Conventionally fixed thresholds are used to detect chromosome aberrations in CGH.
Ratio profiles that deviate for instance 15% or 20% from
ratio1.0 will typically be regarded as aberrant. It is, however, common knowledge that this method often detects false positive results in certain chromosomal regions. Consequently these regions are often excluded from interpretation.
We have developed a detection method where ratio profiles are evaluated by a dynamic standard reference interval instead of fixed thresholds. The dynamic standard reference interval is based on an average of
16 normal CGH analyses. The basis for using a normal standard is observations of ratio profiles of hybridizations with normal DNA. These
ratio profiles deviate from ratio 1.0 in a systematic pattern that is characteristic for every single chromosome. The degree of deviation differs from analysis to analysis, however, it is always correlated for all chromosomes in the individual analysis.
The characteristic deviations of the dynamic standard reference are
correlated to the banding pattern of inverted DAPI stained chromosomes.
The dynamic standard reference moves normally to the left at dark bands and to
the right at light bands when our CGH-protocols are used (see fig. 1),
but from time to time we also observe cases where it moves in the opposite
direction (like being reflected in the ratio 1.0 axis).
The dynamic standard reference interval can be scaled to fit any analysis (see
HR-CGH aberrations are detected by comparing the 99.5% confidence intervals of the mean ratio profiles of the test sample to the 99.5% dynamic standard reference interval. Regions where the two sets of intervals do not overlap are considered as aberrant (see figure):
More details about the HR-CGH method are described in the section: HR-CGH technical notes
False positive results appear seldom by HR-CGH and no regions need exclusion. The sensitivity is increased compared to detection by fixed thresholds. A 3 Mb deletions on chromosome 11 was detected by HR-CGH and so was several Prader willi/Angelmann deletions at chromosome 15 which are considered to be approximately 4 Mb (Kirchhoff et al., 1998; Kirchhoff et al., 1999)
We have used HR-CGH in several tumor studies as well as in the clinical cytogenetic routine laboratory since 1997.
Both applications have shown that compared to fixed thresholds HR-CGH detects more and smaller aberrations with higher specificity.
The HR-CGH method was developed on a Magican analysis system (Applied Imaging) and most of the cases up to the end of year 2002 were analyzed on this system. In collaboration with Applied Imaging (Genetix), the concept of dynamic standard reference intervals has been implemented in the CGH software of the CytoVision and is thus commercially available (CytoVision™ cytogenetic workstation Genetix - Applied Imaging).