| |
Incidence of Osteonecrosis of the Jaw in Women With Postmenopausal Osteoporosis in the Health Outcomes and Reduced Incidence With Zoledronic Acid Once Yearly Pivotal Fracture Trial
|
| |
| |
J Am Dent Assoc, Jan 2008 Vol 139, No 1, 32-40.
John T. Grbic, DMD, MS, M Med Sc, Regina Landesberg, DMD, PhD, Shou-Qing Lin, MD, Peter Mesenbrink, PhD, Ian R. Reid, MD, Ping-Chung Leung, MD, Noemi Casas, MD, Christopher P. Recknor, MD, Ye Hua, MD, MPH, Pierre D. Delmas, MD, PhD, Erik F. Eriksen, MD, DMSc; AND for the Health Outcomes Reduced Incidence With Zoledronic Acid Once Yearly Pivotal Fracture Trial Research Group
AUTHOR CONCLUSIONS
Evidence from randomized, placebo-controlled clinical trials with prospectively adjudicated adverse events needs to be considered when providing scientifically appropriate guidance regarding bisphosphonate use and the risk of developing ONJ to physicians, dentists and patients. Furthermore, such guidance should distinguish between patients at higher risk of developing ONJ (for example, patients with cancer) and those at lower risk (for example, patients with osteoporosis).
Our study is the largest trial to date to provide prospective data based on stringent diagnostic criteria and expert adjudication. From the available data, it appears that the development of ONJ in a population of women with osteoporosis is rare. In our study, the occurrence of ONJ in participants who received zoledronic acid was similar to that in participants who received the placebo. As with all therapies, dentists should consider the risks and benefits of using bisphosphonates in relation to the risk of developing ONJ. Based on available evidence, no alterations to current treatment practices for osteoporosis are warranted, and concepts such as so-called drug holidays (temporary cessation of bisphosphonate therapy) have the potential to harm patients at risk of experiencing fractures.
Based on the data from the HORIZON-PFT, the use of zoledronic acid as a once yearly 5-mg infusion significantly reduces the incidence of hip, vertebral and nonvertebral fractures, as well as the morbidity associated with them, including days of disability due to fracture or back pain. These findings need to be weighed against the rare occurrence of ONJ in the noncancer population.
The data we obtained in this study provide no evidence that healthy patients with osteoporosis require special treatment beyond routine dental care.
ABSTRACT
Background. The authors determined incidence of osteonecrosis of the jaw (ONJ) in a large, prospective three-year clinical trial of zoledronic acid in women with postmenopausal osteoporosis (PMO).
Methods. A total of 7,714 women with PMO received intravenous zoledronic acid 5 mg or a placebo. No spontaneous reports of ONJ were received. An independent, blinded adjudication committee searched the trialfs adverse event database by using 60 terms. On an ongoing basis, the committee reviewed the identified events, and it defined ONJ as exposed bone in the maxillofacial area with delayed healing for more than six weeks despite appropriate care.
Results. One participant who received a placebo and one participant who received zoledronic acid experienced delayed healing associated with infection. Both conditions resolved after antibiotic therapy, debridement or both.
Conclusion. The occurrence of ONJ is rare in a PMO population, and delayed healing of lesions can occur with and without bisphosphonate use over three years.
Clinical Implications. The low incidence of ONJ must be assessed in the context of the clinical benefit of zoledronic acid therapy in reducing hip, vertebral and nonvertebral fractures in this at-risk population. There is no evidence to suggest that healthy patients with osteoporosis who are receiving bisphosphonates require any special treatment beyond routine dental care or to support altering standard treatment practices.
