Manufacturing

Ion Trap Mobility Spectrometry (ITMS) represents afast and specific technique for the analysis of samplesfor cleaning validation and verification in thepharmaceutical industry. Furthermore, at-lineanalysis has the potential for greatly improving theefficiency of analysing cleaning results andimproving equipment turn-around. This articlereviews a study on the use of this technology for‘direct swabbing,’ or directly sampling and analysingthe equipment of interest. Recovery results fromstainless steel surfaces for two different compounds,cefuroxime sodium and pseudoephedrine HCl, arepresented.

ITMS APPLICATIONS

Ion Mobility Spectrometry (IMS) and Ion Trap MobilitySpectrometry (ITMS) are built on the principle ofmeasuring the drift velocity of ions as they are propelledthrough a ‘drift gas’ at ambient pressure via the force of an electric field (1). The technology has been in

use for over 30 years, primarily applied in detecting trace amounts of narcotics and explosives (2), and is found at most airports as part of their securityscreening procedures.

More recently, ITMS has found application in thepharmaceutical industry, mainly focusing on cleaningvalidation and verification. While the technology hasbeen reviewed (3) and specific applications described(4,5), the data published to date has focused on resultsgenerated from extracted solutions, rather than directsampling of a surface of interest.

The act of taking a sample directly from the surface ofequipment has been termed ‘direct swabbing’ in thatthe sample is analysed directly instead of via anintermediate extraction step. Similar to the use ofITMS in security applications, the advantage of directswabbing is that it allows the user to generate results without the need to send samples back to ananalytical laboratory. Additionally, the portability of

ITMS: Applications in At-Line Cleaning Validation and VerificationIon Trap Mobility Spectrometry (ITMS) offers a fast, specific analyticaltechnology for at-line measurements, with the potential to deliversubstantial improvements in cleaning analysis and monitoring efficiency

By Derek Brand, Mei Guo, Ralf Wottrich and Tim Wortley at GE Sensing

Derek Brand received his bachelors in Biology at Hamilton College (Clinton, NY) and received his MBA from Babson College (Wellesley, MA). He has held roles in research and engineering in the development of spectroscopicinstrumentation and imaging systems in the medical device industry. He has also held roles in business development,marketing, and analysis over seven years in the medical device, pharmaceutical, and biotech industries. In his role at GE Sensing as Global Product Manager, Pharmaceutical, he is involved with the launch and commercialisation of the Kaye Validator ITMS system for cleaning validation and verification in the pharmaceutical industry.

Mei Guo is an experienced Engineering Technician. In this role, she has assisted the engineering department in the building and testing of new products. She has supported both the flow and the moisture engineers. Recently, she has also taken on the role of conducting tests and experiments for the ITMS-NPI project as TECHTechnician – D at GE Sensing.

Ralf Wottrich studied at the University of Karlsruhe and received a diploma in Biology. He accumulated experience in several fields of biology and biochemistry and has achieved a Ph.D. in environmental toxicology. He has held roles in research and product development in the pharmaceutical and biotech industries before joining GE Sensing.Working in Application Engineering/Commercial Engineering, he is now involved in the commercialisation and applicationdevelopment of the Kaye Validator ITMS system for cleaning validation and verification in the pharmaceutical industry.

Tim Wortley has an Electrical & Electronics Engineering degree from Heriot-Watt University in Edinburgh, Scotland. He has been with GE Sensing (formally Kaye Instruments) for 12 years developing products and solutions for thepharmaceutical industry. His background is in the design of precision instrumentation, data acquisition and motorcontrol products. He is currently the Global Product Manager for Kaye & Pharmaceutical products at GE Sensing.

90 Innovations in Pharmaceutical Technology

IPT 25 2008 24/4/08 12:36 Page 90

commercially available ITMS instrumentation allowsthe testing to be completed at-line.

The FDA’s Guide to Inspections Validation of CleaningProcesses discusses the sampling methods applied to thecleaning process – rinse and swab (direct) sampling – aswell as the analytical methods necessary to measure thesamples taken. Specifically, these sampling andanalytical methods need to be challenged and a‘recovery’ that describes the effectiveness of thesampling/analytical combination needs to “show thatcontaminants can be removed from the equipmentsurface and at what level, that is 50% recovery, 90%recovery, etc” (6).

The guide also discusses cleaning limits, and whilepurposefully staying away from tangential description, it puts forth that the limits for a particular compoundand process must be “practical, achievable, andverifiable”, and that the analytical method used tomeasure them needs to have the requisite level ofsensitivity for these measurements.

The determination of carryover limits for a particularcompound has been described using both themaximum allowable dose carried over to the nextproduct batch (7), as well as the use of acute data suchas LD50 values (the amount/dose of a substance thatproduces death in half of the animals tested) (8,9). It isimportant to note that the limits referred to in thepresent study are the maximum allowable amount ofresidue on the equipment surface, as opposed to thelimit in the subsequent product or the limit in ananalytical sample.

The wide range of potential carryover limits inpharmaceutical cleaning challenges the analyticalmethods used to measure the limits, regardless ofwhether the method used is a direct swabbing methodor one that relies on extraction and dilution. Theanalytical method needs to have the appropriatedynamic range to measure the substance at its cleaninglimit with an appropriate linear range that ensures theability to effectively differentiate a passing result from afailure (10).

EXPERIMENTATION

In this series of experiments, we demonstrated the abilityto recover the residues of two compounds from stainless steel surfaces and analyse the results directlyusing ITMS. One of the substances selected wascefuroxime sodium, classified as a β-lactam antibiotic

with typically very low carryover limits due to potentiallysevere allergic reactions (11) and anaphylactic shock (12)in some cases of ingestion. The second compound waspseudoephedrine HCl, a common decongestant withcleaning limits significantly higher than cefuroximesodium (12). The chemical structures of these moleculesare shown in Figure 1.

The goals of this experiment were to demonstrate thatITMS can be used in a direct swabbing capacity togenerate acceptable recovery levels across a wide range of carryover limits. The experiment used the KayeValidator® ITMS for sample measurement, and sampleswere prepared using USP-grade cefuroxime sodium and pseudoephedrine, with dilutions being prepared in methanol.

The experimentation included analysis of compounds todetermine their time of flight (TOF), generation ofcalibration curves and determination of the linear ranges,and finally measurements of samples taken directly fromthe steel coupons in order to determine recoverypercentage. As ITMS uses the TOF as a metric foridentifying a molecule, the first stage of the study was to determine the TOF for both cefuroxime and pseudoephedrine.

Determining Time of FlightThe instrument used for the determination of time offlight has the ability to collect data for both positive andnegative ions within a single measurement. This bringsseveral potential advantages – among them the ability to

91Innovations in Pharmaceutical Technology

Figure 1: Molecularstructures of cefuroximesodium (top) andpseudoephedrine HCl(bottom). The molecularweights of these compounds are 446.4 and 201.7, respectively

+

OC COHN

OH

CH CH CH3 CI-

NH2 CH3

N N

C OCH2OCNH2

ONa

S

OCH3

O

O

IPT 25 2008 24/4/08 12:37 Page 91

detect multiple ion speciesregardless of the charge on the‘preferred’ ion state in a singlescan (a ‘single mode’ instrumentwould require two separatemeasurements).

Additionally, as there is no need to switch modes in theinstrumentation, the ValidatorITMS eliminates re-equilibrationtime associated with switchingmodes, shortening the amount of time necessary to develop a method for a particularsubstance.

Using samples of the pure API (active pharmaceuticalingredient) dissolved in methanol,aliquots were spiked directly onto

the swabs used in the instrument, the swabs were thenanalysed and the resulting peaks recorded. In addition,measurements were taken on: (a) swabs without anysubstance present; (b) swabs that were spiked with 100µlof methanol and allowed to dry; and (c) with theinstrument having no swab inserted, in order to accountfor background peaks. Finally, a very small sample of thedry API powder was swiped directly onto the swab. Thiswould highlight any differences seen due to interactionswith the solvent.

The time of flight for cefuroxime sodium wasdetermined to be a positive ion complex at 7.790ms,with the time of flight for pseudoephedrine

determined to be a positive ion complex at 5.885ms. Representative plasmagrams (similar to achromatogram in HPLC) with locations of therepresentative API peaks, as well as the locations of thedrift gas peak and common fragments in thecefuroxime data, are shown in Figure 2.

For the remainder of the analysis, cefuroxime wasidentified as a positive ion with a time of flight of 7.790+/-0.04ms; pseudoephedrine was identified as a positiveion with a time of flight of 5.885ms +/- 0.04ms. Noinstances of a peak potentially associated with the maincefuroxime ion or pseudoephedrine ion occurred outsidethese windows of detection.

Determining Quantitative ResponseAfter determining the time of flight for each API, thequantitative instrument response for each compoundand the linear range were determined. The carryoverlimits for cefuroxime and pseudoephedrine used in thisexperimentation were 1µg and 20µg per 25cm2,respectively. Figure 3 shows the instrument responsecurve for both cefuroxime and pseudoephedrine. Theparameters of the instrumentation were adjusted in orderto establish the appropriate linear range for eachcompound (described previously).

