ 0.15). Given the small sample size and variability, MRI findings should be interpreted cautiously.

Discussion

In this 12-month proof-of-concept trial, adjunctive dulaglutide in patients with relapsing–remitting MS (RRMS) was associated with significant improvements in anthropometric and metabolic measures, including reduced weight, BMI, and markers of adiposity, as well as improved glucose tolerance (reduced glucose AUC). The intervention was well tolerated at a weekly dose of 0.75 mg, with no discontinuations due to adverse effects and no serious adverse events.

It is important to distinguish clearly between the metabolic improvements demonstrated in this trial and the concept of true neuroprotection. The reductions in weight, adiposity, and glucose AUC are direct metabolic effects of GLP-1RA therapy consistent with its established pharmacology in type 2 diabetes. Whether these metabolic changes translate into neuroprotective benefit in MS remains speculative and is not supported by the biomarker data in this study. Statements regarding potential neuroprotective mechanisms should therefore be interpreted as hypotheses for future investigation rather than conclusions supported by the present data.

The decrease in resting metabolic rate (RMR) observed in the active group is plausibly explained by concurrent weight loss, given the dependence of resting energy expenditure on body mass and composition. Consistent with this, the absence of a significant group × time interaction suggests that dulaglutide did not exert a specific effect on RMR independent of weight loss.

The reduction in glucose AUC without a concomitant increase in insulin AUC is consistent with improved glucose handling without increased insulin secretion. Potential mechanisms include modest improvement in insulin sensitivity and non-insulin-mediated effects, such as delayed gastric emptying and altered postprandial glucose dynamics. Although clear between-group differences were limited, changes in insulin sensitivity indices were directionally consistent with improvement.

In addition to metabolic effects, exploratory analyses suggested modest improvement in functional performance in the dulaglutide group. The non-dominant hand 9HPT showed a favorable group × time interaction, and the T25FW showed significant time and group × time effects, indicating greater improvement in walking speed over 12 months in the dulaglutide group. These findings require cautious interpretation given the small sample size and open-label design, the restriction of the 9HPT signal to the non-dominant hand, and the susceptibility of performance-based outcomes to practice effects and day-to-day variability (e.g., fatigue, temperature sensitivity, and motivation) and expectation bias in the absence of assessor blinding; and the fact that the observed T25FW and 9HPT improvement did not meet the established minimum clinically important difference threshold. Furthermore, and critically, the lack of parallel changes in biomarkers of neurodegeneration (plasma NfL and MRI volumetry) argues against attributing these functional improvements to neuroprotection. It is more plausible that GLP-1RA-associated weight loss and metabolic changes—such as reduced adiposity, improved energy metabolism, and potentially reduced fatigue—could translate into modest gains in mobility and manual dexterity through non-neuroprotective mechanisms. This hypothesis warrants testing in larger, blinded trials with prespecified functional endpoints and standardized exercise and fatigue monitoring.

Despite these metabolic and functional signals, dulaglutide was not associated with improvement in the prespecified neurodegeneration-related outcomes (plasma NfL and MRI volumetry) over 12 months. Several factors may explain these neutral findings. First, the study was substantially underpowered for neurodegeneration endpoints, as confirmed by post hoc power calculations. Second, the selection of natalizumab-treated patients—while methodologically advantageous for reasons of homogeneity, clinical stability, and mechanistic clarity—may have paradoxically reduced the sensitivity of NfL and MRI outcomes to detect adjunctive neuroprotective effects. Natalizumab is one of the most effective approved therapies for RRMS and substantially lowers NfL toward the normal range in well-controlled patients [ 20 , 21 ]. Baseline NfL values in our cohort were low (active group, 8.1 ± 3.7 pg/mL; control group, 7.6 ± 2.9 pg/mL), and no gadolinium-enhancing lesions were present at baseline in either group, indicating a narrow dynamic range available for further biomarker improvement. This floor/ceiling effect limits the ability to detect any additional neuroprotective benefit from dulaglutide in this specific population. Third, the 0.75-mg weekly dose may have been insufficient to achieve CNS concentrations necessary for neuroprotective effects. Fourth, 12 months may be insufficient to detect changes in slowly evolving structural MRI outcomes, particularly brain volume, which is further confounded by pseudoatrophy effects in the first year [ 22 ].

The choice of natalizumab as the required background therapy was deliberate and scientifically motivated. Natalizumab’s exclusive peripheral mechanism of action—blocking α4-integrin-mediated lymphocyte trafficking across the blood–brain barrier—does not involve direct metabolic effects, minimizing pharmacological confounding when interpreting GLP-1RA-related metabolic outcomes [ 9 , 21 , 23 ]. This is in contrast to other DMTs such as interferons, dimethyl fumarate, or sphingosine-1-phosphate receptor modulators, which have pleiotropic metabolic, antioxidant, or immunomodulatory effects that could confound the interpretation of adjunctive GLP-1RA therapy. Restricting enrollment to a single background DMT also enhanced population homogeneity, which is particularly important in a small proof-of-concept trial. Furthermore, patients on natalizumab continue to experience residual neurodegeneration independent of relapses, representing a clinically relevant therapeutic target for adjunctive neuroprotective strategies. Future studies evaluating GLP-1RA neuroprotection in MS might alternatively consider enrolling patients on moderately effective DMTs with higher residual disease activity and NfL levels, where the dynamic range for biomarker improvement is greater, or specifically targeting patients with progression independent of relapse activity (PIRA), where neurodegeneration continues despite inflammatory control.

