- Email: [email protected]
Impact of automation on mass spectrometry
Impact of automation on mass spectrometry Yan Victoria Zhang, Alan Rockwood PII: DOI: Reference:
S0009-8981(15)00407-6 doi: 10.1016/j.cca.2015.08.027 CCA 14089
To appear in:
Clinica Chimica Acta
Received date: Revised date: Accepted date:
21 June 2015 13 August 2015 30 August 2015
Please cite this article as: Zhang Yan Victoria, Rockwood Alan, Impact of automation on mass spectrometry, Clinica Chimica Acta (2015), doi: 10.1016/j.cca.2015.08.027
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Impact of Automation on Mass Spectrometry
PT
Yan Victoria Zhang1, Alan Rockwood2,3
NU
SC
RI
1. Department of Pathology and Laboratory Medicine, University of Rochester Medical Center, Rochester, NY 14642; 2. Department of Pathology, University of Utah School of Medicine, Salt Lake City, UT 84112; 3. ARUP Laboratories, Salt Lake City, UT, 84108
MA
Corresponding Author
Yan Victoria Zhang, Ph.D., DABCC
AC CE P
Rochester, NY 14642
TE
601 Elmwood Avenue Box 608
D
University of Rochester Medical Center
Phone: 585-276-4192
Email: [email protected]
ACCEPTED MANUSCRIPT Abstract
AC CE P
TE
D
MA
NU
SC
RI
PT
Mass spectrometry coupled to liquid chromatography (LC-MS and LC-MS/MS) is an analytical technique that has rapidly grown in popularity in clinical practice. In contrast to traditional technology, mass spectrometry is superior in many respects including resolution, specificity, multiplex capability and has the ability to measure analytes in various matrices. Despite these advantages, LC-MS/MS remains high cost, labor intensive and has limited throughput. This specialized technology requires highly trained personnel and therefore has largely been limited to large institutions, academic organizations and reference laboratories. Advances in automation will be paramount to break through this bottleneck and increase its appeal for routine use. This article reviews these challenges, shares perspectives on essential features for LC-MS/MS total automation and proposes a step-wise and incremental approach to achieve total automation through reducing human intervention, increasing throughput and eventually integrating the LC-MS/MS system into the automated clinical laboratory operations.
Keywords: mass spectrometry, liquid chromatography, clinical practices, automation, routine clinical laboratories, high throughput
ACCEPTED MANUSCRIPT 1. Introduction
NU
SC
RI
PT
Mass spectrometry (MS) has been proven to be a very powerful analytical platform and has been applied in several clinical fields over the last few decades. MS was first used in diagnosing inborn errors of metabolism [1-3] and then became the technology of choice for forensic and clinical toxicology [4, 5]. The combination of liquid chromatography and tandem mass spectrometry (LC-MS/MS) provided even faster growth when soft ionization techniques became available[6]. LC-MS/MS is now commonly used in many clinical specialties such as endocrinology[7], immunosuppressant and therapeutic drug monitoring [8], small molecule and peptide and protein marker analysis [9-12]. Recently, MS has shown great progress in microorganism identification [13], initiating the exploration of a new clinical area.
TE
D
MA
The expansion of MS in clinical practice is mainly due to its several advantages over traditional immunoassay, ultraviolet or florescence-based technologies. MS itself is a high resolution technique with high specificity that allows positive identification of compounds of interest [14]. It is able to detect several or even hundreds of analytes simultaneously. MS is also compatible with samples in a variety of matrices including serum, plasma, urine and saliva. In addition, most MS-based assays do not rely on raising antibodies and therefore the method development time has been greatly reduced.
AC CE P
While MS is making its way into many clinical laboratories, it is still limited to specialty laboratories and many challenges that prevent it from being implemented routinely. The overall workflow is labor intensive and manual process-driven; it requires highly trained technical staff to perform daily operations, regular troubleshooting and assay development and validation [15-18]. Limited access to those expertise and extensive technical training requirement has hindered the further growth and implementation of this platform [19-21]. The throughput is lower than other chemistry or immunoassay analyzers in clinical laboratories, which provides less desirable productivity and turnaround time for many clinical applications. [21, 22]. In the near term, MS will remain a specialty instrument. We believe, however, that automation on many important external features will be essential enabling this platform to be more applicable in the routine clinical use. The ideal mass spectrometer will be accessible to those who desire its use without the requirement of becoming an expert in its technical aspects. This article will explore the desirable automation features for MS. Recent developments in these areas and the practical interim steps that clinical laboratories can take to improve the efficiency and productivity of the platform to achieve the ultimate goal of developing MS into a total automation, high throughput, continuous and random access platform. 2. Current status of LC-MS/MS workflow
ACCEPTED MANUSCRIPT
PT
A typically LC-MS/MS workflow includes sample receiving/acquisition, sample preparation, LCMS/MS analysis, data review and results reporting steps (Figure 1). Though LC-MS/MS analysis itself is fairly automated, the overall workflow requires manual processes which are laborintensive and time consuming.
NU
SC
RI
The sample receiving and acquisition step includes retrieving or obtaining samples from the central processing site or another part of the core laboratory, sample identification against a working or sample list, centrifugation if necessary, decapping, aliquoting and re-labeling steps. Those steps are typically processed manually. Samples are processed in batches to increase productivity. Processing samples for a typical 96-well plate, the sample receiving step can take 10 to 30 minutes depending on the specific processes involved.
AC CE P
TE
D
MA
Sample preparation involves a single or multiple steps for samples to be LC-MS/MS compatible. For instance, whole blood samples require lysis of the cells to release the analyte before further clean up. Other body fluid samples such as serum or plasma may be treated with protein precipitation to remove large molecules in the sample. Liquid-liquid extraction or solid phase extraction (SPE) can be used alone or combined with other methods such as protein precipitation to make samples compatible with mass spectrometric analysis. Those steps can be manual or semi-automated with liquid handlers or SPE handlers. The time required for one batch (96 well plate) is at the scale of 10 to 30 minutes. Liquid handler can reduce human involvement, but cannot significantly reduce the overall time requirement for this step. Although being the essence of the platform, LC separation and MS analysis is not the laborintensive step or mostly the rate-limiting step in the entire workflow. The LC separation takes a few minutes to typically less than 10 minutes and the MS analysis is as fast as a fraction of a second for most instruments and most applications. However, before LC-MS/MS analysis, sample information typically needs to be manually entered to the MS analytical software to create a work list. After analysis, the results are reviewed by the staff for any adjustments or reintegration of chromatograms or other data review required before the results are transferred into the laboratory information system and then reported to electronic health record. Data review and results reporting can take another 10 – 20 minutes for one batch of samples. Current MS practice does not necessarily fit the overall workflow in a clinical laboratory. It is labor intensive, has limited throughput, suffers from a lack of robustness, and has no random access. As a result, MS requires a high level of human capital besides its high capital expenses and is under pressure to become more automated with high throughput to reduce overall cost [23]. 3. Key features for future automated LC-MS/MS workflow
ACCEPTED MANUSCRIPT
SC
RI
PT
Higher level of automation will reduce the barriers of entry, increase scalability and enable the LC-MS/MS platform to be accessible and implemented in routine clinical laboratories. Automated mass spectrometry platform can in the future either be a standalone floor model (Figure 2A) or integrated into the core laboratory function as one of the automated platform along with chemistry analyzers, hematology analyzers, coagulation analyzers and others (Figure 2B). In either case, key features include automated pre-analytical process, bar coding system, a bi-directional interface, an automated sample preparation system, and advanced software for data review process.
