Research articlesInvestigation of a bone lesion in a gorgonopsian (Synapsida) from the Permian of Zambia and periosteal reactions in fossil non-mammalian tetrapods
Abstract
While only distantly related to mammals, the anatomy of Permian gorgonopsians has shed light on the functional biology of non-mammalian synapsids and on the origins of iconic ‘mammal-like’ anatomical traits. However, little is known of gorgonopsian behaviour or physiology, which would aid in reconstructing the paleobiological context in which familiar mammalian features arose. Using multi-modal imaging, we report a discrete osseous lesion in the forelimb of a late Permian-aged gorgonopsian synapsid, recording reactive periosteal bone deposition and providing insights into the origins and diversity of skeletal healing responses in premammalian synapsids. We suggest that the localized lesion on the anterolateral (preaxial) shaft of the left radius represents acute periostitis and, conservatively, most likely developed as a subperiosteal haematoma with subsequent bone deposition and limited internal remodelling. The site records an inner zone of reactive cortical bone forming irregular to radial bony spicules and an outer, denser zone of slowed subperiosteal bone apposition, all of which likely occurred within a single growing season. In surveys of modern reptiles—crocodylians, varanids—such haematomas are rare compared to other documented osteopathologies. The extent and rapidity of the healing response is reminiscent of mammalian and dinosaurian bone pathologies, and may indicate differing behaviour or bone physiology compared to non-dinosaurian reptiles. This report adds to a growing list of putative disease entities recognized in early synapsids and broadens comparative baselines for pathologies and the evolution of bone response to disease in mammalian forebears.
This article is part of the theme issue ‘Vertebrate palaeophysiology’.
1. Introduction
Gorgonopsians were a group of carnivorous non-mammalian synapsids that thrived during the middle to late Permian (ca 265–252 Ma). Like other contemporaneous therapsids, these saber-toothed apex predators exhibited an amalgam of plesiomorphic reptile-like and mammal-like traits, making them an important fossil group for understanding the origins of mammalian behaviour and physiology [1,2]. Previous studies of gorgonopsian functional biology have largely emphasized the cranium and jaw mechanics [3,4]. Studies on the appendicular skeleton have suggested a plausible wide range of fore- and hind limb postures, underscoring the important transition from a sprawling to a more upright limb posture in the predecessors to mammals [1,5–7]. Moreover, the phylogenetic proximity of gorgonopsians to eutheriodonts—the synapsid subgroup that includes mammals [2,8–10]—has further fuelled interest in investigating their paleobiology. Nevertheless, gorgonopsian fossils remain relatively rare, and the majority of those that have been described in detail represent incomplete or isolated cranial specimens [5,8].
Prior paleopathological studies on non-mammalian synapsids have found bony abnormalities having the gross appearance of lesions and concluded that they resulted from either healing fracture/trauma or severe infection in the form of osteomyelitis. For example, both fracture calluses and callus-like structures resulting from microlamellar slippage were documented in the basal synapsid families Edaphosauridae [11–13] and Sphenacodontidae [11,13–15], whereas osteomyelitis has been suggested in Sphenacodontidae (spine: [11]), Stahleckeriidae (femur and tibia: [16,17]) and Titanosuchidae (femur: [18]). Among these, Moodie [11] described the oldest known synapsid paleopathology, a case of osteomyelitis at the base of a neural spine in the sail-backed sphenacodontid Dimetrodon [11]. The middle Permian titanosuchid described by Shelton et al. [18] represents the oldest pathology recorded in the mammal-like therapsids that was consistent with a diagnosis of osteomyelitis.
While these studies provide precious calibration points for investigating ancient skeletal diseases, surveys of non-mammalian bone pathologies have been limited in their taxonomic scope, especially among Permian therapsids. Gaps in our understanding of paleopathology have not only made difficult the identification and diagnosis of ancient diseases, but also form a barrier to reconstructing the evolutionary history of mammalian healing responses. Such gaps are due to the comparative rarity of pathological specimens themselves, as well as the paucity of systematic documentation of osseous lesions outside of a few animals of economic importance, such as horses and poultry. Thus, our study of a well-preserved gorgonopsian with an osseous lesion provides a new datum for the evolution of mammalian healing responses, particularly within Permian theriodonts. In this study, we used a combination of high-resolution X-ray micro-computed tomography (micro-CT) and histologic sectioning, in conjunction with extant comparative baselines, to describe and diagnose a case of periostitis in the forelimb of a gorgonopsian. We propose that the resulting periosteal reaction is best attributed to an ossifying subperiosteal haematoma, the first documented case in a gorgonopsian and the oldest in any theriodont therapsid.
