Longitudinal cohort study of skeletal-related events (SRE) and long-term prognosis in patients with multiple myeloma (MM) with bisphosphonate-related osteonecrosis of the jaw (ONJ).

Abstract

e19645 Background: Nitrogen containing bisphosphonates (BIS) have been shown to decrease SRE and improve survival in patients with MM. BIS use has been tempered by the occurrence of ONJ. Little is known about the prognosis incidence of SRE in MM patients with ONJ. Could the development of ONJ be a biomarker of enhanced sensitivity to BIS, with reduced SRE and better survival? Methods: In this retrospective cohort study, 11 MM patients who met diagnostic criteria for ONJ between June 2005 and June 2010 were matched to 39 MM patients without ONJ from our institution’s database. Patients were matched based on age, stage, histology, gender, ethnicity, and performance status. Skeletal surveys were used to evaluate bone disease at the start of BIS infusions and during follow-up. SRE’s were defined as: a pathological fracture, surgery to treat pathological fractures, radiation to bone, spinal cord compression or hypercalcemia. SRE were recorded from start of BIS, events needed to be > 21 days apart to be counted. All data was censored at either death or last clinic visit. Results: ONJ patients were followed for median 4.3 ± 2.4 years, Controls 3.9 ± 2.1 years (p =0.6). Time from diagnosis to transplant was non significant (NS) between ONJ and controls. Diffuse vs. limited bone disease on skeletal survey at start of BIS was NS. ONJ patients got 20.8 ± 18.0 infusions of BIS, cases 25.6 ± 17.7 (p=0.4). Only one patient in the ONJ group died (9%), 14 of the 39 control patients died (36%) (p = 0.13). Odds ratio (OR) of death in the ONJ group was 0.179 (with 95% CI (0.021, 1.544). Log-rank test to examine overall survival between ONJ and control group tended to significance (p= 0.08). SMR (skeletal morbidity rate = SRE/ patient-year) was 0.1019 events/patient-year in ONJ patients (95% C.I. 0, 0.26) and was 0.1620 events/patient-yr in controls (95% C.I. 0.05, 0.26). The odds ratio for SRE in patients with ONJ was estimated to be 0.54 (95% C.I. 0.13-2.15) p-= 0.38. Conclusions: The low odds ratios of SRE and deaths in the ONJ group support the hypothesis that ONJ might mark for increased BIS sensitivity (and hence different prognosis), but small sample sizes underpower the result.