- Email: [email protected]
Role of raw materials in biopharmaceutical manufacturing: risk analysis and fingerprinting
Available online at www.sciencedirect.com
ScienceDirect Role of raw materials in biopharmaceutical manufacturing: risk analysis and fingerprinting Anurag S Rathore, Deepak Kumar and Nikhil Kateja Accurate fingerprinting of critical raw materials that have significant impact on process performance and product quality is a necessary precursor for implementation of QbD in process and product development. This article presents a review of major developments in this space in the last 10 years, with a special emphasis on those in last 5 years. A step by step approach for managing raw materials in the QbD paradigm has been proposed. We think that it is necessary for the biotech industry to better manage variability originating from raw materials if holistic implementation of QbD is to be achieved.
Address Department of Chemical Engineering, Indian Institute of Technology, Delhi, India
Raw materials used in biopharmaceutical manufacturing are a diverse source of materials and include media components for cell culture and fermentation, fine chemicals for purification and chemical modification processes, and excipients used in a formulation of a final drug product [3]. They also include product contact materials like the plastic used in disposable bags. Variability in raw material can come from a change in a chemical or physical characteristic of the material [4]. This variability can affect characteristics and quality of drug product and potentially impact the product’s safety, stability, and efficacy [5]. At times, this has resulted in major adverse events and even resulted in drug recalls [6]. For instance, presence of glass particulates in drug product have led to more than 20 product recalls in the last few years, including the recall of Procrit and Epogen injections [7].
Corresponding author: Rathore, Anurag S ([email protected])
Types of raw material variability Current Opinion in Biotechnology 2018, 53:99–105 This review comes from a themed issue on Pharmaceutical biotechnology Edited by Amanda Lewis and Nripen Singh
https://doi.org/10.1016/j.copbio.2017.12.022 0958-1669/ã 2017 Published by Elsevier Ltd.
Variability in raw materials can be subdivided into three broad categories. The first category includes trace impurities that alter the quality of the biotherapeutic, either by directly modifying it or by catalysing its modification such as peroxides, aldehydes, reducing sugars, and catalytically active metal ions. The second category consists of trace impurities that are themselves toxic to humans, such as lead and aluminium. The third group comprises of microorganism contaminants (and their associated endotoxins) that lead to variabilities in the bioburden of raw materials and can cause severe immunological responses in patients [5]. Each of these require their own approach for monitoring and control.
Sources of variability Introduction Biopharmaceutical manufacturing, because of its complex nonlinear nature, is fraught with a myriad of process variations that can impact safety and efficacy of the drug. Since the introduction of concepts such as process characterization and design of experiments (DoE) over two decades ago, the biopharmaceutical industry has created and demonstrated considerable expertise in unravelling how the process affects the product. However, the role of raw materials (RM) has been somewhat overlooked and as a result has become the primary source of variability in process performance and product quality. The growing significance of the role of raw materials in the process control strategy is evident from the ICH Q8 guideline, which suggests that in the Quality by Design (QbD) framework the manufacturer must understand all sources of variability including the raw materials [1,2]. www.sciencedirect.com
Cell culture processes used to make recombinant proteins use complex growth media. Although some cells can be maintained in a basal medium with no supplementation, majority of cells require addition of up to 100 components such as hormones, growth factors, vitamins, peptides, amino acids and hydrolysates to grow [8]. Naturally derived media can contain a large number of compounds [9,10]. They also have micronutrients in trace amounts. Variability may creep in the media because of several reasons. Degradation of raw materials, impurities, and contaminants present in the media; non-uniformity in milling and blending during manufacturing of large batches of media due to loss of micronutrients; and inconsistency of content levels and changes in raw material sourcing due to constraints in availability can result in raw material variability. Though there has been a shift Current Opinion in Biotechnology 2018, 53:99–105
100 Pharmaceutical biotechnology
towards use of chemically defined media in the last decade, natural media continues to be used by the industry and continues to offer a challenge in maintaining a consistent product quality. Even chemically defined media have a complex composition which necessitates its characterization to achieve consistent process performance. Inconsistency in excipient quality can have an adverse impact on drug product quality. Excipient-derived contaminants can be listed in the following categories — trace metals, peroxides, aldehydes, reducing sugars/polyols, and organic acids. The purity levels of these excipients may vary significantly and affect the product quality/ stability to different degrees. Even with pure excipients, contaminants may be generated during storage through various degradation pathways [11]. Polysorbates, for instance, undergo autooxidation which can influence the stability of a biopharmaceutical product [12]. Leachates from the container/closure system have been shown to have significant impact on product quality as evidenced by recent issues associated with glass lamellae in glass vials [13], tungsten with prefilled syringes [14], silicon from vial stoppers or prefilled syringe barrels [15], leachates from filters [16], and shedding of nanoparticles from a filling pump’s solution-contact surfaces [17,18]. Lot-to-lot variability in chromatography resins can also result in unanticipated changes in yield or product quality [19].
QbD based approach for managing raw materials International Conference on Harmonization (ICH) in its Q7 good manufacturing practice [20], Q8 pharmaceutical development [2], Q9 quality risk assessment [21] and Q10 pharmaceutical quality system [22] guidelines have stipulated stringent requirements regarding product quality. Current practice for raw material analysis has been described in ICH Q7 Guideline [20]. The document states that materials used to prepare active pharmaceutical ingredients (both small molecules and biologics) need to have the identity of each batch confirmed on receipt and a Certificate of Analysis (C of A) provided from the supplier. Pharmacopeial and formulary monographs such as the USP/NF, EuP, JP, and BP provide standardized test methods for the most common and widely used materials. Manufacturers take various approaches towards testing compliance of raw materials. Some qualify a raw materials supplier by performing an initial detailed vendor audit followed by an annual qualification consisting of testing as per the pharmacopeial monograph on three lots of raw material. If the qualification lots test successfully, then subsequent material shipments will require only monograph identification testing. However, companies that take a more conservative approach to raw materials release require full monograph testing for each lot of supplied material [16]. In addition, the supplier must be qualified as suitable based on audits of their facility, Current Opinion in Biotechnology 2018, 53:99–105
Figure 1
Critical/ key raw material
Understand “how” it affects the process and the product
Identify critical attributes of the raw material
Create an analytical method for characterizing the critical attribute
Collect data to fingerprint the raw material Implement approach to enable identification of acceptable and unacceptable raw material lots Current Opinion in Biotechnology
Managing raw materials in the QbD paradigm.
their analytical results must be confirmed to be reliable, and a sampling plan is needed for each incoming material [19]. A key challenge is the relatively large number of raw materials that are used in biopharmaceutical manufacturing (typically > 100). To effectively deal with such a large number of raw materials, a multistep Quality by Design (QbD) based approach has been proposed [23,24] (Figure 1). It is recommended that this evaluation be performed at every major milestone of product development (First in Animal, Phase I, Phase II, Phase III, Validation). Although an arduous task when performed the first time, the effort significantly reduces in the following times as most of the raw materials are the same amongst products of similar kind (e.g. monoclonal antibodies): A risk assessment is performed encompassing all raw materials used in the process. There are a number of different risk assessment tools available in a range of detail and complexity, and it is important to use a methodology suited to the purpose of the assessment [24]. Appropriate stakeholders including process development, manufacturing, quality assurance, and quality control are included in this assessment. The team discusses if the raw material is likely to impact the process performance and if it is likely to impact product quality. Those who are likely to impact both are termed as critical raw materials, those who just impact the process and not the product are termed as key raw materials, and those who are not expected to impact either are called as non-key raw materials [23]. Critical raw materials are therefore thoroughly characterized and their mechanisms of process interactions www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 101
well understood. The manufacturer should understand ‘how’ the raw material is impacting the process and the product. Raw material attributes that can be used to characterize this impact need to be identified and acceptance criteria for these attributes need to be established. Analytical tests need to be developed to monitor these attributes to ensure that each critical raw material lot meets the respective acceptance criteria before its use in the process [23]. Key raw materials can be characterized same as critical raw materials, but while critical raw materials undergo this analysis for every product, key raw materials need to be characterized once for a platform. Every time a new key raw material is added to the platform, it undergoes a thorough characterization before its use [23]. Non-key raw materials include the remaining (and majority of) raw materials and are handled via the internal Quality System of the biopharmaceutical manufacturer as in traditional pharmaceutical manufacturing [23].
