Abbreviations

BNM: Bioassay in Normocythaemic Mice; BRP: Biological Reference Preparation; c BRP: Candidate BRP; CV: Coefficient of Variation; CZE: Capillary Zone Electrophoresis; EDQM: European Directorate for the Quality of Medicines; EPO: Erythropoietin; HPAEC: High-performance anion-exchange chromatography; I-number: Isoform number calculated via the peak numbering in CZE; IF: Isoform; in: Individual Isoform number shares; IU: International unit; MV: Mean value; Ph. Eur.: European Pharmacopoeia; pn: Peak area percent shares; RP-HPLCE: Reversed-phase high-performance liquid chromatography.

Introduction

The reduction, refinement and replacement (3Rs) of bioassays by physicochemical or cell based assays or other alternative methods is a common goal of the European Commission, health care and animal welfare societies, regulatory authorities, expert groups and pharmaceutical companies [1], in particular with erythropoietin [2]. Three physicochemical assays based on capillary zone electrophoresis data have up till now aimed at replacing the mouse bioassay of erythropoietin batch release according to Ph. Eur. [3], i) the “I-number assay” introduced in 2006 for the quality control of EPO medicinal products [4-6] and recently used to calculate and compare the potencies of the biological reference preparations of erythropoietin [7], ii) the “Ibio-number assay” that has likewise been introduced in 2005/2006 [5-6] however only recently been termed as such when the assay was applied to calculate the potencies of various EPO drug substance and drug product samples [8] respectively the various (candidate) EPO BRPs [9], and iii) the “reduced linear regression model” from Roche introduced in 2010/2011 [10-12] that meanwhile has been accepted by the European authorities to replace the Normocythaemic mouse bioassay of epoetin beta drug substance batch release [13]. Whereas the “I-number assay” uses the peak numbering of the EPO Isoform separated in CZE for “I-number” calculation [4-7], the “Ibio-number assay” uses factors that reflect the bioactivity of the EPO iso forms [5-6] for direct potency calculation [8-9]. And whereas these two assays use all eight EPO Isoform separated in CZE for bioactivity calculation (where applicable), the Roche assay uses only five Isoform and adds instead a term reflecting sialylation [10]. The aim of this current study was to prove that the “I-number assay”, like the “Ibio-number assay” [8-9], is suitability to predict the potency of EPO medicinal products with high precision and accuracy.

Note: It should be mentioned at this stage that the current “I-number” study in a way parallels the “Ibio-number” study reported elsewhere [8] as it is based on the same set of CZE data. Hence, the final potencies calculated via the two assays are quite comparable (mean difference = 1.2%), and the arguments in the two papers are likewise very similar (if not identical). But the “I-number assay” deserves its own presentation in a separate paper, as potency calculation according to the “I-number assay” is based on the peak numbering in CZE of the EPO Isoform, whereas potency calculation according to the “Ibio-number assay” is based on the bioactivities of these Isoform.

Materials and Methods

Materials

1. EPO cBRP3 was as described in a collaborative study published in 2007 [14], and the CZE data were taken from this study. EPO cBRP3 consisted in “a 50:50 (weight / weight) blending of 2 erythropoietin preparations corresponding to products currently available on the European market (erythropoietin-Alfa and erythropoietin-beta), with each vial containing approximately 250 µg erythropoietin” [14] – with “a potency of 35,280 IU/vial” [14], i.e. 141,120 IU/mg or 141.1 IU/µg, used throughout herein.
2. Epoetin Alfa secondary standard was from Centocor as described elsewhere [15], and the CZE data were taken from this study. Briefly, the epoetin Alfa sample from Centocor consisted in epoetin Alfa obtained from Johnson & Johnson Ortho Biotech and was used to prepare a secondary standard that has been calibrated against EPO BRP1 [15].
3. An artificial epoetin beta material was hypothetically constructed from a pool of 40 epoetin beta training batches from Roche, “derived from several variants of a well-defined manufacturing procedure which is tightly controlled within defined limits” [10]. The mean Isoform percentages measured by Roche in CZE of these 40 epoetin beta training batches were used to calculate the mean I-number (Iß_40tr) of this hypothetical epoetin beta mixture. Noteworthy, in these 40 epoetin beta training batches “batches were also included with markedly low biological potencies” [10], determined in Normocythaemic mice.
4. A hypothetical epoetin Alfa / beta 1:1 mixture was constructed from Centocor’sepoetin Alfa secondary standard [15] and the mean of Roche’s 40 epoetin beta training batches [10].

