Skip to main content

Cytogenetics

Score & Signal Patterns

The analysis phase of Fluorescence In Situ Hybridization (FISH) is the translation of fluorescent optical data into a genetic diagnosis. Unlike the black-and-white bands of a karyotype, FISH signals are discrete points of light that represent specific DNA sequences. The laboratory scientist must strictly adhere to scoring criteria to distinguish between true genetic abnormalities (signal patterns) and technical artifacts (noise), applying specific logic based on the probe strategy used

General Scoring Principles (The Rules of Analysis)

To ensure statistical validity and reproducibility, laboratories follow strict protocols defined by the American College of Medical Genetics (ACMG) and internal validations

  • Cell Selection Criteria
    • Non-Overlapping: Nuclei must be distinct. If two nuclei overlap, their signals merge, creating a “False Polysomy” (e.g., counting 4 signals in what looks like one cell but is actually two cells)
    • Intact Boundaries: The nuclear border must be visible to ensure no signals were lost due to nuclear rupture (False Monosomy)
    • Signal Quality: Signals must be bright, distinct, and of similar intensity. Diffuse “hazy” spots or tiny “pinpricks” of background noise are excluded
  • The Scoring Count
    • Constitutional/Prenatal: Typically 50 cells are scored
    • Oncology (Leukemia/Lymphoma): Typically 200 cells are scored. This higher number is required to detect low-level residual disease or minor clones (mosaicism)
    • Tissue (FFPE): Typically 20–60 cells are scored due to the difficulty of finding intact, non-truncated nuclei
  • The Cutoff Value
    • A sample is not “Positive” just because one cell has an abnormal pattern. Every probe has a validated Normal Cutoff Value (NCV)
    • Example: If the NCV for BCR:ABL1 fusion is 1.5%, a patient with 0.5% positive cells is reported as Normal: (within the range of technical background error), whereas a patient with 5.0% is reported as Abnormal

Enumeration Strategy (Counting Aneuploidy)

This strategy uses Centromere Enumeration Probes (CEP) to determine chromosome copy number. The interpretation is a direct count

  • Normal Pattern (Disomy): Two signals (2R)
  • Abnormal Patterns
    • Monosomy (Loss): One signal (1R). Note: In FFPE tissue, monosomy must be interpreted with caution due to nuclear slicing (truncation).
    • Trisomy (Gain): Three signals (3R)
    • Tetrasomy: Four signals (4R)
  • Interpretation Nuance (Split Signals/Doublets)
    • In Interphase, cells in the S-Phase: or G2-Phase of the cell cycle have replicated their DNA. A single centromere may appear as two small dots connected by a thread or positioned very close together
    • The Rule: If the distance between two spots is less than the width: of the spots themselves, they are counted as one signal (a doublet). If they are further apart, they are counted as two distinct signals. Counting doublets as two signals leads to False Trisomy

Dual-Color Dual-Fusion (Translocation)

Used for reciprocal translocations (e.g., BCR:ABL1). The probe contains two colors (Red and Green)

  • Normal Pattern: Two Red, Two Green (2R, 2G). The chromosomes are separate
  • Classic Positive Pattern (Dual Fusion): One Red, One Green, Two Fusions (1R, 1G, 2F)
    • Interpretation: The “Fusion” (Yellow) signal is created by the overlay of Red and Green. The “Dual Fusion” indicates a balanced reciprocal translocation where both derivative chromosomes [e.g., der(9) and der(22)] are present
  • Atypical Positive Pattern (Single Fusion): One Red, One Green, One Fusion (1R, 1G, 1F)
    • Interpretation: This represents a translocation with a deletion: of the reciprocal breakpoint. One derivative chromosome was lost or the breakpoint occurred outside the probe region. This is still a “Positive” result but may carry a worse prognosis (e.g., deletion on the der(9) in CML)

Break-Apart Strategy (Rearrangement)

Used for genes with multiple partners (e.g., KMT2A or MLL). The probe flanks the gene with Red (5’ end) and Green (3’ end)

  • Normal Pattern: Two Fusions (2F). The 5’ and 3’ ends are adjacent, so the colors merge into yellow
  • Classic Positive Pattern: One Fusion, One Red, One Green (1F, 1R, 1G)
    • Interpretation: One allele is normal (Fusion). The other allele has broken apart; the 5’ end (Red) moved to one chromosome, and the 3’ end (Green) moved to another
  • Deletion Pattern: One Fusion, One Red (1F, 1R)
    • Interpretation: This indicates the 3’ end of the gene (Green) was deleted. While technically a “rearrangement,” it is scored as an abnormality involving the target gene

Amplification Strategy (Copy Number)

Used for oncogenes (e.g., HER2, MYCN). Requires counting the Target (Gene) and the Control (Centromere)

  • Normal Pattern: 2 Targets, 2 Controls (Ratio = 1.0)
  • Polysomy: 3 Targets, 3 Controls (Ratio = 1.0)
    • Interpretation: The cell has an extra whole chromosome (Trisomy). The gene is elevated, but it is not: amplified relative to the chromosome
  • True Amplification: Many Targets, 2 Controls
    • Interpretation: The gene copy number is disproportionately high
    • Cluster Method: If the gene signals are so numerous they form a giant glowing cloud, it is scored as a “Cluster” (impossible to count distinct dots) and is Positive
    • Ratio Method: The analyst counts signals in 20 cells and calculates the ratio: Total Target Signals / Total Control Signals
      • Ratio < 2.0: Negative (Non-amplified)
      • Ratio \(\ge\) 2.0: Positive (Amplified)

Troubleshooting Signal Artifacts

  • Cross-Hybridization: If a probe binds to a similar sequence on a different chromosome (homology), faint “background” signals appear. These are typically smaller and dimmer than true signals and should be ignored based on intensity
  • Autofluorescence: Red blood cells or necrotic debris may glow. If a signal is visible through both the Red and Green filter simultaneously (appearing orange/brown), it is likely autofluorescent debris, not a probe signal. True probes usually emit in only one distinct wavelength (unless it is a fusion)
  • Doughnuts: If a signal looks like a hollow ring, the chromatin was over-denatured (heat damage). These signals are often unreliable and the slide may need to be repeated