H5N1 Genetic Predisposition Media Myth
Recombinomics Commentary
December 23, 2007
Some experts believe that a sick person is more likely to pass the infection on to a blood relative rather than an in law if both are in close and continuous physical contact with him/her.
The above comment is another widely circulated media myth. It shows up in wire service stories, and well as ProMed commentaries. However, the speculation has little scientific basis.
The speculation became more widespread after the Indonesia cluster in Karo. In that cluster, there were 8 family members who died or were H5N1 confirmed. Since all infected were blood relatives, and genetic predisposition was postulated and the lack of clusters involving a husband and wife was cited in support.
However, the Karo cluster is easily explained by close contact with patients who are fatally infected with a high H5N1 viral load Data supporting close contact comes from the WHO update:
The newly confirmed case is a brother of the initial case. Specimens were taken on 21 May and flown the same day to Jakarta. Tests run overnight confirmed his infection. His 10-year-old son died of H5N1 infection on 13 May. The father was closely involved in caring for his son, and this contact is considered a possible source of infection.
Although the investigation is continuing, preliminary findings indicate that three of the confirmed cases spent the night of 29 April in a small room together with the initial case at a time when she was symptomatic and coughing frequently. These cases include the woman’s two sons and a second brother, aged 25 years, who is the sole surviving case among infected members of this family. Other infected family members lived in adjacent homes.
All confirmed cases in the cluster can be directly linked to close and prolonged exposure to a patient during a phase of severe illness. Although human-to-human transmission cannot be ruled out, the search for a possible alternative source of exposure is continuing.
The index case had developed symptoms May 24 from an unknown source. A family gathering was planned for May 30. As noted in the update, four of the eight members of the cluster slept in the same small room on May 29. They developed symptoms a few days later, as did the index case’s daughter, niece, and nephew. The family gathering and the location of their housing allowed for additional close contact on or about May 30, which is the approximate date of exposure based on symptoms a few days later. Another brother developed symptoms after his son (who was the nephew of the index case) died.
The development of symptoms close to the date of death of the index case is common. Just prior to death the viral load is usually at a peak. Most H5N1 isolates come from samples collect just prior to or on the date of death. Thus, the infection of family members, including blood relatives who cared for the infected relative, is not unexpected.
The number of large H5N1 clusters is relatively small. The index case frequently is a child or teenager, which is likely due to an increased exposure risk. A requirement of infection of three family members or more is required for infection of husband and wife.
However, a cluster of five in Haiphong did involve a husband, wife, and three daughters. Similarly, there have been clusters that involved a health care worker in Vietnam, and a friend in Azerbaijan. Thus, although the number of larger clusters is limited, they contain clear examples of infections in contacts that are not blood relatives.
Similarly, media reports suggest that one or two health care workers were infected through contact with the larger familial cluster in Pakistan, and health care workers linked to patients infected in 1997 in Hong Kong had antibodies, indicating the were infected with H5N1 by patients who were not blood relatives.
Moreover, many of the outbreaks are multi-focal. In Pakistan in addition to the large familial cluster, there was a smaller familial cluster, as well as at least one additional patient who did not appear to be a blood relative of any of the cluster members. The relationships in the Pakistan outbreak will be clearer when the situation update is released.
However, earlier clusters do not support the genetic predisposition that is cited in media reports.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings