Analytical validation / assay validation notes

Analytical Validation Table Continuation

visible category / row	transcribed content	caveat
Stability of BM & reagents / Process stability	2-8도 & ambient temperature	Korean unit/handwritten-like spacing needs manual check.
Short & long term storage	Eg. GBA activity assay: storage for 10 and 20 weeks at either 4°, -20° or -80° + addition of protease-inhibitors?	Temperature symbols and line breaks checked visually but should be confirmed.
GBA activity assay	Residual activity (%) > 80%.	Right-side table cell.
Freeze-thaw cycles	Eg. cathepsin D activity assay: 3 cycles. Minimum 12h between cycles	Exact punctuation/spacing pending.
acceptance / error note	relative error in a calibration curve: absolute value of ((measured concentration) / (nominal concentration) * 100 - 100); It should be within 20% and 25% in LLOQ.?	Category label at far left includes Korean note 어느 category 불확실?; exact interpretation not normalized.
Calibrator	Calibrator (= calibration standard) = reference standard (ie known amount) + targeted matrix (ie the same matrix as the study samples to be assayed)	Long line visible and OCR-supported.
Kamiguchi	3 concentrations of matrix based QCs spiked to be higher than naive condition (Low, Mid, High). Low: 3 fold of endogenous level, Mid: IC50 of calibration curve, High: 1/3 of ULOQ	Citation/name and concentration wording require manual confirmation before citation use.
FDA	5 QC levels (LLOQ QC, Low QC, Mid QC, High QC, and ULOQ QC)	trailing per appears in OCR/image but context continues off-row.
EMEA 2011	(행간?) 4 levels:: the LLOQ, within three times the LLOQ (low QC), around 30 - 50% of the calibration curve range (medium QC), and at least at 75% of the upper calibration curve range (high QC).	Korean parenthetical and exact punctuation are high-risk.
QCs	QCs (ie Quality control samples) = reference standard (ie known amount) + targeted matrix
In-study sample	In-study sample	Header/row label only.
A&P run	LLOQ, low, mid and high QCs

ELISA Validation

1. Determine buffer-based LLoQ and detection feasibility in plasma and CSF (8x dilutions (1-128)). Milestone 1: CV under 5% (inter-assay), dilution linearity within 20% of expected values on at least some of the dilutions.

2. Determine MRD (4 different dilutions (based on experiment 1) with and without spike-in). Milestone 2: Spike-in recovery within 80-120% on at least some of the dilutions. Spike-in is recombinant protein from a different source than the ELISA-kit.

3. Prepare QC samples (aliquots of recombinant protein with a concentration similar to human healthy controls). Run 3 Independent QC runs on separate days. Milestone 3: Complete the QC runs.

www.aacrjournals.org
5973
Clin Cancer Res 2008;14(19) October 1, 2008
Quantitative
Qualitative
Immunoassays
Immunohistochemistry

MJFF Phases of Assay Development

Phase	visible heading	visible content
Phase 1	Antibody pair finding -> initial assay optimiation	sensitivity testing using recombinant protein; specificity testing using KO cells; spike recovery & dilution linearity in HC biofluid samples.
Phase 2	ASSAY VALIDATION FOR PRECLINICAL AND EARLY CLINICAL EVALUATION	assay sensitivity, dilution linearity, stability, intra/inter- robustness/reproducibility testing, etc.; assay testing in patient samples.
Phase 3	GLP ASSAY DEVELOPMENT FOR LATE-STAGE CLINICAL TRIAL USE	transfer of the assay to a separate CRO for GLP-certification and reevaluation.

Examples of Biomarker

The lower half is a wide table. Because several cells are narrow, vertically wrapped, Korean-heavy, or cut at the photo edges, the table is summarized conservatively below and should be verified from image crops before KB reuse.

row / type	example	pathophysiology / discovery	MOA of drug	clinical trial / utility
Patient selection BM	Herceptin with Korean note roughly indicating it is target rather than BM	HER2 is a major oncogene; HER2 is activated/up-regulated in subgroup (20%); HER2 overexpression/amplification notes including MAC 117, five- to tenfold (26), Korean note about an animal cancer-causing virus hijacking part of EGFR; discovery as candidate biomarker using patient-derived sample and patient-record comparison; overexpression in breast cancer associated with poor prognosis (Slamon et al. 1987, PMID); amplification/copies correlated with positive lymph node count and decreased survival.	HER2 internalization/degradation; ADCC recruiting immune cells (NK Cells); Korean explanation of generating/selecting HER2-specific antibody, including 4D5.	Clinical trial cell cites (Smith 2001, PMID 11989525) and notes trials in HER2+ patients, HER2 overexpression (3+ by IHC) and gene amplification by FISH; clinical utility Korean text describes routine Her2 testing by IHC and interpretation of 0, 1+, 2+, 3+, with FISH follow-up for 2+.
Patient selection / prediction BM	keytruda; pembrolizumab, a therapeutic antibody that binds to and blocks PD-1 located on lymphocytes	Mechanism note: increased PD-1 interacting with PD-L1/CD80 could induce increased expression of Bim leading to apoptosis of CD8+ T-cells; Korean note indicates cancer-cell survival/proliferation without interference.	Keytruda vs SOC (Docetaxel) appears in the trial/treatment area.	P3 KEYNOTE-024 (Reck et al. 2016, PMID 27718847) with Only in PD-1>50%; phase 2/3 KEYNOTE-010 (Herbst et al. 2016, PMID 26712084) advanced non-small-cell lung cancer, with PD-L1 expression >= 1% of tumour cells, but analysis in patients >50% of PD-L1.

Uncertain Spans

• The page starts mid-table and ends mid-table; adjacent photos may contain continuation rows and clearer edges.
• The top analytical validation table contains exact acceptance criteria, units, and Korean notes that require manual verification before use as canonical values.
• The small biomarker assay figure under ELISA validation is preserved as image evidence; exact plus/minus entries are not reliably transcribed.
• The Examples of biomarker table is structurally visible but dense. Korean vertical text, clinical utility wording, arrows, PMIDs, trial names, and cutoff expressions need manual confirmation.
• PD-1>50% may be intended as PD-L1 >50% in the visible Keytruda row; the photographed text/OCR should be checked against surrounding pages before correction.
• OCR reports page navigation context as Genetic prevalence, but the visible status/navigation strip indicates Analytical (=Technical) validation; this draft uses the visible status strip for nav_path.