Introduction
Bone metastases are a common cause of pain in patients with metastatic cancer, negatively impacting their quality of life [1,2]. Palliative radiotherapy is one of the main treatment options for pain relief in bone metastases. Trials have shown that single-fraction radiotherapy (1×8 Gy) is an appropriate and effective alternative to longer regimens for pain relief in bone metastases [3]. These studies have consistently shown equal pain response rates and toxicity with single fraction versus multi-fraction regimens. As a result, single-fraction radiotherapy has become the internationally recommended treatment for bone metastases [4-6].
Several factors favor the use of hypofractionated regimens consisting of one or two fractions over longer regimens. Given the limited survival of many patients with bone metastases, hypofractionated regimens may benefit patients and their families by reducing the treatment burden, and the time and indirect costs spent on treatment [7], allowing them to spend their remaining time in more meaningful ways. In addition, shorter treatment regimens allow radiotherapy departments to use resources more efficiently.
Despite these benefits and the evidence supporting hypofractionated regimens, studies have shown wide variations in the uptake of hypofractionated regimens, with some countries adopting hypofractionation more routinely while others continue to rely predominantly on conventional fractionation schedules [8,9]. However, evidence on recent trends in the uptake of hypofractionated treatment regimens in daily clinical practice, particularly since the coronavirus disease 2019 (COVID-19) pandemic, remains limited. During the pandemic, radiotherapy practices shifted considerably globally, including increased use of telemedicine, reduced patient volumes, postponement of treatments, and a push toward shorter treatment schedules [10,11]. The use of hypofractionated regimens was specifically advocated to reduce hospital visits and optimize the use of limited resources [12], and the pandemic may therefore have further accelerated their adoption [12]. In the Netherlands, the use of one or two fractions of 8 Gy as a substitute for longer regimens (e.g., 5×4 Gy or 10×3 Gy) was advised by the Dutch Society for Radiotherapy and Oncology [13]. Therefore, this study aimed to evaluate trends in fractionation patterns including data from multiple radiotherapy departments in the Netherlands before, during, and after the COVID-19 pandemic. Moreover, we performed an exploratory analysis in a subset of the total cohort to assess pain response in patients treated with hypofractionated palliative regimens versus those receiving multi-fraction regimens in daily clinical practice.
Materials and Methods
1. Study population and data
Eleven out of the 22 academic and non-academic radiotherapy departments in the Netherlands participated in this study. Data on all treatment regimens for bone metastases between 2018 and 2022 were collected from the local registration systems, including age at time of treatment, sex, performance status, primary tumor localization, treatment site, and radiotherapy treatment details (total dose, number of fractions, technique, previous radiation to the same site). Stereotactic treatment regimens with high ablative total doses were excluded either through the variable indicating the radiotherapy technique when available or, when absent, by identifying stereotactic dose‑fractionation schedules based on expert opinion of a senior radiation oncologist (YML). Pain outcome data were collected for all patients from the total study cohort who were previously enrolled in the PRospective Evaluation of interventional StudiEs on boNe meTastases (PRESENT) cohort. All adult patients with bone metastases undergoing radiotherapy were eligible for inclusion in the PRESENT study and had to provide informed consent. Exclusion criteria were mentally incompetent patients and a life expectancy of <1 year. More detailed inclusion and exclusion criteria for enrollment in the PRESENT cohort were described previously [14]. The pseudonymized data were combined and collected in a central database at the Netherlands Comprehensive Cancer Organisation (IKNL). As the initiators of the study, IKNL and Leiden University Medical Centre (LUMC) obtained approval from the Scientific Committee of the Department of Radiotherapy at LUMC. The study was exempted from review by the Medical Research Ethics Committee of the LUMC and the Amsterdam University Medical Centers because the Medical Research Involving Human Subjects Act did not apply to the study.