2. Material and methods
A partial skeleton of a mid-sized non-rubidgine gorgonopid, including partial cranium, mandible with dentition, vertebral column, and both forelimbs and portions of the hind limb, was collected near the southern border of North Luangwa National Park (Northern Province, Zambia) by a team led by C.A.S. in 2014 and is now catalogued at the National Heritage Conservation Commission (Lusaka) as NHCC LB396. The specimen was recovered from light brown mudrocks of the upper Permian Madumabisa Mudstone Formation (Luangwa Basin); detailed locality information is available from C.A.S. The left radius of the specimen was well preserved and presented a rugose lesion on the anterolateral face of the mid-diaphysis upon inspection (figure 1). The radius was further prepared in the paleohistology laboratory of A.K.H. at University of Southern California's Keck School of Medicine for further radiographic and histologic analysis (figures 2 and 3).
Figure 1. Forelimb of the gorgonopsian (NHCC LB396) presenting a bony lesion on the left radius. (a) Anterior view of the left radius showing the extent of the lesion (black arrows). The radius is oriented with the distal end toward the top of the page and proximal at the bottom. (b) Dorsal view of the pathological radius showing the extent and gross appearance of the lesion. (c) Articulated right humerus, radius and ulna of the same specimen, showing a slender (non-pathological) appearance of the radial diaphysis. (Online version in colour.) Figure 2. Microanatomy of a lesion in a gorgonopsian (NHCC LB396). (a) CT image of the pathologic radial diaphysis demonstrating the lesion in longitudinal section. (b) Dorsal view of a three-dimensional model of the radial fragment for orientation of the CT slices. (c) Transverse section of the distal region of the radius showing the large diameter and high vascularity of the lesion. (d) Transverse, mid-lesion CT image showing the high vascularity of the healing response and large diameter of the lesion. (e) Transverse CT image of the proximal portion of the lesion showing breakage of the radius. Figure 3. Histological sections of radius, NHCC LB396, showing distal (a,c) and proximal (b,d) portions of the lesion. Note the radial spicules are localized within a single growth zone, supporting the view that the injury was not chronic and confined to a single growth season (white arrows in (a) denote growth marks bounding growth zones). Putative cracks in the proximal lesion show jagged edges and calcite infilling, indicating that the breakage in this area was postmortem and unrelated to the injury. Circularly polarized light (b,d) reveals mud infilling (m.i.) and calcite precipitation around the lesion and other porous regions of the radius in cortical and subcortical areas. a, anterior; d, dorsal; p, posterior; v, ventral. (Online version in colour.)
NHCC LB396 was micro-CT scanned at the University of Southern California Molecular Imaging Center using a Nikon XTH 225 ST system. The resulting data were visualized in NIH ImageJ and Amira 5.3.3, revealing some breakage and compaction of the bone wall into the medullary cavity, although it could not be determined whether this breakage was due to fracture or occurred postmortem, necessitating thin-sectioning and polarized light microscopy to inspect the cortical bone tissue (figures 2 and 3). Computed tomography as a supplement to histology has become widely used in paleopathologic investigations [19], offering a guide for histologic sectioning and further analysis at the tissue level.
Thin-sectioning of the radius followed standard procedures outlined by Chinsamy & Raath [20], Wilson et al. [21] and Lamm [22]. After scanning, moulding and casting the specimen, the shaft of the radius was embedded in a cold-mounting resin and cut into 1-mm wafers on a Buehler IsoMet low-speed saw with a diamond blade. Representative transverse sections were cut from the proximal, middle and distal portions of the shaft, as well as a longitudinal section spanning the lesion. Wafers were glued onto frosted petrographic slides with an epoxy resin and then ground to a thickness of approximately 100 µm using a Buehler MetaServ 250 grinder/polisher, then polished by hand. Thin sections were imaged using normal and circularly polarized light settings with a Leica DM 2700 microscope fitted with a motorized stage and digital image capture system.
3. Results
The lesion was localized on the anterolateral (preaxial) edge of the left radial shaft, and presents as an oblong, raised rugosity measuring approximately 6 cm in length (figure 1). Because most appendicular skeletal elements are represented in NHCC LB396, it was possible to determine that this was a solitary, monostotic lesion. No other irregularities were noted, which indicates that the affected area was localized within the left forelimb and rules out a systemic or metabolic disease.
