Safety and Efficacy of Switching SAR341402 Insulin Aspart and Originator Insulin Aspart vs Continuous Use of Originator Insulin Aspart in Adults With Type 1 Diabetes: The GEMELLI X Trial

Abstract

Background:

SAR341402 insulin aspart (SAR-Asp) is a rapid-acting insulin analog developed as an interchangeable biosimilar to the marketed insulin aspart reference product (NovoLog; NN-Asp). GEMELLI X was a randomized controlled trial to assess outcomes with a biosimilar in line with the US Food and Drug Administration requirements for designation as an interchangeable biosimilar. This report assessed whether multiple switches between SAR-Asp and NN-Asp lead to equivalent safety and efficacy compared with continuous use of NN-Asp in adults with type 1 diabetes (T1D) treated with multiple daily injections, using once-daily insulin glargine U100 (Lantus) as the basal insulin.

Methods:

This open-label randomized (1:1), parallel-group, phase 3 trial compared four × four weeks of alternating use of individually titrated SAR-Asp and NN-Asp (NN-Asp for first four weeks, SAR-Asp in last four weeks; switching group) vs 16 weeks of continuous use of NN-Asp (nonswitching group). End points included pharmacokinetics, immunogenicity, adverse events, hypoglycemia, insulin dose, and change in efficacy parameters.

Results:

Of the 210 patients randomized, 200 (95.5%) completed the trial. Patients assigned to switching group (n = 104) and nonswitching group (n = 106) showed similar safety and tolerability, including anti-insulin aspart antibody responses, adverse events, and hypoglycemia. At week 16, there was no relevant difference between switching vs nonswitching groups in the change from baseline in glycated hemoglobin (least square [LS] mean difference = 0.05% [95% confidence interval [CI] = −0.13, 0.22]; 0.50 mmol/mol [−1.40, 2.39]), fasting plasma glucose (LS mean difference = 0.23 mmol/L [95% CI = −1.08, 1.53]; 4.12 mg/dL [−19.38, 27.62]), and changes in insulin dosages.

Conclusions:

Alternating doses of SAR-Asp and NN-Asp compared with continuous use of NN-Asp showed similar safety, immunogenicity, and clinical efficacy in adults with T1D. This study supports interchangeability between SAR-Asp and NN-Asp in T1D management.

Keywords

Introduction

The cost of diabetes care continues to rise, primarily due to the increased number of diagnosed patients. The price of insulin remains a concern for many patients, with insulin manufacturers taking steps to reduce the price of insulin products.¹ The availability of biosimilar insulin products may help to further lower costs by providing equally efficacious alternatives that may also reduce treatment costs.² In 2021, the World Health Organization prioritized access to insulin analogs and their biosimilars in their list of essential medicines, thereby increasing access to diabetes treatments by expanding the choice of treatment.³

SAR341402 insulin aspart (SAR-Asp) is a rapid-acting insulin analog developed as a proposed biosimilar to the marketed insulin aspart product (NovoLog/NovoRapid, Novo Nordisk, Bagsværd, Denmark).^4,5 Sanofi’s insulin aspart has the same amino acid sequence and structure as insulin aspart in NovoRapid/NovoLog. With any biosimilar, subtle differences may exist among protein products manufactured in living cells via different processes that can result in different clinical effects. The clinical development program of SAR-Asp, including physicochemical analyses, nonclinical, clinical phase 1 and phase 3 studies, was designed to demonstrate that SAR-Asp and NovoRapid/NovoLog are highly similar. SAR341402 insulin aspart has been shown to have similar pharmacokinetic exposure and pharmacodynamic activity to the reference insulin aspart product sourced from the European Union (EU) (NovoRapid),⁶ the United States (NovoLog),⁶ and Japan (NovoRapid).⁷ A phase 3 trial (GEMELLI 1) showed equivalent efficacy and similar safety/immunogenicity profiles for SAR-Asp and NovoLog/NovoRapid in people with type 1 or type 2 diabetes treated for 6 and 12 months.^8,9 SAR341402 insulin aspart and NovoLog (NN-Asp) were also well tolerated in insulin pump users with T1D treated for four weeks.¹⁰ SAR341402 insulin aspart has received regulatory approval in the EU and other countries as a biosimilar to NovoRapid.

Despite being highly similar to the reference insulin with regards to safety and efficacy, a biosimilar in the Unites States cannot currently be substituted for the reference product at the pharmacy without consulting the prescriber. However, the pharmacist may substitute the product without having to notify the prescribing physician if the biosimilar is approved as an interchangeable biosimilar. This interchangeability designation requires additional evidence to show that the biosimilar product, when administered in an alternating sequence with the reference product, results in the same clinical outcome (immunogenicity, safety, efficacy) as the reference product without alternation or switch.¹¹

Using draft guidance issued by the US Food and Drug Administration (FDA) in 2017¹² (subsequently approved in May 2019)¹¹ about demonstrating interchangeability of a biosimilar with its reference product (originator), we designed this switching study to evaluate efficacy, safety, and immunogenicity of interchangeability between SAR-Asp and originator NovoLog.

