16 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING Joseph Etse 16.1 INTRODUCTION Analytical technology transfer and manufacturing is the mechanism by which knowledge acquired about a process for making a pharmaceutical active ingre- dient or dosage form during the clinical development phase is transferred from research and development to commercial scale-up operation or shared between internal groups or with third parties. Analytical technology transfer guarantees that laboratories can routinely execute tests, obtain acceptable results, and be able to accurately and independently judge the quality of com- mercial batches. One of the most important analytical technology transfers is high-performance liquid chromatography (HPLC) methods. The success or failure of analytical technology transfers are judged on the merits of data gen- erated using HPLC. Consequently, a major focus of regulatory authorities [1–3] is on methods transfer as a critical link in the drug development con- tinuum. Depending on the structure of the pharmaceutical organization,trans- fer of analytical technology and manufacturing may occur at the end of the phase II clinical studies or during the transition from phase II to phase III. However, for a successful transfer of analytical technology to occur, the exis- tence of HPLC methods that have been fully validated in accordance with the ICH guidelines on validation will be required [4–7]. A full description of method validation is provided in Chapter 9. 735 HPLC for Pharmaceutical Scientists, Edited by Yuri Kazakevich and Rosario LoBrutto Copyright © 2007 by John Wiley & Sons, Inc. 16.2 PREREQUISITES FOR TRANSFER OF HPLC METHODS 16.2.1 Availability of Either Fully or Partially Validated Methods A prerequisite for the transfer of analytical technology is the establishment of fully validated methods in accordance with the International Council on Har- monization (ICH) [4, 5],United States Pharmacopeia (USP) [6] and European Pharmacopeia (EP) [7] guidelines for method validation, the existence of a final synthetic process for the active pharmaceutical ingredients (APIs), and final market image of the pharmaceutical dosage form. Method development and validation usually parallels the API and pharmaceutical dosage form development. It progresses from very rudimentary Tier 1 methods with limited validation as shown in Table 16-1 through to Tier 2 methods and culminating in Tier 3/registration-type methods [8–10]. Differences between methods from Tier 1 through Tier 3 are due to the extent of validation of the analytical figures of merit that is performed [3]. During early development of the active pharmaceutical ingredient and early dosage form development, emphasis is placed on speed and quantitation of the API. At this stage, methods rely on the use of short columns, fast flows, and very minimum validation to quickly identify the most desirable synthetic route for the API that will produce an adequate impurity profile (overall yield may not be optimized at this stage) and most desirable prototype formulations and excipients that will ultimately lead to the selection of the final formula- 736 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING TABLE 16-1. Progressive Validation of Analytical Figures of Merit Progressive Method Development Discovery/ Phase I Phase II Phase III Phase IV Analytical Figures of Merit Tier 1 Tier 2 Tier 3 Registration 1. Linearity √√√ √ 2. Range √√√ √ 3. Accuracy — √√ √ 4. Specificity/stress studies — √√ √ 5. Precision • Repeatability (injection) √√√ √ • Intermediate precision (API) — √√ √ • Intermediate precision (RS) — — √√ 6. Robustness — — √√ 7. Solution stability √√√ √ 8. Limit of detection (LOD) — — √√ 9. Limit of quantitation (LOQ) — √√ √ √, Validated; —, not validated; R&D, discovery research, API, active pharmaceutical ingredient; RS, related substance. Source: Reprinted from Am. Pharm. Rev. Vol. 8(1), (2005), 76, with permission. tion. Typically, Tier 1 and Tier 2 methods include validation of some, but not all, of the analytical figures of merit as shown in Table 16-1. Tier 1 methods are the simplest methods in the sense that only linearity and precision may have been validated. As the synthesis scheme for the API becomes optimized with respect to improving the overall API yield and as the dosage form devel- opment evolves from prototype formulations to the more robust final market image formulations, the analytical methods employed also evolve and become increasingly robust and optimized for the quantitation of the API as well as degradation products and related substances. Once the final synthesis is set and final formulations are selected, more robust and fully validated Tier 3 methods [3] are established to ensure the successful transfer of analytical tech- nology from research and development to commercial operation. Supple- mentary to this prerequisite is the identification of the commercial production site or launch site where the pharmaceutical dosage form will be manufac- tured. This is usually the stage in which all the drug development activities come together in a New Drug Application (NDA) for regulatory approval (Figures 16-1 and 16-2). Phase IV methods are usually slight variations of Tier 3 methods which include but are not limited to calculation formulas, the number of sample preparations for API, and the number of dosage units. PREREQUISITES FOR TRANSFER OF HPLC METHODS 737 Figure 16-1. Drug development and method transfer continuum for pharmaceutical active ingredients (APIs). 