The burden of skeletal-related events in four Latin American countries: Argentina, Brazil, Colombia, and Mexico

Abstract

Aim

Skeletal-related events (SREs) are major bone complications that frequently occur in patients with solid tumors (ST) and bone metastases, and in patients with multiple myeloma (MM). SREs include pathological fracture, spinal cord compression, radiation to bone, and surgery to bone. Limited data are available regarding the burden of SREs in Latin America. We built an economic model to quantify the current and future economic burden of SREs among adults in four Latin American countries: Argentina, Brazil, Colombia, and Mexico.

Methods

A comprehensive literature review with a systematic search strategy was conducted to parameterize the economic burden of illness (BOI) model. Economic analyses were conducted using a prevalence-based model. Aggregate SRE costs obtained from country-specific sources were used. We also included patient productivity losses. Costs were expressed in 2020 USD for the total annual burden, annual burden per 1,000 at risk, and projected five-year burden.

Results

The estimated total number of SREs was 251,503 in 2020, amounting to a total annual cost of USD 1.4 billion. The total projected five-year cost was USD 6.9 billion. Annual costs were highest in Brazil (USD 779.1 million), followed by Mexico (USD 281.8 million), Argentina (USD 174.6 million), and Colombia (USD 120.1 million). The average financial burden per 1,000 at risk was greatest in Brazil (USD 3.6 million), followed by Mexico (USD 3.4 million), Colombia (USD 2.9 million), and Argentina (USD 2.7 million).

Conclusion

Despite recommendations by medical societies for the use of bone-targeted agents in patients with solid tumors and bone metastasis or with multiple myeloma and bone lesions, a large proportion of patients at risk of experiencing SREs are not treated. Early detection of bone metastases and SREs and the use of the most effective preventative treatments are needed to decrease the clinical and economic burden of SREs in Latin America.

Keywords:

• Skeletal-related events
• burden of illness
• Argentina
• Brazil
• Colombia
• Mexico

JEL CLASSIFICATION CODES:

• I10
• I1
• I
• I12

Introduction

Bone is one of the most common sites of metastasis among patients with solid tumor (ST) cancers. Up to 75% of patients with advanced prostate or breast cancer will develop bone metastases during the course of their disease^1–4. Given that patients with breast or prostate cancer have a relatively long survival time after diagnosis of bone metastases, the prevalence of bone metastases is high. For other types of ST cancers, such as lung, bladder, kidney, thyroid, and melanoma, the incidence of bone metastases with advanced disease ranges from 14 to 60%¹. Multiple myeloma (MM) is an aggressive cancer of the bone marrow plasma cells and is also characterized by lytic bone lesions. Approximately 90% of patients with MM will develop lytic bone lesions during the course of their disease⁵. Patients with ST and bone metastases or MM are highly susceptible to bone complications known as skeletal-related events (SREs). SREs include pathologic fractures, spinal cord compression, radiation to bone, and surgery to bone. Up to 64% of patients with ST and bone metastases, and up to 95% of patients with MM and lytic bone lesions will experience an SRE over the course of their disease^6,7. All SREs are associated with pain progression, impaired quality of life, and an increased risk of death^8–11. Based on the critical and potentially fatal effects of SREs, it is important to seek effective preventative options.

Bone-targeted agents (BTA) are recognized as the current standard of care for treating bone metastases and preventing SREs. Following the diagnosis of bone metastases, it is recommended to initiate BTA, such as zoledronic acid (a bisphosphonate) or denosumab (a human monoclonal antibody against RANKL)¹². A study investigating the use of intravenous bisphosphonates in patients with bone metastases secondary to breast, lung, or prostate cancer reported that discontinuation rates remain high in clinical practice¹³. A phase III clinical trial demonstrated that, despite receiving zoledronic acid, 36–39% of patients with bone metastases secondary to ST still experienced at least one SRE during the 21 months of treatment¹⁴. A recent meta-analysis that included three large phase 3 trials (one trial in prostate cancer¹⁵, one in breast cancer¹⁶, and one in other ST and MM¹⁷) and compared denosumab vs. zoledronic acid for the prevention of pathological fracture in patients with bone metastases from advanced cancer concluded that denosumab reduces the likelihood of pathological fractures compared to zoledronic acid (odds ratio [OR] 0.86, 95% confidence interval [CI], 0.74–0.99, p = .04)¹⁸. The main end-point for the randomized controlled trials (RCTs) included in this meta-analysis was the first on-study SRE^15–17,19. Denosumab was demonstrated to be superior to zoledronic acid in patients with prostate cancer (hazard ratio [HR], 0.82 [95% CI, 0.71–0.95; p = .008])¹⁵ and patients with breast cancer (HR, 0.82 [95% CI, 0.71–0.95; p = .01])¹⁶, and non-inferior to zoledronic acid in patients with other ST (i.e. not breast or prostate cancer) or MM (HR, 0.84 [95% CI, 0.71–0.98; p = .0007])¹⁷ in delaying time to first on-study SRE.

