Predictive Factors for Failed Nonsurgical Management of Long Bone Metastasis and Myeloma.

Bone is the third most common organ affected by metastatic carcinoma, with the incidence of metastatic bone disease estimated to be more than 280,000 patients per year in the United States. 1 Osseous metastases can cause substantial morbidity related to pain, impaired mobility, spinal cord compression, and hypercalcemia. 2 In addition, pathologic fractures afflict up to one third of cancer survivors and may necessitate reconstructive surgeries, subjecting patients to potential complications. 3 , 4 Systemic treatment alone is typically not sufficient to eradicate bone metastasis, so additional local control is often needed to relieve pain, avoid fracture, and allow continued mobility. 5 Radiation is helpful in this respect, but local progression and pathologic fracture can occur even after radiation therapy; for example, postradiation fracture has been reported to occur in 7.7% of femoral metastasis. 6 As such, long bone metastases that are considered at risk for fracture are typically treated with a combination of surgical stabilization and radiation therapy. When indicated, surgical intervention should be performed sooner rather than later to reduce the morbidity of activity restriction and potential fracture. However, prophylactic fixation carries its own morbidity and risk, so appropriate patient selection is critical.

Over the past several decades, multiple systems have been proposed to assess fracture risk and thereby help determine those patients who may benefit from prophylactic stabilization. Mirels criteria, originally published in 1989, has become a historical standard of practice and may still be the most commonly used system; however, it does not account for primary tumor histology, bone modifying agents, and other patient-specific factors that may influence fracture risk. 7 Furthermore, some argue that Mirels score lacks specificity, leading to overtreatment, and high interobserver variance results in poor reproducibility. 8 , 9 Indeed, when measuring the lesion size, Howard et al. 10 claims that many practicing physicians are unclear if the “thirds” in the Mirels size criterion is referring to the proportion of bone shaft or cortical involvement. Newer published guidelines describe femoral shaft metastatic lesions of 30 mm or more in the greatest dimension of axial cortical involvement on CT as predictive of fracture risk; however, this criterion is less practical as CT imaging is not routinely performed during the assessment of fracture risk, and most clinical decisions are made using plain radiographs. 9 , 11 , 12

The primary aim of this study was to improve understanding of patients who will ultimately be indicated for surgery in addition to radiation. To accomplish this, we evaluated patient-, lesion-, and disease-specific characteristics as potential risk factors for postradiation fracture and/or stabilization. Our secondary aim was to report the overall survival of our patients relative to these characteristics, given that life expectancy should also be considered in determining patients who are likely to benefit from surgery. The ultimate goal was to improve care of patients with metastatic disease by preventing pathologic fractures, expediting mobilization, and avoiding unnecessary intervention.

Methods and Sample Size

Patient Selection and Inclusion/Exclusion Criteria

This retrospective cohort study used the institutional review board–approved Palliative Radiation Registry database, including all patients who underwent radiation therapy for bone tumors at the University of Pennsylvania Perelman Center of Advanced Medicine. We included patients >18 years of age who underwent radiation therapy for metastatic long bone lesions between May 2011 and February 2020. Notably, patients were referred for local control radiation therapy based on medical and surgical decision making by a multidisciplinary team of radiation oncologists and orthopaedic oncologists. This study excluded patients with primary bone tumors and those who fractured or elected surgical management before radiation therapy. Two hundred ninety-three lesions underwent plain radiograph image review, after which metastatic bone lesions that were less than 0.5 cm in maximum length, lesions not discernible by at least one reviewer, and lesions without two orthogonal radiographs were excluded, leaving 186 lesions (55 upper extremity and 131 lower extremity) available for study (Figure 1 ). The primary outcomes of this study were prophylactic fixation, fracture, or incidence of any skeletal-related events (SRE) representing prophylactic fixation or fracture after intention to treat with radiation therapy alone. The secondary outcomes were time from radiation therapy to prophylactic fixation, fracture, and/or death.

Medical Record Data Extraction

For eligible patients, we supplemented the Palliative Radiation Registry database at our tertiary care institution with relevant patient medical records from electronic medical record (EMR). Associated patient records were reviewed to determine clinical outcomes, including (1) incidence of prophylactic surgery following radiation therapy for persistent, progressive, or recurrent local disease or (2) incidence of pathologic fracture following radiation therapy. Timing of these events and overall survival relative to radiation therapy were noted. Furthermore, we recorded the date of death or last EMR appointment within our hospital system to control for unrecorded deaths and patients who no longer receive care from our network.