The relatively preserved metabolic status of our cohort at baseline—with mean HOMA-IR within the normal range and no participants meeting criteria for type 2 diabetes or metabolic syndrome—may also have limited the magnitude of GLP-1RA-mediated metabolic and potentially neuroprotective effects. GLP-1RAs are known to produce larger metabolic improvements in individuals with higher baseline BMI, insulin resistance, or dysglycemia. Future trials should consider stratifying by baseline metabolic phenotype or specifically recruiting patients with MS and comorbid metabolic syndrome or insulin resistance, in whom GLP-1RA therapy may exert more pronounced effects.

Mechanistically, GLP-1 receptors are expressed in multiple brain regions [ 24 , 25 ], and GLP-1RAs may exert central effects that require CNS access. Dulaglutide is a long-acting GLP-1RA with a half-life of approximately 4–7 days and a molecular weight of 59.7 kDa [ 24 ]. Several GLP-1RAs, including dulaglutide, have been reported to cross the blood–brain barrier, potentially with enhanced transport under conditions of barrier dysfunction [ 26 – 28 ].

In experimental autoimmune encephalomyelitis, GLP-1RAs show anti-inflammatory and neuroprotective effects, potentially via modulation of NF-κB and AMPK pathways, support of oligodendrocyte lineage biology, and pro-myelinating signaling such as PI3K/AKT [ 29 – 37 ]. Although these preclinical data are encouraging, our results did not demonstrate measurable neuroprotection as assessed by plasma NfL and MRI volumetry over 12 months in this cohort.

Several limitations of this study must be acknowledged. First, the small sample size ( n  = 28) resulted in substantial underpowering for neurodegeneration endpoints, as confirmed by post hoc power calculations; the neutral NfL and MRI findings should be interpreted as inconclusive rather than as evidence of absence of biological effect. Second, the open-label design introduces risk of performance bias for functional outcomes, particularly in the absence of assessor blinding; future trials should incorporate blinded assessors and, where feasible, a sham injection control. Third, the 12-month follow-up may be insufficient to detect treatment effects on slowly evolving neurodegeneration biomarkers, which typically require longer observation periods. Fourth, the 0.75-mg weekly dose of dulaglutide may have been insufficient to achieve neuroprotectively relevant CNS concentrations; dose-optimization studies are required. Fifth, the selection of well-controlled natalizumab-treated patients, while methodologically advantageous, created a floor effect for NfL and MRI outcomes that limited biomarker sensitivity. Sixth, no correction for multiple comparisons was applied and repeated-measures analyses were not adjusted for baseline covariates (such as BMI, EDSS, or RMR) given the exploratory nature of the trial. Seventh, the relatively preserved metabolic status of the cohort at baseline may have limited the magnitude of GLP-1RA effects.

Conclusions

Adjunctive dulaglutide in natalizumab-treated RRMS was feasible, well tolerated, produced metabolic improvements, and showed exploratory signals of functional benefit, but did not modify plasma NfL or MRI volumetric measures over 1 year. These results should be interpreted in the context of the study’s limitations, particularly its small sample size, underpowering for neurodegeneration endpoints, and the floor effect created by well-controlled disease on natalizumab. Larger, adequately powered, blinded studies with higher-dose GLP-1RA regimens, longer follow-up, and inclusion of patients with residual metabolic dysfunction or higher neurodegeneration burden are warranted to confirm these functional benefits and fully evaluate the neuroprotective potential of GLP-1RA therapy in MS.

Acknowledgements

We thank the participants of the study.

Author Contributions

Pavel Šiarnik: Conceptualization, Investigation, Writing—original draft, Formal analysis; Žofia Rádiková: Investigation, Writing—review & editing, Data curation, Supervision, Formal analysis; Branislav Kollár: Investigation, Supervision, Project administration; Miroslava Kutšálová: Investigation, Project administration; Adela Penesová: Investigation; Andrea Havranová: Investigation; Richard Imrich: Investigation, Methodology; Miroslav Vlček: Investigation, Project administration; Ingrid Žitňanová: Investigation; Georgi Krastev: Investigation; Peter Turčáni: Writing—review & editing, Conceptualization, Methodology, Supervision.

Funding

Funding for this study and the journal's Rapid Service Feewas provided by the Ministry of Health of the Slovak Republic Grant (Grant No. MZ SR 2018/37-LFUK-11) and the SlovakResearch and Development Agency Grants (APVV; Grant Nos. APVV-15-0228 and APVV-21-0261).

Data Availability

The data presented in this study are available on request from the corresponding author.

Declarations

Conflict of Interest

The authors (Pavel Šiarnik, Žofia Rádiková, Branislav Kollár, Miroslava Kutšálová, Adela Penesová, Andrea Havranová, Richard Imrich, Miroslav Vlček, Ingrid Žitňanová, Georgi Krastev, Peter Turčáni) declare no conflict of interest.

Ethical Approval

The trial was registered in the EU Clinical Trials Register (EudraCT 2019-003001-94) and approved by the Ethics Committee of the Old Town Hospital, University Hospital Bratislava, on 25 September 2019, under reference VV-2018-R-EK. The study was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent after receiving detailed study information and training in the administration of dulaglutide.

Footnotes

Prior Publication Conference presentation: ASENT 2024 Annual Meeting, March 12–14, 2024 at the Hyatt Regency Bethesda in Bethesda, MD, USA. Reference: Imrich, R., et al., Effects of GLP-1 Receptor Agonist on Chronic Axonal Damage, Neurodegeneration, and Cognitive Function in Relapsing-remitting MS (MSGLP): Exploratory, Open Label, Randomized Controlled Trial. Neurotherapeutics, 2024. 21(4): p. e00403. 10.1016/j.neurot.2024.e00403.

References

Associated Data

Data Availability Statement

The data presented in this study are available on request from the corresponding author.