SREs represent an economic burden to society due to the associated increased health care resource utilization and costs. Patients with bone metastases and SREs incur significantly higher medical care costs than patients without SREs²⁰. Based on an observational European study, 21–48% of patients with SREs require an inpatient hospital stay with an average duration ranging from 8.4 to 41.1 days²¹. SREs are also associated with numerous diagnostic tests, procedures, and outpatient visits. Direct health care costs (i.e. inpatient stays, outpatient visits, home/long-term care, facility stays, daycare visits, emergency room visits, and procedures) for SREs in patients with ST and bone metastases in Europe range from €704 to €51,188^22,23, while the per-patient-per-year cost for an SRE in patients with MM can be up to $80,000 in United States dollars (USD)²⁴. Furthermore, a retrospective study based in the United States (US) showed that 67% of patients with SREs experienced disability or the inability to complete daily routine tasks, and 35% lost their jobs due to the physical inability to complete work functions^25,26. Loss of income due to patients’ inability to work can have detrimental financial impacts on patients and their families.

While the burden of SREs has been well-defined in the US^24,27–31 and several European countries^{21,22,32–34}, this burden is not well-understood in Latin America. Therefore, the objective of this study was to determine the economic burden of SREs in Latin American countries: Brazil, Mexico, Colombia, and Argentina.

Methods

Comprehensive literature review

We conducted a comprehensive literature review to collect relevant evidence for the four Latin American countries of interest: Argentina, Brazil, Colombia, and Mexico. Our population of interest was adults (≥18 years old) with ST with bone metastases and SREs, or with MM and SREs. We sought information on the prevalence and incidence of SREs in patients with ST with bone metastasis, and in patients with MM with lytic bone lesions, as well as the associated economic burden (e.g. health care resource utilization, direct health care costs, and indirect/societal costs). We did not limit study selection by study design, or by intervention or comparators. We focused on literature published in English, Spanish, and Portuguese from 1 January 2010 to 5 September 2019. An experienced medical information specialist developed and tested the search strategy, which was also peer-reviewed by another senior information specialist using the Peer Review of Electronic Search Strategies (PRESS) Checklist³⁵. Using the OVID platform, we searched Ovid MEDLINE, Embase, Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, Health Technology Assessment Database, NHS Economic Evaluation Database, and the Cochrane Central Register of Controlled Trials. We also undertook a search of four relevant Latin American databases: LILACS, Scielo, BVS, and Redalyc (see Supplementary Appendix S1). We performed grey literature searches using a modified version of the tool Grey Matters: a practical tool for searching health-related grey literature³⁶ (see Supplementary Appendix S2), as well as targeted searches for relevant data from other countries.

Key opinion leader interviews

Between November 2019 and March 2020, we conducted interviews with one oncologist and one hematologist from each country of interest to supplement the literature review.

Burden of illness model

A prevalence-based burden of illness (BOI) model was built for Argentina, Brazil, Colombia, and Mexico using data collected in our comprehensive literature review, targeted literature searches, and key opinion leader (KOL) interviews (i.e. interviews with oncologists and hematologists). Wherever possible, country-specific inputs were used to parameterize the BOI model. Costs are presented in 2020 US dollars (USD). Currencies were converted using exchange rates effective 24 March 2020^37,38. The population of interest was split into four patient groups: prostate cancer, breast cancer, other solid tumors, and MM.

Epidemiology of skeletal-related events

To calculate the annual number of pathological fractures, radiation to bone, spinal cord compressions, and surgeries to the bone for each patient group (prostate cancer, breast cancer, other solid tumors, and MM) in each country, we used estimates of the prevalence of patients with bone metastases from prostate cancer, breast cancer, and other solid tumors, and the prevalence of patients with MM and bone lesions, along with the annual rates of SREs for each group.

