HPLC method development can be accelerated by using three distinct method templates of increasing complexity.
High performance liquid chromatography (HPLC) method development is an arduous process requiring considerable experience and scientific judgment. This column installment presents a road map for rapid HPLC method development using a three-pronged approach. In this approach, three distinct method templates of increasing complexity are defined: fast LC isocratic methods, generic broad gradient methods, and multi-segment gradient stability-indicating methods. The characteristics, advantages, and limitations of each template are described, and case studies are used to illustrate their applications. The use of this template approach is expected to expedite the method development process, particularly during early-phase drug development.
High performance liquid chromatography (HPLC) method development is a labor-intensive and time-consuming task performed mostly by more-experienced scientists. Paradoxically, the best method development strategy is actually no method development — that is, if an acceptable method can be found elsewhere. In many cases, method development is unavoidable such as to support the development of new chemical entities (NCE) (for example, new drugs or chemicals). An HPLC analytical method typically consists of two major parts: the sample preparation procedure and the HPLC operating conditions. This column installment focuses on the latter part of the process by first summarizing a systematic method development strategy and by proposing a simple three-pronged template approach.
There is no shortage of information on HPLC method development. Useful information can be found in chapters of HPLC books (1,2), specialized books on method development (3–5) and pharmaceutical analysis (6,7), journal publications, short courses, and web resources.
An abbreviated synopsis of a "common method development strategy," extracted from references 2 and 3, is included here for the convenience of our readers. The reader is referred to the original sources for a fuller description of the steps highlighted below.
The most important question for the analytical method's goal is: Is the method for quantitation of the main component only (potency methods) or for purity determination (stability-indicating)? Another important question is: Is the method for quality control (QC)? QC methods and testing in a regulated environment typically have more-stringent method performance and robustness requirements. Sample type is also important because sample complexity dictates the column length and HPLC operating conditions, whereas the sample matrix may impose additional requirements in sample preparation and detection.
A thorough literature search can often provide a ready-to-use method or at least some useful starting points for method development. Knowledge of the analytes such as their chemical structure, molecular weight, purity, solubility, LogD value, number of acidic and basic functional groups and chiral centers, p _K_ a, absorbance maximum (λ max), toxicity, degradative pathways, reactivity, and stability are useful. Other valuable resources are material safety data sheets (MSDS), certificates of analysis (COA), and suppliers' technical packages. Unfortunately, information about the physicochemical properties of the molecule, while useful to avoid pitfalls, does not necessarily lead to an easier path for method development.
Before initiating method development to obtain the first chromatograms of the sample, some immediate decisions are needed: selection of HPLC mode, detection technique, column, and mobile phases. In many cases, some default or standard conditions appear to work well for the first "scouting" method; for instance, reversed-phase LC mode with a C18 column, UV detection (for chromophoric compounds), mobile-phase A: 0.1% acid in water and mobile-phase B: acetonitrile or methanol. The most common mobile-phase A additives are: trifluoroacetic acid, formic acid, or phosphoric acid. The actual column dimension and particle size of the packing are dictated by the complexity of the sample and the method's goals (2,3). For nonchromophoric analytes, refractive index, evaporative light-scattering, charged-aerosol, or mass spectrometry (MS) detection can be used. For isocratic analysis, a process of sequential isocratic steps is generally effective (2,8). This is accomplished by lowering the solvent strength of the mobile phase until all key analytes are retained and resolved. For purity testing or separation of more complex mixtures, a broad linear gradient separation from 5% to 100% mobile-phase B is used to generate the first chromatograms. Typically, this approach reveals the number of components and overall purity of the sample. The molecular weights and λ max values of the analytes are easily obtained using MS and photodiode-array detection, respectively.
For purity analysis, test mixtures (cocktails) of process precursors, impurities, degradants, additives, and excipients are used to confirm that the method is able to separate all of these components (that is, to establish that the method is specific). Mother liquors from the final crystallization step and forced degradation samples are used if reference standards of impurities are not available (2,6,7). The use of forced degradation samples is required to establish the stability-indicating nature of a method. Screening of different columns and mobile phases (pH, buffers, organic solvents), and adjustment of operating conditions (temperature, flow rate, gradient time [_t_ G], gradient range — that is, starting and ending %B, and single or multiple-segment gradients) are used to resolve all key analytes. Finally, the HPLC conditions are optimized for sensitivity, peak shapes, and analysis time.
The fine-tuning step, often called "selectivity tuning," is the most time-consuming step in the method development process. According to Lloyd Snyder, a pioneer in HPLC and a champion in simulation software, the "trial-and-error" approach of varying one factor at a time (OFAT) still reigns in HPLC method development (9). This often "graduates" to an "enlightened trial-and-error" approach for the experienced scientist with a good understanding of chromatographic principles who can establish the final method conditions in fewer steps. With the numerous factors controlling retention during gradient elution, this OFAT strategy is clearly not a very efficient process, even for the expert. Here, the use of simulation software (such as DryLab [Molnar Institut or ChromSword Group]) and automation systems (for example, a column–mobile phase screening system, Fusion AE QbD [S-Matrix], or AutoChrom [ACD/Labs]) is particularly helpful for the optimization process of methods for difficult samples (2,8–10). For the primary stability-indicating assays used in drug development and QC, a systematic and thorough method development process is increasingly becoming a regulatory expectation if not yet a requirement (11). It is worthwhile to note that HPLC method development during clinical development of new drugs is often an iterative process, repeated many times to accommodate the resolution of unexpected impurities found in new synthetic routes or formulations. To conserve resources, a proactive "stage-appropriate method development and validation approach" is adopted by many pharmaceutical laboratories (2,6).