2. Definitions
Treatment regimens were categorized as hypofractionated (1×8 or 2×8 Gy) or multi-fraction (≥5 fractions). A small proportion of the regimens (2%) consisted of <5 fractions but other than 1×8 Gy or 2×8 Gy and were excluded from the analyses. Localization of the primary tumor was categorized into lung, prostate, breast, other, or unknown. Since the treatment site was recorded as open text data, artificial intelligence (AI: GPT-4o-Mini) was used to categorize it into one of the following groups: cervical spine, thoracic spine, lumbar spine, ribs, sacrum (including os coccygis), pelvis, proximal femur/acetabulum, humerus, or other. After categorization by AI, the first 100 observations per radiotherapy department were manually checked. This showed a correct categorization in approximately 95% of the cases. In about a quarter of the regimens, the treatment site consisted of a combination of these categories, most commonly (1) a larger treatment area in the spine, (2) spine and/or pelvis and/or sacrum, (3) pelvis and/or sacrum and/or proximal femur/acetabulum, and (4) spine and ribs. These categories were added to the original nine categories. The distribution of the treatment sites of the included treatment regimens across these categories is shown in Supplementary Table S1. For the analyses, the treatment site was grouped into five broader categories, namely spine, pelvic region, extremities, ribs, or other. Data on re-irradiation were available at two centers and were categorized as yes or no based on whether a patient had received a previous course of radiotherapy to the same treatment site. Performance status was grouped into Eastern Cooperative Oncology Group (ECOG) 0, 1, ≥2, or unknown. If performance status was documented as Karnofsky performance status (KPS), it was converted to ECOG (KPS 90–100 to ECOG 0, KPS 70–80 to ECOG 1, KPS 10–60 to ECOG ≥2) [15]. Pain scores were self-reported using the Brief Pain Inventory at baseline (a maximum of plus or minus 7 days from date of first fraction), and 4 and 8 weeks and 3 months after treatment. The outcome measure was the worst pain a patient experienced in the previous 3 days, measured on the Numeric Rating Scale ranging from 0 (no pain) to 10 (worst pain imaginable).
3. Statistical analyses
Characteristics of the study population were analyzed descriptively. Trends in the use of hypofractionated (1×8 or 2×8 Gy) and multi-fraction (≥5 fractions) regimens were analyzed over time for the entire cohort and stratified by radiotherapy department to assess variation. Trends in fractionation patterns were additionally stratified into 1×8 Gy, 2×8 Gy, 5×4 Gy, 10×3 or 13×3 Gy, or other multi-fraction (≥5). To evaluate trends in hypofractionation use over time and account for demographic shifts (e.g., aging population), a logistic regression was performed including year of treatment, age, sex, and primary tumor site. Logistic regression analyses were also performed to assess factors independently associated with hypofractionation. First, univariable analyses were performed for the variables age, sex, radiotherapy department, primary tumor localization, treatment site, and re-irradiation. Variables significantly associated with hypofractionation in the univariable analyses (p < 0.1) were included in a multivariable model. A separate multivariable model was performed in the subset of patients with information on re-irradiation (n = 3,893). Because most variables had very few missing observations (<1%), complete‑case analyses were used for all regression models. Since performance status was missing frequently, and not at random due to systematic differences in registration between radiotherapy departments, it was excluded from the regression analyses.
An exploratory analysis of pain outcomes was performed in patients for whom pain scores were available through inclusion in the PRESENT cohort. Mean pain scores and pain response were calculated over the 3-month post-treatment period for patients with baseline pain scores ≥2. Pain response was based on the difference between baseline pain scores and pain scores at week 4, week 8, and 3 months, irrespective of changes in analgesic use, as no data on this were available. Therefore, calculations of pain response could not be fully performed according to the international consensus criteria [16]. Patients who achieved a complete (pain score 0) or partial (reduction ≥2 points) response were classified as responders; others were classified as non-responders. Mean pain scores were compared between hypofractionated and multi-fraction regimens using the Wilcoxon rank sum tests. Pain response was compared using chi-squared tests. Statistical analyses were performed using Stata version 17.0 software (Stata Corp., College Station, TX, USA). A two-tailed p-value of <0.05 was considered statistically significant.
