The use of Low Copy Number (LCN) DNA in forensic science, and the criminal justice system, has been affected in the wake of Justice Weir’s ruling in the trial of Mr Sean Hoey (The Omagh Bombing Trial)1; “I am not satisfied that the publishing of two journal articles describing a process invented by the authors can be regarded, without more, as having "validated" that process for the purpose of its being confidently used for evidential purposes.”

So began a sequence of events that leaves the UK as one of the only adversarial justice systems in the world to routinely use LCN DNA as evidence. Professor Brian Caddy recently released his much anticipated review of “Low template” DNA profiling,2 a technique pioneered by the Forensic Science Service Ltd® and used in thousands of prosecutions across Great Britain. The Review was commissioned and has now apparently been endorsed by the new Forensic Regulator in response to the intense criticism leveled at the technique and its purveyors, and voiced by Weir J in the Omagh Bombing Trial. His Honour found that there were “very many unsatisfactory matters” to do with the DNA evidence and LTDNA in particular. This article reviews the problems with LTDNA and explores how they have now been compounded by the Review, which in our opinion failed in its remit to the Forensic Regulator and has provided an unsound basis for the continued use of LTDNA profiles in criminal prosecutions.

The LTDNA technique, at least so far as the UK and the Review are concerned, encompasses two different approaches to the recovery and analysis of extremely low amounts of DNA. The technique called LCN was developed and is currently sold by the Forensic Science Service Ltd®. Another technique, using a different principle, was developed by LGC Forensics Ltd. The standard DNA profiling kit used in the UK (SGM+) is designed and validated to work using a specific amount of DNA (nanogram amounts, or 0.000000001g). Low Template DNA profiling pushes the amount of DNA used as a starting point to 1,000 times below that limit (picogram amounts, or 0.000000000001).

Three problems emerge from this extreme sensitivity. First, it is frequently impossible to know how such a small amount of DNA came to be in the place where it was discovered and second, it is difficult to achieve reproducible results, and even then, to know what the results mean in terms of the contributing profile(s).

Transfer and Contamination

The first problem arises because each of us has about 1014 cells in our body, each with a full DNA profile packed inside them. Everywhere you go you leave your DNA, and your DNA goes places you have never been. DNA can be transferred from you to someone else, and from that someone to someone or somewhere else that you may have never been. This process is called transfer. It is yet to be established how extensive transfer can be and what factors and influence it and to what extent. The LTDNA technique is therefore very susceptible to identifying DNA that is not truly related to the crime under investigation.

Furthermore, if crime scene samples are not handled using extremely thorough precautions, the risk of contaminating a sample with DNA from other crime scene samples, from police, from forensic scientists or from anyone who may have had (perfectly innocent) contact with the scene or sample is extremely high. The defence must now carefully examine the continuity of such evidence (and indeed this was a focal point of the outcome in the Omagh Bombing Trial).

Reproducibility

No agreement exists, even among the few providers of the service, about how the results of LTDNA profiling should be interpreted. In effect, the DNA profiles reported for LTDNA cases are likely to depend on which laboratory the material is sent to, which is clearly not the hallmark of a “robust” scientific technique.3

There is no doubt that current technology is capable of amplifying very small amounts of DNA, and certainly other laboratories have developed methods to do so. The question is whether we allow our enthusiasm for ground-breaking science to overcome the fundamental principles of good science – that a technique should be validated, reliable and reproducible. The Sally Clark case, the Birmingham Six and the Guildford Four are just a few of the cases which provide a stark reminder of the weight which scientific evidence and experts can carry in criminal trials.

The Review

The Caddy review was released in April 2008. Despite the limited time frame and the fact that the review panel chose to consult only with the FSS Ltd® and the police, it concluded that the FSS Ltd’s® LTDNA technique was “robust” and “fit for purpose”, although that purpose was never defined.

Worryingly, the contents of the Review read like a directory of the problems with, and shortcomings of, the LTDNA technique. Of even more concern is the fact that the conclusions of the Review appear to ignore the contents in order to provide a completely unjustified clean bill of health to the technique. If this Review is to represent the scope and the depth of the work of the new office of the Forensic Regulator then we should not be optimistic for the future of the quality, and in particular the reliability, of science apparently approved by the Crown for use in British courts.

The Review team did not consult anyone who had expressed contrary opinion on the merits of the FSS Ltd’s® LTDNA technique; in fact, the panel admitted to speaking with only the organisations selling the technique and to the police as ‘customers’. This is despite the Home Office’s own stated view that where commercial products are being “sold” to the police, “the police and others are not well placed to evaluate the quality of the service provided across the range of scientific disciplines…there needs to be a mechanism to identify poor providers or services and protect the police and Criminal Justice System (CJS) from them before procurement…and the police are not the only user of forensic science and the quality standards must reflect the needs of other stakeholders in the CJS.”4

The degree of involvement between the FSS Ltd® and the Review is not clear, however a scientific report produced for a criminal appeal case by the FSS Ltd® in March 2008 contained the statement: “Preliminary indications are that [the Review] makes no significant criticisms of the LCN technique”. This means that the FSS Ltd® and its staff had knowledge of the results of the Review at least three weeks prior to its release. This despite the fact that the Home Office specifically called for the Regulator to be “independent of any forensic science provider”,5 which presumably was meant to include the purveyors of the techniques and theories being scrutinised.

Not only was the review conducted on the basis of an unsatisfactory consultation process, but there has been no opportunity for the international scientific community, nor anyone other than the three members of the Review panel, to assess the data claimed to support the validation of the technique.

Many jurisdictions apply a series of tests to all scientific evidence before it is allowed into court:6 Has the science been tested? Has it been published and peer reviewed? Is the error rate of the technique measurable and known? Is the theory or technique generally accepted within the relevant scientific community? Although these tests are not directly applied in the UK, this kind of scrutiny is used by the law to try to identify when a scientific technique or theory has not been properly verified or validated.

Except for the Netherlands, LTDNA profiling has not been purchased by any forensic science service providers outside the United Kingdom and is not used for evidentiary purposes by laboratories in any of the jurisdictions with equivalent legal systems to ours (such as the USA, Canada, Australia). Claims that the UK is simply leading the field are scoffed at by scientists internationally (and by some in the UK). Scottish and Irish state laboratories have no plans to introduce the LTDNA technique. The senior scientist at the FBI warned of the problems years ago. To suggest that the FBI and similar organisations have neither the resources nor the expertise to do what they already do, (as the LCN technique requires no more than the routine equipment), and that they have lagged behind for the last ten years, is stretching credibility to the limit.

Mixtures and Interpretation

Among several problems identified by critics, and repeated by the Review, DNA in forensic work frequently involves mixtures. Even standard DNA profiling methods (using large samples of DNA) present difficulties for scientists trying to determine how many contributors there may be in a mixed DNA sample. The problems are exacerbated in LTDNA profiling when swabs are taken to detect samples which aren’t even visible to the naked eye. If further evidence was needed, the Regulator kindly provides it;

“2.5.6. I agree there needs to be an agreed approach to the interpretation of such

profiles. The intention will be to produce a single proposal which addresses the issues of stochastic effects and mixture interpretation. This will include, but not be limited to, the following:

• the process by which the analytical results are interpreted to produce profiles;

• the manner in which profiles, and in particular mixed profiles, are interpreted to generate evidential weight;

• the manner in which issues around transfer and persistence of DNA are addressed;

• the manner in which issues of extrinsic or contaminant DNA are addressed;

• the way in which all of these factors are considered in relation to the circumstances of the case;

• the reservations or limitations that have to be considered and how these are to be reported to the court in the light of the case circumstances; and

• the scientific and statistical basis for the approach adopted.”
  

Only the devotee could, as stated by the FSS representative on Radio 4’s Today programme, consider this a ‘ringing endorsement’ of the method.

Despite this, the Review mentions mixtures only three times in 35 pages, and airily recommends “more work” be done on the interpretation of mixtures. In these circumstances, surely LTDNA profiling is far from being a “robust” technique that is fit for the purpose of identifying and prosecuting alleged offenders.

Conclusions

The Review lacks sufficient authority to allow any weight to be attached to its findings until all of the defects identified by the Review and other scientists have been rectified, and the clear disagreements among providers and scientists nationally and internationally have been resolved. The Report can not be accorded any scientific significance until the data upon which the opinions are based are made available to all, and have met general scientific approval.

Even by the most generous interpretation, there is clearly no general agreement in the scientific community about the reliability of LTDNA analyses as performed by the FSS Ltd®. The most basic legal tenets regarding the acceptability of forensic science evidence in court, such as peer review, measurable errors and acceptance by the scientific community, have not been met in relation to LTDNA. The puzzling endorsement of LTDNA by the Review may now mean that the UK will see evidence presented that would not meet the standard of other comparable technologically and legally advanced systems.

The description of the LTDNA technique sold by the FSS Ltd® as “robust” or “fit for purpose” is a denial of the serious scientific questions which remain about the reliability and validity of the technique. Taking the review as the ‘final word’ on the technique is an error with potentially serious consequences for the reputation of British science and for the Criminal Justice System.

The Home Office and CPS appear to have adopted an attitude of judging LTDNA (LCN) on a case by case basis. To paraphrase, “It’s evidence, Jim, but not as we know it.”

Professor Allan Jamieson
Dr Rhonda Wheate

The Forensic Institute
www.theforensicinstitute.com