Annals of Oncology Advance Access originally published online on October 26, 2008
Annals of Oncology 2009 20(2):331-336; doi:10.1093/annonc/mdn630
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
quality of life/supportive care/palliative care |
Initial experience with conservative treatment in cancer patients with osteonecrosis of the jaw (ONJ) and predictors of outcome
1 Department of Medical Oncology
2 Department of Nuclear Medicine, Antwerp University Hospital, Edegem
3 Department of Oral and Maxillo-Facial Surgery, ZNA Middelheim, Wilrijk, Belgium
* Correspondence to: Dr T. Van den Wyngaert, Department of Medical Oncology, Antwerp University Hospital, Wilrijkstraat 10, 2650 Edegem, Belgium. Tel: +32-473-542124; Fax: +32-3-8253308; E-mail: tim.van.den.wyngaert{at}skynet.be
 |
abstract |
|---|
 Top abstract introductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
Background: Overall survival (OS) and outcome of cancer patients with bisphosphonate-associated osteonecrosis of the jaw (ONJ) using conservative treatment (chlorhexidine 0.12% rinse, intermittent antibiotics, and careful sequestrectomy) are unknown.
Design: In all, 33 ONJ patients were studied for OS and ONJ outcome.
Results: Median duration of bisphosphonate treatment was 27 months (range 4–115) and was stopped in 25 (76%) patients. Nine (27%) cases presented with stage 1, 21 (64%) with stage 2, and 3 (9%) with stage 3 disease. During median follow-up of 23 months, 11 patients (33%) died (median survival 39 months), with no ONJ-related fatalities. Out of 30 assessable patients, 53% no longer had exposed bone, 37% had stable lesions, and 10% showed progressive necrosis. The hazard ratio for healing with doubling of bisphosphonate exposure was 0.419 [95% confidence interval (CI) 0.178–0.982; P = 0.045], stage 2 versus stage 1 disease 0.216 (95% CI 0.063–0.738; P = 0.015), and stage 3 versus stage 1 disease 0.084 (95% CI 0.008–0.913; P = 0.042). Cessation of bisphosphonate treatment did not influence outcome.
Conclusions: Conservative treatment of ONJ leads to mucosal closure in 53% of patients. Doubling the exposure time to bisphosphonates and higher stages of ONJ significantly reduce ONJ healing rates.
Key words: bisphosphonates, jaw, ONJ, osteonecrosis, outcome study
|
introduction |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
Osteonecrosis of the jaw (ONJ) is a recently recognized clinical entity that has been associated with the use of bisphosphonates, a class of therapeutic agents that is frequently used in the treatment of among others osteoporosis, skeletal metastases from solid tumors, hypercalcemia of any cause, multiple myeloma, and Paget's disease [1–3]. By consensus, ONJ has been defined as the persistence of exposed bone in the oral cavity, despite an adequate treatment for 8 weeks, without local evidence of malignancy and no prior radiotherapy to the affected region [4].
Approximately 95% of ONJ patients suffer from cancer and receive bisphosphonates for the prevention or treatment of skeletal-related events [5]. The prevalence of ONJ among patients with malignancy has been reported to vary between 
1.0% to 9.9% depending on the studied population, possibly because of differing underlying baseline risks for osteonecrosis [6–9].
Particularly the group of the potent nitrogen-containing bisphosphonates (e.g. pamidronate, zoledronic acid, alendronate, risedronate, and ibandronate) has been associated with this disease [5]. Although the underlying pathophysiology is not well understood, a number of theories have been proposed. The powerful inhibition of bone resorption by the bisphosphonates and the subsequent reduction of remodeling capacity are believed to leave the jaw unable to adapt to external stresses such as the repeated trauma of mastication. The particular vulnerability of the jaws is theorized to stem from a higher dependence on active bone turnover compared with other parts of the skeleton and the thinness of the mucosa separating the bone from a microbiologically heavily colonized oral cavity [10]. Other factors including secondary infection, inhibition of angiogenesis, and impaired wound healing caused by cytotoxic or steroid therapy have also been suggested to play a role [11].
