Burdens of SREs Associated With Bone Metastases
Burdens of SREs Associated With Bone Metastases
It has already been recognized that patients experiencing SREs have decreased survival, increased treatment costs and worse quality of life compared with those without SREs. However, far less research has been dedicated to delineating those outcomes with respect to individual SRE types. Nonetheless, the review presented herein illustrates a limited number of studies in which researchers sought to draw distinctions between SRE types rather than treating them in a composite manner.
The differing effect of SRE types on per-event medical care costs was the most widely reported outcome reviewed, and data have been reported for all major SRE types. While further analysis was limited by the heterogeneous nature of restricted data, one can surmise that SRE treatment costs are reportedly greater in the USA and that bone surgery and spinal cord compression are the most expensive. Although costly, it is important to note that these are the least incident SREs reported in clinical trials and, therefore, their effect on total treatment costs may be limited. The increased costs in the USA may also be, in part, due to the limited nature of our data. That is, the SRE cost data reported from a US perspective was disproportionally comprised of direct observations (57% of US data were direct observations). This is contrasted with non-US data, in which only 3% of the data were direct observations.
Across all SREs, the mean costs per SRE in direct observations versus estimations were US$24,303 and 7798, respectively. Therefore, the significant difference between US-reported and non-US reported SRE costs may be a reflection of the method used to assess costs (direct observations in US vs indirect estimation in non-US countries). The cost methodology being coincidented with the country of analysis may explain the lack of significant difference between direct observation and estimation values, despite the large difference between mean costs of the two groups. The comparison of US and non-US per-event SRE costs may have also been limited by uncertainty as to whether the US costs were truly costs rather than charges, claims or reimbursements. The lack of clarity with regard to the types of fees for US data should be considered a minor limitation for this end point.
Additionally, the impact of SREs in the real-world clinical setting may be further limited by the fact that they are often asymptomatic and/or require little more than oral analgesics. In fact, Hillner et al. found in an economic analysis of pamidronate for breast cancer patients with bone metastases that the cost of fractures was much lower (i.e., ~US$600) than reported in this review. A separate economic analysis costs of admission in Spanish patients with bone metastatic breast, lung and prostate cancers was conducted by Pockett et al.. In that analysis, SRE types were delineated for the purpose of calculating their respective individual rates within each tumor type; however, SRE costs were reported indirectly and in a composite manner. For example, the difference in hospital admission costs between patients with metastatic bone disease without an SRE and patients with metastatic bone disease and an SRE were €242 and €405 for breast cancer and prostate cancer patients, respectively. Interestingly, the mean hospital admission cost for lung cancer patients without metastatic bone disease was €696 greater than for patients with metastatic disease and an SRE. These results point to the great variability of the nature of SRE costs.
The effects of SRE types on survival were less defined than economic outcomes, and while some conclusions could be drawn from indirect comparisons of patients experiencing spinal cord compression or pathologic fracture (vertebral and nonvertebral), the most interesting and comprehensive analysis was reported by Woodward et al.. This was the only analysis to directly compare SRE types on the basis of overall patient survival, and found that spinal cord compression and hypercalcemia were most strongly associated with poor survival. The authors suggested that the effect of spinal cord compression may have been particularly negative given that it is often associated with emergency treatment and profound consequences such as paralysis. With regard to these findings, one might also consider the possibility that SRE-type severity is associated with underlying disease severity such that those experiencing the most severe SRE-types are also the most progressed, and are therefore closest to death. Due to the restricted availability of studies assessing the differential effects of SREs on survival, conclusions should be drawn carefully. Furthermore, whether SREs in composite or by type are useful to healthcare providers in a prognostic capacity in clinical practice has not been adequately assessed in the literature, although it is likely that the experience of an SRE is a meaningful indication of worsening bone-metastatic disease.
