Efficacy and safety of the combined metabolic medication, containing inosine, nicotinamide, riboflavin and succinic acid, for the treatment of diabetic neuropathy: a multicenter randomized, double-blind, placebo-controlled parallel group clinical trial (CYLINDER)

Study overview

We performed a randomized double-blind, placebo-controlled clinical trial in 10 clinical centers in Russian Federation.

Study population

We included men and women aged 45–74 years, with diagnosis of type 2 diabetes confirmed ≥1 year before recruitment and previously diagnosed symptomatic distal sensorimotor DPN, as demonstrated by Total Symptom Score19 (TSS) ˃5 points, with a score ≥2 by at least one of the TSS subscales; the severity of pain on the TSS subscale had to be ≤2; and NIS-LL (Neuropathy Impairment Score of the Lower Limb score) ≥2. Diagnosis of DPN was previously established in the ambulatory centers of diabetes control; diagnostic evaluation was not repeated for the study purpose. Other inclusion criteria were body mass index 22–40 kg/m²; hemoglobin A1c (HbA1c) 7.0%–10.0%; regular treatment with oral hypoglycemic drugs and/or insulin of medium, long or ultra-long duration of action and/or agonists of glucagon-like peptide-1 receptors, in stable dosage for at least 12 weeks before screening; consent to maintain a stable diet, exercise, therapy, and diabetes control, and adequate contraceptive methods if fertile. We excluded pregnant or lactating women; patients with type 1 diabetes and other specific types of diabetes; with fasting plasma glucose (FPG) >15 mmol/L; history of ketoacidosis or hyperosmolarity (coma) within 6 months prior to screening, or hypoglycemia within 3 months before screening; severe manifestations of uncontrolled diabetes, including proliferative retinopathy, diabetic foot with limb ischemia or wound defect, neuro-osteoarthropathy; patients on therapy with short-acting and ultra-short-acting insulins, prior or current treatment with systemic corticosteroid drugs, cytostatics, or penicillamine; patients with severe central nervous system disorders (seizures, head trauma, demyelinating disease, tumor), malignancies, decompensated cardiovascular disease currently or within 3 months before screening, kidney disease with a glomerular filtration rate <30 mL/min; active viral hepatitis or cirrhotic liver disease; anemia or acute blood loss or blood transfusion within the previous 12 weeks; HIV infection; any severe infection within 30 days of screening; and history of drug or alcohol abuse. We did not include patients with type 1 diabetes and patients treated with short-acting insulins in order to avoid imbalance and wide variations of insulin daily doses in the study population.

Study medication SINR is a combined product available in two forms: 10 mL vials with solution for intravenous infusion, containing (per 1 L): succinic acid 100 g, nicotinamide 10 g, inosine 20 g, riboflavin sodium phosphate 2 g; and enteric-coated tablets, containing succinic acid 0.3 g, inosine 0.05 g, nicotinamide 0.025 g, and riboflavin sodium phosphate 0.005 g. Intake of inosine, nicotinamide, riboflavin, succinic acid, alpha-lipoic acid, thiamine derivatives, pyridoxine, cyanocobalamin (excluding multivitamins), antidepressants and gabapentin derivatives within 3 months before screening was prohibited during the whole study period. In addition, other drugs with action on the nervous system and metabolic medications were not allowed for use within the whole study period, including antiepileptic drugs, antidepressants, agents affecting neuromuscular transmission, anticholinesterase agents, drugs for the treatment of dementia, psychostimulants and nootropic drugs, centrally acting sympathomimetics and parasympathomimetics, peripheral vasodilators, and drugs containing vitamin B.

Study design

Participants were randomly assigned (1:1) into an experimental group or placebo group using the interactive web response system. Block randomization was used (block sizes of four) stratified by intake of insulin. The randomization sequence was drawn up by an independent, blinded statistician who used a validated pseudo-random number generator. The study consisted of a screening period (2 weeks), a treatment period (12 weeks), and a follow-up period (2 weeks) (figure 1). Treatment was performed in two phases: intravenous infusion of study medication/placebo for 10 days, followed by oral administration of study medication/placebo for 75 days; in total the duration of investigational treatment was 12 weeks or 85 days. During the first phase (days 1–10), the study medication/placebo was administered once a day at a dose of 20 mL intravenously, diluted by 200 mL of normal saline, at a rate of 3–4 mL/min. The solution for infusion was administered in each center by a blinded team, except for the one unblinded person who prepared the medication before infusion; the yellow color of the solution was masked by opaque infusion systems and dark coverings for vials. From day 11, patients were given study medication/placebo for self-administration, to be taken two tablets two times per day orally (in the morning and in the evening no later than 18:00). Blinding of tablets was ensured by masking with a placebo, packaging, and the same labeling. All outcome assessments were performed by the investigators who were blinded to treatment allocation.