No definitive standard of care has been set for ONJ and aggressive surgery has been reported to be counterproductive, often exacerbating bone exposure. Hence, consensus guidelines have been published suggesting a conservative approach, consisting of rinses with chlorhexidine, intermittent antibiotic therapy, and careful removal of sequesters [4, 12].
However, the evolution of ONJ lesions using this conservative therapy and the patient-specific factors influencing the outcome of ONJ have hitherto not been investigated for patients with a malignancy. Additionally, the survival of cancer patients after receiving the diagnosis of ONJ is unknown. Indeed, the assessment of the impact of ONJ on the quality of life is complicated by the fact that the length of time that patients have to cope with this potentially debilitating condition has not been determined.
|
methods |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
patients and end points
A cohort of 33 cancer patients with a history of treatment with an i.v. administered nitrogen-containing bisphosphonate because of metastatic disease to the bones, and fulfilling the diagnostic criteria of ONJ was treated conservatively in keeping with current published guidelines, using nonalcohol-containing chlorhexidine 0.12% rinse tid and intermittent oral antibiotics as clinically required. The extent of ONJ was staged according to the guidelines published by the American Association of Oral and Maxillofacial Surgeons. Patients with stage 1 ONJ have exposed bone in the oral cavity but are asymptomatic and have no signs of infection. Stage 2 disease is characterized by infection, as evidenced by pain and erythema with or without purulent drainage. Finally, stage 3 is diagnosed when pathological fracture, extraoral fistula, or osteolysis extending to the inferior border complicates ONJ [4].
Antimicrobial treatment consisted of amoxicillin–clavulanate 875 mg bid or clindamycin 300 mg three times daily during 10 days followed by a maintenance therapy with doxycycline 100 mg once daily for a minimum of 3 weeks or until improvement of local soft tissue swelling or purulent discharge. Careful debridement of bone sequesters was undertaken only if further trauma to the adjacent bone could be avoided or when protruding bone edges were at risk of damaging surrounding soft tissues. Pain control was established by preferably using oral administration of acetaminophen, nonsteroidal anti-inflammatory drugs, or opioids. Dentures were adjusted if necessary to minimize pressure points and to reduce soft tissue irritation. Hyperbaric oxygen therapy was not used in any patient included in this series. The continuation of bisphosphonate treatment after ONJ was diagnosed was left to the discretion of the treating physician.
Studied end points were the overall survival (OS) after the diagnosis of ONJ was established and the outcome of ONJ lesions, which was coded as the closure of the mucosal defect, persistence of exposed bone, or progressive necrosis. A set of 20 patient-specific variables empirically chosen on the basis of the current understanding of ONJ pathophysiology was recorded for each patient to identify possible predictors of ONJ outcome (Figure 1). The anticancer treatments were coded as ‘prior’ if they were ongoing or completed within 6 months before developing ONJ. Therapies before this period were not considered to have an impact on ONJ healing. Patients who received sequential treatment with different bisphosphonates were analyzed according to the bisphosphonate to which they had the longest exposure time.
|
statistical methods
Survival analysis was carried out with the Kaplan–Meier method. Candidate predictor variables of ONJ outcome were tested using univariate Cox proportional hazards regression analysis. Variables that were significant at P < 0.10 in univariable analysis were included in multiple Cox proportional hazards regression modeling. Crude and adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) are reported. In case of non-normality or heterogeneity of variance, the logarithmically transformed values were used. The base-2 logarithmic transformation was used to ease the interpretation of HRs. Continuous variables entered in Cox regression were standardized about their mean and categorical predictors were indicator coded as appropriate. The assumption of proportional hazards was confirmed for all covariates in the model with time interaction testing and by plotting the Schoenfeld residuals against time. A linear relationship between included predictors and the log-hazard function was verified by plotting Martingale residuals against the studied covariate. The absence of multicollinearity was checked by calculating the correlation matrix of the regression coefficients. Possible outliers were identified with the DfBeta statistic [13]. Finally, all covariates included in the model were tested for interactions and an additional model controlling for gender and age was constructed to rule out the possibility of effect modification that could have gone undetected in univariable analysis. Hypotheses were tested nondirectionally with a P value of <0.05 considered to be significant. All analyses were carried out using SPSS version 16 (SPSS Inc., Chicago, IL).