Data regarding the differential impacts of SRE types on patients' quality of life were quite sparse. Only two of the studies found in our review had elicited data directly via objective means (i.e., the EQ-5D and time-trade-off), while one study, an economic analysis, relied on investigator estimation. Whether author judgment represents a valid method of utility elicitation may be debated. In the absence of more data, it is a reasonable starting point for further exploration. However, from the limited information available, one will find that spinal cord compression is associated with the greatest negative impact on quality of life, which is conceivable given that spinal cord compression was also associated with the highest levels of clinical and economic burden in this review.
A question that arises is whether the evaluation of SRE-related costs and utilities should be based on a societal or a patient perspective? That is, should the utilities be elicited directly from patients who are directly affected by the disease or indirectly from members of society at large, who often carry some of the burden of the sick through taxes and insurance premiums? The choice may affect post hoc analyses such as cost–effectiveness, because patients tend to report a lower disease burden on quality of life than do healthy individuals from society. Likewise, costs might be considered narrowly to include only direct medical expenditures incurred by payers or more broadly to include direct and indirect medical or disease-related costs incurred by payers, patients and caregivers; possibly even employers who are burdened by work days missed.
This review illustrates that while it is possible to differentiate SREs by clinical, economic and humanistic outcomes, substantive data are lacking, particularly with regard to health state utilities. While this may not be a large issue for the medical community as a whole, the lack of quality of life and economic data for SREs limit post hoc analyses comparing SRE-limiting treatments such as zoledronic acid and denosumab. The dearth of data may also be illustrative of the general lack of understanding of the nature of the SREs themselves, as they vary greatly in terms of severity and duration of effect. Such qualities make the application of mean costs and utility decrements to SREs difficult.
The clinical, economic and humanistic impacts of SREs differ by SRE type and, therefore, analyses involving SRE assessment should consider each type individually. That is, in order to fully assess the impact of SREs on patient outcomes, one should not consider SREs in a composite manner. However, survival, cost per event and quality of life data with which to differentiate SREs are scarce, although cost per event to a lesser extent. Consequently, our understanding of SREs' effects is equally limited. This may limit the usefulness and generalizability of post hoc analyses assessing the benefits of SRE-limiting treatment regimens or programs. It is recommended that more prospective studies, powered to detect clinical, economic and humanistic differences between SRE types, be conducted.
Expert Commentary
It has already been recognized that patients experiencing SREs have decreased survival, increased treatment costs and worse quality of life compared with those without SREs. However, far less research has been dedicated to delineating those outcomes with respect to individual SRE types. Nonetheless, the review presented herein illustrates a limited number of studies in which researchers sought to draw distinctions between SRE types rather than treating them in a composite manner.
The differing effect of SRE types on per-event medical care costs was the most widely reported outcome reviewed, and data have been reported for all major SRE types. While further analysis was limited by the heterogeneous nature of restricted data, one can surmise that SRE treatment costs are reportedly greater in the USA and that bone surgery and spinal cord compression are the most expensive. Although costly, it is important to note that these are the least incident SREs reported in clinical trials and, therefore, their effect on total treatment costs may be limited. The increased costs in the USA may also be, in part, due to the limited nature of our data. That is, the SRE cost data reported from a US perspective was disproportionally comprised of direct observations (57% of US data were direct observations). This is contrasted with non-US data, in which only 3% of the data were direct observations.
Across all SREs, the mean costs per SRE in direct observations versus estimations were US$24,303 and 7798, respectively. Therefore, the significant difference between US-reported and non-US reported SRE costs may be a reflection of the method used to assess costs (direct observations in US vs indirect estimation in non-US countries). The cost methodology being coincidented with the country of analysis may explain the lack of significant difference between direct observation and estimation values, despite the large difference between mean costs of the two groups. The comparison of US and non-US per-event SRE costs may have also been limited by uncertainty as to whether the US costs were truly costs rather than charges, claims or reimbursements. The lack of clarity with regard to the types of fees for US data should be considered a minor limitation for this end point.