MA
NU
By being part of the core laboratory automation system, the MS platform can take advantage of the existing pre-analytical module for sample login, de-capping, sample sorting, and aliquoting. Those aliquots are then delivered (often through a conveyor belt) to the designated analyzer (Figure 2B). A standalone MS platform must include the bar coding system and interface to get access to patient information and tests that are requested. A standalone MS platform may still need manually aliquot as do most standalone chemistry analyzers.
AC CE P
TE
D
MS samples will need to be prepared with an automated liquid handler or similar means with the designated sample preparation method. The treated samples will then be subjected to the chromatographic separation and MS/MS analysis. Results will be processed by the software and reported directly to the LIS. Any abnormal samples or results will be reported through integrated software or middleware for human intervention. Total automation will also enable random access to reduce turnaround time, a parameter that is of more importance to the client than sample throughput, whereas sample throughput is primarily of interest to the performing laboratory. This future vision of MS seems idealistic and remote, and yet recent developments in MS and the history of automated chemistry analyzers shows that all those goals can be accomplished. Nonetheless, it is not an easy undertaking and will require years of continuous effort. We believe this task can be executed through a stepwise approach and gradually gain productivity during the interim. Those steps include reducing manual process and human intervention, increasing throughput and decreasing operational complexity for total automation (Figure 3). The initial steps can be achieved in the foreseeable future and will bring immediate value and benefit to the overall operation, while total automation will require longer time and more efforts to achieve. These milestones are not necessarily linear and there will be a lot of overlapping developments. Each interim step accomplishes unique focuses to bring efficiency and productivity to the overall workflow. 4. Reducing manual process
ACCEPTED MANUSCRIPT
RI
PT
Reducing manual labor process could decrease staff time requirements, increase walk-away time and increase overall efficiency. Though it takes a significant instrument time, the liquid chromatographic separation and MS analysis requires minimal human intervention. The most labor-intensive steps include sample acquisition, sample preparation, reviewing of data and releasing results. Sample preparation, data review and results reporting can benefit from automation and gain immediate benefits to reduce the human time involvement (Figure 4). 4.1 Sample preparation
MA
NU
SC
MS analysis requires some level of sample preparation. Preparation can be as simple as “diluteand-shoot”, as in many urine drug tests. Most times, however, preparation requires more than one step to remove interferences and/or to concentrate the analyte of interest. Commonly used methods include blood cell lysis, protein precipitation, liquid-liquid extraction, and solid phase extraction.
TE
D
Automation of sample preparation may require several platforms to fulfill the specific needs for different protocols. Off line processing using liquid handler and automated solid phase extraction has gained in popularity to reduce manual processing time [24]. Sample preparation can be further simplified by moving the preparation online with LC separation through approaches like turbo flow and two-dimensional chromatography.
AC CE P
Data entry during sample preparation can be automated by implementing a bar code reader, a common practice in automated laboratory systems. It has been shown that barcode system is a far more accurate and faster and accurate approach than human keystrokes for data entry [25, 26]. Barcode systems can increase data entry speed by more than 6 fold to less than 1 second from 6 seconds by keystrokes for 12 characters [27] and reduce the potential error rate from 1 per 300 characters to estimated 1 per 2.7 X 106 (for code 128) [26]. Simply implementing bar code readers and liquid handlers can significantly reduce labor intensity; increase walk-away time and throughput during sample preparation. Furthermore, automation will reduce error and increase data quality. Ideally, six-sigma quality level is the goal regarding each part of the steps in sample handling process.
4.2 Results releasing and interfacing Chromatographic data review and results releasing are two other labor-intensive steps. Data review process to ensure the quality of overall analysis and peak integration. Software packages that were developed to “automate” the data review process, have shown to reduce the assay time per sample from 25.8 to 9.9 minutes and QA review time from 18.3 to 3.9 minutes [28].
ACCEPTED MANUSCRIPT
RI
PT
The results are currently manually entered into the LIS and it requires another person to check the accuracy of data entry. This manual process can be automated by interfacing with the LIS and electronic health record. However, the interfacing process has been a challenge in the field, largely due to the middleware companies’ unfamiliarity with the MS data format and the MS manufacturer’s lack of experience in this field. It requires close collaboration and teamwork among the lab, the institutional IT group, the mass spectrometry vendor, and the vendor of the hospital laboratory information system.
MA
NU
SC
One-directional interfacing has been implemented in several institutions such as the University of Rochester Medical Center (URMC) and ARUP. The most common practice is through stepwise data transfer approach. The results are exported as ASCII files (.csv) or EXCEL file and then uploaded to the LIS. It is not an ideal situation and yet a workable one as a kind of interim solution. The recent implementation at URMC indicated significantly reduction of the overall results reporting time to less than 5 minutes from 45 minutes (unpublished data) for one batch of 96-well plate samples. The error rate was reduced to essentially zero.
AC CE P
5. Improving throughput
TE
D
Although not yet implemented in clinical laboratories to our knowledge, the ultimate goal, however, is to have a bi-directional interface. The bi-directional inferface will enable downloading test requests from LIS to the local system during sample receiving step and direct transfer of the results to the LIS at the end of the analysis.
Limited throughput has been another shortcoming for mass spectrometry-based technology. Throughput can be categorized into LC-MS/MS throughput and overall throughput. The former focuses on the number of samples processed by the LC-MS/MS on a per unit time basis while the latter relate to overall turnaround time to processing samples from sample receiving to results releasing. Improvements in overall throughput and LC-MS/MS throughput can be achieved by different means. 5.1 LC-MS/MS throughput As the MS/MS analysis only takes a fraction of a second, the rate-limiting step for LC-MS/MS analysis is the chromatographic separation, which typically takes a few minutes for each assay. Several approaches have been reported to improve the LC-MS/MS throughput. Multiplexing the chromatographic front ends to one MS/MS system has been a common practice for the labs with large test volumes. The front end can be a dual channel or quadratic channel system [29, 30] and it can therefore increase the throughput by 2 - 4 fold.
ACCEPTED MANUSCRIPT
RI
PT
Throughput can also be improved by sample multiplexing which allows more than one sample to be injected at once. This requires sample derivation with differential mass taggings [31], so that each sample will have a different m/z and can be detected by the MS simultaneously. It was reported that five differentially tagged samples were analyzed per injection, and the throughput was increased from 60 samples per hour to 300 samples per hour [31]. This technique has a great potential to increase the throughput significantly and up to the point where the dwell time of mass spectrometer becomes the limiting step.
NU
SC
Another approach is to reduce chromatographic separation. Recent development using online solid phase extraction cartridges reduced the HPLC time from a typical 2 minutes to 15 seconds [32], which enables a significant increase of LC-MS/MS throughput.
MA
Alternatively, one may consider de-coupling chromatographic separation from the mass spectrometer, performing chromatographic separation off-line, using fraction collection followed by high throughput flow injection analysis on the mass spectrometer. This would also have the advantage of maximizing sample throughput through the mass spectrometer.
D
5.2 Overall throughput
AC CE P
TE
Improving overall throughput requires considerations for all the steps in the sample analysis process including sample acquisition, sample preparation, LC-MS/MS analysis and data review and result reporting. Improvement of the overall throughput is a dynamic process, which focuses on the rate-limiting step. At different stage of automation, the rate-limiting step may be different. For instance, the overall throughput will increase as the LC-MS/MS throughput increases until the pre- and post- LC-MS/MS steps become the rate-limiting steps. The approaches that reduce manual processing time do not necessarily improve the overall throughput or to the same extent (Figure 4). For instance, liquid handler for sample preparation minimizes human intervention and increases human productivity, and yet it does not significant reduce the total sample processing time that is associated with the overall throughput. However, interface discussed above reduces human involvement and overall sample processing time to improve overall throughput. It is worth mentioning that the time needed to transfer the sample to the lab can take from a few hours to overnight or longer which is another challenge for the overall throughput and turnaround time. Most laboratories have limited impact on improving the efficiency of this step. 6. Other considerations for achieving total automation
ACCEPTED MANUSCRIPT Besides maximizing walk-away time and increasing throughput, total automation requires other considerations such as a broad test menu, the availability of commercial assay kits, standardization and results harmonization, and timely technical services.
PT
6.1 Commercial assay kits
MA
NU
SC
RI
Developing and validating MS assays is a major undertaking in a lab requiring a great deal of time and expertise. MS assays are laboratory developed tests (LDTs), which require the most stringent validation under Clinical Laboratory Improvement Amendments of 1988 (CLIA’88) regulation. Limited access to those who with this expertise has been a challenge to implement the MS platform in many clinical laboratories. Recently, Clinical & Laboratory Standards Institute (CLSI) issued a guideline (C62-A) on method development and validation. Though the guideline does not solve the lack of expertise reality, it provided the best practices and benchmark for method development and validation procedures and some key considerations.
AC CE P
TE
D
Commercially available assay kits will alleviates some of the challenges in developing and validating assays in the clinical labs. Recently the Food and Drug Administration (FDA) decided to exercise their enforcement discretion and extend their regulatory oversight to LDTs [33, 34]. As of the time of this writing, the FDA guideline has not released. However, it is likely that all current MS-based LDT assays will be subjected to this regulation. So far, only one FDA approved commercial assay kit is available on the market. This new regulation will significantly impact the vendors’ decision in moving forward with FDA approved assay kits to clinical laboratories. However, until FDA approved clinical assays are available for mass spectrometry; the regulation may hinder the adoption of this platform into routine clinical practices. 6.2 Standardization and results harmonization The limitations of both inter-laboratory variations of MS assays are well recognized, and in some but not all cases, these variations are even higher than immunoassays [35]. Different platforms may have different responses to the same analyte, which contributes to the variations among or even within vendors. Several approaches have the potential to enhance standardization. Use of a common calibrator has been reported to reduce the variation in results for vitamin D assays [36, 37], but not for other assays [38]. Inter-laboratory performance can be improved through the use of commercial kits, and yet the differences in performance of the same instruments still exist [39, 40]. While LC-MS/MS provides flexibility to develop individual assays with unique sample preparation and analytical parameters, it is key to ensure the harmonization and communicability of results through approaches such as accuracy based proficiency testing. 6.3 Cost and post-sale services
ACCEPTED MANUSCRIPT
RI
PT
High capital cost hinders the acceptance of the technology. However, the costs of the instrument are not likely to come down until the manufacturers reach the economy of scale for production. The economy of scale will rely on the automation steps discussed above. Therefore, it is reasonable to assume that reduced cost will be the result of automation rather than the driver of automation and popularization. Not until then will the manufacturer be able to afford this specialty instrument being sold as a routine platform.
NU
SC
A mass spectrometer is not a black box but rather a very complicated piece of machinery. It requires a sophisticated skill set and expertise to operate and troubleshoot the instrument. Current service contracts from manufacturers tend to be research-oriented and are not appropriate for routine clinical needs. To ensure smooth operation and minimize downtime in regular clinical services, short turnaround-time technical support and 24-hour technical support and short turnaround time service will be essential.
MA
7. Conclusions
AC CE P
TE
D
MS is an emerging technology that has made its way into clinical practices at an accelerating rate, particularly in the past couple of decades. It has been considered an instrument for esoteric tests and applied in specialty laboratories. Many challenges have hindered the implementation of the platform in many clinical laboratories. The field is waiting for the next generation of instruments, improvements in information technology, and vendor support to move mass spectrometry into the mainstream, i.e. into more widespread use by mid-size and smaller labs. In the near future, MS platforms may be differentiated into two categories. One category is the current existing system for esoteric and low volume laboratory developed tests. Another category is a basic model sold as a medical device for high volume and well-defined tests such as vitamin D, drugs of abuse screening and confirmation, and immunosuppressant drugs. The latter category may achieve higher level of automation and standardization for routine clinical usage. Higher level of automation enabling MS a robust, random access, high throughput platform is essential to make MS a technique of choice in routine practices. We believe three major milestones will mitigate those challenges. The first is to increase the walk-away time and reduce manual processing time, the second is to increase overall throughput, and the final step is to achieve total automation, including random access processing and reduced turnaround time. This will be enabled by automation of sample preparation and data analysis, integration with existing automated platforms, Food and Drug Administration clearance/approval, commercially available reagent kits and calibrators and seamless communication with existing electronic systems through bi-directional interfacing. Total automation requires generations of
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
effort and step-wise improvements. The future for MS in the clinical lab is a bright one, and automation will put wings on this giant elephant, enabling it to fly.
ACCEPTED MANUSCRIPT Acknowledgements
AC CE P
TE
D
MA
NU
SC
RI
PT
The authors thank Dr. Thomas Jackson for his discussion and the initial proofreading of the manuscript.
ACCEPTED MANUSCRIPT References
AC CE P
TE
D
MA
NU
SC
RI
PT
[1] Schulze A, Lindner M, Kohlmuller D, Olgemoller K, Mayatepek E, Hoffmann GF. Expanded newborn screening for inborn errors of metabolism by electrospray ionization-tandem mass spectrometry: results, outcome, and implications. Pediatrics 2003; 111:1399-1406. [2] Wilcken B, Wiley V, Hammond J, Carpenter K. Screening newborns for inborn errors of metabolism by tandem mass spectrometry. The New England journal of medicine 2003; 348:2304-2312. [3] Mizuno T, Abe N, Teshima H, et al. Application of a gas chromatography mass spectrometry computer system for clinical diagnosis. Biomedical mass spectrometry 1981; 8:593-597. [4] Marquet P. Progress of liquid chromatography-mass spectrometry in clinical and forensic toxicology. Therapeutic drug monitoring 2002; 24:255-276. [5] Maurer HH. Current role of liquid chromatography-mass spectrometry in clinical and forensic toxicology. Analytical and bioanalytical chemistry 2007; 388:1315-1325. [6] Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization for mass spectrometry of large biomolecules. Science (New York, NY) 1989; 246:64-71. [7] Kushnir MM, Rockwood AL, Bergquist J. Liquid chromatography-tandem mass spectrometry applications in endocrinology. Mass spectrometry reviews 2010; 29:480-502. [8] Streit F, Armstrong VW, Oellerich M. Rapid liquid chromatography-tandem mass spectrometry routine method for simultaneous determination of sirolimus, everolimus, tacrolimus, and cyclosporin A in whole blood. Clinical chemistry 2002; 48:955-958. [9] Clarke NJ, Zhang Y, Reitz RE. A novel mass spectrometry-based assay for the accurate measurement of thyroglobulin from patient samples containing antithyroglobulin autoantibodies. Journal of investigative medicine : the official publication of the American Federation for Clinical Research 2012; 60:1157-1163. [10] Hoofnagle AN, Roth MY. Clinical review: improving the measurement of serum thyroglobulin with mass spectrometry. The Journal of clinical endocrinology and metabolism 2013; 98:1343-1352. [11] Kushnir MM, Rockwood AL, Roberts WL, Abraham D, Hoofnagle AN, Meikle AW. Measurement of thyroglobulin by liquid chromatography-tandem mass spectrometry in serum and plasma in the presence of antithyroglobulin autoantibodies. Clinical chemistry 2013; 59:982-990. [12] Netzel BC, Grebe SK, Algeciras-Schimnich A. Usefulness of a thyroglobulin liquid chromatography-tandem mass spectrometry assay for evaluation of suspected heterophile interference. Clinical chemistry 2014; 60:1016-1018. [13] Claydon MA, Davey SN, Edwards-Jones V, Gordon DB. The rapid identification of intact microorganisms using mass spectrometry. Nature biotechnology 1996; 14:1584-1586. [14] Anderson L, Hunter CL. Quantitative mass spectrometric multiple reaction monitoring assays for major plasma proteins. Molecular & cellular proteomics : MCP 2006; 5:573-588. [15] Vogeser M, Seger C. A decade of HPLC-MS/MS in the routine clinical laboratory--goals for further developments. Clinical biochemistry 2008; 41:649-662. [16] Hetu PO, Robitaille R, Vinet B. Successful and cost-efficient replacement of immunoassays by tandem mass spectrometry for the quantification of immunosuppressants in the clinical laboratory. Journal of chromatography B, Analytical technologies in the biomedical and life sciences 2012; 883884:95-101. [17] Clarke W, Rhea JM, Molinaro R. Challenges in implementing clinical liquid chromatographytandem mass spectrometry methods--the light at the end of the tunnel. Journal of mass spectrometry : JMS 2013; 48:755-767. [18] Honour JW. Development and validation of a quantitative assay based on tandem mass spectrometry. Annals of clinical biochemistry 2011; 48:97-111.
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
[19] Grebe SK, Singh RJ. LC-MS/MS in the Clinical Laboratory - Where to From Here? The Clinical biochemist Reviews / Australian Association of Clinical Biochemists 2011; 32:5-31. [20] Vogeser M, Kirchhoff F. Progress in automation of LC-MS in laboratory medicine. Clinical biochemistry 2011; 44:4-13. [21] Adaway JE, Keevil BG, Owen LJ. Liquid chromatography tandem mass spectrometry in the clinical laboratory. Annals of clinical biochemistry 2015; 52:18-38. [22] van den Ouweland JM, Kema IP. The role of liquid chromatography-tandem mass spectrometry in the clinical laboratory. Journal of chromatography B, Analytical technologies in the biomedical and life sciences 2012; 883-884:18-32. [23] Panteghini M. The future of laboratory medicine: understanding the new pressures. The Clinical biochemist Reviews / Australian Association of Clinical Biochemists 2004; 25:207-215. [24] Robandt PP, Reda LJ, Klette KL. Complete automation of solid-phase extraction with subsequent liquid chromatography-tandem mass spectrometry for the quantification of benzoylecgonine, mhydroxybenzoylecgonine, p-hydroxybenzoylecgonine, and norbenzoylecgonine in urine--application to a high-throughput urine analysis laboratory. Journal of analytical toxicology 2008; 32:577-585. [25] Poon EG, Keohane CA, Yoon CS, et al. Effect of bar-code technology on the safety of medication administration. The New England journal of medicine 2010; 362:1698-1707. [26] Snyder ML, Carter A, Jenkins K, Fantz CR. Patient misidentifications caused by errors in standard bar code technology. Clinical chemistry 2010; 56:1554-1560. [27] Hawker CD. Bar codes may have poorer error rates than commonly believed. Clinical chemistry 2010; 56:1513-1514. [28] Dickerson JA, Schmeling M, Hoofnagle AN, Hoffman NG. Design and implementation of software for automated quality control and data analysis for a complex LC/MS/MS assay for urine opiates and metabolites. Clinica chimica acta; international journal of clinical chemistry 2013; 415:290-294. [29] Wu JT. The development of a staggered parallel separation liquid chromatography/tandem mass spectrometry system with on-line extraction for high-throughout screening of drug candidates in biological fluids. Rapid communications in mass spectrometry : RCM 2001; 15:73-81. [30] Xu RN, Fan L, Rieser MJ, El-Shourbagy TA. Recent advances in high-throughput quantitative bioanalysis by LC-MS/MS. Journal of pharmaceutical and biomedical analysis 2007; 44:342-355. [31] Netzel BC, Cradic KW, Bro ET, et al. Increasing liquid chromatography-tandem mass spectrometry throughput by mass tagging: a sample-multiplexed high-throughput assay for 25hydroxyvitamin D2 and D3. Clinical chemistry 2011; 57:431-440. [32] Grote-Koska D, Czajkowski S, Brand K. Performance of the new RapidFire(R) system for therapeutic monitoring of immunosuppressants. Therapeutic drug monitoring 2014. [33] Gutman S. The role of Food and Drug Administration regulation of in vitro diagnostic devices-applications to genetics testing. Clinical chemistry 1999; 45:746-749. [34] Scott MG, Ashwood ER, Annesley TM, Leonard DG, Burgess MC. FDA oversight of laboratorydeveloped tests: is it necessary, and how would it impact clinical laboratories? Clinical chemistry 2013; 59:1017-1022. [35] Vogeser M, Seger C. Pitfalls associated with the use of liquid chromatography-tandem mass spectrometry in the clinical laboratory. Clinical chemistry 2010; 56:1234-1244. [36] Yates AM, Bowron A, Calton L, et al. Interlaboratory variation in 25-hydroxyvitamin D2 and 25hydroxyvitamin D3 is significantly improved if common calibration material is used. Clinical chemistry 2008; 54:2082-2084. [37] Carter GD. 25-Hydroxyvitamin D assays: the quest for accuracy. Clinical chemistry 2009; 55:1300-1302. [38] Levine DM, Maine GT, Armbruster DA, et al. The need for standardization of tacrolimus assays. Clinical chemistry 2011; 57:1739-1747.
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
[39] Napoli KL, Hammett-Stabler C, Taylor PJ, et al. Multi-center evaluation of a commercial Kit for tacrolimus determination by LC/MS/MS. Clinical biochemistry 2010; 43:910-920. [40] Carter GD, Jones JC. Use of a common standard improves the performance of liquid chromatography-tandem mass spectrometry methods for serum 25-hydroxyvitamin-D. Annals of clinical biochemistry 2009; 46:79-81.
ACCEPTED MANUSCRIPT Figure legends Figure 1. Current LC-MS/MS platform workflow and typical time requirement for each step.
RI
PT
Sample receiving includes sample identification, centrifugation, de-capping, aliquoting and relabeling steps. Time estimate is based on one batch sample for one 96 well plate. Blank box: automated process; dotted line: semi-automated process; dots: manual process
SC
Figure 2. Possible automated LC-MS/MS workflows.
NU
Panel A. Automated LC-MS/MS as a standalone analyzer and the key features and potential solutions for automation. Panel B. LC-MS/MS as part of the automated clinical laboratory system and samples are processed and transferred through the conveyor belt to the designated analyzers.
MA
Figure 3. Interim steps to take to achieve total automation for LC-MS/MS platform and the key features for each step.
TE
D
Figure 4. Impact of automation approaches on manual and total processing time for various steps in the LC-MS/MS workflow.
AC CE P
Major steps are labeled on the y-axis. As MS/MS analysis takes no manual time and very little processing time, it is not indicated in the diagram. Blank box, manual processing time; gray box, total processing time. Time required for each step is labeled in relative scale as the current time requirement being the full scale (indicated at progression point 0). Half size box, significant total processing time required; boxes with diagonal line, eliminate nearly all manual or total processing time. Numbers indicated various approaches. 1. LC-multiplexing and online or offline LC separation approaches; 2. Automated sample preparation approaches such as liquid sampler; 3. Interface connection for results reporting; 4. Automated data review software; 5. Total automation for sample receiving step.
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
TE
D
MA
NU
SC
RI
PT
LC-MS/MS becomes an important analytical platform in clinical practices. Achieving total automation is key to bring LC-MS/MS into routine clinical labs. Incremental improvements are essential for LC-MS/MS to achieve total automation. Automation includes sample acquisition, preparation, data review and reporting. Interfacing with LIS will gain immediate benefits in reducing labor intensity.
