This study used the linked SEER-Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the Applied Research Program, NCI; the Office of Research, Development and Information, CMS; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database.
The collection of the California cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract N01-PC-35136 awarded to the Northern California Cancer Center, contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's National Program of Cancer Registries, under agreement #U55/CCR921930-02 awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and endorsement by the State of California, Department of Public Health the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors is not intended nor should be inferred.
Androgen-deprivation therapy (ADT) causes bone loss and fractures. Guidelines recommend bone density testing before and during ADT to characterize fracture risk. The authors of the current report assessed bone density testing among men who received ADT for ≥ 1 year.
Surveillance, Epidemiology, and End Results/Medicare data were used to identify 28,960 men aged > 65 years with local/regional prostate cancer diagnosed from 2001 to 2007 who were followed through 2009 and who received ≥ 1 year of continuous ADT. Bone density testing was documented in the 18-month period beginning 6 months before ADT initiation. Logistic regression was used to identify the factors associated with bone density testing.
Among men who received ≥ 1 year of ADT, 10.2% had a bone density assessment from 6 months before starting ADT through 1 year after. Bone density testing increased over time (14.5% of men who initiated ADT in 2007-2008 vs 6% of men who initiated ADT in 2001-2002; odds ratio for 2007-2008 vs 2001-2002, 2.29; 95% confidence interval, 1.83-2.85). Less bone density testing was observed among men aged ≥ 85 years versus men ages 66 to 69 years (odds ratio, 0.76; 95% confidence interval, 0.65-0.89), among black men versus white men (odds ratio, 0.72; 95% confidence interval, 0.61-0.86), and among men in areas with lower educational attainment (P < .001). Men who visited a medical oncologist and/or a primary care provider in addition to a urologist had higher odds of testing than men who only consulted a urologist (P < .001).
Few men who received ADT for prostate cancer underwent bone density testing, particularly older men, black men, and those living in areas with low educational attainment. Visiting a medical oncologist was associated with increased odds of testing. Interventions are needed to increase bone density testing among men who receive long-term ADT.
Prostate cancer remains the most commonly diagnosed noncutaneous cancer in men in the United States.1 Although it is also the second leading cause of cancer death in men, most patients with prostate cancer become long-term survivors of the disease. Because of this, awareness of late complications of therapy is critical in treating men with prostate cancer.
Androgen-deprivation therapy (ADT) is the most frequently used systemic therapy for men with prostate cancer. Up to 50% of men with prostate cancer receive ADT during the course of their disease, and an estimated 600,000 American men with prostate cancer are receiving treatment with ADT at any given time.2,3 ADT improves overall survival when given as adjuvant therapy for men with high-risk tumors and improves quality of life among men with metastatic prostate cancer.4-6 However, its use is also high in other settings, for which benefits of treatment are less clear, although some evidence suggests that the use of ADT in these settings is declining.7
Although it is not as obviously toxic as chemotherapy, ADT is not without drawbacks.8 ADT causes a decline in bone mineral density and increases the risk of treatment-related fragility fractures.9-14 Since 2008, guidelines of the National Comprehensive Cancer Network have recommended routine bone density testing before and during treatment with ADT to characterize a man's risk of fracture.15 In addition, the American College of Physicians 2008 guidelines for osteoporosis screening among men recommend bone density testing for all men who are at increased risk of developing osteoporosis, including men who receive treatment with ADT.16
Relatively few data are available describing rates of bone density testing among men who receive ADT. Several single-institution studies of Veteran's Health Administration practices have reported low rates of bone density testing among men receiving ADT.17-19 To our knowledge, the frequency of bone density testing among men with prostate cancer who were treated in nonmilitary institutions in the United States has not been reported.
In the current analysis, we assessed bone density testing in a large, population-based cohort of older men with prostate cancer in the United States who received ADT as continuous treatment for at least 1 year. We also identified patient, physician, and disease factors associated with testing.
MATERIALS AND METHODS
We used Surveillance, Epidemiology, and End Results (SEER)/Medicare data for this analysis. The SEER Program of the National Cancer Institute collects uniformly reported data from population-based cancer registries that cover approximately 28% of the United States population.20 The information collected includes patient demographics, tumor characteristics, and treatment with surgery or radiation for each incident cancer.
Since 1991, SEER data have been merged with Medicare administrative data using a matching algorithm that successfully links files for over 94% of SEER patients aged ≥ 65 years.21 The Medicare claims data used in this study included the Medicare Provider Analysis and Review file (inpatient admissions), the 100% Physician/Supplier file (physicians' services for comorbidity assessment and ascertainment of bone density testing and ADT), and the Hospital Outpatient Standard Analytic file (outpatient facility services to identify comorbidity and bone density testing and ADT).
