фазы 2 Судебный процесс над Фрексалимабом, антагонистом CD40L, среди подростков и взрослых с недавним-оффсетным типом 1 Диабет (ФАБУЛИНУС): Обоснование и проект исследований.

Introduction

Type 1 diabetes (T1D) is a chronic autoimmune disease, characterised by selective destruction of insulin‐producing pancreatic β‐cells that is mediated by autoreactive T and B‐cell lymphocytes [ 1 ]. Over time, which often may be years, a critical mass of β‐cells is destroyed, and physiological glycaemic control can no longer be supported [ 2 ]. A significant β‐cell mass remains functional at the time of clinical diagnosis of autoimmune T1D, contributing in some cases to improved glucose control over a certain time after diagnosis, known as ‘honeymoon period’. However, the vast majority of β‐cells, if not all, will eventually be destroyed as the disease progresses [ 3 , 4 ]. A diagnosis of T1D is associated with decreased lifespan [ 5 ] and a significant psychological burden, particularly affecting children and their caregivers [ 6 , 7 ]. Therefore, there is an urgent need to develop novel therapeutics that can prevent or slow the rate of β‐cell loss to improve glucose control and long‐term outcomes, and reduce the burden and severity of autoimmune T1D [ 8 , 9 ].

Cluster of differentiation 40 (CD40) is a costimulatory member of the tumour necrosis factor receptor superfamily expressed on both immune and non‐immune cells [ 10 ]. Interactions between B cell‐expressed CD40 and its binding ligand, CD40L (or CD154), predominantly expressed on activated CD4+ T cells, play a critical role in regulating the initiation of adaptive and innate immune responses. CD40/CD40L interaction is required for B‐cell activation/maturation, dendritic cell maturation and T‐cell activation including the T‐cell‐dependent antibody response (TDAR). Disruption of the CD40/CD40L pathway in autoimmune diseases, particularly those where pathogenic B‐cell responses are a key hallmark of disease, would be predicted to impact both cellular and humoral responses and have therapeutic benefit.

Frexalimab (SAR441344) is a second‐generation humanised mouse immunoglobulin (Ig) G1 antagonist monoclonal antibody (mAb), which binds to CD40L, thereby blocking binding to CD40 and disrupting the subsequent CD40/CD40L signalling pathway that is important for activation and function of adaptive (T and B cells) and innate (macrophages/cytokines) immunity. The Fc region of frexalimab is modified to prevent FcγRIIa binding and platelet activation, thereby overcoming the risk of thromboembolic events observed with first‐generation anti‐CD40L mAb therapies [ 11 , 12 ].

Frexalimab has shown promising results in a Phase 2 trial in patients with relapsing multiple sclerosis (MS) [ 13 ] that triggered initiation of two Phase 3 clinical trials in non‐relapsing secondary progressive MS ( NCT06141486 ) and relapsing MS ( NCT06141473 ). In addition, frexalimab is being evaluated in Phase 2 clinical trials: completed in Sjögren's syndrome ( NCT04572841 ) and ongoing in systemic lupus erythematosus ( NCT05039840 ) and primary focal segmental glomerulosclerosis/minimal change disease ( NCT06500702 ).

A Phase I multiple ascending dose study showed dose‐dependent inhibition of TDAR (Sanofi, data on file), suggesting that frexalimab administration has the potential to suppress important mechanisms involved in immune‐mediated pancreatic β‐cell destruction in T1D, and thus to preserve capacity to produce and secrete endogenous insulin in clinically significant quantities. Further recent clinical trials provide evidence that targeting critical autoimmune processes in recently diagnosed T1D patients can significantly decrease the rate of β‐cell loss and improve glycaemic control [ 14 , 15 , 16 , 17 , 18 ].

Existing evidence suggests that in children, adolescents and young adults with Stage 3 T1D, disease progression and the rate of β‐cell loss are more than 2‐fold greater than in adults [ 4 , 19 ] and are associated with pro‐inflammatory autoreactive CD4+ T cell responses [ 20 ], effector CD8+ T cell responses that drive β‐cell loss [ 21 ], a more B lymphocyte‐dominated, hyper‐immune islet pathology [ 22 ], and more autoantibodies, particularly against insulin, playing crucial roles in pathogenesis [ 23 ]. C‐peptide decline is age‐dependent, with the rate being slower with age [ 24 ]. Most clinical trials performed only in adult populations have failed to demonstrate clinically meaningful effects on preservation of β‐cell function [ 25 , 26 ]. However, some studies show significant benefit in terms of β‐cell preservation when a significant proportion of participants are children, adolescents or young adults [ 14 , 16 , 27 ] or when analyses were performed in subgroups of children and adolescents [ 25 , 26 ]. Therefore, considering the relatively slow C‐peptide decline and its high variability, as well as lower incidence rates in adults with T1D [ 5 ], reasonably sized proof‐of‐concept or dose‐finding studies are very challenging, if at all feasible in this population. To undertake clinical studies in a more suitable paediatric population, evidence of safety and potential for efficacy in adult participants should first be demonstrated.