Request permission

Study procedures

The following tests and scales were used for study assessments: general scale for the assessment of neurological symptoms (TSS);19 20 objective assessment of neurological symptoms with NIS-LL;21 rough neuropsychological assessment of executive ability by performing a simple test of visual scanning and working memory, Trail Making Test (TMT);22 evaluation of symptoms of chronic asthenia by Fatigue Assessment Scale (FAS).23 All assessments were performed on days 0, 11, 29, 57, and 86. Details of TSS assessment were published elsewhere.19 20 The primary outcome was improvement of DPN symptoms, as measured by change in TSS at week 12 (day 86) from baseline. Secondary outcomes, also assessed at day 86 from baseline, were changes of the individual components on TSS scale, the proportion of patients who achieved ≥50% reduction in TSS total score, change in NIS-LL score, change of the results of the TMT (part A and part B), change in FAS scores, HbA1c and FPG levels, and changes of the atherogenic index and body weight. Safety outcomes were the incidence of hypoglycemia episodes during the 12 weeks of study treatment, and the incidence of adverse events (AEs) and serious AEs (SAEs) based on subjective complaints, laboratory tests, results of physical examination, vital signs, and ECG. Registration of AEs/SAEs was provided from the signing of informed consent until the completion of participation in the study. Cases of AE were coded using the latest MedDRA medical dictionary (V.23.1). Deviations of laboratory findings from normal ranges, and clinically significant deviations in vital signs and ECG records during the study period were reported as AEs.

Statistical analysis

In a double-blind, randomized, placebo-controlled trial SYDNEY 224 alpha-lipoic acid showed superiority over placebo in patients with DPN in terms of a decrease in the total TSS score after 5 weeks of treatment. The minimum decrease in TSS in patients receiving the study drug was 4.5 points, and in the placebo group 2.9 points. The maximum SD of the TSS change was 3.5 points. For this study, the sample size calculation was based on assumption that the difference between SINR and placebo groups in terms of TSS reduction at week 12 from baseline will be 1.5 points or more. To achieve 80% statistical power with α=0.05 and SD 3.5 points, to test the superiority hypothesis, 86 patients in each group must be included in the analysis. Assuming a 20% potential dropout rate, the calculated study population was 216 patients (108 patients in each group). The primary population for efficacy analysis was the full analysis set (ITT), which included all randomized patients who received at least one dose of study medication and had at least one efficacy parameter score after baseline (intent-to-treat analysis, ITT). In addition, efficacy analysis was performed in the per-protocol (PP) population, which embodies patients who received the full course of investigational treatment without significant protocol deviations. All patients who received at least one dose of the study medication were included into the safety analysis.

Data analysis

A t-test was used to assess the primary efficacy endpoint, that is, the change in the total TSS score at week 12 from baseline. Additionally, analysis of covariance was performed using treatment group, stratum, and center factors as fixed parameters to test the null hypothesis of absence of difference between groups. To assess the dynamics of TSS subscores between visits, a mixed model was used, with the subscale score as a dependent variable, group as a fixed factor, center and subject as random factors, and visit number as a repeated factor. The analysis of other numerical parameters of efficiency was carried out in a similar way. For the analysis of secondary efficacy endpoints, which were categorical variables, and for between-group comparison of safety (AE/SAE), the χ2 test or, where appropriate, Fisher’s exact test, was used. A mixed repeated measures model was used to assess the dynamics of study variables between visits. To assess the dynamics of laboratory parameters in relation to the baseline, a paired t-test or Wilcoxon test (depending on the type of data distribution) was used separately for each group. For between-group comparison of changes relative to baseline, the t-test or the Mann-Whitney test was used, depending on the type of data distribution. Testing of quantitative variables for normality of distribution was carried out using the Shapiro-Wilk test and the test for skewness and kurtosis.

The incidence of AEs/SAEs during treatment was calculated for each class of organ systems and the preferred term for each group. Summary data on the severity of AEs, their association with study medication, and outcomes were presented for each organ system class. Data on SAEs, AEs leading to the need for early completion of the study for the subject, and AEs with severity according to National Cancer Institute Common Terminology Criteria for Adverse Events (V.5.0, publish date: November 27, 2017) of grade 3 or higher were analyzed separately. For between-group comparison of safety parameters, the χ2 test and, where appropriate, Fisher’s exact test, was used. All statistical calculations were performed using Stata V.14.

