Bird Flu Secondary Cases Are Predominantly Female
Recombinomics Commentary
March 10, 2005
>> Nearly all cases known to date have involved close contact with sick birds. Human-to-human transmission of H5N1 is extremely rare, though health officials suspect that a mother in Thailand contracted the disease by cradling her sick daughter. <<
The human-to-human transmission described in the familial cluster in Thailand is not extremely rare. Such familial clusters account for over one third of the H5N1 cases reported in Thailand, Vietnam, and Cambodia. The Thailand cluster was rare because the index case was an 11 year old girl who was living with her aunt in Kamphaengphet-Nontaburi. Her mother was an office worker in Bangkok, and therefore was not present when her daughter was infected. The mother visited her daughter in the hospital, and then developed symptoms after the hospital visit. The daughter died. The mother was hospitalized in Bangkok where she also died. The index case's aunt also developed symptoms, but recovered. The cluster was written up in the New England Journal of Medicine because the geographical distance between the mother and daughter at the time of the infection was large, and the mother and aunt developed symptoms after caring for the daughter. Since the mother did not have contact with chickens, and disease onset dates were bimodal, it was concluded that this was human-to-human transmission of H5N1.
However, the circumstances associated with the H5N1 transmission were virtually identical to the other familial clusters. The cluster was bimodal. The secondary cases developed symptoms 1-2 weeks after the index case. The secondary patients were family members who were likely caregivers for the index case. In all instances the index case died and at least one member of the cluster was confirmed to be H5N1 positive. In almost all clusters the lab confirmations were less than ideal. Samples were not collected from several of the index cases and many patients were retested after negative or inconclusive results. Each case that was retested eventually tested positive for H5N1.
The details in the Thailand case were not unusual. The index case was misdiagnosed as dengue fever. She was not tested for H5N1. Her mother was not tested for H5N1 while she was alive. Investigators were gathering data on another case in Bangkok. By sheer chance a nurse alerted them to another unreported flu death. The investigators noticed the common name between the mother and daughter and located the mother just before she was to be cremated. H5N1 sequences were then obtained from cadaver tissues. The aunt also tested negative initially, but was positive on subsequent tests.
H5N1 sequences from the mother have been placed on deposit at GenBank (A/Thailand/Kamphaengphet-Nontaburi/04(H5N1)). There are 857 bp of the NA gene, and only one bp difference between the mother's H5N1 sequence and several isolates from Thailand, including at least one tiger isolated at the beginning of 2004. The sequence is also closely related to sequences from other patients in Thailand and Vietnam. Similarly, 782 bp of the HA gene were also deposited. That sequence has only 2 bp differences between several isolates from Thailand, including the tiger. The sequence is also closely related to other isolates from patients in Thailand and Vietnam. Thus, the available sequence shows that the virus that infected the girl's mother was virtually identical to many other viruses isolated in Thailand and Vietnam last year.
Thus, the virus and transmission circumstances of the cluster in Thailand match the other familial clusters in Vietnam and Cambodia from virus, to bimodal distribution, to infection of close relatives.
However, another way to demonstrate human-to-human transmission in these other familial clusters is to simply compare the gender distribution of the primary cases relative to the secondary. There were 13 index cases in the 12 clusters (in one cluster cousins developed symptoms at the same time). There were 6 females and 7 males indicating the risk to both sexes was similar. In contrast, 11 of the 14 secondaries were female.
If infection was from a common source, the gender differences should have been minimal, as was seen in the primaries. However, since the females were sisters, daughters, or wives of the primaries, they were likely to have been the primary caregivers, traditionally the role of females. Consequently, the secondary cases were mostly female.
However, even the three male secondaries probably were infected by the index cases. Two were young brothers (age 4 and 5) of an older sister who was fatally infected with H5N1. This brother sister relationship probably led to close contact between primary and secondary cases. The third male was the brother of the Thai Binh index case in January. He had gone home after the family gathering, but returned to Hanoi to visit his brother in the hospital and then developed symptoms the day his brother died.
Thus, the gender analysis of primary and secondary cases also strongly implicates human-to-human transmission from primary to secondary case. The bimodal distribution of onset dates supports this transmission route. Since these familial cases represent over one third of the reported bird flu cases, such transmissions are not rare.
Failure to acknowledge and warn of such transmissions increases the likelihood of additional transmissions, because the caregiver relative mistakenly believes that infection from poultry is much more likely than infection from the relative being cared for or visited in the home or hospital.
Media link