AC CE P
1. 2. 3. 4. 5.
S0009-8981(15)00407-6 doi: 10.1016/j.cca.2015.08.027 CCA 14089
To appear in:
Clinica Chimica Acta
Received date: Revised date: Accepted date:
21 June 2015 13 August 2015 30 August 2015
Please cite this article as: Zhang Yan Victoria, Rockwood Alan, Impact of automation on mass spectrometry, Clinica Chimica Acta (2015), doi: 10.1016/j.cca.2015.08.027
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Impact of Automation on Mass Spectrometry
PT
Yan Victoria Zhang1, Alan Rockwood2,3
NU
SC
RI
1. Department of Pathology and Laboratory Medicine, University of Rochester Medical Center, Rochester, NY 14642; 2. Department of Pathology, University of Utah School of Medicine, Salt Lake City, UT 84112; 3. ARUP Laboratories, Salt Lake City, UT, 84108
MA
Corresponding Author
Yan Victoria Zhang, Ph.D., DABCC
AC CE P
Rochester, NY 14642
TE
601 Elmwood Avenue Box 608
D
University of Rochester Medical Center
Phone: 585-276-4192
Email: [email protected]
ACCEPTED MANUSCRIPT Abstract
AC CE P
TE
D
MA
NU
SC
RI
PT
Mass spectrometry coupled to liquid chromatography (LC-MS and LC-MS/MS) is an analytical technique that has rapidly grown in popularity in clinical practice. In contrast to traditional technology, mass spectrometry is superior in many respects including resolution, specificity, multiplex capability and has the ability to measure analytes in various matrices. Despite these advantages, LC-MS/MS remains high cost, labor intensive and has limited throughput. This specialized technology requires highly trained personnel and therefore has largely been limited to large institutions, academic organizations and reference laboratories. Advances in automation will be paramount to break through this bottleneck and increase its appeal for routine use. This article reviews these challenges, shares perspectives on essential features for LC-MS/MS total automation and proposes a step-wise and incremental approach to achieve total automation through reducing human intervention, increasing throughput and eventually integrating the LC-MS/MS system into the automated clinical laboratory operations.
Keywords: mass spectrometry, liquid chromatography, clinical practices, automation, routine clinical laboratories, high throughput
ACCEPTED MANUSCRIPT 1. Introduction
NU
SC
RI
PT
Mass spectrometry (MS) has been proven to be a very powerful analytical platform and has been applied in several clinical fields over the last few decades. MS was first used in diagnosing inborn errors of metabolism [1-3] and then became the technology of choice for forensic and clinical toxicology [4, 5]. The combination of liquid chromatography and tandem mass spectrometry (LC-MS/MS) provided even faster growth when soft ionization techniques became available[6]. LC-MS/MS is now commonly used in many clinical specialties such as endocrinology[7], immunosuppressant and therapeutic drug monitoring [8], small molecule and peptide and protein marker analysis [9-12]. Recently, MS has shown great progress in microorganism identification [13], initiating the exploration of a new clinical area.
TE
D
MA
The expansion of MS in clinical practice is mainly due to its several advantages over traditional immunoassay, ultraviolet or florescence-based technologies. MS itself is a high resolution technique with high specificity that allows positive identification of compounds of interest [14]. It is able to detect several or even hundreds of analytes simultaneously. MS is also compatible with samples in a variety of matrices including serum, plasma, urine and saliva. In addition, most MS-based assays do not rely on raising antibodies and therefore the method development time has been greatly reduced.
AC CE P
While MS is making its way into many clinical laboratories, it is still limited to specialty laboratories and many challenges that prevent it from being implemented routinely. The overall workflow is labor intensive and manual process-driven; it requires highly trained technical staff to perform daily operations, regular troubleshooting and assay development and validation [15-18]. Limited access to those expertise and extensive technical training requirement has hindered the further growth and implementation of this platform [19-21]. The throughput is lower than other chemistry or immunoassay analyzers in clinical laboratories, which provides less desirable productivity and turnaround time for many clinical applications. [21, 22]. In the near term, MS will remain a specialty instrument. We believe, however, that automation on many important external features will be essential enabling this platform to be more applicable in the routine clinical use. The ideal mass spectrometer will be accessible to those who desire its use without the requirement of becoming an expert in its technical aspects. This article will explore the desirable automation features for MS. Recent developments in these areas and the practical interim steps that clinical laboratories can take to improve the efficiency and productivity of the platform to achieve the ultimate goal of developing MS into a total automation, high throughput, continuous and random access platform. 2. Current status of LC-MS/MS workflow
ACCEPTED MANUSCRIPT
PT
A typically LC-MS/MS workflow includes sample receiving/acquisition, sample preparation, LCMS/MS analysis, data review and results reporting steps (Figure 1). Though LC-MS/MS analysis itself is fairly automated, the overall workflow requires manual processes which are laborintensive and time consuming.
NU
SC
RI
The sample receiving and acquisition step includes retrieving or obtaining samples from the central processing site or another part of the core laboratory, sample identification against a working or sample list, centrifugation if necessary, decapping, aliquoting and re-labeling steps. Those steps are typically processed manually. Samples are processed in batches to increase productivity. Processing samples for a typical 96-well plate, the sample receiving step can take 10 to 30 minutes depending on the specific processes involved.
AC CE P
TE
D
MA
Sample preparation involves a single or multiple steps for samples to be LC-MS/MS compatible. For instance, whole blood samples require lysis of the cells to release the analyte before further clean up. Other body fluid samples such as serum or plasma may be treated with protein precipitation to remove large molecules in the sample. Liquid-liquid extraction or solid phase extraction (SPE) can be used alone or combined with other methods such as protein precipitation to make samples compatible with mass spectrometric analysis. Those steps can be manual or semi-automated with liquid handlers or SPE handlers. The time required for one batch (96 well plate) is at the scale of 10 to 30 minutes. Liquid handler can reduce human involvement, but cannot significantly reduce the overall time requirement for this step. Although being the essence of the platform, LC separation and MS analysis is not the laborintensive step or mostly the rate-limiting step in the entire workflow. The LC separation takes a few minutes to typically less than 10 minutes and the MS analysis is as fast as a fraction of a second for most instruments and most applications. However, before LC-MS/MS analysis, sample information typically needs to be manually entered to the MS analytical software to create a work list. After analysis, the results are reviewed by the staff for any adjustments or reintegration of chromatograms or other data review required before the results are transferred into the laboratory information system and then reported to electronic health record. Data review and results reporting can take another 10 – 20 minutes for one batch of samples. Current MS practice does not necessarily fit the overall workflow in a clinical laboratory. It is labor intensive, has limited throughput, suffers from a lack of robustness, and has no random access. As a result, MS requires a high level of human capital besides its high capital expenses and is under pressure to become more automated with high throughput to reduce overall cost [23]. 3. Key features for future automated LC-MS/MS workflow
ACCEPTED MANUSCRIPT
SC
RI
PT
Higher level of automation will reduce the barriers of entry, increase scalability and enable the LC-MS/MS platform to be accessible and implemented in routine clinical laboratories. Automated mass spectrometry platform can in the future either be a standalone floor model (Figure 2A) or integrated into the core laboratory function as one of the automated platform along with chemistry analyzers, hematology analyzers, coagulation analyzers and others (Figure 2B). In either case, key features include automated pre-analytical process, bar coding system, a bi-directional interface, an automated sample preparation system, and advanced software for data review process.
MA
NU
By being part of the core laboratory automation system, the MS platform can take advantage of the existing pre-analytical module for sample login, de-capping, sample sorting, and aliquoting. Those aliquots are then delivered (often through a conveyor belt) to the designated analyzer (Figure 2B). A standalone MS platform must include the bar coding system and interface to get access to patient information and tests that are requested. A standalone MS platform may still need manually aliquot as do most standalone chemistry analyzers.