Description of the study groups

Baseline demographic data, clinical characteristics, and comorbidities assessed in the ITT population were similar between treatment groups (table 1). All study subjects were Caucasians. Statistically significant differences of the incidence of sequelae of diabetes were not observed, except for the presence of diabetic retinopathy, which was more prevalent in the placebo group. Most patients in both groups had type 2 diabetes for more than 10 years. The duration of DPN was, in most cases, longer than 4 years. Only 4 of 108 (3.7%) patients in the SINR group and 3 of 107 (2.8%) patients in the placebo group received concomitant therapy for DPN (p>=0.991). The most frequent comorbidities were arterial hypertension, dyslipidemia, ischemic heart disease, and obesity. All patients received glucose-lowering therapy for type 2 diabetes; also, most patients received antihypertensive therapy and treatment for coronary artery disease. Compliance with the study medication intake in both groups was within the range 80%–120%, as permitted by the study protocol.

Primary endpoint

In the experimental group, a significantly greater reduction in the symptoms of DPN, assessed by the TSS, was achieved, compared with the placebo group (table 2). Reduction of symptoms measured by the TSS scale in subgroups, defined by the degree of compensation for type 2 diabetes (as reflected by glycated hemoglobin level below vs above 8.0%) and by severity of DPN symptoms (as demonstrated by baseline TSS below vs over 7.5), is presented in table 2.

As for the individual components of the TSS, several statistically significant differences were found between groups in terms of reduction of TSS subscores at different visits (figure 3). The analysis revealed the presence of statistically significant differences between the groups by dynamics of the intensity of paresthesia (p=0.004) and the intensity of numbness (p=0.003). There were no statistically significant differences between the groups in the dynamics of the intensity of burning (p=0.245) and the intensity of pain (p=0.578). On day 11, paresthesia intensity was 1.63±0.73 in SINR group and 1.83±0.72 in the control group (p=0.045); on day 31 1.58±0.71 and 1.78±0.66 (p=0.021), on day 57 1.49±0.66 and 1.70±0.69 (p=0.024), and on day 86 1.38±0.67 and 1.66±0.74 (p=0.018), respectively. Intensity of numbness in SINR group versus control group on day 11 was 1.65±0.66 vs 1.82±0.61 (p=0.049), on day 31 1.52±0.62 vs 1.82±0.61 (p=0.006), on day 57 1.45±0.64 vs 1.70±0.65 (p=0.084), and on day 86 1.30±0.60 vs 1.63±0.60 (p=0.022). By the end of treatment, statistically significant differences between the groups were detected also by the intensity of burning: on day 57 1.18±0.67 vs 1.27±0.71 (p=0.001), and on day 86 1.09±0.69 vs 1.18±0.74 (p=0.007) in SINR group versus control group, respectively.

Request permission

Secondary endpoints

We did not observe statistically significant differences by the other secondary study endpoints. The mean decline of NIS-LL score on day 86, that is, at the end of therapy, compared with baseline in the experimental group was 3.2 points, in the placebo group 2.5 points (p=0.098). Neither statistically significant differences by dynamics of changes in the NIS-LL scale were found between the groups throughout the 12 weeks of treatment (p=0.166, mixed model). However, we observed a non-significant trend of improvement of touch pressure sensation, as reflected by lower mean NIS-LL touch pressure subscore, in patients receiving SINR versus placebo on day 86 (1.32±0.9 vs 1.56±0.8, respectively, p=0.055) (online supplemental table 1). Mean HbA1c levels slightly decreased in both study groups; on day 86 compared with baseline, mean change was −0.36 (95% CI −0.53 to −0.19) in SINR group and −0.23 (95% CI −0.42 to −0.05) in placebo group (p=0.306). In the SINR group, there was also a trend to the decrease in the FPG levels (mmol/L) on day 29, day 57, and day 86 relative to the baseline level, but this trend did not reach statistical significance (p=0.334). The performance indicators of the TMT (execution time, number of errors) of part A and part B improved in both groups, but the analysis did not reveal statistically significant differences between the groups by these parameters. In both treatment groups, there was a slight decrease in the atherogenic index and in body weight, but without statistical significance between the treatment groups (p=0.817 and p=0.755, respectively).

Safety endpoints

During 12 weeks of investigational treatment, hypoglycemia occurred in 16 of 108 (14.8%) patients of the experimental group and in 25 of 107 (23.4%) patients of the control group (p=0.105). All episodes did not require special treatment (only intake of carbohydrate food or drinks) and resolved without consequences. In addition, during the study, 173 AEs were recorded, which developed in 87 patients; 32 episodes of reported AEs in 16 patients had at least a possible association with investigational treatment. The most common AEs were increased blood pressure (SINR/placebo=11/11), headache (SINR/placebo=19/8), and dizziness (SINR/placebo=6/4). AEs reported by the investigator as related to study drug administration (definite, probable, possible association categories) were hypertension, headache, and dizziness. There were five episodes of SAE, which required hospitalization of patients: four cases of pneumonia (the study period substantially overlapped with the peak of COVID-19 pandemic) and one case of the newly detected atrial fibrillation paroxysm. No fatal AEs were reported. There were no statistically significant differences between the study groups in terms of AE frequencies, severity, outcomes, and measures taken.