Cleaning Verification: Method Development and Validation Using Ion Mobility Spectrometry2

By: Elizabeth Galella, Scott Jennings, Madhavi Srikoti, Elizabeth Bonasso
IMS distinguishes ions of a given compound on the basis of their velocities through a drift tube under the influence of a weak electric field. Ion velocity (v) is proportional to the applied electric field (E)
in which K is the ion mobility in units of cm² /Vs. Drift time t is proportional to 1/K , ion mass m, and collisional cross section C. Reduced mobility K_o for an analyte is calculated by normalizing its drift time to that of the internal calibrant as in the following equation: in which (K_o^C ) is the reduced mobility of the calibrant and t C is the drift time of the calibrant.

Figure 3: Illustration of ions of varying sizes migrating toward the detector through a constant flow of air.

These velocities, or ion mobilities, are determined and selective for a given compound. The high sensitivity of the instrument (nanogram to picogram range) provides for an excellent technique to quickly verify equipment cleanliness. Figure 3 illustrates the separation process described in the theory above. This technique allows for positive or negative mode of detection. Each mode has a unique internal calibrant used for calculation of the K_o. In the positive mode the internal calibrant is nicotinamide. In the negative mode the internal calibrant is methyl salicylate. 
Advantages of IMS

Figure 4: Efficiency gained by ion mobility spectrometry (IMS) versus high-performance liquid chromatography (HPLC).

Ion mobility has unique advantages over the conventional HPLC technique for cleaning verification. Although the principles of IMS technology have been well established, its uses in the pharmaceutical industry have, until recently, been limited. This has been changed by the development of easy-to-use commercial instruments of moderate cost, with small footprint and high sensitivity (4). As shown in Figure 4, when compared with the typical HPLC method, the IMS technique can save a significant amount of time in the analyzing cleaning verification samples. Furthermore, a quick response time in getting results can reduce down time of key manufacturing and packaging equipment, ultimately leading to significant cost savings and increased productivity for the company (1,3). Method development
As with most analytical methods, the IMS instrument parameters must be examined and optimized for each compound as part of method development. These parameters include ionization mode, desorber temperature, injection volume, post-injection delay, drift flow velocity, and analysis time. The development process begins by examining the selectivity of the target compound in both positive and negative modes. In the authors' experience, most pharmaceutical compounds respond better in the positive mode because of the presence of basic functional groups within the molecule. The desorber temperature for compounds analyzed in the positive mode is typically set at ~290 °C. This temperature should be hot enough to effectively desorb the sample off the substrate but not so hot as to thermally degrade the compound. Typical sample volumes are 1 µL, injected using a 10-µL syringe. The post-injection delay is dependent on this volume and the type of solvent being used. A large injection volume of a solvent of low volatility requires a longer post-injection delay because it will take more time for the solvent to evaporate. Finally, the analysis time and drift flow velocity settings are dependent on the IMS response. A sufficient analysis time and drift flow are necessary to ensure depletion of all sample ions in the drift tube. In a case where the analysis time is too short, or drift flow velocity is too low, sample carryover may be an issue.

Figure 5: Second-order polynomial calibration curve.

The generally accepted practice is that an analytical method must exhibit sufficient sensitivity to measure the concentration at the ARL of the active agent or degradants being investigated in the swab and rinse samples. Typically, a limit test is used to establish a pass–fail criterion for swab samples and determine whether the manufacturing equipment is clean. Determining the pass–fail limit is accomplished by evaluating the instrument response curve and the system's precision. It allows analysts to capture the ARL within a linear calibration range, therefore providing an accurate means of determining the cleanliness of equipment parts. In the response curve, the "target level" represents the highest point within linear range (see Figure 5). The "action level" represents the resulting pass–fail level after adjusting the target level for instrument response variability. Any sample that responds above this action level registers as a failure. Any sample that responds below this action level registers as passing. Consequently, there is an area of uncertainty between the target and action level within which false-positive results may occur. However, these are generally infrequent occurrences. 
Method validation
Typical IMS method validation parameters to be considered include selectivity, linearity, reproducibility, recovery and solution stability. Selectivity of the analyte is determined based on evaluation of the molecular structure dictating the mode of detection. Selectivity from the sample matrix is based on minimizing the interference from product excipients and cleaning detergents, and should be examined before performing the recovery experiment. In addition, the linearity is based on a second-order polynomial curve obtained from the response versus the amount introduced, as described previously. Reproducibility must be used to determine the action level. Therefore, these parameters must be considered during method development and verified during validation.

