RPID ND SENSITIVE PLC/V/MS METD FR SIMVSTTIN ND LV S TT IN IN S P P RT F C L E NING V L I DT I N S T DIES Erin Chambers, Diane Diehl, and Jeff Mazzeo Waters Corporation, Milford, M, S BCKGRND tilizing CQITY PLC Technology with V and MS detection enhances the selectivity, sensitivity, and through-put in cleaning validation studies. This results in improved methods, more confidence in results, and significant cost savings. To highlight this utility, we developed an analytical method in support of a cleaning validation study for simvastatin and lovastatin, two commonly prescribed cholesterol-lowering drugs in the statin class. is newly off patent and has been the focus of several generic companies. For this reason, an analytical method is needed to support cleaning validation studies in a manufacturing environment. NLYTICL CLLENGES Pharmaceutical manufacturing equipment is cleaned after production to avoid cross-contamination in subsequent batches of a different product. Effectiveness of the cleaning process needs to be confirmed by analytical measurements. Therefore, a cleaning validation method must be developed which provides evidence that cleaning processes applied to the equipment are sufficient to remove residues of bulk drug to predetermined safety levels. These methods ensure that subsequent batches of other products are not contaminated by previously manufactured products, or by the cleaning process itself. Cleaning validation methods need to achieve limits of detection which are low enough to detect small quantities of residual drug and they must demonstrate adequate resolution of drug product from interferences from solvents and the swabs or wipes used to sample the equipment surfaces. In addition, the method must be shown to be linear over the range required by the assay. 3 C 3 C C 3 C 3 3 C MW 418.27 Figure 1. Structures of simvastatin and lovastatin. and lovastatin, whose structures are shown in Figure 1 are 3-hydroxy-3-methylglutaryl-coenzyme reductase inhibitors, which significantly impact the synthesis of cholesterol precursors. Drugs of this class represent the most efficient treatment of hypercholesterolemia. is a natural product, while simvastatin is semi-synthetic, being manufactured from lovastatin. s a result, lovastatin and simvastatin need to be detected in the same analytical method. EXPERIMENTL Click on Part Numbers for more information screening approach was utilized to choose the best starting chromatographic conditions. Four different column chemistries were screened at high and low p using two organic modifiers. The resulting chromatograms were evaluated for peak shape, retention, and selectivity. From these data, the column chemistry, p, and mobile phase modifier were chosen. The screening gradient was later optimized to decrease run time if possible, and to improve resolution of simvastatin and lovastatin from interferences extracted from swabs or wipes. 3 C 3 C C 3 MW 44.25 C 3
LC Conditions for Screening LC System: CQITY PLC with Column Manager, CQITY PLC PD and SQ Columns: CQITY PLC BE C 18 2.1 x 5 mm, 1.7 µm Column Temp: Flow Rate: Part Number: 186235 CQITY PLC BE Shield RP18 2.1 x 5 mm, 1.7 µm Part Number:1862853 CQITY PLC BE Phenyl 2.1 x 5 mm, 1.7 µm Part Number:1862884 CQITY PLC SS T3 2.1 x 5 mm, 1.8 µm Part Number:1863538 3 C 6 µl/min Mobile Phase 1:.1 C in 2 (~p 2.7) Mobile Phase 2:.1 N 4 in 2 (~p 11) Mobile Phase B1: Mobile Phase B2: Gradient: Wavelength: MS Conditions MS System: Ionization Mode: Capillary Voltage: Cone Voltage: Desolvation Temp: Desolvation Gas: Source Temp: Me CN 5-95 B in 2 min, hold for.5 min; return to initial conditions 248 nm Waters SQ Detector ESI Positive 3 V 43 V 35 C 55 L/r 12 C SIR Channels: m/z 419.25, Extraction Procedure m/z 45.25 RESLTS ND DISCSSIN Chromatographic Screening The chromatograms resulting from p screening with CN as the organic modifier are shown in Figure 2. n initial PD scan was used from 21 to 3 nm in order to determine the optimal wavelength of 248 nm for detection of simvastatin and lovastatin. ll columns were run at low p (). The C 18, Shield RP18, and Phenyl columns were also run at high p (B). (Note: the SS T3 column is a silica particle and cannot be run at high p.) Chromatograms were evaluated with respect to peak shape, resolution of impurities or degradants from the compounds of interest, and selectivity. Retention of these neutral compounds is relatively unaffected by p. owever, at low p we begin to resolve additional drug compound-related peaks from simvastatin and lovastatin. The slightly broader peak shape observed on some columns at high p is due to coelution of these smaller peaks with the compounds of interest. We chose low p mobile phase due to better peak shape and improved resolution of impurities. We next screened all column chemistries using two organic modifiers. The chromatograms resulting from organic modifier screening are shown in Figure 3. ll columns were run at low p with Me () and CN (B). It is clear that Me is not a selective enough elution solvent to resolve these hydrophobic compounds. and lovastatin elute as a single broad peak when Me is used as the organic modifier. In contrast, the two compounds chromatograph as well-resolved, narrow peaks using CN. The final step involved choosing the optimal column chemistry. Examination of the low p, CN chromatograms indicates that only the SS T3 column is able to adequately resolve all of the impurity peaks from simvastatin and lovastatin. We optimized the chromtographic method using the SS T3 column at low p, with CN as the organic modifier. Cotton swabs (TexWipe ) were sonicated for 45 min in 1 ml of 75:25 CN: 2, or 1 CN. The extract was removed and analyzed for the presence of interferences.
. CN, low p B. CN, high p 1.3.36.38 1.94.55 1.8.5 1.83 2.5 2.4 2.15 C 18 SS T3 RP18 Phenyl Figure 2. PD chromatograms at 248 nm resulting from p screening with CN as the organic modifier. 1.e-2-1. e-2 5.e-3-5. e-3 1.e-2-1. e-2.35.69 2.35 2.5.73 2.43 2.5.57 2.18 2.5 4.e-2 4.e-2. B. Me, low p C 18 SS T3 RP18.33 1.26.33.33 1 1.92 Not applicable 1.8 1.12 1.81.3.36.38 1.4 1.76 1.82 1.94.55 CN, low p 1.8 2.2 2.4 2.5 2.15 Extraction Recovery Recoveries for simvastatin and lovastatin in both the 1 CN and 75:25 CN: 2 extracts were > 95. This was determined by spiking the swabs prior to extraction with known quantities of active drug. Peak area in the extract was then compared to peak area in standards prepared in the identical organic solvent. Limit of Detection (LD) and Linear Dynamic Range The LDs for simvastatin and lovastatin were determined under optimized chromatographic conditions with both V and MS detection. The LD obtained from data acquired using the PD operating at 248 nm was shown to be 5 ng/ml for both analytes (Figure 4). The LD for mass spectrometry operating in SIR mode was 1 ng/ml for each analyte (Figure 4B). The dynamic range of the assay was 5 ng/ml to 5 µg/ml by PD at 248 nm and 1 ng/ml to 1 µg/ml using mass spectrometry detection. Standard curves were linear with 1/x 2 weighting over this range with R 2 values from.983 to.998. ll data points met the criteria for accuracy with all non-ld points having CVs of less than ± 15, and all LD points having CV values of less than ± 2. Sample standard curves for simvastatin and lovastatin using both PD and mass spectrometry detection are shown in Figure 5..57 2.5 Figure 3. PD chromatograms at 248 nm resulting from organic modifier screening at low p. Chromatographic ptimization 2.27 Phenyl.5 1.83. 6.5e-3 6.e-3 5.5e-3 5.e-3 4.5e-3 4.e-3 3.5e-3 3.e-3 2.5e-3 2.e-3 1.5e-3 1.e-3.2 5 ng/ml.39 5.e-4 5.1.2.3.4.5.6.7.8.9 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7.94 1.29 5 ng/ml 1.47 and lovastatin are very hydrophobic compounds as indicated by their late elution time under reversed-phase LC conditions. The generic screening gradient was modified to range from 5 to 95 CN to better reflect the nature of the analytes, and to shorten the chromatographic run time. se of this shallower gradient also allows us to improve resolution from any additional impurities or interferences discovered during analysis of swab samples. In addition, formic acid was added to the CN (mobile phase B) at a concentration of 75 to reduce V baseline drift. B. 1 1.31.46.47.56.57.68.73.77 1. 1.5.91.94 1.12 1.18.1.2.3.4.5.6.7.8.9 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1 5.14.24.17.19.29.34.4.45.53.63.68.7.73.1.2.3.4.5.6.7.8.9 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 Figure 4. Limits of detection were determined for V () and mass spectrometry (B.).79 1 ng/ml.84.85 1 ng/ml.85.99 1.5.88 1.15.93 1.2 1.1 1.25 1.25 1.19 1.25 1.33 1.32 1.32 1.37 1.37 1.41 1.39 1.48 1.45 1.5 1.51 1.5 1.54 1.56 1.6 1.63 1.61 1.62 1.67 1.78 1.75 1.73 1.73 1.79
