Use of HPLC as an Enabler of Process Analytical Technology in

Subscriber access provided by Kaohsiung Medical University

Technical Note

Use of HPLC as an enabler of process analytical technology in process chromatography Anamika Tiwari, Nikhil Kateja, Shubham Chanana, and Anurag S. Rathore Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00897 • Publication Date (Web): 31 May 2018 Downloaded from http://pubs.acs.org on May 31, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Use of HPLC as an enabler of process analytical technology in process chromatography Anamika Tiwari1, Nikhil Kateja1, Shubham Chanana1, Anurag S. Rathore1* Department of Chemical Engineering, Indian Institute of Technology, Hauz Khas, New Delhi, India

*

Corresponding author:

Anurag S Rathore Department of Chemical Engineering Indian Institute of Technology Delhi Hauz Khas, New Delhi, 110016, India Phone: +91-9650770650 Email: [email protected] Website: www.biotechcmz.com

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract On-line monitoring of product quality attributes using high resolution analytical tools is a prerequisite for implementation of process analytical technology (PAT) and thereby ensuring product quality and consistency. On-line high performance liquid chromatography (HPLC) has been established for real time monitoring of product quality attributes. However, requirement of liquid handling system capable of on-line sampling and fractionation and interfacing it with preparative scale chromatography appends to the cost and complexity of the design module of commercially available on-line HPLC. This paper proposes a cost-effective approach for using a traditional, off-line HPLC for on-line analysis using a 2way/6port valve to facilitate simultaneous automated sampling of product stream eluting from a process column and fractionation. No sample dilution is required in the proposed approach. The versatility of the proposed on-line configuration has been verified by demonstrating its use for two of the most common separations required during production of monoclonal antibody therapeutics separation of aggregates and separation of charge variants. Process modeling has been performed to allow interpolation of HPLC data and facilitate pooling to achieve the desired purity (model predicted purity 99.1% vs. achieved purity of 99.3 % for removal of aggregates, model predicted main species yield of 62.4 % vs. achieved main species of 62.9 % for pooling of charge variants). The study thus demonstrates that the proposed on-line HPLC configuration can be used for PAT applications in preparative chromatography.


Page 2 of 15

Page 3 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Monoclonal antibody products have gradually become the largest class of therapeutic products. Due to their inherent complexity and heterogeneity, slight process deviations can affect their product quality profile. So, real time monitoring of its product quality attributes using high resolution analytical tools becomes a prerequisite for detecting any process deviations, thereby ensuring product quality and consistency.1 Spectroscopy based tools such as Ultraviolet–visible (UV), Near-Infrared (NIR), Fourier Transform Infrared (FTIR), and Raman spectroscopy have proven to be powerful techniques for real time monitoring and process control.2-3 However, due to limited specificity that spectroscopic methods offer, considerable analytical development is required for creating methods that are capable of performing the required analysis and even thereafter, the robustness and utility of the final solution is suboptimal in most cases. On-line analytical chromatography overcomes these limitations due to its unparalleled robustness, high resolution and specificity that it offers for different species. On-line high performance liquid chromatography (HPLC) can act as a tool to measure product purity in real time, resulting in significantly higher operational efficiency and reduced variability in product quality, thereby enabling automation and control. Its use has been widely demonstrated for industrial applications including pharmaceuticals4, chemical5, and electronics.6 Case studies examining the use of on-line HPLC for different modalities of chromatography has also been reported.7-9 Additionally, for the case of continuous manufacturing, on-line monitoring of critical quality attribute (CQAs) is a prerequisite for ensuring product quality and consistency.10 Implementation of on-line HPLC typically includes three steps: a mechanism for automated sampling, sample pretreatment such as dilution, and finally the actual chromatographic and data analysis. These steps along with the requirement of liquid handling system capable of on-line sampling and fractionation and interfacing it with preparative scale chromatography appends to the cost and complexity of the design module of commercially available on-line HPLC. Moreover, very few studies have been conducted regarding the integration of on-line HPLC with process chromatography for biopharmaceuticals applications. On premise of these limitations, in this study, we propose a simplistic and economical solution for converting an existing off-line HPLC to an on-line HPLC without the requirement of any sample pretreatment (dilution). The proposed approach utilizes a 2way/6 port valve. Moreover, in the proposed configuration the liquid handling system is bypassed and point sampling of elute is performed directly from the preparative chromatography column. This is different from the case of conventional on-line HPLC where the aliquots of samples are taken from the fractions (inside liquid handling system) 3 ACS Paragon Plus Environment


