Abstract
Cleaning validation (CV) is a Good Manufacturing Practices (GMP) requirement in the manufacturing of drug substances and drug products. High performance liquid chromatography (HPLC) with UV detection is a common choice for cleaning verification studies for most drugs at sensitivity levels (limits of quantitation or LOQ) of ~20 to 50 ng/mL. For highly potent drugs, CV methods to reach sensitivity levels of low ng/mL can be implemented using UV detection with a long-pathlength UV flow cell and/or mass spectrometric (MS) detection. The sensitivity of an ultra-high pressure liquid chromatography (UHPLC)-MS method presented herein using a single quadrupole mass spectrometer (SQMS) provides even lower LOQ of ~0.5 ng/mL. In addition, larger injection volumes can further lower LOQs. In this study, we show a 2-minute generic cleaning verification method having an LOQ range of 0.5 ng/mL to 1000 ng/mL for 5 model drug compounds. This generic cleaning verification assay method can also be adapted to many different drugs with minimum efforts in method development, qualification, and transfer. In this paper, we also demonstrate the identification of unknown compounds using a user-supplemented Mass Spectral Search Program.
Introduction
After the manufacture of a given drug, a cleaning verification (CV) analysis must be performed to demonstrate the “cleanliness” of the production train (vessels and processing equipment) to confirm that the active pharmaceutical ingredient (API) has been adequately removed to pre-established acceptance limits (maximum allowable carryover) before a new pharmaceutical can be made in the same equipment. The development of a high-sensitivity CV method is especially challenging for highly potent drugs (SafeBridge classifications 3B or 4)1 or for those possessing weak chromophores. For high potency drugs, acceptance criteria at low ng/mL detection limits are often required. The limit of quantitation (LOQ) of the traditional liquid chromatography with ultraviolet detection (LC-UV) method can be lowered by employing mass spectrometric detection.2 Additionally, unknown compounds are occasionally observed in CV studies, whose identification is often necessary. Our objective is to demonstrate a fast, sensitive, generic ultra-high pressure liquid chromatograph with ultraviolet detection mass spectrometric (UHPLC-UV-MS) method for performing CV of highly potent drugs, and an easier means to identify unknown compounds.
In earlier studies, a 60-mm long-path flow cell enabled 2x to 5x lower LOQ than a standard 10-mm flow cell in cleaning validation studies using conventional HPLC columns.3,4 We confirm similar results here for the faster UHPLC method. Additionally, a high flow rate used in CV method can reduce analysis time without sacrificing peak capacities or resolution. A single generic UHPLC-UV-MS CV method increases productivity for a laboratory pilot-scale or manufacturing facilities,5 which place a heavy emphasis on quick sample turnarounds.
In CV samples, different residual compounds can often be detected, such as an API from the current or previous batches, by-products, degradants, excipients, or even cleaning agents. The use of an LC-UVMS assay, along with a user-created mass spectral library, can make identification of unknown residual compounds much easier.
In this work, we document the development of a high-throughput UHPLC-UV-MS generic method for 5 common drugs using a short C18 column (3.0 x 30 mm) packed with sub-3-μm particles. This LC column, packed with superficially porous particles (SPP), has improved kinetic performance vs. those with fully porous particles and is operated at 1 mL/min for a 2-minute run time. Using this procedure, mixtures of the 5 model drugs at a concentration range of 0.5 ng/mL to 1000 ng/mL were measured using both UV and MS detection. Method performance criteria, such as specificity, precision, accuracy, limits of detection (LOD), LOQ, and linearity, were determined and found to be acceptable for intended use. The creation of a custom mass spectral library is described, and its searching capability is also demonstrated. This method can potentially be applied for a wide variety of drugs for cleaning verification with minimal efforts in method development, qualification, and transfer.
Experimental
A generic UHPLC-UV-MS method was developed and demonstrated for 5 model drugs: sulfamethizol, sulfamethoxazole, propranolol, imipramine, and amitriptyline. Note that these drugs are selected to illustrate method performance and system capability even though they are not considered to be highly potent. A model 1290 UHPLC system equipped with a binary pump, an autosampler, a column oven, a diode array detector (with 10-mm and 60-mm flow cells), and a 6150 Series Single Quadrupole Mass Spectrometer with a Jet Stream source was used (Agilent Technologies, Santa Clara, CA, U.S.A.). The dwell volume of the system is ~100 μL and the dispersion of the system is ~16 μL2 (variance). Linearity was shown at various concentrations using 10-mm standard and 60-mm long-pathlength UV flow cells and MS-SIM or SCAN mode. Studies using higher-volume injections were also performed to determine improvements in signal-to-noise (S/N) ratio for sulfamethoxazole. Table 1 shows the details of the instrumental parameters of the generic method developed.
