Musculoskeletal adverse events in dogs receiving bedinvetmab (Librela)

Abstract

Objectives: To conduct a specialist-led disproportionality analysis of musculoskeletal adverse event reports (MSAERs) in dogs treated with bedinvetmab (Librela™) compared to six comparator drugs with the same indication. Furthermore, to report the findings from a subset of dogs whose adverse event (AE) data underwent independent adjudication by an expert panel.

Study design: Case-control study and case series analysis.

Sample population: The European Medicines Agency's EudraVigilance database (2004-2024) and 19 client-owned dogs.

Methods: An EBVS^® Veterinary Specialist in Surgery individually reviewed all MSAERs to Librela™, Rimadyl^®, Metacam^®, Previcox^®, Onsior^®, Galliprant^®, and Daxocox^® (2004-2024). The primary null hypothesis was that Librela's MSAER rate would not exceed that of comparator drugs by more than 50%. The secondary hypothesis was that MSAER would surge and taper following the launch of new drugs.

Results: The disproportionality analysis did not support the hypotheses. Ligament/tendon injury, polyarthritis, fracture, musculoskeletal neoplasia, and septic arthritis were reported ~9-times more frequently in Librela-treated dogs than the combined total of dogs treated with the comparator drugs. A review of 19 suspected musculoskeletal adverse events (MSAEs) by an 18-member expert panel unanimously concluded a strong suspicion of a causal association between bedinvetmab and accelerated joint destruction.

Conclusion: This study supports recent FDA analyses by demonstrating an increased reporting rate of musculoskeletal adverse events in dogs treated with Librela. Further investigation and close clinical monitoring of treated dogs are warranted.

Impact: Our findings should serve as a catalyst for large-scale investigations into bedinvetmab's risks and pharmacovigilance.

Keywords: Librela; NGF; RPOA; accelerated joint destruction; bedinvetmab; rapidly progressive osteoarthritis.

Figure 1

This diagram is adapted from the European Medicines Agency’s pharmacovigilance data flow chart (36). Although the Agency encourages reporting via the attending veterinarian (shown in the upper half of the diagram), an Adverse Event Report (AER) may be initiated by anyone directly involved with the event. Crucially, this system lacks a direct feedback loop to the first-hand reporter. A lack of feedback hinders transparency, increases the risk of errors, and has the potential to discourage future reporting. Note that 88% of AERs were submitted by veterinarians or other healthcare professionals (HCPs), a finding that does not support Zoetis’ concerns regarding potential over-reporting influenced by negative social media activity directed at Librela (22).

Figure 2

Accumulated MSAER comparing Librela to six medications with the same indication, focusing on specific reaction types. Ligament/tendon injury, polyarthritis, fracture, musculoskeletal neoplasia, and septic arthritis were reported ~9 times more frequently in dogs treated with Librela than the combined total of dogs treated with the comparator medications. No qualifying MSAER were found for Daxocox^® (enflicoxib), a weekly NSAID released in April 2021, 2 months after Librela’s European launch.

Figure 3

Accumulated MSAER for Librela and the same comparator medications as shown in Figure 2. There was a dramatic increase in MSAER to Librela in the second half of 2024, reaching ~4-times the level observed in the 6-months following its 2023 U.S. release and ~59-times the level observed in the 6-months following its 2021 European release. To test the hypothesis that MSAER would surge and taper following the launch of new drugs (indicated by color-coded arrows), we analyzed data from the EudraVigilance database since its 2004 inception. Neither this hypothesis nor the primary null hypothesis—that Librela’s MSAER would not exceed those of comparator drugs by more than 50%—were supported by the data. In fact, Librela’s accumulated MSAER measured over 45 months were ~3-times higher than the combined accumulated MSAER of all comparator drugs measured over 240 months. The February 2022 MSAER spike might reflect the EMA’s mandate for MAHs to submit all previously un-submitted AERs at regular intervals (43). Post-release MSAER trends for Rimadyl^® and Metacam^® are unavailable because they were released in 1997 and 1998, respectively.

Figure 4

Severity and outcome data for MSAER associated with Librela. Unexpected findings are highlighted in red. These include a significant proportion of severe MSAER, such as ligament ruptures, luxations, fractures, limb collapse, and septic arthritis, filed as “not serious”. In addition, several dogs diagnosed with bone cancer were reported as “recovered/resolving”. The EMA defines a serious adverse event as “any adverse event which results in death, is life-threatening, results in persistent or significant disability/incapacity, or a congenital anomaly or birth defect.” However, a more intuitive and clinically relevant definition includes events causing permanent disability (44), requiring surgical intervention and/or prolonged hospitalization (12). Importantly, published AER data are subject to change, but only if translation errors are recognized and reported (see Figure 1).