Methods

EPO sample preparation and analytical conditions of CZE were as first described in 2002 [16] and kept unchanged up to date [3].The I numbers were calculated as earlier described [4-6]. Thus, “the overall Isoform number I of EPO is the sum of the products of the individual CZE peak area percent shares (pn) and the corresponding individual Isoform numbers (n)” [4] (Formula 1)

I  =  p₁ × 1 + p₂ × 2 + p₃ × 3 + p₄ × 4+ p₅ × 5+ p₆ × 6 + p₇ × 7+ p₈ × 8 (Formula 1).

The bioactivity of a sample (Biosample) was calculated from the I-number of the sample (I_sample), the I-number of cBRP3 (I_cBRP3) and the stated bioactivity of BRP3 (expressed by the author as 141.1 IU/µg), as shown in (Formula 2).

Biosample = I_sample / I_cBRP3 * Bio_BRP3 (Formula 2)

The I-number of cBRP3(I_cBRP3) was retrospectively calculated by the author according to Formula 1 from the CZE data given in the collaborative study published in 2007 [14]. This I_cBRP3-number was correlated with the stated bioactivity of EPO BRP3. Note: The outlier according to Grubbs [17] and according to Dixon [18] revealed by the author in the course of this retrospective I-number calculation was disregarded, which is specified in the footnote to (Table 1). The I number of Centocor’sepoetin Alfa secondary standard (Iα) was retrospectively calculated by the author according to (Formula 1), using the CZE data published by Centocor [15]. The mean I number of Roche’s 40 epoetin beta training batches (Iß_40tr) was retrospectively calculated by the author according to (Formula 1), using the mean CZE data published by Roche [10]. Thus, the average peak area values of Isoform 1-8 (numbering according to Ph. Eur. [3,16]) respectively IF_8 – IF_1 (numbering by Roche [10] ) of these 40 epoetin beta training batches were regarded as the values of an artificial epoetin beta concentrated solution pool material. Note: Compared with the Roche papers [10-12], the Isoform are numbered in reverse order in this paper, which uses the order of the Ph. Eur. monograph [3,16] (Figure 1). The I-number of the hypothetical epoetin Alfa / beta 1:1 mixture (Iα/ß_40tr) was gained from the I-number of Centocor’sepoetin Alfa secondary standard (Iα) and the mean I-number of Roche’s 40 epoetin beta training batches (Iß_40tr) as the arithmetic mean (Formula 3).

1. α/ß_40tr = (Iα + Iß_40tr) / 2 (Formula 3)

The bioactivity of this hypothetical epoetin Alfa / beta 1:1 sample was again calculated via IcBRP3 and the stated bioactivity of BRP3 (cp. Formula 2).

Results

A reference electropherogram is shown in (Figure 1).

Note: As mentioned in the “Methods”, the peak numbering used by Roche [10-12] was inverse, compared to the peak numbering according to Ph. Eur. [3,16] respectively the Ibio-number assay [8-9], and the isoform numbers had to be considered accordingly (Table 3). Centocor’sepoetin Alfa secondary standard and Roche’s mean of the 40 epoetin beta training batches, when theoretically mixed 1:1, yielded a mean I-number of Iα/ß_40tr = 527.3 which, calculated via (Formula 2), corresponded with a bioactivity of 140.0 IU/µg that matched the stated bioactivity of BRP3with an accuracy of 99.2% (difference = ‑0.8%) (Table 4)

