Skip to Content

Are you based outside North America? If yes, switch to our global website →

Validation of signal pattern cut-off levels

Want to download this as a PDF? Download now


In December 2018 Oxford Gene Technology received FDA clearance for eight Cytocell Aquarius® fluorescence in situ hybridization (FISH) probes:


These kits are FISH tests used to detect common chromosomal rearrangements found in 3:1 methanol/acetic acid fixed bone marrow specimens from patients with AML and MDS.

In order to fulfil the high standards required for an FDA clearance, all eight Cytocell FISH probes in this submission underwent stringent performance assessments to underpin the probe performance and ensure that the products were safe and effective for their intended use. A summary of these performance assessments can be found in each of the pack inserts of these probe sets.

Practical Advice

The pack inserts for these probe kits contain information including guideline cut-off values for the expected abnormal signal pattern(s) for each probe set, but recognize that other signal patterns may be possible in the presence of aneuploid or unbalanced rearrangements.

The purpose of this support document is to aid customers wishing to introduce these tests into their laboratory, by providing calculation guidance and giving example cut-off values for non-standard signal patterns calculated from the analysis data included in the Cytocell probe FDA submission. This data is intended to be used by laboratories solely as a benchmark when determining their own internally-generated cut-off values. These cut-off values will be required when assessing and reporting samples with non-standard signal patterns in accordance with standard laboratory accreditation guidelines.

When introducing new FISH tests into the laboratory, cut-off values for each probe set should be generated for use specifically in the validating laboratory. There are a number of published methodologies available and it is up to the
laboratory to select the appropriate one in accordance with their standard practices1,2,3. This assessment will result in a data set of analyses which detail the frequency of all signal patterns seen on these specimens. This data set can then be used to generate cut-off values for any signal pattern required. If a signal pattern has not been seen in the dataset, then the frequency is counted as zero; if a particular signal pattern was seen in the dataset, record the highest number of cells seen in one individual specimen. One available method to calculate the cut-off from the data obtained is to use the BETAINV Microsoft Excel formula:

Table 2For example, to calculate a cut-off with a 95% confidence level in which six false-positive cells for a given signal pattern were seen in a specimen with 200 nuclei examined, using Microsoft EXCEL enter: (BETA INV(0.95,6+1,200))*100; the result is 5.67% cut-off.

If sample results are tabulated in Microsoft Excel with the signal patterns seen, the BETAINV function can be embedded at the bottom of the spreadsheet to provide a comprehensive set of cut-off values for all signal patterns.

Tables 1-8 show illustrative data generated during the FDA submission process for these probes.

Further Guidance

It important to note that reporting the presence of non-standard signal patterns in a specimen will be at the discretion of the testing laboratory and should be provided on the laboratory report for informational purposes only. Similarly, the presence of abnormal signal patterns should not be extrapolated to indicate the presence of particular karyotype or rearrangement, for example, when using the USA-LPH026 AML1/ETO (RUNX1/RUNX1T1) Translocation, Dual Fusion FISH Probe Kit, the presence of cells with a 2 red, 3 green signal pattern higher than the cut-off level for this pattern will be consistent with, but not indicative of the presence of trisomy 8.

Regulatory Disclaimers

This technical supplement is not intended to replace the information in the pack insert. Cut-offs for all FISH probe kits should always be validated by the user. These FISH probe kits are not validated for use outside of the stated
intended use.


  1. Wiktor AE, Van Dyke DL, Stupca PJ, et al. Preclinical validation of fluorescence in situ hybridization assays for clinical practice. Genet Med.2006;8(1):16-23.
  2. Mascarello JT, Hirsch B, Kearney HM, et al. Section E9 of the American College of Medical Genetics technical standards and guidelines: fluorescence in situ hybridization. Genet Med. 2011;13(7):667-675.
  3. CLSI Fluorescent in-situ Hybridization Methods for Clinical Laboratories; Approved guideline – Second Edition. CLSI Document MM07-A2. Wayne, PA: Clinical and Laboratory standards institute 2013.

Example Data

Table 3 v2Table 1. Cut-off values for signal patterns seen in USA-LPH013: MLL (KMT2A) Breakapart FISH Probe Kit 

Table 4Table 2. Cut-off values for signal patterns seen in USA-LPH017: P53 (TP53) Deletion FISH Probe Kit

Table 5Table 3. Cut-off values for signal patterns seen in USA-LPH020: Del(20q) Deletion FISH Probe Kit

Table 6Table 4. Cut-off values for signal patterns seen in USA-LPH022: CBFβ (CBFB) / MYH11 Translocation, Dual Fusion FISH Probe Kit

Table 7Table 5. Cut-off values for signal patterns seen in USA-LPH024: Del(5q) Deletion FISH Probe Kit

Table 8Table 6. Cut-off values for signal patterns seen in USA-LPH025: Del(7q) Deletion FISH Probe Kit

Table 9Table 7. Cut-off values for signal patterns seen in USA-LPH026: AML1/ETO (RUNX1/RUNX1T1) Translocation, Dual Fusion FISH Probe Kit

Table 10Table 8. Cut-off values for signal patterns seen in USA-LPH036: EVI1 (MECOM) Breakapart FISH Probe Kit

*Cut-off rounded up to the next whole decimal

This document and its contents are © Oxford Gene Technology IP Limited – 2019. All rights reserved. Trademarks: OGT™ (Oxford Gene Technology IP Ltd); Cytocell®, Aquarius® (Cytocell Ltd), Cytocell FDA cleared probes: For In Vitro Diagnostic Use. Rx only. Product availability may vary from country to country and is subject to varying regulatory requirements. Please contact your local representative for availability.