Preparation of Standards
The stock solutions of sulfamethizole, sulfamethoxazole, propranolol, imipramine, and amitriptyline were prepared at 1000 μg/mL in 100% methanol. The stock solution was diluted to 10 μg/mL working solution using 50:50 (methanol: mobile phase A [MPA]). The linearity solutions at various concentrations were prepared by serial dilution. The concentrations in ng/mL were: 0.2, 0.3, 0.5, 1, 5, 10, 20, 30, 50, 100, 500, 1000, 2000, 5000, and 10 000.
Results and Discussions
UHPLC-UV Detection
A UV detector with a long-path UV flow cell and standard flow cell were used alternatively in the generic UHPLC/UV method (Table1). The resulting chromatograms obtained in 6 replicate injections of a 5-compound mixture using a 10-mm flow cell are overlaid (Figure 1). Linearity experiments were performed up to concentration 10 000 ng/mL using both these flow cells. In the case of sulfamethizole, LOQs of 5 ng/mL and 20 ng/mL were found using a long-path flow cell and a standard 10-mm flow cell, respectively. As expected, a 60- mm long-path flow cell produced 3x to 4x lower LODs than that of a standard flow cell. These results were comparable to an earlier study.2 The concentration ranges, in terms of LOD/LOQ, for both flow cell types are shown in Table 2. LOQs of 20 to 50 ng/mL are achievable using UV detection with a standard flow cell. The linearity of the compounds in the experimental concentration levels shows coefficient of determination r2 values >0.99 for the linearity ranges studied (Figure 3).
Table 1. The experimental parameters and instrumentation used in UHPLC-UV-MS generic method
Figure 1. UHPLC-UV results.
Figure 3. UHPLC-UV linearity runs using 60-mm long-path flow cell.Table 2. UHPLC-UV method validation results using 10-mm standard and 60-mm long-path flow cells
Single Quadrupole Mass Spectrometer detection
A 2-minute generic gradient method as described in Table 1 was applied using the Agilent 1290 attached to a Single Quadrupole Mass Spectrometer. A chromatogram generated from the analysis of a 30 ng/mL mixture is shown in Figure 4. The linearity experiments performed at 0.5 ng/mL to 100 ng/mL solution demonstrated the excellent linearity achievable in this (concentration) range. The LOQ of 0.5 ng/mL obtained using this UHPLC-MS method is comparable or better to the LOQ for a longer LC-MS method as described by Liu, et al. for highly potent drugs.6 The relative standard deviations (RSDs) of peak areas and retention times at the LOQ were found to be <5% and <0.1%, respectively (Table 3). The linearity for all compounds in the experimental concentration levels gave coefficients of determination or r2 values of >0.98 (Figure 5). The accuracy values, back calculated from the linearity equation, were within 15% using a weighting of 1/x2 (see Table 4). The LOQ obtained from Single Ion Monitoring (SIM) method was also compared with MS scan data where both SIM and full scan were acquired within the same run. The LOQs from extracted ion chromatogram (EIC) signals from full scan data were 5 ng/mL for all compounds (data not shown), suggesting that the SIM mode is ~10x more sensitive than that from the full scan mode.
Figure 4. UHPLC/MS (SIM) results of 30 ng/mL mix solution.Table 3. The UHPLC-MS (SIM) method validation results
Figure 5. UHPLC-MS (SIM) method linearity of the drug compounds.Table 4. UHPLC-MS (SIM) method accuracy
Large Volume Injections for LC-MS Runs
The possibility to achieve even lower LOQs for a sensitive CV method was studied by using larger volume injections. This is useful when additional sensitivity is required. The use of larger injection volumes increased both peak area and S/N ratios, but may also increase peak broadening. This may not be a major concern in CV determination since peaks associated with multiple APIs are rarely observed in actual CV sample solutions (except for combinational drug products). The results for higher volume injections for 3 different concentrations of sulfamethoxazole using the generic method is shown in Table 5. The results showed that, with the increase in injection volumes on 30-mm column, from 5 μL to 12 μL, the peak areas, peak widths, and S/N ratios increased by 57%, 31%, and 41%, respectively. The sulfamethoxazole study showed that another 2-fold increase in S/N ratio could be achieved by using higher-volume injections of 10 μL. Higher volume injections beyond 12 μL led to peak splitting due to the organic strength of the sample solution used in this study.