Tools for assessment of variability Raw material fingerprinting based on detailed characterization of how the raw material impacts the process and product quality offers a potential solution to the abovementioned conundrum. It can facilitate efficient and effective monitoring of raw material quality and thus prevent significant adverse impact on product quality.
Analytical tools Assessment of raw materials thus involves use of a wide spectrum of analytical techniques like spectroscopic, chromatographic, and enzyme-based tools. Spectroscopic techniques are rapid and have been demonstrated to be quite effective in monitoring variation in raw material quality and hence been the tools of choice. Raman [25,26], near-infrared (NIR) [27–29], Fourier-transform infrared (FTIR), mid-infrared (MIR) [30,31], fluorescence spectroscopy [32] and nuclear magnetic resonance (NMR) spectroscopy [33,34] have been used for identification and characterization of raw materials. Researchers have recently assessed the capability and suitability of several spectroscopic techniques for adulteration detection in tromethamine (Tris) and tromethamine hydrochloride (Tris-HCl), two of the commonly used excipients in biopharmaceutical formulations [35]. Near-infrared (NIR), Raman (handheld Raman and Raman-microscope), Fourier Transform Infrared (FTIR), and Nuclear Magnetic Resonance (NMR) have been evaluated and their merits and demerits have been highlighted in terms of speed of analysis and ability to differentiate adulterated from unadulterated samples [35]. In other studies, mass spectroscopy (MS) has been shown to offer significantly high sensitivity, but is expensive and also cumbersome to perform for routine raw material release testing. Surface enhanced Raman www.sciencedirect.com
spectroscopy (SERS) with gold nanoparticles has been successfully used for detection of trace melamine in raw materials used for protein pharmaceutical manufacturing [36]. A recent review provides an extensive list of tools for examining the sources of variability in cell culture media used for biopharmaceutical manufacturing along with their merits and demerits [8]. This information has been summarized in Table 1. Other alternatives include blotting, capillary electrophoresis (CE) [37], enzymatic methods (e.g. enzyme-linked immunosorbent assay (ELISA)), high-performance liquid chromatography (HPLC) [38], gas chromatography, and SDS-PAGE [39]. Proteomic analysis has also been proposed as a tool for assessing complex raw materials such as serum. Proteomic techniques were used to understand the lot-to-lot variability with respect to impact on growth properties of the cell culture process [40]. As an extension of this approach, proteomics combined with metabolomics has been used to identify the root cause for the lower productivity during scale up of a cell culture process. The analysis attributed hypoxia, resulting due to elevated Cu levels (as a contaminant in media component) as the contributing factor for the observed decline in the performance of the cell culture [41].
Role of statistics in data analysis Datasets originating from the manufacturing of biopharmaceuticals, including spectroscopic tools, are often complex and hence univariate or bivariate analysis is insufficient and likely to result in erroneous conclusions [42]. Multivariate data analysis (MVDA) is required in most cases to delineate relevant information from large multifactorial and multi-collinear data sets. MVDA involves use of tools for exploratory data analysis (data mining), classification (cluster analysis), regression analysis, and predictive modelling [43]. Projection methods are commonly used to overcome multidimensionality of the data set, multicollinearity, missing data, and other variations introduced by deviating factors such as experimental error and noise [44]. One such technique is Principal Component Analysis (PCA), which finds the major variability directions within the data set to explain the overall system behaviour with a few components by decomposing the covariance matrix of the data set [19]. When there is a need to explore multivariate correlations against one or more response variables, a variety of regression techniques are used like principal components regression, partial least squares (PLS), and ridge regression. PLS rearranges process variables space and response variables space by reducing their size while maximizing the covariance between them [19]. Characterization data of raw material library may include one or more categorical response variables and this necessitates the use of a variant of PLS called PLS Discriminant Analysis (PLSDA) [3]. Continuous variable data contains both predictive and uncorrelated information, Orthogonal-PLS (OPLS) is used which leads to improved diagnostics Current Opinion in Biotechnology 2018, 53:99–105
102 Pharmaceutical biotechnology
Table 1 Comparison of different analytical tools for testing raw material variability. Techniques Spectroscopic analytical tools
Methods MIR
NIR
Raman
Fluorescence
NMR Chromatographic techniques
LC, GC, CE, IC
Pros
Cons
Broad range of wavelength, non-invasive, ability to quantify analytes present in very low concentrations, real time monitoring Higher penetration, less absorbance compared to MIR, probes are more durable and cheaper compared to those of MIR, minimal requirement for sample preparation, hand-held, real time monitoring
Less penetration, probes are more fragile and expensive
Can be used for powder and liquid samples, selective bands higher than in case of NIR, sub-sampling possible, can be used for low concentrations by virtue of SERS, hand held, real time monitoring Effective at low concentrations, useful for samples that have low spectral resolution and signal strength by means of fluorescence anisotropy, real time monitoring High specificity and reproducibility, fast and higher throughput, wide dynamic range Liberty to use suitable detection method (MS, spectroscopic) along with separation technique so that specific media components can be addressed, detection of trace samples is possible using more sensitive capillary methods
More convenient only for powdered samples, not recommended for low concentration samples, not suitable for sensitive samples owing to highenergy laser Wavelength selection is crucial due to possibility of interference, high variability causing less reproducibility Possibility of quenching and overlap, can be used only for samples with fluorescent properties High operational cost, requires space Involves invasion and destruction, high cost of equipment, comprehensive mass spectral databases for metabolites is lacking
Source: Adapted from [8].
and interpretability of models [3]. Batch modelling based on PLS and PCA are commonly used in biotech industries to deal with batch operation data [3]. Design of experiments (DoE) is the approach of choice for planning experimental studies such that the relevant process information can be obtained using a minimum number of experiments. It provides information about the interaction of factors and the way total system behaves. It shows how interconnected factors respond over a wide range of values, without requiring the testing of all possible values directly. Thus, it decreases quantity of effort and cost of many runs without compromising the quality of results [45]. MVDA provides an efficient approach for analysis of spectral data. Applications include pre-processing methods to reduce and correct interferences in spectroscopic data such as overlapped bands, baseline drifts, scattering, and path length variation [46]. Statistical approaches are also used for calibration and diagnostics and variable selection, statistic result calculation to build representative and reliable models, and model validation and integration for achieving rigorous prediction and real-time product quality and process monitoring [3,43].
Recent trends in raw material fingerprinting In recent studies, a combination of spectroscopy and chemometrics is used to characterize raw material variability and to predict final product quality based on information about the raw material. Since cell culture Current Opinion in Biotechnology 2018, 53:99–105
raw materials are complex mixtures, it is difficult to identify and quantify every compound present. Spectroscopic tools, such as Raman, MIR, NIR and fluorescence have shown to be adequate when combined with chemometrics for characterizing complex raw materials [8,47]. NIR spectroscopy coupled with PCA has been demonstrated to be effective in fingerprinting a cell culture media component and thereby distinguish between good and poor performing media lots [27]. In another recent study by Lee et al., NIR spectra of multiple soy hydrolysate lots were analyzed and by using a chemometrics approach, it was shown that NIR spectra could be used to reveal lot-to-lot variability as well as vendor-to-vendor differences. Partial least square regression (PLSR) based prediction models for estimating cell growth and productivity of mammalian cell cultures were constructed from the near-infrared spectra [48]. Furthermore, it is important to assess the performance of multiple spectroscopic platforms when determining the impact of a new raw material on cell culture performance. Researchers have demonstrated that both NIR and 2D fluorescence spectroscopy were able to detect compositional changes in basal media and that aging of basal media was found to affect cell culture performance, concluding that both NIR and 2D fluorescence spectroscopy can assess quality of basal media after storage [49]. Another similar study demonstrated capability of NIR and 2D fluorescence spectroscopy for detection of lot-to-lot variability in cultivation media component [50]. www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 103
Figure 2
(b)
Raw Materials:
(a) Group 1
P GMP
Pilot Runs Clinical Runs
Basal Powders
(c) Group 2
(d)
Group B
Group A
Current Opinion in Biotechnology
Case study in raw material management in the QbD paradigm. (a) Multivariate data analysis indicated differences in clinical manufacturing (Group 2) and commercial manufacturing (Group 1). (b) Root cause analysis indicating differences in basal media as the primary source of process variability. (c) Development of a NIR based method for fingerprinting of raw material. (d) NIR method is able to distinguish between ‘good’ (Group B) and ‘bad’ (Group A) lots. Source: Adapted from [16,30].
Researchers argue that in a lot of cases a single spectroscopy may be insufficient for characterization of a raw material and multiple, orthogonal spectroscopy techniques may be required in conjunction to achieve the required specificity and accuracy [47]. A variety of combinations of spectroscopic and non-spectroscopic techniques have been suggested for characterization of complex raw materials. Researchers have demonstrated use of a gas chromatography–mass spectrophotometry (GC–MS) method for the micro quantitation of lipids, determined as www.sciencedirect.com
fatty acid methyl esters (FAME), in complex serum-free cell culture media [51]. Other researchers have suggested the use of a combination of affinity chromatography (Protein G or Protein A), SDS-PAGE, and peptide mass fingerprinting for analyzing cell culture media for the presence of contaminating bovine antibodies [39]. Liquid chromatography–mass spectroscopy (LC–MS) techniques have also been applied for the simultaneous quantification of multiple media components [52]. Recently, combinations of different spectroscopic tools supported Current Opinion in Biotechnology 2018, 53:99–105
104 Pharmaceutical biotechnology
by chemometrics based data fusion approach have also been explored. In one such study multiple spectral platforms (NIR, Raman, fluorescence and X-ray fluorescence) were compared with each other in terms of their estimation capability and then integrated into a unifying prediction model. The results showed that data fusion models exhibit better prediction accuracy than individual spectroscopic methods, demonstrating the synergetic effects of data fusion in characterizing the raw material quality [53].
Case study in raw material fingerprinting Researchers from a major biotechnology company reported an instance of significant issues faced during technology transfer and scale-up of a mammalian cell culture step from clinical scale (2000 L) to commercial scale (15 000 L). This case study is illustrated in Figure 2. The CQA of the product manufactured at commercial scale was significantly different from that manufactured at clinical scale. An extensive investigation was performed to identify the root cause [30]. The root cause was identified as the difference in the basal media supplied by the same vendor for the two scales of operation. Further investigation at the vendor site yielded the information that grouping of medium components during the milling and blending process varied with the scale of production and media type, resulting in different characteristics. Considering the significant impact of the raw material, a NIR based method was created for screening of lots of basal medium powders based on their impact on process performance and product attributes [16]. A combined NIR/MVDA approach made it possible to finger print the raw material to distinguish between good and poor performing media lots.
Summary Regulatory feedback has resulted in a major shift in how the biopharmaceutical industry looks at raw materials and their management. Most major manufacturers are applying the principles of risk analysis and good science towards identification of critical raw materials and their analytical characterization. The concept of accurate fingerprinting of critical raw materials has also gained steam over the last decade. The step by step approach for managing raw materials in the QbD paradigm as proposed in this article could result in a significant improvement in our understanding of this topic. Although a lot has been accomplished in the last decade, we are still far from reaching the same rigor with respect to raw material characterization as we are with process characterization. Constant addition of new raw materials to the system will ensure that the task remains challenging.
Conflict of interest statement The authors declare no financial or commercial conflict of interest. Current Opinion in Biotechnology 2018, 53:99–105
Acknowledgements This work was funded by the Center of Excellence for Biopharmaceutical Technology grant from Department of Biotechnology, Government of India (number BT/COE/34/SP15097/2015).
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest 1. Rathore AS, Winkle H: Quality by design for biopharmaceuticals. Nat Biotechnol 2009, 27:26-34. Seminal paper on application of QbD for biopharmaceuticals. The paper presents the underlying concepts of QbD as well as presents a step by step approach for its application. 2.
Guideline ICHHT: Pharmaceutical development Q8 (R2). Curr Step 2009, 4.
3.
Undey C, Low DC, Menezes J, Koch JM (Eds): PAT Applied in Biopharmaceutical Process Development and Manufacturing. CRC Press; 2012.
4.
Chalk S: Raw material variability. Biopharm Int 2014, 27:38-39.
5.
Challenger CA: Biopharma takes on raw material variability. Biopharm Int 2016, 29:20-24.
6.
Ebbers HC, de Tienda NF, Hoefnagel MC, Nibbeling R, MantelTeeuwisse AK: Characteristics of product recalls of biopharmaceuticals and small-molecule drugs in the USA. Drug Discov Today 2016, 21:536-539.
7.
Reynolds G, Paskiet D: Glass delamination and breakage. BioProcess Int 2011, 9:52-57.
8.
McGillicuddy N, Floris P, Albrecht S, Bones J: Examining the sources of variability in cell culture media used for biopharmaceutical production. Biotechnol Lett 2017, 39:1-17. Contribution of lot of lot variability in quality of serum and other animal or plant derived components to variability in cell culture processes is well known. This preview presents a summary of the various sources of variability. 9.
Yao T, Asayama Y: Animal-cell culture media: history, characteristics, and current issues. Reprod Med Biol 2017 http://dx.doi.org/10.1002/rmb2.12024.
