• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Medical Devices

  • Print
  • Share
  • E-mail

Draft Guidance for Industry and Food and Drug Administration Staff - Recommendations for Premarket Notifications for Lamotrigine and Zonisamide Assays

PDF Printer VersionDRAFT GUIDANCE

This guidance document is being distributed for comment purposes only.
Document issued on: August 6, 2010

You should submit comments and suggestions regarding this draft document within 90 days of publication in the Federal Register of the notice announcing the availability of the draft guidance. Submit written comments to the Division of Dockets Management (HFA-305), Food and Drug Administration, 5630 Fishers Lane, rm. 1061, Rockville, MD 20852. Submit electronic comments to http://www.regulations.gov. Identify all comments with the docket number listed in the notice of availability that publishes in the Federal Register.

For questions regarding this document, contact Avis Danishefsky at 301-796-6142 or by email at avis.danishefsky@fda.hhs.gov.

CDRH Logo

U.S. Department of Health and Human Services
Food and Drug Administration
Center for Devices and Radiological Health
Division of Chemistry and Toxicology Devices
Office of In Vitro Diagnostics Evaluation and Safety

Preface

Additional Copies

Additional copies are available from the Internet. You may also send an e-mail request to CDRH-Guidance@fda.hhs.gov to receive a copy of the guidance. Please use the document number 1654 to identify the guidance you are requesting.  


Table of Contents

  1. INTRODUCTION
  2. BACKGROUND - PREMARKET NOTIFICATIONS
  3. ASSAYS WITHIN THE SCOPE OF THIS GUIDANCE
  4. BACKGROUND LAMOTRIGINE AND ZONISAMIDE MEASUREMENTS
  5. DEVICE DESCRIPTION
  6. PERFORMANCE CHARACTERISTICS
    • 6.A General Study Recommendations
    • 6.B Precision
    • 6.C Trueness/Bias
    • 6.D Linearity
    • 6.E Lower Limits of the Assay
    • 6.F Endogenous and Exogenous Interference
    • 6.G Matrix Comparison
    • 6.H Specimen Storage Recommendations
    • 6.I Method Comparison
    • 6.J Calibrators and Control Materials
    • 6.K Studies at External Sites
  7. LABELING

Draft Guidance for Industry and FDA Staff - Recommendations for Premarket Notifications for Lamotrigine and Zonisamide Assays

This draft guidance, when finalized, will represent the Food and Drug Administration's (FDA's) current thinking on this topic. It does not create or confer any rights for or on any person and does not operate to bind FDA or the public. You can use an alternative approach if the approach satisfies the requirements of the applicable statutes and regulations. If you want to discuss an alternative approach, contact the FDA staff responsible for implementing this guidance. If you cannot identify the appropriate FDA staff, call the number listed on the title page of this guidance.

1. Introduction

This guidance document contains recommendations to manufacturers and FDA reviewers concerning information to include in premarket notifications for assays for the anti-seizure drugs, lamotrigine and zonisamide. These assays are intended to quantitatively measure concentrations of the respective drugs in serum or plasma as an aid in the management of patients treated with lamotrigine or zonisamide. The recommendations are based on current review practices as well as information regarding zonisamide and lamotrigine that were submitted to FDA by the Therapeutic Drug Monitoring Roundtable1. Some of the general concepts in this guidance may also be helpful in preparing 510(k) submissions for other therapeutic drug assays previously cleared by FDA, and classified within 21 CFR 862 subpart D.2,3

FDA's guidance documents, including this guidance, do not establish legally enforceable responsibilities. Instead, guidances describe the Agency's current thinking on a topic and should be viewed only as recommendations, unless specific regulatory or statutory requirements are cited. The use of the word should in Agency guidances means that something is suggested or recommended, but not required.

2. Background – Premarket Notifications

A manufacturer who intends to market a device of this generic type must conform to the general controls of the Federal Food, Drug, and Cosmetic Act (the act), including the premarket notification requirements and obtaining a substantial equivalence determination from FDA prior to marketing the device (see 21 CFR 807.81 and 807.87).

This document supplements other FDA documents regarding the specific content requirements of a premarket notification submission. You should also refer to 21 CFR 807.87. Other sources of information include the guidance document “Format for Traditional and Abbreviated 510(k)s,” the section of CDRH’s Device Advice web page entitled “Premarket Notification 510(k),” and the guidance document “The New 510(k) Paradigm – Alternate Approaches to Demonstrating Substantial Equivalence in Premarket Notifications; Final Guidance.

3. Assays Within the Scope of this Guidance

Lamotrigine and zonisamide assays are regulated under 21 CFR 862.3350, Diphenylhydantoin test system, Class II. The product codes for these devices are:

ORH for lamotrigine assays, and
NWM for zonisamide assays.

This guidance addresses lamotrigine or zonisamide assays that employ immunoassay technologies; are for use in central clinical laboratories; and are for use with serum or plasma samples. For assays that do not have these features, different studies or additional studies may be appropriate3.

4. Background - Lamotrigine and Zonisamide Assays

Assays for lamotrigine or zonisamide are indicated as an aid in management of individuals treated with these anti-seizure drugs. For example, the assays might be useful for purposes such as helping to diagnose or monitor drug overdose; managing individual patients who might experience one or more drug-drug interactions (DDIs) that can alter the pharmacokinetics of these drugs resulting in changes in drug levels; and providing a relative reference value for an individual after achieving a steady concentration at a specific daily dose judged to be either effective or ineffective4.

Caution should always be exercised regarding individual patient management based upon observed reference ranges for zonisamide or lamotrigine published in the literature to date. There is a lack of documented correlation between serum concentrations and therapeutic effects or toxicity, and overlap has been observed in serum concentrations between responders and non-responders as well as between levels associated with seizure control and adverse reaction. Randomized studies investigating the usefulness of routine monitoring of anti-seizure drugs are few, and evidence regarding the usefulness of routine monitoring is lacking. Nonetheless, therapeutic drug monitoring for these drugs may be helpful, especially in specific situations (as noted above).5

Assay results should always be interpreted in combination with careful observation of clinical response and with an understanding of the limitations of any specified reference ranges. Discussion of these types of issues should be included in the labeling for zonisamide or lamotrigine assays. (See Section 7 - Labeling.)

Assays for lamotrigine or zonisamide are expected to adequately measure the concentration of the specified drug with defined performance characteristics. A false result (falsely high or falsely low) or misinterpretation of results may contribute to inappropriate patient dosing, which in turn could lead to either toxicity or lack of seizure control.

5. Device Description

You should include descriptions of the following device features in your 510(k):

  • All reagents (analytical and pre-analytical), including controls and calibrators marketed with the assay.
  • Assay technology, including specific reactions, and the method of detection.
  • Instrumentation, such as specific analyzers, needed to run the assay.

If your device includes software specific to your assay, see “Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices,” regarding software information that you should include in your 510(k).

6. Performance Characteristics

6.A General Study Recommendations

You should perform the evaluations described in this guidance in a manner consistent with the specific procedures that you recommend to users in the labeling, in order to reflect performance to be expected by the user. For example, samples should be stored as you recommend in the package insert. Quality control measures should be those you plan to recommend to assay users.

So results can be correctly interpreted during review of your 510(k), you should provide clear descriptions of the study designs. In addition, when referring to Clinical and Laboratory Standards Institute (CLSI) guidelines, we recommend you indicate which specific aspects of the guidelines you followed and which you modified (if any). Similarly, we recommend that you provide a summary of this type of information in the device labeling, so device users can correctly interpret results.

When describing studies using spiked or other prepared samples, you should summarize how the samples were prepared, including materials used for spiking or diluting. For studies using patient samples, you should clarify the source of the samples and any sample selection criteria applied.

For analytical evaluations involving estimates of assay bias (e.g., recovery, interference, linearity, matrix evaluations) we recommend you statistically determine the appropriate number of replicates to include at each concentration, based on considerations of factors such as assay repeatability and your acceptance criteria for assay bias.

6.B Precision

You should characterize precision of your test system. We recommend that you follow the guideline “Evaluation of Precision Performance of Quantitative Measurement Methods;” CLSI Document EP5-A2, regarding statistical approach and statement of claims. We recommend you perform this evaluation over a 20 day period to evaluate within-run, between-run, between-day, and any other conditions relevant to your assay.

We recommend that you evaluate precision for at least three drug concentrations. The concentrations should be chosen to span your claimed assay reportable range (i.e., include a high and a low sample) and to include potential medical decision point(s). You should perform the precision evaluations using samples in the intended matrix. We recommend that you include stable samples or pools from patients treated with zonisamide or lamotrigine within the evaluation. Samples diluted or spiked into drug-free serum or plasma are appropriate to supplement the evaluation with patient samples. We recommend that you also run control materials within this evaluation. When feasible, we recommend you also include evaluation of multiple reagent lots measuring the same sample(s).

We recommend that you include the following in the description of your precision evaluation in the 510(k):

  • Description of samples tested, including preparation or source of samples, matrix, and target concentrations.
  • Site or sites at which the precision protocol was run (e.g., manufacturer’s site).
  • Number of days, runs, and observations.
  • Identification of other factors that were held constant and those that were varied during the evaluation (e.g., number of instruments, calibration cycles, reagent lots, and operators).
  • Description of computational methods used (including equations, if they were different or modified from those described in CLSI EP5-A2). All results should be included in the calculations. If you deleted any outliers from your calculations, you should state this and include your justification, for review.
  • Results, including mean, coefficients of variation (CV), and standard deviations (SD).

6.C Trueness/Bias

You should evaluate bias of your zonisamide or lamotrigine assay in the intended matrix. Typically, this involves recovery studies in which known amounts of a reference material6 (whose concentration is known independently of your assay) are spiked into drug-free clinical samples or pools and the samples are measured using the new assay. Sample matrices used in this evaluation should mimic patient samples as closely as possible. Therefore we recommend spiking into drug-free patient serum or plasma. (We do not recommend spiking the drug into only calibrator or control material.) We recommend that you evaluate recovery for replicate samples at a minimum of three concentrations spanning the reportable range.

In cases in which the 510(k) submission includes calibrators, we recommend spiking the same reference materials similarly into both calibrator and serum samples in order to demonstrate whether there is any matrix effect for the calibrator material relative to clinical samples.

Some immunoassays may exhibit a "high dose hook effect" in which there is a decrease in response of the assay at high concentrations. Whenever appropriate (e.g., for two-site or sandwich immunoassays), you should extend the studies beyond the reportable range to the highest concentrations that might possibly be encountered in clinical settings in order to evaluate whether the device exhibits a high dose hook effect.

We recommend that you include the following in the description of your recovery evaluation:

  • Matrix, and source or preparation of the samples and materials used for spiking.
  • Target/expected concentrations of the samples and the methods by which these were determined, independently of the new assay.
  • Number of replicates measured, and calculations or statistical methods used.
  • Results of the new assay measurements relative to target (expected) results. Results should be represented separately for each concentration level tested.

6.D Linearity

You should characterize linearity of your assay by evaluating samples, with concentrations distributed across the entire reportable range. The preferred samples to use, in order to mimic the intended sample matrix as closely as possible, are patient samples or pools containing elevated drug concentrations diluted with drug-free patient samples. In cases where unaltered patient specimens or pools are not available at concentrations at or above the high end of the assay range, we recommend spiking the highest available patient sample with a stock solution of the drug. The guideline “Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach”, CLSI document EP6-A, recommends evaluating 7-11 concentrations. We recommend that you prepare these samples to be evenly distributed across the claimed measuring range.7 It may be possible to perform your linearity and recovery evaluations together as one study.

We recommend that you include the following in the description of your linearity evaluation:

  • Sample matrix, and method of preparation or origin.
  • Number of samples and replicates evaluated.
  • Calculations and statistical methods used.
  • Results, including the known sample concentrations, and the concentrations measured with your assay for each sample.
  • Results of linear regression analysis.

If you are following the statistical approach recommended in CLSI guideline “Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach”; CLSI document EP6-A, you should include the statement of claims and the analyses recommended in that guideline.

If you recommend to users that samples outside the range can be diluted, you should describe your validation of the procedure. Samples in the intended matrix spiked to high concentrations can be used for this purpose. The (high) target/expected concentration should be determined independently of the new assay.

6.E Lower Limits of the Assay

You should evaluate assay performance at the claimed lower limits of your assay. We recommend you follow “Protocols for Determination of Limits of Detection and Limits of Quantitation,” CLSI document EP17-A. As defined in that document, the limit of quantitation (LoQ) is the lowest amount of analyte in a sample that can be quantitatively determined with stated acceptable precision and trueness, under stated experimental conditions (modified from WHO-BS/95.1793). The limit of detection (LoD) is the lowest amount of analyte in a sample that can be detected with (stated) probability, although perhaps not quantified as an exact value. The target concentrations of the samples should be known by a reliable method independent of the assay being evaluated, so that bias (as well as precision) can be determined. We recommend that you perform the study over a minimum of five days and measure multiple replicates per day for each sample. We recommend you vary other parameters as well, such as instruments or lots, to help ensure the estimates reflect the lower limits to be expected once the test is marketed. CLSI guideline EP17-A recommends a minimum of 40 total replicates for each sample.

In your 510(k), you should describe your study design for determining the lower limits of the assay, including the types of information listed above in the bullets in Section 6B- Precision and Section 6C-Trueness/Bias, above. You should describe results for both precision and accuracy at this lower limit.

LoQ evaluations are especially relevant in cases where clinical decisions could potentially be based on values close to the lower limits of the assay range. When considering clinical decision points, it is important to keep in mind that lower limits for clinical decisions may vary depending on co-administered drugs and other clinical factors.

6.F Endogenous and Exogenous Interference

You should characterize the effects of potential interferents on assay performance. We recommend that you follow the guideline “Interference Testing in Clinical Chemistry; Approved Guideline,” CLSI Document EP7-A2, regarding types of interferents to consider for testing and study designs, including specified test concentrations (of analyte and interferent) and statistical considerations.8 Studies should be designed so that both positive and negative bias may be detected. We recommend you determine the appropriate number of replicates so that testing is performed with sufficient power to detect any clinically significant interference and with a sufficient confidence level to recognize when there is no significant interference.

For all interference testing, we recommend you describe the following in your 510(k):

  • Potential interferents and concentrations at which they were tested.
  • The concentrations of lamotrigine or zonisamide in the samples and how samples were prepared (e.g., by spiking drug-free serum sample pools).
  • Definitions and/or calculations for computing interference (and cross-reactivity where applicable).
  • Results of test and control sample measurements and calculated interference (or cross-reactivity as relevant).

Common examples of compounds to test are listed below. In addition, manufacturers should carefully consider other substances that may potentially interfere based on knowledge of the chemistry of the procedure, the indications for use, and structure of the analyte. Structural databases may be helpful for consideration of the latter.

Examples of endogenous compounds:
  • Bilirubin
  • Triglycerides
  • Cholesterol
  • Uric acid
  • Rheumatoid factor
  • Albumin
  • Gamma globulin
  • Human anti-mouse antibodies, HAMA
  • Hemoglobin
Examples of types of commonly co-administered drugs:
  • All antiepileptic drugs (and relevant metabolites when available).
  • Available antipsychotic and antidepressant drugs.
  • Common tranquilizers and hypnotics.
  • Commonly prescribed antibiotics.
  • Common over-the-counter drugs.

In accordance with CLSI guideline EP7-A2, we recommend that you test up to levels at least three times that of the highest acute peak concentration reported following therapeutic dosage.

Specificity for parent compound

You should characterize cross-reactivity with drug metabolites. For lamotrigine assays, primary known metabolites you should test include N-2 glucuronide, N-5 glucuronide, N-2 methyl metabolite, and N-2 oxide. For zonisamide assays, metabolites to test include N-acetyl zonisamide (NAZ) and 2-sulfamoylacetyl phenol (SMAP). The highest metabolite concentration should challenge the assay; the concentration should be at least three times higher than the peak metabolite concentrations found in patient samples, including samples from patients with conditions that may cause metabolite accumulation. We recommend that you include samples negative for parent drug, as well as samples containing parent drug concentrations near the potential medical decision point(s).

6.G Matrix Comparison

You should evaluate performance of your assay with all serum and plasma matrices that you recommend for use with your assay, including specified anticoagulants or types of collection tubes you recommend. As noted in Section 6I – Method Comparison, below, you should identify the matrix for samples used in the method comparison study. If you recommend use of your assay with additional matrices, anticoagulants or collection tubes, which were not fully evaluated in your method comparison study, you should submit matrix comparison evaluations. In any case, you should check that the collection tubes are FDA cleared for TDM use.

You should compare results of patient samples in the test matrix to results of patient samples with the same drug concentration in the control matrix. (The control matrix in this case is the one that is fully validated in the method comparison study.) We recommend that you compare a minimum of 15-20 samples at multiple concentrations spanning the assay range. In many cases, it is acceptable to test drug-free patient samples or pools that were collected in the appropriate collection tube type and then spiked with zonisamide or lamotrigine. This approach of spiking samples is acceptable in cases in which (1) the predicate assay (with the same technology and intended use) has been cleared for use with the matrix or tube type in question, and (2) there are no known issues regarding performance for samples in that matrix.

If you choose to include claims for use of your assay with particular matrices or tube types that were not previously cleared for zonisamide or lamotrigine immunoassays, you should include a more comprehensive study rather than relying solely on spiked samples, so that the evaluation incorporates any effect of pre-analytical steps. For example, samples from patients treated with zonisamide or lamotrigine may be drawn into “control” tubes and “new matrix” tubes. Studies of this type described in the literature may be used as supporting information if the study is well-designed and the assay used for testing is similar to your assay. You should describe the study design including sample types and preparation, calculations or statistical methods, and specific collection tubes used. Results should include recovery relative to the control matrix for each concentration level tested, as well as results of linear regression analysis. You should also clarify how you determined that assay precision is not affected by sample matrix.

Because tube components may sometimes affect assay results, we encourage manufacturers to clarify in the package insert the specific types of tubes used in the method comparison and matrix comparison studies.

6.H Specimen Storage Recommendations

You should describe in your 510(k) the study design and results supporting the conditions for specimen storage and transport that you recommend in your package insert. The study should include assessment of whether the assay maintains acceptable precision and accuracy when patient specimens are stored at the limits of the storage times and temperatures (including freeze/thaw cycles if applicable). We recommend you present results in terms of concentration measured and percent recovery for the various time points and explain any trends observed.

6.I Method Comparison

You should compare patient sample measurements using your assay to those measured with a previously cleared assay. It is important to use unaltered patient samples from the intended use population because these reflect the relevant proportions of free and bound drug, metabolites, and other drugs commonly co-administered to patients treated with the target drug. Banked (retrospective) samples are typically appropriate for these studies. However, we recommend you provide general information concerning known overall sample characteristics, selection methods, and criteria (see bullets below). Samples should be distributed throughout the claimed analytical measurement range of your assay. We recommend you follow the CLSI guideline EP-9A2, “Method Comparison and Bias Estimation Using Patient Samples” regarding your study design.

Appropriate sample size for this evaluation may depend on factors such as assay precision, interference, and assay range. The number of patients should also be large enough to observe any effects from inter-individual variations. Consistent with CLSI document EP9-A2, we recommend testing a minimum of 100 patient samples across the measuring range of the assay. We recommend you provide information supporting the study sample size in the 510(k).

If you choose to include multiple specimens from some of the individual patients you should summarize your results with appropriate statistical analyses to account for correlation of repeat measurements within patients in the study.9

We recommend that you include the following when describing the method comparison study design:

  • Selection methods used to obtain the samples. You should describe the origin of the sample population as a whole and any specific criteria (inclusion or exclusion) you applied in obtaining samples. We also recommend you describe summary information of the sample population as a whole, such as age range (e.g., adults) and time of blood draw with respect to drug administration (e.g., pre-dose), as well as sample matrix. If samples were from clinical studies or sites using specific or atypical drug regimens or sampling times, you should indicate this.
  • The number of data points and number of samples measured. If multiple specimens were selected from individual patients, we recommend you indicate the number of patients represented by the sample population.
  • The sites at which the study was conducted (e.g., manufacturer’s site).
  • The comparator method used and the type of site at which that method was performed (e.g., manufacturer’s site, clinical laboratory).
  • The time over which samples were measured using the new test method and the number of calibration cycles performed.
  • Statistical methods used to analyze the data.
  • If any data were omitted from any of your analyses, this should be described for review and justified.

We recommend you include analyses of data based on single measurements of the new device (for example, the first replicate of each set of measurements taken with the new device) in the 510(k) and the labeling since this may best depict variability of results under conditions that will be used for reporting test results (i.e., results based on one measurement). If you collected duplicate measurements with each system (as recommended in CLSI EP9-A2), we recommend you also include these results since this may provide a clearer estimate of bias relative to the predicate device.

We recommend you include the following when presenting results of your method comparison evaluation in the 510(k):

  • Range of data (the highest and the lowest value by the comparator).
  • The method used to fit the linear regression line (ordinary least squares, weighted regression, Deming, Orthogonal regression) and a scatter plot of new assay versus the comparator method illustrating the line of best fit along with the line of identity.
  • Results of regression analysis including the slope and intercept with 95% confidence intervals, the standard error of the estimate (calculated in the y direction – rather than orthogonally), and correlation coefficient should be included. If the comparator method is subject to measurement error, then Deming regression may be appropriate in stead of ordinary least squares regression. We recommend you present method comparison results based on single measurements of each sample with your new assay.
  • A bias/difference plot of difference in measurements (i.e., new method minus the comparator method on the y-axis) versus the drug concentrations (on the x-axis). We recommend you also present related summary results including the mean bias and standard deviation.
  • A line listing of all measurements from both methods, for reference.

6.J Calibrators and Control Materials

This section summarizes information we recommend you include in your 510(k) to describe calibrator and control materials you plan to market for zonisamide and lamotrigine assays.10

General Description

You should describe the components of the calibrators and control materials, including analyte concentrations, matrix, and preservatives, as well as any specific components that may call for special handling (e.g., preservatives such as sodium azide.) You should also describe if any matrix effects are observed, as compared to patient specimens. The latter can be incorporated into the recovery studies (see Section 6C- Trueness/Bias, above) by spiking patient samples and calibrator samples with equal amounts of the drug in question.

When blood products are used as components, each donation of human blood or blood component intended for use in preparing a product must be tested for communicable disease agents specified in 21 CFR 610.40. The material must be tested by an FDA approved (or equivalently recognized) assay and found to be negative for the communicable disease agents.

Traceability and Value assignment

You should describe the material to which your calibrator is traceable. We recommend that calibrators should be traceable to the best reference material available (for example, USP material) when it is available. You should include sufficient information to clearly characterize the accuracy of the assigned value of the marketed calibrator11.

The following types of information may be helpful to present for this purpose:

  • A summary of your process for value assignment and verification.
  • The uncertainty of measurement of the product calibrator assigned values.12
  • Evaluation of calibrator recovery relative to the higher level standard to which the calibrator is traceable.

You should also describe how you assign values for control materials.

Stability

You should describe your design and results for stability studies under closed and opened conditions for calibrator and control materials. “Closed” conditions refer to the shelf-life conditions you recommend to end-users prior to opening the product. “Opened” refers to the time after first opening the vial, which may include on-board or other recommended storage and use conditions. For long term studies for which results are not yet available, acceptance criteria may be provided in place of results.

We recommend that you present results for stability in terms of analyte concentration recovery since this is most understandable for FDA review. One approach is to measure a panel of samples, or control materials that are kept under known stable conditions, while calibrating the system at the various test time points, using the calibrators under evaluation. Another approach may be to compare measurements of calibrators under test conditions to measurements of the same calibrator material kept under known stable conditions at the various time points. We recommend that the other assay reagents should be kept under known stable conditions during calibrator stability studies. You should include testing of samples at multiple concentrations across the assay range and at multiple time points.

If you present results or acceptance criteria for calibrator stability in terms of control material recovery, please note that acceptable criteria for control ranges are often much wider than those that would be acceptable for calibrator stability. The results or acceptance criteria you specify for calibrator stability should ensure continued accuracy commensurate with performance claimed in the package insert.

We recommend that you include descriptions of the following for both opened and closed calibrator or control material stability testing:

  • Test samples and control/reference samples measured.
  • Materials used for calibration.
  • Time intervals of testing.
  • Temperature (and any other relevant environmental conditions).
  • Results (and/or acceptance criteria when long-term results are not available). If you conclude that no change is detected, you should also clarify in the 510(k) the difference (in terms of recovery) that your study was designed to detect.
  • Expiration dating currently specified in the product label.