Prevalence of prostate cancer, breast cancer, and other solid tumors

To estimate the prevalence of patients with ST (prostate cancer, breast cancer, and other solid tumors) and bone metastasis, we followed the method described by Cristino et al.³⁹ using country-specific mortality rates reported by the Institute for Health Metrics and Evaluation⁴⁰. The method described by Cristino et al. assumes that annual cancer mortality is equal to the annual incidence of metastatic disease. For other solid tumors, the annual mortality rate was obtained by subtracting the annual mortality rate of breast cancer, prostate cancer, leukemia, Hodgkin lymphoma, non-Hodgkin lymphoma, and MM from the overall cancer mortality rate in each country. Based on data from Coleman¹, we assumed that 70% of patients who die of prostate or breast cancer, and 35% (median value reported for tumors other than prostate and breast by Coleman¹) of patients who die of other solid tumors have bone metastasis. These mortalities were assumed as the annual incidence of patients with ST and bone metastasis. Lastly, we estimated the prevalence of patients with ST and bone metastasis as the multiplication of the annual incidence by mean survival time (prevalence = incidence × survival), consistent with the method described by Cristino et al.³⁹. Values obtained from published literature for mean survival time were 2.24, 3.73, and 1.73 years for prostate cancer, breast cancer, and other solid tumors, respectively^41,42. See Table 1 for the calculation of the prevalence of ST and bone metastasis.

Table 1. Prevalence of solid tumors with bone metastasis.

	Argentina			Brazil			Colombia			Mexico
	Prostate cancer	Breast cancer	OSTs	Prostate cancer	Breast cancer	OSTs	Prostate cancer	Breast cancer	OSTs	Prostate cancer	Breast cancer	OSTs
Annual mortality^a (n)	4,938	6,778	58,859	19,712	18,818	188,625	3,743	3,385	35,112	7,556	7,776	73,223
Proportion with BM^1	70%	70%	35%	70%	70%	35%	70%	70%	35%	70%	70%	35%
Mortality with BM (n) (assumed equal to incidence of cancer with BM)	3,457	4,744	20,601	13,798	13,172	66,019	2,620	2,369	12,289	5,289	5,443	25,628
Survival (years)^41^,^42	2.24	3.73	1.73	2.24	3.73	1.73	2.24	3.73	1.73	2.24	3.73	1.73
Prevalence of cancer with BM (n) (incidence × survival)	7,743	17,697	35,639	30,908	49,133	114,212	5,869	8,838	21,260	11,848	20,302	44,337

Abbreviations. BM, bone metastasis; OST, other solid tumors.

^aEstimated mortality for 2020. Based on mortality rates⁴⁰ and population data from the World Bank Group (https://databank.worldbank.org).

Prevalence of multiple myeloma

No country-specific data were identified to inform the total prevalence of MM in the Latin American countries of interest. To estimate the total prevalence of MM in Argentina, Brazil, Colombia, and Mexico, we calculated the ratio of MM prevalence between the United States (US) and each of the Latin American countries by using the country-specific 5-year prevalence reported by the World Health Organization⁴³. Then, we used the calculated ratios (US prevalence to Latin American country prevalence) to estimate the MM prevalence in our countries of interest-based on the prevalence of MM in the US reported by the National Cancer Institute in 2016⁴⁴. Estimated Latin American prevalent cases were extrapolated to 2020 values based on population growth⁴⁵. Lastly, we assumed that 86.4% of patients with MM have lytic bone lesions as reported by Hungria et al.⁴⁶. See Table 2 for the calculation of the prevalence of MM with lytic bone lesions.

Table 2. Prevalence of multiple myeloma with lytic bone lesions.

	Argentina	Brazil	Colombia	Mexico
(i) Ratio of MM prevalence in US to MM prevalence in LATAM country^a	28.8	5.4	23.6	18.05
(ii) Total prevalent cases in US in 2016^44	131,392	131,392	131,392	131,392
Total prevalent cases in LATAM country in 2016 (ii/i)	4,562	24,332	5,567	7,279
Estimated prevalence for LATAM country in 2020^b	4,754	25,124	5,910	7,623
Proportion with lytic bone lesions^46	86.4%	86.4%	86.4%	86.4%
Estimated prevalence of MM with lytic bone lesions (2020)	4,108	21,707	5,106	6,587

Abbreviations. LATAM, Latin-America; MM, multiple myeloma; US, United States.

^a5-year prevalence in US/5-year prevalence in the respective LATAM country (i.e. Argentina, Brazil, Colombia, or Mexico)⁴³.

^bBased on population growth (https://databank.worldbank.org).

Treatment distribution

The estimated number of patients with ST and bone metastases and patients with MM and lytic bone lesions were stratified based on treatment for the prevention of SREs (i.e. treated with bisphosphonates, treated with denosumab, or untreated). The treatment distribution was informed by KOL feedback for patients with ST and patients with MM for Brazil, Colombia, and Mexico, and patients with MM for Argentina. Treatment distribution of Argentinian patients with ST was informed by local market research⁴⁷.