|
results |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
The general patient characteristics and treatment history of the studied group are summarized in Table 1. Overall, patients were treated with bisphosphonates for a median of 27 months (range 4–115) at the time of ONJ diagnosis (Table 2). Of the six patients who received sequential bisphosphonate therapy, three were assigned to the pamidronate group and three to the zoledronic acid group, according to the longest duration of treatment on either bisphosphonate. The precipitating event leading to ONJ was a dental extraction in 25 cases (76%) and affected the mandible in 17 (52%), the maxilla in 11 (33%), or both jaws in five patients (15%). No traumatic event leading to ONJ could be identified in eight cases. The ONJ lesions were unilateral in 25 patients (76%), while bilateral ulcers were present in eight cases (24%). Using the proposed staging mechanism for ONJ, nine patients (27%) presented with stage 1 disease, 21 (64%) with stage 2, and three cases (9%) with stage 3 ONJ. The bisphosphonate treatment was stopped by the treating physician in 25 patients (76%).
|
|
The oncological treatment modalities in the studied group varied widely and included chemotherapeutical agents (anthracyclines, alkalyting agents, antimetabolites, taxanes, vinca alkaloids, and platinum compounds), hormonal treatments (antiandrogens, antiestrogens, aromatase inhibitors, and gonadotropin-releasing hormone agonists), and inhibitors of angiogenesis (thalidomide and bortezomib).
The median time of follow-up for the cohort was 23 months (range 2–66), during which 11 patients (33%) succumbed to complications directly related to their malignancy, giving a median OS of 39 (95% CI 21–56) months. No recorded fatalities could be attributed directly to ONJ, its complications, or the instituted treatment.
With a strictly conservative treatment, closure of the mucosal lesions with complete covering of the previously exposed bone was accomplished in 16 patients (53%), with a median time until mucosal healing of 20 months (range 2–54). In 11 patients (37%), the oral lesions remained stable or showed only partial closure. In contrast, three cases (10%) developed progressive exposure of necrotic bone or additional complications related to ONJ, such as a pathological fracture or extraoral fistula. The effect of the treatment of ONJ could not be evaluated in three patients (9%) due to rapidly progressive disease and subsequent death.
Stratified by the stage of ONJ at presentation, six of seven patients with stage 1 ONJ had mucosal healing, compared with 9 of 20 and one of three with stage 2 and 3, respectively. All patients in whom progressive necrosis was seen had initially presented with stage 2 disease. In the group of patients with stage 1 or 3 disease, no worsening of osteonecrosis was recorded.
Visual inspection of the histograms of the predictor variables collected for each patient revealed a highly positively skewed distribution of the duration of bisphosphonate therapy, which was corrected by logarithmic transformation.
The results of univariate Cox proportional hazards regression analysis of the studied variables are summarized in Figure 1. The crude HR for the healing of mucosal lesions associated with a doubling of the exposure time to bisphosphonates at the time of ONJ diagnosis was 0.389 (95% CI 0.187–0.809; P = 0.012). Compared with cases who presented with stage 1 disease, patients with stage 2 or 3 ONJ had crude HRs for mucosal healing of 0.165 (95% CI 0.049–0.552; P = 0.003) and 0.054 (95% CI 0.005–0.573; P = 0.015), respectively. The use of chemotherapy before the development of ONJ was linked with a crude HR of healing of 0.323 (95% CI 0.118–0.884; P = 0.028). A linear relationship with the log-hazard function and the assumption of proportional hazards were confirmed for these covariates. Additionally, no significant outliers were identified.
Subsequent multiple Cox proportional hazards regression using the significant covariates from univariable analysis yielded a model (overall –2 log-likelihood test P = 0.002) with adjusted HRs for healing with the doubling of bisphosphonate exposure time of 0.419 (95% CI 0.178–0.982; P = 0.045), stage 2 versus stage 1 disease 0.216 (95% CI 0.063–0.738; P = 0.015), stage 3 versus stage 1 disease 0.084 (95% CI 0.008–0.913; P = 0.042) (Table 3). Controlling for the stage of ONJ and the duration of bisphosphonate treatment, the use of chemotherapy before the onset of ONJ was no longer a statistically significant covariate in the model (P = 0.169). No multicollinearity or significant interactions were detected between the included significant covariates. Additional testing was carried out to detect effect modification by gender and age, by forcing these variables into the model, even though the preset threshold of significance was not reached in univariable regression for these covariates. In accordance with univariable regression, no statistically significant HRs were obtained for gender (HR 0.361; 95% CI 0.076–1.715; P = 0.200) or age (HR 0.948; 95% CI 0.883–1.017; P = 0.136), and the statistical significance of the HRs for the other included covariates in the model was not changed, ruling out a possible interaction.
