Chlamydia pecorum is the single most destructive disease facing koalas in eastern Australia. Infection rates hit 66% in some populations. The disease causes blindness, infertility, and emaciation — and antibiotic treatment frequently kills the animals by destroying the gut microbiome they need to digest eucalypt leaves. Since 2022, koalas in Queensland, New South Wales, and the ACT have been listed as endangered. A vaccine has been the obvious answer for over a decade. Two are in development; one was approved in September 2025. The approved single-dose vaccine, developed by University of the Sunshine Coast researchers led by Prof Peter Timms, rests on a body of 29 published papers. But the keystone 2024 paper in npj Vaccines — which reported a 64% reduction in chlamydia deaths over 10 years — contains what University of Queensland molecular virologist Prof Keith Chappell calls a 'silly' data error. In a pre-print posted to bioRxiv, Chappell and colleague Dr Michaela Blyton identified that survival from birth to time of vaccination was incorrectly attributed to vaccination itself. After correcting, they found no significant difference in survival or disease rates between vaccinated and unvaccinated koalas. A separate randomised controlled trial published last year — the gold standard the original vaccine studies lacked — delivered the same verdict from a different angle. Conducted by University of Sydney researchers on three wild NSW koala populations, the study found the UniSC vaccine stimulated an immune response but 'did not translate to any effect on chlamydiosis.' Dr Valentina Mella was blunt: 'We wanted a very strong effect of the vaccine and we didn't get it.' The Australian Pesticides and Veterinary Medicines Authority approved the vaccine with carefully qualified language — the permitted claim states only that it 'may aid in the reduction of clinical signs of the disease,' not that it prevents infection or mortality. The APVMA rejected the characterisation that its approval constituted regulatory failure. But the gap between what qualified regulatory language permits and what field veterinarians and the public understand is doing real work here. A vaccine that 'may aid' is functionally indistinguishable from a vaccine that doesn't work, if the only randomised trial shows no disease effect. Meanwhile, a second vaccine developed at Queensland University of Technology by Dr Freya Russell's team has vaccinated more than 500 koalas in the Gold Coast's Currumbin region and reports reduced chlamydia-related hospital admissions. But this team has not yet conducted placebo-controlled studies either — they judged it unethical to withhold treatment from a rapidly declining population. They have, however, developed a biodegradable implant to deliver the booster dose without recapturing wild koalas, which a captive study found equivalent to the two-shot protocol in stimulating antibody responses. The implant is not yet APVMA-approved. The structural problem is clear: conservation urgency is running ahead of evidence quality. Both vaccine teams face legitimate ethical constraints — you cannot easily randomise endangered wild animals into placebo groups when populations are crashing. But approval without randomised evidence, followed by a data error in the pivotal study, followed by a randomised trial showing no effect, creates a situation where regulatory approval may be providing false confidence that the chlamydia problem is being addressed. UniSC defends its broader research programme, noting the data dispute concerns 'one aspect of a much larger research program' and does not 'invalidate more than a decade of peer-reviewed science.' Timms frames it as iterative: 'It's a journey, right? There'll be ups and downs.' Both framings are reasonable. But neither addresses the core question: is the approved product doing anything useful in wild koalas right now? The only randomised evidence says no.