6.K Studies at External Sites

For fully automated immunoassays that use standard technologies, studies using the new system performed at the manufacturer’s site are typically sufficient. For assays that may be complex to perform (e.g. complex, non-automated analytical or pre-analytical procedures), we recommend that you evaluate performance at two external clinical laboratory sites in addition to that of the manufacturer. This study could be included as part of the method comparison study described above and should also include precision information from each site. Data from individual sites should initially be analyzed separately to evaluate any inter-site variation. The data summary may be pooled in the labeling if you demonstrate that there are no significant differences in results among sites.

7. Labeling

Labeling for in vitro diagnostic devices must satisfy the requirements found in 21 CFR 809.10. Specific recommendations for zonisamide and lamotrigine assays, regarding some of these requirements, are described below.

Specimen collection and storage:

You must include instructions for specimen collection and preparation for analysis. 21 CFR 809.10(b)(7). We recommend that you include the following:

  • Discussion of any limitations or instructions related to the specimen type, including the appropriate matrix, anticoagulants, preservatives, or collection tubes.
  • Instructions concerning preserving integrity of the specimen, such as required temperatures or materials for collection, transport, storage (short and long term), and assay procedural steps. Storage conditions that you recommend to the user should be based on the conditions you have validated for your test system. You should clearly define acceptance criteria that you apply (e.g., x% recovery) in determining the recommended storage conditions. Information on storage conditions based on literature can be cited if it is applicable to your test system.
  • Discussion of the importance of consistency and accurate recording of time of blood draw with respect to the last dose if this is relevant for interpretation of results.

Assay procedure

You must provide a step by step outline of recommended procedures from reception of the specimen to obtaining results. 21 CFR 809.10(b)(8). We recommend that you include the following as part of the procedure:

  • A validated procedure for dilution, if you instruct users to dilute samples with values above the highest calibrator.
  • Steps, based on procedures you have validated for your test system, that users can take to minimize the effect of carryover or other causes of bias or imprecision.
  • Any special handling instructions for reagents.

Quality control and calibration procedures

The step by step outline of the procedure must include details of kinds of quality control procedures and materials required, as well as details of calibration. 21 CFR 809.10(b)(8)(v) and 21 CFR 809.10(b)(8)(vi). You should advise users of the specifics of calibration and quality control procedures necessary to ensure your stated performance claims. We recommend that you include a statement in your package insert that laboratories should follow federal, state, and laboratory guidelines for quality control. When you specify ranges for your quality control material, you should clarify how these were determined and indicate that laboratories should determine their own acceptable ranges based on their own acceptance criteria. We also recommend that you identify the material to which your calibrator is traceable. If you do not market these materials for your assay, you should include recommendations for appropriate quality control materials.

If your 510(k) includes calibrator or control materials to be sold separately from your assay, you should include the separate labeling for these components.

Warnings and Precautions

When blood products are used as components, you should include a statement in your package insert that the animal/human source components containing any blood-product derived material has been tested by FDA approved (or equivalently recognized) assays and found to be negative for HIV type 1 (HIV-1) and type 2 (HIV-2), as well as for hepatitis B surface antigen and antibody to hepatitis C virus (HCV), and found to be negative (not repeatedly reactive). You should also clarify that, because no testing can offer complete assurance that these or other infectious agents are absent, this material should be handled using good laboratory practice to avoid skin contact and ingestion.

Limitations

You must include a statement of limitations of the procedure. 21 CFR 809.10(b)(10). You should identify any factors known to affect results and describe the effect on results (e.g., highly lipemic samples may cause falsely low results).

It may be appropriate in some cases to recommend that it is generally good practice to use the same assay and matrix consistently for individual patients.

Reference ranges

You must indicate how reference ranges were established, including identifying the population on which it was established. 21 CFR 809.10(b)(11). At this time, we recommend, that you state that therapeutic ranges are not well-established for zonisamide and lamotrigine and include explanations such as the following:

Considerable overlap has been observed between serum levels in responders and non-responders as well as between levels associated with seizures and adverse effects, and there is not a clear relationship between lamotrigine serum concentration and clinical response across patients. Optimal ranges may vary among patients. Caution should always be exercised regarding individual patient management based upon observed reference ranges; assay results should always be interpreted in combination with careful observation of clinical response.