Annual rates of skeletal-related events

We calculated the annual rates of SREs for patients treated with bisphosphonates, patients treated with denosumab, and untreated patients. Rates for patients treated with bisphosphonates or with denosumab, as well as the proportion of patients with vertebral vs. non-vertebral fracture, were informed by unpublished data from four clinical trials of denosumab vs. zoledronic acid^15,16,19,48. SRE rates were calculated as the incidence of SREs divided by follow-up time (subject-years). Relative risk (RR) for incidence of SREs in patients treated with bisphosphonates vs. no treatment were obtained from three Cochrane reviews^49–51 and used to estimate annual SRE rates for untreated patients. Macherey et al.⁴⁹, O'Carrigan et al.⁵¹, and Mhaskar et al.⁵⁰ reported RRs for overall SRE incidence of 0.87 for patients with prostate cancer, 0.85 for patients with breast cancer, and 0.74 for patients with MM, respectively, in patients treated with bisphosphonates vs. no treatment.

Skeletal-related events management cost

Aggregate SRE management costs were obtained from country-specific sources. Costs for Colombia were obtained from the budget impact analysis (BIA) (Colombian healthcare system’s perspective) published by Pérez et al.⁵² and inflated to 2020 values using the Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI-Centre) Cost Converter⁵³. Costs for Brazil were informed by the BIA (private healthcare system’s perspective) published by Tanaka et al.⁵⁴. Details of sensitivity analysis for Brazil using alternative costs from the public healthcare system are provided in Supplementary Appendix S3. Costs for Mexico were informed by data from the Instituto Mexicano del Seguro Social (Mexican Social Security Institute)^55,56. Costs for Argentina were informed by KOL feedback on medical resource utilization and costs from the Programa de Atención Médica Integral (Comprehensive Medical Care Program. PAMI. https://www.pami.org.ar/). Data sources for Argentina and Brazil split costs into two groups: (i) costs for patients with prostate cancer, breast cancer, or MM, and (ii) cost for patients with other solid tumors. For Colombia and Mexico, the same aggregate costs for each SRE were reported across all types of cancer. See Table 3 for SRE management costs.

Table 3. SRE management costs for Argentina, Brazil, Colombia, and Mexico.

SRE	Prostate cancer, breast cancer, and MM		Other solid tumors
	Argentinian Peso (ARS)	US Dollars (USD)^b	Argentinian Peso (ARS)	US Dollars (USD)^b
Argentina^a
Vertebral fracture	228,576	3,468	222,183	3,371
Non-vertebral fracture	209,528	3,179	200,696	3,045
Radiation to bone	352,684	5,351	346,291	5,254
Surgery to bone	387,024	5,872	370,810	5,626
Spinal cord compression	339,437	5,150	319,202	4,843
SRE	Brazilian Real (BRL)	US Dollars (USD)^d	Brazilian Real (BRL)	US Dollars (USD)^d
Brazil^c
Vertebral fracture	28,035	5,173	27,246	5,027
Non-vertebral fracture	18,811	3,471	18,022	3,325
Radiation to bone	43,538	8,033	42,749	7,887
Surgery to bone	18,811	3,471	18,022	3,325
Spinal cord compression	13,260	2,446	12,471	2,301
	All cancers
SRE	Colombian Peso (COP)		US Dollars (USD)^f
Colombia^e
Vertebral fracture	29,135,063		6,739
Non-vertebral fracture	22,738,106		5,259
Radiation to bone	10,991,298		2,542
Surgery to bone	29,135,063		6,739
Spinal cord compression	29,243,311		6,764
SRE	Mexican Peso (MXN)		US Dollars (USD)^h
Mexico^g
Vertebral fracture	78,813		2,971
Non-vertebral fracture	77,462		2,920
Radiation to bone	147,378		5,555
Surgery to bone	256,543		9,670
Spinal cord compression	484,638		18,268

Abbreviation. SRE, skeletal-related event.

^aCosts were informed by KOL feedback on medical resource utilization and costs from the Programa de Atención Médica Integral (Comprehensive Medical Care Program. PAMI. https://www.pami.org.ar/).

^bUSD-ARS exchange rate = 1:65.91³⁸.

^cCosts source: Tanaka et al.⁵⁴.

^dUSD-BRL exchange rate = 1:5.42³⁷.

^eCosts source: Pérez et al.⁵².

^fUSD-COP exchange rate = 1:4,323.39³⁷.

^gCosts were informed by data from the Instituto Mexicano del Seguro Social (Mexican Social Security Institute)^57,58.