Results
1. Cohort characteristics
A total of 17,336 patients were included in this study, receiving 31,677 treatment regimens. During the study period, 9,545 patients (55.1%) received one radiotherapy treatment, 4,249 patients (24.5%) two treatments, and 3,542 patients (20.4%) three or more treatments. At the time of first treatment, the median age of patients was 70 years (interquartile range, 61 to 76 years), and most patients were male (57.5%) (Table 1). The most frequent primary tumor was lung cancer (26.6%), followed by prostate cancer (21.3%) and breast cancer (17.2%), and 34.5% had another primary tumor. Across all treatment regimens, the most common treatment site was the spine (40.9%), followed by the pelvic region (26.1%). Across the radiotherapy departments, the number of patients treated for bone metastases during the study period ranged from 637 to 2,977. Moreover, lung, prostate, and breast cancer were the most common primary tumor localizations across all departments.
2. Trends in radiotherapy fractionation for bone metastases
Among all 31,677 radiotherapy treatment regimens, 81.4% were hypofractionated (n = 25,790), while 18.6% were multi-fraction regimens (n = 5,887). The use of hypofractionated regimens increased slightly from the onset of the COVID-19 pandemic in 2020 onwards (77.9%–79.0% before COVID-19 vs. 82.5%–83.9% since COVID-19, p < 0.001) (Fig. 1A). This trend remained significant when controlling for age, sex, and primary tumor localization (p < 0.001) (Supplementary Table S2). A single fraction of 8 Gy was the most commonly used hypofractionated regimen (72.0%–75.9% over the years) (Fig. 1B). The use of 2 fractions of 8 Gy increased slightly (4.9%–9.2% over the years). In total, 5 fractions of 4 Gy was the most commonly provided multi-fraction regimen, with a slight decrease over the years (16.1% to 11.7% over the years). Between radiotherapy departments, the use of hypofractionated regimens ranged from 33.6% to 98.9% (Fig. 2).
3. Factors associated with receipt of hypofractionated regimens
Age, radiotherapy department, localization of the primary tumor, and treatment site were independently associated with hypofractionation (Table 2). Older patients had higher odds of receiving hypofractionated treatment regimens compared to younger patients (>75 years; odds ratio [OR], 1.61; 95% confidence interval [CI], 1.41 to 1.83; p < 0.001). Patients with a primary tumor in the breast were less likely to receive hypofractionated regimens (OR, 0.66; 95% CI, 0.59 to 0.73; p < 0.001). Bone metastases in the spine and pelvic region had similar odds of being treated with hypofractionated regimens. The odds of hypofractionation were slightly lower when the treatment site was the extremities (OR, 0.88; 95% CI, 0.80 to 0.96; p = 0.009) and higher when the treatment site was the ribs (OR, 1.70; 95% CI, 1.47 to 1.96; p < 0.001). The likelihood of hypofractionation was also higher in case of re-irradiation of a previously treated site (Supplementary Table S3).
4. Pain
Pain scores at baseline were available in a subset of 278 patients. These patients had a median age of 69 years, were predominantly male (65.1%), and most commonly had a primary tumor in the prostate (32.0%), followed by the lung (26.2%) (Supplementary Table S4). Compared to the total study cohort, these patients were more often male, more likely to have prostate cancer, and more frequently treated with multi‑fraction schedules. The majority of these patients received hypofractionated regimens (n = 204, 73.3%). Mean pain scores did not differ significantly between hypofractionated and multi-fraction regimens over the 3-month post-treatment period. Scores declined from 6.5 to 4.0 in the hypofractionated group and from 6.3 to 3.4 in the multi-fraction group (Fig. 3). Pain response at week 4, week 8 and 3 months was similar between patients receiving hypofractionated and multi-fraction regimens (month 3, 55.6% vs. 63.0%; p = 0.406) (Fig. 4).
Discussion and Conclusion
This study evaluated fractionation patterns of palliative radiotherapy for bone metastases in a large cohort of patients in the Netherlands. Hypofractionated regimens were most commonly used, and a modest but significant increase in hypofractionated regimens was observed since the onset of the COVID-19 pandemic. In an exploratory analysis in a subgroup of patients, comparable mean pain scores and pain responses were observed between patients receiving hypofractionated and multi-fraction regimens in routine clinical practice over the 3-month post-treatment period.