Methods

Research Design

This multicenter, open-label, randomized, parallel-group, phase 3 trial (ClinicalTrials.gov identifier: NCT03874715) compared alternate use of SAR-Asp and NN-Asp (switching group) vs continuous use of NovoLog (nonswitching group) in adults with type 1 diabetes (T1D) treated with multiple daily injections (MDIs) using once-daily insulin glargine U100 (Lantus; Gla-100) as the basal insulin.

The study included a two week screening period, a 12 week run-in period for some participants not meeting inclusion criteria regarding pre-treatment, a 16 week treatment period, and a two week post-treatment follow-up period (Supplementary Figure 1). At the end of week 16, participants underwent pharmacokinetic assessment after administration of single doses (0.15 U/kg) of SAR-Asp in the switching group or NN-Asp in the nonswitching group. Results of this pharmacokinetic assessment are reported in a separate article.¹³

The study was conducted at 31 United States sites that screened at least one participant between March 11, 2019 and July 8, 2020, and the protocol was approved by local or central ethics committees. During the last few months of the study, contingency measures were put in place for participant safety, retention, and data capture with the ongoing coronavirus disease 2019 (COVID-19) pandemic (Supplementary Data).

Participants

Adults (≥18 years) with T1D ≥12 months and glycated hemoglobin (HbA_1c) of ≤10% (86 mmol/mol), a body mass index (BMI) ≤35 kg/m², and using an MDI insulin regimen (≥3 times daily) with NovoLog as mealtime insulin and insulin glargine 100 U/mL (Gla-100) as basal insulin for at least 12 weeks were included in this study. Participants receiving twice-daily Gla-100 before the study were switched to once-daily Gla-100. Key exclusion criteria included use of glucose lowering treatments other than MDI including use of insulin pump therapy in past three months.

Procedures and Assessments

Eligible participants were randomized 1:1 to four × four week periods of alternating use of SAR-Asp and NN-Asp (switching group) or 16 weeks of continuous use of NN-Asp (nonswitching group) (Supplementary Figure 1). Randomization was stratified by HbA_1c at screening (<8.0%, ≥8.0%). In the switching group, NN-Asp was initially administered for four weeks, followed by SAR-Asp for four weeks, then NN-Asp for four weeks and ending with SAR-Asp for the last four weeks, for a total of four × four week treatment periods with three switches. The recommended starting dose of NN-Asp after randomization in both groups was a unit-to-unit (1:1) conversion from the NovoLog dose used at the end of the screening or run-in period. NovoLog insulin aspart was supplied as a 100 U/mL insulin solution for subcutaneous (SC) injection in 3 mL NN-Asp FlexPen disposable prefilled pens. SAR341402 insulin aspart was supplied in a 3 mL pre-filled disposable SAR-Asp SoloSTAR pen injector at a concentration of 100 U/mL for SC injection. The starting dose of Gla-100 was the same as the last dose of insulin glargine (100 U/mL) used at the end of the screening or run-in period or was adjusted based on the investigator’s clinical recommendation. Gla-100 was injected once daily at a consistent time (daily time of administration agreed between the participant and the study doctor) using the SoloSTAR pen.

Participants were instructed to measure their blood glucose every day before breakfast, lunch, main evening meal, bedtime, and during suspected hypoglycemic episodes during the trial. Mealtime and basal insulins were titrated based on self-monitored plasma glucose (SMPG) to achieve protocol-specified glycemic targets (preprandial 80-130 mg/dL and 2 hours postprandial <180 mg/dL), while avoiding hypoglycemia.¹⁴ Glycemic targets could be adapted for individual participants, if deemed necessary.

Blood samples for immunogenicity assessments were collected at least 8 hours after the last SAR-Asp or NN-Asp dose at baseline and weeks 8, 12, and 16, and at any early discontinuation visit. The immunogenicity of SAR-Asp and NN-Asp was assessed using a three-tiered approach with screening-, confirmatory-, and neutralizing assays. The presence of anti-insulin aspart antibodies (AIAs) was determined in a central laboratory (Farmovs, Bloemfontein, South Africa) using a radioimmunoassay validated in accordance with recent recommendations.^15-17 Assessments were blinded to the treatment group and included AIA status (positive or negative), AIA titers, cross-reactivity to human insulin, and neutralizing capacity of confirmed positive AIAs (including status and titers of neutralizing anti-insulin aspart antibodies [NAbs]) at each sampling visit. Neutralizing anti-insulin aspart antibodies were analyzed in a central laboratory (Sanofi, Frankfurt, Germany) using a competitive ligand binding assay for confirmed positive AIA samples.^16,17 Anti-insulin aspart antibody and NAb titers were defined as the reciprocal of the highest dilution that yields a positive result, eg, dilution of 1/100 = titer of 100.¹⁵ Potential hypersensitivity or hypersensitivity-like events and events of potential loss of efficacy were adjudicated by an Allergic Reaction Assessment Committee of experts, who were independent from the sponsor and the investigators and blinded to the study treatment.

