Limitation of results

Limitations of Genetic tests

In clinical laboratory testing there are potential “uncertainties” that can affect test results (for example; poor specimen collection or transport, patient related factors such as biological variation and the presence of drugs, or other interfering factors).

In addition, the analytical process itself is subject to some degree of inherent variability and this is often referred to as the “reproducibility” or “imprecision” of the method. Laboratories regularly monitor this by the use of internal quality control samples within each batch of analysis and by comparing the results of external quality assurance schemes designed to ensure that results are comparable with others laboratories using similar methods.

Providing relevant clinical details at the time that the request is made can also clarify the significance of a particular result or a change in results.

Measurement uncertainty data for specific tests can be provided to service users upon request.

Where relevant limitations are stated in the footer of the report. Further information is available on request from the lab if required.  Some examples of limitations are listed below.

Cytogenetic Results

Cytogenetic results from blood samples will be based on the analysis of a minimum of 3 banded cells.

The presence of mosaicism for any chromosome abnormality is not routinely investigated by the level of analysis performed unless dictated by the clinical referral reason or suggested by an observation during routine analysis.

The standard count for detection of a clone present at greater than 10% level is 30 cells. This is reported in the karyotype comments if carried out.

The nature of fluorescent in-situ hybridisation (FISH) means that interpretation of FISH signals can be challenging. There is the possibility of split signals (one signal appears as 2), false signals due to background fluorescence and weak signals resulting in them not being detected. For this reason sufficient numbers of metaphases or nuclei are examined to maximise the statistical validity of testing. The number of cells that should be examined depends on the type of FISH being performed.

Prenatal Samples

Microarray (SNP array) testing will detect aneu- and polyploidy but not balanced rearrangements and is limited in detecting mosaicism.  The resolution of SNP array analysis will be stated on the report.

Prenatal reports are based on the analysis of a minimum of three banded cells, and therefore are unlikely to detect mosaicism. It should be noted that the majority of samples sent for prenatal chromosome analysis are taken with a view to screening for common numerical chromosomal abnormalities, particularly Down syndrome. For technical reasons, the quality of structural analysis on such samples is often conducted at a lower level than that which is required to reliably detect small and unexpected chromosomal deletions and other rearrangements. Although many structural chromosomal abnormalities will be detected, those that fall below the limits of resolution of the analysis will be missed.

Oncology Samples

For oncology samples a normal cytogenetic result is based on the complete analysis of a minimum of 10 G-banded cells usually with the examination of at least a further 10 cells. The number of cells analysed in abnormal cases or previously abnormal cases may be increased or decreased according to the reason for referral. A normal in situ hybridisation result is based on the exclusion of a given abnormality in a minimum of 100 interphase cells. An indication will be given if minimum standards are not reached.

For all cytogenetics analyses the ability to detect subtle, unexpected chromosome rearrangements is governed by the resolution of banding achieved  This is intrinsically highly variable. If the quality of the preparation or extent of the analysis performed is not considered adequate for the reason for referral this will be indicated on the report.

DNA Sequencing

Sensitivity of DNA sequencing is over 95%. Rare cases of single nucleotide polymorphisms under the primer binding sites may lead to non-amplification of one allele when using Sanger DNA sequencing (mainly used for familial tests. The specificity is 99% where the variant has been previously reported.

Next generation sequencing technical sensitivity may be reduced for genes with pseudogenes or paralogs, mosaics and for copy-number variation >20 nucleotides. Dosage analysis screening for large deletions and duplications is performed using comparative depth of coverage of NGS data (DeCON software: sensitivity >0.999 and specificity 0.989). Where the variant is novel or there is limited or conflicting data available, it may be necessary to carry out further studies e.g. family testing, and it still may not be possible to reach a conclusion regarding pathogenicity.

All sequence and MLPA variants are classified using practice guidelines for variant interpretation: ACMG/AMP Richards et al 2015 Genet Med. 17(5):405-24 and ACGS Best Practice Guidelines for variant classification (https://www.acgs.uk.com/quality/best-practice-guidelines/). Variants may be subject to reclassification with new evidence or changes in variant classification guidelines. Only clinically relevant results are reported.

Low Level Mutations

In some circumstances it may prove difficult to detect mutations present at a low level, e.g. in cases of mosaicism or mitochondrial heteroplasmy or where there is a mixed cell population due to malignancy. Sensitivity of detection may be tissue-specific and in some cases alternative sample types may be required.

Non-paternity

An error in the diagnosis of disease status may occur if the true biological relationships of the family members being tested are not as stated. For example, non-paternity means that the stated father of an individual is not the true biological father. Any erroneous diagnosis in a family member can lead to an incorrect diagnosis for other related individuals who are being tested.