An important consideration in study design is balancing the need to detect signals with the imperative to avoid unnecessarily prolonged or extensive exposure of participants outside the potential target population. In the T1D area, innovation is needed in the conduct of paediatric studies involving drugs that are still in development and not yet approved in other indications. Many trials to date have focused on approved drugs or those more extensively studied in adults. Our aim here is to bridge scientific understanding with safety for the target population, while maintaining a reasonable probability of achieving clinically meaningful results. Historically, it has been challenging to introduce innovative, novel immunomodulatory therapies into paediatric T1D research.

To address these substantial challenges and to ensure sufficient safety monitoring in a paediatric population, this study will initially assess the safety of the highest dose‐level of frexalimab in adult participants (18–35 years) (Part A). In turn, Part B (the main part of the study) will then assess safety and efficacy (including dose–response relationship) in sequentially escalated age‐adjusted doses of frexalimab in adolescents and young adults (12–21 years). Hereafter are provided the methods and analysis planned to meet the trial objectives.

Methods and Analysis

Trial Design

This is a multicentre, double‐blind, randomised, placebo‐controlled trial to assess the safety, efficacy and tolerability of frexalimab in participants with recently diagnosed T1D on insulin therapy. The trial comprises two parts (Figure  1 ). Part A is exploratory and designed to establish the safety of the highest dose level of frexalimab versus placebo in adults (age range 18–35 years) randomised following a 2:1 ratio (frexalimab: placebo). Efficacy, pharmacokinetics (PK) and effects on pharmacodynamic (PD) biomarkers will also be explored.

Trial design. DMC—Data Monitoring Committee, OLE—open‐label extension for up to 2 years.

Part B is the dose‐finding part intended to evaluate the safety and efficacy of 3 age‐adjusted dose‐levels of frexalimab compared with placebo in adolescents and young adults (12–21 years). PK and effects on PD biomarkers will also be assessed in this part, in which participants are sequentially randomised to frexalimab or placebo (2:1 ratio).

Enrolment of participants in Part B will start with the lowest dose level, considering the Data Monitoring Committee (DMC) recommendations following initial review of safety data from Part A. Enrolment of participants in the next dose level in Part B will start immediately after all the participants have been randomised to the previous dose level.

In each part, there is a 3–5‐week screening period, a 104‐week double‐blind treatment period (main period: 52 weeks, blinded extension: 52 weeks), and a post‐treatment safety follow‐up period of 26 weeks. Blinded extension is intended to evaluate long‐term safety and durability of treatment effect on endogenous insulin production and glycaemic control. Overall, the planned treatment duration in this study is 104 weeks but in case of treatment failure (see below for further details), the treatment may be discontinued earlier by the investigator. Upon completion of Part A or B and blinded extension, participants may enrol in an open‐label extension (OLE) to assess long‐term safety and efficacy of frexalimab.

The trial (DRI17476) was registered at the EU trial register 2022‐500531‐36‐00 and ClinicalTrials.gov ( NCT06111586 ) prior to the inclusion of the first patient. Date and protocol version identifier: electronic 5.0, Date: 10 July 2025. The study started screening on 11 December 2023. The estimated primary completion date is April 2027 and estimated study completion date is October 2030.

Patient Population

Eligibility Criteria

Eligible participants are between 12 and 35 years of age, had received a diagnosis of T1D according to American Diabetes Association (ADA) 2023 criteria [ 28 ], initiated exogenous insulin replacement therapy not longer than 90 days prior to the study, and have a random C‐peptide level of at least 0.2 nmol/L at the screening visit. In addition, all participants require positive results on testing for at least one of the following antibodies associated with T1D: anti–glutamic acid decarboxylase 65, anti–islet antigen 2, anti–zinc transporter 8, or anti‐insulin (as long as the anti‐insulin antibody level was obtained before insulin administration or within 10 days after the initiation of exogenous insulin therapy). All participants are required to be up to date with age‐appropriate immunisations prior to randomisation. Patients with a history of other clinically significant diseases, including autoimmune diseases (except stable autoimmune thyroiditis), are excluded. Full inclusion and exclusion criteria are presented in Table  1 .

