Printer Friendly

Historical response factor-based quantification for LC-MS/MS.

To the Editor:

With great interest we read the article by Olson et al. (1), who thoroughly compared sporadic response factor (RF) [1] calibration to calibration in each run (CR) for the therapeutic drug monitoring of nortriptyline with liquid chromatography-tandem mass spectrometry (LC-MS/MS). They found the results obtained for the 2 quantification methods to be analytically and clinically commensurate. These findings could lead to substantial reductions in costs, workload, and turnaround time.

The savings for sporadic RF quantification are most useful for low-throughput laboratories that run different low-volume tests on the same instrument. This situation is often encountered, especially in Europe, where centralized high-throughput mass spectrometry centers are the exception rather than the rule. As Olson et al. stated, their approach permitted the use of sporadic RF calibration in only 8 of 16 assay run days over a 2-month period, thus limiting the benefit.

We presently have >2 years of experience in clinical laboratory practice with LC-MS/MS quantification (Alliance HPLC 2795 separations module coupled to a Quattro Micro mass spectrometer, both from Waters Corporation) based on sporadic RF. Instead of the criteria used by Olson et al. for applying sporadic RP (1), we decided to make an even greater leap. We derived the sporadic RP from our initial validation experiment and evaluated its use during a 6-month period. During this time, we ran other tests on the same instrument, carried out normal routine maintenance, and even had the manufacturer perform an extended maintenance evaluation. Our results for the quantification of the antifungal agent voriconazole in plasma were published in 2012 (2). Our findings indicated that results obtained with the sporadic RF [which can be called a historical RF (hRF) after a use of >6 months without change] were comparable with the CR for voriconazole, further supporting and expanding the results of Olson et al.

After our successful experience with voriconazole, we decided to use the same approach for the implementation of an LC-MS/MS quantification method for the antiepileptic lamotrigine and the immunosuppressant mycophenolate in plasma. The results for lamotrigine and mycophenolate were in line with those obtained for voriconazole. The hRF quantification and CR values were comparable (Table 1).

In our experience, however, there is a price to pay for the gains in cost, workload, and turnaround time. The CV increased with the hRF compared with the CR (Table 1), a finding not examined in detail by Olson et al. In our opinion, this difference in CVs hampers the use of a hRF or a sporadic RF for analytes requiring accurate quantification at concentrations near the limits of instrument performance, because reliable quantification requires a CV of <20%. Furthermore, some analytes may require a very stringent analytical imprecision (3). Finally, the interpretation of sequential results may become more difficult owing to larger analytical variation. We advise clinical laboratories to prudently evaluate the impact of hRF-based (or sporadic RF-based) quantification on the required limit of quantification and quality requirements of their method. For some analytes, a stringent QC protocol might be needed if the hRF is used.

In conclusion, we would like to endorse and extend the findings of Olson and colleagues in this bigger step toward the dream of a random-access LC-MS/MS platform that uses hRFs. We nevertheless realize this "big step" is relatively small given the above-mentioned remarks on increased analytical variation and the reality of most small- to middle-sized laboratories running different LC-MS/MS methods on the same instrument.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.

Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.

References

(1.) Olson MT, Breaud A, Harlan R, Emezienna N, Schools S, Yergey At, Clarke W. Alternative calibration strategies tor the clinical laboratory: application to nortriptyline therapeutic drug monitoring. Clin Chem 2013;59:920-7.

(2.) Pauwels S, Vermeersch P, Van Eldere J, Desmet K. Fast and simple LC-MS/MS method for quantifying plasma voriconazole. Clin Chim Acta 2012;413:740-3.

(3.) Fraser CG. Desirable standards of performance for therapeutic drug monitoring. Clin Chem 1987;33:387-9.