We recommend that, rather than citing a reference range based on a single study, you cite and summarize balanced and representative studies in the literature concerning serum concentration ranges that have been observed in those studies to be associated with seizure control or with adverse effects. When summarizing studies, we recommend that you include relevant points, such as the sampling time (relative to time of drug dose), the assay methodology, and the nature of the population tested in that study (e.g., newly diagnosed). If studies have shown lack of correlation, we recommend you discuss this as well. We ask that you include in your 510(k) copies of references you use to support statements concerning ranges cited in your labeling. You should clarify that reference ranges should only imply a lower limit below which therapeutic response is relatively unlikely and an upper limit above which toxicity is relatively likely to occur in the specific populations studied. Clinicians using proposed ranges should be aware that because of individual variation some patients may achieve therapeutic benefit with serum drug concentrations outside the range or may experience toxicity with levels below the lower limit of the reference range.

We recommend that you refer to the drug package insert regarding pertinent pharmacokinetic information including: time to steady state, appropriate sample draw times (with respect to dose), and clinical conditions or co-administered drugs that may affect pharmacokinetics.

You should discuss, or refer to the drug label, regarding information such as time to steady-state, or recommended sampling times.

Performance Characteristics

You must include specific performance characteristics of the assay. 21 CFR 809.10(b)(12). We recommend that you summarize the study design and results for each performance characteristic discussed in Section 6. Your representation of study designs and results in the package insert should include information relevant to aid the user in understanding test performance. The reportable range claimed in the package insert should be based on the assay range for which you have validated precision, linearity, and method comparison. The CLSI guidelines, referred to in Section 6 – Performance Characteristics, include instructions and examples for statements of claims in the labeling. We recommend you follow these formats in your labeling. You should include data plots of the method comparison study. If your assay is for use with more than one instrument, you should specify in the package insert the instrument used to evaluate the performance characteristics represented in the package insert.


1 FDA docket: FDA-2004-D-0421.

2 Guidances specifically for sirolimus, cyclosporine and tacrolimus assays are available at http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/ GuidanceDocuments/ucm077267.htm, and http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/ GuidanceDocuments/ucm092778.htm. These guidances address additional issues including sample pre-treatment for whole blood samples and cross-reactivities for immunosuppressant assays.

3 We recommend that you contact OIVD for feedback regarding possible regulatory routes and validation strategies if you are considering submitting premarket submissions for new types of assays.

4 Patsalos et al, "Antiepileptic drugs--best practice guidelines for therapeutic drug monitoring: a position paper by the subcommission on therapeutic drug monitoring, ILAE Commission on Therapeutic Strategies." Epilepsia 49 (2008)1239-76.

5 The current labeling for lamotrigine states “the value of monitoring plasma concentrations of LAMICTAL has not been established. Because of the possible pharmacokinetic interactions between LAMICTAL and other drugs including AEDs, monitoring of the plasma levels of LAMICTAL and concomitant drugs may be indicated, particularly during dosage adjustments. In general, clinical judgment should be exercised regarding monitoring of plasma levels of LAMICTAL and other drugs whether or not dosage adjustments are necessary.”

6 In this context, the term “reference material” refers to material whose properties are sufficiently homogeneous, stable and well established to be used for the calibration of a measuring system, the assessment of a measurement procedure, or for assigning values to materials.

7 Preparation of sample dilutions by consecutive 1:1 dilutions of a high sample may not be sufficiently well-distributed throughout the assay range, especially near the lower end of the assay range; therefore we do not recommend this approach. In addition, we do not recommend the approach of preparing multiple dilution series’, each spanning only part of the assay range.

8 For testing collection tubes/anticoagulants, see Section 6G-Matrix Comparison.

9 We do not recommend selecting multiple samples from a small number of patients in order to achieve the adequate sample size since this may not evaluate the effect of inter-individual variability.

10 Additional FDA guidance documents generally addressing calibrators and controls include “Abbreviated 510(k) Submissions for In Vitro Diagnostic Calibrators; Final Guidance for Industry” and “Assayed and Unassayed Quality Control Material”.

11 The term accuracy in this context is related to both trueness and precision of measurement.

12 See International Standards Organization, 2003 “Metrological Traceability of Values Assigned to Calibrator and Control Materials,” ISO Document 17511.