Superficially, the area exhibits the roughened texture and location frequently found in healed greenstick fractures, without any notable periosteal response on the posterior face of the bone shaft (figure 1). However, micro-CT imaging showed little evidence of fracture trauma, revealing in cross-section a periosteal reaction forming an osseous thickening with limited to no erosion or displacement of the original cortical bone (figure 2). The reactive bone growth surrounds half of the radial midshaft at its extreme and is far more extensive than the gross appearance of the lesion (figures 2 and 3). The reactive bone expresses a continuous, regular spiculated pattern showing no layering, interruption or any concentrated marginal ossifications (such as the Codman's triangle of fast-growing osteosarcoma) and no accompanying lytic lesions in the surrounding bone. A dense outer shell of poorly vascularized, periosteally deposited bone coats the underlying spiculated bone, likely corresponding to the final phase of healing and resumption of normal bone growth. This dense outer layer appears continuous with the remainder of the cortex and is somewhat less porous than the underlying zone (figure 2). The perimedullary region near the distal metaphysis also showed some apparent internal remodelling, indicated by the medullary bone porosity and blurring of the perimedullary margins of the cortical bone.
Histological imaging further revealed that the healing response was confined to a single growth zone and composed of a homogenous meshwork of disorganized bony spicules, primarily fine cancellous spicules with a few radial spicules (figure 3a,c). Though evidence of premortem fracture was poor based on the micro-CT scan data, some minor cortical breakage was noted that infiltrated and compacted some of the cancellous bone within the medullary cavity at the level of the proximal shaft (figure 2e). This was further inspected in histological thin section. Fragmented and compacted bony tissue within the deep cortex and medullary cavity showed hard and jagged edges, and lacked any apparent healing response (figure 3b,d). Polarized light microscopy showed that the medullary cavity and endosteal resorption cavities were primarily sites of calcite precipitation, whereas the cortical lesion contained a mud-matrix infilling, providing taphonomic evidence that the large breakages in these areas were associated with postmortem crushing and burial.
4. Discussion
(a) Differential diagnosis of skeletal disease in synapsids
Multi-modal imaging approaches to paleopathology have the greatest potential to diagnose putative disease entities in synapsids and other tetrapod groups [19], but have only recently been widely adopted. A recent histologic investigation of the lower jaw and dentition of a gorgonopsian from the late Permian of Tanzania recognized a compound odontoma [23], the first such tumour convincingly demonstrated in a stem mammal. Prior paleopathological investigations of therapsid fossils have relied on combinations of gross observation [16–18] and/or histologic thin-sectioning [18,24], enabling tentative diagnoses of non-traumatic or infectious diseases with varying degrees of confidence. The titanosuchid pathology examined by Shelton et al. [18] was diagnosed as acute osteomyelitis based on radial bony spicules with pockets of empty space in the cortical bone that were interpreted as abscesses or fibriscesses. Whereas the absence of an involucrum and cloaca in NHCC LB396 weaken the potential diagnosis of a pus-producing osteomyelitis in the present gorgonopsian specimen [25,26], its exceptional preservation and multi-modal imaging permit detailed consideration of other candidate diseases in a gorgonopsian for the first time.
Differential diagnoses for periosteal reactions generally include not only osteomyelitis, but also trauma, sometimes resulting in partial or full fracture. Non-traumatic etiologies can include infection, neoplasm, degenerative or metabolic or developmental processes (table 1). The affected region in NHCC LB396 is not a discrete tendon or ligament attachment site, ruling out a diagnosis of enthesopathy or joint-related disease, and without a systemic healing response, additional evidence of neoplasm or malignancy is also lacking [26]. Although the spiculated, rapidly deposited bone is reminiscent of cranial lesions caused by alpha and beta thalassemia, the location, lack of lesion on the contralateral bone and outer shell of dense bone covering the cancellous areas argue against this interpretation [27]. These observations also argue against neoplasm and degenerative etiologies. The superficial appearance of a greenstick fracture is contradicted by radiographic and histologic observations, as in the hard edges of the fragmented and compacted tissue, and preserved calcite and mud-matrix infilling supporting that bone breakage was postmortem (figure 3b,d). There is likewise no evidence of cortical drift or callus formation, further weakening the possibility that the pathology represents a fracture [28].
| category of etiology | disease entity candidate | quality of fit |
|---|---|---|
| fracture trauma | greenstick fracture | poor |
| non-traumatic or infectious | periostitis | good |
| non-pyogenic osteomyelitis | fair | |
| pyogenic osteomyelitis | poor | |
| neoplasm | osteosarcoma | poor |
| degenerative | enthesopathy | poor |
Taking a conservative approach, we attribute this pathology to an acute form of periostitis based on the following diagnostic criteria: (1) contiguous nature of reactive bone to nearby subperiosteal cortical bone within a single growth zone, excluding exostosis; (2) continuous, irregularly and radially oriented bony spicules confined to a relatively radiolucent area apposed against the original cortical bone; (3) a thin, outer radiodense layer of subperiosteal bone; (4) absence of large lytic lesions or a central nidus, ruling out many forms of neoplasm; (5) the absence of marginal features, such as the Codman's triangle, a pathognomic feature of osteosarcoma; and (6) lack of an open cloaca for pus drainage, reducing the likelihood of pyogenic osteomyelitis. Infection cannot be entirely excluded by the lack of pus drainage features. Many taxa do not form pus, but instead react with fibresses featuring caseous necrosis (e.g. birds, crocodylians). However, the regularity of the spiculated bone, its presumably rapid formation and lack of globular lacunae argue against this form of infectious response in the present case.
Given the apparently rapid bone formation and the absence of features often associated with infection, it remains most plausible that the periosteal reaction formed as an ossifying subperiosteal haematoma, which may occur secondary to trauma but not universally. A localized non-pyogenic immune response in this clade remains another possibility, though more difficult to establish without obvious features of infection or other soft-tissue evidence (table 1).
(b) Other cases of non-mammalian subperiosteal haematoma
Simple periosteal reactions as in subperiosteal haematoma can be secondary to minor trauma or stress, and may show slight periosteal elevation and roughening on macroscopic inspection [29] and a somewhat solid or uninterrupted radiographic appearance [30]. At the histological level, however, ossifying haematomas in mammals are characterized by normal cortical bone associated with the periosteal surface and radiating, fine cancellous bone apposing it [31], underscoring its generally rapid healing response.
As in the gorgonopsian specimen described here, Rothschild [29] attributed an elevated periosteal lesion in the tibia of Varanus komodoensis—American Museum of Natural History (AMNH) 74606—to a subperiosteal haematoma. Given the prior investigation's sole reliance on macroscopic observation, we CT-scanned the tibia of AMNH 74606 for comparative purposes (figure 4) and asked whether similar radiographic characteristics would present in both modern and fossil specimens. As predicted, the periosteal reaction in AMNH 74606 presents as an apparently continuous zone of reactive bone apposed against the original cortical bone. The reactive bone is highly vascularized and contains fine cancellous tissue composed of irregular and radial spicules. Interestingly, there is more porous space but with overall lesser bone density than in the mammal-like response of the gorgonopsian (a phenomenon that has been noted previously in reptilian and amphibian reactive bone growth; [32]). Notably, the collections-based survey of wild and captive reptiles by Rothschild [29] showed that, while rare in non-mammals, non-traumatic osseous changes are more common in some reptile groups than in others. In particular, aside from spondyloarthropathy (the most commonly encountered reptile skeletal pathology) non-traumatic cases constituted only 0.4% of reptile pathologies, but represented 3.8% of crocodylid and 3.1% of varanid cases, as well as a small percentage of iguanians. It is therefore plausible that the surprising occurrence of an ossifying haematoma in a gorgonopsian is suggestive of vague but similar behaviours and ecophysiology as in these reptile groups, which are known to be of substantial body mass, and either active, wide-ranging foragers (varanids; [33]) or belonging to groups suspected of having endothermic ecophysiology ancestrally (crocodylians; [34]).
Figure 4. High-resolution radiograph of a lesion in the tibia of Varanus komodoensis from the American Museum of Natural History, specimen number AMNH 74606, in transverse (a) and longitudinal (b) section. The gross appearance of the lesion was originally described by Rothschild [29] where it was attributed to a subperiosteal haematoma (r.b., reactive bone).
