The current manuscript attempts to compare and contrast the microscopic features from jaw lesions of patients with suppurative osteomyelitis, bisphosphonate-induced osteonecrosis of the jaws (BIONJ) and osteoradionecrosis (ORN). The authors, however, provide no underlying hypothesis or rationale for their research design and statistical evidence to support the claim of a significant number of study cases is never provided (despite use of “statistically” in the abstract). Furthermore, no subsequent statistical analyses of any results are provided. In addition, several other issues confound interpretation of this work. For example, the authors claim that the cases were compared in a “blinded fashion”; however, the method of blinding is never subsequently described in the Materials and Methods section. The authors state that medical records were redacted from the study materials, at least partly to justify their contention that the study was exempt from approval by the local Institutional Review Board (IRB). Yet, it is well known that BIONJ in cancer patients is often associated with significant comorbidities related to malignancy, cancer chemotherapeutic agents or both. Absent correlation between microscopic findings and the complete clinical setting with BIONJ patients, the relative contribution of bisphosphonates to reported tissue alterations remains obscure. This, in turn, makes the authors’ suggestion that the related microscopic findings are “bisphosphonate-specific or characteristic” suspect at best.

The microscopic evidence raises additional specificity issues. For example, the authors highlight separation of osteoclasts from underlying bone in Fig. 11 and imply that this is a feature of BIONJ. A case of osteomyelitis in Fig. 2, however, shows similar osteoclastic separation. Moreover, such separation is recognized by experienced pathologists as a common tissue processing artifact with bone specimens (see Figs. 1, 3, 7 and 13), independent of the presence of inflammatory or necrotic changes. Given the routine nature of this finding, its diagnostic significance or relevance within this study is questionable without more detailed microscopic evidence, including higher-magnification photomicrographs.

Beyond issues of specificity, several unusual microscopic descriptions bring the authors’ histopathologic expertise into question. Returning to Fig. 11, the legend describes osteoclasts in bisphosphonate patients as being larger and having more nuclei, supposedly in comparison to non-bisphosphonate patients. Interestingly, similar changes in osteoclast morphology have been noted by others, including Weinstein et al., but these articles are unreferenced. In describing the osteoclast nuclei, the authors write: “…many of which are disrupted spilling chromatin into the cytoplasm…”. There is, however, no evidence of either “disruption” or “spillage”. Furthermore, as they were not reported by Weinstein et al. and are not characteristic of apoptosis, any link of these “findings” to bisphosphonate use would be purely conjectural. Finally, the legend for Fig. 8 describes a circular space in the connective tissue as an “Empty remnant of a blood vessel with absence of endothelial cells but residual basal lamina…”. Unfortunately, no reference is provided to show that this finding is characteristic of ORN and it’s difficult to understand how such a vascular “cavity” could be maintained without an endothelial lining yet avoid platelet activation and thrombosis. In addition, comparison to Fig. 7 suggests that these circular spaces could simply be artifacts secondary to loss of round trabeculae of bone.

Most importantly, two other relevant studies by Hansen et al . are not cited; and their findings differ markedly from the current work. In the 2007 article, for example, a total of 45 patients diagnosed with osteoradionecrosis (16 patients), bisphosphonate-associated osteonecrosis (27 patients) or non-specific osteonecrosis (3 patients) were examined . Actinomycotic colonization of affected bone was reported in 100% of these patients, compared to 76% of BIONJ cases and 58% of ORN cases in the current study. Numerical discrepancies aside, Actinomyces species are well-known as rapid colonizers of oral hard surfaces, including exposed bone. Given this proclivity, experienced pathologists would predict a result similar to Hansen et al. with colonization approximating 100%. In the 2006 article, an inflammatory infiltrate (consisting of either neutrophilic granulocytes or mixed inflammatory cells) was observed in the medullary spaces of all 18 cases (8 bisphosphonate-associated osteonecrosis, 10 osteoradionecrosis) together with fibrosis. In contrast, the current work found marrow space inflammatory cells in 0 of 37 cases of BIONJ and reported “…There was a notable absence of inflammatory cells… throughout the marrow…” in a total of 45 cases of ORN.

In summary, the authors have provided a descriptive, albeit controversial, analysis of histopathologic specimens from patients with suppurative osteomyelitis, BIONJ and ORN. Unfortunately, the manuscript adds nothing of statistical or scientific significance to the literature. Should the authors wish to pursue related studies in the future, possibly to test a specific hypothesis, they would be wise to first identify a collaborator with experience and board-certification in pathology.