for getting loaded onto the HPLC column. In the latter, process deviations can occur and can avert the purpose of real time monitoring. The versatility of the proposed on-line configuration has been verified by demonstrating its use for two of the most common separations required during production of monoclonal antibody products, separation of aggregates and separation of charge variants. Point sampling of the elute directly from the preparative column into the HPLC system along with the use of shorter analytical method runtime resulted in a remarkably faster analysis, allowing a sufficient time frame available for taking process decision. The proposed system can facilitate real time decision making for pooling during process chromatography. Materials and Methods Materials The mAb sample (IgG 1, 150 KD, pI 7.9-8.1) used in this study was donated by a leading Indian biopharmaceutical manufacturer. The aggregate content in mAb feed sample was 2.3% while its charge variant composition included acidic (12.4%), main (35.2%) and basic species (52.4%). Sodium acetate, acetic acid, sodium di-hydrogen phosphate, di-sodium hydrogen phosphate and sodium chloride were procured from Merck, India. For preparative chromatography, chemicals were of analytical grade and for analytical chromatography of HPLC grade. Buffer solutions were prepared using Mili Q water. Buffers for analytical high-performance liquid chromatography (HPLC) were filtered through a 0.22 µm filter and degassed prior to use. Case Study I: Analytical SEC for analysis of aggregates Tosoh TSKgel SuperSW mAb HTP 4 µm, 4.6 mm ID × 150 mm (Tosoh Bioscience LLC, Part no. 0022855, Tosoh, USA) column was used for size exclusion high performance liquid chromatography (SE-HPLC). Aggregate analysis was performed at a flow rate of 0.5 mL/min at 25°C using 0.2 M phosphate buffer with pH 6.7 as mobile phase with a runtime of 6 minutes. Case Study I: Preparative scale CEX chromatography A 4-mL preparative scale column, 5×200 mm, was packed in Tricon (GE Healthcare Life Sciences, Uppsala, Sweden) column using SP Sepharose Fast Flow (GE Life Sciences) resin. The chromatography process was performed on an Akta Avant 25 (GE Healthcare Life Sciences, Uppsala, Sweden). The process consisted of 5 column volumes (CV) of equilibration with 20 mM, pH 5.5 acetate buffer. It was followed by loading a mAb sample at 20 mg/mL resin capacity followed by a wash of 5 CV with equilibration buffer. Elution was performed using a 25 CV linear gradient with 200mM NaCl in elution buffer and a flow rate of 0.5 mL/min. The pumps and data acquisition were controlled using UnicornTM 6.0 software (GE Healthcare).


Page 4 of 15

Page 5 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Case Study II: Analytical CE-HPLC for analysis of charge variants Analytical cation exchange high performance liquid chromatography (CE-HPLC) was performed using a MAbPaC SCX-10RS column (4.6 x 150 mm, Thermo Fisher Scientific, Waltham, USA). Buffer A (15 mM sodium phosphate, pH 6.2) and buffer B (150 mM sodium phosphate, pH 6.2) were used as mobile phase buffers. Charge variant analysis was carried out with 4 minutes runtime method using sigmoidal gradient at a flow rate of 1.2 mL/min at 25°C as has been described in detail in an earlier publication.11 Case Study II: Preparative scale CEX chromatography Chromatography was performed using Poros HS resin (Applied Biosystems) and a pH gradient with pH 6, 15 mM sodium phosphate buffer containing 35 mM NaCl as equilibration buffer and pH 9, 15 mM sodium phosphate buffer containing 35 mM NaCl as elution buffer. After 5 CVs of equilibration, loading of the mAb sample was followed by 5 CVs of sample wash and thereafter 25 CVs of sample elution with pH gradient in the elution buffer. On-line HPLC Preparative chromatography system was integrated with the analytical HPLC (1260 Infinity II, Agilent Technologies) using software controlled 2 pos/6port valve (1290 Infinity Valve Drive, Agilent Technologies) with a 100 µL external sample loop. The 2pos/6port valve was used to simultaneously control the loading of the analytical column by appropriate switching and peak fractionation. Figures 1B and 1C depict the valve switching positions to ensure sample loop filling in bypass and column loading in main pass, respectively. This activity of loop filling and column loading is performed based on runtime of the methods which is after every 6 minutes for SE-HPLC analysis and 4 minutes for CE-HPLC analysis. Amount of sample to be loaded onto the analytical column is regulated by controlling the valve switching time. The decision for the amount of volume to be injected on the analytical column is made on the basis of absorbance of the elution profile peak. It was found that injection volume between 5 µL to 100 µL containing the product between 5 µg to 40 µg produced acceptable performance. This gave us a wide range of operating window, thereby obviating the need for any sample dilution. Lag time due to tubing hold up volume between the preparative chromatography and the analytical system was taken into account. In order to minimize peak dispersion, tubing hold up volume was kept minimal.



Page 6 of 15

Figure 1. A. Schematic illustration of the set up for On-line-HPLC. B. Configuration of 2 Position 6 port valve for loop filling. C. Configuration of 2 Position 6 port valve for column loading.

Systematic Interpolation Technique of on-line HPLC data for pooling decision making On-line analysis and pooling decision were made according to the proposed method as shown in Figure 2. Sample eluting from process chromatography was diverted to HPLC at every 6 minutes for SE-HPLC and 4 minutes for CE-HPLC analysis. Area for individual species obtained after each HPLC analysis was normalized with its injection volume and utilized for construction of their respective distribution profiles. Systematic interpolation of normalized area was performed with respect to elution CV using piecewise cubic hermite interpolation polynomial (PCHIP) feature of MATLAB (v.8.0.1, Mathworks®).12 PCHIP is a shape preserving interpolation technique which ensures that there is minimal degree of curvature between two data points, resulting in retaining the original maxima and minima of the peak curve. This interpolation technique has been previously utilized for construction of chromatographic peak profiles.13,14 Pooling was performed by utilizing area under the interpolated curve (Figures 4A and 6A). Calculation of pool purity and main species purity were performed as follows: b

∫ f ( x)dx 1

Pool Purity =

× 100

a b

b

∫ f ( x)dx + ∫ f 1

a

2

(1)

( x)dx

a b

∫ g ( x)dx 1

Main species purity =

a b

b

∫ g ( x)dx + ∫ g 1

a

a

×100

b 2

( x)dx + ∫ g 3 ( x)dx 6

a

ACS Paragon Plus Environment

(2)

Page 7 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Where, a and b are start and end CV of pooling, respectively. For case study I, equation 1 was used where, f1(x) and f2(x) represent the interpolated profile for monomer and aggregate, respectively, and

∫ f ( x)dx 1

and

∫f

2

( x ) dx refer to the amount of monomer and aggregate in the

product, respectively. For case study II, equation 2 was used where g1(x), g2(x), and g3(x) represents the interpolated profile for main, acidic and basic species respectively and

∫g

2

∫ g ( x)dx , 1

( x )dx , and ∫ g 3 ( x )dx refers to amount of main, acidic and basic species respectively. Based

on the area under the interpolated curve and pooling criteria defined for aggregate and charge variant analysis, the pool start and end elution CVs were estimated.

Figure 2. Proposed method for on-line analysis and pooling decision

Results and Discussion The PAT initiative under the US Food and Drug Administration promotes real time monitoring and control of product quality attributes.15,16 On-line monitoring of the production process enables us to achieve continuous process verification. Of the various on-line analytical tools that can be used, on-line HPLC scores high due to the unparalleled specificity and selectivity that HPLC offers compared to other spectroscopic tools. In this case we propose a setup that facilitates automated HPLC sampling and analysis from process chromatography stream. On-line data points obtained from aggregate and charge variant analysis were interpolated to get their 7 ACS Paragon Plus Environment


distribution over the entire elution peak and make the pooling decision according to the desired purity level. Case Study I: Separation of mAb aggregates by cation exchange (CEX) process chromatography Preparative CEX runs were performed with SP Sepharose Fast Flow (SPFF) resin using a salt linear gradient. Figure 3 depicts the process CEX chromatogram together with the on-line HPLC chromatograms corresponding at various points of the elution profile. As expected, the earlier fractions contain pure product without any aggregates whereas the latter part of the elution peak contains elevated levels of aggregate species. A total of 12 samples of the elution peak were analyzed, each at an interval of 6 minutes. Further, 11 fractions of 3 mL (0.75 CV) each were collected in the fractionator. Performance of chromatographic separation for both off-line and on-line modes was compared. Figure 3B shows the comparison between the two methods with respect to % monomer area. It can be observed that there is high degree of correlation between the on-line and traditional offline HPLC analysis data, thus confirming its feasibility for use in PAT application for aggregate analysis. The on-line HPLC data obtained for monomer and aggregates were used to construct its distribution profile with respect to elution CV (using PCHIP) as shown in Figure 4A. Pooling criteria was set according to the maximum allowable % for aggregates in mAbs (typically 99%. Predicted pooling was found to be in agreement with the actual values for product purity (Table 1). To further visualize the impact of elution start CV (column volume) and end CV on monomer distribution in the pool, contour plot (Figure 4B) was constructed using the interpolated data. It is evident from the contour plot that pool purity decreases with increase in elution CVs. This is because aggregates elute during the later portion of the elution peak as they exhibit a stronger binding to the CEX resin because of their greater surface charge compare to monomeric mAb.


