Veterinary treatment of livestock with diclofenac, a non-steroidal anti-inflammatory drug (NSAID), has caused catastrophic declines of Gyps vultures in Asia. This has highlighted a lack of knowledge on the potential impacts of NSAIDs on scavenging birds. Surveys of veterinarians and zoos document the outcomes of the treatment of over 870 scavenging birds from 79 species. As well as diclofenac, carprofen and flunixin were associated with mortality, with deaths observed in 13 and 30% of cases, respectively. Mortality was also found following treatment with ibuprofen and phenylbutazone. NSAID toxicity was reported for raptors, storks, cranes and owls, suggesting that the potential conservation impact of NSAIDs may extend beyond Gyps vultures and could be significant for New World vultures. In contrast, there were no reported mortalities for the NSAID meloxicam, which was administered to over 700 birds from 60 species. The relative safety of meloxicam supports other studies indicating the suitability of this NSAID to replace diclofenac in Asia.
Populations of three species of Gyps vulture in the Indian subcontinent have collapsed since the early 1990s and are now at high risk of extinction (IUCN 2004). Veterinary use of diclofenac, a non-steroidal anti-inflammatory drug (NSAID), is a major cause of the observed population declines (Green et al. 2004; Oaks et al. 2004). Vultures are exposed to diclofenac when they consume carcasses of livestock that were treated with the drug shortly before death. Experiments show that they die from kidney failure within days of exposure and have extensive visceral gout at post-mortem (Oaks et al. 2004; Swan et al. 2006a). Identical signs of toxicity have been found in carcasses of wild vultures (Oaks et al. 2004; Shultz et al. 2004).
The collapse in numbers of Gyps vultures across Asia means that other scavenging birds are increasingly exposed to contaminated carcasses. Whether diclofenac is affecting them is unknown, although Indian vultures from other genera are also in rapid decline (Cuthbert et al. 2006). Diclofenac and other veterinary NSAIDs are licensed and used in many areas of the world, including southern Africa and South America (Anderson et al. 2005; J. Parry-Jones 2006, unpublished information). Hence, the use for conservation of ‘vulture restaurants’ in southern Africa and the veterinary use of diclofenac in South America are a major cause for concern.
Steps are now being taken to control the veterinary use of diclofenac in India and the identification of the NSAID meloxicam as an alternative has facilitated this (Swan et al. 2006b). However, the safety of meloxicam to other species of scavenging birds and the potential toxicity and the safety of other NSAIDs have not been reported. In this study, we used questionnaire surveys on the clinical use of NSAIDs to make a preliminary assessment of their safety to vultures, raptors and other scavenging birds.
2. Material and methods
Questionnaires were sent to zoos, wildlife rehabilitation centres and veterinarians worldwide. We requested detailed information on species and number of individuals treated, NSAID or other anti-inflammatory drug used, method of administration, number and frequency of doses, days of treatment, dose level, condition treated and the clinical outcome of treatment. Some survey information could not be completely quantified, particularly for the number of individuals treated. Where respondents replied with ‘several’, ‘many’ or ‘more than 1’, we recorded the number of birds treated as two. Consequently, final sample sizes are likely to be minima. Some birds were treated on multiple occasions. We considered the treatment of an individual (whether single or multiple treatments) with a specific NSAID as the unit of replication. Treatments of the same individual with separate courses of different NSAIDs (n=4) were recorded as separate cases.
A total of 31 veterinarians and institutions responded, providing information on over 870 cases of NSAID treatment for 79 species of birds including Gyps vultures, other raptors, storks, cranes, owls and crows. While owls and cranes are not scavenging birds, the survey provided comprehensive information for owls and one reported an instance of mortality for a crane: consequently the results are presented. Information was also provided on dexamethasone, a steroidal anti-inflammatory drug.
As well as the known diclofenac mortalities, there were 16 instances of mortality with renal disease and gout for a number of NSAIDs across a range of species (figure 1; table 1). Carprofen and flunixin meglumine were associated with mortality of Gyps vultures and other species, with a reported mortality of 13% (5/40 cases) and 30% (7/23), respectively. These figures do not include a Gyps africanus that died after treatment with both carprofen and ketoprofen, and another that died after receiving either flunixin or ketoprofen. There is no indication that the birds which died received a particularly high dose of carprofen (1–3, 4 and 5 mg kg−1, cf. 1.5–7.6 mg kg−1 for all birds treated) or flunixin (1–4.5 mg kg−1, cf. 0.5–12 mg kg−1). Two instances of mortality with renal disease and gout are reported for ibuprofen and phenylbutazone.
There were no mortalities following treatment with meloxicam. For Gyps vultures, 39 individuals from six species (G. africanus, Gyps bengalensis, Gyps coprotheres, Gyps fulvus, Gyps himalayensis and Gyps rueppellii) have been treated and a minimum of 700 birds from 54 other raptors and scavenging species were given meloxicam (see electronic supplementary material). Meloxicam doses ranged from 0.1 to 0.75 mg kg−1 bw, with a median dose of 0.5 mg kg−1. Meloxicam was administered by intramuscular injection (57% of treatments), orally (32%) or through a combination of one intramuscular injection followed by oral dosing (11%). Treatment ranged from 1 to 120 days (median 5 days). Less information is available on the safety of other NSAIDs, although the survey results indicate ten cases where dexamethasone (a steroidal anti-inflammatory) and 20 instances where ketoprofen (when this drug was administered on its own) have been administered with no reported mortalities.
Our results show that certain NSAIDs are toxic to raptors, storks, cranes and owls and suggest that the conservation impact of diclofenac and other NSAIDs may not be restricted to Gyps vultures. Of particular significance is the mortality of a Marabou stork (Leptoptilus crumeniferus) following treatment with flunixin. Storks and New World vultures are phylogenetically closely related (Sibley et al. 1988), and the veterinary use of NSAIDs within South America is consequently of potential conservation concern; testing the toxicity of NSAIDs to New World vultures is a priority. The survey also highlights the relative safety of meloxicam to a wide range of bird species, with over 739 individuals from 60 species treated with no mortalities. More information is required to assess the safety of ketoprofen and dexamethasone since the number of birds treated is small, and there are reported concerns on the safety of ketoprofen in ducks (Mulcahy et al. 2003).
Carprofen and flunixin appear to carry a high risk of renal damage in birds, which supports earlier findings concerning the safety of flunixin (Klein et al. 1994; Clyde & Murphy 1999). Carprofen and flunixin are used for the treatment of livestock within Europe, although they are not yet available in South Asia. The published information on carprofen and flunixin in livestock tissue residues indicate that a vulture consuming a 1 kg meal from an animal that died shortly after a veterinary course of these drugs, could be exposed to doses close to, or within, the range of doses (1–5 mg kg−1) that caused mortality of birds after clinical treatment (see electronic supplementary material). Consequently, the veterinary use of flunixin and carprofen in South Asia could result in similar problems to those caused by diclofenac. Currently, a range of NSAIDs are recommended for veterinary use in India (Anonymous 2002), including drugs this study has found associated with mortality. This highlights the need for robust safety testing before recommending any NSAID as a safe replacement for diclofenac (Swan et al. 2006b).
Knowledge of the mechanism of NSAID toxicity in vultures is currently lacking, although Meteyer et al. (2005) propose that diclofenac toxicity of G. bengalensis is a consequence of renal ischemia through activation of renal portal valves. The same clinical signs at post-mortem (renal disease and visceral gout) are found for diclofenac, carprofen and flunixin; suggesting that the mechanism of toxicity may be similar. NSAIDs operate through the inhibition of the cyclo-oxygenase enzymes, COX-1 and COX-2, and the relative inhibition of these two enzymes is thought to alter the risk of adverse effects on renal function (Brater 2002). The hepatotoxicity of different NSAIDs has also been linked to chemical structure, with evidence for toxicity where there is a carboxylic acid group (–COOH) in combination with a nearby linking –NH group (Sussman & Kelly 2003). Consideration of the eight NSAIDs reported in this study, suggest that there is no simple relationship between NSAID toxicity and COX-1/COX-2 inhibition (table 2). However, there is some support that the presence of both –COOH and –NH groups is associated with toxicity, as these structures are present in the NSAIDs most associated with mortality and are absent from those NSAIDs that exhibited no signs of toxicity (table 2). However, ibuprofen and phenylbutazone do not conform to this pattern and this hypothesis requires further investigation.
In conclusion, our survey suggests that widespread use of NSAIDs may be having impacts on bird populations in addition to the known effect of diclofenac on Gyps vultures. At least two NSAIDs, in addition to diclofenac, show evidence of toxicity to scavenging birds. However, the conclusion that meloxicam is not toxic to scavenging birds at concentrations likely to be encountered is supported by the survey and supports the use of this drug as an alternative for diclofenac.
We are very grateful for the veterinary, zoo and raptor rehabilitation communities for their response to this survey. We would like to thank the following individuals and institutions: Simon Girling; Arnaud Van Wettere; Jose Lopez Cerezuela; Chester Zoo; Cheyenne Mountain Zoo; Franklin Park Zoo Hospital, Zoo of New England, USA; Louisville Zoological Garden; Zoo-Aquarian de Madrid; Andres Montesinos; Oregon Zoo; Ursula and Juan Manuel; Nick Lindsay; Wildvets S.C.; Willemien van Wyk; Whipsnade Wild Animal Park, Zoological Society of London; Breeding Centre for Endangered Arabian Wildlife; Calgary Zoo; San Diego Zoo; Klinik fur Vögel; Lincoln Park Zoo; Maryland Zoo, Baltimore; San Antonio Zoo; Tulsa Zoo and Living Museum; Tri-State Bird Rescue & Research Inc.; John Chitty; Halley D Buckanoff; Chaffee Zoological Gardens, California; Cascades Raptor Centre, Oregon; Wild at Heart Wildlife Rehabilitation, North Carolina; Smithsonian's National Zoological Park; National Zoo, Santiago; The Hawk Conservancy; and the Vulture Conservation Breeding Centre, Haryana, India. We also thank the UK Government's Darwin Initiative for the Survival of Species and the RSPB for financial support.