TABLE 1 Inclusion and exclusion criteria. Inclusion criteria Exclusion criteria Male or female aged 18–35 years (Part A), aged 12–21 years (Part B). Participants can only be enrolled in Part B after review of safety data from the first 18 participants of Part A who complete 12 weeks of treatment. Serious systemic viral, bacterial, or fungal infection, infection requiring hospitalisation or IV antibiotics or significant chronic viral (including history of recurrent or active herpes zoster, acute or active CMV, EBV as determined at screening), bacterial, or fungal infection (e.g., osteomyelitis) 30 days before and during screening. Participants who meet the criteria of T1D according to ADA [ 28 ]. Participants with a history of invasive opportunistic infections, such as, but not limited to histoplasmosis, listeriosis, coccidioidomycosis, candidiasis, pneumocystis jirovecii and aspergillosis, regardless of resolution. Initiated exogenous insulin replacement therapy within 90 days prior to screening visit (Visit 1) at which random C‐peptide will be assessed. Evidence of active or latent tuberculosis (TB) as documented by medical history and examination, chest X‐rays and/or TB testing. Blood testing (e.g., QuantiFERON TB Gold test) is strongly preferred; if not available, any local approved TB test is allowed. Receiving at least one of the following T1D SOC, insulin hormone replacement therapy—one or MDI of basal insulin, prandial insulin and/or premixed insulin, or CSII. Evidence of any clinically significant, severe, or unstable, acute or chronically progressive, uncontrolled infection, medical or surgical condition, or any condition that may affect participant safety in the judgement of the Investigator (including vaccinations which are not updated based on local regulation). Participants must be positive for at least 1 of the following T1D autoantibodies confirmed by medical history and/or obtained at study screening: GAD‐65, IA‐2, ZnT8 or insulin (if obtained not later than 10 days after initiation of exogenous insulin therapy). History of a systemic hypersensitivity reaction or significant allergies, other than localised injection site reaction, to any humanised mAb. Clinically significant multiple or severe drug allergies, intolerance to topical corticosteroids, or severe post‐treatment hypersensitivity reactions (including, but not limited to, erythema multiforme major, linear IgA dermatosis, toxic epidermal necrolysis and exfoliative dermatitis). Have random C‐peptide levels ≥ 0.2 nmol/L at screening. History or current hypogammaglobulinemia. Be vaccinated according to the local vaccination schedule. Any vaccinations should take place at least 28 days prior to randomisation for non‐live vaccines and at least 3 months prior to randomisation for live vaccines. Has other autoimmune diseases (e.g., rheumatoid arthritis [RA], polyarticular juvenile idiopathic arthritis [pJIA], psoriatic arthritis [PsA], ankylosing spondylitis [AS], MS, SLE), except autoimmune thyroiditis with controlled function of thyroid gland and celiac disease (at discretion of investigator). Contraceptive use by men and women consistent with local regulations regarding the methods of contraception for those participating in clinical studies. History, clinical evidence, suspicion or significant risk for thromboembolic events, as well as myocardial infarction, stroke, antiphospholipid syndrome, other prothrombotic disorders (e.g., congenital or inherited deficiency of antithrombin III, protein C, protein S and Factor V Leiden mutations) and/or participants requiring antithrombotic treatment. Diabetes of forms other than autoimmune T1D, including but not limited to genetic forms of diabetes, MODY, LADA, secondary to medications or surgery, T2D by investigator judgement. History of malignancy of any organ system, treated or untreated, within 5 years of screening, regardless of whether there is evidence of local recurrence or metastases. Systemic corticosteroids (duration > 7 days), adrenocorticotropic hormone 1 month prior to screening. Any IV, IM or SC administered biologic treatments, < 3 months or < than 5 half‐lives (whichever is longer), prior to randomisation. Any live (attenuated or viral‐vector) vaccine (including but not limited to varicella zoster, oral polio, nasal influenza, rabies) within 3 months prior to randomisation. Any non‐live (inactivated, mRNA, recombinant, conjugate, toxoid) vaccine administered less than 28 days prior to randomisation. Other medications not compatible or interfering with IMP at discretion of investigator. Any immunosuppressive therapy within 12 weeks prior to randomisation. Course of Thymoglobulin, teplizumab or other immunomodulatory treatments at any time. Any GLP‐1 agonists, SGLT2/1 inhibitor and verapamil within 2 weeks prior to screening. Abnormal laboratory test(s) at screening. Abbreviations: CMV, cytomegalovirus; CSII, continuous subcutaneous insulin infusion; EBV, Epstein–Barr virus; GAD‐65, glutamic acid decarboxylase; GLP‐1, glucagon‐like peptide 1; IA‐2, insulinoma antigen‐2; IMP, investigational medicinal product; LADA, latent autoimmune diabetes of the adult; MDI, multiple daily injection; MODY, maturity‐onset diabetes of the young; MS, multiple sclerosis; SGLT2/1, sodium–glucose co‐transporter‐2 and 1; SLE, systemic lupus erythematosus; T1D, Type 1 diabetes; T2D, Type 2 diabetes; ZnT8, zinc‐transporter 8.

Inclusion and exclusion criteria.

Abbreviations: CMV, cytomegalovirus; CSII, continuous subcutaneous insulin infusion; EBV, Epstein–Barr virus; GAD‐65, glutamic acid decarboxylase; GLP‐1, glucagon‐like peptide 1; IA‐2, insulinoma antigen‐2; IMP, investigational medicinal product; LADA, latent autoimmune diabetes of the adult; MDI, multiple daily injection; MODY, maturity‐onset diabetes of the young; MS, multiple sclerosis; SGLT2/1, sodium–glucose co‐transporter‐2 and 1; SLE, systemic lupus erythematosus; T1D, Type 1 diabetes; T2D, Type 2 diabetes; ZnT8, zinc‐transporter 8.

