Collaborative study for establishment of an HPLC-method for batch consistency control of recombinant interferon-alfa-2.

For interferon-alfa-2 (IFN-alfa-2) finished products a simple potency assay for batch consistency control or market surveillance studies is lacking. The European Pharmacopoeia monograph "Interferon alfa-2 concentrated solution" applies to bulk product and requires a cell culture potency assay which is neither straightforward nor are its precision and interlaboratory variability satisfying. Thus a project was initiated within the Biological Standardisation Programme to establish an HPLC assay for batch potency testing of recombinant IFN-alfa-2 finished products. In the collaborative study the suitability and transferability of an HPLC method developed in the project leaders' laboratory was checked. Twelve laboratories participated in the study. Based on a standardised method, calibration curves were established using the European Pharmacopoeia reference substances for IFN-alfa-2a and IFN-alfa-2b. The resolution between IFN-alfa and the oxidised forms of IFN-alfa was evaluated. Three marketed IFN-alfa preparations (one containing albumin) with known IFN-alfa content were used for the quantitative assay. The reproducibility of the method was not satisfactory but appears improvable with a slight modification of the method. Most laboratories succeeded in separating oxidised IFN from the main IFN peak. Five laboratories achieved very good linearity of the calibration curves while poor linearity was observed in three, mainly due to flattening of the curve at the lowest concentrations. The intercepts of the calibration curves were negative in all cases which might be due to a substantial degree of adsorption of IFN on glass and plastic surfaces. The robustness of the method concerning the choice of chromatographic column seems to be low. Using a corresponding column from a different manufacturer instead of the one proposed resulted in a loss of resolution between oxidised IFN and the main peak. Even the use of different batches of the prescribed column seemed to influence the result. The repeatability for the four consecutive injections of the three commercial IFN products does not indicate a systematic methodological error. The reproducibility (variability between laboratories) of the originally proposed method was large. Thus, the originally proposed method appears less suitable. However, the values obtained with a modified technique were less variable; in addition, an improvement of peak symmetry was obtained. Based on the results obtained with the original and the adapted method, a follow-up study with a modified technique (e.g. changes in liquid phase, application of system suitability criteria) is proposed.