Metrological Assessment of Analytical Procedures Used for Semi-micro and Micro Determination of Water in Lyophilisates

INTRODUCTION. The metrological characterisation of analytical procedures helps in selecting a procedure that will provide reliable and consistent results. This is particularly relevant in cases where a pharmacopoeial standard includes several test methods for a quality parameter of a medicinal product. For example, the water content of a lyophilisate can be measured by volumetric or coulometric Karl Fischer titration (semi-micro and micro determination) and by the loss on drying.

AIM. This study aimed to investigate the correlation between the method selected for the determination of water content in a lyophilisate and the results obtained by this method via analysis of variance (ANOVA), as well as to evaluate the uncertainty of water content measurements by Karl Fischer titration.

MATERIALS AND METHODS. The study focused on a lyophilised micafungin medicinal product. The water content in the lyophilisate samples was measured by Karl Fischer titration (semi-micro and micro determination) and, additionally, by the loss on drying as outlined in the Pharmacopoeia of the Eurasian Economic Union (General Monograph 2.5.1.14 Lyophilisates). The authors performed ANOVA using Microsoft Excel to assess the effect of the test method on the results. The authors conducted statistical analysis of the experimental results, including Fisher's test to assess the reproducibility and Student's test to compare the mean results obtained by two different methods. To assess measurement uncertainty, the authors calculated confidence interval limits at confidence levels of 0.95 and 0.99.

RESULTS. ANOVA demonstrated that the method selected to determine the lyophilisate water content had an effect on the results obtained. Karl Fischer titration (semi-micro and micro determination) provided reproducible results; the F-values calculated for four lyophilisate batches (F₁=2.3440, F₂=1.0762, F₃=3.5302, and F₄=1.0989) were lower than the critical F-value (F_{crit(0.95;2;2)}=19.000). The mean results of different methods were statistically dissimilar; the calculated t-values (t₁=9.2391, t₂=11.4847, t₃=11.0041, and t₄=33.6502) exceeded the critical t-value (t_crit(0.95;2)=4.3027). To assess the measurement uncertainty, the authors calculated two-sided confidence interval limits using Student's test at confidence levels of 0.95 and 0.99. The measurement uncertainty of the semi-micro method was lower than that of the micro method.

CONCLUSIONS. The results of water determination in the micafungin lyophilisate depend on the method used. Micafungin water content is not commonly quantified by the loss on drying, and the methods of semi-micro and micro determination are not interchangeable for this purpose. Further studies are required to estimate bias and include the estimate in the measurement uncertainty budget.

1. Fadeikina OV, Voropaev AA, Davydov DS, Volkova RA. Estimation of measurement uncertainty for the determination of loss on drying of biologicals. Biological Products. Prevention, Diagnosis, Treatment. 2023;23(3–1):452–62 (In Russ.). https://doi.org/10.30895/2221-996X-2023-23-3-1-452-462

2. Abdullina SG, Lira OA, Petrova IK. Coulonometric deter mination of water in medications. Pharmacy. 2011;(5):15– 6 (In Russ.). EDN: OCQRBT

3. Turundaeva AA, Olefir YuV, Luttseva AI, Gromova NN, Mamashina EA. Metrological aspects of expert evaluation of medicinal products’ quality. Bulletin of the Scientific Centre for Expert Evaluation of Medicinal Products. 2017;7(4):228–32 (In Russ.). EDN: ZWDDTR