Patient demographic variables were collected through retrospective chart review and included patient age, sex, and race, as well as socioeconomic status, which was approximated using zip code. Several comorbidities were collected from EMR and scored using the Charlson comorbidity index (CCI). Fall risk was defined by designated fall risk in EMR, documented previous falls, difficulty with gait, or use of assistive walking devices. Insurance coverage was also noted.

Clinical characteristics of each patient's metastatic disease and its treatment were recorded, including use of disease modifying antiresorptive medications (denosumab, zoledronic acid, and/or pamidronate) and ECOG score as an estimation of severity of illness. Pain scores and qualitative descriptions of pain were obtained from the EMR and categorized into mild, moderate, or functional by an orthopaedic oncologist. Local metastatic lesion radiation therapy was characterized by single-fraction (8 Gy in one fraction) or multiple-fraction (30 Gy in 10 fractions, 20 Gy in five fractions, 24 Gy in six fractions) treatment course.

Primary histologies were categorized according to general responsiveness of bone disease to radiation therapy by a radiation oncologist and orthopaedic oncologist. High-risk histology included renal cell carcinoma, melanoma, and colon carcinoma. Intermediate-risk histology included non–small cell lung cancer, breast cancer, thyroid cancer, bladder cancer, urothelial cancer, and pancreatic cancer. Low-risk histologies included multiple myeloma, prostate cancer, diffuse large B-cell lymphoma, and small cell lung cancer. Histologies less typically seen with bone metastases were assigned into “other” risk, and these included esophageal cancer, pyriform sinus cancer, neuroendocrine cancer, pheochromocytoma, adrenocortical carcinoma, hepatic cell carcinoma, vaginal cancer, endometrial cancer, malignant teratoma, and MALT lymphoma.

Radiographic Review

Blinded radiograph review was independently done by multiple members of the study team, including two orthopaedic oncologists and two musculoskeletal radiologists, to achieve consensus on lesion size and cortical involvement. Based on this and the clinical pain assessments described above, Mirels score was calculated for each lesion.

Plain radiographs of the limb at presentation were reviewed, and digital measurements were taking using Sectra IDS-7 for tumor characteristics including location, lytic/blastic/mixed appearance, diffuse/solid infiltration, percentage width of cortical involvement, maximal length of the lesion, percentage width of the lesion across the bone shaft, area on anterior/posterior axis, and area on medial/lateral axis (Figure 2 , A and B). Ill-defined lesions or lesions that resulted in differing measurements over +/− 10% were then re-reviewed by an orthopaedic oncologist. For quality control, one out of every five images were selected to be re-reviewed by a musculoskeletal radiologist. Blastic/lytic/mixed lesion type, lesion location, and percentage width of the lesion were integrated along with pain rating into our Mirels score calculation for each lesion for comparison.

Statistical Analysis

Observations with missing variables were analyzed using an intent-to-treat paradigm. Regression analyses for dichotomous outcome variables was done using logistic regression and backward variable selection to estimate the risk of prophylactic fixation and fracture while accounting for characteristics of the tumor, patient, and treatment, per above. Interreviewer radiograph measurement reliability was assessed using intraclass correlation coefficient. Time-to-event plots were created using Kaplan-Meier methodology, and regression was done using Cox proportional hazard analysis. All statistical analyses were done using the R programming language. 13 Alpha value of 0.05 was used.

Results

One hundred eighty-six lesions in 171 patients met the inclusion and exclusion criteria for this study (Table 1 ). Sex distribution was relatively even (48.4% were female), and most patients (70.4%) presented with lower extremity lesions. The median age of the cohort was 63 years (interquartile range 54 to 70 years) (Table 1 ). Most lesions had a primary histology of carcinoma 153 (82.3%), whereas 28 lesions (15.1%) were myeloma, and the remaining 5 lesions (2.7%) were lymphoma or melanoma.