Page 8 of 15

Page 9 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. A. Preparative CEX chromatogram showing separation of aggregates from the mAb product. The insets show on-line SE-HPLC chromatograms at various points of the elution profile. It is evident that aggregate species elute in the later part of the product peak. B. Comparison of off-line and on-line HPLC analysis for monomer peak %.

Figure 4. A. Interpolated graph of on-line data showing distribution of aggregates and monomer with elution CVs. B. Contour plots indicating impact of elution start and end CV on % pool purity. The reddish area shows the acceptable range for pooling.

1



Case study II: Separation of charge variants by cation exchange (CEX) process chromatography Separation of charge variants on preparative cation exchange chromatography was achieved by using a pH gradient. Figure 5 shows the process chromatogram together with the analytical CEHPLC chromatograms depicting acidic, main and basic species eluting at various points of the chromatographic profile. As can be seen from the chromatograms, the acidic species elute in earlier fractions with the main species appearing primarily towards the peak maxima of the elution peak and finally towards the later part of the peak, fractions contain basic species. Analysis was performed at a total of 23 points of the elution profile with 22 fractions of 2 mL (0.5 CV) each collected simultaneously in the fractionator. It is evident from Figure 5B that there is good agreement between on-line and off-line fraction data for % main peak area . The number of online data points obtained to regenerate the profile is dictated by run time of the analytical method. As a result, in regions where there is a rapid change in concentration of protein or poor resolution of species, it is likely that there will be disparity between the off-line and on-line data. Of course, the faster the time of analysis, this disparity will be reduced. In our case, the offline and online data were quite comparable for acidic and basic species (data not shown). Interpolation of the on-line CE-HPLC data was performed to construct the distribution profile for individual charge variants across the elution peak following the same approach as before (Figure 6A). Pooling decision was made on the basis of main species content by using equation 2. Since more than two components were present, two distinct cut points were required (both elution start and end CV). For example, pooling from 18.5 CV to 19.5 CV was predicted to obtain more than 60% of the main product species. Once again, the predictions were found to be in agreement with the actual column performance (Table 1). Further, contour plots were created to understand the impact of changes in start and end CV of pooling on target % of main peak. It is observed that % of main species decreases towards the initial and end stages of the elution CVs as the early eluting CEX fractions are rich in acidic variants due to their lower isoelectric point (pI) while the late eluting CEX fractions are rich in basic variants due to their relatively higher pI compared to the main species.


Page 10 of 15

Page 11 of 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. A. Preparative CEX chromatogram showing separation of charge variants of the mAb product. The insets show on-line CE-HPLC chromatograms at various points of the elution profile. B. Comparison of off-line and on-line HPLC analysis for main species peak %.

Figure 6. A. Interpolated graph of on-line data showing distribution of acidic, main and basic species with elution CVs. B. Contour plots indicating impact of elution start and end CV on % main species pool purity. The reddish area shows the acceptable range for pooling.

Benefits of the proposed configuration The proposed on-line HPLC set up offers significant advantages over traditional offline systems. The on-line set up has the ability to withdraw eluent from preparative chromatography in continuous manner compared to analyzing the individual fraction off-line post the preparative 11 ACS Paragon Plus Environment


run. This will result in a substantial reduction in the overall process time as well as in the risk of external contamination. On-line analysis also made it feasible for real time decision making for chromatographic pooling based on % purity. Automated injection without any in-process sample dilution and rapid analytical methods allowed for easier and faster monitoring of CQAs resulting in consistent product quality. Additionally, point analysis of the sample instead of fraction analysis can eventually result in greater process understanding due to real time monitoring. This will be helpful in detecting any process deviation because such deviations will be averaged out during fractional analysis. This will not only result in improved process modeling but also help in implementing process control strategies. Although not included in this study, the proposed online configuration can also be implemented in continuous manufacturing for improved process control which is required for maintaining steady state over longer duration of time.

Conclusions A setup to perform on-line sampling of process chromatography elute followed by HPLC analysis has been suggested as an approach to achieve consistent pooling. Advantages of the proposed setup include automated injection without any need of sample dilution followed by rapid HPLC analysis to allow for real time process monitoring. Interpolation of the HPLC data allowed us to obtain distribution of aggregate species as well as charge variants across the elution peak. This approach allowed us to predict the start and end CV to be pooled to achieve the desired level of pool purity. To test the validity of the predicted values, pseudo pools were examined to obtain the actual % pool purity. It is evident from Table 1 that, in cases of both aggregate and charge variant analysis, the predicted pooling purity was in agreement with the actual experimental values. All the deviations were within

Use of HPLC as an Enabler of Process Analytical Technology in

Recommend Documents