Recruitment, Consent and Enrolment

Recruitment will be through diabetes clinics at participating centres. Currently, approximately 88 clinical sites across Austria, Belgium, Canada, Denmark, Finland, France, Germany, Hungary, Italy, Poland, Slovenia, Spain, Sweden, United Kingdom and United States are planned. Written informed consent will be obtained for all participants at the first screening visit. For participants under the legal age of majority (adulthood), written assent will be obtained in addition to a written consent from a parent/caregiver. Re‐consent will be requested when participants reach the legal age of majority.

Study Schedule

An overview of the study schedules is provided in Table  S2 .

Randomisation and Blinding

Participants will be centrally randomised at visit 2 using an interactive response technology (IRT) system.

For Part A, participants will be randomised in a 2:1 ratio to receive either high dose frexalimab or matching placebo. For Part B, participants will be randomised in a 2:1 ratio at each dose level opened sequentially: frexalimab dose 1 (low dose) or matching placebo; frexalimab dose 2 (medium dose) or matching placebo, then frexalimab dose 3 (high dose) or matching placebo. In Part B, randomisation will be stratified by age group (12–15 or 16–21 years).

Trial participants and research teams involved in the conduct of the trial will be blinded to the treatment allocation for the duration of the trial. There are no apparent differences between vials of frexalimab and placebo, so they cannot be distinguished by the site staff or participant. Study drug volume (i.e., dose of active drug or matching placebo) is not blinded.

Emergency unblinding will be managed by the IRT randomisation system. Participants whose treatment assignment has been unblinded before completion of trial treatment must be withdrawn from study treatment but should continue to be monitored in the study and attend study visits as per the trial plan.

Study Interventions

There are two treatment groups in Part A (frexalimab solution for injection and matching placebo) and four treatment groups (frexalimab solution low, medium and high doses for injection and matching placebo) in Part B. Frexalimab solution for injection is supplied in single‐use vials containing frexalimab. Placebo is prepared using the same formulation as frexalimab without the active protein and supplied in single‐use vials that are indistinguishable from active treatment. Insulin is used as standard of care background therapy as either multiple daily injections or continuous subcutaneous infusion. Switching between both is at the investigator's discretion.

Frexalimab is administered as intravenous (IV) infusion (loading dose) on Day 1 and as a subcutaneous (SC) injection every 14 ± 3 days from W2 to W102 by qualified and authorised site staff or home nurse. Self‐administration of the SC injections following adequate training is allowed. When participants self‐administer the investigational medicinal product (IMP) at home, compliance will be assessed at each visit by a review of electronic diary record (Section  2.6.5 ) completed by the participant at home and counting returned boxes during the site visits and documented in the source.

Trial Assessments and Procedures

Table  S2 provides an overview of the trial procedures. Further details are provided below for each assessment contributing towards the objectives.

Mixed Meal Tolerance Test

The mixed‐meal tolerance test (MMTT) stimulated C‐peptide is used as the primary outcome measure, as it is considered the gold standard for measuring endogenous insulin production among people with T1D to assess residual β‐cell function [ 29 ]. Participants will undergo a MMTT to assess mean stimulated C‐peptide concentration over 2 h at baseline (D1), at weeks 12, 26, 38, 52, 78, 104 and 130 to allow estimation of change from baseline in C‐peptide mean concentration and the proportion of participants who remain C‐peptide positive (mean concentration ≥ 0.2 nmol/L) over time. Participants who retain C‐peptide above the treatment failure threshold (see details below) at Week 104 will be invited to participate in the OLE, upon their consent. On the days of the MMTT, participants will come to the investigational site in the morning, in a fasting state for at least 8 h (no food or drinks during this time except water), before the scheduled MMTT. The participant will be asked to ingest within 15 min a standardised liquid meal (6 mL/kg, but not more than 360 mL) provided by the trial containing approximately 150 kcal, 6 g of proteins, 5 g of lipids and 20 g of carbohydrates per 100 mL of product. Blood samples for plasma glucose and C‐peptide will be collected at 10 min prior to the meal (−10 min), at the time of ingestion (0 min), and 15, 30, 60, 90, and 120 min after mixed meal consumption. Administration of bolus insulin will be held during MMTT.

Glucose Monitoring

All participants will be provided with a continuous glucose monitoring (CGM) device and a glucometer device for self‐measured plasma glucose (SMPG) measurements as necessary throughout the duration of the study. Participants are expected to wear the CGM for at least 20 days between the screening visit and randomisation, and prior to study visits at weeks 12, 26, 38, 52, 78, 104 and 130 (Table  S2 ) to generate a minimum of 10 days of useable CGM records (useable days being defined as days [00:00–23:59] with ≥ 80% data completeness). The windows considered for the analysis of CGM data will be within 28 days before/after the post‐baseline actual visit date (and from screening visit up to the first IMP intake at baseline).