Hypoglycemic events were documented by the participants in a diary. Additional SMPG measurements were performed to document hypoglycemia in case a patient experienced symptoms that were suggestive for hypoglycemia. Glycated hemoglobin and fasting plasma glucose (FPG) values were determined in a central laboratory (Central Laboratory Services LP, Indianapolis, Indiana) blinded to the treatment group, with samples collected at screening (HbA_1c only), randomization (day 1), and weeks 8, 12 (FPG only), and 16, and at early discontinuation.

In this article, we report clinical end points that include efficacy, safety, and immunogenicity between switching and nonswitching groups. The safety end points included hypoglycemic events, treatment-emergent adverse events (TEAEs), serious TEAEs, and AEs requiring special monitoring (injection site and hypersensitivity reactions). Hypoglycemia episodes were categorized based on the American Diabetes Association classifications.^18-20 Treatment-emergent adverse events were defined as AEs that developed, worsened, or became serious during the 16 week on-treatment period (defined as the time from the first injection of study medication up to the last injection + 1 day). Immunogenicity end points included AIA and NAb response from baseline to week 16. Clinical efficacy end points included the change from baseline to week 16 in insulin dose, HbA_1c, and FPG. Supplementary Table 1 provides further details of the trial end points.

Data Analysis and Statistics

Efficacy end points were analyzed in the intent-to-treat population (all randomized participants), whereas safety end points were analyzed using the safety population (all randomized participants who received at least one dose of study medication). Immunogenicity was analyzed in the AIA population that included all participants from the safety population with at least one AIA sample available for analysis. There was no formal sample size or power analysis for safety, efficacy, or immunogenicity end points as these are secondary or tertiary/exploratory end points. All statistical analyses were descriptive. The primary end point of the study was area under plasma concentration curve and maximum plasma concentration of SAR-Asp vs NN-Asp after a single dose (0.15 U/kg) of SAR-Asp in the switching group or NN-Asp in the nonswitching group at week 16. This pharmacokinetic data from the same study are reported in a separate publication.¹³

As a post hoc efficacy analysis and for descriptive purpose, the difference between treatment groups in the adjusted least square (LS) mean change from baseline to week 16 in HbA_1c and FPG were calculated with their associated two-sided 95% confidence interval (CI) from an analysis of covariance (ANCOVA) model with fixed categorical effect for treatment group and the randomization strata (only for FPG), and the end point baseline value as fixed covariate. All HbA_1c and FPG values collected during the 16 week period were considered, regardless of adherence to treatment. Missing data were imputed using a return-to-baseline multiple imputation approach (missing values at week 16 imputed as equal to the participant’s baseline value plus an error). Results were combined using the Rubin formula.²¹

The difference between treatment groups in the percentage of participants with treatment-emergent AIAs (those with newly positive post-baseline or who developed a fourfold or greater increase in AIA titer compared with baseline) was calculated with its associated two-sided 90% CI based on a binomial regression model with an identity-link function and fixed categorical effects for treatment group and randomization strata. A subgroup analysis evaluated the potential impact of treatment-emergent AIAs on HbA_1c, insulin dose, and safety (hypoglycemia, hypersensitivity reaction, injection site reaction) using descriptive statistics.

Data and Resource Availability

The data sets generated during and/or analyzed during the current trial are available from the corresponding author on reasonable request.

Results

Baseline Characteristics

Of 279 participants screened, 210 were randomized (104 switching group, 106 nonswitching group) and received at least one dose of study medication (Supplementary Figure 2). Five participants randomized to the switching group discontinued study medication during the first treatment period before switching to SAR-Asp and were analyzed in the nonswitching group for the AIA and safety evaluations. In total, 91.3% (95/104) of participants in the switching group and 99.1% (105/106) in the nonswitching group completed the 16 week treatment period. No participants were considered as having trial impact (disruption) due to COVID-19 (Supplementary Data).

Demographics and baseline characteristics were similar in the two treatment groups (Table 1). Participants had a mean age of 44.1 years, were predominantly white (86.2%), and had a mean duration of diabetes of 21.1 years with mean BMI of 26.7 kg/m².

Table 1.

Baseline Demographics and Clinical Characteristics (Randomized Population).

Characteristic	Switching group (N = 104)	Nonswitching group (N = 106)	All participants (N = 210)
Age, years	43.8 ± 16.2	44.3 ± 15.8	44.1 ± 16.0
Male, n (%)	60 (57.7)	66 (62.3)	126 (60.0)
Race, n (%)
White/Caucasian	91 (87.5)	90 (84.9)	181 (86.2)
African American	8 (7.7)	11 (10.4)	19 (9.0)
Asian	3 (2.9)	2 (1.9)	5 (2.4)
Other^a	2 (2.0)	3 (2.8)	5 (2.4)
Hispanic or Latino, n (%)	20 (19.2)	14 (13.2)	34 (16.2)
BMI, kg/m²	26.8 ± 4.2	26.6 ± 4.3	26.7 ± 4.2
Diabetes duration, years	22.4 ± 15.1	19.9 ± 13.2	21.1 ± 14.2
HbA_1c, %	7.70 ± 1.05	7.69 ± 0.96	7.70 ± 1.01
HbA_1c, mmol/mol	60.67 ± 11.52	60.58 ± 10.50	60.62 ± 10.99

Data are mean ± SD unless otherwise indicated.

