Use of Orthogonal Methods During Pharmaceutical Development: Case Studies

During early phase pharmaceutical development, drug substance synthetic routes, and drug product dosage forms are selected. This is typically an iterative process that from a chromatographic perspective requires high performance liquid chromatography (HPLC) methods to separate a potentially different set of impurities and degradation products as drug substance and drug product development advance. At the same time, methods are required for release and stability testing of clinical supplies to assure that the products are safe and effective in vivo. For this application, the methods need to unequivocally monitor all impurities and degradation products to assure that guidelines for reporting, identification and toxicological qualification (1,2) are met.

Results and Discussion

Our approach to method development, in context of the cited objectives, has been described in detail previously (3,4). In summary, the approach consists of:

• Obtaining all available batches of drug substances (and drug product, where available) to assure that all synthetic impurities are assessed as part of method development. Additionally, potential degradation products are generated via forced decomposition studies (5,6). Forced decomposition can be performed manually, or can be automated (7,8). Following forced decomposition, samples are stored at –20 °C to minimize further degradation and allow for future use (5).

• The samples generated in the first step are screened by a single chromatographic method. This can be either a method established during drug discovery or a generic broad gradient method. The study is conducted to identify samples for further method development. The selected samples based upon this first screen consist of each drug substance (drug product) lot with a unique impurity profile, and samples of interest from the forced degradation studies.

Where multiple samples are generated for a single degradation pathway, samples degraded 5–15% are selected. Solutions degraded above 15% have a larger risk of containing secondary degradation products that might not be formed under less stringent stress conditions. Accordingly, results for these samples can be misleading (4). If no degradation is observed for a condition, further studies on solutions obtained via the condition are deferred for later analysis. Notably, at this point, only a single chromatographic method has been used for evaluation. This method has not been demonstrated to be stability indicating, but simply has been used to select the samples for method development. Accordingly, all samples should be retained.

Table I: Orthogonal Screening: Method Description

• The samples of interest from the second step are screened using six broad gradients on each of six columns (that is, 36 conditions for each sample). The mobile phases are chosen to be broad gradients to minimize elution at the solvent front (no retention) or nonelution of sample components. Typical conditions are shown in Table I. In each experiment the gradient is kept constant, but a different pH modifier is used. The modifier is prepared at 20× the required concentration and added to the mobile phase at a constant 5% (v/v). The modifiers typically used and the associated pH, expressed at the nominal concentration (that is, 5% of stock concentration), are provided in Table II.

Table II: Orthogonal screening: mobile phase modifiers

Columns are selected based upon anticipated selectivity differences. Accordingly, columns with different bonded phases are evaluated. A possible column set is provided in Table III. It is emphasized that this is not a stagnant set of columns. As new columns are introduced, particularly columns that provide novel selectivity, revision to the set should be considered.

Table III: Orthogonal screening: columns

• Based upon the results of the experiments in the third step, a full picture of the components that need to be separated emerges. Typically, a condition from the screen will separate all the components of interest. If this is not achieved further method development, or modification, is required. As important as obtaining the separation described, using our approach, is to obtain an additional method that provides very different selectivity to the primary method, that is, is orthogonal. The orthogonal method can be used to evaluate the primary method on an ongoing basis, to assure that the primary method remains specific if new synthetic impurities or degradation products are formed in subsequent batches of drug substance or drug product.

• To further examine the two selected methods, the samples previously identified as containing degradation products, along with the most stressed samples obtained under other stress conditions are analyzed under both sets of conditions, to assure that no peaks were missed by the initial generic gradient.

Following the completion of orthogonal screening, software tools such as DryLab (LC Resources, Walnut Creek, California) (9–11) are used to help optimize both the primary and orthogonal methods. The optimization includes changing column conditions such as column dimensions, particle size, flow rate, and column temperature; changing solvent strength by adjusting the gradient steepness or replacing acetonitrile with methanol or acetonitrile–methanol mixtures; and changing the concentration of the modifier.

Figure 1: Chromatograms of two API batches of compound A analyzed by the primary method. Column: 150 mm × 4.6 mm, 5-μm Zorbax XDB-C8; temperature: 25 °C; gradient: 25 min with acetonitrile and water, with 0.1% formic acid as a mobile phase modifier.

Following optimization, the primary method is validated (12) and is used for release and stability testing. The orthogonal method is used to screen samples from new synthetic routes and pivotal stability samples of both drug substance and drug product. This ensures, as a first pass, that all peaks of interest are reported using the release method and triggers the need for method redevelopment or use of additional control methods if additional, new peaks are observed with the orthogonal method. Test cases follow to illustrate the point.

Figure 2: Chromatograms of two API batches of compound A analyzed by orthogonal method. Column: 150 mm × 4.6 mm, 3-μm Phenomenex Curosil PFP; temperature: 25 °C; gradient: 30 min with acetonitrile, methanol, and water and 0.1% trifluoroacetic acid as a mobile phase modifier.

Case 1

Based upon the evaluation of the screening results, a selected and optimized HPLC method was used to release batches of compound A. No new impurities were detected when a new active pharmaceutical ingredient (API) batch was analyzed by the primary method (Figure 1). With the orthogonal method, however, the coelution of impurities (A1 and A2) and the highly retained compounds (dimer 1 and dimer 2) were detected in the new API batch (Figure 2).

