Phylogenetic analyses that helped convict a Spanish anesthetist of infecting 275 of his patients with Hepatitis C virus over a ten year period fifteen years ago have just been published in BMC Biology. The results of these analyses were presented as evidence in a high-profile court case that itself spanned two years (September 2005 to May 2007) and in which the accused protested his innocence and subsequently appealed. The sentence of nearly 2,000 years’ imprisonment was upheld however, and with legal restrictions now lifted, the authors have published a full account of their work.
The outbreak of HCV infection in Valencia that led to the court case was exceptionally large and complex, and the phylogenetic evidence would not have been sufficient by itself to identify the source. The issues raised by the use of phylogenetic inference in court are discussed more generally in an accompanying commentary by Anne-Mieke Vandamme, who two years ago convened a working group of the world’s leading experts on this subject, and Oliver Pybus, who uses phylogenetic analyses in his own work on the molecular epidemiology and epidemic history of HIV and HCV.
Convictions obtained through DNA fingerprinting, where samples from the same person are identical, are now commonplace, but Vandamme and Pybus point out that sequence-based phylogenetic analysis of highly mutable viruses such as HIV and Hepatitis C virus is much less straightforward. These methods have become increasingly sophisticated, and in addition to proving invaluable in reconstructing the likely evolutionary history of viral epidemics, they are starting to be used more frequently in forensic settings. Tracing the source of the infection through a phylogenetic analysis, however, presents challenges of interpretation that are formidable, because HCV (like HIV) evolves within a single infected individual. This challenge is particularly compounded when, as in the Valencian case, the outbreak occurred over a considerable length of time.
Phylogenetic analyses reconstruct the most likely path of evolution leading to the diversity of viral sequences present at any one time, allowing an estimation of the evolutionary distance between the viruses carried by any two individuals, and the probability that they derive from a common source. In some cases, when the sequences found in one individual nest within a greater diversity of sequences found in a putative source, they provide evidence for the direction of transmission from one individual to another. But as Vandamme and Pybus point out, it is always possible that sequences present in infected individuals have come from some source that hasn’t been sampled. A further complication arises from the fact that HCV replicates in the liver, so samples taken from blood (as in the Valencian case) may not reflect all of the sequences present in the individual. This could explain why the phylogenetic tree constructed by Gonzalez-Candelas and colleagues did not in fact point to the anaesthetist as the source of the outbreak (though he was clearly a part of it). Strong evidence from other sources however clearly pointed to the anaesthetist as the source; and indeed, Vandamme and Pybus emphasize that it is generally the case that sequence-based viral phylogenies are used only to support evidence of other kinds in convictions.
The sequence-based analysis did however clearly indicate that 47 patients, who on the basis of their medical history could have been infected with HCV-1a by the anesthetist, in fact carried sequence variants indicating they were far more likely to have been infected from an alternative local source.
Not surprisingly then, some experts argue that viral phylogenies should only be used to rule putative sources of infection out (or cast serious doubt upon them), not to assign culpability.
Written by Penelope Austin, Associate Editor for BMC Biology.