Table 5. Peak area, peak width and S/N ratio for higher volume injections of sulfamethoxazole solutions
User Supplemented MS Unit z Mass Library
In cleaning verification using a UHPLC-MS method, it is also possible to detect unknown compounds in full scan mode. An unknown peak found may trigger further investigation. A software-assisted identification of the compound from its mass spectrum can also facilitate quick identification. A user-created unit-mass MS spectral library was developed to augment the NIST Mass Spectral Search Program. The spectra of 5 standard compounds were incorporated in the library. Sulfamethoxazole compound was injected, and a mass chromatogram acquired in scan mode. The scan mode spectrum was searched within the user created library. The results (Figure 6) show the correct identification of sulfamethoxazole. Therefore, unknown peaks can be readily identified from the NIST Mass Spectral Search Program supplemented with user-generated spectral information from the reference standards.
Figure 6. NIST mass spectral search program.Summary and Conclusion
Cleaning verification for high potency drugs demands highly sensitive, high throughput methods. A 2-minute generic method for CV was developed and validated for 5 common drugs to illustrate method performance for this application. LOQs of 20 to 50 ng/ mL and 0.5 ng/mL were achievable using UV and MS (SIM) detection, respectively. Further enhancements can be made using a 60-mm-pathlength flow cell (2-4 fold) and larger volume injections (2-fold using 10-μL injection). A UHPLC-MS system, operating under SIM or SCAN mode, can be a versatile tool for cleaning verification studies for quantifying known compounds in the solvent rinsates or swabbing extracts from manufacturing equipment. In addition, the identification of unknown compounds can be readily performed by searching within a useraugmented MS spectral library.
Acknowledgments
The authors are grateful to many colleagues for their useful inputs and suggestions: Dr. Jackson Pellet and Chris Goretski of Genentech, Drs. Davy Guillarme and Szabolcs Fekete of University of Geneva, and Dr. Tom Waeghe from MAC-MOD Analytical.
References
- Farris JP, Ader AW, Ku RH. History, implementation and evolution of the pharmaceutical hazard categorization and control system. Chemistry Today. 2006; 24(2):5- 10.
- Dong MW, Zhao EX, Yazzie DT, Gu CC, Pellet JD. A generic HPLC/UV platform method for cleaning verification. Amer. Pharm. Rev. 2012; 15(6):10-17.
- Naegele E, Kornetzky K. Highly Sensitive UV Analysis with the Agilent 1290 Infinity LC System for Fast and Reliable Cleaning Validation – Part 1: Measurement of calibration curves, determination of LOD and LOQ and method validation using a DAD equipped with standard or high sensitivity flow cell. Agilent Application Note. 2010; 5990-6929EN.
- Naegele E, Kornetzky K. Highly Sensitive UV Analysis with the Agilent 1290 Infinity LC for Fast and Reliable Cleaning Validation – Part 2: Monitoring a cleaning validation procedure using a DAD equipped with standard or high sensitivity flow cell. Agilent Application Note. 2010; 5990-6930EN.
- Dong MW. A three-pronged template approach for rapid HPLC method development. LCGC North America. 2013; 31(8):612-621.
- Liu L, Pack, BW. Cleaning verification assays for highly potent compounds by high performance liquid chromatography mass spectrometry: Strategy, validation, and long-term performance. J. Pharm Biomed. Anal. 2007; 43:1206-1212.
Author Biographies
Syed Salman Lateef, PhD, is lead application scientist in Life Science Center India at Agilent Technologies India Pvt Ltd. He is responsible for developing applications and workflow solutions across analytical platforms. He holds a PhD in Analytical Chemistry from University of Illinois at Chicago.
Vinayak AK is an application scientist in Life Science Center India at Agilent Technologies India Pvt Ltd. He is responsible for developing applications and workflow solutions across analytical platforms. He holds a Master’s degree in Biotechnology from Amrita University, Kerala, India.
Michael W. Dong, PhD, is senior scientist in Small Molecule Analytical Chemistry and Quality Control (SMACQC), Small Molecule Drug Discovery, at Genentech in South San Francisco, where he is responsible for new technologies, automation and analytical support for multiple late-stage research projects. He holds a PhD in Analytical Chemistry from City University of New York and a certificate in biotechnology from U. California at Santa Cruz. He has 100+ publications and 3 books including a bestseller in chromatography—Modern HPLC for Practicing Scientists, Wiley, 2006. He teaches short courses in HPLC/UHPLC, drug development process, and drug quality fundamentals at national meetings. He is a member of the editorial advisory boards of American Pharmaceutical Review and LCGC North America.
Christine C. Gu, PhD, is a scientist in SMACQC in Genentech responsible for identification and structure elucidation of impurities and degradants in multiple development projects. She is a subject matter expert in mass spectrometry and holds a PhD degree in Toxicology at U. California at Riverside.