Applied Molecular Evolution: The Evidence in Favor of the Tripoli Six

The five nurses and one doctor (the Tripoli Six) accused of infecting hundreds of patients with HIV in Libya are awaiting the verdict of their trial, expected to be handed down on December 19. The second trial concluded on November 4 -- the original guilty verdict was overturned. I previously mentioned that the molecular evidence (DNA sequences from HIV taken from the patients) did not support a guilty verdict. Nature has published the results of an analysis of those DNA sequences that suggests the doctors were not responsible for the HIV infection. A description of that evidence is below the fold.

Along with the HIV infection, many of the children were also infected with the hepatitis C virus (HCV). The researchers examined HIV sequences from 44 infected children and 61 HCV sequences. In addition, they collected other, closely related HIV and HCV sequences from available data bases (this is why all DNA sequence data should be made publicly available) and constructed phylogenies to examine how the sequences are related.

Two types of phylogenetic evidence would exonerate the Tripoli Six. First, if the sequences from the Libyan children clustered into multiple clades, it would suggest that there were multiple sources of infection. A deliberate attempt to infect the children from a single source should lead to a single clade containing all the sequences from the affected children. Second, one can use the amount of sequence diversity of the virus from the children to date the first introduction of a particular viral lineage into the hospital. The greater the sequence diversity, the further in the past the infections began.

The phylogeny to the right shows the relationships of the HIV strains taken from the children and closely related strains retrieved from databases. The 44 HIV sequences from the Libyan hospital form a monophyletic clade (the big red triangle). Therefore, there is no evidence for multiple sources of infection -- the virus invaded the population only once.

The researchers also performed a series of analyses to estimate the most recent common ancestor (MRCA) of that clade. The accused medics arrived in March of 1998, but all possible models used to estimate the date of the MRCA indicate that the infection was introduced prior to 1998. In fact, most of the analyses estimate the probability that the clusters originated after March of 1998 as zero. That is, there is near certainty that the HIV infection began prior to the arrival of the medics.

The HCV phylogenies tell a slightly different story. Three monophyletic clusters contain the majority of the HCV sequences, with four other sequences scattered throughout the tree. The HCV sequences are so diverse that they must be put into two separate trees, as shown below.

As with the analysis of the HIV sequences, these clusters all originated prior to March 1998. Three distinct HCV strains were circulating in the population prior to the arrival of the foreign medics. Additionally, 70% of the lineages samples were present prior to 1998. Either the medics infected the children with an extremely diverse sample of viral strains or the infection had begun way before their arrival.

The evidence presented above indicates that the Libyan hospital had a major problem maintaining a sterile environment. The hospital was a haven for HIV and HCV infection, but this was not due to the six foreign medics. They could not have introduced these viruses into the hospital -- the chronology does not work out. The Libyan judicial system has no evidence, other than their presence at the hospital, to convict the medics of infecting the children.

Many thanks go out to Declan Butler and revere for spreading the word.

More coverage can be found at the following places:

• Declan Butler's news item in Nature.

• Feedback from top researchers on the new data.

• At ScienceBlogs.