Steven Pauwels [2,3] *

Koen Poesen [2,4]

Johan Van Eldere [2,5]

Koen Desmet [2,3]

Pieter Vermeersch [2,3]

[1] Nonstandard abbreviations: RF, response factor; CR, calibration in each run; LS-MS/MS, liquid chromatography-tandem mass spectrometry; hRF, historical RF

[2] Clinical Department of Laboratory Medicine University Hospitals Leuven Leuven, Belgium Departments of

[3] Cardiovascular Sciences,

[4] Neurology, and

[5] Microbiology and Immunology KU Leuven Leuven, Belgium

* Address correspondence to this author at: Laboratory Medicine University Hospitals Leuven Herestraat 49 B-3000 Leuven, Belgium Fax16-34-79-31 E-mail steven.pauwels@uzleuven.be

Previously published online at DOI: 10.1373/clinchem.2013.212761

In Reply

We are grateful to Pauwels et al. for their favorable commentary on our description of calibration based on the response factor (RF) [1] (1), and we agree with them that the cost reductions can be substantial, especially for low-volume tests. We also agree that the ideal clinical liquid chromatography-tandem mass spectrometry platform would involve random-access assays and that a substantial step toward this goal would be calibration based on the RF, rather than on the calibration curve. As we have mentioned, RF-based calibration strategies are certainly not new. In communications that have been published after the publication of our report, including this exchange, we are pleased to have encountered other investigators, including Pauwels et al. (2) and Taylor et al. (3), who have also shown excellent performance with RF-based calibration in clinical assays. Our oversight of these important works in no way indicates our underestimation of their importance. On the contrary, we are pleased that we now know of 2 independent research groups that have corroborated the basic premises and findings of our work.

As a related comment, the oversight of 2 articles on the topic of RF-based calibration--despite a familiarity with the literature, a reasonably thorough literature search to find corroborating studies, conversations with numerous experienced colleagues in our field about our work, and a rigorous review process--underscores the need for our community to adopt a standardized terminology for RF-based calibration. For reasons detailed in our report (1), we suggest that "sporadic RF" (sRF) calibration be used universally to designate any RF-based calibration strategy that does not determine the RF routinely with each assay batch. Similarly, we suggest that "contemporaneous RF" (cRF) calibration be used to indicate any RF-based calibration strategy that does determine the RF routinely with each assay batch. Although the "historical" RF nomenclature was fitting in the work described by Pauwels et al. (2), they readily acknowledge that stringent QC procedures would be necessary if this calibration strategy were to be applied in routine practice. Given that QC failures are caused by a variety of sources of variance, both systematic and random, they are by definition unpredictable at times. Our hope is that such failures would lead to reevaluating the RF (sporadically), so we feel sRF is the most inclusive term for all calibration strategies of this nature.

As with any assay in the clinical laboratory, the ideal for guiding assay design should be "fit-for-purpose," and this ideal applies to evaluation of the acceptable variance as well as to the analytical platform. The study described in our report demonstrated that RF-based calibration methods are indeed fit for the purpose of nortriptyline therapeutic drug monitoring by their basing of the choice of internal-standard concentration on the clinical decision points of the analyte. We also compared discrete cases in which different calibration schemes led to potentially different clinically actionable values; however, we noted that these cases would have been close to the decision point in any calibration scheme. Beyond demonstrations of statistical interchangeability and the standard performance features, we feel that any assay changes should be accompanied by a detailed and rigorous analysis of the actual effects on patients (4, 5). Our published study remains the only one to have evaluated the data on RF-based calibration in this way, and we hope that future publications on the topic of assay improvements will include this straightforward, patient-centered analysis with standard analytical-validation statistics.

As our astute colleagues have pointed out, our experimental design allowed the sRF to be used on only half of the days that the nortriptyline assay was run. This design did indeed limit the power for comparing sRF and cRF assay performance, a point we were keen to emphasize at the end of our discussion. The experience of Pauwels et al. involved a longer time period and more analytes, and their sRF approach had a higher CV. This result should not surprise anyone, given that the multiple factors causing the variance in the RF are poorly understood. It was indeed clear to us that the RF was not particularly stable; therefore, we do not plan to use sRF calibration without restrictive QC protocols.

In conclusion, we are pleased with the growing interest in RF-based calibration strategies, and we look forward to more investigators taking a critical look at this exciting new twist on old calibration.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation ofdata; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.

Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:

Employment or Leadership: None declared.

Consultant or Advisory Role: W. Clarke, Thermo Fisher Scientific.