Figure 3: Chromatograms of two API batches of compound B analyzed by the primary method. Column: 150 mm × 4.6 mm, 3-μm YMC-Pack Pro C18; temperature: 25 °C; gradient: 35 min with acetonitrile and water with 0.1% trifluoroacetic acid as a mobile phase modifier.

Case 2

Analysis of a new drug substance lot of compound B with the primary method (Figure 3) showed the appearance of a 0.40% impurity. Using the orthogonal method (Figure 4), this peak was shown to be the result of coeluted compounds (impurity A and impurity B). A previously unknown isomer of the API was additionally detected.

Figure 4: Chromatograms of two API batches of compound B analyzed by orthogonal method. Column: 150 mm × 4.6 mm, 3-μm YMC-Pack Pro C18; temperature: 50 °C; gradient: 35 min with methanol and water and 0.02% trifluoroacetic acid as a mobile phase modifier.

Case 3

For compound C, one impurity (impurity 1) with w/w% greater than 0.05 was detected when the first drug substance batch was analyzed by the primary method (Figure 5). The same impurity (impurity 1) also was observed by the orthogonal method (Figure 6). In the new batch, a second impurity (impurity 2) not previously observed, is similarly observed by both methods. However, the orthogonal method detects a third component (impurity 3) at 0.10% (w/w). This component is coeluted with the API in the primary method.

Figure 5: Chromatograms of two API batches of compound C analyzed by the primary method. Column: 150 mm × 4.6 mm, 3.5-μm Zorbax XDB-C8; temperature: 30 °C; gradient: 30 min with acetonitrile and water and 0.05% trifluoroacetic acid as a mobile phase modifier.

Conclusions

HPLC methods are necessary to support clinical release and stability activities and evolutionary synthesis and formulation optimization. To address changes in the impurities and degradation product profiles generated during these activities, a systematic approach to method development, using an array of methods, is advocated as a means of obtaining full knowledge of drug substance and drug product chemistry. The approach places heavy emphasis on the use of orthogonal screening as a central activity, to gather the knowledge required as a prerequisite to developing methods suitable for the control of commercial supplies of drug substance and drug product. Key elements of the approach include the generation of degradation products via forced decomposition and a continual evaluation of samples generated during the early development cycle.

Figure 6: Chromatograms of two API batches of compound C analyzed by the orthogonal method. Column: 150 mm × 4.6 mm, 3-μm Phenomenex Gemini C18; temperature: 40 °C; gradient: 30 min with acetonitrile and water with 5 mM ammonium acetate as a mobile phase modifier.

Acknowledgments

The authors wish to thank the numerous members of the Global Johnson & Johnson Pharmaceutical Research and Development, LLC Analytical staff for their contributions to the design of the processes discussed herein.

Henrik T. Rasmussen and Fengmei Zheng

Johnson and Johnson PharmaceuticalResearch and Development, LLC, Raritan, NJ

Please direct correspondence to HRasmuss@its.jnj.com

References

(1) ICH Guideline Q3A (R1). Impurities in New Drug Substances (2002). http://www.ich.org/LOB/media/MEDIA422.pdf.

(2) ICH Guideline Q3B (R1). Impurities in New Drug Products (2003). http://www.ich.org/LOB/media/MEDIA421.pdf.

(3) H.T. Rasmussen, W. Li, D. Redlich, and M.I. Jimidar, "HPLC Method Development," in Handbook of Pharmaceutical Analysis by HPLC, S. Ahuja and M.W. Dong, Eds. (Elsevier, Amsterdam, 2005), Chapter 6.

(4) H.T. Rasmussen, K.A. Swinney, and S. Gaiki, "HPLC Method Development in Early Phase Pharmaceutical Development," in Handbook of HPLC Method Development for Pharmaceuticals, S. Ahuja and H.T. Rasmussen, Eds. (Elsevier, Amsterdam, 2007), Chapter 12.

(5) M. Bakshi and S. Singh, J. Pharm. Biomed. Anal. 28, 1011 (2002).

(6) R.A. Jackson, Am. Pharm. Rev. 10(5), 59 (2007).

(7) A.M. Fermier, B.L. Armstrong, A.R. Oyler, J.V. Weber, and J. Nalasco, J. Automated Degradation Instrument. US Patent Apllication Serial No. 09/816,787 (2001).

(8) A.M. Fermier, A.R., Oyler, B.L. Armstrong, B.A. Weber, R.L. Rodriguez, J.V. Weber, and J.A. Nalasco, JALA 7(1), 68 (2002).

(9) J.W. Dolan, L.R. Snyder, N.M. Djordjevic, D.W. Hill, and T.J. Waeghe, J. Chromatogr., A 857, 1 (1999).

(10) J.W. Dolan, L.R. Snyder, N.M. Djordjevic, D.W. Hill, and T.J. Waeghe, J. Chromatogr., A 857, 21 (1999).

(11) L.R. Snyder and J.W. Dolan, J. Chromatogr., A. 892, 107 (2000).

(12) ICH Guideline Q2(R1). Validation of Analytical Procedures: Text and Methodology (2005). http://www.ich.org/LOB/media/MEDIA417.pdf.