Selectivity The method was shown to be selective for the analytes of interest. Swabs were extracted as above, the extract removed, and then analyzed to confirm lack of interferences from swabs and extraction solvents. Figures 6 and 7 demonstrate that both analytes are resolved from interferences when either 1 CN or 75:25 CN: 2 are used as swab extraction solvents. Figure 6 demonstrates minimal interferences using V detection, while figure 7 demonstrates a much higher degree of selectivity for the analyte using mass spectrometry.. : PD 248 nm Standard curve is linear from 5 ng/ml to 5 µg/ml 1/x 2 weighting, r 2 =.997, CV from -.4 to 8 6 B. : SQ in SIR mode Standard curve is linear, 1 ng/ml to 1 µg/ml 1/x 2 weighting, r 2 =.99, CV from -.2 to -15.7 75 5 25 ng/ml 5 1 15 2 25 3 35 4 45 5 4 2 Conc 5 1 15 2 25 3 35 4 45 5 55 6 65 7 75 8 85 9 951 : PD 248 nm Standard curve is linear from 5 ng/ml to 5 µg/ml 1/x 2 weighting, r 2 =.998, CV from.4 to 6.7 : SQ in SIR mode Standard curve is linear from 1 ng/ml to 1 µg/ml 1/x 2 weighting, r 2 =.983, CV -.6 to 18 1 5 4 2 ng/ml 5 1 15 2 25 3 35 4 45 5 Conc 5 1 15 2 25 3 35 4 45 5 55 6 65 7 75 8 85 9 95 1 Figure 5. Standard curves for simvastatin (top) and lovastatin (bottom) using V detection at 248 nm () or mass spectrometry in SI R mode (B). 7.e-2 Swab extracted with CN 1: SIR of 2 Channels ES+ TIC 8.4e6 6.e-2 5.e-2 5ng/mL (1) lovastin and (2) simvastatin Swab extracted with 75:25 CN: 2 4.e-2 Swab extracted with 75:25 CN: 2 5 ng/ml (1) lovastatin and (2) simvastatin 1.63 3.e-2 Swab extracted with CN Minimial interference No interference 1.e-2 1.29 1.47.2 1.68-1.e-2.2.4.6.8 1. 1.2 1.4 1.6 1.8 2..1.2.3.4.5.6.7.8.9 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2. Figure 6. V swab interferences. Figure 7. Mass spectrometry swab interferences.
Conclusion We have successfully employed a chromatographic screening strategy to facilitate fast development of a rapid, selective and sensitive LC/MS/V analytical method to support cleaning validation studies. The method achieves LDs with V detection that meet the requirements for most cleaning validation protocols. Single quadrupole mass spectrometry yields lower limits of detection, suitable for the testing of manufacturing equipment used to prepare even very potent drugs. The increased resolution afforded by PLC technology allows us to separate the two analytes of interest from swab and solvent interferences in under 3 minutes. dditional selectivity is realized with mass spectrometric analysis. This rapid analysis time enables manufacturing plants to move to the next production batch faster, thus increasing their productivity and efficiency. In addition, increased simplicity and cost savings are achieved by having a single method for multiple analytes. References dditional information relating to the use of CQITY PLC systems for cleaning validation studies can be found in our document entitled CQITY PLC System: Better utilizing assets in a small laboratory environment with PLC, literature code 722328EN. Waters, The Science of What s Possible, CQITY PLC and PLC are trademarks of Waters Corporation. ll other trademarks are the property of their respective owners. 27 Waters Corporation. ctober 27 722337EN VW-PDF Waters Corporation 34 Maple Street Milford, M 1757.S.. T: 1 58 478 2 F: 1 58 872 199 www.waters.com