Low copy number DNA: The urgent need for standards in interpretation

Tweet

Previous

Eleven years has passed since the technique to develop profiles of ‘low copy number’ (LCN) DNAfirst emerged.Compared to the so-called ‘standard’ DNA test (SGM+), LCN testing potentially offers a much more sensitive testing technique for processing DNA. Low number or otherwise poor quality profiles can be enhanced through increasing the amplification cycles throughwhich DNAis processed using LCN testing. Forensic scientists remain deeply divided as to the integrity and reliability of the LCN testing technique.Conflict arises: whereas some scientists are prepared to analyse and offer an interpretation of samples where the amount of DNA available for analysis is very low, others discredit the task of attempting any such analysis where a sample is so limited. What is proposed here is that current problems with LNC testing can and should be addressed by introducing as a minimum standard a requirement that the amount of low level DNA is quantified.

             The difficulties to which LCN testing gives rise can be briefly outlined.The first problem is this. Only a limited quantity of DNA material is available for analysis in samples with low amounts of DNA – so-called ‘low template DNA’ (LTDNA). It is the very feature that the DNA available is limited which causes the rate of allele drop-out to increase. This results in failure to detect expected DNA components. An additional or separate problem where the amount of DNA is limited is that random imbalances among allelic pairs in a profile can result. Either problem results in difficulties with interpretation. 

 Few areas of forensic evidence give rise to the controversy that LCN testing does.Disagreement among experts is usually as to the application of a scientific technique. Differently, what is at stake with LCN testing is the very concept of deriving results from a sample where the amount of DNA (cellular material) is itself an unknown quantum. The LCN testing technique having been developed in 1999 by the Forensic Science Service (FSS), it has not been stewarded well by the FSS. In 2007, the Forensic Science Regulator reviewed the process for LNC testing. The remit of that review covered the use of all testing of low amounts of DNA. A particular flaw in the approach taken in that review lay with failing to appraise the LNC technique in the light of important developments in LNC processing which had taken place after 1999. Specifically, the chemicals used in the DNA analysing process had changed. As a result, the sensitivity of DNA had increased significantly. In approaching the review task without acknowledging this change, the Regulator overlooked an opportunity to recommend that parliament or a specialist advisory panel consider how best to implement minimum standards in LCN testing. 

 In an important editorial piece published in April this year, Denise Syndercombe Court, a leading forensic scientist based at Barts and The London School of Medicine and Dentistry,renewed the call for standardising the approach to LCN testing. She confirms that the lack of minimum standards in LCN testing persists as a problem:

 ‘Application of LCN without knowledge of the amount of DNA recovered from an item can lead to problems and errors in interpretation, and this approach [occurs] in LCN analysis undertaken today.’[1]

 It is precisely the potential forwide variation among expert opinion as to LCN testing that first prompted the courts to express serious concern as to the reliability of the LCN testing. Analysis of DNA by way of LCN testing was stopped in December 2007, directly as a result of concerns raised by the Crown Court in Northern Ireland in R v Hoey, NICC 49 (20 December 2007) in connection with DNA collected after the Omagh bombing incidents.[2]

             In 2008, the Caddy report,[3] which reviewed low template DNA analysis, concluded that the multiple approaches to analysing LTDNA were fit for purpose. Nevertheless, that report specifically recommended that the amount of DNA in a sample undergoing a testing process should be quantified. Quantification would serve to differentiate between sample results and allow for informed, properly evidenced interpretations of those results. This is key to assisting the jury in their task of interpreting the results of LCN testing. To date, there has been no move to standardise LCN testing by way of a quantification requirement. Following judgment in a conjoined appeal heard last year, the Court of Appeal (Thomas LJ, Kitchin and Holroyde JJ) took the opportunity to closely examine LTDNA techniques.[4]The court did not specifically need to consider or rule on evidence concerning LTDNA. The first appellant’s general attack on the reliability of low template DNA obtained using the low copy number process resolved itself. Post-hoc quantification of the remaining DNA in that appeal was not below the permitted ‘stochastic threshold’. While a precise stochastic threshold does not currently exist, agreement clusters around a threshold DNA level of 100 and 200 pictograms (pg).[5] (The case also establishes that a forensic science officer with scenes of crime experience could give an opinion as to possible explanations for the presence and site of DNA).

 It is regrettable, having pursued an examination of the LTDNAprocess, thatthe court did not go further. It could have endorsed the findings of the Caddy report and itself laid down that quantification of low level DNA is a criterion to its admissibility. Despite thefresh call by Syndercombe Court for requiring as a minimum standard that samples are quantified, the CPScontinues to make use of undifferentiated LCN testing. The CPS webpage on LCN testing omits any reference to distinguishing between samples that have and have not been quantified:

 ‘The [LCN] test can obtain a profile from as few as 5 – 10 cells, or from DNA that is in poor condition. This could be the amount of DNA left on a cup by drinking from it or on a pen by writing with it.’[6]

 A change of culture is required. The problem is that there is no standardised approach to LCN testing. Unlike fingerprint evidence, the admissibility of DNA evidence is not currently subject to a standardised approach. Unless a fingerprint lift exhibits at least the required ‘7 ridge detail’, it cannot be adduced at criminal trial. This stark result is warranted by insistence on a minimum standard.

 Conclusions

 The potential of LCN testing has been heralded as a radical, promising development. This billing has a particular context, DNA having long been heralded the ‘gold standard’ of forensic techniques’.[7] However, LCN testing stands apart from other forensic techniques in terms of its regulation and admissibility in the criminal courts. There is for LCN testing no equivalent of the minimum admissibility threshold standard which the ‘7-ridge detail’. Fresh calls for concern that LCN testing is still being undertaken in the absence of a safeguarding minimum threshold – that low copy number samples are quantified –render this untenable.

The effectiveness of an interpretative forensic technique is premised on the reliability of that technique. In the criminal courts, the requirement for reliability translates into a requirement that a testing technique can be reproduced. This is the basis on which expert evidence potentially has probative value: results can be attributed to a method and process. That technique can then be understood in terms of strengths and limitations. To understand, one must be able to follow.

Processes and results in a criminal trial need to be understood and followed by a jury. It is crucial that results are attributed to a particular process which can be explained to twelve lay persons. Criminal lawyers are familiar with this concept since the concern with attribution manifests itself in other guises – such as continuity of evidence. It follows that adducing evidence of LCN testing should be subject to the minimum standard that the amount of DNA recovered is quantified, rather than generalised as ‘low copy number’. This standard should be recognised by the courts as a criterion for the admissibility of results generated by LCN testing (subject to relevance and fairness).   

 Abigail Bright is a pupil at QEB Hollis Whiteman. She was a teaching fellow in criminal law at UCL (2009-2010) and recently completed the Dip.FMS(diploma in forensic medical sciences) at Barts and The London School of Medicine