Cefuroxime DataThe cefuroxime measurements encompassed sampleamounts between 250ng and 3µg. As the instrumentwas able to give a repeatable response at 250ng thatcould be used for quantification, and cefuroxime wasdetectable at sample amounts lower than 250ng, forthe purposes of this experiment 250ng was consideredthe limit of quantification (LOQ) and it was assumed

92 Innovations in Pharmaceutical Technology

Figure 3: Quantitative response of cefuroxime sodium (top) and pseudoephedrine HCl (bottom) in the ITMS instrument. Data shown is average value at each sample amount with error bars representing one standard deviation from the mean. R2 values were determined

using a scatter plot encompassing all of the data in the linear range

a) Cefuroxime data3,000

2,000

1,000

0

4,000

3,000

2,000

1,000

0

b) Pseudoephedrine data

Peak

are

aPe

ak a

rea

Time of flight (ms)

Drift gas Fragments

Drift gas Fragments

(+) Cefuroxime ion

(+) PseudoephedrineHCI ion

0 2 4 6 8

0 2 4 6 8

Peak

are

a

Sample amount (Micrograms)

1,600

1,400

1,200

1,000

800

600

400

200

0

Peak

are

a

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

Linear Range: 0.5-1.5µgR2 > 0.95Average RSD = 8.9%

Linear Range: 10-25µgR2 > 0.95Average RSD = 4.9%

0 0.5 1 1.5 2 2.5 3 3.5

Sample amount (Micrograms)

0 5 10 15 20 25 30

Figure 2: Plasmagram of cefuroxime (top) and pseudoephedrine (bottom) measurements.

Positive ion data is shown, indicating the primary ion complex, drift gas and fragments

Time of flight (ms)

IPT 25 2008 24/4/08 12:37 Page 92

that the limit of detection (LOD) was below 250ng.The linear range was considered to be between 500ngand 1.5µg, values corresponding to 50% and 150% ofthe carryover limit, respectively. This is a greatertolerance than called for normally, as cefuroxime’s low carryover limits appropriate a wider window of measurement.

Additionally, 500ng was twice the value of the limit ofquantification and more than twice the level of the limitof detection. The R2 value for the linear range of thiscalibration curve was >0.95.

For the purposes of recovery, the spiked samplesrepresented 100% recovery for the API. This wasvalidated with two sets of measurements: (1)measuring for any residual cefuroxime on trapscontaining 1.5µg and 3µg after they had been sampled for the calibration curve; and (2) measuring asample of five glass fibre traps coated withpolytetrafluoroethelylene (PTFE) that were placedunderneath the sample traps as they were spiked withcefuroxime. Both sets of measurements failed to showany presence of residual cefuroxime.

Pseudoephedrine DataThe pseudoephedrine measurements encompassedsample amounts between 5µg and 25µg. The limits ofdetection and quantification with these instrumentsettings were well below 5µg, and the lower bound ofthe linear range (10µg) was therefore greater than twicethe amount of both the LOD and LOQ. The linearrange of 10-25µg encompassed more than +/-25% ofthe carryover limit of 20µg. Again, the R2 value for thelinear range of this calibration curve was >0.95, and thetests mentioned above for validating 100% recovery ofthe spiked samples were performed as describedpreviously.

Measurements of swabs after they had been sampledproduced no trace of pseudoephedrine. Measurementof the PTFE traps placed underneath the 20µg sampleyielded trace amounts (under 100 instrument counts,representing under 100ng of pseudoephedrine) in twoout of five samples. As this represents less than 0.4%of the total sample, the spiked samples wereconsidered to be representative of 100% recovery forthis experiment.

SWAB RECOVERY

Swabbing was performed on 316 stainless steelcoupons with a #7 finish, in an area of 25cm2.

Aliquots of each sample were spiked onto the couponsand allowed to dry before swabbing. The materialused for swabbing was a specialised polyimidematerial developed for use with the Kaye ValidatorITMS instrument. The swabs have a specific‘sampling area’ that comprises the area of the swabthat is fully sampled by the instrument.

This area was wet with 200µl of methanol and, using aPTFE barrier between the swabber’s finger and theswabbing material, the trap was applied to the surfaceand swabbing commenced with overlapping verticalstrokes across the surface. The swabber performed eightstrokes in a vertical motion, followed by eightoverlapping strokes in a horizontal motion. Figure 4shows these motions.

After swabbing, the traps were allowed to dry and weremeasured with the ITMS system. The areas for the APIpeaks were recorded and the amount of API presentdetermined through the equation generated by thelinear fits of the data shown in Figure 3. Table 1 (seepage 94) shows the calculated recovery percentages foreach level.