The primary outcomes of this study were SRE (prophylactic fixation or fracture) after intention to treat with radiation therapy alone. In total, 16 lesions (8.6%) underwent prophylactic fixation and 26 lesions (14.0%) fractured after radiation therapy (Table 2 ). As secondary outcomes, the median time between radiation therapy and fracture was 209 days, and the median time between radiation therapy and death was 283 days.

Clinical and Demographic Factors Associated with Prophylactic Fixation vs Pathologic Fracture Risk

Overall, radiograph image interreviewer reliability had an intraclass correlation coefficient of 0.99 ( P value < 0.001). Prophylactic fixation was associated with higher Mirels score (OR = 1.98, P = 0.025), higher lesion cortical involvement (%) (OR = 16.96, P = 0.010), and lower age (OR = 0.93, P = 0.024) (Table 3 ). No statistical difference was identified in insurance type, histology risk, or location of lesion.

Pathologic fracture was associated with higher lesion cortical involvement (%) (OR = 10.16, P = 0.003). Pathologic fracture was also associated with location in the humerus (OR = 3.54, P = 0.079) and low-risk histology (OR = 9.01, P = 0.057), although these trends were not statistically significant. No statistical difference was identified in Mirels score, antiresorptive medication use, fall risk, radiation therapy course, or age.

The combined end point of any SRE (prophylactic surgery or pathologic fracture) after radiation therapy was associated with higher Mirels score (OR = 1.62, P = 0.015), higher lesion cortical involvement (%) (OR = 8.94, P = 0.002), location in the humerus (OR = 4.19, P = 0.042), and private insurance (OR = 5.69, P = 0.022). In addition, a negligible trend toward Medicare over Medicaid insurance was found (OR = 4.12, P = 0.062). No statistical difference was identified in BMI, histology risk, or location of lower extremity lesions.

Our results from the cox proportional hazard model found that Mirels score (OR = 1.38, P = 0.013) was the only predictor of time from radiation therapy treatment to any SRE (Table 4 ). The associated Kaplan-Meier curve also did not find a notable difference between time from radiation therapy to prophylactic surgery or pathologic fracture as separate end points ( P = 0.8; Figure 3 ).

Clinical and Demographic Factors Associated with Mortality

Our results from the cox proportional hazard model found that time from radiation therapy treatment to death was associated with ECOG score (OR = 1.28, P = 0.003) as a risk factor, whereas low-risk histology (OR = 0.51, P = 0.029), mixed lesion type (OR = 0.34, P = 0.006), and BMI (OR = 0.95, P = 0.001) were protective factors (Table 5 ).

Discussion

Prophylactic stabilization for impending fractures is less complicated and more controlled than postfracture fixation, with lower associated hospital costs, length of hospital stay, complication risk, and mortality. 14 - 18 Image-based risk-assessment models have been developed to provide guidance for prophylactic treatment; however, they generally do not account for patient- and treatment-specific factors that also influence pathologic fracture risk. 9 , 19 - 22 Multiple studies have emphasized the limitations of the Mirels score, as lesion size and radiographic appearance have limited specificity. In addition, patient pain can be subjective with high interobserver variance, especially in the context of multifocal metastatic disease and patient-specific pain regimens. 9 , 10 , 23 Moreover, Cumming et al. 24 found that less than 20% of oncologists used a fracture-prediction score and less than 25% of patients with at-risk lesions were subsequently referred to orthopaedics. Our study found that, as expected, Mirels score highly correlated with patients undergoing prophylactic fixation after radiation therapy; however, it did not predict pathologic fracture risk after radiation therapy. We found that although the cortical involvement on radiograph was the most predictive determinant of fracture risk, other factors are also notable and should therefore be considered. In particular, when relying on standard predictive models, prophylactic fixation may be underperformed in primary tumor histologies that are considered low risk based on anticipated response to radiation therapy.

Radiographic Factors

Our study found that cortical involvement on plain radiographs had the most notable correlation with postradiotherapy fracture risk, with high interreviewer reliability in radiograph measurements. This is consistent with Mirels criteria and other standard systems. Although improved accuracy has been reported with more recent CT-based algorithms, CTs are more costly and challenging to obtain compared with radiographs. 9 Furthermore, these models often require multiple rounds of imaging for risk assessment both before and after radiotherapy. Our findings support the predictive value of radiographs with emphasis on cortical involvement as an accurate and cost-effective means of evaluating the risk of pathologic fracture.