16.2.2 Availability of the Finalized Pharmaceutical Active Ingredient (API), Known Degradation Products, By-products and Reference Standards Besides the existence of validated HPLC methods, the availability of a final- ized API synthetic scheme and optimized formulations is another prerequisite for ensuring a successful transfer of analytical technology from research and development to commercial operation (Figure 16-1). Evaluation of data gen- erated from the HPLC analysis of the API provides the means by which deter- mination is made about whether a validated API synthetic scheme exists and if the API can be made reproducibly during commercial operation. Confirmation of the existence of a validated API synthetic process is based on the interpretation of acquired nuclear magnetic resonance (NMR) spec- trum of the API in conjunction with definitive molecular formula for the API and its components based on mass spectroscopy (MS). Concluding that a reproducible API manufacturing process exists is based on whether purity profiles of subsequent drug substance batches retain the same profile of the API and its related substance peaks as was in the reference material. In that sense, data generated using HPLC and hyphenated HPLC techniques such as HPLC/MS, HPLC-MS/MS, and HPLC/NMR serve as the foundation for declaring the existence of reproducible API manufacturing process [11–13]. 738 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING Figure 16-2. Drug development and method transfer continuum for pharmaceutical dosage forms. 16.2.3 Availability of Drug Products Made by the Definitive Manufacturing Process Once the API has been selected for further development, a clear definition and demonstration of the validated status of the manufacturing process for the pharmaceutical dosage form is required in order to initiate transfer of ana- lytical technology and manufacturing process. Development of the phar- maceutical dosage form consists of a series of experimental activities that ultimately result in the transformation of the API into a dosage form (i.e., tablet, capsule, suspension, injectable, patch, creams, inhalation product) suit- able for human use [14]. This is achieved through the manufacture of batches at the chosen commercial site for scale-up production (Figure 16-2). As the formulation development process evolves into the final market image dosage form, more robust HPLC methods as shown in Table 16-1 are established [3]. Assessment of whether or not the transfer of manufacturing process has been successful is usually done by sampling and analyzing batches for blend uni- formity [15, 16]. Sampling of the batches can be done in a number of ways. Sampling by variables or the Bergum approach [17] are examples of sampling techniques that can be used to measure the quality characteristics of the batch on a continuous statistical scale. A simple method of measurement used to judge the quality of the batch is the sample mean value (SMV) and its corre- sponding percent relative standard deviation (%RSD). A sample mean value that is close to the expected target value of 100% with a corresponding low %RSD serves to prove the attainment of a homogeneous product that is uniform and not variable in the content of the API. Assessment of product quality using the Bergum approach relies on whether selected samples of a batch has an SMV and %RSD that match the Bergum acceptance criteria.The process is rejected if the SMV and %RSD lie outside the allowable Bergum acceptance criteria [17–19]. An example of how the Bergum acceptance crite- ria are applied for evaluating content uniformity data generated using HPLC is shown in Figure 16-3. A sample with an SMV of 95% will pass the Bergum criteria, provided that its corresponding %RSD is less than or equal to 3.5. Similarly, a sample with a %RSD of 2.0 must have a corresponding SMV of either 90% or 109% to pass the Bergum acceptance criteria.The general trend as seen with Bergum acceptance criteria is that an increase in the magnitude of the %RSD and its corresponding sample mean value (SMV) follow a bell- shaped distribution in such way that %RSD decreases to zero as the SMV becomes greater than or less than 100%. 16.2.4 Availability of Suitable Instruments and Personnel A critical activity that precedes the start of analytical technology transfer is to assess whether suitable instrumentation and qualified personnel are available at the receiving laboratory.The receiving laboratory is the laboratory to which the analytical methods are transferred to. This assessment is accomplished through the organization of an analytical challenge meeting. The purpose of PREREQUISITES FOR TRANSFER OF HPLC METHODS 739 this meeting is usually to open up channels of communication between the participating laboratories to discuss the methods that are going to be trans- ferred, to share and exchange knowledge about the idiosyncrasies of the methods, and to agree on the types of tests to perform and samples that will be used for the cross-over testing. Because of the potential impact of sample- to-sample variability on the agreement between data generated, participating labs should agree to use identical batches and also decide on the number and types of batches that should be tested [17–20]. Additionally, identification of key contact persons and assignment of responsibilities are other items that can be agreed upon during the analytical challenge meeting. 