Bone 41 (2007) 318 – 320

www.elsevier.com/locate/bone

Perspective
Is bisphosphonate-associated osteonecrosis of the jaw caused by
soft tissue toxicity?

Abstract

Osteonecrosis of the jaw (ONJ) is a complication of high-dose bisphosphonate use, characterized by the finding of exposed bone in the oral
cavity. It has been assumed that the primary lesion lies in bone and is related to over-suppression of bone turnover, but it is unclear why such a
lesion should present with loss of the soft tissue covering of the mandible or maxilla as the primary clinical feature. A possible explanation of this
paradox is that bisphosphonate is accumulated in bone in concentrations sufficient to be directly toxic to the oral epithelium. This would result in
the failure of healing of soft tissue lesions (such as those caused by invasive dental procedures or by subclinical trauma from dentures) leading to
secondary infection of the underlying bone. This model would explain why bone resection is unhelpful in managing this problem, suggests that
low bone turnover caused by non-bisphosphonate drugs should not cause the same problem, and raises the possibility that agents which reverse
bisphosphonate effects in vitro might have a role in the management of ONJ.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Alendronate; Zoledronate; Pamidronate

Introduction over-suppression of bone turnover, but it is unclear why such a

Bisphosphonates are the most widely used pharmaceuticals
in the management of osteoporosis. In general, they are well
tolerated, the only common side effect from their oral use being
inflammation or ulceration of the upper gastrointestinal tract [1].
This side effect is circumvented by the use of intravenous
bisphosphonates, which has been widespread in oncology for a
number of years, and is now a route of administration being
developed for use in benign bone diseases such as osteoporosis
and Paget's disease. Oncologists use intravenous bisphospho-
nates in much higher doses than are necessary for the
management of benign conditions, and this has been accompa-
nied by the development of a new adverse effect attributed to
these drugs, osteonecrosis of the jaw (ONJ) [2,3]. The hallmark
of this condition is the finding of exposed bone in the oral cavity
[4]. Almost all cases have occurred in oncology patients after
several years' therapy, usually with monthly doses of pami-
dronate or zoledronate. In many cases, the precipitating event
appears to have been a dental extraction or other invasive
procedure. With local and/or systemic antibacterial therapy,
most lesions stabilize and some show healing. The limited
histology data available show evidence of bone necrosis and,
often, infection, but not usually of ischemia [5]. It has been
assumed that the primary lesion lies in bone and is related to
8756-3282/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.bone.2007.04.196
lesion should present with loss of the soft tissue covering of the
mandible or maxilla as the primary clinical feature. A possible
explanation of this paradox is that bisphosphonate accumulated
in bone is directly toxic to the oral epithelium, resulting in a
failure of healing of soft tissue lesions, whether caused by
invasive dental procedures or by subclinical trauma, such as that
from dentures.
Bisphosphonate toxicity to soft tissues has been known for
many years and is the reason that extravasation of bispho-
sphonate must be avoided during intravenous infusions.
Alendronate damage to subcutaneous tissue has recently
been documented again in an animal model [6]. Toxicity to
the soft tissues of the digestive tract has received more detailed
attention as a result of investigations into the mechanism of the
esophageal and gastric side effects of orally administered
preparations of these drugs. There is clear documentation of
bisphosphonate toxicity to gastrointestinal epithelia. For
example, alendronate 30 nmol/ml and risedronate 10 nmol/
ml inhibit epithelial cell proliferation in vitro by means of
inhibition of farnesyl diphosphate synthase, the same enzyme
which is the target of bisphosphonates in osteoclasts [7]. The
clinical sequelae of this phenomenon are well recognized in
the form of upper gastrointestinal side effects from bispho-
sphonate use, and also as oral ulceration occurring in patients
 Perspective
who suck bisphosphonate tablets [8]. Since high concentra-
tions of bisphosphonate in the oral cavity disrupt the oral
mucosa, it is reasonable that comparably high concentra-
tions in the underlying bone might produce a similar effect.
Tooth extraction or other dental trauma will result in local
release of bisphosphonate which will inhibit proliferation
of adjacent epithelial cells and slow healing of the physical
breach in the mucosal barrier, if the local concentration of
drug is high enough. Secondary bone infection is likely to
ensue.
The concentration of bisphosphonate reached in tissue
fluid adjacent to a dental extraction site can only be
estimated, but is likely to be in the toxic range in patients
receiving long-term high-dose intravenous treatment. Half of
an intravenous dose of bisphosphonate is taken up by the
skeleton, where it is retained long-term until that bone is
resorbed [9]. Therefore, after 4-year use of zoledronate in the
standard oncology dose of 4 mg/month, the average skeletal
content will be 70 nmol/g of bone tissue. Bisphosphonate
uptake into bone is in direct proportion to the local rate of
bone turnover (hence the value of these agents in bone
scintigraphy) and the alveolar ridges have high turnover [10]
so they will accumulate rather more zoledronate than other