Additionally, the impact of SREs in the real-world clinical setting may be further limited by the fact that they are often asymptomatic and/or require little more than oral analgesics. In fact, Hillner et al. found in an economic analysis of pamidronate for breast cancer patients with bone metastases that the cost of fractures was much lower (i.e., ~US$600) than reported in this review. A separate economic analysis costs of admission in Spanish patients with bone metastatic breast, lung and prostate cancers was conducted by Pockett et al.. In that analysis, SRE types were delineated for the purpose of calculating their respective individual rates within each tumor type; however, SRE costs were reported indirectly and in a composite manner. For example, the difference in hospital admission costs between patients with metastatic bone disease without an SRE and patients with metastatic bone disease and an SRE were €242 and €405 for breast cancer and prostate cancer patients, respectively. Interestingly, the mean hospital admission cost for lung cancer patients without metastatic bone disease was €696 greater than for patients with metastatic disease and an SRE. These results point to the great variability of the nature of SRE costs.
The effects of SRE types on survival were less defined than economic outcomes, and while some conclusions could be drawn from indirect comparisons of patients experiencing spinal cord compression or pathologic fracture (vertebral and nonvertebral), the most interesting and comprehensive analysis was reported by Woodward et al.. This was the only analysis to directly compare SRE types on the basis of overall patient survival, and found that spinal cord compression and hypercalcemia were most strongly associated with poor survival. The authors suggested that the effect of spinal cord compression may have been particularly negative given that it is often associated with emergency treatment and profound consequences such as paralysis. With regard to these findings, one might also consider the possibility that SRE-type severity is associated with underlying disease severity such that those experiencing the most severe SRE-types are also the most progressed, and are therefore closest to death. Due to the restricted availability of studies assessing the differential effects of SREs on survival, conclusions should be drawn carefully. Furthermore, whether SREs in composite or by type are useful to healthcare providers in a prognostic capacity in clinical practice has not been adequately assessed in the literature, although it is likely that the experience of an SRE is a meaningful indication of worsening bone-metastatic disease.
Data regarding the differential impacts of SRE types on patients' quality of life were quite sparse. Only two of the studies found in our review had elicited data directly via objective means (i.e., the EQ-5D and time-trade-off), while one study, an economic analysis, relied on investigator estimation. Whether author judgment represents a valid method of utility elicitation may be debated. In the absence of more data, it is a reasonable starting point for further exploration. However, from the limited information available, one will find that spinal cord compression is associated with the greatest negative impact on quality of life, which is conceivable given that spinal cord compression was also associated with the highest levels of clinical and economic burden in this review.
A question that arises is whether the evaluation of SRE-related costs and utilities should be based on a societal or a patient perspective? That is, should the utilities be elicited directly from patients who are directly affected by the disease or indirectly from members of society at large, who often carry some of the burden of the sick through taxes and insurance premiums? The choice may affect post hoc analyses such as cost–effectiveness, because patients tend to report a lower disease burden on quality of life than do healthy individuals from society. Likewise, costs might be considered narrowly to include only direct medical expenditures incurred by payers or more broadly to include direct and indirect medical or disease-related costs incurred by payers, patients and caregivers; possibly even employers who are burdened by work days missed.
This review illustrates that while it is possible to differentiate SREs by clinical, economic and humanistic outcomes, substantive data are lacking, particularly with regard to health state utilities. While this may not be a large issue for the medical community as a whole, the lack of quality of life and economic data for SREs limit post hoc analyses comparing SRE-limiting treatments such as zoledronic acid and denosumab. The dearth of data may also be illustrative of the general lack of understanding of the nature of the SREs themselves, as they vary greatly in terms of severity and duration of effect. Such qualities make the application of mean costs and utility decrements to SREs difficult.
The clinical, economic and humanistic impacts of SREs differ by SRE type and, therefore, analyses involving SRE assessment should consider each type individually. That is, in order to fully assess the impact of SREs on patient outcomes, one should not consider SREs in a composite manner. However, survival, cost per event and quality of life data with which to differentiate SREs are scarce, although cost per event to a lesser extent. Consequently, our understanding of SREs' effects is equally limited. This may limit the usefulness and generalizability of post hoc analyses assessing the benefits of SRE-limiting treatment regimens or programs. It is recommended that more prospective studies, powered to detect clinical, economic and humanistic differences between SRE types, be conducted.
Source...