Stock Ownership: None declared.

Honoraria: W. Clarke, Thermo Fisher Scientific.

Research Funding: W. Clarke, Thermo Fisher Scientific.

Expert Testimony: None declared.

Patents: None declared.

References

(1.) Olson MT, Breaud A, Harlan R, Emezienna N, Schools S, Yergey AL, Clarke W. Alternative calibration strategies for the clinical laboratory: application to nortriptyline therapeutic drug monitoring. Clin Chem 2013;59:920-7.

(2.) Pauwels S, Vermeersch P, Van Eldere J, Desmet K. Fast and simple LC-MS/MS method for quantifying plasma voriconazole. Clin Chim Acta 2012;413: 740-3.

(3.) Taylor PJ, Forrest KK, Salm P, Pillans PI. Single-point calibration for sirolimus quantification. Ther Drug Monit 2001;23:726-7.

(4.) Olson MT, Kickler TS, Lawson JA, McLean RC, Jani J, FitzGerald GA, Rade JJ. Effect of assay specificity on the association of urine 11-dehydro thromboxane B2 determination with cardiovascular risk. J Thromb Haemost 2012;10:2462-9.

(5.) Olson MT, Lombardi L, Clarke W. Clinical consequences of analytical variance and calculation strategy in oral busulfan pharmacokinetics. Clin Chim Acta 2011;412:2316-21.

Matthew T. Olson [2] William Clarke [2] *

[1] Nonstandard abbreviations: RF, response factor; sRF, sporadic RF; cRF, contemporaneous RF.

[2] Department of Pathology The Johns Hopkins University School of Medicine Baltimore, MD

* Address correspondence to this author at: Sheikh Zayed Tower B-1020F 1800 Orleans St. Baltimore, MD 21287 Fax 410-614-7609 E-mail wclarke@ jhmi.edu

Previously published online at DOI: 10.1373/clinchem.2013.213637
Table 1. CVs for QCs and Deming regression (with CR as reference
method) for the quantification of voriconazole, lamotrigine, and
mycophenolate. (a)

                       Voriconazole, mg/L       Lamotrigine, mg/L
                            (CV) (b)                  (CV)

QC

  n                            53                      35

  Low


    CR                     2.0 (3.3%)              2.9 (4.8%)
    hRF                    2.0 (6.3%)               2.7(13%)

  Medium

    CR
    hRF

  High

    CR                     5.8 (4.3%)              14.1 (4.1%)
    hRF                    6.0 (6.3%)               14.0(10%)

Deming regression

  n                           227                      172
  Slope (95% Cl)        1.07 (1.05-1.09)        0.98 (0.94-1.03)
  Intercept          -0.06 (-0.10 to -0.01)   -0.05 (-0.25 to 0.15)
    (95% Cl), mg/L

                        Mycophenolate,
                          mg/L (CV)

QC

  n                          104

  Low

    CR                    1.0 (7.6%)
    hRF                    1.0(10%)

  Medium

    CR                    7.2 (6.8%)
    hRF                   7.3 (13%)

  High

    CR                   11.5 (7.8%)
    hRF                   11.7 (11%)

Deming regression

  n                          1071
  Slope (95% Cl)       0.98 (0.92-1.04)
  Intercept          0.05 (-0.07 to 0.17)
    (95% Cl), mg/L

(a) Quantifications were carried out both with calibration used
for each run (CR) and with hRF-based quantification evaluated
over a 6-month period. QC values were quantified every assay run
day. QC data are expressed as mean concentrations.

(b) Voriconazole results are taken from Pauwels et al. (2).
COPYRIGHT 2013 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Letters to the Editor
Author:Pauwels, Steven; Van Eldere, Koen Poesen Johan; Vermeersch, Koen Desmet Pieter
Publication:Clinical Chemistry
Article Type:Letter to the editor
Date:Nov 1, 2013
Words:2084
Previous Article:Point-of-care and over-the-counter qualitative human chorionic gonadotropin (hCG) devices remain susceptible to false-negative results caused by...
Next Article:Abdominal pain and increased CA19-9.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters