Abstract
Background: Measuring pharmacodynamic biomarkers near the therapeutic site of action presents considerable challenges for sites with limited matrix volume or difficult access. Bioanalytical method qualification requires the use of numerous matrix samples, which is problematic for rare matrices. The aim of this study was to design and implement a streamlined, fit-for-purpose strategy for qualification of biomarker assays in rare matrices. Materials & methods: A multiplexed biomarker immunoassay was developed in human aqueous humor. Results: Our strategy was successfully implemented, providing characterization of assay performance while reducing number of samples in assay qualification. Our assay was used in clinical trial support for an ophthalmic drug candidate. Conclusion: Our results indicate this approach can be applied to other early stage drug development programs facing similar challenges.
Keywords: : aqueous humor, biomarker assay qualification, Ella, pharmacodynamic biomarker, rare matrix
Tweetable abstract
Bioanalytical assay development and qualification requires the use of numerous matrix samples, which is problematic for rare matrices. We designed and successfully implemented a streamlined strategy for qualification of biomarker assays in rare matrices.
Plain language summary
Summary points.
To address the challenge of a limited number and extremely low volume of rare matrix samples.
We developed a streamline strategy to qualify clinical biomarkers assays in those matrices.
This strategy was successfully applied to an Ella immunoassay measuring four cytokines in aqueous humor.
The Ella immunoassay was subsequently used to support the pharmacodynamics analysis for a clinical trial of an ophthalmologic drug.
All four cytokines were detectable at an equivalent frequency in a limited number of commercially sourced aqueous humor samples compared with an expanded panel.
A similar streamlined biomarker assay qualification strategy can be applied to other rare human matrices.
Biomarkers are critical in early clinical development as they can demonstrate target engagement and confirm the mechanism of action of new drug candidates prior to availability of efficacy data [1]. Pharmacodynamic (PD) biomarker data can also inform dose selection and de-risk subsequent efficacy analysis from mid-stage clinical trials. As drug candidate complexity increases, PD biomarker data become critical for decision-making.
Ideally, a PD biomarker is measured at, or near, the site of action of a therapeutic. Data from traditional matrices (e.g., serum and plasma) can be supplemented or replaced by data from more challenging, rare and potentially complex matrices closer to the site of action. Examples of such matrices include cerebrospinal fluid (CSF), bone marrow/bone marrow aspirates, synovial fluid, skin/tape strip or suction blister, tracheal aspirates, bronchoalveolar lavage effluent, nasal swabs, tumor/tissue biopsy and aqueous humor [2]. Each of these matrices presents unique bioanalytical challenges such as invasive collection, limited sample volumes, chemical complexity, high viscosity and difficulty in sourcing. Rare matrices specifically refer to samples with small volumes that are challenging to collect and commercially source, resulting in a limited number and volume of samples available for bioanalytical method development and assay qualification or validation. The COVID-19 pandemic resulted in supply chain issues that caused shortages in multiple matrices, such as nonhuman primate CSF, highlighting the need for strategies to conserve rare matrices [3].
Aqueous humor can be used for measuring PD biomarkers of ophthalmic drugs delivered by intravitreal administration, as well as other analytes. It is a clear fluid found between the lens and cornea near the front of the eye that maintains intraocular pressure and delivers nutrients to the eye [4]. We consider aqueous humor a rare matrix due to multiple factors including small sample volumes, invasive collection and difficulty in sourcing [5]. Often, only up to 100 μl of aqueous humor can be collected from the anterior chamber of a patient's eye. Aqueous humor is commercially collected from cadavers or during cataract surgery from patients with or without other diagnosed ocular diseases. In most of our studies, however, we avoid the use of cadaveric aqueous humor due to the potential release and/or degradation of cytokines, as well as the undesirable implications of a compromised blood–ocular barrier.
Biomarkers measured in aqueous humor have successfully informed decisions during drug development, even with the inherent challenges of rare matrices. For example, a free VEGF-A assay with low pg/ml sensitivity demonstrated target engagement in patients with diabetic macular edema, as free levels of VEGF-A decreased with increasing levels of an anti-VEGF-bispecific antibody in aqueous humor [6]. DKK3, a substrate for the serine protease HTRA1, is a PD biomarker in aqueous humor that demonstrated target engagement and dose response of an anti-HTRA1 Fab in a phase I study and informed the dose levels and dosing frequency for a phase II trial [7,8].
We developed a fit-for-purpose strategy for the qualification of biomarker assays utilizing rare matrices to support early stage clinical trials. We balanced characterizing performance properties of the biomarker assay while conserving the number of samples needed for assay qualification. Our strategy combines the expertise and experience from industry and health authorities on biomarker assay strategy and qualification from white papers, research publications, guidances and meeting reports [9–14]. Here, we describe the assay qualification plan, acceptance criteria and results from a case study measuring four PD biomarkers in human aqueous humor samples using the Ella platform.
Materials & methods
Matrix samples
Human aqueous humor samples were commercially procured from BioIVT (NY, USA) and Hemoline Ltd (Tbilisi, Georgia). Samples were collected from individuals with dry age-related macular degeneration.
Ella platform
Four cytokines of interest were measured using the Ella platform from ProteinSimple (Bio-Techne, MN, USA). Ella is a microfluidic-based immunoassay platform in which multiple single-plexed assays can be run concurrently. Diluted samples are routed through microfluidic channels, each composed of three separate glass nanoreactors (GNRs) coated with an antibody directed toward the analyte of interest. Unbound material is washed away and a fluorescently labeled detection antibody, specific for that same analyte, is introduced. Following a final wash step, the GNRs are scanned for fluorescence intensity, giving a triplicate result. A cartridge can measure up to eight different analytes; crosstalk and interference are minimized as the sample is partitioned into separate channels [15]. The measured fluorescence from each GNR is back-calculated against a lot-specific, factory-set calibration curve, resulting in a concentration value.
In-house assay qualification
Cartridges were purchased from ProteinSimple to measure four inflammatory cytokines in aqueous humor samples. Assay controls prepared at low and high concentrations, using recombinant analyte purchased from a commercial source (Bio-Techne), were run on each plate. All samples were tested in a single well. Samples to assess detectability were prepared at the minimum required dilution (MRD) of 1/5 and analyzed. Samples to assess parallelism were first prepared at the MRD, followed by a series of three, 1/2 serial dilutions in sample diluent (SD13; Bio-Techne, cat. no. 896098). To confirm specificity, an aqueous humor sample from a disease-state individual and a sample of pooled human serum from normal individuals were analyzed at the MRD in the presence or absence of antibodies against each analyte at 100 μg/ml. The antianalyte antibodies were provided by ProteinSimple (Bio-Techne) and used at the recommended dilution. Assay sensitivity samples were prepared by spiking recombinant analytes in sample diluent at concentrations matching the upper and lower limits of quantitation of the assay for each analyte. Assay sensitivity samples were prepared using the same recombinant analytes used to prepare assay controls.
Approach
Rare matrix biomarker assay qualification strategy
We devised an assay qualification strategy to account for the unique challenges inherent to aqueous humor, but these guidelines can be applied to any rare matrix. The assay qualification has two stages, with the initial assay development and qualification completed in-house using three disease state or healthy human samples; this is considerably fewer samples compared with the 10+ matrix individual samples commonly tested [9,13,14,16]. The decision to substitute samples from healthy donors for disease-state samples will depend upon the biomarker of interest. An additional five disease-state matrix samples are tested at a contract research organization (CRO) during the second stage of the assay qualification to prepare for clinical sample testing. Specifically, a priori acceptance criteria for each parameter are set based on fit-for-purpose assay strategies including how the data will be used, whether biomarker levels are expected to increase or decrease with treatment, the anticipated fold changes in concentration, as well as if longitudinal samples will be analyzed [11].
The first step in assay development for a custom assay, or for feasibility assessment of a commercial kit, is determining the assay range using the standard curve and then setting the quality control (QC) concentrations. The assay's accuracy and precision are then assessed using these QC samples. The standard qualification plan and our streamlined strategy are similar, except for a reduction in the number of runs required for accuracy and precision (from six to three runs; Table 1). There should also be special consideration given to the use of an endogenous QC. Ideally, a human matrix containing the endogenous biomarker is included as the third QC along with the two QCs prepared with recombinant biomarker spiked into assay buffer at low and high concentrations. We recommend considering healthy human serum or plasma as a surrogate for a rare matrix for the endogenous QC, if the biomarker of interest is present at detectable concentrations, since these matrices are more readily available. These two surrogate matrices are also biochemically complex, likely more complex than the rare matrix of interest, and may contain higher concentrations of biomolecules, thus setting a high bar for monitoring the accuracy and robustness of the assay.
Table 1.
Comparing the standard biomarker assay qualification plan to the streamlined plan for rare matrices.
| Experiment | Standard qualification plan | Streamlined plan for rare matrices |
|---|---|---|
| Standard curve accuracy, precision, and quantitative range | ≥6 runs | 3 runs |
| Endogenous quality control sample | 1 (in same matrix) | 1/none (optional; consider use of surrogate matrix sample) |
| Specificity | 6 individuals | 2 samples (use a surrogate matrix if possible) |
| Detectability | 10–15 individuals | 3 individuals (minimum, expand up to 10, if available) |
| Parallelism | ≥6 individuals | 3 individuals |
| Reproducibility | ≥6 individuals | 3 individuals (combined with parallelism/detectability) |
| Stability | 3 individuals | Pool samples (test limited conditions) |
The specificity of the biomarker assay is checked early during our assay qualification to confirm that the capture antibody binds specifically to the biomarker of interest. The addition of excess capture antibody to the surrogate matrix should reduce or eliminate signal output relative to a matrix without the spiked capture antibody. To conserve rare matrix samples, specificity assessments can use a surrogate matrix containing the endogenous biomarker (Table 1). Surrogate matrix options include human plasma, human serum, cell culture supernatants and the same rare matrix from nonhuman primates. In this regard, the choice of surrogate matrix will depend upon the biomarker to be measured. Similarly to the considerations described above for choice of alternative matrices for EC, the surrogate matrix used here would also ideally be more biochemically complex compared with the rare matrix, and/or have higher concentrations of biomolecules. This would therefore set a higher bar for specificity of the capture antibody. For example, the total protein content in plasma is approximately 60–80 mg/ml compared with approximately 0.1 mg/ml in aqueous humor, making plasma an appropriate surrogate matrix for specificity experiments due to its higher relative protein content [17–19].
Detectability is critical for biomarker assays and is defined as the ability of the assay to detect the endogenous biomarker in disease-state samples. We recommend using a minimum of three and up to ten samples for these experiments, in contrast to the standard 10–15 individual samples (Table 1). While determining the a priori criteria for detectability, it is important to consider whether the biomarker is expected to increase or decrease upon drug treatment, and whether the data analysis will require normalizing biomarker concentrations to baseline levels. This information enables a thoughtful approach on the required percentage of samples with detectable levels at baseline because the overall number of samples tested will be smaller.
Parallelism is another critical parameter for biomarker assays and is assessed by serially diluting matrix samples containing the endogenous analyte. The three parallelism samples can be chosen from the detectability experiments. Parallelism samples should contain sufficiently high biomarker levels to elucidate whether the assay measures consistent biomarker levels across multiple matrix dilutions; this experiment also serves to confirm whether the selected MRD is within that range (Table 1). Setting the MRD requires special considerations for rare matrix biomarker assays to balance sample volume limitations with detectability. For example, aqueous humor samples are often limited in volume to less than 100 μl. Thus, a 1/5 dilution may be required just to achieve the volume needed for the assay, and further dilution may be beneficial to enable repeat or additional exploratory analysis, while a higher MRD could negatively impact detectability.
Spike recovery is generally not assessed in this context, since nonspecific background or matrix interference will be apparent in the parallelism experiments [10]. If parallelism is not achievable due to low concentrations of the endogenous biomarker, then spike recovery using recombinant biomarker spiked into the rare matrix could be performed as part of assay qualification. We recommend testing parallelism again later, using a limited number of study samples that have higher concentrations of endogenous biomarkers.
The reproducibility of the biomarker assay is assessed by combining the detectability and parallelism data from a minimum of two runs to further conserve samples (Table 1). An additional consideration for commercially available biomarker kits is lot-to-lot variability; this can be analyzed on an as-needed basis. Finally, for analyte/sample stability, pools of rare matrix samples should be considered as an option to enable testing a few key conditions.
If used, a CRO would conduct further assay qualification experiments, following the successful initial in-house qualification and demonstration of proof of utility for the chosen analytes. For that, five additional disease-state rare matrix samples are recommended to further assess detectability, parallelism and reproducibility. At this stage, additional assessment of sample stability could also be considered. The extent of stability testing would depend on several factors, including sample storage temperature, stage of the program, testing schedule, biomarker stability, reanalysis plans and batching of data.
Results & discussion
Case study: aqueous humor biomarker assay qualification
For an early stage clinical development ophthalmology program, assays directly measuring the drug target may prove to be challenging, often dealing with invasive sample collection, limited sample volumes and difficulty in sourcing commercial samples. Thus, measurement of downstream biomarkers would be required to demonstrate target engagement and subsequent downregulation of the target's signaling pathway. In our case study, we applied the described streamlined assay qualification plan to a multiplexed PD biomarker assay on the Ella platform. The assay measured four different inflammatory cytokines in aqueous humor samples. Sampling of aqueous humor from patients yields limited amounts of fluid, averaging <100 μl per collection per eye. Samples would also need to be divided among several other assays including those assessing pharmacokinetics (PK) and immunogenicity. The Ella platform was chosen for its degree of automation, low sample volume requirements, potential for multianalyte measurements, availability of relevant biomarkers, sensitivity and precision [20]. Importantly, the reliability and low run failure rate of the Ella minimizes the need for sample reanalysis. Furthermore, the Ella measures multiple analytes simultaneously in separate channels, thereby reducing the risk of analyte interference and cross-reactivity.
The acceptance criteria and results from the assay qualification are outlined in Table 2. A total of ten aqueous humor samples from ten patients with age-related macular degeneration were analyzed to establish detectability. Due to the importance placed on the results from this assay, the number of samples analyzed was increased from the minimum three samples to ten. For three of the four analytes tested, 100% (10/10) of the samples had measurable concentrations of analyte while for the fourth analyte, 90% (9/10) of the samples had measurable concentration of the analyte (Figure 1). Based upon the results from the detectability experiment, three samples with concentrations of all four analytes high enough to be serially diluted were analyzed to demonstrate parallelism. Measurements of all four analytes from all samples exhibited acceptable parallelism (Figure 2). The percent relative error for each subsequent dilution was ≤30% and the %CV between dilutions was ≤30%. To demonstrate reproducibility, the results from the MRD of the three samples used in the parallelism experiment were compared against the results for the respective samples from the detectability experiment. The %CV between the corresponding results for each analyte was ≤30% for all three samples.
Table 2.
In-house qualification acceptance criteria for the Ella multiplexed immunoassay and the qualification results.
| Assay parameter | Criteria | Result |
|---|---|---|
| Standard curve accuracy and precision | Mean %RE and CV ≤20% (25% at LLOQ) and total error ≤30% (3 runs) | NA (Ella uses a predetermined, lot-specific standard curve) |
| Assay quantitative range | LLOQ (pg/ml): %RE | Analyte A (1.52): -6% |
| Analyte B (0.6): -12% | ||
| Analyte C (4.1): -15% | ||
| Analyte D (0.28): 21% | ||
| ULOQ %RE range | -12 to 15% | |
| Quality controls precision and accuracy | Mean % CV ≤30% Mean % RE ≤30% Total error ≤40% (3 runs) |
Acceptable for low and high concentration controls for all analytes |
| Specificity | 2 samples show >80% signal inhibition with the capture antibody in solution. | Acceptable (also see Table 3) |
| Detectability | Fit-for-purpose acceptance criteria for detectability, e.g., >80% or 2/3 | 10/10 samples with detectable levels of analytes A, B and C 9/10 samples with detectable levels of analyte D |
| Parallelism | (2/3 samples) %CV and %RE ≤30% between dilutions within range | 3/3 samples %RE ≤30% and %CV ≤30% for all analytes |
| Reproducibility | %CV ≤30% between two values for 2 out of 3 samples | 3/3 samples %CV ≤30% for all analytes |
| Lot-to-lot variability | Controls within range as determined during qualification | Acceptable |
CV: Coefficient of variation; LLOQ: Lower limit of quantitation; NA: Not available; RE: Relative error; ULOQ: Upper limit of quantification.
Figure 1.
Detectability with a limited number of aqueous humor samples.
Ten samples were analyzed at the minimum required dilution (1/5) to assess detectability of the four analytes of interest. Nine of ten samples had levels of all four analytes above the respective assay lower limit of quantitation shown by the red lines.
LLOQ: Lower limit of quantitation.
Figure 2.
Parallelism for all four analytes upon serial dilution of three aqueous humor samples.
An initial 1/5 MRD was followed by a series of three, 1/2 dilutions in sample diluent for a final dilution of 1/40. The percent relative error from the MRD for all dilutions was within the ±30% acceptance criteria, with the exception of one sample at the 1/20 dilution for analyte C.
MRD: Minimum required dilution.
The assay quantitative range and sensitivity were confirmed by analyzing samples prepared at the manufacturer-assigned upper and lower limits of quantitation. Samples were prepared in sample diluent. The percent relative error for each sample was ≤30%, thus meeting the recommended acceptance criteria (data not shown). Two different lots of Ella kits were compared, for lot-to-lot variability, and passed that assessment (Table 2).
Specificity was determined by the addition of monoclonal antibodies against each analyte to one aqueous humor sample and one sample from a surrogate matrix (pooled normal human serum). In the presence of antianalyte antibodies, the measured concentration of each corresponding analyte was inhibited by ≥80%, meeting the recommended acceptance criteria (Table 3). Additional crossreactivity and interference testing was performed for each assay/analyte by the vendor (ProteinSimple) using panels of relevant analytes; these experiments were not repeated during assay qualification, in an effort to conserve matrix samples.
Table 3.
Results from specificity testing for each analyte.
| Sample | Analyte A (measured concentration†) | Analyte B (measured concentration) | Analyte C (measured concentration) | Analyte D (measured concentration) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| w/o Ab | w/Ab | %Diff | w/o Ab | w/Ab | %Diff | w/o Ab | w/Ab | %Diff | w/o Ab | w/Ab | %Diff | |
| Dry AMD | 679 | 0 | -100% | 16.6 | 0.333 | -98% | 592 | 2.39 | -100% | 113 | 0.316 | -100% |
| Pooled NHS | 114 | 0 | -100% | 106 | 0 | -100% | 488,733 | 337 | -100% | 4.12 | 0.325 | -92% |
One rare matrix sample (aqueous humor from a patient with dry age-related macular degeneration) and one surrogate matrix sample (pooled normal human serum) were analyzed. Specificity is demonstrated via a reduction in measured concentration in the presence of 100 μg/ml of capture antibody for each respective analyte.
Measured concentrations are in pg/ml.
%Diff: Percent difference between ‘w/o Ab’ and ‘w/Ab’; AMD: Age-related macular degeneration; NHS: Normal human serum; w/o Ab: Without competing antibody; w/A: With competing antibody.
The assay was subsequently used to test panels of clinically relevant, well-characterized aqueous humor samples from two studies, as well as in support of an early stage clinical trial. In all those cases, the assay performed as expected (data not shown), confirming its clinical utility and demonstrating the appropriateness of the assay qualification strategy described here.
Additional considerations
Considering the context-of-use is recommended for all biomarker assays, including those using rare matrices, and having the context-of-use clearly defined enables a successful assay qualification and subsequent decision making from the biomarker data [21]. Regulators encourage sponsors to consult with relevant health authorities when utilizing rare matrices for alignment on assay qualification requirements [2], especially when obtaining critical data in support of registrational studies.
In preparation for clinical trials, careful consideration of aqueous humor sample volumes (similar to other low-volume matrices) must be taken into account, as a single sample is often used for both PK and biomarker analysis of multiple analytes. Measuring small-molecule drugs in aqueous humor and ocular tissues for PK analysis is also challenging, but strategies and case studies have been described in the literature [5,22,23] for those situations. A similar assay qualification approach with a reduced number of total individual samples could be used for a large-molecule PK assay in a rare matrix to complement serum PK measurements.
Drug developers must consider additional challenges when measuring biomarkers in a rare matrix. In some cases, the site of collection of the rare matrix may not exactly overlap with the site of action, such as: 1) when measuring biomarkers in aqueous humor, which are postulated to diffuse from the retina, at the back of the eye, to the anterior chamber in the front of the eye; or 2) when analyzing biomarkers that diffuse from the patient's brain to CSF, resulting in a time delay in the response of the biomarkers to the treatment, requiring incorporation of an additional compartment to the corresponding PK/PD model [8]. An additional challenge is artificial fluctuations of biomarker concentrations due to sample collections designed for longitudinal monitoring of biomarkers. As an example, repeated lumbar punctures to collect CSF in close succession increased the levels of common biomarkers of Alzheimer's disease, including Aβ42 and tau [24].
