The authors have declared that no competing interests exist.
The Northern Territory of Australia (NT) is currently regarded as free of the vectors of dengue
On 20 July 2010, a 34-year-old man presented to the public hospital in Darwin on the north coast of the NT with a one day history of rash and fever, retro-orbital headache, and prominent myalgia. He denied any travel outside the Northern Territory in the previous month. On examination he was febrile (38.3°C) and had an extensive blanching erythematous rash on his trunk, back, and upper and lower limbs.
A serum sample referred to PathWest Laboratory Medicine WA and tested for NS1 antigen by a commercial enzyme immunoassay (Bio-Rad Platelia Dengue NS1 Ag) was strongly positive. Dengue virus (DENV) was detected using a type-specific semi-nested polymerase chain reaction (PCR). The first round used consensus outer primers that detect all four dengue serotypes, followed by a second round using type-specific inner primers
Samples were also sent to a second reference laboratory (Queensland Health Forensic and Scientific Services) for confirmation and sequencing of the envelope gene. Nucleotide sequencing and phylogenetic analysis of the DENV-1 E gene (1,485 base pairs) was carried out comparing the NT 2010 sequence with 53 other sequences derived from local, imported, or global origins (A. Pyke, unpublished data). The NT 2010 sequence (GenBank accession number HQ871946) was designated within DENV-1 genotype IV, which contains predominantly Pacific strains, but initial nucleotide comparisons did not reveal highly similar sequences amongst other available strains. However, in September 2010, analysis of virus obtained from a symptomatic traveller recently returned from Bali, Indonesia (Bali 2010a,
Strain names and respective genotypes are given including locally transmitted viruses (Townsville, Cairns, and Mareeba). GenBank accession numbers of retrieved global strains are shown in brackets. Bootstrap support values derived from 1,000 replicate NJ trees are represented for principal nodes >70%. DENV-2 strain Townsville 2004 was used as an outgroup control.
The case gave consent for the details of his illness to be published. He lived in an outer suburb of Darwin and worked in an industrial zone on the southern boundary of the Darwin Airport, which functions as both a civilian and military airport (
Shown are: the case's workplace (cross in circle), containerised cargo delivery locations (dash in circle), apron boundaries (dashed line), southern airport boundary (dot and dash line), airplane arrival areas (airplane symbol), routine adult mosquito traps (red circle), and extra case adult mosquito traps (green circle). Image supplied courtesy of Northern Territory Government.
Eight mosquito larval surveys, four egg trapping surveys using ovitraps, and 12 adult mosquito trapping surveys using Biogents and CO2-baited traps were carried out at the case's residence, places visited, workplace, and nearby premises, including four locations within the airport's southern boundary from 5 August until the early wet season rains had been established (October). These failed to reveal any dengue vectors. All recent adult mosquito collections in Darwin, part of the routine mosquito surveillance program, were re-examined for any evidence of exotic mosquitoes but none were found. Public Health Units in Queensland were contacted to establish the nature of any recent outbreaks of type 1 dengue.
During July, Darwin Airport received approximately 20 flights from north Queensland and 50 international flights per week including 14 per week from Bali
Despite heightened awareness and active case-seeking, no further cases were detected in Darwin.
Prior to this case, there had not been a documented case of locally acquired dengue in the NT since the 1950s. We consider it highly unlikely that there were undetected
The proximity of the case's workplace to the RAAF base, and the arrival there of a C-130 aircraft from Bali just prior to the most likely date of acquisition, raise the possibility that it was acquired from an infective mosquito which escaped this aircraft.
Alternatively, it is possible that an
Mosquito ecology would also suggest that air transport is associated with a higher risk of introducing a live infected mosquito than sea or road transport, and therefore the most likely source of dengue infection for this case is an infected mosquito arriving via an aircraft. The proximity of the case-patient's workplace to the RAAF base implicates that as a possible route, but the evidence is nevertheless circumstantial.
The Australian Quarantine and Inspection Service (AQIS) disinsection procedures are mandatory for all international aircraft arriving in Australia
The genotyping of this virus was important in assisting in the investigation of the case. As the likely origin was only determined when a Bali-acquired case presented in another Australian state 6 weeks after this case, it demonstrated the value of centralising sequence data for viruses of public health significance. There is nevertheless always some uncertainty about the exact origin of viral strains. There were multiple strains emerging from Bali during the 2010 season, which might reflect the increase in travel to this destination, but also might mean that some of these strains were acquired from other locations but labelled as Bali strains, as Bali is often the last stopover before leaving South East Asia.
In summary, this is the first recognised case of locally transmitted dengue in the NT since the 1950s. We consider the most likely source of dengue to be a mosquito which alighted from a C-130 military aircraft arriving from Bali in early July and that this may be a case of “airport dengue” equivalent to previous reported cases of airport malaria
Dengue can be transmitted to non-endemic regions through the transport of infected mosquito vectors in the same way as airport malaria.
Clinicians should be aware that, with the increase in global travel, international trade, and military movements, dengue can appear in places far removed from regions where vectors are established.
The centralisation of sequence data on viruses allows for comparisons between isolates and greatly assists in the analysis of transmission dynamics in viral diseases.
We wish to acknowledge the following: the staff of Medical Entomology for mosquito trapping and surveys, staff of the Centre for Disease Control and the Northern Territory Department of Health, Dr Rob Baird and pathology staff at Royal Darwin Hospital, virology staff at the Queensland Health Forensic and Scientific Services, staff at PathWest, RAAF Darwin, RAAF Air Movements and HQ Air Command NSW, Kevin Langham and Mary Finlay-Doney at AQIS, the Tropical Public Health Unit, Queensland, staff and residents of Dundee Beach and Melville Island Lodge, industrial businesses in Winnellie industrial area, Perkins Shipping, and the patient and his family for assistance. Thanks also for comments by William Pettit of ME DoH and Mary Finlay-Doney of AQIS Darwin.