Regulatory guidelines for insulin products (IPs) development require to test immunogenicity in a reasonable number of patients with type 1 diabetes mellitus (T1DM). The problem of sample size for immunogenicity comparison is substantial in insulin biosimilar development since the required samples are sizeable however still incapable to detect any immunogenicity difference, because the biosimilar and a reference drug are very similar in all terms. The aim of this study was to estimate the input of IP into immune response development.
Blood samples of 180 T1DM patients were taken to estimate an anti-insulin antibody concentration (AIAC) via enzyme-linked immunosorbent assay. The information on insulin international non-proprietory name (INN) was gathered. Influence of basal and bolus insulin INNs on AIAC was estimated with ANOVA. One-to-one comparisons were made with Mann-Whitney U-test.
Patients were administered 4 INNs of basal insulin (glargine, detemir, degludec, isofan) and 4 INNs of bolus insulin (lispro, aspart, apidra, human soluble). AIAC did not depend on INN of insulin administered to a patient (p=0.765 for basal insulin, p=0.283 for bolus insulin). AIAC did not differ in patients receiving insulin analogues and human insulins (p=0.374 for basal insulin, p=0.751 for bolus insulin).
IP’s input into immune response development was nonsignificant, thus the huge sample is required to detect immunogenicity difference. This leads to dramatic increase of clinical trials cost and breaks the idea of cost-affordable biosimilars. Thus immunogenicity risk assessment should be provided mostly at preclinical stage, and immunogenicity should be studied via pharmacovigilance.