The case report by Vandecandelaere et al. (2026) describing two fatalities following administration of extended-release injectable omeprazole (ERIO) provides an important opportunity to reflect on the inherent risks associated with parenteral drug administration in horses. While these deaths are undoubtedly tragic, it is essential to contextualise this risk within the broader landscape of equine injectable therapeutics to provide balanced clinical guidance. Vandecandelaere et al. (2026) describe two Warmblood horses that developed peracute systemic reactions immediately after intramuscular administration of compounded extended-release injectable omeprazole (ERIO), with one horse recovering and the second dying within 40 min of injection. The reactions developed during or within seconds of injection, with collapse, paddling, tachycardia, dyspnoea and cyanosis, making a causal relationship with the injection event highly likely. Product analysis did not reveal major impurities, although soil-associated organisms were cultured from residual product in the fatal case and the authors discuss inadvertent intra-arterial or subarachnoid injection, anaphylaxis and contamination as potential mechanisms. The author speculates inadvertent intra-arterial administration as being most likely. Commercial sales and pharmacovigilance data from Australia and the UK document over 210,000 units of LAIO have been sold between 2019 and 2025. Within these markets, three deaths have been reported following ERIO injection: the fatal gelding in the case report plus two additional horses that collapsed and died after cervical intramuscular administration, and a further horse that collapsed but recovered fully. One of the three deaths followed an 11-day course of systemic illness associated with a contaminated injection site, representing the development of sepsis rather than a direct pharmacological effect. Even assuming complete administration of all sold doses and underreporting of adverse events, these figures suggest an apparent death prevalence no higher than approximately 1 in 70,000 injections and a very small number of life-threatening nonfatal reactions, which falls within or below the ‘very-rare’ category used in summary of product characteristics (SPCs) for licensed products (<1 animal in 10,000 treated). Procaine penicillin is a useful comparator because it is administered intramuscularly in relatively large oily volumes and has long been recognised as a cause of peracute systemic reactions in horses. Olsen et al. (2007) collated 59 reports of adverse reactions to procaine benzylpenicillin or sodium/potassium benzylpenicillin in Swedish horses, with most reactions occurring within minutes of injection and dominated by locomotor and behavioural disturbances consistent with central nervous system excitation. Although the total number of doses administered was not reported, practice-based experience from the author suggests approximately one death per 2000–3000 procaine penicillin injections, at least an order of magnitude higher than the observed ERIO death prevalence, recognising that both figures are subject to reporting bias. Ivermectin, frequently considered a low-risk routine treatment, also carries measurable risk. In the study by Karns and Luther (1984), 366 of 3316 horses (11%) developed an adverse reaction following ivermectin administration, with 332 reactions (91% of all reactions) comprising ventral midline pruritus or oedema. Fifteen horses (0.45%) developed transient injection-site swelling or stiffness, eleven (0.33%) developed limb oedema and four (0.12%) eyelid oedema. More serious events were rare but included two horses that developed increased respiratory rate and depth, fever, disorientation and colic, and one horse (0.03%) died shortly after injection (1 in approximately 3000). These data underline that deaths temporally associated with accepted injectable medications are not unique to omeprazole. Modern equine vaccines provide another benchmark for tolerated injection risk. SPCs for widely used licensed products define adverse-reaction frequencies as very common (more than 1 in 10 animals treated), common (more than 1 but less than 10 animals in 100), uncommon (more than 1 but less than 10 animals in 1000), rare (more than 1 but less than 10 in 10,000) and very rare (less than 1 in 10,000 animals treated, including isolated reports). When considering two commonly administered vaccines, Strangvac and combined influenza and tetanus vaccine Equip FT, we can evaluate reported adverse events. For Strangvac, transient fever and local swelling at the injection site are described as common, while anaphylactic-like reactions are very rare. Published data (Rendle et al., 2025) from more than 20,000 doses indicate reported transient reactions in 560 horses (5.6%), mostly mild and self-limiting, with one injection-site abscess (0.01%) and a single suspected immune-mediated response among approximately 10,000 treated horses. For Equip FT, transient fever and pyrexia are reportedly rare whereas injection-site pain, anorexia, lethargy and hypersensitivity are very rare (Zoetis UK Limited, 2024). These figures emphasise that even rigorously tested biologicals are associated with occasional, sometimes serious reactions. Against this background, three ERIO-associated deaths and a handful of severe systemic reactions among at least 210,000 doses sold appear broadly compatible with the very-rare adverse-event profile accepted for many established therapies in equine practice. The clinical