10. Arora M: Cell culture media: a review. Mater Methods 2013, 3:175. 11. Wang W, Ignatius AA, Thakkar SV: Impact of residual impurities and contaminants on protein stability. J Pharm Sci 2014, 103:1315-1330. 12. Kerwin BA: Polysorbates 20 and 80 used in the formulation of protein biotherapeutics: structure and degradation pathways. J Pharm Sci 2008, 97:2924-2935. 13. Wen Z-Q, Torraca G, Masatani P, Sloey C, Phillips J: Nondestructive detection of glass vial inner surface morphology with differential interference contrast microscopy. J Pharm Sci 2012, 101:1378-1384. 14. Liu W, Swift R, Torraca G, Nashed-Samuel Y, Wen Z-Q, Jiang Y, Vance A, Mire-Sluis A, Freund E, Davis J et al.: Root cause analysis of tungsten-induced protein aggregation in pre-filled syringes. PDA J Pharm Sci Technol 2010, 64:11-19. 15. Thirumangalathu R, Krishnan S, Ricci MS, Brems DN, Randolph TW, Carpenter JF: Silicone oil- and agitation-induced aggregation of a monoclonal antibody in aqueous solution. J Pharm Sci 2009, 98:3167-3181. 16. Jameel F, Hershenson S, Khan MA, Martin-Moe S (Eds): Quality by Design for Biopharmaceutical Drug Product Development. Springer; 2015. 17. Tyagi AK, Randolph TW, Dong A, Maloney KM, Hitscherich C, Carpenter JF: IgG particle formation during filling pump operation: a case study of heterogeneous nucleation on stainless steel nanoparticles. J Pharm Sci 2009, 98:94-104. www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 105
18. Carpenter JF, Randolph TW, Jiskoot WIM, Crommelin DJA, Middaugh CR, Winter G, Fan Y-X, Kirshner S, Verthelyi D, Kozlowski S et al.: Overlooking subvisible particles in therapeutic protein products: gaps that may compromise product quality. J Pharm Sci 2009, 98:1201-1205. 19. Rathore AS, Mhatre R: Quality by Design for Biopharmaceuticals: Principles and Case Studies. Wiley; 2009. 20. Guideline ICH: Q7 good manufacturing practice guide for active pharmaceutical ingredients. Curr Step 2000, 4. 21. Guideline ICH: Q9 quality risk management. Curr Step 2005, 4. 22. Guideline ICH: Q10 pharmaceutical quality system. International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use. 2008. 23. Rathore AS, Low D: Managing raw materials in the QbD paradigm, part 1: understanding risks. Biopharm Int 2010, 23:34-40. The authors present a QbD based approach for management of raw materials for the biopharmaceutical industry. 24. Rathore AS, Low D: Managing raw materials in the QbD paradigm, part 2: risk assessment and communication. Biopharm Int 2010, 23:48-57. The authors present a approach towards risk assessment for biopharmaceutical raw materials, along with examples of risk assessment tools that can be used. 25. Herrero AM, Jime´nez-Colmenero F, Carmona P: Elucidation of structural changes in soy protein isolate upon heating by Raman spectroscopy. Int J Food Sci Technol 2009, 44:711-717. 26. Li B, Ryan PW, Ray BH, Leister KJ, Sirimuthu N, Ryder AG: Rapid characterization and quality control of complex cell culture media solutions using Raman spectroscopy and chemometrics. Biotechnol Bioeng 2010, 107:290-301. 27. Kirdar AO, Chen G, Weidner J, Rathore AS: Application of nearinfrared (NIR) spectroscopy for screening of raw materials used in the cell culture medium for the production of a recombinant therapeutic protein. Biotechnol Progr 2010, 26:527-531. 28. Rhiel M, Ducommun P, Bolzonella I, Marison I, Von Stockar U: Real-time in situ monitoring of freely suspended and immobilized cell cultures based on mid-infrared spectroscopic measurements. Biotechnol Bioeng 2002, 77:174-185. 29. Rhiel M, Cohen MB, Murhammer DW, Arnold MA: Nondestructive near-infrared spectroscopic measurement of multiple analytes in undiluted samples of serum-based cell culture media. Biotechnol Bioeng 2002, 77:73-82. 30. Roychoudhury P, Harvey LM, McNeil B: The potential of mid infrared spectroscopy (MIRS) for real time bioprocess monitoring. Anal Chim Acta 2006, 571:159-166. 31. Hashimoto A, Yamanaka A, Kanou M, Nakanishi K, Kameoka T: Simple and rapid determination of metabolite content in plant cell culture medium using an FT-IR/ATR method. Bioprocess Biosyst Eng 2005, 27:115-123. 32. Ryan PW, Li B, Shanahan M, Leister KJ, Ryder AG: Prediction of cell culture media performance using fluorescence spectroscopy. Anal Chem 2010, 82:1311-1317.
using surface-enhanced Raman spectroscopy (SERS) with gold nanoparticles. Appl Spectrosc 2011, 65:514-521. 37. Simonelli JK, Lee AH, Gallea PE, Cheng KC: Mapping and identification of biopharmaceutical grade cell culture media by MEKC. Chromatographia 2007, 66:977-981. 38. Rathore AS, Bhambure R, Ghare V: Process analytical technology (PAT) for biopharmaceutical products. Anal Bioanal Chem 2010, 398:137-154. 39. Rasmussen LK, Larsen YB, Hojrup P: Characterization of different cell culture media for expression of recombinant antibodies in mammalian cells: presence of contaminating bovine antibodies. Protein Expr Purif 2005, 41:373-377. 40. Zheng X, Baker H, Hancock WS, Fawaz F, McCaman M, Pungor E: Proteomic analysis for the assessment of different lots of fetal bovine serum as a raw material for cell culture. Part IV. Application of proteomics to the manufacture of biological drugs. Biotechnol Prog 2006, 22:1294-1300. 41. Gao Y, Ray S, Dai S, Ivanov AR, Abu-Absi NR, Lewis AM, Huang Z, Xing Z, Borys MC, Li ZJ, Karger BL: Combined metabolomics and proteomics reveals hypoxia as a cause of lower productivity on scale-up to a 5000-liter CHO bioprocess. Biotechnol J 2016, 11:1190-1200. 42. Kourti T: The process analytical technology initiative and multivariate process analysis, monitoring and control. Anal Bioanal Chem 2006, 384:1043-1048. 43. Kirdar AO, Green KD, Rathore AS: Application of multivariate data analysis for identification and successful resolution of a root cause for a bioprocessing application. Biotechnol Prog 2008, 24:720-726. 44. Martin EB, Morris AJ: Enhanced bio-manufacturing through advanced multivariate statistical technologies. J Biotechnol 2002, 99:223-235. 45. Kumar V, Bhalla A, Rathore AS: Design of experiments applications in bioprocessing: concepts and approach. Biotechnol Prog 2014, 30:86-99. 46. Vaidyanathan S, Arnold SA, Matheson L, Mohan P, McNeil B, Harvey LM: Assessment of near-infrared spectral information for rapid monitoring of bioprocess quality. Biotechnol Bioeng 2001, 74:376-388. 47. Trunfio N, Lee H, Starkey J, Agarabi C, Liu J, Yoon S: Characterization of mammalian cell culture raw materials by combining spectroscopy and chemometrics. Biotechnol Prog 2017 http://dx.doi.org/10.1002/btpr.2480. The paper presents an interesting application of spectroscopic tools for monitoring of cell culture raw materials. 48. Lee HW, Liu JJ, Yoon S: Estimation of raw material performance in mammalian cell culture using near infrared spectra combined with chemometrics approaches. Biotechnol Prog 2012, 28:824-832. 49. Hakemeyer C, Strauss U, Werz S, Folque F, Menezes JC: Nearinfrared and two-dimensional fluorescence spectroscopy monitoring of monoclonal antibody fermentation media quality: aged media decreases cell growth. Biotechnol J 2013, 8:835-846.
33. Wishart DS: Characterization of biopharmaceuticals by NMR spectroscopy. Trends Anal Chem 2013, 48:96-111.
50. Jose GE, Folque F, Menezes JC, Werz S, Strauss U, Hakemeyer C: Predicting mab product yields from cultivation media components, using near-infrared and 2D-fluorescence spectroscopies. Biotechnol Prog 2011, 27:1339-1346.
34. Meriage D, Rogers G, Phillips J: Routine screening for the presence of adulteration in raw materials using automated nuclear magnetic resonance spectroscopy. PDA J Pharma Sci Technol 2012, 66:547-559.
51. Shen CF, Hawari J, Kamen A: Micro-quantitation of lipids in serum-free cell culture media: a critical aspect is the minimization of interference from medium components and chemical reagents. J Chromatogr B 2004, 810:119-127.
35. Wen ZQ, Chen G, Luo Y, Li G, Masatani P, Cao X, Bondarenko P, Phillips J: Adulteration detection-assessment of spectroscopic techniques for adulteration detection of raw materials used in biopharmaceutical manufacturing. Am Pharm Rev 2012, 15:26.
52. Qiu JS, Chan PK, Bondarenko PV: Monitoring utilizations of amino acids and vitamins in culture media and Chinese hamster ovary cells by liquid chromatography tandem mass spectrometry. J Pharm Biomed Anal 2016, 117:163-172.
36. Wen ZQ, Li G, Ren D: Detection of trace melamine in raw materials used for protein pharmaceutical manufacturing
www.sciencedirect.com
53. Lee HW, Christie A, Xu J, Yoon S: Data fusion-based assessment of raw materials in mammalian cell culture. Biotechnol Bioeng 2012, 109:2819-2828.
Current Opinion in Biotechnology 2018, 53:99–105
ScienceDirect Role of raw materials in biopharmaceutical manufacturing: risk analysis and fingerprinting Anurag S Rathore, Deepak Kumar and Nikhil Kateja Accurate fingerprinting of critical raw materials that have significant impact on process performance and product quality is a necessary precursor for implementation of QbD in process and product development. This article presents a review of major developments in this space in the last 10 years, with a special emphasis on those in last 5 years. A step by step approach for managing raw materials in the QbD paradigm has been proposed. We think that it is necessary for the biotech industry to better manage variability originating from raw materials if holistic implementation of QbD is to be achieved.
Address Department of Chemical Engineering, Indian Institute of Technology, Delhi, India
Raw materials used in biopharmaceutical manufacturing are a diverse source of materials and include media components for cell culture and fermentation, fine chemicals for purification and chemical modification processes, and excipients used in a formulation of a final drug product [3]. They also include product contact materials like the plastic used in disposable bags. Variability in raw material can come from a change in a chemical or physical characteristic of the material [4]. This variability can affect characteristics and quality of drug product and potentially impact the product’s safety, stability, and efficacy [5]. At times, this has resulted in major adverse events and even resulted in drug recalls [6]. For instance, presence of glass particulates in drug product have led to more than 20 product recalls in the last few years, including the recall of Procrit and Epogen injections [7].
Corresponding author: Rathore, Anurag S ([email protected])
Types of raw material variability Current Opinion in Biotechnology 2018, 53:99–105 This review comes from a themed issue on Pharmaceutical biotechnology Edited by Amanda Lewis and Nripen Singh
https://doi.org/10.1016/j.copbio.2017.12.022 0958-1669/ã 2017 Published by Elsevier Ltd.
Variability in raw materials can be subdivided into three broad categories. The first category includes trace impurities that alter the quality of the biotherapeutic, either by directly modifying it or by catalysing its modification such as peroxides, aldehydes, reducing sugars, and catalytically active metal ions. The second category consists of trace impurities that are themselves toxic to humans, such as lead and aluminium. The third group comprises of microorganism contaminants (and their associated endotoxins) that lead to variabilities in the bioburden of raw materials and can cause severe immunological responses in patients [5]. Each of these require their own approach for monitoring and control.
Sources of variability Introduction Biopharmaceutical manufacturing, because of its complex nonlinear nature, is fraught with a myriad of process variations that can impact safety and efficacy of the drug. Since the introduction of concepts such as process characterization and design of experiments (DoE) over two decades ago, the biopharmaceutical industry has created and demonstrated considerable expertise in unravelling how the process affects the product. However, the role of raw materials (RM) has been somewhat overlooked and as a result has become the primary source of variability in process performance and product quality. The growing significance of the role of raw materials in the process control strategy is evident from the ICH Q8 guideline, which suggests that in the Quality by Design (QbD) framework the manufacturer must understand all sources of variability including the raw materials [1,2]. www.sciencedirect.com
Cell culture processes used to make recombinant proteins use complex growth media. Although some cells can be maintained in a basal medium with no supplementation, majority of cells require addition of up to 100 components such as hormones, growth factors, vitamins, peptides, amino acids and hydrolysates to grow [8]. Naturally derived media can contain a large number of compounds [9,10]. They also have micronutrients in trace amounts. Variability may creep in the media because of several reasons. Degradation of raw materials, impurities, and contaminants present in the media; non-uniformity in milling and blending during manufacturing of large batches of media due to loss of micronutrients; and inconsistency of content levels and changes in raw material sourcing due to constraints in availability can result in raw material variability. Though there has been a shift Current Opinion in Biotechnology 2018, 53:99–105
100 Pharmaceutical biotechnology
towards use of chemically defined media in the last decade, natural media continues to be used by the industry and continues to offer a challenge in maintaining a consistent product quality. Even chemically defined media have a complex composition which necessitates its characterization to achieve consistent process performance. Inconsistency in excipient quality can have an adverse impact on drug product quality. Excipient-derived contaminants can be listed in the following categories — trace metals, peroxides, aldehydes, reducing sugars/polyols, and organic acids. The purity levels of these excipients may vary significantly and affect the product quality/ stability to different degrees. Even with pure excipients, contaminants may be generated during storage through various degradation pathways [11]. Polysorbates, for instance, undergo autooxidation which can influence the stability of a biopharmaceutical product [12]. Leachates from the container/closure system have been shown to have significant impact on product quality as evidenced by recent issues associated with glass lamellae in glass vials [13], tungsten with prefilled syringes [14], silicon from vial stoppers or prefilled syringe barrels [15], leachates from filters [16], and shedding of nanoparticles from a filling pump’s solution-contact surfaces [17,18]. Lot-to-lot variability in chromatography resins can also result in unanticipated changes in yield or product quality [19].
QbD based approach for managing raw materials International Conference on Harmonization (ICH) in its Q7 good manufacturing practice [20], Q8 pharmaceutical development [2], Q9 quality risk assessment [21] and Q10 pharmaceutical quality system [22] guidelines have stipulated stringent requirements regarding product quality. Current practice for raw material analysis has been described in ICH Q7 Guideline [20]. The document states that materials used to prepare active pharmaceutical ingredients (both small molecules and biologics) need to have the identity of each batch confirmed on receipt and a Certificate of Analysis (C of A) provided from the supplier. Pharmacopeial and formulary monographs such as the USP/NF, EuP, JP, and BP provide standardized test methods for the most common and widely used materials. Manufacturers take various approaches towards testing compliance of raw materials. Some qualify a raw materials supplier by performing an initial detailed vendor audit followed by an annual qualification consisting of testing as per the pharmacopeial monograph on three lots of raw material. If the qualification lots test successfully, then subsequent material shipments will require only monograph identification testing. However, companies that take a more conservative approach to raw materials release require full monograph testing for each lot of supplied material [16]. In addition, the supplier must be qualified as suitable based on audits of their facility, Current Opinion in Biotechnology 2018, 53:99–105
Figure 1
Critical/ key raw material
Understand “how” it affects the process and the product
Identify critical attributes of the raw material
Create an analytical method for characterizing the critical attribute
Collect data to fingerprint the raw material Implement approach to enable identification of acceptable and unacceptable raw material lots Current Opinion in Biotechnology
Managing raw materials in the QbD paradigm.
their analytical results must be confirmed to be reliable, and a sampling plan is needed for each incoming material [19]. A key challenge is the relatively large number of raw materials that are used in biopharmaceutical manufacturing (typically > 100). To effectively deal with such a large number of raw materials, a multistep Quality by Design (QbD) based approach has been proposed [23,24] (Figure 1). It is recommended that this evaluation be performed at every major milestone of product development (First in Animal, Phase I, Phase II, Phase III, Validation). Although an arduous task when performed the first time, the effort significantly reduces in the following times as most of the raw materials are the same amongst products of similar kind (e.g. monoclonal antibodies): A risk assessment is performed encompassing all raw materials used in the process. There are a number of different risk assessment tools available in a range of detail and complexity, and it is important to use a methodology suited to the purpose of the assessment [24]. Appropriate stakeholders including process development, manufacturing, quality assurance, and quality control are included in this assessment. The team discusses if the raw material is likely to impact the process performance and if it is likely to impact product quality. Those who are likely to impact both are termed as critical raw materials, those who just impact the process and not the product are termed as key raw materials, and those who are not expected to impact either are called as non-key raw materials [23]. Critical raw materials are therefore thoroughly characterized and their mechanisms of process interactions www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 101
well understood. The manufacturer should understand ‘how’ the raw material is impacting the process and the product. Raw material attributes that can be used to characterize this impact need to be identified and acceptance criteria for these attributes need to be established. Analytical tests need to be developed to monitor these attributes to ensure that each critical raw material lot meets the respective acceptance criteria before its use in the process [23]. Key raw materials can be characterized same as critical raw materials, but while critical raw materials undergo this analysis for every product, key raw materials need to be characterized once for a platform. Every time a new key raw material is added to the platform, it undergoes a thorough characterization before its use [23]. Non-key raw materials include the remaining (and majority of) raw materials and are handled via the internal Quality System of the biopharmaceutical manufacturer as in traditional pharmaceutical manufacturing [23].