Glycated Haemoglobin A1c

Glycated haemoglobin A1c (HbA1c) is assayed at baseline, and at every on‐site visit from Week 12. HbA1c values are determined in a central laboratory using an assay certified by the National Glycohemoglobin Standardization Program (NGSP) ( Q
2 , Valencia, CA, USA).

Insulin Dose (Background Therapy)

Total dose and doses of each formulation, as well as the number of injections per day for background insulin therapy (injections of basal, mealtime insulin or continuous subcutaneous insulin infusion) will be documented in the study e‐diary or participant's personal diary.

Participant E‐Diary

Each participant will be given an electronic diary installed on a study smartphone and will be instructed to enter specific data into it. At the end of the study, the information from the subject electronic diaries will be kept as source documentation in the subject files at the investigational sites. All subjects will be required to collect specific information into the subject e‐diary during the trial. The e‐diary includes sections for: IMP volumes and time of injections, insulin doses, any change in concomitant treatment, adverse event (AE) recording, symptomatic hypoglycaemia recording, hyperglycaemia and diabetic ketoacidosis (DKA) recording, local tolerability assessment, electronic versions of participants'/caregivers' questionnaires (reference to Section  2.6.8 ), and pregnancy tests, if applicable.

Safety Assessments

Safety outcomes assessed during the treatment period include hypoglycaemic events, thromboembolic events, infections, occurrence of treatment‐emergent AEs (TEAEs) including local (injection site) and systemic (hypersensitivity) allergic reactions, serious AEs (SAEs), and routine laboratory tests. All symptomatic cases of hypoglycaemia should be confirmed by glucose values (CGM and/or SMBG) and reported in the e‐diary. Hypoglycaemia episodes are categorised based on ADA classifications [ 30 , 31 , 32 , 33 ].

Immunogenicity Assessments

Antibodies to frexalimab will be evaluated in serum samples collected from all participants according to the schedule of activities shown in Table  S2 . A 3‐tiered approach will be employed to assess the immunogenicity of frexalimab when applicable: serum samples will be screened; then confirmed for antibodies binding to frexalimab; and then the titre of confirmed positive samples will be reported. Other analyses may also be performed to further characterise the immunogenicity of frexalimab.

Patient and Parent‐Reported Outcome Measures

Patient and parent‐reported outcome measures for participants and their parent/caregiver will be assessed at baseline, and weeks 12, 26, 52, 78 and 104 by validated questionnaires including quality of life, assessed using the Pediatric Quality of Life (PedsQL) Questionnaire [ 34 ] and Problem Areas in Diabetes (PAID) Questionnaire [ 35 , 36 , 37 ], treatment satisfaction, assessed using the Diabetes Treatment Satisfaction Questionnaire (DTSQ) [ 38 , 39 ].

Objectives and Endpoints

The primary objective of the trial is to demonstrate the efficacy of different doses of frexalimab in comparison with placebo and in addition to standard of care on endogenous insulin secretion in participants (12–21 years; Part B) with recently diagnosed T1D over a 52‐week period (Table  S1 ). The secondary objective is to evaluate the efficacy of frexalimab in comparison to placebo and on top of standard of care on glycaemic control and endogenous insulin secretion in participants (12–21 years; Part B), to evaluate safety and tolerability of frexalimab versus placebo, to characterise PK, to evaluate potential immunogenicity and to assess the impact of frexalimab treatment on caregiver and/or patient‐reported outcomes in all participants.

Primary Endpoint

The primary efficacy endpoint is the change from baseline to Week 52 in mean 2‐h MMTT stimulated C‐peptide concentration, calculated from the area under the concentration‐time curve [AUC] (i.e., AUC divided by the duration of the test: 120 min) (Table  S3 ). This endpoint is used for determination of the dose–response relationship in these participants.

Secondary Endpoints

Parameters characterising glycaemic control and diabetes progression, including CGM glucose time in range (TIR, 70–180 mg/dL), time in tight range (TITR, 70–140 mg/dL), changes in C‐peptide levels, insulin dose, HbA1c and insulin dose adjusted HbA1c (IDAAC) [ 40 ] will be evaluated throughout the study as secondary endpoints (Table  S1 ).

Secondary safety and tolerability endpoints assessed during the 12‐month treatment period include hypoglycaemic events, occurrence of TEAEs including local (injection site) and systemic (hypersensitivity) allergic reactions, treatment‐emergent SAEs, change in body weight and routine laboratory tests. The potential for immunogenicity is assessed by the presence of anti‐drug antibodies (ADAs). Additional secondary endpoints include the PK of frexalimab and the impact of frexalimab treatment on caregiver‐ and/or patient‐reported clinical outcome measurements (see Table  S1 for a complete list and Table  S4 : SPIRIT 2013 Checklist).