pattern in the ERIO deaths which include collapse, paddling, tachycardia and rapid progression to cyanosis either during injection or soon after injection is most consistent with inadvertent intra-arterial injection of a particulate oily suspension. The ERIO formulation is designed to form a depot in muscle. If accidentally introduced into the circulation it could act as a lipid embolus to the pulmonary vasculature or disseminate smaller emboli to cerebral and coronary vessels, analogous to reported injuries after intra-arterial injection of particulate corticosteroids in people. Repeated injections into the same region may promote local vascular proliferation and increase the likelihood of intravascular placement, a mechanism also proposed for procaine penicillin reactions. True anaphylaxis cannot be excluded, especially given multiple reports of anaphylactic reactions to oral and intravenous omeprazole in humans, but no such reactions have been documented in horses to date. Bacterial contamination of residual product in one fatal ERIO case, with growth of Bacillus cereus and Cellulosimicrobium cellulans, points to realistic risks of environmental contamination when injections are performed through hair coats, although these organisms are not typically associated with peracute collapse. Horses are inherently challenging patients to inject. Sometimes thick, long, dusty coats are difficult to decontaminate in field conditions, and standard cervical injection landmarks lie near major vascular structures. Behavioural reactions at the time of needle placement or injection (flinching, jumping, head movements) can shift or bend needles from intramuscular to intravascular or extra-dural spaces without obvious warning. Puschmann and Ohnesorge (2015) reported that complications following intramuscular injections are most frequent in the pectoral region and higher in the neck than in hamstring muscles, a pattern echoed in ERIO and Strangvac studies that have reported fewer issues when gluteal or hamstring sites are used and injections are performed by veterinary surgeons rather than owners. Intravenous administration of long-acting injectable omeprazole has been reported without adverse effects (E. Jones, personal communication), which supports the inference that the active molecule itself is unlikely to be directly cardiotoxic or neurotoxic at therapeutic doses and that the principal risk arises from the combination of an oily particulate vehicle, a high-risk injection site and the intrinsic challenges of equine field practice. Specific evidence comparing different skin-preparation regimes for equine intramuscular injections is lacking, but the identification of soil organisms in residual ERIO highlights the potential for environmental contamination of needles, syringes and vials when injections are performed through contaminated coats. Rubbing and scrubbing the skin in a nonsterile manner has been reported to loosen epidermal cells and drive resident bacteria into the needle track and increase infection risk in small animals (Ford & Mazzaferro, 2006). SPCs for vaccines and other injectables emphasise aseptic technique and avoidance of visibly soiled sites, for routine intramuscular use. Needle size is controversial and there is little evidence in any species regarding local trauma and infection risk. One study (Diggle & Deeks, 2001) in human infants highlights more tissue trauma with smaller gauge needles (25G vs. 23G) due to higher jet pressure, whereas larger physical size in food animals is suggested to increase the risk of absolute tissue trauma and risk of inoculating cores of skin and hair (Alabama Cooperative Extension System, 2025). The authors of the ERIO case report highlight that compounded medications have not undergone the same pre-registration safety and efficacy testing as licensed products. It is acknowledged that compounding pharmacies are not equal, but many, especially those in the UK, are compliant with good manufacturing (cGMP) and laboratory practices (GMP) as per the European Medicine Standards (2023). The data reported and presented here does not show an increased risk of severe reactions or sudden death when compared with licensed products administered intramuscularly. In practice, long-acting omeprazole preparations are increasingly used to address compliance issues of daily oral administration and to treat glandular disease where injectable formulations may achieve superior outcomes. The deaths reported by Vandecandelaere et al. (2026) serve as a sobering reminder that no injectable medication is without risk. However, when contextualised against surveillance data and comparative risk from other commonly used injectables, the safety profile of injectable omeprazole appears acceptable for judicious clinical use. The most likely mechanism - inadvertent intra-arterial injection of an oil-based formulation, represents a risk inherent to the route of administration rather than a specific drug toxicity. Continued pharmacovigilance, adherence to best practices for injection technique and site preparation, and transparent communication with clients remain essential to minimising avoidable harm. Always report adverse drug reactions to appropriate authorities. Thanks to BOVA UK and Australia for providing sales and pharmacovigilance data to better inform this commentary and for the personal communication from Emma Jones. There are no funders to report for this commentary. G. Hallowell provides educational CPD for BOVA.
G. D. Hallowell (Sun,) studied this question.