|
In summary, from the studied set of 20 variables only the exposure time to bisphosphonates and the stage of ONJ at diagnosis were found to be significant predictors of the outcome of ONJ lesions in multiple regression analysis. The fitted model shows that in the studied cohort and controlling for the stage of ONJ, a doubling of the exposure time to bisphosphonates reduces the rate of healing with 58%. Conversely, controlling for bisphosphonate exposure patients who present with stage 2 or 3 disease have a 78% and 92% reduction in the rate of healing, respectively, compared with patients with stage 1 ONJ.
|
discussion |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
ONJ is a novel clinical entity that could possibly be considered a late toxicity of the treatment with potent nitrogen-containing bisphosphonates, even though the true pathophysiological disease model is not well understood and may be multifactorial. Other serious side-effects of bisphosphonate therapy, including acute renal failure after i.v. administration and gastrointestinal toxic effects associated with oral use, can be effectively prevented by respecting infusion times, adjusting the dosage according to the renal function, maintaining adequate hydration, and adhering to postdose posture requirements. Nevertheless, this class of therapeutic agents is in general well tolerated, which has undoubtedly contributed to their widespread use [14].
In oncology, the primary goal of bisphosphonate therapy is to maintain or improve the quality of life of patients, as the disease is usually incurable when bone metastases are present [15]. However, the occurrence of ONJ can lead to recurrent episodes of pain, swelling, and infection, reducing the quality of life, which the treatment with bisphosphonates was intended to safeguard. In particular, the inability to wear dentures can have an important functional and psychological impact on patients.
The general population characteristics of the presented cohort are comparable to those previously reported in the literature, warranting further investigation of prognostic factors. Indeed, our data confirm the high prevalence of trauma as the precipitating event of ONJ (76%), compared with 80% reported in other published series, which we reviewed elsewhere [5].
However, it should be recognized that the reported OS of 39 months in our cohort may not necessarily be applicable to other ONJ populations because of differences in the underlying malignancies and the impact of individual disease control. Nonetheless, our data illustrate that cancer patients can be exposed to the complications and symptoms of ONJ for many years during their illness. Additionally, we show that with conservative treatment ONJ itself should not be a cause of mortality in cancer patients.
Although the difference in exposure time to bisphosphonates between the patients treated with pamidronate alone or pamidronate followed by zoledronic acid was not statistically significant (P = 0.972), both these groups had a significantly longer duration of bisphosphonate treatment at the time of ONJ diagnosis compared with zoledronic acid users (P = 0.005), in a one-way analysis of variance (ANOVA) with Bonferroni correction (Table 2). However, the lack of a control group of non-ONJ patients in this study precludes any conclusions on the earlier occurrence of ONJ in the zoledronic acid group.
From the 20 patient-specific variables that were examined as possible predictors of ONJ outcome, three were statistically significant in univariate analysis: the total duration of bisphosphonate therapy, the presenting stage of ONJ, and the use of chemotherapy before developing ONJ (Figure 1). With respect to bisphosphonate exposure, the new finding that with longer duration of therapy the rate of healing of lesions diminishes could provide intriguing insights into the underlying disease process. In particular, an excessively strong inhibition of bone turnover has been suggested as the core mechanism for the development of ONJ, preventing osteoclast-mediated removal of debris after trauma and inhibiting the natural bone remodeling necessary to repair microfractures [1]. Moreover, the pharmacokinetics of i.v. administered bisphosphonates is characterized by a prompt plasma clearance, but shows a terminal half-life in the skeleton that can be in excess of 8 years [16]. Our data would support a disease model of ONJ where there is accumulation of the agent in the jaws that increases with the length of treatment, resulting in further inhibition of regenerative capacity and subsequently requiring a longer time until healing of the lesions after ONJ has occurred.