^hUSD-MXN exchange rate = 1:26.53³⁷.

Productivity loss associated with skeletal-related events

Productivity loss associated with SREs was mainly informed by KOL feedback due to a lack of published literature on this topic. See Supplementary Appendix S4 for a summary of the methods and inputs that were used to estimate productivity loss.

Economic burden of SREs

To estimate the economic burden of SREs, we multiplied the annual rates of SREs by the estimates of patients at risk (patients with ST and bone metastasis, and patients with MM and lytic bone lesions) to obtain the annual number of SREs for each country. Next, we multiplied the country-specific SRE management costs by the estimated annual number of SREs. Lastly, we added the costs associated with productivity loss to the total economic burden for each country.

Results

We screened 2,486 records identified through database searching (MEDLINE, Embase, Cochrane, LILACS, Scielo, BVS, Redalyc). Including 20 records identified through additional grey literature and targeted searches, 26 sources of data were included for informing the BOI model.

The BOI model predicted 251,503 SREs would occur in 2020 in the four Latin American countries analyzed, based on prevalence of disease (Table 1, 2) and rates of SREs (unpublished data)^15,16,19,48. Expected cases were highest for Brazil (total number: 133,653; annual incidence: 0.63 per 1,000), followed by Mexico (total number: 51,972; annual incidence: 0.40 per 1,000), Argentina (total number: 39,653; annual incidence: 0.87 per 1,000), and Colombia (total number: 26,254; annual incidence: 0.51 per 1,000).

Table 4 depicts the annual (2020) BOI of SREs in the analyzed countries (see Table 3 for SRE management costs). In Brazil, the largest expenditure was radiation to bone (60.7% of burden), followed by vertebral fracture (18.1% of burden), and non-vertebral fracture (14% of burden). In Mexico, the largest expenditure was radiation to bone (46.2% of burden), followed by spinal cord compression (18.7% of burden), and non-vertebral fracture (12.9% of burden). In Argentina. the largest expenditure was radiation to bone (54.4% of burden), followed by non-vertebral fracture (17% of burden), and vertebral fracture (15.5% of burden). In Colombia, the largest expenditure was vertebral fracture (31% of burden), followed by non-vertebral fracture (27.9% of burden), and radiation to bone (24.5% of burden).

Table 4. One-year burden of illness of SRE in Argentina, Brazil, Colombia, and Mexico.

Cost category	1-year burden (Argentina)	1-year burden (Brazil)	1-year burden (Colombia)	1-year burden (Mexico)	1-year burden (total LATAM)
Vertebral fracture	$27,061,296	$140,855,041	$37,223,768	$31,318,732	$236,458,837
Non-vertebral fracture	$29,705,304	$109,433,597	$33,536,449	$36,363,098	$209,038,449
Radiation to bone	$94,995,836	$472,545,159	$29,389,300	$130,234,653	$727,164,949
Spinal cord compression	$10,608,735	$17,408,107	$9,749,707	$52,752,360	$90,518,909
Surgery to bone	$11,523,001	$23,140,439	$9,112,997	$25,565,502	$69,341,939
Productivity costs	$710,355	$15,702,107	$1,099,437	$5,571,418	$23,083,317
Total	$174,604,527	$779,084,451	$120,111,658	$281,805,763	$1,355,606,399

Abbreviations. LATAM, Latin America; SRE, skeletal-related event.

Table 5 shows the annual BOI of SREs stratified by type of cancer (prostate cancer, breast cancer, other solid tumors, and MM). Other ST constituted ∼62% of the total BOI for the four Latin American countries analyzed, prostate cancer (16%) and breast cancer (17%) together constituted approximately a third of the burden, and MM constituted 6% of the burden. These proportions were relatively similar across the four countries for the BOI of SREs associated with other ST (range from 55% [Colombia] to 64% [Mexico]), prostate cancer (range from 13% [Argentina] to 16% [Brazil and Mexico]), and breast cancer (range from 15% [Mexico] to 19% [Argentina]). A greater variation was observed for the BOI associated with MM: 4, 6, 12, and 5% of the total BOI for Argentina, Brazil, Colombia, and Mexico, respectively. The greater relative BOI of SREs associated with MM in Colombia (12 vs. 4–6% in other countries) is mainly explained by two factors: (i) higher rates of fractures and lower rates of radiation to bone among patients with MM compared to ST, and (ii) a higher ratio fracture cost to radiation cost in Colombia than in Argentina, Brazil, and Mexico (Table 3).

Table 5. One-year burden of illness of SRE in Argentina, Brazil, Colombia, and Mexico stratified by type of cancer.

Cost category	Argentina $ (%)	Brazil $ (%)	Colombia) $ (%)	Mexico $ (%)	Total LATAM $ (%)
Vertebral fracture	27,061,296 (100)	140,855,041 (100)	37,223,768 (100)	31,318,732 (100)	236,458,837 (100)
Prostate cancer	3,487,435 (13)	22,237,947 (16)	5,865,856 (16)	4,909,724 (16)	36,500,962 (15)
Breast cancer	7,783,294 (29)	32,709,078 (23)	8,005,311 (22)	7,852,960 (25)	56,350,643 (24)
Other ST	12,899,043 (48)	62,645,710 (44)	16,203,989 (44)	14,490,785 (46)	106,239,526 (45)
MM	2,891,524 (11)	23,262,307 (17)	7,148,612 (19)	4,065,263 (13)	37,367,707 (16)
Non-vertebral fracture	29,705,304 (100)	109,433,597 (100)	33,536,449 (100)	36,363,098 (100)	209,038,449 (100)
Prostate cancer	2,615,577 (9)	12,208,318 (11)	3,745,584 (11)	3,948,187 (11)	22,517,666 (11)
Breast cancer	9,852,661 (33)	30,308,073 (28)	8,627,704 (26)	10,658,668 (29)	59,447,106 (28)
Other ST	14,829,303 (50)	52,738,396 (48)	16,095,172 (48)	18,126,673 (50)	101,789,543 (49)
MM	2,407,764 (8)	14,178,810 (13)	5,067,990 (15)	3,629,570 (10)	25,284,133 (12)
Radiation to bone	94,995,836 (100)	472,545,159 (100)	29,389,300 (100)	130,234,653 (100)	727,164,949 (100)
Prostate cancer	14,092,302 (15)	84,304,254 (18)	5,265,210 (18)	22,687,237 (17)	126,349,003 (17)
Breast cancer	13,853,050 (15)	57,972,389 (12)	3,484,748 (12)	16,887,873 (13)	92,198,059 (13)
Other ST	66,260,108 (70)	323,530,105 (68)	20,148,090 (69)	89,274,784 (69)	499,213,087 (69)
MM	790,376 (1)	6,738,411 (1)	491,253 (2)	1,384,759 (1)	9,404,800 (1)
Spinal cord compression	10,608,735 (100)	17,408,107 (100)	9,749,707 (100)	52,752,360 (100)	90,518,909 (100)
Prostate cancer	2,046,114 (19)	4,141,139 (24)	2,337,689 (24)	11,939,081 (23)	20,464,024
Breast cancer	866,085 (8)	1,198,435 (7)	601,497 (6)	3,653,352 (7)	6,319,369
Other ST	7,627,651 (72)	11,885,569 (68)	6,693,165 (69)	36,751,065 (70)	62,957,450
MM	68,884 (1)	182,964 (1)	117,355 (1)	408,863 (1)	778,066
Surgery to bone	11,523,001 (100)	23,140,439 (100)	9,112,997 (100)	25,565,502 (100)	69,341,939 (100)
Prostate cancer	426,263 (4)	1,132,448 (5)	468,081 (5)	1,219,427 (5)	3,246,219 (5)
Breast cancer	1,562,910 (14)	2,688,726 (12)	948,479 (10)	3,060,075 (12)	8,260,190 (12)
Other ST	8,826,955 (77)	16,983,244 (73)	6,644,151 (73)	19,338,117 (76)	51,792,468 (75)
MM	706,873 (6)	2,336,021 (10)	1,052,285 (12)	1,947,882 (8)	6,043,061 (9)
Productivity costs	710,355 (100)	15,702,107 (100)	1,099,437 (100)	5,571,418 (100)	23,083,317 (100)
Prostate cancer	56,936 (8)	2,178,026 (14)	34,468 (3)	591,565 (11)	2,860,995 (12)
Breast cancer	55,969 (8)	2,512,407 (16)	40,178 (4)	574,099 (10)	3,182,653 (14)
Other ST	272,648 (38)	9,417,410 (60)	133,091 (12)	2,615,608 (47)	12,438,757 (54)
MM	324,801 (46)	1,594,264 (10)	891,700 (81)	1,790,147 (32)	$4,600,912 (20)
All costs	174,604,527 (100)	779,084,451 (100)	120,111,658 (100)	281,805,763 (100)	1,355,606,399 (100)
Prostate cancer	22,724,627 (13)	126,202,132 (16)	17,716,888 (15)	45,295,221 (16)	211,938,869 (16)
Breast cancer	33,973,969 (19)	127,389,108 (16)	21,707,917 (18)	42,687,027 (15)	225,758,020 (17)
Other ST	110,715,708 (63)	477,200,434 (61)	65,917,658 (55)	180,597,032 (64)	834,430,831 (62)
MM	7,190,222 (4)	48,292,777 (6)	14,769,195 (12)	13,226,484 (5)	83,478,679 (6)

Abbreviations. LATAM, Latin America; MM, multiple myeloma; ST, solid tumors; SRE, skeletal-related event.

As the countries included in this study have different population sizes, comparing the burden per 1,000 at risk provided more meaningful insight into the differences in per-patient expenditures in these countries. The average burden per 1,000 at risk in Latin America was USD 3,150,673. The country with the greatest burden was Brazil (USD 3,607,528), followed by Mexico (USD 3,392,247), Colombia (USD 2,924,375), and Argentina (USD 2,678,544) (Figure 1).

Figure 1. Burden of illness of SREs per 1,000 at risk in Argentina, Brazil, Colombia, and Mexico.

When the four Latin American countries were considered together, the one-year (2020) BOI was USD 1.4 billion, and the projected five-year BOI was USD 6.9 billion. The country that contributed most to the burden was Brazil, followed by Mexico, Argentina, and Colombia. As expected, Brazil and Mexico contributed the most as they have the largest populations (212,871,096 and 129,164,010, respectively [2020]) and the greatest standardized costs (per 1,000 at risk) for SRE management. Although Argentina has a smaller population compared to Colombia (45,426,785 and 51,131,075, respectively [2020]) and the standardized cost for SRE management in Argentina is lower than in Colombia, its contribution to the total burden is greater than Colombia’s contribution due to a greater prevalence of patients with solid tumors and bone metastasis (Table 1).

Discussion

This BOI model demonstrates that SREs are associated with a substantial economic burden in Argentina, Brazil, Colombia, and Mexico. The estimated total number of SREs is 251,503 in 2020, resulting in a total annual cost of approximately USD 1.4 billion, and a projected economic burden of USD 6.9 billion over the next five years. Both pathological fracture (vertebral and non-vertebral) and radiation to bone each accounted for ∼45% of the total number of SREs (and together accounted for almost 90% of total SREs).

The relative costs for the different types of SREs varied across countries. For instance, from the five types of SREs that we considered (vertebral fracture, non-vertebral fracture, radiation to bone, surgery to bone, and spinal cord compression), radiation to bone was the least expensive SRE in Colombia, the second and third most expensive SRE in Argentina and Mexico, respectively, and the most expensive SRE in Brazil. Likewise, a vertebral fracture was the second most expensive SRE for both Brazil and Colombia, and the fourth most expensive for both Argentina and Mexico (see Table 3 for SRE management costs).

In terms of the contribution of different types of SREs to the total economic burden, for Argentina, Brazil, and Mexico, the highest proportion of the economic burden was from radiation to bone (54, 61, and 46%, respectively), followed by non-vertebral fracture in Argentina (17%), vertebral fracture in Brazil (18%), and spinal cord compression in Mexico (19%). For Colombia, the highest proportion of the economic burden was from vertebral fracture (31%), followed by non-vertebral fracture (28%).

Variation in SRE-associated costs across countries has been reported in Europe. A cost of illness analysis in Austria, the Czech Republic, Finland, Greece, Portugal, and Sweden noted variations in the mean SRE-associated costs across countries (even with adjustment for purchasing power parity). The authors stated that cost differences were likely driven by differences in both unit costs and treatment practices across countries³⁴. The use of less invasive surgical procedures in countries, such as Finland and Portugal has been noted as a potential variation in treatment practices that may impact the healthcare resource utilization costs associated with SREs⁵⁹.

Despite recommendations for the initiation of a bisphosphonate or denosumab at the time of diagnosis of bone metastasis to prevent or delay the first or subsequent bone complications^60,61, unmet needs remain about the proportion of patients with the malignant bone disease treated with BTA. A European cost of illness study reported that 60% of patients with bone metastases or lesions secondary to breast cancer, prostate cancer, lung cancer, or multiple myeloma did not receive BTA. Of the patients who experienced an SRE, 37% remained untreated⁵⁹. A large chart-review study conducted in France, Germany, Italy, Spain, and the United Kingdom found that 44% of patients with lung cancer and bone metastases were not receiving BTA, and 23% were predicted never to receive them⁶². Based on KOL feedback, we estimated that most patients with ST and bone metastasis in Brazil, Colombia, and Mexico do not receive treatment for the prevention of SREs (55, 70, and 54%, respectively). Likewise, local market research in Argentina determined that 35% of these patients remain untreated⁴⁷. Early detection of bone metastases and SREs and the use of the most effective preventative treatments are needed to decrease the economic burden of SREs and prevent the worsening of patient quality of life⁶⁰.