Finally, we present a survey of osseous paleopathologies documented in fossil tetrapods in appendix A and in figure 5, demonstrating gaps in our knowledge of non-traumatic osseous changes in non-mammals, especially in reptiles and early synapsids. Although there are more than 45 published cases of paleopathologies consistent with the differential diagnosis for our lesion in NHCC LB396, these cases are distributed unevenly across anamniote, reptilian and synapsid subgroups (figure 5). Most of these published reports were based on macroscopic or radiographic observations, and not analysed at the microanatomical or biogeochemical levels, leaving substantial room to improve rigour in their differential diagnosis and interpretation of underlying disease [35]. Consequently, more samples and improved analytical techniques are needed. Nevertheless, the results of our investigation agree with the broad conclusions of Shelton et al. [18]—that mammalian forebears presented remarkable similarities to mammalian and dinosaurian disease responses by as recently as the middle to late Permian.
Figure 5. Published description of skeletal disease in late Paleozoic–Mesozoic amniote and anamniote stem groups and their extant relatives (data from appendix A). Mapped occurrences are limited to primary disease candidates for the bony lesion of NHCC LB396: fracture, periostitis, osteomyelitis, spondyloarthropathy, enthesopathy, neoplasm and other. (Online version in colour.)
5. Conclusion
Here, we described in detail the first well-preserved osseous lesion identified in the postcranial skeleton of a Permian gorgonopsian synapsid. The solitary lesion is best diagnosed as an instance of periostitis, plausibly related to a subperiosteal haematoma as in similar lesions reported rarely in non-mammals. The healing response appeared to be relatively rapid, with highly vascularized radial bone formation, thus necessitating comparisons to similar pathologies in non-mammals, such as varanids and crocodylians, where the tissue-level processes are not as well understood. At the moment, it is apparent that healing responses of ossifying haematomas may be somewhat conserved across tetrapods, while also being variable in the rate and extent of reactive bone formation during the ossification phase [32]. Further conclusions regarding synapsid physiology, however, will require additional baselines from other fossil and extant non-mammals.
Data accessibility
Virtual data are available with permission through MorphoSource.org (https://doi.org/10.17602/M2/M97687) or from the corresponding author (A.K.H.) by request.
Authors' contributions
K.M.K. analysed the data, prepared figures and led writing of the paper; E.A.R. analysed the data, consulted on its interpretation and wrote portions of the paper; C.A.S. contributed data; A.K.H. contributed and analysed data, prepared figures and wrote portions of the paper.
Competing interests
The authors declare no competing interests.
Funding
K.M.K. was supported by the University of Southern California Undergraduate Research Associates Program (URAP).
Acknowledgements
The gorgonopsian fossil described here was collected with the support of NSF EAR-1337569 (to C.A.S.) and prepared at the Burke Museum (Seattle) by G. Livingston. We thank the National Heritage Conservation Commission (Lusaka) and the field team for assisting with research in Zambia. We thank D. Kizirian and the American Museum of Natural History, New York, for the loan of V. komodoensis AMNH 79606. We also thank T. Jashashvili and the University of Southern California's Molecular Imaging Center in the Keck School of Medicine for assistance with micro-CT scanning NHCC LB396. CT scanning of AMNH 79606 was performed by S. Merchant and E. Hsu of the University of Utah's Small Animal Imaging Center, Salt Lake City. K.M.K. thanks the University of Southern California Undergraduate Research Associates Program (URAP) for funding.