We also found that location in the humerus (vs other long bones) was associated with likelihood of SRE after radiation therapy. However, we did not find a notable association between humerus lesions and prophylactic surgery. This suggests that although Mirels criteria assigns a lower score to upper extremity lesions, humerus lesions with extensive cortical involvement should be considered for prophylactic surgery because patients often rely on their upper extremities for activities of daily living and mobility. Patients with metastatic disease may be required to bear weight with their arms because of impending fractures or recent surgery in their legs. They may also be deconditioned or have notable fall risk, increasing the likelihood of eventual fracture. Indeed, Mirels score has previously been criticized as being insufficient in characterizing impending upper extremity pathologic fractures. 23 , 25 Humerus fractures do carry notable and potentially underappreciated consequences; a 2016 retrospective study found that although upper extremity pathologic fractures secondary to metastatic cancer make up a lower proportion of pathologic fracture sites than lower extremity pathologic fractures (8.7% compared with 25%, respectively), the upper extremity lesions that do fracture have a higher same admission mortality than lower extremity pathologic fractures (8.4% compared with 6.3%, respectively). 26 Based on the findings in prior literature and this study, humeral lesions may warrant surgical intervention more commonly than that traditionally suggested.

In this study, neither lesion appearance (lytic, blastic, or mixed) nor infiltration (well-defined solid or diffuse) was found to be markedly associated with prophylactic fixation or fracture risk when controlling for other variables, such as cortical involvement. This is a departure from Mirels criteria and other historic classifications; however, it is concordant with the multicenter prospective study by Crenn et al, 23 which also did not find a difference in pathologic fracture presentation across osteoblastic or osteolytic metastatic long bone lesions. According to these data, the risk of SRE in mixed or blastic lesions should not be underestimated.

Disease Factors

Our study found a trend toward increased postradiotherapy fracture risk in metastasis from “low-risk” primary histologies, suggesting that surgery may be underperformed for these lesions. Fracture can occur because of disease persistence or progression following radiation therapy, particularly with primary histologies that are more aggressive or less responsive to treatment. Metastases from differing primary tumors are known to have varying sensitivities to local radiation therapy and systemic treatments. For example, bone metastasis arising from thyroid, hepatocellular, renal cell, and cholangiocarcinoma have been shown to be less responsive to radiation therapy and therefore are generally considered more likely to necessitate surgical intervention. 6 , 18 , 27 , 28 Our findings suggest that, with the focus on these “high-risk” histologies, risk may currently be underestimated, and surgery may therefore be underperformed, for “low-risk” histologies.

In this study, we did not find an association between fracture risk and palliative radiation therapy schedules. Histology-specific palliative radiation therapy schedules for bone metastasis have been highly debated in the literature. 29 Some reports suggest that bone recalcification seems to be equivalent after low-dose versus high-dose radiation therapy for multiple myeloma, but higher doses seem to lead to more recalcification for solid malignancies. 30 , 31 Shorter courses may be more favorable for patient quality of life given that only half of the patients receiving external beam radiation therapy (EBRT) for bone metastasis are estimated to survive for one year; the goal of care for patients with limited life expectancy, as approximated by ECOG score, is to administer high-dose radiation therapy in one fraction to minimize hospitalization. 29 Our findings support using factors such as convenience and quality of life to dictate radiation therapy schedules because local control efficacy was equivalent regardless of treatment course.

Patient Factors

Interestingly, our data suggest that patients with private insurance were markedly more likely to undergo an intervention after radiation therapy when compared with patients with Medicaid insurance. This has also been described in the previous literature, which shows that Medicaid and uninsured patients are more likely to have limited access to care, resulting in decreased rates of prophylactic fixation and increased disease burden at time of diagnosis. 32 , 33 Our findings add to the body of literature, confirming that greater efforts are needed to address health disparities, particularly among cancer patients. We did not find a notable association between fracture risk and other patient-related factors, such as patient age, BMI, fall risk, ECOG score, and concurrent antiresorptive medication.

Mortality

We did observe that higher BMI, low-risk histology, and mixed lesion appearance were found to be negatively correlated with overall patient mortality. As the focus of this study is on fracture risk, additional research would be needed to confirm the influence of these risk factors on patient mortality.

Limitations

Given the nature of our retrospective study design, our data were based on the available documentation collected during radiation therapy and orthopaedic oncology clinic visits. In particular, Mirels score was calculated retrospectively; to minimize bias in this process, the authors involved were blinded to patient outcomes. Second, this study was conducted at a single-center tertiary institution, so patient care was influenced by the available resources and practice patterns at our cancer center. As such, decisions about treatment may have been different elsewhere, and the findings may not be generalizable to all centers. Moreover, metastatic disease is inherently complex in terms of pathophysiology and management (eg, number of lesions, treatment cycles, and clinical responses), and it was not feasible to capture or control for this heterogeneity with our sample size. In particular, we were unable to make statistically significant comparisons between EBRT and stereotactic body radiation therapy (SBRT) given the relatively infrequent utilization of SBRT due to insurance restrictions. Although challenging to execute, a large-scale multi-institutional study with greater statistical power would be needed to more definitively confirm these findings.

Conclusions

Surgical intervention inherently has risks, and the benefits of prophylactic fixation should be considered in the context of remaining life expectancy. In addition, the risks and recovery associated with prophylactic surgery must be weighed relative to the morbidity of prolonged immobilization and/or pathologic fracture that can occur with nonsurgical management. As such, an accurate method of predicting patients who can be successfully managed with radiation therapy alone is critical to maximizing quality of life for patients with metastatic bone disease.

The current work adds meaningfully to the existing literature on this topic in that it studies a large number of radiation therapy treated lesions in both the upper and lower extremities, integrating the histologic diagnosis, new measures of fracture risk on plain radiographs with a multidisciplinary team of orthopaedic oncologists and musculoskeletal radiologists (lesion size, cortical involvement, and permeative/complete characterization), and patient-specific variables (comorbidities, fall risk, body mass index [BMI], Eastern Cooperative Oncology Group [ECOG] score) as a more reliable and accurate model of fracture risk.

In this study, extensive cortical involvement on plain radiograph was found to be the most accurate predictor of fracture risk. Importantly, although Mirels score predicted prophylactic surgical intervention, neither the score itself nor its individual components (pain, lesion appearance, location, and size in relation to bone diameter) were found to be notable predictors of fracture risk. In addition, prophylactic surgery may be underperformed for primary tumor histologies that are traditionally considered radioresponsive, such as multiple myeloma, prostate cancer, diffuse large B-cell lymphoma, and small cell lung cancer. These findings suggest that Mirels criteria should not be considered the sole or optimal determinant of fracture risk in metastatic long bone lesions. Simplifying and clarifying the indications for prophylactic fixation to the most notable predictors of fracture risk, such as lesion cortical involvement, could encourage more widespread, consistent adoption of clinical decision-making guidelines for orthopaedic oncologists and other members of the multidisciplinary treatment team. This shift would ultimately minimize surgical interventions, mobility restrictions, and pathologic fractures to optimize patient outcomes.

Contents of this manuscript have been previously presented at Musculoskeletal Tumor Society 2023 Annual Meeting. Contents of this manuscript have not been previously published elsewhere.

None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Ms. Pei, Dr. Karnuta, Dr. Jones, Dr. Griffith, Dr. Jia, Dr. Behzad, Dr. Weber, and Dr. Cipriano.

References

1. Li S Peng Y Weinhandl ED : Estimated number of prevalent cases of metastatic bone disease in the US adult population . Clin Epidemiol 2012 ; 4 : 87 - 93 . 22570568 10.2147/CLEP.S28339 PMC3345874

2. Macedo F Ladeira K Pinho F : Bone metastases: An overview . Oncol Rev 2017 ; 11 : 321 . 28584570 10.4081/oncol.2017.321 PMC5444408

3. Mavrogenis AF Pala E Romagnoli C Romantini M Calabro T Ruggieri P : Survival analysis of patients with femoral metastases . J Surg Oncol 2012 ; 105 : 135 - 141 . 21815154 10.1002/jso.22061

4. Ratasvuori M Wedin R Keller J : Insight opinion to surgically treated metastatic bone disease: Scandinavian Sarcoma Group Skeletal Metastasis Registry report of 1195 operated skeletal metastasis . Surg Oncol 2013 ; 22 : 132 - 138 . 23562148 10.1016/j.suronc.2013.02.008

5. Gillespie EF Mathis NJ Marine C : Prophylactic radiation therapy vs. standard-of-care for patients with high-risk, asymptomatic bone metastases: A multicenter, randomized phase II trial . Int J Radiat Oncol Biol Phys 2022 ; 114 : 1059 . 10.1200/JCO.23.00753 PMC10730067 37748124

6. Shimoyama T Katagiri H Harada H : Fracture after radiation therapy for femoral metastasis: Incidence, timing and clinical features . J Radiat Res 2017 ; 58 : 661 - 668 . 28992299 10.1093/jrr/rrx038 PMC5737329

7. Ramsey DC Lam PW Hayden J Doung YC Gundle KR : Mirels scores in patients undergoing prophylactic stabilization for femoral metastatic bone disease in the veterans administration healthcare system . J Am Acad Orthop Surg Glob Res Rev 2020 ; 4 : e20.00141 . 10.5435/JAAOSGlobal-D-20-00141 PMC7469986 32890013

8. Van der Linden YM Dijkstra PD Kroon HM : Comparative analysis of risk factors for pathological fracture with femoral metastases: Results based on a randomised trial of radiotherapy . J Bone Joint Surg Br Vol 2004 ; 86 : 566 - 573 . 15174555

9. Nguyen MV Carlier C Nich C Gouin F Crenn V : Fracture risk of long bone metastases: A review of current and new decision-making tools for prophylactic surgery . Cancers 2021 ; 13 : 3662 . 34359563 10.3390/cancers13153662 PMC8345078

10. Howard EL Shepherd KL Cribb G Cool P : The validity of the Mirels score for predicting impending pathological fractures of the lower limb . The Bone Joint J 2018 ; 100-B : 1100 - 1105 . 30062934 10.1302/0301-620X.100B8.BJJ-2018-0300.R1

11. Tatar Z Soubrier M Dillies AF Verrelle P Boisgard S Lapeyre M : Assessment of the risk factors for impending fractures following radiotherapy for long bone metastases using CT scan-based virtual simulation: A retrospective study . Radiat Oncol 2014 ; 9 : 227 - 310 . 25319635 10.1186/s13014-014-0227-1 PMC4205287

12. Damron TA Nazarian A Entezari V : CT-based structural rigidity analysis is more accurate than Mirels scoring for fracture prediction in metastatic femoral lesions . Clin Orthop Relat Res 2016 ; 474 : 643 - 651 . 26169800 10.1007/s11999-015-4453-0 PMC4746194

13. R Core Team : R: A Language and Environment for Statistical Computing . Vienna, Austria , R Foundation for Statistical Computing , 2014 , Available at: http://www.R-project.org/

14. Agarawal JP Swangsilpa T Van der Linden Y Rades D Jeremic B Hoskin PJ : The role of external beam radiotherapy in the management of bone metastases . Clin Oncol 2006 ; 18 : 747 - 760 . 10.1016/j.clon.2006.09.007 17168210

15. Bauer HC : Controversies in the surgical management of skeletal metastases . J Bone Joint Surg Br Vol 2005 ; 87 : 608 - 617 . 10.1302/0301-620X.87B5.16021 15855359

16. Blank AT Lerman DM Patel NM Rapp TB : Is prophylactic intervention more cost-effective than the treatment of pathologic fractures in metastatic bone disease? . Clin Orthop Relat Res 2016 ; 474 : 1563 - 1570 . 27020430 10.1007/s11999-016-4739-x PMC4887361

17. El Abiad JM Raad M Puvanesarajah V Rao SS Morris CD Levin AS : Prophylactic versus postfracture stabilization for metastatic lesions of the long bones: A comparison of 30-day postoperative outcomes . J Am Acad Orthop Surg 2019 ; 27 : e709 - e716 . 31344005 10.5435/JAAOS-D-18-00345

18. Arpornsuksant P Morris CD Forsberg JA Levin AS : What factors are associated with local metastatic lesion progression after intramedullary nail stabilization? . Clin Orthop Relat Res 2022 ; 480 : 932 - 945 . 34962492 10.1097/CORR.0000000000002104 PMC9007189

19. Hipp JA Springfield DS Hayes WC : Predicting pathologic fracture risk in the management of metastatic bone defects . Clin Orthop Relat Res 1995 ; 312 : 120 - 135 . 7634597

20. Mirels H : The classic: Metastatic disease in long bones A proposed scoring system for diagnosing impending pathologic fractures . Clin Orthop Relat Res 2003 ; 415 : S4 - S13 . 10.1097/01.blo.0000093045.56370.dd 14600587

21. Damron TA Mann KA : Fracture risk assessment and clinical decision making for patients with metastatic bone disease . J orthopaedic Res official Publ Orthopaedic Res Soc 2020 ; 38 : 1175 - 1190 . 10.1002/jor.24660 PMC7225068 32162711

22. Toci GR Bressner JA Morris CD Fayad L Levin AS : Can a novel scoring system improve on the Mirels score in predicting the fracture risk in patients with multiple myeloma? . Clin Orthop Relat Res 2021 ; 479 : 521 - 530 . 32420721 10.1097/CORR.0000000000001303 PMC7899603

23. Crenn V Carlier C Gouin F Sailhan F Bonnevialle P : High rate of fracture in long-bone metastasis: Proposal for an improved Mirels predictive score . Orthop Traumatol Surg Res 2020 ; 106 : 1005 - 1011 . 32782175 10.1016/j.otsr.2020.03.034

24. Cumming D Cumming J Vince A Benson R : Metastatic bone disease: The requirement for improvement in a multidisciplinary approach . Int Orthop 2009 ; 33 : 493 - 496 . 18431570 10.1007/s00264-008-0556-5 PMC2899083

25. Hoban KA Downie S Adamson DJ MacLean JG Cool P Jariwala AC : Mirels' score for upper limb metastatic lesions: Do we need a different cutoff for recommending prophylactic fixation? JSES Int 2022 ; 6 : 675 - 681 . 35813136 10.1016/j.jseint.2022.03.006 PMC9264023

26. Behnke NK Baker DK Xu S Niemeier TE Watson SL Ponce BA : Risk factors for same-admission mortality after pathologic fracture secondary to metastatic cancer . Support Care Cancer 2017 ; 25 : 513 - 521 . 27704262 10.1007/s00520-016-3431-8

27. Grünwald V Eberhardt B Bex A : An interdisciplinary consensus on the management of bone metastases from renal cell carcinoma . Nat Rev Urol 2018 ; 15 : 511 - 521 . 29904105 10.1038/s41585-018-0034-9 PMC7136176

28. Jung IH Yoon SM Kwak J : High-dose radiotherapy is associated with better local control of bone metastasis from hepatocellular carcinoma . Oncotarget 2017 ; 8 : 15182 - 15192 . 28146433 10.18632/oncotarget.14858 PMC5362477

29. Makita K Hamamoto Y Kanzaki H : Local control of bone metastases treated with external beam radiotherapy in recent years: A multicenter retrospective study . Radiat Oncol 2021 ; 16 : 225 - 228 . 34801042 10.1186/s13014-021-01940-0 PMC8605549

30. Rudzianskiene M Inciura A Gerbutavicius R : Single vs. multiple fraction regimens for palliative radiotherapy treatment of multiple myeloma: A prospective randomised study . Strahlentherapie und Onkologie 2017 ; 193 : 742 - 749 . 28573476 10.1007/s00066-017-1154-5 PMC5570777

31. Koswig S Budach V : Remineralization and pain relief in bone metastases after after different radiotherapy fractions (10 times 3 Gy vs. 1 time 8 Gy). A prospective study . Strahlentherapie und Onkologie 1999 ; 175 : 500 - 508 . 10554645 10.1007/s000660050061

32. Ward EM Fedewa SA Cokkinides V Virgo K : The association of insurance and stage at diagnosis among patients aged 55 to 74 years in the national cancer database . Cancer J 2010 ; 16 : 614 - 621 . 21131794 10.1097/PPO.0b013e3181ff2aec

33. Mosher ZA Patel H Ewing MA : Early clinical and economic outcomes of prophylactic and acute pathologic fracture treatment . J Oncol Pract 2019 ; 15 : e132 - e140 . 30523763 10.1200/JOP.18.00431