"This is a highly fatal disease and we know there is a real risk that it can reappear in an infected animal at a later time," said Dr Ron Glanville. "It was considered that the human health and safety considerations must come first."
Tamworth, reported to be worth $A200,000 was among five infected horses at the Redlands Veterinary Clinic in Brisbane.
While he recovered, authorities enforced biosecurity regulations requiring that the horse be euthanized.
Dr Glanville said testing on Tamworth had detected genetic material from the Hendra virus, indicating recent infection.
"Tamworth also tested positive to serological tests indicating past infection with Hendra virus. The serum neutralisation test used is regarded as the gold standard for detection of antibody response to Hendra virus infection.
"While the national policy for Hendra virus is, and has always been, that blood-test-positive horses should be put down, we did take the time to review the policy in this case," he said.
"We also seriously considered the option of long-term quarantine and monitoring of the surviving horse. However we came to the final view that the horse should be put down largely based on the advice of medical authorities that it was not worth the risk to human life - particularly the people caring for the horse.
"It was considered that the human health and safety considerations must come first."
Dr Glanville said the decision was supported by chief veterinary officers from all states and the Commonwealth, the Head of the Australian Animal Health Laboratory in Geelong, the Australian Racing Board, the Australian Horse Industry Council, the federal Department of Health and Aging, Queensland Health and the Australian Veterinary Association.
"In making this decision, it was also noted that the actual information to be gleaned from longer term monitoring the horse would be relatively limited. However, we will learn as much as possible from this horse.
"People from the Australian Animal Health Laboratory attended the autopsy and a wide range of tissues were collected for further analysis."
Dr Glanville said there are two primary pieces of evidence indicating that Hendra virus infection may remain dormant in the body and re-appear at a later time:
History of Hendra virus incidents:
Tests on tissue samples obtained from Tamworth's body returned positive tests in a range of tissues - kidney, urine, spleen, bronchial lymph node, two head lymph nodes, renal lymph node, brain and spinal cord.
"This test detects the presence of viral RNA, not necessarily live virus. However the range of positive tissues 40-plus days following putative infection is noted as surprising."
The autopsy was conducted by a team of expert veterinary pathologists under high level biosecurity containment measures.
Dr Glanville said the Primary Industries Department has reviewed Hendra guidelines for veterinarians with considerable input from Australian Veterinary Association and Equine Veterinarians Australia.
The Redlands cases, he said, raised a range of issues regarding how private veterinarians will approach possible Hendra virus cases in the future, including case definition and the use of personal protection equipment and appropriate biosecurity measures.
The document has been circulated to vets.
Hendra virus guidelines for horse owners is under final review with input from the horse industry.
The Redlands outbreak resulted in two people catching the virus. Veterinarian Ben Cunneen died five weeks later in hospital, becoming the third person to die among the six known to have contracted the virus since it was first isolated in 1994.
The virus is carried naturally by native fruit bats and spillover to horses is rare.
"The bats appear to be susceptible to infection with Hendra virus, but do not themselves become ill," Dr Glanville said.
The route of infection between bats and horses is believed to be from bat bodily fluids, including saliva, urine and birthing fluids contaminating horse feed or water.
Transfer of Hendra virus from human to human, or from human to horse, has not been recorded.
There have been 11 clusters of Hendra virus recorded in horses since the virus was first identified in 1994. All but one of these cases has been in Queensland. The case fatality rate in horses is high (greater than 70%). A table of Hendra virus incidents is below.
All human cases of the disease have been in Queensland and were in people who had in close contact with Hendra-infected horses.
Dr Glanville, who said the Redlands cluster of cases was "largely resolved" on August 22, also spoke of the three unrelated cases of Hendra virus at the Proserpine property, in northern Queensland.
Two horses have been euthanized, with a decision pending regarding the recovered horse.