Skip to main content

Cytogenetics

FISH Quality Control

Quality Control (QC) is the framework that ensures the accuracy, reliability, and reproducibility of FISH testing. Because FISH results drive critical clinical decisions (e.g., chemotherapy selection), the laboratory must rigorously validate every probe before use and monitor every run with appropriate controls

Validate Probes & Establish Reference Ranges

Before a probe is used for diagnosis, it undergoes Validation to define its performance characteristics

  • Performance Metrics
    • Specificity (Mapping): Hybridizing to normal metaphases to confirm the probe binds only to the correct locus
    • Accuracy: Testing known positive and negative samples to confirm the probe yields the correct diagnosis
    • Sensitivity/Precision: Measuring how often the probe works and how consistent the results are between laboratory scientist s
  • Establish Cut-offs (NCV)
    • The Problem: FISH is not perfect. Technical noise (overlaps, truncation) causes “False Positive” patterns in normal cells
    • The Solution: The lab runs the probe on 20 Normal Samples to quantify this background noise
    • Calculation: Using statistical methods (Mean + 3SD or Beta-Inverse), a Normal Cut-off Value (NCV) is set
      • Example: If the NCV for a deletion probe is 6.0%, a patient with 4.0% deleted cells is reported as Normal, while a patient with 8.0% is Abnormal

Positive/Negative Controls

Controls are run concurrently with patient samples to verify that the assay conditions (Denaturation, Hybridization, Wash) worked correctly on that specific day

  • Negative Controls (Specificity)
    • Internal Control (CEP): A second probe (e.g., Centromere) mixed with the target probe. If the target is missing (deletion), the presence of the Control signal proves the assay worked (it wasn’t just a “no signal” failure)
    • Biological Negative: A normal sample run with the batch to monitor daily background noise levels
  • Positive Controls (Sensitivity)
    • Internal Normal Cells: Normal cells (e.g., T-cells in a myeloma sample) within the patient sample act as built-in positive controls. They should show the expected normal (2R, 2G) pattern
    • External Positive: A known abnormal sample (cell line) run to verify the probe can detect the specific abnormality
  • Batch Rule: If the daily Batch Control fails (e.g., no signal), the entire patient run is invalid and must be repeated