Conclusion
Here, we outlined a biomarker assay qualification strategy for rare matrices and presented a case study applying this strategy to measure PD biomarkers in human aqueous humor. The strategy effectively characterizes key parameters for the assay while reducing the number of individual samples required and minimizing the utilized volume of those samples. The aqueous humor assay described in our case study was subsequently transferred to a CRO that performed additional qualification experiments, which confirmed and extended the conclusions from the in-house assay qualification, as well as the robustness and suitability of the in-house developed method.
Future perspective
We envision an increasing focus on bioanalysis in rare matrices as the use of diverse drug modalities and unconventional modes of drug delivery to peripheral compartments expand. More assays in rare matrices will present unique challenges for both the sourcing and bioanalysis of analytes. We anticipate the strategies for assay qualification, assay validation and the use of surrogate matrices to continue to evolve.
Acknowledgments
The authors thank Vahan Indjeian, Michael Elliott, Wayne Kung, May Chen, Henry Wiley, Hao Chen and Surinder Kaur for valuable feedback on various topics addressed in this paper. They acknowledge Violet Lee for initial discussions on the bioanalytical strategy, and Jose Diaz for support in sourcing commercially-available aqueous humor samples.
Funding Statement
Financial and material support from Genentech was received by all authors for this research and the creation of this work.
Financial disclosure
Financial and material support from Genentech was received by all authors for this research and the creation of this work. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Competing interests disclosure
All authors are employees of Genentech, Inc. and stockholders of the Roche group. The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed.
Writing disclosure
Medical writing support was provided by Anshin BioSolutions and was funded by Genentech.
References
Papers of special note have been highlighted as: • of interest
- 1.FDA-NIH Biomarker Working Group . BEST (Biomarkers, EndpointS, and other Tools) Resource [Internet]. US FDA, MD, USA. http://www.ncbi.nlm.nih.gov/books/NBK326791/ [PubMed] [Google Scholar]
- 2.Kaur S, Alley SC, Szapacs Met al. 2021 White Paper on Recent Issues in Bioanalysis: Mass Spec of Proteins, Extracellular Vesicles, CRISPR, Chiral Assays, Oligos; Nanomedicines Bioanalysis; ICH M10 Section 7.1; Non-Liquid & Rare Matrices; Regulatory Inputs (Part 1A – Recommendations on Endogenous Compounds, Small Molecules, Complex Methods, Regulated Mass Spec of Large Molecules, Small Molecule, PoC & Part 1B – Regulatory Agencies' Inputs on Bioanalysis, Biomarkers, Immunogenicity, Gene & Cell Therapy and Vaccine). Bioanalysis 14(9), 505–580 (2022). [DOI] [PubMed] [Google Scholar]
- 3.Dubiel EA, Myler H, Arnold MEet al. Biological matrix supply chain shortages: more matrices are now rare-the case for surrogate matrices. AAPS J. 24(2), 42 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tamhane M. Review of biomarkers in ocular matrices: challenges and opportunities. Pharm. Res. 36(3), 35 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Heinig K, Bertran E, Potter J, Fraier D. Ocular bioanalysis: challenges and advancements in recent years for these rare matrices. Bioanalysis 9(24), 1997–2014 (2017). [DOI] [PubMed] [Google Scholar]; • Review covering challenges with four ocular matrices and biomarkers measured in each matrix.
- 6.Göpfert JC, Reiser A, Carcamo Yañez VAet al. Development and evaluation of an ultrasensitive free VEGF-A immunoassay for analysis of human aqueous humor. Bioanalysis 11(9), 875–886 (2019). [DOI] [PubMed] [Google Scholar]
- 7.Tom I, Pham VC, Katschke KJet al. Development of a therapeutic anti-HtrA1 antibody and the identification of DKK3 as a pharmacodynamic biomarker in geographic atrophy. Proc. Natl Acad. Sci. USA 117(18), 9952–9963 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Genentech. Maass, Katie. 2019 AAPS PharmSci 360. https://www.eventscribe.com/2019/PharmSci360/fsPopup.asp?efp=SUlFUEhHSFQ4MDkx&PosterID=235059&rnd=0.8233907&mode=posterinfo ; • Novel biomarker in aqueous humor that informed the dosing schedule early on in clinical development.