16.2.5 Availability of a Protocol Containing Predetermined Acceptance Criteria The methods transfer protocol is the main driver that governs the conduct of the experiments and ensures that assessment of results generated is not unduly influenced by biases due to either (a) the analytical method or (b) inherent batch-to-batch variability of the active pharmaceutical ingredient or pharma- ceutical dosage form. The methods transfer protocol establishes the predeter- mined acceptance criteria by which results will be judged to have either passed or failed the methods transfer.The criteria for assessment of success or failure contained in the methods transfer protocol is achieved through an iterative 740 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING Figure 16-3. Bergum acceptance curve for evaluating content uniformity (CU) data. Note:The curve shows that for a relative standard deviation (%RSD) of 4.5%, a sample mean value (SMV) of 100% must be achieved for the manufacturing process to be judged as validated. process of exchange of ideas and comments between the originating lab and the participating or receiving labs . Since the aim of the protocol is to ensure the mitigation of problems, the essential elements of the protocol consists of sections that include (a) an Introduction, (b) treatment and disposition of data,(c) types of methods being transferred, (d) materials, reference standards, and reagents being used, (e) recommended type of equipment, (f) sample handling, (g) predetermined acceptance criteria, and (h) an Acknowledgment section. An example of a typical table of contents (TOC) of an analytical methods transfer protocol is discussed in Table 16-2. PREREQUISITES FOR TRANSFER OF HPLC METHODS 741 TABLE 16-2. Table of Contents of an Example of Analytical Transfer Protocol Section Description 1. Introduction • Aim and scope of the protocol • Brief description of dosage forms • Arguments for waivers, bracketing, and matrixing (if applicable) 2. Treatment and disposition of data • Format for reporting data (i.e., number of decimal places • Archival of data • Handling and resolution of OOS or OOT results 3. Materials, standards, and reagents • Batches to be tested including batch- specific data (storage conditions, shelf life, etc.) • Reference substances (including storage conditions and expiry dates) • Special handling instructions or precautions (if applicable) • Reagents and source of supplier(s) 4. Test methods and specifications • Lists all applicable test methods and specifications • List methods that will not be transferred by cross-over testing 5. Acceptance criteria • Lists pass or fail requirements • Enumerates the release acceptance requirements • Stipulates the number of required replicate determinations • Establishes the statistical assessment requirements 6. Method acknowledgment • Feedback from participating labs • Signature of acceptance/approval (receiving and transfer laboratory) OOS, out-of-specification results; OOT, out-of-trend results. Source: Adapted from Am. Pharm. Rev. 8(1) (2005), 76, with permission. 16.2.5.1 Introduction. T he introduction section lays down the purpose of the transfer and clearly identifies the originator lab where the method was developed as the reference lab (also designated as center A or originator lab) and identifies all the other labs participating in the transfer as participating or receiving labs. A brief description of the pharmaceutical dosage forms being transferred as well as any arguments for bracketing and matrixing of the testing plan is included in this section. Because a minimum of three batches of each dosage strength of a pharmaceutical dosage form is normally required to be to tested during the transfer activities [17–19], any argument that serves as justification for omitting the testing of certain dosage strengths through the application of bracketing and matrixing strategies becomes an important con- sideration that is addressed in this section. Bracketing and matrix testing is the approach in which only the highest and lowest dosage strengths are tested in the cross-over experiments and carry the benefits of reducing the amount of testing required. When this approach is applied to a pharmaceutical dosage form consisting of 50-, 75-, 100-, 125-, and 150-mg dosage strengths, cross-over testing is performed using only the 50-, 100-, and 150-mg dose strengths. In that sense, passing results generated for the 50-, 100-,and 150-mg strengths automatically becomes surrogate and proof of the transfer for the 75- and 125-mg dose strengths. Usually, arguments for bracketing and matrixing are easier to justify if the analytical sample prepa- ration steps for all dosage strengths are similar and if the dosage strengths are made from similar or identical granulation and the excipient-to-drug content ratio is dose and/or weight proportional. 16.2.5.2 Treatment and Disposition of Data. This section discusses treat- ment and disposition of data and establishes the mechanism by which data will be assessed as having passed or failed the predetermined acceptance criteria. It is important that this section also address (a) the mechanism by which out- of-specification (OOS) or out-of-trend (OOT) results will be handled and (b) procedures for reporting and archiving of data. Stipulation of how assessment of reported results and the archival and disposition of data will be handled should also be discussed under this section. Clear rules governing how assess- ment of data is handled takes on an ever-increasing significance because of the intense scrutiny regulatory agencies apply when reviewing any docu- mented report that claims equivalency based on data comparison between different laboratories [20–22]. 