Figure 5

(A) Case 1, a young adult Labrador Retriever. A,B: Orthogonal CT slices demonstrating medial compartment pathology within the right elbow joint. Arrows indicate a medial coronoid process (MCP) fissure and humeral condylar kissing lesion. There is no evidence of a humeral intracondylar fissure (HIF). C,D: Preoperative radiographs obtained after 6 Librela injections reveal a lateral humeral condylar fracture (HCF) with subsequent malunion. A chevron symbol highlights atypical periosteal new bone growth along the medial epicondyle [see Panel 5B(F)]. E,F: Postoperative radiographs acquired 9 months after surgery show successful fracture union, although subchondral bone resorption had progressed. G,H: Severe bilateral seromas that developed after two Librela injections and persisted despite treatment. (B) Case 1—contralateral elbow. A–D: Pre-treatment radiographs and CT scans show MCP remodeling and a humeral condylar lesion consistent with osteochondrosis. E,F: Severe proximal ulnar sclerosis and atypical medial epicondylar periosteal new bone formation were observed after 6 doses of Librela. G,H: Surgical intervention was performed 3 months later. Surgery resulted in successful fracture union, but persistent clinical dysfunction necessitated euthanasia. Histopathological examination: Surgical site infection and septic arthritis were excluded. Findings were consistent with those described in an article on RPOA submitted to the pathologist by the attending specialist (20). AER: The MAH filed a report with an incorrect diagnosis of septic arthritis (Supplementary Figure S1).

Figure 6

Case 4 is a Labrador Retriever with no history of trauma. (A,B) Pre-treatment radiographs taken aged 8.5-years. A pre-treatment left craniocaudal radiograph was not acquired. (C,D) Post-Librela radiographs demonstrate a left lateral HCF with an atypical medial epicondylar periosteal reaction (C) and a right HIF (D). (E,F) Immediate post-operative radiographs. AER: After considering the lack of trauma and atypical signalment for HIF, the attending specialist filed an AER to the Veterinary Medicines Directorate (VMD) for suspected RPOA. This report was shared with the MAH, who filed a report for non-serious “arthritis”, with an outcome of recovered/resolving (Supplementary Figure S3).

Figure 7

Left (L): Normal pre-treatment CT scans from Case 11, a 7.5-year-old Labrador Retriever. Post-Librela CT scans revealing fulminant periarticular osteophytosis. Right (R): Post-Librela CT scans revealing fulminant periarticular osteophytosis. AER: The attending specialist filed an AER to the VMD specifying their suspicion of RPOA. This report was shared with the MAH, who filed an AER for non-serious arthritis, with an outcome of recovered/resolving (Supplementary Figure S5).

Figure 8

Case 13 is a Staffordshire Bull Terrier with no history of trauma. (A–D) Pre-treatment CT scans showed mild elbow arthrosis and excluded HIF. (E–H) Post-Librela radiographs demonstrated severe left elbow arthrosis and a right lateral HCF with an atypical medial epicondylar periosteal reaction. AER: Given the patient’s age and breed, which is not predisposed to HCF, the attending specialist submitted a report to the MAH specifying a suspicion of RPOA. The MAH filed a report with an incorrect diagnosis of “osteosarcoma” (Supplementary Figure S7).

Figure 9

Case 14, a 5.5-year-old Labrador Retriever, received Librela to treat mild elbow dysplasia. Tarsal swelling and hindlimb lameness developed, and this CT scan was acquired after 16 doses. Note the marked bilateral soft tissue swelling and palisading calcaneal periosteal reaction. AER: A report was filed for suspected RPOA because both hocks were clinically normal before treatment and developed severe OA with bilateral talar insufficiency fractures after treatment.

Figure 10

Case 15, an 8-year-old English Bull Terrier treated with Librela for elbow dysplasia developed tarsal swelling and hindlimb lameness within 3 weeks of the first dose. (A,B) Radiographs acquired after 2 Librela injections show moderate bilateral arthrosis which progressed rapidly, culminating in left tibiotarsal luxation (C–E). Synovial fluid analysis ruled out inflammatory arthropathy or tick-borne disease. AER: The MAH filed a report using the diagnostic term “swollen joint”, and misclassified the reaction as not serious and recovered/resolving. Following communication from the attending veterinarian regarding the translation error, the MAH escalated the enquiry to their Global Pharmacovigilance Team, who concluded that the observed pathology was consistent with erosive immune-mediated polyarthritis (IMPA), despite normal synovial fluid test results (Supplementary Figure S8).

Figure 11

Radiographs from Case 17, a 5-year-old Springer Spaniel, before and after 20 Librela injections.

Figure 12

Case 18, a 5.5-year-old Australian Shepherd, had stable stifle joints before starting Librela. AER: The attending specialist filed a report for suspected RPOA to the MAH, who submitted an AER for ligament ruptures, fractures, and joint subluxation/luxations. Their report designated this reaction as not serious (Supplementary Figure S9).

Figure 13

Case 19, a 10.5-year-old mixed-breed dog, had 7 Librela injections following right tibial plateau leveling osteotomy (TPLO). Severe joint erosion is evident in the medial femoral condyle (arrow) and tibial plateau, causing screw exposure (chevron). Intractable pain necessitated euthanasia. Surgical site infection and septic arthritis were excluded by necropsy, which revealed findings similar to those reported in human RPOA (20). AER: The attending specialist filed a report for suspected RPOA to the MAH, who submitted an AER using the diagnostic terms “limb non-weight bearing” and “abnormal radiograph finding”.