Appendix A
| higher taxon | etiology | pathology | casesa | reference(s)b |
|---|---|---|---|---|
| Anamniotes | ||||
| Bufonidae | traumatic | facture callus | 2 | [36] |
| infectious | unidentified infection | 2 | [36] | |
| Cryptobranchidae | traumatic | fracture callus | numerous | [37] |
| non-traumatic | spondylarthropathy | 8 | [37] | |
| †Diadectidae | non-traumatic | spondyloarthropathy | 1 | [36] |
| Ranidae | traumatic | fracture callus | 2 | [36] |
| †Stereospondyli | non-traumatic | congenital vertebra fusions | numerous | [38] |
| †Temnospondyli indet. | neoplasm | parostotic osteosarcoma | 1 | [39] |
| †Basal tetrapod indet. | traumatic | fracture callus | 1 | [40] |
| reptiles | ||||
| Agamidae | traumatic | fracture callus | numerous | [36] |
| non-traumatic | exostosis | 1 | [29] | |
| non-traumatic | osteochondroma | 1 | [36] | |
| non-traumatic | osteoma | 1 | [36] | |
| non-traumatic | pseudogout | 1 | [36] | |
| Alligatoridae | non-traumatic | articular gout | 3 | [36] |
| †Allosauridae | traumatic | avulsion | 1 | [41] |
| traumatic | fracture callus | 2 | [41,42] | |
| traumatic | pseudarthrosis | 2 | [42] | |
| non-traumatic | idiopathic | 2 | [42] | |
| infectious | osteomyelitis | 1 | [42] | |
| infectious | pyogenic osteomyelitis | 1 | [41] | |
| degenerative | enthesopathy | 2 | [42] | |
| †Baenidae | non-traumatic | avascular necrosis | 1 | [36] |
| †Camptosauridae | infectious | necrosis | 1 | [11] |
| †Captorhinidae | infectious | osteomyelitis | 1 | [43] |
| Cathartidae | non-traumatic | hypertrophic osteopathy | 1 | [44] |
| †Ceratopidae | traumatic | greenstick fracture | 1 | [11] |
| non-traumatic | hyperostosis | 1 | [11] | |
| Chamaeleonidae | non-traumatic | spondyloarthropathy | 1 | [29] |
| infectious | unidentified joint infection | 1 | [29] | |
| Cheloniidae | non-traumatic | avascular necrosis | 12 | [36] |
| neoplasm | osteoblastoma | 1 | [36] | |
| †Clevosauridae | infectious | osteomyelitis | 1 | [45] |
| Colubridae | neoplasm | chondrosarcoma | 1 | [36] |
| neoplasm | osteosarcoma | 2 | [36] | |
| Corytophanidae | infectious | unidentified | 1 | [29] |
| Crocodylidae | non-traumatic | articular gout | 6 | [36] |
| non-traumatic | congenital abnormality | 3 | [29] | |
| non-traumatic | exostosis | 2 | [29] | |
| non-traumatic | gout | 1 | [29] | |
| non-traumatic | osteochondroma | 1 | [29] | |
| non-traumatic | osteochondrosis | 1 | [29] | |
| non-traumatic | periosteal reaction | 1 | [29] | |
| non-traumatic | spondyloarthropathy | 7 | [29] | |
| infectious | unidentified infectious bone | 6 | [29] | |
| Dermochelyidae | non-traumatic | avascular necrosis | 4 | [36] |
| †Desmatochelyidae | non-traumatic | avascular necrosis | 2 | [36] |
| †Dinosauria (unidentified) | non-traumatic | hypertrophic osteopathy | 1 | 44 |
| (non-avian dinosaur) | traumatic | fracture callus | 1 | 11 |
| (horned dinosaur) | infectious | subperiosteal abscess | 1 | [11] |
| †Diplodocidae | traumatic | fracture callus | 1 | [11] |
| non-traumatic | spondylitis deformans | 1 | [11] | |
| infectious | osteomyelitis | 1 | [11] | |
| Elapidae | non-traumatic | osteochondroma | 1 | [36] |
| neoplasm | osteochondrosarcoma | 1 | [36] | |
| neoplasm | numerous | numerous | [46] | |
| †Elasmosauridae | non-traumatic | congenital abnormality | 1 | [36] |
| Gavialidae | non-traumatic | articular gout | 2 | [36] |
| Gekkonidae | non-traumatic | spondyloarthropathy | 2 | [29] |
| †Hadrosauridae | traumatic | fracture | 2 | [19] |
| Helodermatidae | non-traumatic | osteomalacia | 1 | [29] |
| non-traumatic | spondyloarthropathy | 1 | [29] | |
| Iguanidae | non-traumatic | articular gout | 1 | [36] |
| non-traumatic | chondro-osteofibroma | 1 | [36] | |
| non-traumatic | congenital abnormality | 2 | [29] | |