16.2.5.3 List of Materials, Standards, and Reagents. The impact that batch-to-batch variability of a pharmaceutical dosage form can have on the interpretation of results can be very challenging [17, 18]. The same challenges are also present for the API because different vendors may be used for the raw materials that may lead to different impurity profiles of the API. Hence, the purpose of this section of the protocol is to describe in unequivocal terms 742 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING the analytical methods, test samples, standards, and reagents that should be used. An approach that greatly helps to alleviate potential differences that lead to non-agreement of data between the participating and originator labs due to the potential impact of batch-to-batch variability is to use either freshly manufactured or well-characterized batches.This can be accomplished by pro- viding all the participating labs with identical samples derived from previously prepared composite samples or aliquots from the same batches of drug product (DP) and drug substance (DS). The prepared composite samples for the DP or DS are subdivided evenly into smaller lots that are distributed to all participating laboratories. Because the ultimate goal of the technical trans- fer is the transfer of analytical test methods and not the identification of inher- ent batch-to-batch variability that may exist among batches, it is essential to stipulate the lot number, date of manufacture and the required storage con- dition of the batches selected. For example, a storage condition of “5°C” or “Do not store above 30°C,” together with any special handling instructions, and the appropriate “Re-test” or “Expiry date (i.e., 2 years, etc.),” should be provided [23, 24].Additionally, it may also be necessary to stipulate and restrict the number of vendors/suppliers of reagents and reference standards as a means by which potential biases from such sources on the results generated can be eliminated or minimized. For example, if in-house reference standards are used, it may be necessary to quarantine that particular lot of reference standard material to ensure that adequate supplies are available at all times for the duration of the analytical methods transfer. Material Safety and Data Sheet (MSDS) classification of the reagents, standards, and the API should be provided to assure the types of adequate precaution that should be taken to avoid unintended exposure of analysts to potentially dangerous materials [23, 24]. 16.2.5.4 Test Methods and Specifications. This section tabulates all the test methods and specifications that are being transferred via an interlaboratory cross-over testing plan. In addition, a discussion of the rationale for transfer- ring the methods should be provided in this section. When transfer of related substances methods is required, a number of steps should be considered to ensure that data generated will be reasonable and meaningful for the purposes of comparing data between the different laboratories.Transfer of related sub- stances methods often presents a challenging situation because related sub- stances tend to occur in low levels,especially in recently manufactured batches, or may be present at levels that are close to the quantitation limits of the method. To circumvent this, a very well-defined strategy can adopted by the participating labs to mitigate difficulties associated with the transfer of related substances. For example, clear instructions regarding whether a spike experiment involving the use of pre-prepared samples containing a known amount of the degradation product or the use of samples containing well- characterized amount of the degradation product of interest can be stipulated. Also, clear definition of the sample preparation in regard to sonication (power PREREQUISITES FOR TRANSFER OF HPLC METHODS 743 output, water level in batch), shaking, filtration steps, and so on, must be communicated [25–29]. Another aspect of this section of the protocol is to discuss the rationale for not transferring certain methods . In general, test methods such as microbial limit tests (MLT) and other types of test methods based on universally accepted pharmacopeia procedures are not typically subjected to interlabora- tory cross-over testing. Instead, such methods are validated locally by demon- strating suitability for their intended purpose. Similarly, test methods that rely on the use of well-established techniques that are considered routine such as “appearance by visual examination” and “uniformity of dosage units by weights” are not normally subjected to interlaboratory cross-over testing. However, if some of the more common tests such as “Appearance of Solution” tests require special pharmacopeia procedures that are not routine, it may be necessary to consider the inclusion of such methods in the interlaboratory cross-over testing scheme.Any decision that is contemplated to either include or exclude a test method or methods from conventional cross-over testing should clearly be justified on the basis of sound scientific argument(s). 