Key Words: Bisphosphonate-associated osteonecrosis; jaw; bone; drug therapy; pharmacology; zoledronic acid
Abbreviations: ASBMR: American Society for Bone and Mineral Research ¥ HORIZON-PFT: Health Outcomes and Reduced Incidence With Zoledronic Acid Once Yearly Pivotal Fracture Trial ¥ MedDRA: Medical Dictionary for Regulatory Activities ¥ ONJ: Osteonecrosis of the jaw ¥ PMO: Postmenopausal osteoporosis
Osteoporosis is a disease caused by decreased bone mass, which leads to an increased risk of experiencing bone fractures, particularly at the hip and vertebrae. Approximately 300,000 hip fractures occur annually in the United States,1 and approximately one-fifth of patients who have experienced these fractures will die within a year after their hip fracture.2 Vertebral fractures are more than twice as common and are associated with increased morbidity and mortality.3,4
Antiresorptive therapy that includes bisphosphonates reduces excessive bone turnover to maintain or increase bone mass with a goal of decreasing the risk of experiencing vertebral, hip and other nonvertebral fractures. Clinical studies have shown that nitrogen-containing bisphosphonates are effective in reducing fracture risk in patients with osteoporosis.5-10 These drugs also are used clinically to treat Paget disease of bone, hypercalcemia of malignancy and skeletal-related events associated with bone metastases. Most bisphosphonates have been developed as oral compounds, but several studies have demonstrated suboptimal persistence of and compliance with the oral regimens, thus compromising the drugfs effectiveness.11-13 Intravenous administration of long-acting bisphosphonates such as zoledronic acid overcomes the problems identified for oral compounds, which could lead to better clinical outcomes.
In 2003, initial case reports of osteonecrosis of the jaw (ONJ) occurring in patients taking bisphosphonates were published.14 These initial reports, subsequent case reports15 and spontaneous reports of ONJ have occurred primarily in patients with metastatic cancer who were taking high cumulative doses of intravenous bisphosphonates to prevent skeletal-related events associated with bone metastases or hypercalcemia of malignancy. The patients described in the case reports typically were in the advanced stages of cancer and had received cancer chemotherapy, including cytotoxic agents and corticosteroids; all of these factors have a negative impact on wound healing. Therefore, it is unknown to what extent bisphosphonates may have contributed to these patientsf risk of developing ONJ. A small number of case reports have described similar lesions in patients without cancer who were being treated for osteoporosis14,16-19 or Paget disease.19,20
Based on data from case report studies, the estimated prevalence of ONJ in patients with cancer who are receiving intravenous bisphosphonates ranged from 0.8 to 10 percent.15 This wide discrepancy is due to differences in the definition of ONJ and the methodology used to define the lack of a case, as well as the inability to accurately quantify the number of patients who are receiving bisphosphonate therapy. Since there is not a clear definition of ONJ, patients who experience delayed healing after dental procedures may have received diagnoses of ONJ, especially if no minimum duration of exposed bone was considered before the diagnoses were made. Recently, guidelines published by the American Association of Oral and Maxillofacial Surgeons suggested that bone be exposed a minimum of eight weeks before a lesion is diagnosed as ONJ.21
Although ONJ has been reported in patients with osteoporosis who are receiving oral bisphosphonate therapy,14,16,17 the number of cases is too small to provide accurate prevalence estimates. A retrospective study in Germany found that of 780,000 patients who received bisphosphonate therapy for osteoporosis, three had ONJ, yielding an estimated prevalence rate of 0.00038 percent or a risk of less than one in 100,000 patients.22,23 This estimate is similar to the reported risk of 0.7 cases per 100,000 patient-years with oral alendronate use.24 A higher prevalence rate was reported in a study that solicited cases of ONJ via a two-page postal survey sent to oral and maxillofacial surgeons in Australia and New Zealand. The studyfs authors compared the number of cases with the total number of bisphosphonate prescriptions in the region and calculated an estimated prevalence rate among patients with osteoporosis of 0.01 to 0.04 percent or one case per 2,260 to 8,470 prescriptions.25 Although prevalence estimates vary, these studies suggest that the risk of ONJ in patients without cancer receiving bisphosphonates is low.