However, at the moment our findings do not suggest that the cessation of bisphosphonate therapy has a positive effect on the healing rate of ONJ lesions. Nonetheless, we recognize that if such an effect were to exist, it could have gone undetected in the current cohort as the bisphosphonate treatment was stopped in the majority of patients (76%). In addition, compared with the sometimes long duration of bisphosphonate therapy before the onset of ONJ (up to 9.5 years), the current median time of follow-up of 23 months could be insufficient to detect a beneficial effect.
The rate of healing of ONJ lesions in the studied group was also significantly associated with the stage of ONJ at presentation, with higher stages linked with lower healing rates. This result suggests that the proposed staging system, which is based on reported symptoms and clinical findings, is not only valid to guide the treatment but also can be used to predict outcome.
In univariable analysis, the use of chemotherapy before the onset of ONJ was found to reduce the healing rate in this cohort with 67%, but was no longer a statistically significant covariate when controlling for the length of bisphosphonate therapy and ONJ stage. A possible explanation might be that the patients in our cohort who were treated with chemotherapy before developing ONJ tended to have a higher stage of ONJ at presentation (15% stage 1, 70% stage 2, and 15% stage 3 versus 46% stage 1, 54% stage 2, and 0% stage 3) and had received bisphosphonate therapy for a longer time (median 33 versus 25 months), although these differences were not statistically significant (P = 0.150) in an exploratory two-way ANOVA.
At this time, there is no accepted standard for treating patients with ONJ. In fact, multiple therapeutic strategies have been proposed in the literature, including hyperbaric oxygen therapy, ozone therapy, and laser treatment [17–19]. However, most reported series are limited in size and feature no comparator treatment. The benefit of these therapies will have to be evaluated prospectively in the light of our findings to confirm their superior outcome and warrant future use because a strictly conservative treatment alone can lead to healing of mucosal lesions in about half of patients. Nonetheless, patients presenting with stage 2 or 3 ONJ could potentially benefit from further research to identify additional therapeutic strategies.
In the meantime, the prevention of ONJ should receive priority, where dentists should actively enquire about the possible use of bisphosphonates, be aware of the early clinical signs of ONJ, and avoid invasive dental treatments. Implementing these strategies could potentially lead to a reduction of 80% in incident ONJ cases. Similarly, prescribing physicians should adhere to the proposed guidelines where every patient who requires bisphosphonate therapy is screened by a dental professional before initiating the treatment [4]. Finally, the threshold for seeking dental advice in the oncologic patient treated with bisphosphonates should be low and multidisciplinary consultation is strongly encouraged to facilitate the early recognition and treatment of ONJ.
|
acknowledgements |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
We gratefully acknowledge the statistical assistance provided by Prof. J. Weyler from the Department of Epidemiology and Social Medicine (University of Antwerp, Wilrijk, Belgium). Disclosure: Dr Van den Wyngaert has received lecture fees from Novartis Pharmaceuticals Corporation. Dr Van den Wyngaert and Prof. Fossion have participated in scientific advisory boards for Novartis Pharmaceuticals Corporation. Statement on originality: The results presented in this paper have been published in part as abstract in the ASCO Proceedings of the 2008 Annual Meeting and were presented in part at the World Cancer Congress 2008. This manuscript or the data contained within are not under consideration for publication by another journal.
Received for publication June 25, 2008. Revision received August 8, 2008. Accepted for publication August 11, 2008.
|
references |
|---|
|
Topabstractintroductionmethodsresultsdiscussionacknowledgementsreferences |
|---|
1. Marx RE. Pamidronate (Aredia) and zoledronate (Zometa) induced avascular necrosis of the jaws: a growing epidemic. J Oral Maxillofac Surg (2003) 61(9):1115–1117.[CrossRef][Web of Science][Medline]
2. Devogelaer JP. Clinical use of bisphosphonates. Curr Opin Rheumatol (1996) 8(4):384–391.[Medline]
3. Ross JR, Saunders Y, Edmonds PM, et al. Systematic review of role of bisphosphonates on skeletal morbidity in metastatic cancer. BMJ (2003) 327(7413):469–474.