Statistics and Data Analysis

General Considerations

For Part B, the primary dose–response analysis using the Multiple Comparison Procedure—Modelling (MCP‐Mod) approach (Step 1) will control the Type I error rate at a two‐sided significance level of 0.05. No formal hypothesis testing is planned for Part A, which is exploratory in nature. Analyses of secondary endpoints will also be considered exploratory and interpreted descriptively. All efficacy analyses will be performed on the intent‐to‐treat (ITT) population, which includes all randomised participants irrespective of compliance with the study protocol and procedures. The per‐protocol population may be used for primary and secondary efficacy analyses to evaluate the influence of major protocol violators and as a sensitivity check of the ITT analysis. All safety analyses will be performed on the safety population, comprising all randomised participants who received at least one dose of study treatment, analysed according to the treatment received.

Unless otherwise specified, analyses will be performed by each dose level of frexalimab and placebo (and overall, for baseline and demographics characteristics) in Parts A and B separately.

Sample Size Considerations

Approximately 192 participants will be randomised to study interventions in Part A ( n  = 30) and Part B ( n  = 162). There is no formal hypothesis testing in Part A, which is an exploratory initial assessment of safety in adults. In Part B, the sample‐size calculation is based on the number of participants who would need to be enrolled to detect a dose–response signal with respect to the primary endpoint [ 41 ].

In Part B, participants will be allocated at unequal numbers to respective frexalimab age‐adjusted dose‐levels (low, medium, high) and or matching placebo (2:1 ratio of frexalimab to placebo) including expected approximately 15% of dropouts. An average of > 90% power will be achieved to detect a dose–response signal via the MCP‐Mod method [ 42 ], considering assumptions based on the findings from the T1GER study [ NCT02846545 ] comparing golimumab with placebo [ 16 ]. The power will be ~73% if the analysis is performed on age group 12–15 or 16–21 of equal size separately.

The sample size calculation in Part B is based on a multiple contrast test at two‐sided significance level of 0.05 for absence of any dose–response among 7 pre‐assumed dose response curves (an E
max model, a sigmoid E
max model, a linear model, two logistic models, a Beta model and an exponential model) over the pre‐specified dose range (from placebo, low to high dose of frexalimab). The R package Dose Finding (version 1.0‐2) from the Comprehensive R Archive Network (CRAN) was used to perform the calculations.

Statistical Analysis of Primary Endpoint

The primary estimand will be the treatment difference between frexalimab and placebo in the change from baseline to 52 weeks in mean 2 h MMTT stimulated C‐peptide concentration, calculated from AUC (i.e., AUC divided by the duration of the test: 120 min). The AUC is calculated using the trapezoidal rule and is considered estimable only if at least one baseline sample (Time (T)−10 or T0 min) is available and no more than one post‐baseline time point (among T15, T30, T60, T90 and T120 min) is missing. Due to skewed C‐peptide AUC data, a logarithmic transformation (log[AUC + 1]) will be performed to normalise its distribution before further analysis of the primary endpoint [ 41 ]. The log‐transformed AUC values (log[AUC + 1]) will be analysed using a mixed model for repeated measures including fixed effects for treatment group, visit, treatment‐by‐visit interaction, age stratification factor and baseline log(AUC + 1) as a covariate. Intercurrent events will be handled according to pre‐specified strategies. Treatment discontinuation due to treatment failure, or the initiation of prohibited medication will be addressed using a treatment‐policy strategy, whereby all available data up to Week 52 will be included in the primary analysis regardless of treatment discontinuation. Treatment discontinuation due to a global pandemic (including COVID‐19) will be addressed using a hypothetical strategy, with post‐treatment discontinuation data set to missing. The dose–response relationship in Part B will be evaluated using a 2‐step MCP‐Mod approach [ 42 ]. The first step is to apply a trend test to detect an efficacy signal (compared to the null hypothesis of a flat, none dose–response curve) from 7 pre‐specified candidate dose–response curves with family‐wise control of Type I error at two‐sided significance level of 0.05. The candidate models and their shape parameters are pre‐specified to represent a range of plausible dose–response relationships. Placebo participants will be pooled across cohorts for this analysis. If a statistically significant signal is present, the second step will be estimation of the dose–response model and dose selection based primarily on the estimated dose–response relationship (considering the minimum effective dose), while also considering clinical relevance and the overall safety and pharmacokinetic profiles. Missing data will be handled using a likelihood‐based mixed model for repeated measures (MMRM), which inherently accommodates missing data under the assumption of Missing At Random (MAR) and uses all available post‐baseline data up to Week 52. A sensitivity analysis exploring departures from the MAR assumption (i.e., considering Missing Not At Random approach) is planned.