AC CE P
TE
D
MS samples will need to be prepared with an automated liquid handler or similar means with the designated sample preparation method. The treated samples will then be subjected to the chromatographic separation and MS/MS analysis. Results will be processed by the software and reported directly to the LIS. Any abnormal samples or results will be reported through integrated software or middleware for human intervention. Total automation will also enable random access to reduce turnaround time, a parameter that is of more importance to the client than sample throughput, whereas sample throughput is primarily of interest to the performing laboratory. This future vision of MS seems idealistic and remote, and yet recent developments in MS and the history of automated chemistry analyzers shows that all those goals can be accomplished. Nonetheless, it is not an easy undertaking and will require years of continuous effort. We believe this task can be executed through a stepwise approach and gradually gain productivity during the interim. Those steps include reducing manual process and human intervention, increasing throughput and decreasing operational complexity for total automation (Figure 3). The initial steps can be achieved in the foreseeable future and will bring immediate value and benefit to the overall operation, while total automation will require longer time and more efforts to achieve. These milestones are not necessarily linear and there will be a lot of overlapping developments. Each interim step accomplishes unique focuses to bring efficiency and productivity to the overall workflow. 4. Reducing manual process
ACCEPTED MANUSCRIPT
RI
PT
Reducing manual labor process could decrease staff time requirements, increase walk-away time and increase overall efficiency. Though it takes a significant instrument time, the liquid chromatographic separation and MS analysis requires minimal human intervention. The most labor-intensive steps include sample acquisition, sample preparation, reviewing of data and releasing results. Sample preparation, data review and results reporting can benefit from automation and gain immediate benefits to reduce the human time involvement (Figure 4). 4.1 Sample preparation
MA
NU
SC
MS analysis requires some level of sample preparation. Preparation can be as simple as “diluteand-shoot”, as in many urine drug tests. Most times, however, preparation requires more than one step to remove interferences and/or to concentrate the analyte of interest. Commonly used methods include blood cell lysis, protein precipitation, liquid-liquid extraction, and solid phase extraction.
TE
D
Automation of sample preparation may require several platforms to fulfill the specific needs for different protocols. Off line processing using liquid handler and automated solid phase extraction has gained in popularity to reduce manual processing time [24]. Sample preparation can be further simplified by moving the preparation online with LC separation through approaches like turbo flow and two-dimensional chromatography.
AC CE P
Data entry during sample preparation can be automated by implementing a bar code reader, a common practice in automated laboratory systems. It has been shown that barcode system is a far more accurate and faster and accurate approach than human keystrokes for data entry [25, 26]. Barcode systems can increase data entry speed by more than 6 fold to less than 1 second from 6 seconds by keystrokes for 12 characters [27] and reduce the potential error rate from 1 per 300 characters to estimated 1 per 2.7 X 106 (for code 128) [26]. Simply implementing bar code readers and liquid handlers can significantly reduce labor intensity; increase walk-away time and throughput during sample preparation. Furthermore, automation will reduce error and increase data quality. Ideally, six-sigma quality level is the goal regarding each part of the steps in sample handling process.
4.2 Results releasing and interfacing Chromatographic data review and results releasing are two other labor-intensive steps. Data review process to ensure the quality of overall analysis and peak integration. Software packages that were developed to “automate” the data review process, have shown to reduce the assay time per sample from 25.8 to 9.9 minutes and QA review time from 18.3 to 3.9 minutes [28].
ACCEPTED MANUSCRIPT
RI
PT
The results are currently manually entered into the LIS and it requires another person to check the accuracy of data entry. This manual process can be automated by interfacing with the LIS and electronic health record. However, the interfacing process has been a challenge in the field, largely due to the middleware companies’ unfamiliarity with the MS data format and the MS manufacturer’s lack of experience in this field. It requires close collaboration and teamwork among the lab, the institutional IT group, the mass spectrometry vendor, and the vendor of the hospital laboratory information system.
MA
NU
SC
One-directional interfacing has been implemented in several institutions such as the University of Rochester Medical Center (URMC) and ARUP. The most common practice is through stepwise data transfer approach. The results are exported as ASCII files (.csv) or EXCEL file and then uploaded to the LIS. It is not an ideal situation and yet a workable one as a kind of interim solution. The recent implementation at URMC indicated significantly reduction of the overall results reporting time to less than 5 minutes from 45 minutes (unpublished data) for one batch of 96-well plate samples. The error rate was reduced to essentially zero.
AC CE P
5. Improving throughput
TE
D
Although not yet implemented in clinical laboratories to our knowledge, the ultimate goal, however, is to have a bi-directional interface. The bi-directional inferface will enable downloading test requests from LIS to the local system during sample receiving step and direct transfer of the results to the LIS at the end of the analysis.
Limited throughput has been another shortcoming for mass spectrometry-based technology. Throughput can be categorized into LC-MS/MS throughput and overall throughput. The former focuses on the number of samples processed by the LC-MS/MS on a per unit time basis while the latter relate to overall turnaround time to processing samples from sample receiving to results releasing. Improvements in overall throughput and LC-MS/MS throughput can be achieved by different means. 5.1 LC-MS/MS throughput As the MS/MS analysis only takes a fraction of a second, the rate-limiting step for LC-MS/MS analysis is the chromatographic separation, which typically takes a few minutes for each assay. Several approaches have been reported to improve the LC-MS/MS throughput. Multiplexing the chromatographic front ends to one MS/MS system has been a common practice for the labs with large test volumes. The front end can be a dual channel or quadratic channel system [29, 30] and it can therefore increase the throughput by 2 - 4 fold.
ACCEPTED MANUSCRIPT
RI
PT
Throughput can also be improved by sample multiplexing which allows more than one sample to be injected at once. This requires sample derivation with differential mass taggings [31], so that each sample will have a different m/z and can be detected by the MS simultaneously. It was reported that five differentially tagged samples were analyzed per injection, and the throughput was increased from 60 samples per hour to 300 samples per hour [31]. This technique has a great potential to increase the throughput significantly and up to the point where the dwell time of mass spectrometer becomes the limiting step.
NU
SC
Another approach is to reduce chromatographic separation. Recent development using online solid phase extraction cartridges reduced the HPLC time from a typical 2 minutes to 15 seconds [32], which enables a significant increase of LC-MS/MS throughput.
MA
Alternatively, one may consider de-coupling chromatographic separation from the mass spectrometer, performing chromatographic separation off-line, using fraction collection followed by high throughput flow injection analysis on the mass spectrometer. This would also have the advantage of maximizing sample throughput through the mass spectrometer.
D
5.2 Overall throughput
AC CE P
TE
Improving overall throughput requires considerations for all the steps in the sample analysis process including sample acquisition, sample preparation, LC-MS/MS analysis and data review and result reporting. Improvement of the overall throughput is a dynamic process, which focuses on the rate-limiting step. At different stage of automation, the rate-limiting step may be different. For instance, the overall throughput will increase as the LC-MS/MS throughput increases until the pre- and post- LC-MS/MS steps become the rate-limiting steps. The approaches that reduce manual processing time do not necessarily improve the overall throughput or to the same extent (Figure 4). For instance, liquid handler for sample preparation minimizes human intervention and increases human productivity, and yet it does not significant reduce the total sample processing time that is associated with the overall throughput. However, interface discussed above reduces human involvement and overall sample processing time to improve overall throughput. It is worth mentioning that the time needed to transfer the sample to the lab can take from a few hours to overnight or longer which is another challenge for the overall throughput and turnaround time. Most laboratories have limited impact on improving the efficiency of this step. 6. Other considerations for achieving total automation
ACCEPTED MANUSCRIPT Besides maximizing walk-away time and increasing throughput, total automation requires other considerations such as a broad test menu, the availability of commercial assay kits, standardization and results harmonization, and timely technical services.
