ADC Bioanalysis Knowledge Base

FDA_BMV_2018

5 min read

FDA Bioanalytical Method Validation Guidance for Industry (2018)

概述 Overview

Document: Bioanalytical Method Validation — Guidance for Industry
Agency: FDA (CDER / CVM)
Published: May 2018 (Final)
Status: Active for US submissions; ICH M10 (2022) is the harmonized successor — FDA now accepts and prefers M10 for new programs
Source: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/bioanalytical-method-validation-guidance-industry
PDF: https://www.fda.gov/files/drugs/published/Bioanalytical-Method-Validation-Guidance-for-Industry.pdf

The 2018 guidance replaced the 2001 BMV guidance, adding an explicit ligand-binding assay (LBA) chapter and harmonizing with the evolving biologic/biotherapeutic landscape. It was the primary US BMV standard until ICH M10 Step 4 adoption in 2022.

核心要点 Key Points

1. Two-Method Framework: Chromatographic vs. LBA

The guidance formally bifurcates validation requirements by platform:

Chromatographic methods (LC-MS/MS):

  • LLOQ: ±20% accuracy, ≤20% CV
  • QC levels (LQC, MQC, HQC): ±15% accuracy, ≤15% CV
  • ISR: ≥67% of repeated samples within ±15% of original result
  • Calibration: ≥6 non-zero standards; linear or 1/x² weighting

Ligand-binding assays (ELISA/ECL):

  • LLOQ: ±20% accuracy, ≤20% CV
  • QC levels: ±20% accuracy, ≤20% CV (less stringent than LC-MS/MS, reflecting inherent LBA variability)
  • ISR: ≥67% of repeated samples within ±20% of original result
  • Calibration: 4PL or 5PL fit; ≥6 non-zero standards

2. Matrix Requirements

  • Selectivity: Test ≥6 individual matrix lots; specifically include:
    • Hemolyzed (2% hemolyzed plasma, typically ~2 g/L Hb)
    • Lipemic (triglycerides ≥1000 mg/dL)
    • Protein-deficient
  • Frozen matrix lots: Use individual donor lots, not pooled matrix, for selectivity assessment to capture individual variability

3. Stability Requirements

The 2018 guidance explicitly requires:

Stability TypeConditionDuration
Bench-topRT (room temperature)Minimum 24 h
Freeze-thaw−70°C ↔ RT≥3 cycles
Long-term frozen−70°C or −80°CMust cover entire study duration
Processed sample4°C or RT in autosamplerDuration of longest analytical run
Stock solution−20°C and −70°C6 months minimum
Diluted matrixSame as long-termIf diluted samples stored

ADC-specific stability caveat (not explicit in guidance but well established): Stability at room temperature for cleavable-linker ADC samples may be dramatically shorter than for a conventional mAb. The guidance’s “minimum 24 h bench-top” criterion must be tested rigorously; many Val-Cit-linked ADC samples degrade within 1–2 h at RT if not properly stabilized (EDTA + ice processing).

4. ISR Requirements

ISR is a required post-validation quality control step to confirm method performance on actual study (incurred) samples:

  • When required: Pivotal studies; GLP toxicokinetic studies; Phase 2b/3 clinical studies
  • Number of samples: ≥10% of total samples (minimum 20 samples in small studies; 100 samples in large)
  • Acceptance: ≥67% of ISR samples within ±20% of original result (LBA) or ±15% (chromatographic)
  • Timing: ISR must be done within the bioanalytical stability window; cannot be done after samples have degraded

5. Partial Validation

When making changes to a validated method, partial re-validation is required. The guidance lists common change types:

ChangeRequired Parameters to Re-Validate
New species/matrixSelectivity, LLOQ, matrix effects, stability
New anticoagulant (e.g., heparin → EDTA)Selectivity, matrix effects; stability if anticoagulant affects analyte
New extraction procedureExtraction recovery, matrix effects, LLOQ, QC accuracy/precision
New concentration range (extending)Additional calibrators, QC at new levels
New analyte form (e.g., total vs. free)Full validation if new analyte; partial if same antibody, new detection

常见问题和挑战，具体案例和解决方案

Challenge 1: LBA Acceptance Criteria Are Wider Than LC-MS/MS — When Is This Appropriate?

Problem: ±20% CV and ±20% bias for LBA QCs is significantly looser than ±15% for LC-MS/MS. Some programs apply LC-MS/MS criteria to LBA assays (internally). What does FDA expect?

FDA position: The 2018 guidance explicitly uses ±20% for LBA. FDA review divisions generally accept this. However, if a program achieves ±15% in LBA and uses that as the internal acceptance criterion, FDA considers it acceptable and may view it favorably in a review.

Solution: Document the rationale for acceptance criteria in the method validation report. If the LBA is performing at ±15%, do not artificially widen to ±20% — it undermines confidence and increases variance.

Challenge 2: Stability Testing Window Shorter Than Study Duration

Problem: A long oncology trial (18–24 months of dosing + 12-month follow-up = 36 months total) requires long-term stability data covering 36 months. Stability testing cannot keep pace with the study.

FDA expectation: Long-term stability must be confirmed “before the long-term storage stability period claimed.” In practice, FDA accepts a risk-based approach where sponsors commit to ongoing stability monitoring and address the gap prospectively in the validation report.

Solution:

  1. Start long-term stability testing before Phase 1 initiation (at validation)
  2. Report stability data at 6-, 12-, 24-month checkpoints in integrated study reports
  3. If samples must be analyzed before stability is confirmed for their storage duration: pre-approve a risk statement in the bioanalytical plan with the clinical pharmacology team

Challenge 3: Switching from FDA BMV 2018 to ICH M10 Mid-Program

Problem: A program validated under FDA BMV 2018 for Phase 1 now needs to prepare for an NDA/BLA that will be filed globally (FDA + EMA + PMDA). ICH M10 requires ≥10 matrix lots (vs. ≥6 in FDA BMV 2018) for selectivity. The Phase 1 validation only tested 6 lots.

FDA/EMA position: ICH M10 does not require retrospective re-validation of studies conducted under prior guidances. However, any new studies (Phase 2b/3 supportive studies) should use M10-compliant methods.

Solution: Conduct partial re-validation under M10 (4 additional matrix lots for selectivity; parallelism assessment if not previously done) before Phase 3 initiation. Document the transition in the bioanalytical plan. Use cross-validation to show equivalency between Phase 1 method (FDA BMV 2018) and Phase 3 method (M10).

Graph View

Backlinks