Table I: Recovery of active from factory-finished, stainless-steel plates.

Recovery. After determination of the action level, sample recovery is determined from a 10-in.² stainless-steel coupon. A stock sample solution is prepared at 100% of the API concentration at the target level. The solution is then spiked onto three separate stainless steel coupons. The surfaces are allowed to dry before swabbing with a clean Texwipe alpha polyester swab. According to industry guidelines, the swabbing technique includes horizontal strokes to swab the entire designated area with one side of the swab head and then vertical strokes with the other side of the swab head. Each swab is then placed back into the vial containing sample diluent, extracted, and later analyzed for active content. Table I shows the recovery values from stainless-steel surfaces and corresponding percent relative standard deviation (%RSD) at the action level for a novel pharmaceutical compound under development. Table I shows the recovery results, which are typical of those generated by swabbing three different stainless-steel coupons. The observed variation is primarily a result of the inherent variability in the surface finish of the coupon. These values are independent of the analysis technique and may be optimized by adjusting the swabbing procedure and/or materials. However, application of a predetermined correction factor to swab assay results can be used to compensate for recovery values as low as 50% during method validation. Solution stability. Sample and standard solution stability studies are typically performed to cover 24 h at room temperature. Additional validation studies, such as robustness and intermediate precision, are not included in the authors' validation plan for a limit test. All of the parameters described previously must meet pre-established criteria as defined in the validation protocol. The criteria and validation protocol must conform to International Conference on Harmonization guidelines as well as internal standard operating procedures.
Conclusion
The ease of use and the small footprint of ion mobility spectrometry instrument allows for the system to be implemented in various work environments, such as quality control, and to report results with great sensitivity (nanogram to picogram range). Moreover, data reduction software package upgrade ensures 21 CFR Part 11 compliance. The use of this software, in addition to a limit test, has simplified the process of data manipulation, resulting in high-confidence passes for each clean sample analyzed.
Elizabeth Galella* is a research scientist, Scott Jennings is a senior research scientist, Madhavi Srikoti is an associate research scientist, Elizabeth Bonasso is a research scientist, all at the analytical research and development unit of the Pharmaceutical Research Institute, Bristol-Myers Squibb, One Squibb Drive, New Brunswick, NJ 08903, elizabeth.galella@bms.com
References
1. G. Walia et al., "Using Ion Mobility Spectrometry for Cleaning Verification in Pharmaceutical Manufacturing," Pharm. Technol. 27 (4), 72–78 (2002).
2. G. Walia et al., "Implementing Ion Mobility Spectrometry as a Cleaning Verification Method," Pharm. Technol.27 (3), 22–25 (2002).
3. D. Brand, X. Li, and T. Wortley, "Ion Trap Mobility Spectrometry: Reducing Downtime in Cleaning Validation and Verification," GE Sensing, http://www.gesensing.com/toolsupport/whitepaper.htm accessed June 16, 2009.
4. D. Brand et al., "Direct Swabbing and Surface Recovery with Ion Trap Mobility Spectrometry," GE Sensing, http://www.gesensing.com/products/resources/datasheets/GEsensing_whitepaperfinal.pdf accessed June 2, 2009.
5. R. DeBono, "Ion Mobility Spectrometry: A Fast, Sensitive, and Robust HPLC Alternative," Appl. in Chrom. March, 20–23 (2002), http://trace.smithsdetection.com/Documents/LifeSciences/LE203sBarringer.pdf.
6. G.A. Eiceman, "Advances in Ion Mobility Spectrometry," Crit. Rev. Anal. Chem.22 (1, 2), 471–490 (1990).