Statistical Analysis of Secondary Endpoints

For the secondary endpoints, continuous efficacy endpoints with repeated measurement will be analysed by using a MMRM to test the difference between treatment by each dose level of frexalimab and placebo. The MMRM will use all observed data from scheduled visits during treatment with terms for treatment, visit, treatment by visit interaction, randomisation stratification factor (when applicable). Baseline value, baseline by visit interaction and/or other covariates may be added if applicable. Binary efficacy endpoints will be analysed by using a logistic regression in general, with treatment and randomisation stratification factor (when applicable) as fixed effects and applicable covariates.

Procedure to Account for Missing or Spurious Data

Every attempt will be made to minimise missing data, encouraging participants to provide Week 52 data even if they are no longer taking the interventional medication. Patterns and levels of missing data will be examined. No imputation will be performed for missing data in the primary analysis.

Monitoring

The Sponsor or its authorised representative contract research organisation (CRO) is responsible for the trial management and monitoring of this study, including controlling compliance with the protocol and reviewing the study progress and all AEs. The Sponsor reserves the option to terminate the study at any time.

Data Monitoring Committee

A DMC of three members (two of whom are experts in diabetes and one statistician), independent from the study team and the investigators, will make appropriate recommendations on the conduct of the clinical trial for ensuring the protection and the safety of the enrolled participants in the study (i.e., benefit/risk evaluation). The DMC will review and analyse, on a regular basis and throughout the study, unblinded safety data (plus efficacy data when deemed appropriate) provided by an unblinded statistical group working independently to the study team. A detailed charter outlines the activities of the DMC. DMC will review the safety data from Part A and will make recommendations about starting enrolment of adolescents and young adults (12–21 years) in Part B with the lowest age‐adjusted dose level.

Treatment Failures

The C‐peptide concentration of 0.2 nmol/L and above is acknowledged to reflect a clinically relevant level of β‐cell function [ 3 , 43 ]. When mean 2‐h C‐peptide concentration after MMTT decreases to the levels of 0.1 nmol/L (50% of 0.2 nmol/L) or lower, the potential benefit of treatment becomes unlikely as the vast majority of β‐cells are already lost. Therefore, a treatment failure for this study is defined as mean 2‐h MMTT stimulated C‐peptide concentration of 0.1 nmol/L or lower, confirmed by two consecutive on‐site visits where C‐peptide is evaluated. In cases of treatment failure, participants will discontinue injections of either frexalimab or placebo. The participant and/or his/her caregivers will be informed and instructed as appropriate, and injections will be discontinued not later than during the next on‐site visit after the results are available for the investigator. The participant will continue to participate in the study and complete the scheduled visits.

Ethics and Dissemination

This study is being conducted in accordance with the International Conference on Harmonisation Guidelines for Good Clinical Practice and the Declaration of Helsinki of 1964 and its later amendments. The protocol has been approved by an independent ethics committee or institutional review board as applicable according to local regulations; written informed consent will be obtained from each patient before any trial‐related activities. The study results will be disseminated by peer‐reviewed publications and conference presentations.

Patient and Public Involvement

Representatives of the patient community from European countries were consulted on the study design to assess the feasibility of the study and collect advice aiming to optimise the study. The Advisory Board gathered adults as well as young adults and parents of minor people diagnosed for T1D. The Sponsor provided an overview of the clinical study design to obtain feedback concerning the study design, endpoints, eligibility criteria, study visits, endpoints and the recruitment strategy. Based on feedback from these panels, all suggestions were given careful consideration.

Procedures and frequency of assessments were consequently adapted wherever possible. The MMTT has been limited to a 2‐h procedure on Day 1, followed by testing every 3 months during the first year of study treatment period, then every 6 months up to the end of the study. In addition, study participants will be offered support for injecting the study drug, which may be at the site or remotely. Furthermore, the treatment failure process will allow avoiding unnecessary exposure to IMP when it happens to not be effective for these participants, ensuring ethical conduct of clinical trial of extended duration (104‐week double‐blind treatment period).

Study Management

Adverse Events

Data on AEs will be collected during the trial and coded using Medical Dictionary for Regulatory Activities (MedDRA).

Data Management

All study participants will be assigned a study identification number. Data will then be entered into an electronic case record form (eCRF) and stored in a password‐protected electronic database. The reason for discontinuation of study treatment must be recorded in the eCRFs. Participants who prematurely discontinue treatment are encouraged to continue with regular visits on‐site or by phone. If possible, the on‐site early treatment termination visit should be completed at the time of treatment discontinuation.

Discussion

This is a combined proof‐of‐concept and dose‐finding study that will evaluate the safety and efficacy of frexalimab on preservation of endogenous insulin secretion, as measured by mean 2‐h C‐peptide concentration after MMTT, derived from AUC, in participants with recently diagnosed T1D on insulin therapy, versus placebo. It is hypothesised that C‐peptide in frexalimab treated participants will decline slower than in the placebo group and will remain close to baseline levels or slightly decrease over time [ 14 , 15 , 16 , 17 , 18 ]. Therefore, in order to demonstrate a difference from the placebo group, a diminish in the decline of C‐peptide in the group receiving the IMP is required. The evidence from the literature shows that this process often takes place relatively slowly [ 4 , 19 ]. Thus, a sufficient duration of study is needed to demonstrate differences between frexalimab and placebo, thereby requiring 52 weeks of treatment for the primary endpoint analysis [ 44 ]. According to the natural course of disease, it is expected that any differences between frexalimab and placebo in terms of change from baseline in C‐peptide and glycaemic control will increase over time. In the blinded 52‐week extension, the durability of the effect of frexalimab on C‐peptide preservation and glycaemic variables, as well as long‐term safety, will be assessed.

Considering the small natural loss of C‐peptide each year in adults with T1D, a long period of time may be required to demonstrate a significant difference, if one exists, between placebo and active drug in this age group. At the same time, inclusion of children younger than 12 years in clinical trials is justified only after positive results are obtained in older populations. Therefore, considering the age‐dependent disease progression rate, high incidence and poor glycaemic control (representing highest unmet need) in this population, adolescents and young adults (12–21 years) comprise the most relevant population for generation of proof‐of‐concept and dose‐finding. As this will be the first study of frexalimab in a paediatric population, additional safety measures are implemented in the study design. First, paediatric participants will only be included in Part B after initial safety data in adults are generated and evaluated by a DMC. Second, participants will start enrolment in Part B using a low dose (or age‐adjusted equivalent) and dose escalation will be implemented only if no safety concerns are identified by the DMC during regular reviews. Third, implementation of treatment failure, as described above, will ensure that participants with clinically insignificant remaining C‐peptide level can discontinue the IMP as likelihood of benefit from the IMP will be minimal at this stage of the disease.

Combining proof‐of‐concept with dose‐finding in the study design allows a safe approach to opening cohorts that will include paediatric participants. Sequential dose escalation will ensure that adolescents are enrolled in a high‐dose cohort only when significant safety data are generated and reviewed (Part A and low and middle dose cohorts in Part B). This 2‐studies‐in‐1 approach will optimise the number of participants and overall development time, making frexalimab Phase 3 ready should the study indicate a positive outcome in terms of efficacy. Moreover, such a positive outcome in recently diagnosed Stage 3 T1D could trigger efforts to develop frexalimab in pre‐symptomatic Stage 2 T1D. Further, the MCP‐mode analysis allows further optimisation of the number of participants required in the dose‐finding part. Accurate and robust dose‐finding is critical for development of assets aimed to be used continuously in paediatric participants to ensure optimal benefit/risk balance.

A limitation of this study is the restriction of the lower age range of participants to 12 years old and above in the main (efficacy) part of the study (Part B). Expanding the study population to 6–11 years old participants would substantially increase the robustness and wider applicability of the eventual findings, but at the expense of delayed timelines due to the need for further age step‐downs.

It is important to mention that stimulated C‐peptide, which is used in virtually all Phase 2 clinical trials in recently diagnosed Stage 3 T1D as the primary endpoint, has not been formally qualified as a validated surrogate endpoint by health authorities and supportive data on glycaemic parameters is needed for drug approval [ 44 ]. Further steps and data from prospective studies are needed before it will be accepted as a primary endpoint in registration (Phase 3) trials.

Conclusions

The FABULINUS study design enables the integration of proof‐of‐concept and dose‐finding phases in a predominantly paediatric population. It ensures safety by initiating enrolment in a low‐dose cohort, allowing for the gradual accumulation of safety and efficacy data before progressing to a high‐dose cohort. Given the absence of approved immunomodulatory treatments for Stage 3 T1D, innovative study designs like FABULINUS are essential to accelerate the development and delivery of new therapies to individuals with T1D.

Contributors

A.C., B.R., N.P., M.P. and G.B. conceptualised and designed the study. F.R. and M.J.H. provided expert consultation on study design, endpoints, protocol and conduct. A.C., B.R., M.C., P.L., M.B.‐D., M.‐L.N.‐P., G.B. and M.P. drafted the paper. All authors critically revised the paper. All authors have read and agreed to the published version of the manuscript and to be accountable for all aspects of the work.

Funding

This work was supported by Sanofi.

Disclosure

Patients and/or the public were involved in the design, conduct, reporting, or dissemination plans of this research. Refer to Section  6 for further details.

Consent

The authors have nothing to report.

Conflicts of Interest

A.C., B.R., M.C., P.L., M.B.‐D., M.P., M.‐L.N.‐P. and G.B. are employees of Sanofi. F.R., Advisory board: Boehringer Ingelheim, Speaker fees: Kyowa Kirin, Novo Nordisk, Sanofi. M.J.H., Consultant: Sanofi, MannKind Corporation. Advisory Panel: SAB Biotherapeutics Inc.

Supporting information

Table S1: Objectives and trial endpoints.

Table S2: Schedule of activities.

Table S3: Evaluation of the mean 2‐h MMTT simulated C‐peptide concentration, calculated from AUC.

Table S4: SPIRIT 2013 checklist.