The Journal of Periodontology
Should Retract This Article

Bone grafts are designed to modify our patients’ healing and it is therefore critical that only accurate information is disseminated to the practicing clinicians. Unfortunately, our journals often lack the guidance to screen for misinformation about bone grafts, which can lead to the use of materials that in the best case scenario only fail and in the worst case harm the patient. Here, we look at the Journal of Periodontology as an example:

Comparing the histological assessment following ridge preservation using a composite bovine-derived xenograft versus an alloplast hydroxyapatite-sugar cross-linked collagen matrix
J Periodontol. 2022 Nov; 93(11):1691-1700.

This article starts off wrong by stating, “Autogenous bone is considered the gold standard for bone grafting due to its osteogenic, osteoinductive, and osteoconductive properties, and the absence of immunologic reactions with its use.” This is a false statement that is continually repeated and has no scientific support. Studies have shown that autografts are not osteogenic or osteoinductive. The authors reference this false statement by citing the following study:

The authors of this systematic review conclude, “The results of this systematic review of the literature suggested no clinically significant differences between autogenous bone grafts and other bone substitutes in terms of implant and prosthesis survival and success. Future studies should be promoted to compare outcomes of dental implants between non-autogenous bone grafts.” They also found that there was no more bone gained with autogenous bone. So, the article starts by making false claims about autografts and then references an article that contradicts their statement that autogenous bone is superior in any way to other graft materials. Did the Journal of Periodontology even read the article before they published it?

The article we are discussing today compared the degree of mineralization of two bone grafts after 4 months in extraction sockets. One of the grafts used was Bio-Oss Collagen, and the second graft was Ossix Bone, described as a sponge-like matrix of 80% synthetic microparticulate unsintered hydroxyapatite alloplast graft with 20% sugar cross-linked porcine type 1 collagen.

The study found that Bio-Oss produced 26.8% vital bone and Ossix produced 39.3% vital bone. The authors stated that both grafts produced low tactile bone density (soft bone). The study found that Bio-Oss produced an insertion torque between 20-40 NCM and Ossix Bone produced an insertion torque between 5-35 NCM, which baffled the authors. How can bone with much lower mineralization have higher insertion torque? The answer to that question is, the authors misread the histology and greatly overstated the amount of mineralization in the Ossix Bone histology.