Discussion

Although the I-number assay has been published more than 10 years ago [4], no data has been provided in the literature thus far that shows that the I-number assay allows to calculate the potency of EPO medicinal products, as suggested [4-6], and this was the aim of the current study. To this, the bioactivity of the samples was calculated via the I-number of cBRP3and the stated bioactivity of BRP3 [14], which laid the basis for the potency calculations of the other samples (cp. Formula 2). This basis is justified by the facts that i) the I-number of cBRP3 could be determined from the data of the 2007 collaborative study of 10 participating laboratories with an inter-laboratory precision of CV = 0.7% (Table 1) and ii) the potency stated for BRP3 was without any doubt – in contrast to the potencies stated for EPO BRP1 and EPO BRP2, as shown elsewhere [7]. This precise set of data primarily relies on the precision and accuracy of CZE. But exactly for that reason, the precision of the I-number determination of EPO samples simultaneously reflects the bioactivity of the samples in CZE in a likewise precise manner [8], which was particularly apparent from the hypothetical epoetinalfa / beta 1:1 mixture that met the stated bioactivity of EPO BRP3 with an amazing accuracy of 99.2% (Table 4). Attempts to calculate the bioactivity of EPO samples via the stated potency of BRP2 and the I-number of BRP2 (cp. Formula 2) have not been satisfactory, most likely due to the fact that the bioactivity of EPO BRP2 has been stated about 10% too low [7,9]. And attempts to compare the calculated bioactivity of Centocor’s secondary standard (144.8 IU/µg; cp. Table 2) with the bioactivity given by Centocor [15] were deliberately omitted as Centocor’sepoetin Alfa secondary standard has been calibrated against EPO BRP1, the stated bioactivity of which has been stated about 5% too low [7,9].

As already mentioned elsewhere [8], “the mean biological activity determined in Normocythaemic mice (BNM) of Roche’s 40 epoetin beta training batches was communicated as 196 “IU/µg” Although not explicitly mentioned by Roche, this value must be read as 196 “IU/µg protein”, concluded by the author from reference [20], and must therefore be multiplied by 0.60 in order to get “IU/µg EPO” glycoprotein, reflecting the protein content of the erythropoietin glycoprotein”[8]. Thus, the mean epoetin beta sample ended up with a BNM of 117.6 “IU/µg EPO” that differed from the bioactivity calculated via the I-number assay (135.1 IU/µg = 100%; cp. Table 3) by ‑14.9%. This difference may be due to the fact that in these 40 epoetin beta training batches “batches were also included with markedly low biological potencies” [10]. This difference may, however, also (and more likely) be attributable to the fact the majority of the 40 epoetin beta training batches (if not all) have been calibrated against EPO BRP2, the stated potency of which has been questioned [7,9]. Considering these inconsistencies revealed for Centocor’sepoetin Alfa secondary standard (due to its calibration against EPO BRP1) respectively for Roche’s mean of the 40 epoetin beta training batches (due to their likely calibration against EPO BRP2), it is quite amazing that the bioactivity of the epoetin Alfa / beta 1:1 mixture calculated via the I-number assay (140.0 IU/µg) was in accordance with the stated bioactivity of EPO BRP3 (141.1 IU/µg) (difference = -0.8%; Table 4), which may be taken as a proof of principle that the bioactivity of EPO samples may be precisely calculated via the I-number assay as previously suggested [4].

Conclusion

The data presented in this retrospective analysis by the author of the CZE data of epoetin Alfa and epoetin beta samples has shown that the “I-number assay” - like the “Ibio-number assay” [8] –is suitable to assess the bioactivity of EPO medicinal products with high precision and accuracy, which should be confirmed in further studies comprising in particular samples from EPO batch release. In combination with other physicochemical assays such as RP-HPLCE, peptide mapping and HPAEC, the “I-number assay” – like the “Ibio-number assay” [8] – offers an opportunity to replace the highly contested (consumption of animals) and highly inaccurate mouse bioassay of EPO batch release.

Note: Lab 4 proved to be an outlier according to Grubbs (α = 0.01) and according to Dixon (α = 0.02) and was therefore disregarded. I_cBRP3 = 531.5 is related to the stated potency of BRP3 [14], expressed as 141.1 IU/µg.

Table 1: cBRP3 I-number calculation and inter-laboratory precision

Basis: (Table 3) dof Behr-Gross et al. 2007 [14] - Isoform distribution (in %) of cBRP3 uncorrected for migration time Roche’s 40 epoetin beta training batches provided a mean I-number of Iß_40tr = 509.0 which, calculated via (Formula 2), corresponded with a bioactivity of 135.1 IU/µg (Table 3).