Includes American Indian or Alaska native, multiple, or not reported.

Efficacy (Glycated Hemoglobin and Fasting Plasma Glucose) and Insulin Doses

There was no relevant difference between switching vs nonswitching groups in the change from baseline to week 16 in HbA_1c (LS mean difference = 0.05% [95% CI = −0.13, 0.22]; 0.50 mmol/mol [−1.40, 2.39]) (Figure 1a and b) or FPG (LS mean difference = 0.23 mmol/L [95% CI = −1.08, 1.53]; 4.12 mg/dL [−19.38, 27.62]) (Supplementary Figure 3). In addition, no clinically relevant changes in basal and mealtime insulin doses were observed in both groups over the 16 week treatment period (Figure 1c and d). Changes in total insulin doses were also small, with a mean increase from baseline to week 16 of 0.018 U/kg/day in the switching group and a mean decrease of −0.004 U/kg/day in the nonswitching group. Changes in prandial insulin doses were also small in the two groups, with a mean increase from baseline to week 16 of 0.008 U/kg in the switching group and a mean decrease of −0.012 U/kg in the nonswitching group.

Figure 1.

HbA_1c over 16 weeks (a) and least squares mean change from baseline to week 16 in HbA_1c (b) (intent-to-treat population), daily basal and mealtime insulin doses by study visit (c-d) and change in daily insulin doses from baseline to week 16 (d) (safety population). (b) Missing data in the HbA_1c at week 16 were imputed using a return-to-baseline multiple imputation approach (10 000 imputations as equal to the participant’s baseline value plus an error). This was followed by an analysis of covariance (ANCOVA) with fixed categorical effect for treatment group, and the end point baseline value as fixed covariate. Results were combined using the Rubin formulae.²¹ Baseline insulin doses were defined as the median of daily doses available in the week before the first injection of study medication. Insulin doses at day 1 were defined as the median of daily doses available in the week after the first injection of study medication. For weeks 4, 8, 12, and 16, insulin dose values were reported as the median of daily doses available in the week before the study visit. BL, baseline; D: day; LS, least squares; MT, mealtime; W, week.

Hypoglycemia

There were no differences in documented symptomatic or asymptomatic hypoglycemia between switching and nonswitching group; however, the rate of severe hypoglycemia per participant-year of exposure was numerically higher in the switching group, whereas rates were similar between groups for other hypoglycemia categories (Supplementary Figure 4 and Supplementary Table 2). Overall, no clinically meaningful difference was observed in the hypoglycemia event rates.

Adverse Events

The two insulin aspart products were well tolerated (Supplementary Table 3) without differences in serious TEAEs between groups. Treatment-emergent adverse events were reported by a lower percentage of participants in the switching group compared with nonswitching group (25.3% vs 42.3%) mainly due to higher frequency of infections and infestations of mild-to-moderate intensity in the nonswitching group. No injection site reactions or hypersensitivity reactions were reported.

Immunogenicity

There were no clinically meaningful differences between switching and nonswitching groups in AIA at baseline, AIA at 16 weeks, and AIA titers over 16 weeks (Supplementary Table 4). The percentage of participants with a treatment-emergent AIA response was similar in the switching and nonswitching groups (8.2% and 10.3%, respectively), with an absolute risk difference between switching and nonswitching of −2.0% (90% CI = [−8.71, 4.76]). The percentage of participants positive for AIAs remained relatively stable in both treatment groups during the on-treatment period, with 41.9% of switching participants and 41.2% of nonswitching participants being AIA positive at week 16 (Supplementary Figure 5).

Anti-insulin aspart antibodies were cross-reactive to human insulin in most of the participants (range = 88.1%-97.6%) in both treatment groups. Subgroup analyses did not suggest an effect of treatment-emergent AIAs on efficacy in terms of mean change of HbA_1c or insulin dose from baseline to week 16, or on safety in terms of hypoglycemia, TEAEs, or serious TEAEs (Table 2). In addition, there was no relationship between the maximal AIA titers and insulin doses or efficacy and safety parameters, regardless of treatment-emergent AIA status or for participants with treatment-emergent AIAs (data not shown). There was a single participant with treatment-emergent NAbs in the nonswitching group.

Table 2.

Efficacy and Safety Parameters by Presence or Absence of Treatment-Emergent AIA (AIA Population).