CONCLUSION

These data show a recovery percentage of greater than65% and strong repeatability, with a relative standarddeviation (RSD) of 17.4% for cefuroxime and arecovery percentage of greater than 87% forpseudoephedrine with an RSD of below 15%.Additionally, the recovery percentages at varying levels

93Innovations in Pharmaceutical Technology

Figure 4: Swabbing motion on the steel coupons (left), where strokes are initiated in a vertical direction and are then followed by strokes

in a horizontal direction

Steel coupons, 316 finish

1) Vertical, overlapping strokes:moving left � right

2) Horizontal, overlapping strokes:moving bottom � top

IPT 25 2008 24/4/08 12:38 Page 93

of sample for this experiment were consistent. Thesedata demonstrated the desired result of thisexperimentation – namely that it is possible torepeatably generate acceptable recovery of residues andmeasure the samples directly using ITMS.

While this experiment shows the feasibility of thetechnique, the method itself has the potential to beimproved so that higher recoveries are possible. Potentialalterations in the pressure and speed of the swabbing, theorientation or the ‘leading edge’ used with each swabbingstroke, and the amount of solvent used could lead tohigher recovery percentages.

This study demonstrates the feasibility of using theKaye Validator ITMS for the direct sampling ofequipment in the pharmaceutical industry. Whilecleaning validation and verification of equipmentinvolves increased layers of complexity – one of thembeing the different types of surfaces that are likely to beencountered during cleaning – the Validator ITMSdemonstrated the ability to produce acceptable levels ofrecovery and repeatability with a technique that is farfaster than the technology currently used by most of the industry.

Given the high costs associated with pharmaceuticalmanufacture, as well as the push for greater processunderstanding through the FDA’s PAT initiative, theimplementation of ITMS as a fast, specific analyticaltechnology for at-line measurements has the potential todeliver substantial improvements in cleaning analysis andmonitoring efficiency.

The author can be contacted at ralf.wottrich@ge.com

References

1. Eiceman GA and Karpas Z. Ion Mobility

Spectrometry. Taylor and Francis

Group. Boca Raton, FL, 2005

2. Parmeter JE and Eiceman GA. Trace Detection

of Narcotics Using a Preconcentrator/Ion

Mobility Spectrometer System. NIJ Report 602-00.

April 2001

3. Brand D, Li X and Wortley T. Ion Trap

Mobility Spectrometry – Reducing Downtime

in Cleaning Validation and Verification.

www.pharmamanufacturing.com. February 2006

4. Munden R et al. IMS Limit Test Improves Cleaning

Verification and Method Development. Pharmaceutical

Technology Europe. October 2002

5. Peterson DE, et al. Ion Mobility Spectrometry for

Determination of Active Drug in Blinded Dosage

Forms. AAPS. pp18-19, February 2005

6. FDA. Guide to Inspections of Validation of Cleaning

Processes. July 1993

7. LeBlanc D. Establishing Scientifically Justified

Acceptance Criteria for Cleaning Validation of

Finished Drug Products. Pharmaceutical Technology,

Volume 22 (10). October 1998

8. LeBlanc DA. Setting Dose Limits Without Dosing

Information. www.cleaningvalidation.com, Cleaning

Memos, May 2001

9. Kramer, et al. Conversion Factors Estimating

Indicative Chronic No-Observed-Adverse-Effect Levels

from Short-Term Toxicity Data. Regulatory Toxicology

and Pharmacology. Volume 23. pp249-255. 1996

10. Swartz ME and Krull IS. Analytical Method

Development and Validation. Marcel Dekker, Inc.

New York, 1997

11. Romano A et al. Immediate hypersensitivity

to cephalosporins. Allergy (57) Supplement 72.

pp52-57. 2002

12. Rossi S (Ed.) 2004. Australian Medicines Handbook

2004. Adelaide, Australia. ISBN 0-9578521-4-2.

13. Cleaning limits provided in private communications

94 Innovations in Pharmaceutical Technology

Cefuroxime recovery data

Amount on coupon (n) Mean amount recovered Recovery %

1.5 micrograms (n = 7) 1.02 micrograms 67.2%

1 microgram (n = 11) 660 nanograms 65.6%

500 nanograms (n = 7) 310 nanograms 62%

Average swab recovery 65.1%

Swab recovery RSD% 17.4%

Pseudoephedrine recovery data

Amount on coupon (n) Mean amount recovered Recovery %

15 micrograms (n = 8) 13.34 micrograms 88.9%

20 micrograms (n = 8) 17.07 micrograms 85.3%

Average swab recovery 87.1%

Swab recovery RSD% 14.2%

Table 1: Cefuroxime recovery and pseudoephedrine recovery data

IPT 25 2008 24/4/08 12:38 Page 94