The potential risk factors for the development of ONJ in patients who are receiving bisphosphonate therapy include cancer, cancer chemotherapy including the use of cytostatic agents and corticosteroids, dental extractions, periodontal disease and dental trauma.26 Studies have suggested that the risk of developing ONJ increases with multiple comorbidities.26,27
We conducted a large prospective three-year clinical trial (The Health Outcomes and Reduced Incidence With Zoledronic Acid Once Yearly Pivotal Fracture Trial [HORIZON-PFT]) to determine the incidence of ONJ in women with post-menopausal osteoporosis (PMO) who received a once-yearly infusion of 5 milligrams of zoledronic acid to reduce the incidence of hip, vertebral and nonvertebral fractures. We also assessed the incidence of ONJ in a placebo control group.
METHODS
Clinical trial design. HORIZON-PFT was a three-year, international, multicenter, randomized, double-blind, placebo-controlled clinical trial that enrolled women with PMO. The methods we used for primary efficacy and safety analyses are published elsewhere.28 We designed HORIZON-PFT to test the effectiveness and safety of a once-yearly infusion of 5 mg of zoledronic acid administered over 15 minutes and compared them with those of a once-yearly placebo infusion. In each group (zoledronic acid and placebo), we further stratified the participants into one of two groups on the basis of their usual care, including concomitant osteoporosis medication use at or before randomization. We allowed participants who had previously received bisphosphonates to be in our study after a prespecified washout period; for example, previous use of bisphosphonates for 48 weeks or more required two years of washout. Participants were not allowed to receive bisphosphonates other than the study drug during the study, but participants in Stratum II were allowed to continue taking other osteoporosis medications (for example, raloxifene, calcitonin, hormone therapy, tibolone). Participants in Stratum I were not permitted to take any of the available osteoporosis medications. All participants received calcium and vitamin D supplementation.
The incidence of morphometric vertebral fractures and hip fractures were the primary outcome variables. During the trial, we coded all adverse reactions using Medical Dictionary for Regulatory Activities (MedDRA) (International Federation of Pharmaceutical Manufacturers and Associations, Geneva) terminology.
Identification of potential ONJ cases. The occurrence of ONJ was not a primary endpoint of the study. To assess maxillofacial adverse events objectively and independently and to identify a possible association, we established an adverse event review process as the study was ongoing, and we identified potential cases on a prospective basis, while the adjudicators were blinded to whether the participant received the zoledronic acid or placebo.
An independent adjudication committee of five experienced dental specialists (two oral and maxillofacial surgeons, two oral pathologists and one periodontist) conducted the maxillofacial adverse event review. The members of the adjudication committee remained blinded to study treatment throughout their review of events. The committee followed up on the potential cases they identified during a search of the trialfs database by making queries of the database to secure a more detailed dental history and supportive documentation, such as photos and radiographs of the lesions or data on the clinical outcome. The committee did not report any of its findings to us or to any of the clinical investigators involved with the trial, nor did they make treatment recommendations. The committee met at regular intervals in person or by telephone whenever enough potential cases of ONJ had accumulated for review and discussed each case in detail. In addition to using 60 MedDRA terms (BoxGo) to search our adverse event database, they searched the concomitant medication and nondrug therapy panel in the study database to identify any surgical procedures that might be associated with maxillofacial complications. The committee defined potential cases of ONJ as exposed bone in the maxillofacial area that had delayed healing for more than six weeks despite appropriate care.
RESULTS
When recruitment was completed, we randomized 3,889 participants (mean age 73.1 years) into the zoledronic acid group and 3,876 participants (mean age 73.0 years) into the placebo group. We excluded 29 patients (14 in the zoledronic acid group and 15 in the placebo group) from all analyses because the clinical centerfs participation was terminated owing to major protocol deviations. In addition, we excluded 22 subjects (13 in the zoledronic acid group and nine in the placebo group) who were randomized but failed to receive treatment from the safety analysis. The demographics of the population we studied are presented in Table 1Go. We found no significant differences between the placebo group and the zoledronic acid group with respect to any parameter we examined at baseline. A total of 3,862 participants in the zoledronic acid group and 3,852 participants in the placebo group received at least one infusion of zoledronic acid or the placebo, respectively. A total of 3,409 participants in the zoledronic acid group and 3,517 participants in the placebo group received at least two infusions of zoledronic acid or the placebo, respectively, and 3,105 participants in the zoledronic acid group and 3,189 participants in the placebo group received all three infusions of the study drug or the placebo, respectively.
We received no spontaneous reports of ONJ during the study. After conducting searches of our studyfs database, we found that 101 of the participants in the zoledronic acid group (2.62 percent) and 127 of the participants in the placebo group (3.30 percent) had potential maxillofacial adverse events that met our criteria for further investigation and adjudication. The incidence of specific adverse events was similar between the zoledronic acid and placebo groups (Table 2Go). The most commonly reported adverse event in both the zoledronic acid group and the placebo group was sinusitis (86 cases [2.2 percent] and 103 cases [2.7 percent], respectively). Sinusitis was the only maxillofacial adverse event that occurred in more than one percent of participants in either group. The next most common maxillofacial adverse event was tooth abscess. After the adjudication committee conducted a thorough blinded investigation of all adverse events meeting criteria for adjudication, it found that only one participant in the zoledronic acid group and one in the placebo group had a lesion that met its definition for ONJ (FigureGo, page 37).
Details of ONJ in a placebo group participant. A 67-year-old woman in the placebo group who had ONJ had a lesion in the left maxilla that was inflamed and painful on palpation. Osteitis was diagnosed without any clinical evidence of a soft-tissue infection, and exposed bone was visible before debridement was performed. Two debridements of the infected area were performed, and the participant received antibiotic therapy to treat the condition. The time to resolution was approximately eight months. The participant reported receiving regular annual preventive dental care and taking multiple medications, including cromolyn sodium and prednisone. She had not previously taken bisphosphonates.
Details of ONJ in a zoledronic acid group participant. A 75-year-old woman with poorly controlled type 2 diabetes, end-stage organ complications, retinopathy and neuropathy had not received regular dental care. She also reported having a history of myocardial infarction. The participant initially had an abscess in a retained root from a previous tooth extraction. The residual root was extracted, and during the healing period, but before complete resolution, 12 additional extractions and curettage were performed. Within one week after the multiple tooth extractions, the participant became extremely ill and a severe periodontal infection was diagnosed. Hospitalization was recommended, but the participant refused. The infection spread to the mandibular bone and osteomyelitis was diagnosed. The reported osteomyelitis resulted in necrosis of part of the mandible, which was confirmed radiographically. The participant was treated with a course of antibiotics. Complete resolution occurred, which was documented both radiographically and with photographs. The patient reported taking multiple concomitant medications, including insulin, paracetamol, aminophenazone, barbital, acetylsalicylic acid, ibuprofen, fosinopril, indomethacin, belladonna alkyloids and heparin. She had no history of bisphosphonate use.
Effect of previous bisphosphonate use. Although a majority of the participants in our study received bisphosphonates for a maximum of three years, a substantial number of them had taken oral bisphosphonates before they were enrolled in this trial. Of the participants randomized into the zoledronic acid group, 565 (14.6 percent) had previously taken bisphosphonates for a mean of 1.14 years, and 205 (5.29 percent) had received bisphosphonates for four years or more before the start of our study. In the placebo group, 557 (14.4 percent) participants had received bisphosphonate therapy, of whom 209 had received bisphosphonates for four years or more. There were no cases of possible ONJ reported in the subgroups that had a longer duration of bisphosphonate exposure and no cases of possible ONJ reported among participants who had previously received bisphosphonates.
DISCUSSION
The clinical results of the HORIZON-PFT demonstrated that in women with PMO, a once-yearly infusion of zoledronic acid was able to reduce the risk of experiencing vertebral fractures by 70 percent, hip fractures by 41 percent and nonvertebral fractures by 25 percent over three years.28 Moreover, reductions in days of disability owing to fracture or back pain were demonstrated.
The incidence of ONJ in a cohort of patients who received bisphosphonates for three years for osteoporosis was low (< one in 10,000 patient-years), although estimates of incidence in which the number of cases is small are problematic. Furthermore, the finding of ONJ in a participant who received the placebo suggests that this disease does not occur only in patients who take bisphosphonates and indicates that ONJ lesions may have causative factors other than bisphosphonate use. Because of the low incidence of ONJ in both the zoledronic acid and the placebo groups, it may be difficult to prove that once-yearly intravenous administration of a bisphosphonate increases the risk of ONJ in women with PMO over three years. Furthermore, a significant number of participants had received bisphosphonate therapy for more than seven years before enrollment, and none developed ONJ. The case definition of ONJ used in this trial was broad and could have included similar-appearing lesions such as those of osteomyelitis.
The participant who received the placebo and developed ONJ received regular dental care, prompt diagnosis of her maxillofacial complications, and appropriate wound management with local and antimicrobial treatment. In contrast, the participant in the zoledronic acid group who had a case of ONJ had not received regular dental care, had a delayed diagnosis and chronic local dental complications, refused to comply with recommended treatment and had had surgery in an infected area. Her poorly controlled type 2 diabetes with end-stage organ complications created further challenges to her immune status and wound-healing capacity. Her condition was most likely a combination of her underlying comorbidities, multiple surgical extractions in an infected area and refusal to be hospitalized. Despite her compromised physical condition, her lesion healed when the ONJ was treated appropriately. It is not possible to determine the contribution of the bisphosphonate therapy to the development of ONJ in this participant.
Our results and those of earlier research indicate that the incidence of ONJ is at least 100-fold lower in the osteoporosis population than in the oncology population.29-31 The marked difference in the incidence of ONJ may be related to the difference in the immunological competence and wound-healing ability of the two populations. Additionally, the larger cumulative dose of bisphosphonates administered to patients with cancer, which can be five to 10 times greater than those used in osteoporosis, may be a significant factor. The use of cytotoxic drugs and corticosteroids is known to decrease the numbers and functions of cells involved in the inflammatory response of wound healing, as well as epithelial cells, which may alter the wound-healing process and lead to the development of chronic non-healing wounds. In our study, both participants who may have had ONJ were taking medications (for example, the corticosteroid prednisone) or had diseases (for example, diabetes mellitus) that have been associated with poor wound healing, immunological compromise or both. Further research is required before the role bisphosphonates play in the settings of such comorbidities can be determined.
Possible effects on wound healing, infection, bone turnover and angiogenesis have been suggested as factors contributing to the pathogenesis of ONJ,15,17 but they remain poorly elucidated. No data on excessive reduction of bone turnover in patients with ONJ have been published to date. Thorough analysis of bone fractures in the HORIZON-PFT did not reveal that zoledronic acid had any adverse effects on fracture healing.28 The antiangiogeneic effect of bisphosphonates has been reported in in vitro and animal models.32-36 However, we found no evidence related to this in our study, and we are unaware of any studies demonstrating effects of bisphosphonates on angiogenesis in bone.
Bisphosphonates have been studied extensively in animal models of periodontal disease, and, despite dental interventions such as tooth ligation and inoculation of pathogenic bacteria, ONJ has not been reported in association with bisphosphonate treatment.37-42 These studies were performed in healthy animals. Therefore, it seems likely that development of appropriate models for ONJ may require induction of disease in an immunocompromised host. Histologic examination of early human ONJ lesions may be of value in addressing this issue. Such investigations, however, are hampered by reluctance among dentists and maxillofacial surgeons to perform biopsies of suspected ONJ lesions because they are concerned that the biopsy will further exacerbate the condition.26
ONJ could be considered to be an alteration in the normal sequence of wound healing, which results in a delayed or nonhealing wound with exposed bone. Chronic wounds have low cell mitoses and decreased proliferation rates, resulting in delayed or absent epithelialization and exposed bone. The presence of an infection within a wound will compromise its ability to heal further and may lead to bone necrosis, as is seen in many histologic evaluations of osteomyelitis. Therefore, regardless of the factors leading to ONJ, it appears that in a number of cases the associated infectious process needs to be managed appropriately. In HORIZON-PFT, both participants with ONJ had resolution of their lesions with appropriate care, including antibiotic treatment. Antibiotic therapy, however, may not be as effective in patients with cancer, whose level of immunocompetence may be significantly less than that typically seen in patients with osteoporosis.
Our study had certain limitations. The adjudication committee ascertained ONJ cases on the basis of a review of adverse event reports, and we did not include specific prospective dental examinations in the study design. Another limitation was that the majority of participants received bisphosphonates for only three years. It is unknown whether longer-term treatment with zoledronic acid in patients with osteoporosis would increase their risk of developing ONJ. However, we are conducting a longer-term three-year extension study to gather efficacy and safety data. In this study, participants who received zoledronic acid and completed the original three-year study will be randomized to either zoledronic acid or placebo groups for three years, and participants who received the placebo and completed the three-year study will receive zoledronic acid for three years.
The lack of an accepted and consistently applied case definition for ONJ has contributed to the variability in incidences reported in different studies.15,22,23,25,29 The definition of ONJ we used in our study was not based on a scientifically validated diagnosis. Instead, it was based on consensus opinion and generally conforms with the definition recently published by a task force of the American Society for Bone and Mineral Research (ASBMR); ASBMR recommends that ONJ be defined as a lesion that does not heal within eight weeks after identification by a health care provider.43 The definition we used was more stringent in that we specified the presence of a non-healing lesion of more than six weeks.
FOOTNOTES
DISCLOSURE: Novartis Pharma AG, Basel, Switzerland, provided financial support and the zoledronic acid used in this study. Dr. Grbic is a consultant to Novartis Pharma AG. Dr. Landesberg is on the Novartis Pharmaceuticals Advisory Board (East Hanover, N.J.). Dr. Mesenbrink is an employee of Novartis Pharmaceuticals and owns shares of and has options in Novartis Pharmaceuticals. Dr. Reid is a consultant to Novartis Pharma AG, Merck (Whitehouse Station, N.J.) and sanofiaventis (Paris). Dr. Casas is a consultant to Novartis de Colombia S.A. (Bogota, Colombia). Dr. Hua is an employee of Novartis Pharmaceuticals and own shares of Novartis. Dr. Delmas is a consultant to and speaker for Novartis Pharma AG. Dr. Eriksen is an employee of Novartis Pharma.
Dr. Grbic is a professor of clinical dentistry and the director, Division of Oral Biology and The Center for Clinical Research in Dentistry, Columbia University College of Dental Medicine, 630 West 168th St., Box 20, New York, N.Y. 10032, e-mail "jtg2@columbia.edu". Address reprint requests to Dr. Grbic.
Dr. Landesberg is an associate professor, Columbia University College of Dental Medicine, New York City.
Dr. Lin is a professor, Peking Union Medical College Hospital, Beijing.
Dr. Mesenbrink is a senior statistical expert, Novartis Pharmaceuticals, East Hanover, N.J.
Dr. Reid is a professor, Faculty of Medical and Health Sciences, University of Auckland, New Zealand.
Dr. Leung is the founding chairman and emeritus professor of Orthopaedics and Traumatology, Department of Orthopaedics and Traumatology, Chinese University of Hong Kong, Prince of Wales Hospital.
Dr. Casas is a professor, Universidad Nacional de Colombia, Bogota.
Dr. Recknor is the medical director, United Osteoporosis Centers, Gainesville, Ga.
Dr. Hua is an assistant medical director, Novartis Pharmaceuticals, East Hanover, N.J.
Dr. Delmas is a professor of medicine, University of Lyon, France; and the director, INSERM Research Unit 831, Lyon, France.
Dr. Eriksen is the global brand medical director, Novartis Pharma AG, Basel, Switzerland.
The authors thank John Cooke, BioScience Communications, New York City, for his editorial assistance in the preparation of this article.
The members of the adjudication committee were Dr. Sidney Eisig, Columbia
University College of Dental Medicine, New York City; Dr. Ellen Eisenberg,
University of Connecticut Health Center, Farmington; Dr. Maurizio Tonetti,
University of Connecticut Health Center, Farmington; Dr. David Shafer,
University of Connecticut Health Center, Farmington; Dr. Joan Phelan, New York
University College of Dentistry, New York City.
|
|
| |
| |
|
|
|