During the study period, one fraction of 8 Gy was the most commonly used regimen for the palliation of bone metastases, with its use ranging from 72%–76% over the years. Treatment with two fractions of 8 Gy increased from 5% to 9% over the years, which may indicate a replacement of 5 fractions of 4 Gy with 2 fractions of 8 Gy, as the use of 5×4 Gy decreased slightly during the study period. These findings are in line with established clinical guidelines that support the use of hypofractionated regimens as they provide optimal treatment for many patients [4-6]. The use of hypofractionated regimens observed in this study is relatively high compared to international practice patterns, with international literature reviews showing use of single-fraction regimens mostly ranging between 5% and 50% [8,17]. This may be related to extensive research in the Netherlands into the effectiveness of single-fraction radiotherapy combined with a nationally applied evidence-based medicine approach that ensures that these findings are integrated into clinical practice. Additionally, differences in financing may play a role, as reimbursement models in some countries are more dependent on fractionation schedules, potentially incentivizing longer treatment courses [18-20]. The high adoption of hypofractionated regimens also suggests a patient-centered care approach, as hypofractionated regimens reduce the burden of treatment, which can ultimately improve patient wellbeing and quality of life [21,22]. In this study, older patients were more likely to receive hypofractionated regimens, supporting the fact that patient-centered care is being delivered to more frail patients by reducing the intensity of treatment. However, since results from the Dutch Bone Metastasis Study showed that response rates with single- or multi-fraction regimens are independent of age [23], hypofractionated regimens may be just as beneficial for younger patients by reducing the treatment and time burden while maintaining the same level of treatment efficacy if pain response is the treatment goal.
This study showed a slight but significant increase in hypofractionation in the treatment of bone metastases since the onset of the COVID-19 pandemic. Previous studies have shown changes in radiotherapy practices during the pandemic, including the increased use of telemedicine and a decrease in patient volumes, postponement of treatments, and shortages of staff and equipment [10,24]. During the pandemic, it was recommended to increase the use of hypofractionated regimens to reduce hospital visits and optimize the use of limited resources [12]. The slight increase in the use of hypofractionated regimens after 2020 may indicate that these recommendations were at least partially adopted by radiation oncologists and that this effect may have continued after the pandemic.
Despite the overall trend toward hypofractionated regimens, this study observed considerable variation between radiotherapy departments in the uptake of these regimens. This is consistent with the variability observed internationally in the use of hypofractionated regimens [8,9,25]. After adjusting for key patient and disease characteristics (age, sex, primary tumor localization, and localization of bone metastases), the differences between departments remained statistically significant, indicating that case mix alone is unlikely to fully account for the observed differences. Previous studies showed that several other factors may contribute to this variability, including beliefs in the clinical evidence surrounding hypofractionation, waiting times at the radiotherapy department, and physician or institutional preferences [25-27]. Individual institutional protocols of the departments may vary in how strongly hypofractionated regimens are advocated, and longstanding departmental practice patterns may shape treatment decisions. These factors likely contribute to the observation that even within a country with a limited number of radiotherapy departments, united within a well‑organized national radiotherapy society, the approach to the radiation treatment of bone metastases remains subject to individualization. Although some variation in clinical practice can be an indicator of effective patient-centered care, unwarranted variation in hypofractionation may impede the delivery of affordable and accessible radiotherapy. Therefore, the practice variation observed in this study may serve as valuable information for benchmarking and targeting implementation strategies.