PT
6.1 Commercial assay kits
MA
NU
SC
RI
Developing and validating MS assays is a major undertaking in a lab requiring a great deal of time and expertise. MS assays are laboratory developed tests (LDTs), which require the most stringent validation under Clinical Laboratory Improvement Amendments of 1988 (CLIA’88) regulation. Limited access to those who with this expertise has been a challenge to implement the MS platform in many clinical laboratories. Recently, Clinical & Laboratory Standards Institute (CLSI) issued a guideline (C62-A) on method development and validation. Though the guideline does not solve the lack of expertise reality, it provided the best practices and benchmark for method development and validation procedures and some key considerations.
AC CE P
TE
D
Commercially available assay kits will alleviates some of the challenges in developing and validating assays in the clinical labs. Recently the Food and Drug Administration (FDA) decided to exercise their enforcement discretion and extend their regulatory oversight to LDTs [33, 34]. As of the time of this writing, the FDA guideline has not released. However, it is likely that all current MS-based LDT assays will be subjected to this regulation. So far, only one FDA approved commercial assay kit is available on the market. This new regulation will significantly impact the vendors’ decision in moving forward with FDA approved assay kits to clinical laboratories. However, until FDA approved clinical assays are available for mass spectrometry; the regulation may hinder the adoption of this platform into routine clinical practices. 6.2 Standardization and results harmonization The limitations of both inter-laboratory variations of MS assays are well recognized, and in some but not all cases, these variations are even higher than immunoassays [35]. Different platforms may have different responses to the same analyte, which contributes to the variations among or even within vendors. Several approaches have the potential to enhance standardization. Use of a common calibrator has been reported to reduce the variation in results for vitamin D assays [36, 37], but not for other assays [38]. Inter-laboratory performance can be improved through the use of commercial kits, and yet the differences in performance of the same instruments still exist [39, 40]. While LC-MS/MS provides flexibility to develop individual assays with unique sample preparation and analytical parameters, it is key to ensure the harmonization and communicability of results through approaches such as accuracy based proficiency testing. 6.3 Cost and post-sale services
ACCEPTED MANUSCRIPT
RI
PT
High capital cost hinders the acceptance of the technology. However, the costs of the instrument are not likely to come down until the manufacturers reach the economy of scale for production. The economy of scale will rely on the automation steps discussed above. Therefore, it is reasonable to assume that reduced cost will be the result of automation rather than the driver of automation and popularization. Not until then will the manufacturer be able to afford this specialty instrument being sold as a routine platform.
NU
SC
A mass spectrometer is not a black box but rather a very complicated piece of machinery. It requires a sophisticated skill set and expertise to operate and troubleshoot the instrument. Current service contracts from manufacturers tend to be research-oriented and are not appropriate for routine clinical needs. To ensure smooth operation and minimize downtime in regular clinical services, short turnaround-time technical support and 24-hour technical support and short turnaround time service will be essential.
MA
7. Conclusions
AC CE P
TE
D
MS is an emerging technology that has made its way into clinical practices at an accelerating rate, particularly in the past couple of decades. It has been considered an instrument for esoteric tests and applied in specialty laboratories. Many challenges have hindered the implementation of the platform in many clinical laboratories. The field is waiting for the next generation of instruments, improvements in information technology, and vendor support to move mass spectrometry into the mainstream, i.e. into more widespread use by mid-size and smaller labs. In the near future, MS platforms may be differentiated into two categories. One category is the current existing system for esoteric and low volume laboratory developed tests. Another category is a basic model sold as a medical device for high volume and well-defined tests such as vitamin D, drugs of abuse screening and confirmation, and immunosuppressant drugs. The latter category may achieve higher level of automation and standardization for routine clinical usage. Higher level of automation enabling MS a robust, random access, high throughput platform is essential to make MS a technique of choice in routine practices. We believe three major milestones will mitigate those challenges. The first is to increase the walk-away time and reduce manual processing time, the second is to increase overall throughput, and the final step is to achieve total automation, including random access processing and reduced turnaround time. This will be enabled by automation of sample preparation and data analysis, integration with existing automated platforms, Food and Drug Administration clearance/approval, commercially available reagent kits and calibrators and seamless communication with existing electronic systems through bi-directional interfacing. Total automation requires generations of
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
effort and step-wise improvements. The future for MS in the clinical lab is a bright one, and automation will put wings on this giant elephant, enabling it to fly.
ACCEPTED MANUSCRIPT Acknowledgements
AC CE P
TE
D
MA
NU
SC
RI
PT
The authors thank Dr. Thomas Jackson for his discussion and the initial proofreading of the manuscript.
ACCEPTED MANUSCRIPT References
AC CE P
TE
D
MA
NU
SC
RI
PT
[1] Schulze A, Lindner M, Kohlmuller D, Olgemoller K, Mayatepek E, Hoffmann GF. Expanded newborn screening for inborn errors of metabolism by electrospray ionization-tandem mass spectrometry: results, outcome, and implications. Pediatrics 2003; 111:1399-1406. [2] Wilcken B, Wiley V, Hammond J, Carpenter K. Screening newborns for inborn errors of metabolism by tandem mass spectrometry. The New England journal of medicine 2003; 348:2304-2312. [3] Mizuno T, Abe N, Teshima H, et al. Application of a gas chromatography mass spectrometry computer system for clinical diagnosis. Biomedical mass spectrometry 1981; 8:593-597. [4] Marquet P. Progress of liquid chromatography-mass spectrometry in clinical and forensic toxicology. Therapeutic drug monitoring 2002; 24:255-276. [5] Maurer HH. Current role of liquid chromatography-mass spectrometry in clinical and forensic toxicology. Analytical and bioanalytical chemistry 2007; 388:1315-1325. [6] Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization for mass spectrometry of large biomolecules. Science (New York, NY) 1989; 246:64-71. [7] Kushnir MM, Rockwood AL, Bergquist J. Liquid chromatography-tandem mass spectrometry applications in endocrinology. Mass spectrometry reviews 2010; 29:480-502. [8] Streit F, Armstrong VW, Oellerich M. Rapid liquid chromatography-tandem mass spectrometry routine method for simultaneous determination of sirolimus, everolimus, tacrolimus, and cyclosporin A in whole blood. Clinical chemistry 2002; 48:955-958. [9] Clarke NJ, Zhang Y, Reitz RE. A novel mass spectrometry-based assay for the accurate measurement of thyroglobulin from patient samples containing antithyroglobulin autoantibodies. Journal of investigative medicine : the official publication of the American Federation for Clinical Research 2012; 60:1157-1163. [10] Hoofnagle AN, Roth MY. Clinical review: improving the measurement of serum thyroglobulin with mass spectrometry. The Journal of clinical endocrinology and metabolism 2013; 98:1343-1352. [11] Kushnir MM, Rockwood AL, Roberts WL, Abraham D, Hoofnagle AN, Meikle AW. Measurement of thyroglobulin by liquid chromatography-tandem mass spectrometry in serum and plasma in the presence of antithyroglobulin autoantibodies. Clinical chemistry 2013; 59:982-990. [12] Netzel BC, Grebe SK, Algeciras-Schimnich A. Usefulness of a thyroglobulin liquid chromatography-tandem mass spectrometry assay for evaluation of suspected heterophile interference. Clinical chemistry 2014; 60:1016-1018. [13] Claydon MA, Davey SN, Edwards-Jones V, Gordon DB. The rapid identification of intact microorganisms using mass spectrometry. Nature biotechnology 1996; 14:1584-1586. [14] Anderson L, Hunter CL. Quantitative mass spectrometric multiple reaction monitoring assays for major plasma proteins. Molecular & cellular proteomics : MCP 2006; 5:573-588. [15] Vogeser M, Seger C. A decade of HPLC-MS/MS in the routine clinical laboratory--goals for further developments. Clinical biochemistry 2008; 41:649-662. [16] Hetu PO, Robitaille R, Vinet B. Successful and cost-efficient replacement of immunoassays by tandem mass spectrometry for the quantification of immunosuppressants in the clinical laboratory. Journal of chromatography B, Analytical technologies in the biomedical and life sciences 2012; 883884:95-101. [17] Clarke W, Rhea JM, Molinaro R. Challenges in implementing clinical liquid chromatographytandem mass spectrometry methods--the light at the end of the tunnel. Journal of mass spectrometry : JMS 2013; 48:755-767. [18] Honour JW. Development and validation of a quantitative assay based on tandem mass spectrometry. Annals of clinical biochemistry 2011; 48:97-111.
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
[19] Grebe SK, Singh RJ. LC-MS/MS in the Clinical Laboratory - Where to From Here? The Clinical biochemist Reviews / Australian Association of Clinical Biochemists 2011; 32:5-31. [20] Vogeser M, Kirchhoff F. Progress in automation of LC-MS in laboratory medicine. Clinical biochemistry 2011; 44:4-13. [21] Adaway JE, Keevil BG, Owen LJ. Liquid chromatography tandem mass spectrometry in the clinical laboratory. Annals of clinical biochemistry 2015; 52:18-38. [22] van den Ouweland JM, Kema IP. The role of liquid chromatography-tandem mass spectrometry in the clinical laboratory. Journal of chromatography B, Analytical technologies in the biomedical and life sciences 2012; 883-884:18-32. [23] Panteghini M. The future of laboratory medicine: understanding the new pressures. The Clinical biochemist Reviews / Australian Association of Clinical Biochemists 2004; 25:207-215. [24] Robandt PP, Reda LJ, Klette KL. Complete automation of solid-phase extraction with subsequent liquid chromatography-tandem mass spectrometry for the quantification of benzoylecgonine, mhydroxybenzoylecgonine, p-hydroxybenzoylecgonine, and norbenzoylecgonine in urine--application to a high-throughput urine analysis laboratory. Journal of analytical toxicology 2008; 32:577-585. [25] Poon EG, Keohane CA, Yoon CS, et al. Effect of bar-code technology on the safety of medication administration. The New England journal of medicine 2010; 362:1698-1707. [26] Snyder ML, Carter A, Jenkins K, Fantz CR. Patient misidentifications caused by errors in standard bar code technology. Clinical chemistry 2010; 56:1554-1560. [27] Hawker CD. Bar codes may have poorer error rates than commonly believed. Clinical chemistry 2010; 56:1513-1514. [28] Dickerson JA, Schmeling M, Hoofnagle AN, Hoffman NG. Design and implementation of software for automated quality control and data analysis for a complex LC/MS/MS assay for urine opiates and metabolites. Clinica chimica acta; international journal of clinical chemistry 2013; 415:290-294. [29] Wu JT. The development of a staggered parallel separation liquid chromatography/tandem mass spectrometry system with on-line extraction for high-throughout screening of drug candidates in biological fluids. Rapid communications in mass spectrometry : RCM 2001; 15:73-81. [30] Xu RN, Fan L, Rieser MJ, El-Shourbagy TA. Recent advances in high-throughput quantitative bioanalysis by LC-MS/MS. Journal of pharmaceutical and biomedical analysis 2007; 44:342-355. [31] Netzel BC, Cradic KW, Bro ET, et al. Increasing liquid chromatography-tandem mass spectrometry throughput by mass tagging: a sample-multiplexed high-throughput assay for 25hydroxyvitamin D2 and D3. Clinical chemistry 2011; 57:431-440. [32] Grote-Koska D, Czajkowski S, Brand K. Performance of the new RapidFire(R) system for therapeutic monitoring of immunosuppressants. Therapeutic drug monitoring 2014. [33] Gutman S. The role of Food and Drug Administration regulation of in vitro diagnostic devices-applications to genetics testing. Clinical chemistry 1999; 45:746-749. [34] Scott MG, Ashwood ER, Annesley TM, Leonard DG, Burgess MC. FDA oversight of laboratorydeveloped tests: is it necessary, and how would it impact clinical laboratories? Clinical chemistry 2013; 59:1017-1022. [35] Vogeser M, Seger C. Pitfalls associated with the use of liquid chromatography-tandem mass spectrometry in the clinical laboratory. Clinical chemistry 2010; 56:1234-1244. [36] Yates AM, Bowron A, Calton L, et al. Interlaboratory variation in 25-hydroxyvitamin D2 and 25hydroxyvitamin D3 is significantly improved if common calibration material is used. Clinical chemistry 2008; 54:2082-2084. [37] Carter GD. 25-Hydroxyvitamin D assays: the quest for accuracy. Clinical chemistry 2009; 55:1300-1302. [38] Levine DM, Maine GT, Armbruster DA, et al. The need for standardization of tacrolimus assays. Clinical chemistry 2011; 57:1739-1747.
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
[39] Napoli KL, Hammett-Stabler C, Taylor PJ, et al. Multi-center evaluation of a commercial Kit for tacrolimus determination by LC/MS/MS. Clinical biochemistry 2010; 43:910-920. [40] Carter GD, Jones JC. Use of a common standard improves the performance of liquid chromatography-tandem mass spectrometry methods for serum 25-hydroxyvitamin-D. Annals of clinical biochemistry 2009; 46:79-81.
ACCEPTED MANUSCRIPT Figure legends Figure 1. Current LC-MS/MS platform workflow and typical time requirement for each step.
RI
PT
Sample receiving includes sample identification, centrifugation, de-capping, aliquoting and relabeling steps. Time estimate is based on one batch sample for one 96 well plate. Blank box: automated process; dotted line: semi-automated process; dots: manual process
SC
Figure 2. Possible automated LC-MS/MS workflows.
NU
Panel A. Automated LC-MS/MS as a standalone analyzer and the key features and potential solutions for automation. Panel B. LC-MS/MS as part of the automated clinical laboratory system and samples are processed and transferred through the conveyor belt to the designated analyzers.
MA
Figure 3. Interim steps to take to achieve total automation for LC-MS/MS platform and the key features for each step.
TE
D
Figure 4. Impact of automation approaches on manual and total processing time for various steps in the LC-MS/MS workflow.
AC CE P
Major steps are labeled on the y-axis. As MS/MS analysis takes no manual time and very little processing time, it is not indicated in the diagram. Blank box, manual processing time; gray box, total processing time. Time required for each step is labeled in relative scale as the current time requirement being the full scale (indicated at progression point 0). Half size box, significant total processing time required; boxes with diagonal line, eliminate nearly all manual or total processing time. Numbers indicated various approaches. 1. LC-multiplexing and online or offline LC separation approaches; 2. Automated sample preparation approaches such as liquid sampler; 3. Interface connection for results reporting; 4. Automated data review software; 5. Total automation for sample receiving step.
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights
TE
D
MA
NU
SC
RI
PT
LC-MS/MS becomes an important analytical platform in clinical practices. Achieving total automation is key to bring LC-MS/MS into routine clinical labs. Incremental improvements are essential for LC-MS/MS to achieve total automation. Automation includes sample acquisition, preparation, data review and reporting. Interfacing with LIS will gain immediate benefits in reducing labor intensity.
AC CE P
1. 2. 3. 4. 5.