Parameter	Presence of treatment-emergent AIA
	Yes^a		No
	Switching group (N = 8)	Nonswitching group (N = 11)	Switching group (N = 90)	Nonswitching group (N = 96)
HbA_1c, %
Baseline [n]	7.55 ± 1.41 [8]	7.47 ± 1.14 [11]	7.71 ± 0.97 [90]	7.70 ± 0.95 [96]
Week 16 [n]	7.37 ± 2.03 [7]	7.01 ± 0.74 [11]	7.50 ± 0.95 [89]	7.52 ± 0.85 [94]
Change from baseline [n]	–0.20 ± 1.16 [7]	–0.46 ± 0.67 [11]	–0.20 ± 0.54 [89]	–0.21 ± 0.65 [94]
Total insulin dose, U/kg/day ^b
Baseline [n]	0.741 ± 0.275 [5]	0.630 ± 0.146 [7]	0.761 ± 0.333 [78]	0.756 ± 0.325 [81]
Week 16 [n]	0.715 ± 0.254 [6]	0.642 ± 0.234 [10]	0.780 ± 0.286 [83]	0.753 ± 0.347 [88]
Change from baseline [n]	–0.035 ± 0.075 [5]	–0.050 ± 0.125 [6]	0.022 ± 0.127 [75]	–0.000 ± 0.107 [76]
Hypoglycemia, n (%) ^c
Any	6 (75.0)	11 (100.0)	88 (97.8)	93 (96.9)
Severe	0 (0.0)	0 (0.0)	6 (6.7)	4 (4.2)
Documented symptomatic
≤70 mg/dL (3.9 mmol/L)	5 (62.5)	11 (100.0)	83 (92.2)	86 (89.6)
<54 mg/dL (3.0 mmol/L)	4 (50.0)	9 (81.8)	73 (81.1)	80 (83.3)
Hypersensitivity/injection site reactions, n (%)
Any hypersensitivity reaction	0	0	0	0
Any injection site reaction	0	0	0	0
TEAEs and SAEs, n (%)
Any TEAE	4 (50.0)	5 (45.5)	20 (22.2)	41 (42.7)
Any SAE	1 (12.5)	0 (0.0)	3 (3.3)	4 (4.2)

Abbreviations: AIA, anti-insulin aspart antibody; SAE, serious adverse event; SD, standard deviation; TEAE, treatment-emergent adverse event.

All data mean ± SD or n (%) unless stated otherwise.

Patients with pre-existing AIAs that were boosted to a significantly higher titer (at least fourfold increase) compared with baseline or patients without pre-existing AIA (or missing baseline) and with at least one positive AIA sample.

Baseline insulin dose was defined as the median of daily doses available in the week before the first injection of study medication. The value at week 16 was the median of daily doses available in the week before the visit.

Number and percentage of participants with at least one treatment-emergent event.

Discussion

This study showed no clinically meaningful differences in the efficacy, safety, and immunogenicity of switching between the biosimilar insulin SAR-Asp and its reference product NN-Asp in individuals with T1D vs continuous use of the reference NN-Asp product. The pharmacokinetic profiles of SAR-Asp and NN-Asp from this study are reported in a separate publication¹³ and confirm similar pharmacokinetic exposure of the two treatment groups. Taken together with similarity demonstrated in physicochemical analyses, nonclinical studies, and clinical studies, this confirms that SAR-Asp is interchangeable with NN-Asp for managing patients with diabetes. To our knowledge, this is the first study to evaluate interchangeability between two insulin aspart products.

Our study has several important clinical implications. First, the results confirm that the two insulin aspart products used in this study are interchangeable without any increased risk for the patient in terms of safety and immunogenicity or diminished efficacy. Second, our report confirms previous study findings that show AIAs, although common in patients with T1D on insulin therapy, did not have a relevant impact on clinical outcomes. Third, the availability of interchangeable insulin therapies would result in accessible future insulin treatments.

After initiation of this study in March 2019, the FDA revised their draft regulatory guidance for interchangeable insulin biosimilars in November 2019 to say that comparative clinical immunogenicity studies are generally not necessary for proposed highly similar interchangeable insulin products that have little or no residual uncertainty regarding immunogenicity.²² As such, given the high similarity between these two insulin aspart products, a study like this may not be a regulatory requirement in the future. Since 2021, two interchangeable insulin products have now been approved in the United States without undertaking a switching study described in FDA’s interchangeability guidance.^23-25 Since October 2022, biosimilar medicines approved in the EU are considered to be interchangeable with their reference medicine or with an equivalent biosimilar and decisions on how to implement interchangeability either through switching and/or substitution are managed by individual member states.²⁶

The evolving landscape of biosimilars means that switching between an originator and a biosimilar product, and even pharmacist-led substitution, is expected to become common in daily clinical practice. It is inevitable that clinicians will be concerned about switching patients on stable treatment with the reference insulin medicine to a biosimilar.^27-29 It should be noted that the results of this study are applicable only to switching between this single originator (NN-Asp) and biosimilar product (SAR-Asp) and cannot be generalized to biosimilar insulins developed by different manufacturers or to other insulin products.