Bioactivity calculation via (Formula 2)

Bio_α = I_α /I_cBRP3 * Bio_BRP3= 545.6 / 531.5 *141.1 IU/µg = 144.8 IU/µg

Table 2: Epoetin alpha secondary standard, I-number calculation

Basis: CZE data from (Table 2) of Zhang et al. [15]

1. Balls M, Halder M (2002) Progress in applying the three Rs of Russel & Burch to the testing of biological products. Dev Biol (Basel) 111: 3-13.
2. Bristow A (2008) Possible approaches for replacement of erythropoietin in vivo potency assay. In: EDQM, editors. Alternatives to animal testing: New approaches in the development and control of biologicals. Proceedings of the International Symposium: Dubrovnik, Croatia; 2008 Apr 23-24. Strasbourg, France: Council of Europe, 87-92.
3. (2017) Erythropoietin concentrated solution, monograph 1316, Ph. Eur. 9.0. Strasbourg, France: Council of Europe, 2391-2395.
4. Hermentin P (2006) Isoform number I – a new tool to evaluate the quality of erythropoietin. Pharmeuropa Scientific Notes 1: 37-40.
5. Hermentin P (2009) Method for characterizing the glycosylation of sialoglycoproteins via an isoform number I.  EP 1 943 524 B1.
6. Hermentin P (2012) Method for characterisation of the quality of sialoglycoproteins via an isoform number I. US 8150630 B2.
7. Hermentin P (2017) Evidence that the potencies of erythropoietin biological reference preparations batch 1 and batch 2 have been stated too low. J Chromatogr Sep Tech, accepted
8. Hermentin P (2017) Ibio-number assay: A physicochemical assay that predicts the bioactivity of erythropoietin with high precision and accuracy and may replace the mouse bioassay in the quality control of EPO batch release. Pharm Anal Acta 8: 533.
9. Hermentin P (2017) Ibio-Number Assay and Erythropoietin Bioactivity: Comparison of the Calculated and the Stated Potencies of the Biological Reference Preparations of Erythropoietin. Pharm Anal Acta 8: 542.
10. Zimmermann H, Gerhard D, Hothorn LA, Dingermann T (2011) An Alternative to Animal Testing in the Quality Control of Erythropoietin. PharmeuropaBio&SN 1: 66-80.
11. Zimmermann H, Gerhard D, Hothorn LA, Dingermann T (2010) Eine Alternative zum Tierversuch in der Qualitätskontrolle von Erythropoietin (Teil 1). Pharmind 72: 708-712.
12. Zimmermann H, Gerhard D, Hothorn LA, Dingermann T (2010) Eine Alternative zum Tierversuch in der Qualitätskontrolle von Erythropoietin (Teil 2). Pharmind 5: 884-896.
13. Schnueriger F (2013) CMC Strategy Forum Europe, May 6-8; Hilton Old Town, (Prague, Czech Republic)
14. Behr-Gross ME, Daas A, Burns C, Bristow AF (2007) Collaborative study for the establishment of erythropoietin BRP batch 3. Pharmeuropa Bio 1: 49-66.
15. Zhang J, Chakraborty U, Villalobos AP, Brown JM, Foley JP (2009) Optimization and qualification of capillary zone electrophoresis method for glycoprotein isoform distribution of erythropoietin for quality control laboratory. J Pharm Biomed Anal 50: 538-543.
16. (2002) Erythropoietin concentrated solution, monograph 1316, Ph. Eur. 4.0. Strasbourg, France: Council of Europe, 1123-1128.
17. Grubbs FE (1969) Procedures for detecting outlying observations in samples. Technometrics 11: 1-21.
18. Dixon WJ (1950) Analysis of extreme values. Ann Math Stat 21: 488-506.
19. Behr-Gross ME, Daas A, Bristow AF (2004) Collaborative study for the establishment of erythropoietin BRP Batch 2. Pharmeuropa Bio 1: 23-33.
20. Burg J, Sellinger KH, Haselbeck A, Koll H (2014) Erythropoietin with high specific activity. EP 1 037 921 B1; German patent WO1999/28346 (1998)