Tools for assessment of variability Raw material fingerprinting based on detailed characterization of how the raw material impacts the process and product quality offers a potential solution to the abovementioned conundrum. It can facilitate efficient and effective monitoring of raw material quality and thus prevent significant adverse impact on product quality.
Analytical tools Assessment of raw materials thus involves use of a wide spectrum of analytical techniques like spectroscopic, chromatographic, and enzyme-based tools. Spectroscopic techniques are rapid and have been demonstrated to be quite effective in monitoring variation in raw material quality and hence been the tools of choice. Raman [25,26], near-infrared (NIR) [27–29], Fourier-transform infrared (FTIR), mid-infrared (MIR) [30,31], fluorescence spectroscopy [32] and nuclear magnetic resonance (NMR) spectroscopy [33,34] have been used for identification and characterization of raw materials. Researchers have recently assessed the capability and suitability of several spectroscopic techniques for adulteration detection in tromethamine (Tris) and tromethamine hydrochloride (Tris-HCl), two of the commonly used excipients in biopharmaceutical formulations [35]. Near-infrared (NIR), Raman (handheld Raman and Raman-microscope), Fourier Transform Infrared (FTIR), and Nuclear Magnetic Resonance (NMR) have been evaluated and their merits and demerits have been highlighted in terms of speed of analysis and ability to differentiate adulterated from unadulterated samples [35]. In other studies, mass spectroscopy (MS) has been shown to offer significantly high sensitivity, but is expensive and also cumbersome to perform for routine raw material release testing. Surface enhanced Raman www.sciencedirect.com
spectroscopy (SERS) with gold nanoparticles has been successfully used for detection of trace melamine in raw materials used for protein pharmaceutical manufacturing [36]. A recent review provides an extensive list of tools for examining the sources of variability in cell culture media used for biopharmaceutical manufacturing along with their merits and demerits [8]. This information has been summarized in Table 1. Other alternatives include blotting, capillary electrophoresis (CE) [37], enzymatic methods (e.g. enzyme-linked immunosorbent assay (ELISA)), high-performance liquid chromatography (HPLC) [38], gas chromatography, and SDS-PAGE [39]. Proteomic analysis has also been proposed as a tool for assessing complex raw materials such as serum. Proteomic techniques were used to understand the lot-to-lot variability with respect to impact on growth properties of the cell culture process [40]. As an extension of this approach, proteomics combined with metabolomics has been used to identify the root cause for the lower productivity during scale up of a cell culture process. The analysis attributed hypoxia, resulting due to elevated Cu levels (as a contaminant in media component) as the contributing factor for the observed decline in the performance of the cell culture [41].
Role of statistics in data analysis Datasets originating from the manufacturing of biopharmaceuticals, including spectroscopic tools, are often complex and hence univariate or bivariate analysis is insufficient and likely to result in erroneous conclusions [42]. Multivariate data analysis (MVDA) is required in most cases to delineate relevant information from large multifactorial and multi-collinear data sets. MVDA involves use of tools for exploratory data analysis (data mining), classification (cluster analysis), regression analysis, and predictive modelling [43]. Projection methods are commonly used to overcome multidimensionality of the data set, multicollinearity, missing data, and other variations introduced by deviating factors such as experimental error and noise [44]. One such technique is Principal Component Analysis (PCA), which finds the major variability directions within the data set to explain the overall system behaviour with a few components by decomposing the covariance matrix of the data set [19]. When there is a need to explore multivariate correlations against one or more response variables, a variety of regression techniques are used like principal components regression, partial least squares (PLS), and ridge regression. PLS rearranges process variables space and response variables space by reducing their size while maximizing the covariance between them [19]. Characterization data of raw material library may include one or more categorical response variables and this necessitates the use of a variant of PLS called PLS Discriminant Analysis (PLSDA) [3]. Continuous variable data contains both predictive and uncorrelated information, Orthogonal-PLS (OPLS) is used which leads to improved diagnostics Current Opinion in Biotechnology 2018, 53:99–105
102 Pharmaceutical biotechnology
Table 1 Comparison of different analytical tools for testing raw material variability. Techniques Spectroscopic analytical tools
Methods MIR
NIR
Raman
Fluorescence
NMR Chromatographic techniques
LC, GC, CE, IC
Pros
Cons
Broad range of wavelength, non-invasive, ability to quantify analytes present in very low concentrations, real time monitoring Higher penetration, less absorbance compared to MIR, probes are more durable and cheaper compared to those of MIR, minimal requirement for sample preparation, hand-held, real time monitoring
Less penetration, probes are more fragile and expensive
Can be used for powder and liquid samples, selective bands higher than in case of NIR, sub-sampling possible, can be used for low concentrations by virtue of SERS, hand held, real time monitoring Effective at low concentrations, useful for samples that have low spectral resolution and signal strength by means of fluorescence anisotropy, real time monitoring High specificity and reproducibility, fast and higher throughput, wide dynamic range Liberty to use suitable detection method (MS, spectroscopic) along with separation technique so that specific media components can be addressed, detection of trace samples is possible using more sensitive capillary methods
More convenient only for powdered samples, not recommended for low concentration samples, not suitable for sensitive samples owing to highenergy laser Wavelength selection is crucial due to possibility of interference, high variability causing less reproducibility Possibility of quenching and overlap, can be used only for samples with fluorescent properties High operational cost, requires space Involves invasion and destruction, high cost of equipment, comprehensive mass spectral databases for metabolites is lacking
Source: Adapted from [8].
and interpretability of models [3]. Batch modelling based on PLS and PCA are commonly used in biotech industries to deal with batch operation data [3]. Design of experiments (DoE) is the approach of choice for planning experimental studies such that the relevant process information can be obtained using a minimum number of experiments. It provides information about the interaction of factors and the way total system behaves. It shows how interconnected factors respond over a wide range of values, without requiring the testing of all possible values directly. Thus, it decreases quantity of effort and cost of many runs without compromising the quality of results [45]. MVDA provides an efficient approach for analysis of spectral data. Applications include pre-processing methods to reduce and correct interferences in spectroscopic data such as overlapped bands, baseline drifts, scattering, and path length variation [46]. Statistical approaches are also used for calibration and diagnostics and variable selection, statistic result calculation to build representative and reliable models, and model validation and integration for achieving rigorous prediction and real-time product quality and process monitoring [3,43].
Recent trends in raw material fingerprinting In recent studies, a combination of spectroscopy and chemometrics is used to characterize raw material variability and to predict final product quality based on information about the raw material. Since cell culture Current Opinion in Biotechnology 2018, 53:99–105
raw materials are complex mixtures, it is difficult to identify and quantify every compound present. Spectroscopic tools, such as Raman, MIR, NIR and fluorescence have shown to be adequate when combined with chemometrics for characterizing complex raw materials [8,47]. NIR spectroscopy coupled with PCA has been demonstrated to be effective in fingerprinting a cell culture media component and thereby distinguish between good and poor performing media lots [27]. In another recent study by Lee et al., NIR spectra of multiple soy hydrolysate lots were analyzed and by using a chemometrics approach, it was shown that NIR spectra could be used to reveal lot-to-lot variability as well as vendor-to-vendor differences. Partial least square regression (PLSR) based prediction models for estimating cell growth and productivity of mammalian cell cultures were constructed from the near-infrared spectra [48]. Furthermore, it is important to assess the performance of multiple spectroscopic platforms when determining the impact of a new raw material on cell culture performance. Researchers have demonstrated that both NIR and 2D fluorescence spectroscopy were able to detect compositional changes in basal media and that aging of basal media was found to affect cell culture performance, concluding that both NIR and 2D fluorescence spectroscopy can assess quality of basal media after storage [49]. Another similar study demonstrated capability of NIR and 2D fluorescence spectroscopy for detection of lot-to-lot variability in cultivation media component [50]. www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 103
Figure 2
(b)
Raw Materials:
(a) Group 1
P GMP
Pilot Runs Clinical Runs
Basal Powders
(c) Group 2
(d)
Group B
Group A
Current Opinion in Biotechnology
Case study in raw material management in the QbD paradigm. (a) Multivariate data analysis indicated differences in clinical manufacturing (Group 2) and commercial manufacturing (Group 1). (b) Root cause analysis indicating differences in basal media as the primary source of process variability. (c) Development of a NIR based method for fingerprinting of raw material. (d) NIR method is able to distinguish between ‘good’ (Group B) and ‘bad’ (Group A) lots. Source: Adapted from [16,30].