The strengths of this study are that we evaluated the use of multiple switches in insulin treatment in a relatively large number of participants across multiple centers. The study design was in line with regulatory guidance at the time of initiation.¹² With switching, multiple exposure to each product can prime the immune system to recognize subtle structural differences between products, which may theoretically increase the overall immune response. The immunogenic potential of SAR-Asp was therefore assessed as a key secondary end point in the study. Efficacy and immunogenicity assessments were based on objectively collected data that was analyzed by central laboratories who were blinded to the study treatment to minimize bias. The 16 week study duration was chosen to ensure a sufficiently long treatment duration in each switching cycle to assess development of AIA and the number of switching cycles. Potential allergic reactions may occur within hours post injection at any time during the treatment; however, these reactions are very rare for highly purified insulin analogs including NovoLog.³⁰ Anti-insulin aspart antibodies of the immunoglobulin G isotype usually appear about two weeks after first exposure to the antigen, or, in case of a re-exposure to the antigen (secondary response), even earlier within one week.^31-33 In addition, the 16 week study duration is considered sufficient long to detect differences in HbA_1c as the efficacy parameter. The study included only participants with T1D as due to their underlying autoimmune disorder, this population is more sensitive to evaluate potential differences in the immune response and the effects of AIAs on efficacy and safety, as compared with the type 2 diabetes population.³⁴

However, the study findings should be interpreted considering a few limitations such as open-label design and the fact that clinical outcomes were secondary or tertiary outcomes in the study with no formal statistical testing. Patient blinding of the study treatments was not possible as SAR-Asp was administered via a pre-filled disposable pen that was different from the approved pre-filled disposable pen used for NN-Asp. To partially overcome this limitation, assessments were based on objectively collected data that were analyzed by central laboratories and an independent adjudication committee that were all blinded to the study treatment. The small number of participants in the subgroup with treatment-emergent AIAs (8 and 11, respectively, in the switching and nonswitching group) limits definitive conclusions about the immunogenicity outcomes. Despite its increasing use in T1D management, continuous glucose monitoring (CGM) was not used as a tool to assess the efficacy and safety of the two treatment regimens in this study. The study was designed in 2018 before CGM use in clinical trials became routine for participants with T1D. Hypoglycemia events were defined based on SMPG, and therefore, time below range (ie, percentage of time when blood glucose is below 70 mg/dL) may have been different if CGM had been used in this study. It is recognized that CGM use would have been potentially helpful in the assessment of the numerical differences in the frequency of severe hypoglycemia events between groups.

Conclusions

In conclusion, SAR-Asp and NN-Asp in this study are interchangeable and switching between these two products results in similar efficacy, safety, and immunogenicity than continuous use of NN-Asp.

Supplemental Material

sj-docx-1-dst-10.1177_19322968241232709 – Supplemental material for Safety and Efficacy of Switching SAR341402 Insulin Aspart and Originator Insulin Aspart vs Continuous Use of Originator Insulin Aspart in Adults With Type 1 Diabetes: The GEMELLI X Trial

Supplemental material, sj-docx-1-dst-10.1177_19322968241232709 for Safety and Efficacy of Switching SAR341402 Insulin Aspart and Originator Insulin Aspart vs Continuous Use of Originator Insulin Aspart in Adults With Type 1 Diabetes: The GEMELLI X Trial by Viral N. Shah, Amer Al-Karadsheh, Cathy Barnes, Jose Mandry, Samer Nakhle, Karin Wernicke-Panten, Daniel Kramer, Wolfgang Schmider, Suzanne Pierre, Lenore Teichert, Baerbel Rotthaeuser, Bhaswati Mukherjee and Timothy S. Bailey in Journal of Diabetes Science and Technology

Footnotes

Acknowledgements

The authors thank the study participants, trial staff, and investigators for their participation. The authors thank Gabriela Turcanu for her clinical input, along with Irma Slomp, Moritz Heinz, Valérie Dyevre, Monica Li, and Catherine Lelouet (all employees of Sanofi) who supported the statistical analysis of this trial. Sanofi was the sponsor of the study and was responsible for the design and coordination of the trial, monitoring clinical sites, collecting, and managing data, and performing all statistical analyses. Editorial assistance provided by DJ Quinlan of Oberon Ltd (London, UK) was funded by Sanofi.

Abbreviations

AE, adverse event; AIAs anti-insulin aspart antibodies; ANCOVA, analysis of covariance; BL, baseline; BMI, body mass index; COVID-19, coronavirus disease 2019; D: day; FDA, Food and Drug Administration; FPG, fasting plasma glucose; Gla-100, insulin glargine 100 U/mL; HbA_1c, glycated hemoglobin; LS, least squares; MDI, multiple daily injections; MT, mealtime; NAbs; neutralizing anti-insulin aspart antibodies; NN-Asp, NovoLog insulin aspart; SAR-Asp, SAR341402 insulin aspart; T1D, type 1 diabetes; TEAE, treatment-emergent adverse event; W, week.

Author Contributions

K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., and B.M. contributed to the conception and design of the study, V.N.S., A.Al-K., C.B., J.M., S.N., and T.S.B. contributed to the acquisition of data (as trial investigators). K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., and B.M. contributed to the analysis of data. K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., and B.M. contributed to the interpretation of the data. K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., and B.M. contributed to the drafting, and V.N.S., K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., A.Al-K., C.B., J.M., S.N., B.M., and T.S.B. contributed to the critical revision of the work for important intellectual content. B.R. is the guarantor of this work and confirms access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors approved the manuscript for publication.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: V.N.S. and V.N.S.s’ previous employer have received research funding from Sanofi US, Eli Lilly, Insulet, Dexcom Inc, Novo Nordisk, Mylan GmbH, and vTv Therapeutics. V.N.S. has served on an advisory board for Sanofi US, Medscape LLC, and LifeScan and received speaking fees from Dexcom Inc and Insulet. K.W.-P., G.T., D.K., W.S., S.P., L.T., B.R., and B.M. are employees and stockholders of Sanofi. A.Al-K., C.B., J.M., and S.N. declare no conflict of interests and no disclosures. T.S.B. has received research support from Abbott Diabetes, Abbott Rapid Diagnostics, Biolinq, Capillary Biomedical, Dexcom, Eli Lilly, Kowa, LifePlus, Livongo, Mannkind, Medtronic, Novo Nordisk, REMD, Sanofi, Sanvita, Senseonics, Viacyte, vTv Therapeutics, and Zealand Pharma; consulting honoraria from Abbott, CeQur, LifeScan, Mannkind, Medtronic, Novo Nordisk, and Sanofi; and speaking honoraria from Mannkind, Medtronic, and Sanofi.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by Sanofi.

ORCID iDs

Viral N. Shah

Bhaswati Mukherjee

Timothy S. Bailey

Data Availability Statement

Qualified researchers may request access to patient-level data and related documents. Patient-level data will be anonymized, and study documents will be redacted to protect the privacy of trial participants. Further details on Sanofi’s data-sharing criteria, eligible studies, and process for requesting access can be found at .

Supplemental Material

Supplemental material for this article is available online.

References

Dafny

. Falling insulin prices—what just happened? N Engl J Med. 2023;388(18):1636-1639. doi:10.1056/NEJMp2303279.

White

Wagner

Patel

. The impact of biosimilar insulins on the diabetes landscape. J Manag Care Spec Pharm. 2022;28(1):91-98. doi:10.18553/jmcp.2021.21253.

World Health Organization. WHO prioritizes access to diabetes and cancer treatments in new Essential Medicines Lists. https://www.who.int/news/item/01-10-2021-who-prioritizes-access-to-diabetes-and-cancer-treatments-in-new-essential-medicines-lists. Accessed November 17, 2023.

Novo Nordisk. NovoLog, insulin aspart injection 100 units/mL. NovoNordisk. https://www.novo-pi.com/novolog.pdf. Published February 2023. Accessed November 17, 2023.

European Medicines Agency. NovoRapid. Summary of product characteristics. https://www.ema.europa.eu/documents/product-information/novorapid-epar-product-information_en.pdf. Published March 22, 2023. Accessed November 17, 2023.

Kapitza

Nosek

Schmider

Teichert

Nowotny

. Single-dose euglycemic clamp study demonstrating pharmacokinetic and pharmacodynamic similarity between SAR341402 insulin aspart and US- and EU-approved versions of insulin aspart in subjects with type 1 diabetes. Diabetes Technol Ther. 2020;22(4):278-284. doi:10.1089/dia.2019.0351.

Shiramoto

Yoshihara

Schmider

, et al. Pharmacokinetic and pharmacodynamic similarity between SAR341402 insulin aspart and Japan-approved NovoRapid in healthy Japanese subjects. Sci Rep. 2021;11(1):22931. doi:10.1038/s41598-021-02410-z.

Garg

Wernicke-Panten

Wardecki

, et al. Efficacy and safety of insulin aspart biosimilar SAR341402 versus originator insulin aspart in people with diabetes treated for 26 weeks with multiple daily injections in combination with insulin glargine: a randomized open-label trial (GEMELLI 1). Diabetes Technol Ther. 2020;22(2):85-95. doi:10.1089/dia.2019.0382.

Garg

Wernicke-Panten

Wardecki

, et al. Safety, immunogenicity, and glycemic control of insulin aspart biosimilar SAR341402 versus originator insulin aspart in people with diabetes also using insulin glargine: 12-month results from the GEMELLI 1 trial. Diabetes Technol Ther. 2020;22(7):516-526. doi:10.1089/dia.2020.0008.

10.

Thrasher

Polsky

Hovsepian

, et al. Safety and tolerability of insulin aspart biosimilar SAR341402 versus originator insulin aspart (NovoLog) when used in insulin pumps in adults with type 1 diabetes: a randomized, open-label clinical trial. Diabetes Technol Ther. 2020;22(9):666-673. doi:10.1089/dia.2019.0446.

11.

US Food and Drug Administration Center for Drug Evaluation and Research (CDER). Guidance for industry. Considerations in demonstrating interchangeability with a reference product; 2019:1-20. https://www.fda.gov/media/124907/download. Accessed November 17, 2023.

12.

US Food and Drug Administration Center for Drug Evaluation and Research (CDER). Draft guidance for industry. Considerations in demonstrating interchangeability with a reference product; 2017:1-30. https://www.fda.gov/media/102592/download. Accessed November 17, 2023.

13.

Shah

Al-Karadsheh

Barnes

, et al. Pharmacokinetic similarity of switching SAR341402 insulin aspart biosimilar and NovoLog insulin aspart versus continuous use of NovoLog in adults with type 1 diabetes: the GEMELLI X trial. Diabetes Obes Metab. 2024;26(2):540-547. doi:10.1111/dom.15341.

14.

