Objective: Fluorescence in situ Hybridisation (FISH) is a widely used and useful cytogenetic technique for diagnosis and monitoring treatment responses in chronic myeloid leukaemia (CML). The positive or negative FISH result is interpreted based on individual standard normal cut off percentage, but universal cut off is unaddressed. We aimed to determine the performance of the FISH technique based on our routine normal cut off.
Materials and Methods: A retrospective descriptive and analytical study was conducted on CML patients followed-up at Srinagarind Hospital. Data has been collected from laboratory records over the past ten years. We excluded patients who were not completely tested FISH and chromosome or RQ-PCR.
Results: 675 FISH tests from 255 CML patients were analyzed. Specimens were mainly from bone marrow (98.2%) Chromosome analysis (G-banding) showed no metaphase in 31.9%. FISH test was positive in 99 samples with normal cut off at 9.7%. The falsenegative rate of FISH was 2.6% (0.65% by using the standard cut off at 1%) and the false positive rate was 3%.
Conclusion: The FISH results should be interpreted carefully. Undue high level of normal cut off causes a high false-negative rate of FISH. The standard normal cut off for double signal FISH (D-FISH) was 1%. However, the cutoff FISH interpretation should be individually set and regularly validated. The FISH result should also be construed with G-banding, and RQ-PCR, which will improve the accuracy and will prevent misdiagnosed of CML.

Keywords: Fluorescence in situ Hybridisation (FISH); Chronic Myeloid Leukemia (CML); Cut off; False positive; False negative