番茄社区app

Fractures due to osteoporosis are a serious health concern in millions of Americans living with osteoporosis.¹ Bone mineral density (BMD), the main screening marker for osteoporosis in the past, is not ideal because many fractures occur above the BMD threshold for osteoporosis. 

To address that gap, the Fracture Risk Assessment Tool (FRAX庐) was developed by the World Health Organization (WHO) in 2008. FRAX includes multiple clinical risk factors in its assessment and is now the most widely used tool to predict 10-year probability of major osteoporotic fractures (MOF).²

Patients with cancer are typically at greater risk of osteoporosis and fracture compared with individuals without cancer, even after adjusting for older age.³ Bone loss is not exclusively due to the disease itself, but also to therapies used to treat it.⁴

A recent study was carried out to evaluate FRAX as a prediction tool for incident fractures in patients with cancer.⁵

Principal investigator, Carrie Ye, MD, FRCPC, MPH, and colleagues published their findings in JAMA Oncology in October 2024. In the study, entitled 鈥淔racture Risk Prediction Using the Fracture Risk Assessment Tool in Individuals With Cancer,鈥� the authors reported that although FRAX does not have specific inputs for cancer or cancer treatments, it still appeared to provide accurate fracture prediction in cancer patients. 

Dr. Ye, a member of the McCraig Institute for Bone and Joint Health in Calgary, Alberta, Canada and an assistant professor at the University of Alberta, elaborated further during an interview with 番茄社区app. 鈥淲hile FRAX has generally performed quite well across many unique populations, I was surprised at how accurately FRAX predicted fractures in individuals with cancer when BMD is included, given the unique risk factors that cancer patients have for fracture,鈥� she explained. 鈥淎lso interesting was our finding that FRAX without BMD actually overestimates fracture risk, which underscores the importance of DXA [dual energy x-ray absorptiometry] scans in risk-stratifying individuals with cancer,鈥� she added. 

Study design and data

This retrospective cohort study was carried out using population-level healthcare administrative databases in the province of Manitoba in Canada. Adult patients with first cancer diagnoses other than nonmelanoma skin cancer between 1987 and 2014 were identified. Up to four controls were matched to each case based on age, sex, and area of residence on the date of cancer diagnosis. In both the cancer and non-cancer groups, only individuals aged 鈮�40 years who had received DXA scans were included in the final evaluation, resulting in 9877 individuals in the cancer group matched with 45,875 individual controls without cancer.

On average, the patients with cancer were slightly older than their non-cancer counterparts, more likely to be male, had a slightly higher body mass index, more likely to have a parent with a history of hip fracture (P<.001 for all), and were more likely to be current smokers (P=.004). Patients in that group also were more likely to have secondary causes of osteoporosis (P<.001) compared with the non-cancer group, which the authors theorized may have been a result of higher aromatase inhibitor use in the cancer cohort.

Osteoporosis undertreated in cancer patients

Study investigators also found a discord in osteoporosis management between the two groups. Despite similar rates of prior fracture history and FRAX MOF probability with BMD and higher rates of FRAX MOF without BMD in the cancer group, only 4.2% of cancer patients were on osteoporosis medications compared with 9.9% of non-cancer patients (P<.001). 

Follow-up times were also significantly shorter in the cancer cohort compared with the non-cancer cohort (mean [SD] 7.6 years [3.1] vs 8.5 years [2.6]; P<.001).

The authors reported a significantly higher incidence of MOF per 1000 person-years in the cancer cohort (14.5 vs 12.9; P<.001) with higher rates of hip fractures (4.2 vs 3.5 per 1000 person-years; P=.002) and humerus fractures (2.9 vs 2.4 per 1000 person-years; P=.02). Significantly more patients with cancer also were reported to have died during the study period (26.8% vs 12.4%; P<.001).

Data analysis

Using the fracture incidence and mortality data, the authors calculated 10-year cumulative probabilities of MOFs and hip fractures, which were then plotted against FRAX-predicted MOFs and hip fractures with and without BMD.

FRAX demonstrated a strong ability to predict incident fractures in all analyses. Significant effect modification by cancer status was noted for both the FRAX values for MOF with BMD (hazard ratio [HR] 2.05; 95% confidence interval [CI] 1.99鈥�2.11 vs HR 1.84; 95% CI 1.74鈥�1.95; P for interaction=.001) and without BMD (HR 1.93; 95% CI 1.88鈥�1.99 vs HR 1.73; 95% CI, 1.63鈥�1.84; P for interaction=.001) as well as FRAX values for hip fractures without BMD (HR 3.44; 95% CI 3.22鈥�3.67 vs HR 2.95; 95% CI 2.59鈥�3.37; P for interaction=.04). Conversely, the effect modification by cancer status for hip fractures with BMD was noted to be non-significant (HR 4.10; 95% CI 3.82鈥�4.41 vs HR 3.61; 95% CI 3.13鈥�4.15; P=.11).

FRAX with BMD predicted rates of incident MOF and hip fractures that were lower in the cancer cohort compared with the non-cancer group. The relationship was inverse when BMD was included in the FRAX score, with consistently higher rates in both cancer and non-cancer cohorts. Given these results, the authors noted the importance of BMD testing in patients with cancer, an obviously at-risk population. However, they recognize the need for more targeted investigation. 

鈥淲hile it is reassuring that FRAX predicts fracture risk well in the general cancer population, we know that this is a heterogeneous population with unique risk factors for fracture across cancer types, stages and treatments,鈥� Dr. Ye explained to 番茄社区app. 鈥淔urther validation in specific high-risk groups and for specific cancer therapies will further add confidence in our fracture risk stratification in individuals with cancer,鈥� she added.

Strengths of this study included the population size used in both cohorts in addition to the duration of follow-up. A limitation was that the study only included cancer patients who had undergone DXA scan, which may not be representative of a general cancer population. The majority of the study cohort were female, which may limit generalizability to male cancer patients. Finally, the authors point out that given the heterogeneity of cancer, their results may not be equally applicable to all cancer subgroups. 

FRAX may be a valuable addition to the oncology armamentarium 

In this study, researchers demonstrated that fracture risk could accurately be predicted in patients with cancer using FRAX, especially when BMD was included. 鈥淭he cornerstone of osteoporosis treatment is identifying those at risk. Now clinicians can confidently risk-stratify their patients with cancer using FRAX so that they can make the appropriate recommendations regarding management of their patients' bone health,鈥� Dr. Ye said.

Published: December 27, 2024