Poorly mineralized bone crushes and becomes condensed when harvested with a trephine. When poorly mineralized bone is crushed with the pressure of a trephine, the soft tissue collapses and the mineralized tissue remains, producing a much higher percentage of mineralized tissue per area. This is a common finding in extraction sites that receive no treatment and it is what happened to the histology produced by Ossix Bone.
In the Ossix Bone histology below, the crest is mineralized. However, just below the crest, the bone is poorly mineralized and when drilling with a trephine, the crestal bone cleaves away from the underlying bone (dashed red line). The poorly mineralized tissue under the crest is crushed and condensed, giving the appearance of a higher percent mineralized tissue. In this article, the coronal 8 mm of the core samples were used to calculate the percent mineralization. Because the core sample for Ossix Bone was crushed, the actual depth of the core was likely 10 to 12 millimeters, rather than the 8 mm.
Ossix Bone 4 power. The crest is to the left of the sample (A). A few millimeters below the crest is a fracture line that shows where the mineralized crestal tissue separates from the poorly mineralized tissue below. In the trephine, the mineralized crest binds in the trephine and begins to spin with the trephine and crushes and condenses the poorly mineralized tissue below (area B) until solid bone is encountered deeper in the jaw (area C). Additionally, the authors claim that this histology contains no residual graft material. However, they fail to mention the large amount of amorphous nonvital inclusions found in the histology. While this material is unknown, it is feasible that this could likely be residual graft material. There is nothing normal about this tissue when compared to normal bone tissue.
Ossix Bone 10 power. It is obvious that there is no normal bone present and all of the tissue in this image is crushed and condensed, making the sample to look like it has a higher percentage of mineralized tissue. The authors failed to note the nonbiologic amorphous purple inclusion material (yellow arrows).
Bio-Oss 4 power. The authors state that the Bio-Oss particles are surrounded by bone. However, in this histology, which they say is a representative sample, shows that very few Bio-Oss particles are covered with bone. The majority of particles are encased in connective tissue. There is nothing normal about this tissue.
Bio-Oss with collagen 10 power. The bone is classified as vital, which is accurate, but no mention is made of the type of bone. This bone is commonly, yet mistakenly called woven bone. However this is sclerotic bone. While it takes a trained pathologist to discern the difference between woven bone and sclerotic bone, there is an easy way for authors to identify sclerotic bone. Separation artifacts (yellow arrows) between the Bio-Oss particle and the newly formed sclerotic bone is pathognomonic of sclerotic bone.
When sclerotic bone forms, it merely covers the particle in an effort to isolate the inflammatory proteins contained in the Bio-Oss particle. When a noninflammatory resorbable synthetic bone graft particle is covered by newly formed normal bone, no separation artifacts are found because the bone is integrated to the graft particle in the same way bone integrates to an implant with no separation artifact present in the histology.
In the 10 power histology, because the Bio-Oss particle is demineralized, the only remaining material is the inflammatory bovine protein matrix (RG). It is these proteins that produce the inflammation and the resultant sclerotic bone. In one respect, we must give credit to the authors in that they do not claim that Bio-Oss is deproteinized. Bio-Oss is full of bovine proteins and is not deproteinized. In this histology, the bovine protein matrix is clearly visible because they used a normal method of demineralization for 2 hours. In publications paid for by Bio-Oss, they incorrectly claim that Bio-Oss is deproteinized because the histology is commonly demineralized for up to a month. This destroys the bovine collagen matrix in the histology so the reader cannot see the bovine proteins. This publication was paid for by Ossix Bone. Hopefully, someday, the Journal of Periodontology will stop allowing authors to falsely claim that Bio-Oss is deproteinized.
The tissue is clearly crushed and condensed in the Ossix Bone histology,
but not crushed in the Bio-Oss histology.
While the authors claim that there is vital bone in the histologic samples, they accurately do not state that the bone is normal. The bone is clearly sclerotic in the Bio-Oss sample as evidenced by the lack of remodeling and the presence of separation artifacts between the newly formed mineralized tissue and the Bio-Oss particles. The Bio-Oss graft had higher insertion torque than Ossix Bone because it actually had higher percent mineralization. In the materials and methods, the authors outline how the bone density was graded (Lekholm and Zarb) and how the bone density was verified. However, the authors omitted this data. In our opinion, the omission was intentional because of the poor results and this is another reason for this article to be retracted.
While the discussion here is focused on the inappropriate methods for calculating percent mineralization, the most important aspect is the pathologic tissue these two graft materials produced. There is nothing normal about this tissue and it is another example of the Journal of Periodontology publishing articles with histology that has been shown to result in implant loss. The techniques used to calculate the percentage of mineralization are not accurate, resulting in false conclusions, so this article should be retracted.
One of the reasons for the pathologic histology is due to the addition of collagen to the graft material. Collagen should never be put in or over a bone graft as we outline in the following post:
One benefit of this publication is the presentation of the pathologic histology that explains why cadaver bone grafts and xenografts in particular are the only variable associated with early implant failure. Read more:

Our profession is woefully deficient in bone education. Our dental schools teach nothing about bone. Our specialty programs only teach how to do a bone graft, but provide minimal education about bone biology and never any meaningful education about bone graft biology. Our professors have no more knowledge about bone and bone graft biology than the average clinician. As a result, we continually see publications in what are intended to be scientific journals that lack any insight into the subject matter and like this article, the findings are completely wrong. Professors in the dental specialties should not publish studies on bone without the input of allied specialties. The most appropriate method of producing honest quality research on bone is to include authors who specialize in the fields of bone biology, pathology, and bone graft biology. At a very minimum, the journals should have these specialties review the papers prior to publication. No one with any amount of knowledge would think the tissue represented in this paper would be capable of implant integration let alone long term implant function.

For professors who are interested in improving the quality of their research and the clinician who is tired of being sold the latest bone graft fad, you are invited to attend an in-person 6-hour course on bone biology, bone graft biology and the clinical application of science based bone grafts to be held in Miami early March 2023. Contact SteinerBio for details.