| non-traumatic | haematoma | 3 | [29] | |
| non-traumatic | hypertrophic osteoarthropathy | 1 | [29] | |
| non-traumatic | osteoarthropathy | 1 | [36] | |
| non-traumatic | osteochondroma | 2 | [36] | |
| non-traumatic | periosteal reaction | 2 | [29] | |
| non-traumatic | spondyloarthropathy | 2 | [29] | |
| infectious | unidentified infectious bone | 2 | [29] | |
| neoplasm | osteosarcoma | 1 | [36] | |
| Lacertidae | neoplasm | chondro-osteofibroma | 1 | [36] |
| neoplasm | osteoma | 1 | [36] | |
| neoplasm | osteosarcoma | 1 | [36] | |
| †Metriorhynchidae | infectious | fungal infection | 1 | [36] |
| †Mosasauridae | non-traumatic | alveolar osteitis | 1 | [11] |
| non-traumatic | rheumatoid arthritis | 1 | [11] | |
| infectious | necrosis | 1 | [11] | |
| neoplasm | osteoma | 1 | [11] | |
| †Pachypleurosauridae | non-traumatic | avascular necrosis | 1 | [36] |
| †Pappochelys | neoplasm | periosteal osteosarcoma | 1 | [47] |
| †Parasuchidae | traumatic | fracture callus | 1 | [11] |
| †Pareiasauridae | non-traumatic | spondyloarthropathy | 1 | [48] |
| †Phytosauria | non-traumatic | necrosis | 1 | [11] |
| non-traumatic | spondyloarthropathy | 1 | [49] | |
| †Pleurosternidae | non-traumatic | avascular necrosis | 1 | [36] |
| †Protostegidae | non-traumatic | avascular necrosis | 4 | [36] |
| †Pterodactyloidae | non-traumatic | osteoarthritis | 1 | [36] |
| Pythonidae | neoplasm | chondro-osteofibroma | 1 | [36] |
| neoplasm | osteosarcoma | 1 | [36] | |
| Scincidae | infectious | unidentified infectious bone | 2 | [29] |
| Sphenodontidae | infectious | osteomyelitis | numerous | [45] |
| †Stegosauridae | traumatic | necrosis | 1 | [11] |
| Teiidae | non-traumatic | articular gout | 1 | [36] |
| non-traumatic | spondyloarthropathy | 1 | [29] | |
| infectious | unidentified bone infection | 1 | [29] | |
| neoplasm | unidentified tumour | 1 | [29] | |
| †Toxochelyidae | non-traumatic | avascular necrosis | 5 | [36] |
| Trionychidae | non-traumatic | avascular necrosis | 2 | [36] |
| unidentified reptile | traumatic | fracture | 1 | [11] |
| infectious | osteomyelitis | 1 | [11] | |
| Varanidae | non-traumatic | articular gout | 1 | [36] |
| non-traumatic | metabolic bone disease | 8 | [29] | |
| non-traumatic | enchondroma | 2 | [36] | |
| non-traumatic | haematoma | 3 | [29] | |
| non-traumatic | hypertrophic osteoarthropathy | 1 | [29] | |
| non-traumatic | osteoarthritis | 1 | [29] | |
| non-traumatic | osteochondritis | 1 | [29] | |
| non-traumatic | osteochondroma | 2 | [36] | |
| non-traumatic | osteoma | 1 | [36] | |
| non-traumatic | pseudogout | 1 | [36] | |
| non-traumatic | spondyloarthropathy | 27 | [29] | |
| infectious | unidentified infectious bone | 3 | [29] | |
| neoplasm | osteosarcoma | 1 | [29] | |
| non-mammalian synapsids | ||||
| †Edaphosauridae | traumatic | fracture callus | 2 | [11,13] |
| †Emydopidae | non-traumatic | double tusks | 1 | [50] |
| †Eodicynodontidae | non-traumatic | double tusks | 1 | [51] |
| †Geikiidae | non-traumatic | epideral inclusion cyst | 1 | [17] |
| †Gorgonopidae | benign tumour | odontoma | 1 | [23] |
| †Kannemeyeriidae | non-traumatic | double tusks | 1 | [52] |
| †Sphenacodontidae | traumatic | fracture callus | 5 | [11,13,15] |
| traumatic | greenstick fracture | 2 | [14] | |
| non-traumatic | osteoclerosis | 1 | [11] | |
| non-traumatic | osteohypertrophy | 1 | [11] | |
| non-traumatic | spondyloarthropathy | 2 | [36] | |
| infectious | osteomyelitis | 1 | [11] | |
| †Stahleckeriidae | traumatic | muscular avulsion | 2 | [17] |
| non-traumatic | epideral inclusion cyst | 1 | [17] | |
| infectious | fungal disease | 6 | [17] | |
| infectious | hydatid disease | 2 | [17] | |
| infectious | osteomyelitis | 3 | [16,17] | |
| †Titanosuchidae | infectious | osteomyelitis | 1 | [18] |