4. Advisory Task Force on Bisphosphonate-Related Osteonecrosis of the Jaws. American Association of Oral and Maxillofacial Surgeons position paper on bisphosphonate-related osteonecrosis of the jaws. J Oral Maxillofac Surg (2007) 65(3):369–376.[CrossRef][Web of Science][Medline]
5. Van den Wyngaert T, Huizing MT, Vermorken JB. Bisphosphonates and osteonecrosis of the jaw: cause and effect or a post hoc fallacy? Ann Oncol (2006) 17(8):1197–1204.
6. Hoff AO, Toth BB, Altundag K, et al. Frequency and risk factors associated with osteonecrosis of the jaw in cancer patients treated with intravenous bisphosphonates. J Bone Miner Res (2008) 23(6):826–836.[CrossRef][Web of Science][Medline]
7. Bamias A, Kastritis E, Bamia C, et al. Osteonecrosis of the jaw in cancer after treatment with bisphosphonates: incidence and risk factors. J Clin Oncol (2005) 23(34):8580–8587.
8. Wang EP, Kaban LB, Strewler GJ, et al. Incidence of osteonecrosis of the jaw in patients with multiple myeloma and breast or prostate cancer on intravenous bisphosphonate therapy. J Oral Maxillofac Surg (2007) 65(7):1328–1331.[CrossRef][Web of Science][Medline]
9. Van den Wyngaert T. Osteonecrosis of the jaw (ONJ) might explain the increased oral surgery risk in cancer patients treated with bisphosphonates. J Evid Based Dent Pract (2007) 7(3):132–135.[CrossRef][Medline]
10. Woo SB, Hellstein JW, Kalmar JR. Narrative [corrected] review: bisphosphonates and osteonecrosis of the jaws. Ann Intern Med (2006) 144(10):753–761.
11. Van den Wyngaert T, Huizing MT, Vermorken JB. Osteonecrosis of the jaw related to the use of bisphosphonates. Curr Opin Oncol (2007) 19(4):315–322.[CrossRef][Web of Science][Medline]
12. Ruggiero S, Gralow J, Marx R, et al. Practical guidelines for the prevention, diagnosis, and treatment of osteonecrosis of the jaw in patients with cancer. J Oncol Pract (2006) 2(1):7–14.
13. Landau S, Everitt B. Survival analysis. In: A Handbook of Statistical Analyses Using SPSS—Landau S, Everitt B, eds. (2004) Boca Raton, FL: Chapman & Hall/CRC Press. 245–276.
14. Conte P, Guarneri V. Safety of intravenous and oral bisphosphonates and compliance with dosing regimens. Oncologist (2004) 9(Suppl 4):28–37.
15. Brown JE, Neville-Webbe H, Coleman RE. The role of bisphosphonates in breast and prostate cancers. Endocr Relat Cancer (2004) 11(2):207–224.[Abstract]
16. Papapoulos SE, Cremers SC. Prolonged bisphosphonate release after treatment in children. N Engl J Med (2007) 356(10):1075–1076.
17. Freiberger JJ, Padilla-Burgos R, Chhoeu AH, et al. Hyperbaric oxygen treatment and bisphosphonate-induced osteonecrosis of the jaw: a case series. J Oral Maxillofac Surg (2007) 65(7):1321–1327.[CrossRef][Web of Science][Medline]
18. Agrillo A, Petrucci MT, Tedaldi M, et al. New therapeutic protocol in the treatment of avascular necrosis of the jaws. J Craniofac Surg (2006) 17(6):1080–1083.[CrossRef][Web of Science][Medline]
19. Vescovi P, Merigo E, Manfredi M, et al. Nd:YAG laser biostimulation in the treatment of bisphosphonate-associated osteonecrosis of the jaw: clinical experience in 28 cases. Photomed Laser Surg (2008) 26(1):37–46.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. A. Almazrooa and S.-B. Woo Bisphosphonate and Nonbisphosphonate-Associated Osteonecrosis of the Jaw: A Review J Am Dent Assoc, July 1, 2009; 140(7): 864 - 875. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||