Researchers argue that in a lot of cases a single spectroscopy may be insufficient for characterization of a raw material and multiple, orthogonal spectroscopy techniques may be required in conjunction to achieve the required specificity and accuracy [47]. A variety of combinations of spectroscopic and non-spectroscopic techniques have been suggested for characterization of complex raw materials. Researchers have demonstrated use of a gas chromatography–mass spectrophotometry (GC–MS) method for the micro quantitation of lipids, determined as www.sciencedirect.com
fatty acid methyl esters (FAME), in complex serum-free cell culture media [51]. Other researchers have suggested the use of a combination of affinity chromatography (Protein G or Protein A), SDS-PAGE, and peptide mass fingerprinting for analyzing cell culture media for the presence of contaminating bovine antibodies [39]. Liquid chromatography–mass spectroscopy (LC–MS) techniques have also been applied for the simultaneous quantification of multiple media components [52]. Recently, combinations of different spectroscopic tools supported Current Opinion in Biotechnology 2018, 53:99–105
104 Pharmaceutical biotechnology
by chemometrics based data fusion approach have also been explored. In one such study multiple spectral platforms (NIR, Raman, fluorescence and X-ray fluorescence) were compared with each other in terms of their estimation capability and then integrated into a unifying prediction model. The results showed that data fusion models exhibit better prediction accuracy than individual spectroscopic methods, demonstrating the synergetic effects of data fusion in characterizing the raw material quality [53].
Case study in raw material fingerprinting Researchers from a major biotechnology company reported an instance of significant issues faced during technology transfer and scale-up of a mammalian cell culture step from clinical scale (2000 L) to commercial scale (15 000 L). This case study is illustrated in Figure 2. The CQA of the product manufactured at commercial scale was significantly different from that manufactured at clinical scale. An extensive investigation was performed to identify the root cause [30]. The root cause was identified as the difference in the basal media supplied by the same vendor for the two scales of operation. Further investigation at the vendor site yielded the information that grouping of medium components during the milling and blending process varied with the scale of production and media type, resulting in different characteristics. Considering the significant impact of the raw material, a NIR based method was created for screening of lots of basal medium powders based on their impact on process performance and product attributes [16]. A combined NIR/MVDA approach made it possible to finger print the raw material to distinguish between good and poor performing media lots.
Summary Regulatory feedback has resulted in a major shift in how the biopharmaceutical industry looks at raw materials and their management. Most major manufacturers are applying the principles of risk analysis and good science towards identification of critical raw materials and their analytical characterization. The concept of accurate fingerprinting of critical raw materials has also gained steam over the last decade. The step by step approach for managing raw materials in the QbD paradigm as proposed in this article could result in a significant improvement in our understanding of this topic. Although a lot has been accomplished in the last decade, we are still far from reaching the same rigor with respect to raw material characterization as we are with process characterization. Constant addition of new raw materials to the system will ensure that the task remains challenging.
Conflict of interest statement The authors declare no financial or commercial conflict of interest. Current Opinion in Biotechnology 2018, 53:99–105
Acknowledgements This work was funded by the Center of Excellence for Biopharmaceutical Technology grant from Department of Biotechnology, Government of India (number BT/COE/34/SP15097/2015).
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest 1. Rathore AS, Winkle H: Quality by design for biopharmaceuticals. Nat Biotechnol 2009, 27:26-34. Seminal paper on application of QbD for biopharmaceuticals. The paper presents the underlying concepts of QbD as well as presents a step by step approach for its application. 2.
Guideline ICHHT: Pharmaceutical development Q8 (R2). Curr Step 2009, 4.
3.
Undey C, Low DC, Menezes J, Koch JM (Eds): PAT Applied in Biopharmaceutical Process Development and Manufacturing. CRC Press; 2012.
4.
Chalk S: Raw material variability. Biopharm Int 2014, 27:38-39.
5.
Challenger CA: Biopharma takes on raw material variability. Biopharm Int 2016, 29:20-24.
6.
Ebbers HC, de Tienda NF, Hoefnagel MC, Nibbeling R, MantelTeeuwisse AK: Characteristics of product recalls of biopharmaceuticals and small-molecule drugs in the USA. Drug Discov Today 2016, 21:536-539.
7.
Reynolds G, Paskiet D: Glass delamination and breakage. BioProcess Int 2011, 9:52-57.
8.
McGillicuddy N, Floris P, Albrecht S, Bones J: Examining the sources of variability in cell culture media used for biopharmaceutical production. Biotechnol Lett 2017, 39:1-17. Contribution of lot of lot variability in quality of serum and other animal or plant derived components to variability in cell culture processes is well known. This preview presents a summary of the various sources of variability. 9.
Yao T, Asayama Y: Animal-cell culture media: history, characteristics, and current issues. Reprod Med Biol 2017 http://dx.doi.org/10.1002/rmb2.12024.