A limitation of this study is that our estimates of SRE burden used data from clinical trials that implemented skeletal surveys for fracture assessment^15,16,19,48. Carter et al. stated that the use of frequent clinical trial skeletal surveillance may result in overestimation of SREs compared to a real-world treatment scenario, mainly due to the identification of asymptomatic vertebral fractures that would not ordinarily be noticed in regular clinical settings⁶³. Liauw et al. conducted a retrospective chart review study to ascertain if the proportion of patients with metastatic breast cancer who had an SRE in a sample of patients in routine clinical practice was comparable to that seen in phase III RCTs of bisphosphonate therapy. The authors found that the proportion of patients reported to experience an SRE in routine clinical practice was 30.2% lower compared with the incidence of SREs observed in the treatment arm of RCTs of intravenous bisphosphonates (30 vs. 43%)⁶⁴. Since clinical trial data were used to inform the burden on SREs in this study, there is the risk of potential overestimation of the BOI. For our study, the lack of data on SRE rates in Latin America hampered the estimates of country-specific SRE burden.

Several publications based on real-world data have confirmed the substantial burden associated with SREs in Europe^34,59,65,66. SREs are associated with significant healthcare resource utilization related to an increase in the frequency and length of hospitalizations, as well as in the number of medical procedures and outpatient, emergency room, and day-care hospital visits^34,59,65,66. Developing countries (including Latin American countries) may face limitations in health care access⁶⁷, which may result in a relatively lower rate of medical procedures (including those classified as SREs, such as radiation to bone or surgery to bone, and those used to treat SREs) compared to developed countries. However, limitations in health care access for patients with SREs would likely be associated with other sources of burden that were not captured in this study, including pain and loss of mobility and independence, and associated costs. Future studies should seek to determine the potential impact of care patterns on the burden of SREs in Latin America.

To our knowledge, this is the first study aimed at quantifying the economic burden of SREs in Latin America. We conducted a comprehensive literature review to find the most relevant data to our BOI model. The model took into account the expected differences in rates of SREs for patients who are treated vs. not treated for prevention of SREs. As part of the model inputs, we estimated the proportions of patients who are treated with bisphosphonates, treated with denosumab, or untreated, and the rates of SREs specific for each group. Lower rates of SREs were applied for the treated populations compared to the untreated population. Despite its strengths, there were some other limitations associated with our analysis. Latin American country-specific data were limited, both in terms of availability and quality. For some model inputs (i.e. treatment distribution and productivity losses), we relied on KOL feedback only. Treatment distribution for Mexico was assumed as the average of the other countries, given that country-specific data were unavailable. Finally, there was variability in the methodology used to cost the management of SREs between the limited sources available.

Conclusion

In the Latin American countries included in our analysis, the one-year burden of SREs in 2020 is approximately USD 1.4 billion, and the projected five-year burden is USD 6.9 billion. Despite recommendations for the use of BTA in patients with ST and bone metastasis and with MM and lytic bone lesions, a large proportion of patients who are at risk of experiencing SREs are not treated. Early detection of bone metastases and SREs and the use of the most effective preventative treatments are needed to decrease the clinical and economic burden of SREs in Latin America.

Transparency

Declaration of funding

This study was funded by Amgen Inc.

Declaration of financial/other relationships

CRG-EVERSANA received financial support from Amgen for the conduct of this study. JGP, RA, VDS, JA, JC, MA, JF, ST, and GS are employees of Amgen.

AZ and DG are employees of CRG-EVERSANA. CRG-EVERSANA consults for various pharmaceutical, medical device, and biotech companies.

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

All authors were involved in the conception and design of the study. AZ and DG were involved in the analysis of data and all authors contributed to the interpretation of data. All authors participated in the drafting of the paper and critical revisions for intellectual content. All authors approved the final version. All authors agree to be accountable for all aspects of the work.

Acknowledgements

We wish to thank the clinicians in Latin America who contributed invaluable information during the interviews. From Argentina, Dr. Valeria Caceres and Dr. Claudia Shanley. From Brazil, Dr. Bruno Dias and Dr. Solange Sanches. From Colombia, Dr. Francisco Linares and Dr. Bonell Patiño. From Mexico, Dr. Martha Alvarado and Dr. Paula Cabrera. We are grateful for their insightful contributions.