skeletal sites. Based on the potency with which zoledronate
inhibits farnesyl diphosphate synthase [11], it would be
expected to be toxic to the oral epithelium at concentrations
of 1 nmol/ml. Therefore, its presence in alveolar bone at 100
times this level would make it plausible that local epithelial
toxicity will occur after release from adjacent injured bone.
The resulting slower healing of the mucosal breach will
prolong exposure of the underlying bone to oral microflora,
facilitating secondary infection which will further compound
the problem.
This postulated mechanism resolves the apparent anomaly of
a bone-targeted drug producing a lesion with a prominent soft
tissue component and provides an explanation of why bone
lesions are not present elsewhere. If high bisphosphonate content
resulted in generalized bone necrosis then we would expect to
see evidence of toxicity throughout the skeleton, in proportion to
bone turnover. Thus, patients treated with sustained high doses
of bisphosphonates might be expected to manifest pathological
vertebral or rib fractures, or collapse of the femoral head (as
occurs with other causes of necrosis at this site), neither of which
occurs. The present hypothesis also has some other important
implications. First, it explains why bone resection is unhelpful in
the management of this condition. Since this procedure would
increase local bisphosphonate release, it would be expected to
make matters worse, consistent with the clinical experience.
Second, ONJ would not be expected to arise in response to other
agents such as denosumab, which comparably depress bone
turnover but are not toxic to soft tissues. To date, ONJ has not
been reported with this agent but clinical experience with it is
still relatively limited. Third, this problem might arise in other
parts of the body where epithelial damage occurs immediately
adjacent to a bone surface. The case report of a syndrome
comparable to ONJ in the external auditory meatus [12] is
therefore of great interest. The nasal sinuses would be another
319
potential site where this might occur. Finally, the epithelial
toxicity of bisphosphonates is reversible in vitro by the addition
of geranylgeraniol or farnesol [7], the downstream products of
farnesyl diphosphate synthase. Topical application of these
compounds might accelerate healing of the epithelial lesions of
ONJ if bisphosphonate toxicity to soft tissue is a significant
etiological factor. There are no reports on the use of such
interventions, but assessment of these agents would be of great
interest in a clinical problem which often proves intractable.
The hypothesis that high bone content of bisphosphonate
results in toxicity to overlying soft tissue provides a promising
line of investigation which is consistent with the available data
relating to this condition. It now requires testing in in vitro an
in vivo models.
Acknowledgments
Dr. Reid has acted as a consultant for and received research
funding from Novartis, Merck, Amgen, Procter and Gamble,
Sanofi-Aventis, Roche and Lilly.
References
[1] de Groen PC, Lubbe DF, Hirsch LJ, Daifotis A, Stephenson W, Freedholm
D, et al. Esophagitis associated with the use of alendronate. N Engl J Med
1996;335:1016–21.
[2] Ruggiero SL, Mehrotra B, Rosenberg TJ, Engroff SL. Osteonecrosis of the
jaws associated with the use of bisphosphonates: a review of 63 cases. J Oral
Maxillofac Surg 2004;62:527–34.
[3] Marx RE. Pamidronate (Aredia) and zoledronate (Zometa) induced
avascular necrosis of the jaws: a growing epidemic. J Oral Maxillofac
Surg 2003;61:1115–7.
[4] Hoff AO, Toth BB, Altundag K, Guarneri V, Adamus A, Nooka AK, et al.
Osteonecrosis of the jaw in patients receiving intravenous bisphosphonate
therapy. J Clin Oncol 2006;24:8528 [suppl].
[5] Hansen T, Kunkel M, Weber A, Kirkpatrick CJ. Osteonecrosis of the jaws
in patients treated with bisphosphonates—histomorphologic analysis in
comparison with infected osteoradionecrosis. J Oral Pathol Med
2006;35:155–60.
[6] Moreira MS, Katayama E, Bombana AC, Marques MM. Cytotoxicity
analysis of alendronate on cultured endothelial cells and subcutaneous
tissue. A pilot study. Dental Traumatol 2005;21:329–35.
[7] Reszka AA, Halasy-Nagy J, Rodan GA. Nitrogen-bisphosphonates block
retinoblastoma phosphorylation and cell growth by inhibiting the
cholesterol biosynthetic pathway in a keratinocyte model for esophageal
irritation. Mol Pharmacol 2001;59:193–202.
[8] Rubegni P, Fimiani M. Bisphosphonate-associated contact stomatitis. N Engl
J Med 2006;355:E25–E25.
[9] Reid IR, Brown JP, Burckhardt P, Horowitz Z, Richardson P, Trechsel U, et al.
Intravenous zoledronic acid in postmenopausal women with low bone
mineral density. N Engl J Med 2002;346:653–61.
[10] Baron R, Saffar JL, Duflot-Vignery A. Alveolar bone remodelling in
the rat: normal status and effects of PTX and PTH on the remodelling
sequence and the osteoclastic pool. Calcif Tissue Res 1977;22:502–4
[Suppl].
[11] Dunford JE, Thompson K, Coxon FP, Luckman SP, Hahn FM, Poulter CD,
et al. Structure–activity relationships for inhibition of farnesyl diphos-
phate synthase in vitro and inhibition of bone resorption in vivo by
nitrogen-containing bisphosphonates. J Pharmacol Exp Ther 2001;296:
235–42.
[12] Polizzotto MN, Cousins V, Schwarer AP. Bisphosphonate-associated
osteonecrosis of the auditory canal. Brit J Haematol 2006;132:114.
320 Perspective

Ian R. Reid
Faculty of Medical and Health Sciences,
University of Auckland, Private Bag 92019,
Auckland, New Zealand
E-mail address: i.reid@auckland.ac.nz.
Corresponding author. Fax: +64 9 308 2308.

Mark J. Bolland
Andrew B. Grey
Department of Medicine, University of Auckland,
Auckland, New Zealand

31 January 2007