Serological investigation among humans involved in the mass culling
operation at a large-scale poultry farm affected with highly pathogenic
avian influenza H5N1 in Kyoto Prefecture, Japan
-----------------------------------------------------------------------

1. Background
----------------
58 persons who were involved in the operation during the poultry highly
pathogenic avian influenza (HPAI) H5N1 outbreak in Feb 2004 were
investigated. Of which 16 are poultry workers from the affected farm and
the other 42 are prefecture workers who participated in the poultry culling
operation. All participants were asked about the type of work that they did
with poultry, performance of personal infection prevention procedures, and
illnesses experienced during and after the operation. For comparison, serum
specimens were also collected from 33 prefecture workers who did not join
the operation.

The poultry workers had been exposed to ill chickens without protection
before prefectural intervention. The first investigation team consisting of
3 persons sent urgently by the Prefecture wore gowns, goggles, gloves and
masks (non N-95) without prophylactic use of oseltamivir. Then Kyoto
Prefecture government conducted the mass culling operation from 27 Feb to
22 Mar 2004.  All persons involved in the operation including the poultry
workers and the 3 first investigation team members wore personal protective
equipments (PPE) including N-95 mask and were given oseltamivir
chemoprophylaxis during the operation.

2. Results
----------
Serum specimens were tested by a microneutralization assay by use of an
avian influenza virus A/Chicken/Kyoto/3/2004(H5N1) which had been isolated
from a chicken from the affected farm. The assay was performed at
Department of Virology III, National Institute of Infectious Disease,
Tokyo, Japan. Serum specimens were considered to be positive for antibody
to H5 viruses by microneutralization assay if titres of 10 were obtained,
because  all the control sera showed titres of less than 10 and a cut-off
value could not be assigned statistically. All specimens were collected
approximately one month after the end of operation. Frozen serum specimens
which had been taken from the 4 persons about one month earlier were
obtained, therefore we could evaluate their neutralization antibody
seroconversions.

Category / No. of persons enrolled / Negative serum Anti-H5 neutralization
antibodies / Positive serum Anti-H5 neutralization antibodies /
Poultry workers / 16 / 12 / 4* /
Prefectural workers / 42 / 41 / 1 /
Control /       33 / 33 / 0 /

* One of whom seroconverted.

Among 4 persons with paired serum samples, one person seroconverted. This
person claimed sore throat soon after starting chemoprophylactic treatment
with oseltamivir for a couple of days but without any systemic symptoms
such as fever. The other four seropositive persons did not experience any
influenza-like symptoms including fever. All seropositive poultry workers
had histories of unprotected exposure to ill chickens. One seropositive
prefecture worker eas among the three sent as the first investigation team.

3. Discussion
--------------
The seroconverted poultry worker was considered to have been infected with
H5 influenza virus recently. Therefore, he was likely to have been infected
during unprotected contacts with the affected chickens. The other 3
seropositive poultry workers were possibly infected with H5 influenza virus
also during unprotected contacts. The seropositive prefecture worker may
have been infected during the first investigation visit with incomplete PPE
such as without wearing N-95 mask.

It is difficult to evaluate precisely the effectiveness of personal
infection preventive procedures including wearing PPE and chemoprophylaxis
based on this study because paired serum specimens were tested in only 4
persons. However, since all prefecture workers who participated in the
operation with the personal infection prevention procedures were
seronegative except a seropositive person of the first investigation team,
the personal infection prevention procedures were considered to be effective.

Among these 5 seropositive persons, none clinically developed human HPAI
infection compatible symptoms or other febrile illness. They also took
chemoprophylaxis during the period of the operation organized by Kyoto
prefecture. Therefore, the chemoprophylaxis may have prevented them from
developing illness. There are no public health problems in terms of
possibility of their developing illness related to this event and further
HPAI transmission.

There are some limitations of this study to deliver solid conclusions.
First, the validity of the cut-off value at 10 in our assay has not been
not evaluated sufficiently, because of limited knowledge on human H5
infections. Second, paired serum specimens were obtained from only a
limited number of people.

--
Dr.Nobuhiko Okabe and the staffs of Infectious Disease Surveillance Center
Dr.Masato Tashiro and the staffs of Virology III(Influenza CC of WHO)
National institute of Infectious Diseases, Japan

----------------------------------------------------------------------------------------

Review in Science Vol 307, Issue 5706, 27 , 7 January 2005

When the Japanese H5N1 outbreak was confirmed at a chicken farm in Kyoto Prefecture last February, Japan´s National Institute of Infectious Diseases urged local officials to survey farm workers, health inspectors, and those who destroyed the chickens. Institute virologist Masato Tashiro, director of the World Health Organization collaborative center for influenza surveillance and research in Japan, says the difficulties in detecting low levels of antibodies slowed the work, and then prefectural officials dithered over releasing the results.

Out of 7000 people potentially exposed, only 58 agreed to participate in the survey. Those 58 included 17 of 19 people who worked on the infected farm before taking the antiviral medication Tamiflu or wearing protective clothing. The five people who proved to be seropositive were among this group; none of those who took Tamiflu before going to the farm or wore protective gear while there proved positive. "We think this does say something about the value of antiviral medication and proper protection," notes Tashiro.

Albert Osterhaus, a virologist at the Netherlands´ Erasmus University Medical Center in Rotterdam, suggests that the five Japanese could have developed antibodies in response to viral antigens in the farm environment and were never actually infected with the H5N1 virus.

Why the infections, if they did occur, proved so mild is less clear. Tashiro offers several possibilities. For one, the genetic sequences of the viral strain collected in Japan and Korea varies from that of the strain that appeared later in Thailand and Vietnam. Once the presence of H5N1 was confirmed, farm workers and health official who had visited the farm took Tamiflu, perhaps in time to reduce the severity of the infection. Finally, exposure to the virus could have been more limited than among the patients in Thailand and Vietnam, many of whom raised chickens at home.

"We don´t have any controls, so it´s difficult to determine just why these differences occurred," Tashiro says. Scientists hope that surveys of cullers in Thailand and Vietnam who did not take Tamiflu and were often not wearing proper protective gear may answer these questions.