- 9.Cowan K, Gao X, Parab Vet al. Fit-for-purpose biomarker immunoassay qualification and validation: three case studies. Bioanalysis 8(22), 2329–2340 (2016). [DOI] [PubMed] [Google Scholar]
- 10.Arnold ME, Booth B, King L, Ray C. Workshop report: Crystal City VI-bioanalytical method validation for biomarkers. AAPS J. 18(6), 1366–1372 (2016). [DOI] [PubMed] [Google Scholar]
- 11.Timmerman P, Herling C, Stoellner Det al. European Bioanalysis Forum recommendation on method establishment and bioanalysis of biomarkers in support of drug development. Bioanalysis 4(15), 1883–1894 (2012). [DOI] [PubMed] [Google Scholar]
- 12.Khan MU, Bowsher RR, Cameron Met al. Recommendations for adaptation and validation of commercial kits for biomarker quantification in drug development. Bioanalysis 7(2), 229–242 (2015). [DOI] [PubMed] [Google Scholar]
- 13.Jani D, Allinson J, Berisha Fet al. Recommendations for use and fit-for-purpose validation of biomarker multiplex ligand binding assays in drug development. AAPS J. 18(1), 1–14 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Piccoli SP, Mehta D, Vitaliti Aet al. 2019 White Paper on Recent Issues in Bioanalysis: FDA Immunogenicity Guidance, Gene Therapy, Critical Reagents, Biomarkers and Flow Cytometry Validation(Part 3 – Recommendations on 2019 FDA Immunogenicity Guidance, Gene Therapy Bioanalytical Challenges, Strategies for Critical Reagent Management, Biomarker Assay Validation, Flow Cytometry Validation & CLSI H62). Bioanalysis 11(24), 2207 (2019). [DOI] [PubMed] [Google Scholar]
- 15.Fischer SK, Carrasco-Triguero M, Hong Ket al. Commercial biomarker assays: friend and foe. Bioanalysis 8(22), 2351–2362 (2016). [DOI] [PubMed] [Google Scholar]
- 16.Lee JW, Devanarayan V, Barrett YCet al. Fit-for-purpose method development and validation for successful biomarker measurement. Pharm. Res. 23(2), 312–328 (2006). [DOI] [PubMed] [Google Scholar]
- 17.Tripathi RC, Millard CB, Tripathi BJ. Protein composition of human aqueous humor: SDS-PAGE analysis of surgical and post-mortem samples. Experiment. Eye Res. 48(1), 117–130 (1989). [DOI] [PubMed] [Google Scholar]
- 18.Goel M, Picciani RG, Lee RK, Bhattacharya SK. Aqueous humor dynamics: a review. Open Ophthalmol. J. 4, 52–59 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Leeman M, Choi J, Hansson S, Storm MU, Nilsson L. Proteins and antibodies in serum, plasma, and whole blood-size characterization using asymmetrical flow field-flow fractionation (AF4). Anal. Bioanal. Chem. 410(20), 4867–4873 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gupta V, Davancaze T, Good Jet al. Bioanalytical qualification of clinical biomarker assays in plasma using a novel multi-analyte Simple Plex™ platform. Bioanalysis 8(23), 2415–2428 (2016). [DOI] [PubMed] [Google Scholar]
- 21.Cowan K, Golob M, Goodman Jet al. Biomarker context-of-use: how organizational design can impact the implementation of the appropriate biomarker assay strategy. Bioanalysis 14(13), 911–917 (2022). • Considerations when defining context of use for a biomarker assay. [DOI] [PubMed] [Google Scholar]
- 22.Velagaleti PR, Buonarati MH. Challenges and strategies in drug residue measurement (bioanalysis) of ocular tissues. Methods Pharmacol. Toxicol. doi: 10.1007/7653-2013-6 (2014). [DOI] [Google Scholar]
- 23.da Silva PHR, Júnior AdSC, Pianetti GA, Fernandes C. Chromatographic bioanalysis of antiglaucoma drugs in ocular tissues. J. Chromatog. B. 1166, 122388 (2021). [DOI] [PubMed] [Google Scholar]
- 24.Olsson M, Årlig J, Hedner J, Blennow K, Zetterberg H. Repeated lumbar punctures within 3 days may affect CSF biomarker levels. Fluids Barriers CNS 16(1), 37 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]