Primary tumor localization and treatment site significantly influenced fractionation patterns. Patients with prostate or lung cancer were more likely to receive hypofractionated regimens than patients with breast cancer. The higher rate of hypofractionated regimens in patients with lung cancer may reflect the relatively poor prognosis of these patients. These patients may have the most to gain in terms of time investment versus life expectancy from hypofractionated regimens. Although both patients with metastatic prostate and breast cancer generally have a better prognosis, there is some evidence that bone metastases from primary breast tumors may have a higher likelihood of pathologic fractures [28,29]. Breast cancer bone metastases are typically osteolytic, which increases the likelihood of pathologic fractures, whereas prostate cancer bone metastases are more likely to be osteoblastic [30]. Since longer treatment regimens may be beneficial in the presence of pathologic fractures [31], this may partly explain the higher rate of multi-fraction regimens in breast cancer patients. Additionally, remineralization has been shown to be significantly higher in breast cancer patients following multi-fraction regimens, contributing to improved bone stabilization [32]. It was also observed that re-irradiation of previously treated metastases was more likely to involve a hypofractionated regimen. This may be related to the fact that physicians are more inclined to retreat after 1 or 2 fractions of 8 Gy [33], and then subsequent re-irradiation is likely to consist of 1 or 2 fractions.
In terms of pain outcomes, the exploratory analysis on pain outcomes in this study showed that mean pain scores and pain response over the 3-month post-treatment period did not differ significantly between hypofractionated and multi-fraction regimens. These results support the generalizability of the clinical trial results showing non-inferiority of hypofractionated regimens for pain relief to real-world settings. It is important to note, however, that selecting an appropriate fractionation schedule can be influenced by expected patient survival, although predicting survival in patients with metastatic disease remains challenging. Several studies have shown that single‑fraction regimens are associated with a greater likelihood of re‑irradiation compared with multi‑fraction regimens [3]. Clinicians may therefore prefer multi‑fraction regimens for patients anticipated to live longer to reduce the risk of pain recurrence. However, the rate of re-irradiation likely does not accurately reflect response duration, since it was shown that re-irradiation is carried out earlier and more often after a single 8 Gy treatment, irrespective of response [33]. It may therefore capture clinicians’ willingness and the feasibility to retreat after a single fraction, and the ongoing disbelief that 8 Gy is as effective as multiple fractions. As advances in systemic therapies extend life expectancy for patients with metastatic disease, the recent availability of stereotactic regimens may provide a chance of more durable pain control in individuals likely to live longer [34]. However, which patients will benefit from stereotactic radiotherapy is still uncertain. Nevertheless, after a single fraction of 8 Gy, re‑irradiation is effective [33,35], and when retreatment is required, a single‑fraction for retreatment still results in less cumulative treatment burden than longer multi‑fraction regimens [33].
The main strength of this study is the inclusion of a large number of treatment regimens for bone metastases from a variety of radiotherapy departments in the Netherlands, thus providing a representative reflection of treatment patterns in clinical practice. However, some limitations should be noted. First, the use of hypofractionated or multi-fraction regimens for bone metastases may depend on variables that were not available for this study. Most importantly, due to the limited structural registration of spinal cord compression and/or pathologic fractures in radiotherapy databases, we were unable to characterize bone metastases as complicated or uncomplicated, which may affect fractionation patterns, as complicated bone metastases may benefit from longer treatment schedules [31]. Other variables that may influence dose fractionation patterns that were not available in this study include factors such as lesion size, soft tissue involvement, travel distance, and concurrent systemic therapy. Second, data on analgesic use were not available for this study, and therefore pain response could only be assessed based on changes in pain scores. Changes in analgesic use could not be accounted for according to the international consensus criteria [16]. This may have led to an underestimation of the response rates in this study. However, in previous analyses of the Dutch Bone Metastasis Study, the effect of reduced analgesic use on response rates appeared to be small (in 3% of patients in both the single and multi-fraction group) [33].
This study shows a high overall adoption of hypofractionated radiotherapy regimens for the palliation of bone metastases in the Netherlands, with a slight increase since the COVID-19 pandemic. In the subgroup with available pain scores, pain outcomes between hypofractionated and multi-fraction regimens were comparable. This exploratory analysis suggests equal palliation with lower care burden in case of hypofractionation. However, variation between radiotherapy departments in the use of hypofractionated regimens exists. These observations support the use of hypofractionation in routine practice to promote convenient and accessible radiotherapy for all patients with bone metastases.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