16.2.5.5 Acceptance Criteria. A most important section of the protocol is the section that outlines all the acceptance criteria by which results generated for each method being transferred will be judged as having fulfilled the requirements of the transfer. Since the interpretation of acceptance criteria is often based on the application of some type of statistical value, an important aspect of this section is to include clear instructions regarding the number of batches and the number of replicate determinations that has to be performed [18, 19]. Table 16-3 shows examples of acceptance criteria that may be applied 744 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING TABLE 16-3. Example of Acceptance Criteria for Assay, Content Uniformity, and Dissolution Test Replicates (n) Acceptance Criteria Assay 3 • Mean difference ≤ 2.0% [%RSD, n = 6) ≤3.0%] Content uniformity (CU) 10 • Results meet current USP <905> and/or harmonized USP, JP, and EP 2.6.1 requirements. • If %RSD at either site is >4.0%, then [STDev (Lab B)/STDev (Lab A)] ≤2.0% • For each site, SMV =±3% of the mean assay within each site Dissolution 12 • Mean difference ≤7.5% SMV, sample mean value; RSD, relative standard deviation; n, number of replicates; USP, United State Pharmacopeia; JP, Japanese Pharmacopeia; EP, European Pharmacopeia. Source: Adapted from Am. Pharm. Rev. 8(1) (2005), 77, with permission. [...]...TYPES OF TECHNICAL TRANSFER 745 for comparing assay, content uniformity, and dissolution data between participating labs For the purposes of comparing assay, content uniformity, and dissolution data, simple statistics such as sample mean value (SMV) and relative standard deviation (%RSD) derived from experience of performing the tests over long periods of time can be used as... divert resources to perform additional experiments in order to resolve the unexpected analytical issue Though not a necessary requirement for this mode of transfer, it may still be prudent for the 752 THE ROLE OF HPLC IN TECHNICAL TRANSFER AND MANUFACTURING participating labs to show that by using their versions of the methods, comparable data with the originator is obtained for the assay of similar... to even cause peak retention reversals [10, 26, 29] Therefore, precise instructions for the preparation of mobile phases including pH adjustment must be provided to ensure that uniform procedures for mobile-phase preparation are followed 16.5.2 Sample Type and Number of Replicate Determination One of the challenges faced during the transfer of HPLC methods is to ensure that samples used in the transfer... selectivity for the separation performed in the pharmaceutical company, since the manufacturing test probes are inept in discriminating between the selectivity differences It is recommended to discuss with the vendor if any manufacturing changes have been implemented for a particular stationary phase If so, columns should be obtained from lots after the manufacturing change was implemented For point... in peak profiles and chromatographic performance (efficiency, peak width, S/N) For example, different values for digital filtering in pre-data processing and differences in values for data smoothing (Savitzy–Golay) in post-data processing can lead to differences in peak profiles It is recommended to do post-data processing so original data is kept and no analytical information is lost 16.6 CONCLUSION Technical... reliance on data generated using HPLC in the go/no-go decision gates along the drug development continuum highway requires that HPLC methods are properly validated and transferred in an efficient fashion Efficient and seamless transfer of HPLC methods can avoid any delays in the regulatory submission process and the ultimate introduction of new pharmaceutical dosage form to the public Hence, it is proper... receiving lab for a considerable period of time Identification by visual examination and uniformity of dosage unit by a weighing operation are two examples that merit consideration for waiver transfer Another circumstance under which a waiver may be justified is when it can be proven that an earlier version of a method has routinely been used to test products at the receiving lab facility A strong case for waiver... sonication time to account for a more robust extraction of analytes A poignant argument that can be made for waiver of transfer is when personnel from the lab where the methods were originally developed and validated are transferred to a brand new lab in a completely different facility In this case the granting of waiver is justified because the expertise required for performing the methods already... the sequence of mixing water and methanol that caused volume expansion was attributed to be the cause for the observed k value differences In another reported example, it was noted that chloroform containing no preservative versus chloroform containing 0.5% ethanol as preservative was responsible for the same types of compounds to elute differently [29–31] using normalphase chromatography Additionally,... testing, which requires the existence of fully validated methods, there is no requirement for the existence of a validated method for transfer based on co-validation In other words, availability of a fully validated method is not a prerequisite for this mode of transfer This type of transfer is usually employed for transferring methods from one AR&D organization to another ARD organization or from an . 9. 735 HPLC for Pharmaceutical Scientists, Edited by Yuri Kazakevich and Rosario LoBrutto Copyright © 2007 by John Wiley & Sons, Inc. 16.2 PREREQUISITES FOR. that sense, data generated using HPLC and hyphenated HPLC techniques such as HPLC/ MS, HPLC- MS/MS, and HPLC/ NMR serve as the foundation for declaring the existence