Method validation (or qualification) is needed to ensure data accuracy and "fit for purpose." It is a mandatory requirement for quality control or regulated testing to demonstrate that the method is "suitable for its intended use." Therefore, the last step of method development is often a prequalification step (checking for specificity, linearity, precision, and sensitivity) to demonstrate that the method is "validatable." Method validation is a good manufacturing practice (GMP) process and method development is not a GMP activity. Thus, it is prudent to make sure that the newly developed method can be validated before the execution of the formal method validation protocol.
The common strategy described above (or other similar process) is generally accepted by most practitioners. Nevertheless, some shortcomings and potential improvements of the process are noted here:
The three-pronged template approach is a concept that helps to expedite the method development process by providing three common method templates and a set of practical ground rules for the selection of templates, columns, and operating conditions. It is a simple, pragmatic approach with a focus on reversed-phase HPLC using UV detection. Figure 1 illustrates the concepts of the three templates of increasing complexity (fast isocratic LC, generic broad gradient, and multisegment gradient). Characteristics, applications, and limitations of each method template are described and illustrated with case studies below.
Figure 1: Diagram depicting the three method templates in the three-pronged approach and their relative ease for implementation. ICH = International Conference on Harmonization. IPC = in-process control.
For potency testing (that is, quantitative determination of the main component), a fast LC isocratic method is recommended for simplicity and speed. Figure 2 is an example of a fast LC isocratic method using a short column (for example, a 50 mm × 4.6 mm C18 column) with UV detection at λ max of the analyte. A recommended retention factor (_k_) of ~1–2 with an analysis time <2 min appears to be feasible in most cases. Application examples are potency assays such as dosing solutions analysis, and content uniformity and dissolution testing of drug products. Method development is very quick using sequential isocratic steps to retain the analytes away from the solvent front (2). Alternately, the optimum isocratic %B can be obtained from a linear broad-gradient (for example, 5–100% B) scouting run. Higher retention may be needed to separate the main component from major interferences. These potency methods can be developed and prequalified in hours. However, they are not stability-indicating methods and have limited resolution or peak capacities. One potential pitfall of a fast LC method is the possible interference from a hydrophobic component such as antioxidant additives in drug product formulations. The use of gradients or isocratic runs with column purging steps is recommended in such cases.
Figure 2: An example of the "fast LC isocratic potency and performance method template" used for content uniformity and dissolution (performance) testing of a drug product (capsule). Column: 50 mm à 4.6 mm, 3-µm dp Waters X-Bridge C18; mobile phase: 30% acetonitrile, 70% water with 0.05% trifluoroacetic acid; flow rate: 1.0 mL/min at 30 °C; detection: 280 nm (λmax of the API); sample: 10 µL of a drug product extract containing ~0.1 mg/mL of API in 1 N hydrochloric acid. Note this potency method is nonstability-indicating with k = 1.0 for the analyte.
The use of generic broad-gradient HPLC methods to monitor reaction progress is very common in organic synthesis laboratories. Broad, linear gradients in reversed-phase mode are typically used (for example, 5–95% B or 100% B) because they usually can elute all components and separate most of them in the sample. Generic gradient methods are also the standard practices in process scale-up laboratories for in-process control (IPC) testing and in medicinal chemistry laboratories for rapid characterization or purification of crude compounds. Multiple UV wavelength monitoring (such as 220, 254, and 280 nm) or mass spectral data are typically recorded.
In contrast, scientists in analytical development or QC laboratories tend to develop custom methods for each NCE in their project, leading to a proliferation of methods with myriad combinations of columns and mobile phases. It appears that the generic gradient approach can be readily applied to purity analysis of many sample types (such as drug substances, drug products, raw or starting materials, and reagents) and other applications (for example, cleaning verification). Figure 3 shows an example of a purity assay method for a starting material (Boc-β-amino acid) using HPLC conditions similar to those used in a typical IPC method. Figure 3a is a chromatogram of a cocktail mixture of the starting material reference standard (F) and all its synthetic precursors. Figure 3b is a chromatogram of a starting material with a purity value of ~98% using an area percent calculation.
Figure 3: A case study of the "generic broad-gradient method template" used for the purity analysis of a starting material used in the synthesis of a new chemical entity: (a) Chromatogram of a standard solution containing the starting material (F) with its precursor compounds (A to E and G) at ~0.1 mg/mL each. (b) Chromatogram showing the purity profile of the starting material (F) at 0.5 mg/mL with an area percent of 97.8%. Column: 150 mm à 4.6 mm, 3-µm dp ACE-3-C18; mobile-phase A: 0.05% trifluoroacetic acid in water; mobile-phase B: 0.03% trifluoroacetic acid in acetonitrile; flow rate: 1.0 mL/min at 35 °C; gradient program: 5â95% B in 15 min, 95% for 2 min, 95â5% B in 0.1 min; detection: 220 nm.
The selection of detection wavelength in HPLC–UV analysis is rarely discus
