Part 3 of the PCR Assay Validation Guide
This post continues the PCR Assay Validation Guide. In Part 1 we introduced the principles of development, qualification and validation and the role of context of use. In Part 2 we covered precision and accuracy.
A validated assay is only as trustworthy as its ability to find what is really there - and nothing else. In regulated bioanalysis, that dual requirement is captured in two performance characteristics: sensitivity and specificity. Together, they determine whether a result can be trusted to reflect biology rather than assay noise or an off-target signal.
For sponsors developing cell and gene therapies, RNA therapeutics, or biomarker-driven programmes, sensitivity and specificity are not abstract statistics. They decide whether a low-copy transgene is detected at all, and whether a near-identical off-target sequence is correctly excluded. At SCD, these characteristics are established and confirmed as part of method development and validation for qPCR, dPCR, NGS, and proteomics assays supporting biodistribution, vector copy number analysis, and biomarker discovery and validation.
What Sensitivity and Specificity Actually Mean in Assay Validation
Sensitivity describes how low a concentration an assay can reliably detect and quantify. Specificity describes how confidently an assay distinguishes the intended target from everything else present in a sample - related sequences, background nucleic acid, or matrix components.
Sensitivity is typically established through three linked measurements:
- Limit of blank (LOB) - the signal expected from a sample containing no target
- Limit of detection (LOD) - the lowest amount reliably distinguished from the blank
- Lower limit of quantification (LLOQ) - the lowest amount that can be measured with acceptable precision and accuracy
Specificity is established through selectivity testing - confirming the assay does not cross-react with related sequences, endogenous genes, or other biological material in the relevant matrix.
Why These Parameters Carry Extra Weight in Cell and Gene Therapy Bioanalysis
Advanced therapy programmes routinely push assays toward their performance limits. Vector genomes may be present at very low copy numbers in non-target tissues. AAV serotypes and transgene cassettes can share substantial sequence homology with related constructs. Engineered cell products and RNA therapeutics often need to be distinguished from endogenous, near-identical biology. In each case, an assay with weak sensitivity risks missing a real signal, and an assay with weak specificity risks reporting one that is not real.
This is why sensitivity and specificity are evaluated together, not separately - and why the appropriate level of validation is selected based on context of use (COU), the same principle SCD applies when matching a technology to a study's intended purpose.
How Sensitivity Is Established: LOB, LOD, and LLOQ
Sensitivity is confirmed using replicate testing across a dilution series, including blank and low-concentration samples. A method's LOD and LLOQ are only considered established once they are confirmed under the same conditions the assay will actually run under - same matrix, same extraction process, same operators - rather than under idealised conditions that will not hold during study sample analysis.
How Specificity and Selectivity Are Established
Specificity is confirmed by testing the assay against unspiked matrix and closely related off-target sequences, and by evaluating recovery in spiked samples at the lower limit of quantification. This step is especially demanding in areas like circulating miRNA biomarker work, where candidate targets can differ from non-target family members by a single nucleotide - a scenario SCD discusses in more detail in its post on Two-Tailed RT-qPCR for siRNA quantification, where single-nucleotide specificity is the deciding factor between a valid result and a false positive.
Platform Considerations: qPCR, dPCR, and NGS
Sensitivity and specificity are not established the same way across platforms:
- qPCR relies on a calibration curve and amplification efficiency, so sensitivity and precision are tied to how well that curve is characterised at low concentrations
- dPCR partitions the sample into thousands of individual reactions and counts positives directly, which often improves sensitivity for low-copy targets without depending on a standard curve
- NGS establishes sensitivity through read depth and specificity through alignment and mapping quality, making coverage design and reference selection central to performance
SCD's post on qPCR, dPCR, and NGS for siRNA bioanalysis walks through how platform choice affects sensitivity and specificity for a specific modality.
Why There Is No Single Regulatory Template - and How the Industry Is Closing the Gap
FDA and EMA guidance on qPCR and dPCR assay validation remains incomplete, leaving sponsors and CROs to reconcile different acceptance criteria across programmes. In response, SCD co-authored Recommendations for Method Development and Validation of qPCR and dPCR Assays in Support of Cell and Gene Therapy Drug Development, published in The AAPS Journal in 2024 alongside 37 industry experts from 24 organisations. The paper sets out shared best practices for assay design, development, and validation - including sensitivity and specificity criteria - for PCR assays supporting biodistribution, transgene expression, viral shedding, and related cell and gene therapy endpoints.
What Sponsors Should Ask When Evaluating Sensitivity and Specificity Claims
- How were the LOD and LLOQ determined, and were they confirmed in the actual study matrix rather than a simplified one?
- What cross-reactivity or selectivity testing supports the specificity claim, and against which related sequences or targets?
- Is the validation package fit-for-purpose or fully validated, and does that match the intended context of use for this data?
- How does the assay perform against near-identical off-target sequences relevant to this programme?
- Were sensitivity and specificity confirmed on the same platform and matrix that will be used for study samples, or extrapolated from a different setup?
From Detection Limits to Defensible Data
Sensitivity and specificity are where assay validation becomes concrete: the point where a method's ability to detect the right signal, and only the right signal, is tested against real data rather than assumed. Read alongside SCD's post on chain of custody in regulated bioanalysis, the two form a fuller picture: one establishes that a result can be trusted to reflect the correct sample, the other that it can be trusted to reflect the correct biology.
FAQ
What is the difference between sensitivity and specificity in bioanalytical assay validation?
Sensitivity is an assay's ability to detect low concentrations of the intended target, measured through the limit of blank, limit of detection, and lower limit of quantification. Specificity is an assay's ability to correctly distinguish that target from related sequences or background material, confirmed through selectivity and cross-reactivity testing.
What is the difference between LOD and LLOQ?
The limit of detection (LOD) is the lowest amount of target that can be reliably distinguished from a blank sample. The lower limit of quantification (LLOQ) is a stricter threshold: the lowest amount that can be measured with acceptable precision and accuracy, and is therefore the point at which a result can be reported as a number rather than simply "detected."
Why is specificity especially important in cell and gene therapy bioanalysis?
Advanced therapy products often share sequence homology with related vectors, endogenous genes, or family-member biomarkers. Without confirmed specificity, an assay risks reporting a signal from an off-target sequence rather than the intended therapeutic construct or biomarker.
Do qPCR, dPCR, and NGS assays establish sensitivity and specificity the same way?
No. qPCR sensitivity depends on calibration curve performance at low concentrations, dPCR establishes sensitivity through direct partition counting, and NGS establishes it through sequencing depth and alignment quality. Each platform requires its own validation approach rather than a shared set of criteria.
How does SCD confirm specificity during method validation?
Specificity is confirmed through testing against unspiked matrix and closely related off-target sequences, alongside recovery testing in spiked samples at the lower limit of quantification, so that results reflect performance under the same conditions used for study samples.
Is there regulatory guidance from FDA or EMA on qPCR and dPCR sensitivity and specificity requirements?
Formal FDA and EMA guidance on qPCR and dPCR validation remains limited. In the absence of a harmonised regulatory template, industry-led frameworks such as the 2024 AAPS Journal recommendations - co-authored by SCD with 37 experts from 24 organisations - are widely used to guide assay design and validation for cell and gene therapy programmes.
Need Support With Assay Validation?
Next in this series: Part 4 — Linearity, Co-linearity, Robustness, Ruggedness and Stability →
SCD offers GLP- and GCLP-compliant method development and validation services for qPCR, dPCR, and NGS assays, covering sensitivity, specificity, and the full suite of validation parameters from early feasibility through to regulatory submission packages.
