Page 229 - Haematologica April 2020

Fractures and survival in MM
Discussion
Our population-based study, including over 14,000 MM patients diagnosed in Sweden from 1990 to 2013, shows that MM patients with a fracture at MM diagnosis are at 28% higher risk of dying than those without a frac- ture at diagnosis. Furthermore, after MM diagnosis, patients who develop a fracture are at a 2-fold increased risk of dying, compared to MM patients who do not develop a fracture. These observations underline the clin- ical importance of fractures in MM patients in a real- world setting, from diagnosis and throughout the course of the disease.
We found that MM patients with a fracture at diagnosis had an inferior survival compared to MM patients with- out a fracture at diagnosis which indicates that fractures are a proxy for a more established and/or active disease at diagnosis. We previously showed in a population-based study that individuals with monoclonal gammopathy of undetermined significance (MGUS), a precursor condition preceding MM, had a 74% increased risk of fractures at five years compared to controls.25 Furthermore, our group and others have shown that a prior diagnosis of MGUS or smoldering MM is associated with improved survival in MM, suggesting that early treatment leads to improved survival.26,27 Possibly, a prior diagnosis of a precursor state could reduce the risk of fractures and/or the impact of fractures on survival after MM diagnosis. Our findings underline the importance of thorough evaluation of bone disease in precursor states as well as in active MM.
Our findings of a 2-fold higher risk of dying for MM patients with a fracture compared to patients without fractures demonstrates the significant impact of fractures on patients with MM, and is a higher risk than previously reported in clinical cohorts.16 This is, to the best of our knowledge, the first real-world data analysis on the asso- ciation of fractures and survival in MM. These results were further confirmed in our landmark analysis, where
MM patients with fractures during the first months of the disease had a poorer survival than those without fractures. Our findings are partially in accordance with what has been observed in the general population, where fractures, especially osteoporotic fractures, have been associated with an increased risk of death, although not to the same extent as in MM patients.28-30 In patients with metastatic or progressed solid tumors, skeletal-related events have been associated with increased mortality, even when adjusted for stage and/or treatment.14,16 Fractures occurring after MM diagnosis may be an indication of aggressive relapse, although we cannot evaluate to what extent fractures are an independent prognostic factor because we do not have clinical data on these patients.
Our analyses on different subgroups of fractures in MM patients and controls indicate that the impact on survival in MM reflects both the effect of the fractures themselves as well as the progression of MM. Femoral fractures are well known to increase mortality in the elderly, especially in the first months after the fracture,29 and were associated with the highest risk of dying in both MM and the con- trols in our study. The impact of humerus fractures on sur- vival in MM, on the other hand, seems to reflect the effect of the progression of the disease, since MM patients who developed a humerus fracture after diagnosis had a 2.6- fold risk of dying compared to an only 1.7-fold risk in the controls with a humerus fracture. In addition to this, we found that ankle fractures were not associated with sur- vival in MM, highlighting the fact that our findings are specific to MM-related or osteoporotic fractures and not all fractures in general. Our results, therefore, indicate that the effect of fractures on survival in MM is both due to the direct impact of fractures as well as through progression of MM, suggesting that patients with extensive MM bone disease have a more aggressive disease.
We did not find any significant change between the two calendar periods in the risk of death after first fracture in MM. More effective treatment agents were introduced in
Figure 4. Risk of death in multiple myeloma (MM) patients who developed a fracture after MM diagnosis compared to patients who did not develop a fracture.
Fractures as time dependent co-variates, adjusted for age, sex, year of diagnosis and previous fractures. *First inpatient fracture. **All fractures registered as patho- logic. ***All vertebral fractures, both pathologic and others. CI: confidence interval; HR: hazard ratio; No: number.
haematologica | 2020; 105(4)
1071