1. Osteonecrosis of the Jaws in Myeloma BRIAN G.M. DURIE, M.D., Michael Katz, Jason McCoy, MS and John Crowley, PhD Hematology/Oncology, Cedars-Sinai Outpatient Cancer Center, Los Angeles, CA, USA; Web Support/Data Analysis, International Myeloma Foundation, Los Angeles, CA, USA; and Statistics, Cancer Research and Biostatistics, Seattle, WA, USA. Time Dependent Correlation with AREDIA and ZOMETA Use

2. Osteonecrosis of the Jaws – What Is It? Exposed bone in the maxilla or mandible Due to disruption of the resorption-remodeling cycle of bone and inhibition of endothelial cell proliferation Poor healing and secondary infection can lead to loss of teeth and segments of jaw bones. Pictures courtesy Dr. Sal Ruggiero

3. How Frequent Is Osteonecrosis? Rare prior to 2001 2003 - Marx* reported 36 patients 2004 - Ruggiero** et al reported 63 patients diagnosed 2001-2003 2004/ 2005 Myeloma specialty groups report an increased frequency; 2-5% of patients at IMF seminars in Dallas/ San Diego/ LA/ Portland/ Tucson indicate osteonecrosis diagnosis * JOMF SURG 61:115 2003 ** JOMF SURG 62:527 2004

4. Questions About Osteonecrosis Was the diagnosis missed prior to 2001? Probably Not It is an obvious dental problem What caused the increased frequency of ONJ? Not Clear Marx and Ruggerio et al proposed an association with bisphosphonate use

5. Important Current Questions/ Issues Is the likelihood of ONJ linked to use of Aredia and/or Zometa? To what extent do other therapies or disease processes have an impact? Are there identifiable risk factors? What is the magnitude/severity of the problem? Are myeloma patients particularly at risk for osteonecrosis (ONJ) e.g. versus breast cancer?

6. OUR STUDY Anonymous WEB Based Survey: Summer 2004 Included 1203 Myeloma(904) and Breast Cancer (299) patients Recruited using IMF email lists/web site plus “ACOR” myeloma and breast Listservs (email), Nexcura (email) and Y-Me National Breast Cancer Organization (web site) Evaluates dates for diagnosis, treatments and complications including dental findings

7. Increase in Treatment Options Over Time 1950-1960s1970-1980s1990s2000s Bortezomib (Velcade) Thalidomide Bisphosphonates Stem cell transplantation High-dose chemotherapy Vincristine, doxorubicin, dexamethasone Radiation Melphalan + Prednisone ALLO CLOD. THAL ALLO CLOD. VELCADE SCT HDC VAD SCT HDC VAD MP RAD STEROIDS SCT HDC VAD ALLO MP RAD STEROIDS RAD STEROIDS MP RAD STEROIDS Myeloma Rx AREDIA ZOMETA THAL AREDIA

8. Numbers of Patients Responding to Survey Total Patients 1203 299904 Osteonecrosis (ONJ) Suspicious findings* (SONJ) ONJ SONJ 13 23 62 54 36 116 * SONJ: Suspicious findings: bone erosions; bone spurs; exposed bone MyelomaBreast

9. Overall Likelihood of ONJ from Time of Diagnosis 904 myeloma patients

10. New Cases of ONJ Each Year Among Respondents 57 patients 12 patients

11. Frequency of Therapeutic Interventions in Myeloma Respondents Bisphosphonates 804/904 (89%) 57/62 (92%) AREDIA (ONLY) 267/904 (30%) 17/62 (27%) ZOMETA (EVER) 515/904 (57%) 40/62 (65%)ZOMETA (EVER) 515/904 (57%) 40/62 (65%) Steroids 738/904 (81%) 55/62 (89%) PREDNISONE 210/904 (23%) 24/62 (39%) DEXAMETHASONE 525/904 (58%) 64/62 (55%) Thalidomide 496/904 (55%) 37/62 (59%) Radiation to head/ neck 61/904 (7%) 3/62 (5%) Stem Cell Transplant 426/904 (47%) 26/62 (42%) OverallONJ

12. Increasing Incidence of ONJ Among Respondents from Date of Diagnosis Months from Diagnosis

13. Mean Time from Myeloma DX to Onset of ONJ or SONJ Zometa only 18 months* 19 months Aredia only 72 months* 32 months ONJSuspicious ONJ Bisphosphonate treatment MONTHS FROM DIAGNOSIS *ONJ: mean times for Aredia only and Zometa only significantly different, p=0.002.

14. 47% Pattern of bisphosphonates in patients with ONJ or SONJ Zometa* Aredia Alone MyelomaBreast 103 27 904 57 46 299 11 16 Overall ONJ SONJ Overall ONJ SONJ * Alone or switched to Zometa Myeloma 81%91%94% 19% 9 % 6 % 70% 68% Breast

15. ONJ Among Respondents vs. Duration of Aredia or Zometa Treatment 0% 20% 40% 60% 80% 100% 024487296120144 Log-rank P=.01 Months from start of Aredia or Zometa Events / N Z only 10 / 211 A only 14 / 231 A and Z 14 / 182

16. 0% 5% 10% 15% 20% 25% 0122436 Months from start of Aredia or Zometa Z only A +/- Z Events / N 10 / 211 10 / 413 36-Month Estimate 10% 4% P =.002 Duration of Aredia and/or Zometa use censored at 3 years

17. Prednisone Does Not Increase the Likelihood of ONJ Months from Diagnosis

18. Thalidomide and Dexamethasone Do Not Increase the Likelihood of ONJ Log-rank P > 0. 5 ThalidomideDexamethasone

19. The Increased Occurrence of ONJ and SONJ Since 2001 CORRELATES WITH The impact of Aredia after 6 years The impact of Zometa after 18 months * The highest risk (45%) is in the group of patients switching from Aredia to Zometa

20. No Difference in Likelihood of ONJ or SONJ in Myeloma versus Breast Cancer Duration of bisphosphonate therapy censored at 3 years

21. Zometa Only is Associated with earlier onset of ONJ or SONJ: MM and breast combined Duration of bisphosphonate therapy censored at 3 years

22. Frequency of Prior Dental Problems* Total Population 396/1203(33%) 59/75 (79%) Myeloma 294/904 (32%) 50/62 (81%) Breast Cancer 102/299 (34%) 9/13 (69%) Two sided P-value for dental problems and osteonecrosis: in Breast: 0.0129 in Myeloma: <0.0001 * Other than Suspicious ONJ findings Overall ONJ Patients

23. Conclusions Amongst the respondents to this survey Duration of bisphosphonate use in myeloma and breast cancer is associated with increased risk of Osteonecrosis (ONJ) 36 month estimates of ONJ are higher for Zometa versus Aredia None of the other therapies analyzed were associated with a time dependent increased risk of ONJ Patients with prior dental problems have a higher risk of ONJ It is likely that precautions related to dental care and bisphosphonates use may reduce the likelihood of ONJ

24. Acknowledgements Special thanks to Judith Peterson Special thanks to Vanessa Bolejack