American Diabetes Association. 8. Pharmacologic approaches to glycemic treatment. In Standards of Medical Care in Diabetes-2017. Diabetes Care. 2017;40(suppl 1):S64-S74. doi:10.2337/dc17-S011.

15.

Shankar

Arkin

Cocea

, et al. Assessment and reporting of the clinical immunogenicity of therapeutic proteins and peptides-harmonized terminology and tactical recommendations. AAPS J. 2014;16(4):658-673. doi:10.1208/s12248-014-9599-2.

16.

European Medicines Agency. Guideline on immunogenicity assessment of biotechnology-derived therapeutic proteins, EMEA/CHMP/BMWP/14327/2006 Rev.1. https://www.ema.europa.eu/documents/scientific-guideline/guideline-immunogenicity-assessment-therapeutic-proteins-revision-1_en.pdf. Published May 18, 2017. Accessed November 17, 2023.

17.

US Food and Drug Administration Center for Drug Evaluation and Research (CDER). Guidance for industry. Scientific considerations in demonstrating biosimilarity to a reference product; April 2015:1-24. https://www.fda.gov/media/82647/download. Accessed November 17, 2023.

18.

Workgroup on Hypoglycemia American Diabetes Association. Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care. 2005;28(5):1245-1249. doi:10.2337/diacare.28.5.1245.

19.

Seaquist

Anderson

Childs

, et al. Hypoglycemia and diabetes: a report of a workgroup of the American Diabetes Association and the Endocrine Society. Diabetes Care. 2013;36(5):1384-1395. doi:10.2337/dc12-2480.

20.

International Hypoglycaemia Study Group. Glucose concentrations of less than 3.0 mmol/L (54 mg/dL) should be reported in clinical trials: a joint position statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2017;40(1):155-157. doi:10.2337/dc16-2215.

21.

Little

RJA

Rubin

. Statistical Analysis with Missing Data. First ed. New York, NY: John Wiley & Sons; 1987.

22.

US Food and Drug Administration Center for Drug Evaluation and Research (CDER). Guidance for industry. Clinical immunogenicity considerations for biosimilar and interchangeable insulin products; November 2019:1-8. https://www.fda.gov/media/133014/download. Accessed November 17, 2023.

23.

Matli

Wilson

Rappsilber

Sheffield

Farlow

Johnson

. The first interchangeable biosimilar insulin: insulin glargine-yfgn. J Diabetes Sci Technol. 2023;17(2):490-494. doi:10.1177/19322968211067511.

24.

In Brief: Rezvoglar—another insulin glargine product interchangeable with Lantus. Med Lett Drugs Ther. 2023;65(1673):56. doi:10.58347/tml.2023.1673c.

25.

Vaidyanathan

Wang

Tran

Seo

. Leveraging clinical pharmacology data to assess biosimilarity and interchangeability of insulin products. Clin Pharmacol Ther. 2023;113(4):794-802. doi:10.1002/cpt.2731.

26.

European Medicines Agency. Biosimilar medicines can be interchanged. https://www.ema.europa.eu/en/news/biosimilar-medicines-can-be-interchanged. Published September 19, 2022. Accessed November 17, 2023.

27.

Hemmington

Dalbeth

Jarrett

, et al. Medical specialists’ attitudes to prescribing biosimilars. Pharmacoepidemiol Drug Saf. 2017;26(5):570-577. doi:10.1002/pds.4186.

28.

Chapman

Fitzpatrick

Aladul

. Knowledge, attitude and practice of healthcare professionals towards infliximab and insulin glargine biosimilars: result of a UK web-based survey. BMJ Open. 2017;7(6):e016730. doi:10.1136/bmjopen-2017-016730.

29.

Sarnola

Merikoski

Jyrkkä

Hämeen-Anttila

. Physicians’ perceptions of the uptake of biosimilars: a systematic review. BMJ Open. 2020;10(5):e034183. doi:10.1136/bmjopen-2019-034183.

30.

Lindholm

Jensen

Home

Raskin

Boehm

Råstam

. Immune responses to insulin aspart and biphasic insulin aspart in people with type 1 and type 2 diabetes. Diabetes Care. 2002;25(5):876-882. doi:10.2337/diacare.25.5.876.

31.

Uhr

Finkelstein

. The kinetics of antibody formation. Prog Allergy. 1967;10:37-83.

32.

Ogra

Karzon

Righthand

MacGillivray

. Immunoglobulin response in serum and secretions after immunization with live and inactivated poliovaccine and natural infection. N Engl J Med. 1968;279(17):893-900. doi:10.1056/nejm196810242791701.

33.

Ogra

Kerr-Grant

Umana

Dzierba

Weintraub

. Antibody response in serum and nasopharynx after naturally acquired and vaccine-induced infection with rubella virus. N Engl J Med. 1971;285(24):1333-1339. doi:10.1056/nejm197112092852401.

34.

Fineberg

Kawabata

Finco-Kent

Fountaine

Finch

Krasner

. Immunological responses to exogenous insulin. Endocr Rev. 2007;28(6):625-652. doi:10.1210/er.2007-0002.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.97 MB