Transition Analysis (TA) is a method traditionally used to evaluate the condition and integrity of a chromatography column. However, this method generally requires conducting complex data post‑processing and calculations which are hard to automate and to run in real time during the manufacturing process. Thus, completing the TA offline leads to time-consuming efforts that may delay the subsequent manufacturing operations. Colleagues from Amgen Manufacturing Limited developed an automated method using SIMCA® and SIMCA®-online to compute the TA curve in real time, allowing them to monitor and verify the condition of their column chromatography steps with limited operator input and no impact on manufacturing rate.
Introduction
Column chromatography steps are among the main unit operations in the purification area of any biopharmaceutical manufacturing setup. To ensure consistent product quality, it is essential to regularly monitor performance for consistency, degradation, efficiency and integrity. There are multiple techniques that have been used to evaluate chromatographic performance. One traditional technique is a pulse injection integrity test, generally used to assess the efficiency of a newly packed column.1 This method requires conducting an additional activity at the column to generate the data needed to calculate the height equivalent of a theoretical plate (HETP) and Asymmetry factor (Af) based on the normal distribution of an elution peak.
This report addresses the key factors shaping pharmaceutical formulation, including regulation, QC and analysis.
Access the full report now to discover the techniques, tools and innovations that are transforming pharmaceutical formulation, and learn how to position your organisation for long-term success.
What you’ll discover:
Key trends shaping the pharmaceutical formulation sector
Innovations leading progress in pharmaceutical formulation and how senior professionals can harness their benefits
Considerations and best practices when utilising QbD during formulation of oral solid dosage forms
Can’t attend live? No worries – register to receive the recording post-event.
Transition Analysis (TA) is another promising technique for evaluating column performance. TA consists of measuring and inspecting the response of a signal (eg, conductivity) at the column outlet to a step change caused by an input solution,2,3 which can be measured during normal column operations, avoiding the additional activity required for the pulse injection integrity test. When performing TA, a first derivative transition analysis curve (TA curve) is calculated on the chromatographic transition,1,4 leading to a bell-shaped curve. HETP and Af can also be computed based on this transition curve.
The TA method described above often poses several challenges. First, it relies on collecting high‑frequency data that needs to be processed with a complex algorithm for filtering and calculating interpolations and derivatives.3 Second, these calculations are generally performed offline due to the challenges of automating them with some of the traditional analytics packages while the chromatographic transition is occurring. Thirdly, the prospects of validating these TA methods were low due to the complexities of validating such a manual and person-dependent approach. Lastly, these complex offline calculations take time and labour, causing commercial manufacturing to delay its subsequent operations. This last challenge is becoming increasingly important as demand for increased manufacturing capacity and efficiency grows.2
There have been recent efforts to automate TA and the calculations of column performance indicators (eg, HETP and Af). J Reckamp,3 for example, presented an automatic approach using the software Seeq®. However, manual TA calculations still prevail in many biotechnology settings. These calculations are often performed using spreadsheet software such as Microsoft (MS) Excel. Although MS Excel provides an extensive list of mathematical functions, some of which are necessary to quickly perform first derivative calculations, this method is prone to errors due to its inherent nature of manually querying and transforming process data. It is also time consuming and delays operation activities.
We developed a method using SIMCA®5 and SIMCA®-online6 to compute in real time the TA curve while the column is either being equilibrated in preparation for use or during any of the post‑elution steps where there is a well-defined output signal transition. The mathematical functions in SIMCA are very limited and lack the functions typically used offline in MS Excel. However, by creatively combining the few mathematical functions available in the software, we were able to replicate the more advanced functions required to generate the TA curve. Additionally, with the use of SIMCA-online, we can now obtain the TA curve in real time, while manufacturing activities on the chromatography column are ongoing, avoiding operational delays due to manual calculations.
Methodology
Translating offline calculations to SIMCA
When performing transition analysis manually, the slope, S, formula7 in (1) is readily available in software such as MS Excel and can be used to obtain the transition curve. To do so, (1) is applied to vectors of transition data to obtain a first derivative of a conductivity signal with respect to totalised flow volume. Thus, when applying (1), V represents totalised flow volume, C represents conductivity, and V_ and C_ represent sample means of totalised flow volume and conductivity, respectively. (1) Although this exact formula is not available in SIMCA, we developed a method that replicates the calculations obtained with (1). Table 1 lists the mathematical functions available in SIMCA that can be combined to mimic the TA curve calculated manually. The key to solving the TA curve, STA, in SIMCA was deciphering how to compute moving averages on any given variable. A moving average, N_Z, can be obtained by applying the Lag function as (2) where N is a vector of any given variable, Z is an odd number representing the number of values used for calculating a sample average, and N_Z is a vector of moving averages of N using Z-values. For example, a three-value moving average on vector N can be expressed as (3)
Calculating the TA curve in SIMCA
The steps defined below can be followed to obtain TA curves in SIMCA that are equivalent to the curves obtained with traditional spreadsheet software. First, the original transition data needs to be post-processed and filtered for noise. The original data on flow rate, Fr, measured at a frequency of s seconds, is converted to totalised flow volume, VT, and normalised by Column Volume, VC, as (4) where frequency s is divided by 60 to convert to minutes. We suggest using data recorded at a frequency of two seconds because it provides sufficient resolution without overwhelming the calculation engine.
Let C be the original conductivity data measured at the same frequency of flow rate. Performing the post-processing and noise filtering on C offline would be achieved by computing a moving average. We achieved this post-processing in SIMCA by employing (2). While there is no standard or recommendation on what value of Z should be used, we used here Z = 11 to match the results from the manual approach presented below in the Results section. Thus, the smoothed conductivity C_11, is obtained as (5) A normalised smoothed conductivity C_N was then obtained as Norm(C_11). Now that the original transition data has been post-processed, we proceeded to solve for each term in the slope equation, S (1). Like the signal smoothing calculations, there is no standard on the sample size when calculating sample means. Hence, using Z = 5 to match the sample size in the manual approach, the sample means V_ and C_ were obtained by applying (2) as (6) (7) The summations in the numerator, dC, in (1) were obtained using once again the Lag function. The number of terms in the summation should match the Z value used when applying (2). Thus, in our example, dC is obtained with a five-term summation as (8) Similarly, the denominator, dF, in (1) can be obtained as (9) Finally, the first derivative TA curve, STA, is obtained as (10)
Results
To demonstrate the proposed method, we illustrate transition data from a commercial process recorded and plotted in real time during a transition activity of a chromatography column. Figure 1 shows the original flow rate (a) and conductivity (b) as it was recorded in real time with SIMCA-online while the process was ongoing. The maturity variable for both trends in Figure 1 is the totalised transition flow normalised to column volume.
Figures 1a and 1b: Original data captured in real time with SIMCA®-online for (a) flow rate and (b) conductivity
Figure 2 illustrates the TA curve that appeared in SIMCA-online based on the same original flow rate and conductivity data shown in Figure 1. Since the TA curve calculations require data processing that depends on normalisation and on the Lag function, the curve appears automatically in SIMCA-online only after the transition phase has concluded.
Figure 2: TA curve (dC/dV) calculated in SIMCA®-online based on the data from Figure 1
Figure 3 compares TA curves obtained in SIMCA‑online to those calculated offline in MS Excel. The TA curves in Figure 3a were obtained from a manufacturing unit and batch different than that of Figure 3b. The TA curves obtained online have a nearly identical shape and size to the curves obtained offline.
Figures 3a and 3b: TA curves (dC/dV) calculated in SIMCA®-online compared to the TA curves obtained offline. TA curves in panel (a) are for a column and lot different than those in panel (b).
Discussion
Value proposition of the real-time TA curve
We presented a method that can be easily adopted by any company that uses SIMCA for real-time monitoring of their column chromatography operations. A model based on these calculations can be developed and quickly deployed into a good manufacturing practice (GMP) and validated production environment for agile analysis of chromatography operations. As such, with minimum effort, it is possible to build a collection of TA curve models for different chromatography units. This presents an opportunity to easily standardise how a manufacturing setting may monitor in real time the condition and efficiency of their chromatography operations with limited operator input and no impact to manufacturing rate.
Next steps
The Direct Transition Analysis (DTA) method developed by Cui et al.1 proposes two performance indicators: a) transition width (TransWidth) and b) direct asymmetry (DirectAf), which can be directly extracted from the transition data without complex signal processing. These two indicators are analogous to the commonly used HETP and Af, respectively.
Therefore, our proposed real-time method for obtaining the TA curve can be greatly enhanced by having real-time and automatic results of HETP and Af, TransWidth, and/or DirectAf. The authors are actively engaged with Sartorius to advance our real-time TA curve method by incorporating online calculations of the above-mentioned performance indicators for a more robust chromatographic analysis.
Acknowledgements
The authors would like to thank their Amgen colleagues: Carlos Domínguez, Christopher Garvin, Daniel Chow, Deborah Soto Ortega, Abhishek Ramanathan, Taína Franco and Michelle Burgos Ortiz. They also want to recognise the contributions from Scott Carpenter, Anna Persson and Amin Ganjian from Sartorius.
Larson T, Davis J, Lam H, Cacia J. Use of process data to assess chromatographic performance in production-scale protein purification columns. Biotechnology Program [Internet]. 2003 [cited 2021 Aug 06]; 19(2):485-92. Available from: https://doi.org/10.1021/bp025639g
Martin A J P, Synge R L M. A new form of chromatogram employing two liquid phases: 1) a theory of chromatography 2) application to the micro-determination of the higher monoamino-acids in proteins. Biochem J [Internet]. 1941 Dec [cited 2021 Aug 06]; 35(12):1358-68. Available from: https://doi.org/10.1042/bj0351358
Pablo J Rosado is a Process Development Data Scientist at Amgen Manufacturing Limited in Juncos, Puerto Rico, supporting initiatives related to process monitoring, data analytics and digital transformation. He has a BS in Mechanical Engineering from the University of Puerto Rico at Mayagüez, as well as a MS and PhD in Mechanical Engineering from the University of California at Berkeley. Pablo joined Amgen in 2017 after completing a Postdoctorate fellowship at Lawrence Berkeley National Laboratory and Postdoctoral fellowship training with Insight Data Science.
Melvin Ortiz is a Principal Engineer at Amgen Manufacturing Limited in Juncos, Puerto Rico. He has a BS in Chemical Engineering from the University of Puerto Rico at Mayagüez, as well as a MS in Chemical Engineering from Purdue University. He has over 23 years of experience in commercial manufacturing with 17 years supporting downstream biotechnology manufacturing.
Badua Merheb is a Data Sciences Senior Manager at Amgen Manufacturing Limited in Juncos, Puerto Rico. He has a Bachelor of Science in Chemical Engineering from University of Puerto Rico at Mayagüez and a Masters in Engineering Management from the Polytechnic University of Puerto Rico. He has more than 24 years of experience in the pharma/biotech industries supporting upstream and downstream processes. He currently leads the Data Science and Analytics team in the Drug Substance Unit.
Alejandro Toro is currently the Executive Director of Drug Substance Process Development at Amgen Manufacturing Limited in Juncos, Puerto Rico supporting commercial operations for bacterial fermentation and mammalian cell culture manufacturing plants. He is responsible for the site’s Process Sciences, Attribute Sciences, Process Engineering (MS&T), Process Validation and Data Science & Analytics teams. Alejandro holds a BS in Chemistry from the Pontifical Catholic University of PR and a PhD in Protein Biochemistry from Rutgers University.
This website uses cookies to enable, optimise and analyse site operations, as well as to provide personalised content and allow you to connect to social media. By clicking "I agree" you consent to the use of cookies for non-essential functions and the related processing of personal data. You can adjust your cookie and associated data processing preferences at any time via our "Cookie Settings". Please view our Cookie Policy to learn more about the use of cookies on our website.
This website uses cookies to improve your experience while you navigate through the website. Out of these cookies, the cookies that are categorised as ”Necessary” are stored on your browser as they are as essential for the working of basic functionalities of the website. For our other types of cookies “Advertising & Targeting”, “Analytics” and “Performance”, these help us analyse and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these different types of cookies. But opting out of some of these cookies may have an effect on your browsing experience. You can adjust the available sliders to ‘Enabled’ or ‘Disabled’, then click ‘Save and Accept’. View our Cookie Policy page.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Cookie
Description
cookielawinfo-checkbox-advertising-targeting
The cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Advertising & Targeting".
cookielawinfo-checkbox-analytics
This cookie is set by GDPR Cookie Consent WordPress Plugin. The cookie is used to remember the user consent for the cookies under the category "Analytics".
cookielawinfo-checkbox-necessary
This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Necessary".
cookielawinfo-checkbox-performance
This cookie is set by GDPR Cookie Consent WordPress Plugin. The cookie is used to remember the user consent for the cookies under the category "Performance".
PHPSESSID
This cookie is native to PHP applications. The cookie is used to store and identify a users' unique session ID for the purpose of managing user session on the website. The cookie is a session cookies and is deleted when all the browser windows are closed.
viewed_cookie_policy
The cookie is set by the GDPR Cookie Consent plugin and is used to store whether or not user has consented to the use of cookies. It does not store any personal data.
zmember_logged
This session cookie is served by our membership/subscription system and controls whether you are able to see content which is only available to logged in users.
Performance cookies are includes cookies that deliver enhanced functionalities of the website, such as caching. These cookies do not store any personal information.
Cookie
Description
cf_ob_info
This cookie is set by Cloudflare content delivery network and, in conjunction with the cookie 'cf_use_ob', is used to determine whether it should continue serving “Always Online” until the cookie expires.
cf_use_ob
This cookie is set by Cloudflare content delivery network and is used to determine whether it should continue serving “Always Online” until the cookie expires.
free_subscription_only
This session cookie is served by our membership/subscription system and controls which types of content you are able to access.
ls_smartpush
This cookie is set by Litespeed Server and allows the server to store settings to help improve performance of the site.
one_signal_sdk_db
This cookie is set by OneSignal push notifications and is used for storing user preferences in connection with their notification permission status.
YSC
This cookie is set by Youtube and is used to track the views of embedded videos.
Analytics cookies collect information about your use of the content, and in combination with previously collected information, are used to measure, understand, and report on your usage of this website.
Cookie
Description
bcookie
This cookie is set by LinkedIn. The purpose of the cookie is to enable LinkedIn functionalities on the page.
GPS
This cookie is set by YouTube and registers a unique ID for tracking users based on their geographical location
lang
This cookie is set by LinkedIn and is used to store the language preferences of a user to serve up content in that stored language the next time user visit the website.
lidc
This cookie is set by LinkedIn and used for routing.
lissc
This cookie is set by LinkedIn share Buttons and ad tags.
vuid
We embed videos from our official Vimeo channel. When you press play, Vimeo will drop third party cookies to enable the video to play and to see how long a viewer has watched the video. This cookie does not track individuals.
wow.anonymousId
This cookie is set by Spotler and tracks an anonymous visitor ID.
wow.schedule
This cookie is set by Spotler and enables it to track the Load Balance Session Queue.
wow.session
This cookie is set by Spotler to track the Internet Information Services (IIS) session state.
wow.utmvalues
This cookie is set by Spotler and stores the UTM values for the session. UTM values are specific text strings that are appended to URLs that allow Communigator to track the URLs and the UTM values when they get clicked on.
_ga
This cookie is set by Google Analytics and is used to calculate visitor, session, campaign data and keep track of site usage for the site's analytics report. It stores information anonymously and assign a randomly generated number to identify unique visitors.
_gat
This cookies is set by Google Universal Analytics to throttle the request rate to limit the collection of data on high traffic sites.
_gid
This cookie is set by Google Analytics and is used to store information of how visitors use a website and helps in creating an analytics report of how the website is doing. The data collected including the number visitors, the source where they have come from, and the pages visited in an anonymous form.
Advertising and targeting cookies help us provide our visitors with relevant ads and marketing campaigns.
Cookie
Description
advanced_ads_browser_width
This cookie is set by Advanced Ads and measures the browser width.
advanced_ads_page_impressions
This cookie is set by Advanced Ads and measures the number of previous page impressions.
advanced_ads_pro_server_info
This cookie is set by Advanced Ads and sets geo-location, user role and user capabilities. It is used by cache busting in Advanced Ads Pro when the appropriate visitor conditions are used.
advanced_ads_pro_visitor_referrer
This cookie is set by Advanced Ads and sets the referrer URL.
bscookie
This cookie is a browser ID cookie set by LinkedIn share Buttons and ad tags.
IDE
This cookie is set by Google DoubleClick and stores information about how the user uses the website and any other advertisement before visiting the website. This is used to present users with ads that are relevant to them according to the user profile.
li_sugr
This cookie is set by LinkedIn and is used for tracking.
UserMatchHistory
This cookie is set by Linkedin and is used to track visitors on multiple websites, in order to present relevant advertisement based on the visitor's preferences.
VISITOR_INFO1_LIVE
This cookie is set by YouTube. Used to track the information of the embedded YouTube videos on a website.