10. Arora M: Cell culture media: a review. Mater Methods 2013, 3:175. 11. Wang W, Ignatius AA, Thakkar SV: Impact of residual impurities and contaminants on protein stability. J Pharm Sci 2014, 103:1315-1330. 12. Kerwin BA: Polysorbates 20 and 80 used in the formulation of protein biotherapeutics: structure and degradation pathways. J Pharm Sci 2008, 97:2924-2935. 13. Wen Z-Q, Torraca G, Masatani P, Sloey C, Phillips J: Nondestructive detection of glass vial inner surface morphology with differential interference contrast microscopy. J Pharm Sci 2012, 101:1378-1384. 14. Liu W, Swift R, Torraca G, Nashed-Samuel Y, Wen Z-Q, Jiang Y, Vance A, Mire-Sluis A, Freund E, Davis J et al.: Root cause analysis of tungsten-induced protein aggregation in pre-filled syringes. PDA J Pharm Sci Technol 2010, 64:11-19. 15. Thirumangalathu R, Krishnan S, Ricci MS, Brems DN, Randolph TW, Carpenter JF: Silicone oil- and agitation-induced aggregation of a monoclonal antibody in aqueous solution. J Pharm Sci 2009, 98:3167-3181. 16. Jameel F, Hershenson S, Khan MA, Martin-Moe S (Eds): Quality by Design for Biopharmaceutical Drug Product Development. Springer; 2015. 17. Tyagi AK, Randolph TW, Dong A, Maloney KM, Hitscherich C, Carpenter JF: IgG particle formation during filling pump operation: a case study of heterogeneous nucleation on stainless steel nanoparticles. J Pharm Sci 2009, 98:94-104. www.sciencedirect.com
Fingerprinting of raw materials for biopharmaceutical manufacturing Rathore, Kumar and Kateja 105
18. Carpenter JF, Randolph TW, Jiskoot WIM, Crommelin DJA, Middaugh CR, Winter G, Fan Y-X, Kirshner S, Verthelyi D, Kozlowski S et al.: Overlooking subvisible particles in therapeutic protein products: gaps that may compromise product quality. J Pharm Sci 2009, 98:1201-1205. 19. Rathore AS, Mhatre R: Quality by Design for Biopharmaceuticals: Principles and Case Studies. Wiley; 2009. 20. Guideline ICH: Q7 good manufacturing practice guide for active pharmaceutical ingredients. Curr Step 2000, 4. 21. Guideline ICH: Q9 quality risk management. Curr Step 2005, 4. 22. Guideline ICH: Q10 pharmaceutical quality system. International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use. 2008. 23. Rathore AS, Low D: Managing raw materials in the QbD paradigm, part 1: understanding risks. Biopharm Int 2010, 23:34-40. The authors present a QbD based approach for management of raw materials for the biopharmaceutical industry. 24. Rathore AS, Low D: Managing raw materials in the QbD paradigm, part 2: risk assessment and communication. Biopharm Int 2010, 23:48-57. The authors present a approach towards risk assessment for biopharmaceutical raw materials, along with examples of risk assessment tools that can be used. 25. Herrero AM, Jime´nez-Colmenero F, Carmona P: Elucidation of structural changes in soy protein isolate upon heating by Raman spectroscopy. Int J Food Sci Technol 2009, 44:711-717. 26. Li B, Ryan PW, Ray BH, Leister KJ, Sirimuthu N, Ryder AG: Rapid characterization and quality control of complex cell culture media solutions using Raman spectroscopy and chemometrics. Biotechnol Bioeng 2010, 107:290-301. 27. Kirdar AO, Chen G, Weidner J, Rathore AS: Application of nearinfrared (NIR) spectroscopy for screening of raw materials used in the cell culture medium for the production of a recombinant therapeutic protein. Biotechnol Progr 2010, 26:527-531. 28. Rhiel M, Ducommun P, Bolzonella I, Marison I, Von Stockar U: Real-time in situ monitoring of freely suspended and immobilized cell cultures based on mid-infrared spectroscopic measurements. Biotechnol Bioeng 2002, 77:174-185. 29. Rhiel M, Cohen MB, Murhammer DW, Arnold MA: Nondestructive near-infrared spectroscopic measurement of multiple analytes in undiluted samples of serum-based cell culture media. Biotechnol Bioeng 2002, 77:73-82. 30. Roychoudhury P, Harvey LM, McNeil B: The potential of mid infrared spectroscopy (MIRS) for real time bioprocess monitoring. Anal Chim Acta 2006, 571:159-166. 31. Hashimoto A, Yamanaka A, Kanou M, Nakanishi K, Kameoka T: Simple and rapid determination of metabolite content in plant cell culture medium using an FT-IR/ATR method. Bioprocess Biosyst Eng 2005, 27:115-123. 32. Ryan PW, Li B, Shanahan M, Leister KJ, Ryder AG: Prediction of cell culture media performance using fluorescence spectroscopy. Anal Chem 2010, 82:1311-1317.
using surface-enhanced Raman spectroscopy (SERS) with gold nanoparticles. Appl Spectrosc 2011, 65:514-521. 37. Simonelli JK, Lee AH, Gallea PE, Cheng KC: Mapping and identification of biopharmaceutical grade cell culture media by MEKC. Chromatographia 2007, 66:977-981. 38. Rathore AS, Bhambure R, Ghare V: Process analytical technology (PAT) for biopharmaceutical products. Anal Bioanal Chem 2010, 398:137-154. 39. Rasmussen LK, Larsen YB, Hojrup P: Characterization of different cell culture media for expression of recombinant antibodies in mammalian cells: presence of contaminating bovine antibodies. Protein Expr Purif 2005, 41:373-377. 40. Zheng X, Baker H, Hancock WS, Fawaz F, McCaman M, Pungor E: Proteomic analysis for the assessment of different lots of fetal bovine serum as a raw material for cell culture. Part IV. Application of proteomics to the manufacture of biological drugs. Biotechnol Prog 2006, 22:1294-1300. 41. Gao Y, Ray S, Dai S, Ivanov AR, Abu-Absi NR, Lewis AM, Huang Z, Xing Z, Borys MC, Li ZJ, Karger BL: Combined metabolomics and proteomics reveals hypoxia as a cause of lower productivity on scale-up to a 5000-liter CHO bioprocess. Biotechnol J 2016, 11:1190-1200. 42. Kourti T: The process analytical technology initiative and multivariate process analysis, monitoring and control. Anal Bioanal Chem 2006, 384:1043-1048. 43. Kirdar AO, Green KD, Rathore AS: Application of multivariate data analysis for identification and successful resolution of a root cause for a bioprocessing application. Biotechnol Prog 2008, 24:720-726. 44. Martin EB, Morris AJ: Enhanced bio-manufacturing through advanced multivariate statistical technologies. J Biotechnol 2002, 99:223-235. 45. Kumar V, Bhalla A, Rathore AS: Design of experiments applications in bioprocessing: concepts and approach. Biotechnol Prog 2014, 30:86-99. 46. Vaidyanathan S, Arnold SA, Matheson L, Mohan P, McNeil B, Harvey LM: Assessment of near-infrared spectral information for rapid monitoring of bioprocess quality. Biotechnol Bioeng 2001, 74:376-388. 47. Trunfio N, Lee H, Starkey J, Agarabi C, Liu J, Yoon S: Characterization of mammalian cell culture raw materials by combining spectroscopy and chemometrics. Biotechnol Prog 2017 http://dx.doi.org/10.1002/btpr.2480. The paper presents an interesting application of spectroscopic tools for monitoring of cell culture raw materials. 48. Lee HW, Liu JJ, Yoon S: Estimation of raw material performance in mammalian cell culture using near infrared spectra combined with chemometrics approaches. Biotechnol Prog 2012, 28:824-832. 49. Hakemeyer C, Strauss U, Werz S, Folque F, Menezes JC: Nearinfrared and two-dimensional fluorescence spectroscopy monitoring of monoclonal antibody fermentation media quality: aged media decreases cell growth. Biotechnol J 2013, 8:835-846.
33. Wishart DS: Characterization of biopharmaceuticals by NMR spectroscopy. Trends Anal Chem 2013, 48:96-111.
50. Jose GE, Folque F, Menezes JC, Werz S, Strauss U, Hakemeyer C: Predicting mab product yields from cultivation media components, using near-infrared and 2D-fluorescence spectroscopies. Biotechnol Prog 2011, 27:1339-1346.
34. Meriage D, Rogers G, Phillips J: Routine screening for the presence of adulteration in raw materials using automated nuclear magnetic resonance spectroscopy. PDA J Pharma Sci Technol 2012, 66:547-559.
51. Shen CF, Hawari J, Kamen A: Micro-quantitation of lipids in serum-free cell culture media: a critical aspect is the minimization of interference from medium components and chemical reagents. J Chromatogr B 2004, 810:119-127.
35. Wen ZQ, Chen G, Luo Y, Li G, Masatani P, Cao X, Bondarenko P, Phillips J: Adulteration detection-assessment of spectroscopic techniques for adulteration detection of raw materials used in biopharmaceutical manufacturing. Am Pharm Rev 2012, 15:26.
52. Qiu JS, Chan PK, Bondarenko PV: Monitoring utilizations of amino acids and vitamins in culture media and Chinese hamster ovary cells by liquid chromatography tandem mass spectrometry. J Pharm Biomed Anal 2016, 117:163-172.
36. Wen ZQ, Li G, Ren D: Detection of trace melamine in raw materials used for protein pharmaceutical manufacturing
www.sciencedirect.com
53. Lee HW, Christie A, Xu J, Yoon S: Data fusion-based assessment of raw materials in mammalian cell culture. Biotechnol Bioeng 2012, 109:2819-2828.
Current Opinion in Biotechnology 2018, 53:99–105