We identified all men with local/regional prostate cancer (based on the SEER summary stage variable, in which metastatic disease is defined as disease with distant site involvement) diagnosed from 2001 to 2007 who were aged ≥ 66 years at diagnosis and were enrolled in Parts A and B of fee-for-service Medicare as of 1 year before diagnosis through 6 months after diagnosis (N = 136,066).22 Men with metastatic disease were not studied because bone density testing is not reliable in bone with metastatic lesions; moreover, guidelines recommend bisphosphonate therapy for metastatic disease without the need for bone density testing. We excluded 1817 men who were diagnosed at autopsy and 3261 men who had no claims from 45 days before diagnosis through 195 days after diagnosis (suggesting incomplete data). We then restricted to 118,839 men with locoregional prostate cancer who were followed through 2009. Among these men, we identified 29,860 who had received ADT continuously for at least 1 year. ADT was ascertained based on claims for gonadotropin-releasing hormone agonists (Health Care Common Procedure Coding System [HCPCS] codes J9217, J9218, J9219, and J1950) or bilateral orchiectomy (Current Procedural Terminology [CPT] codes 54520, 54521, 54522, 54530, 54535, 54690, and 49510; International Classification of Diseases, ninth edition [ICD-9] procedure codes 62.3, 62.4, 62.41, and 62.42). Men were considered hypogonadal for 6 months after the date of their last injection with gonadotropin-releasing hormone agonist therapy because it is known that treatment effects persist for prolonged periods; thus, a man who received 6 months of adjuvant ADT (for example, with two 3-month depot injections of ADT) would be considered on therapy for 9 months and would not be included in these analyses.
Bone Density Testing
We assessed receipt of bone density testing (covered by Medicare every 2 years), including testing with dual-energy x-ray absorptiometry, ultrasound, or computed tomography bone density testing in the 18-month period from 6 months before the first dose of ADT through a 1-year period after the initiation of ADT (CPT codes 76070, 76071, 76075, 76076, 76077, 76078, 76977, 77078, 77079, 77080, 77081, 77082, 77083, 78350, and 78351; ICD-9 procedure code 88.98; HCPCS code G0130; and ICD-9 diagnosis code V82.81). This time frame allowed us to capture baseline bone density testing before the initiation of ADT or up to 1 year after starting therapy.
We characterized patients' age, race/ethnicity, marital status, urban residence, SEER region, comorbid illness at the time of initiation of ADT (based on the Klabunde modification of the Charlson Index23,24), year of initiation of ADT, tumor grade (by Gleason score), primary treatment (surgery, radiation, or neither), median household income and proportion of high school graduates in the census tract of residence (categorized in quartiles within registries), and hospitalizations in the year after starting ADT. We also characterized visits with physicians in the year after starting ADT, focusing on urologists and medical oncologists, who are the primary prescribers of long-term ADT, and primary care physicians (PCPs), defined as general internists, family practitioners, general practitioners, and geriatricians. Variables were categorized as indicated in Table 1. It is noteworthy that nearly all men who were included in the current study visited a urologist in addition to the other providers being assessed.
Table 1. Patient Characteristics and Receipt of Bone Density Testing From 6 Months Before the First Dose of Androgen-Deprivation Therapy Through 1 Year After
P values were based on logistic regression with generalized estimating equations to adjust standard errors for clustering within region and also adjusting for all variables in the table (the Hosmer-Lemeshow goodness-of-fit test suggests good model fit; P = .72).
This value was not reported because of confidentiality issues related to small sample sizes.
Gleason grade 7 was categorized as moderately differentiated before January 1, 2003 and as poorly differentiated as of January 1, 2003.
Primary treatment received in the 6 mo after diagnosis
Physicians seen in period from 6 mo before through 12 mo after first ADT dose
Urologist, no PCP or medical oncologist
Urologist and PCP, no medical oncologist
Urologist, PCP, and medical oncologist
Urologist and medical oncologist, no PCP
No urologist, PCP or medical oncologist
We described receipt of bone density testing from 6 months before the first dose of ADT through 12 months after the first dose by patient characteristics. We then used multivariable logistic regression with generalized estimating equations (to account for clustering by region) to assess the association of patient and tumor characteristics, visits with physicians, and hospitalizations on receipt of bone density testing. Independent variables included all variables listed in Table 1. In sensitivity analyses, we also examined use of bone density testing among the 13,592 men who received ADT for at least 2 years.
All tests of statistical significance were 2-sided. We used the SAS statistical software package (version 9.2; SAS Institute Inc, Cary, NC) for analyses. The study was approved by the institutional review boards at Harvard Medical School and Massachusetts General Hospital.
Characteristics of the 28,960 men with locoregional prostate cancer who received ADT for at least 1 year are included in Table 1. Overall, 10.2% of these men underwent bone density testing during the period from 6 months before through 1 year after initiation of ADT.
Unadjusted rates of bone density testing by patient characteristics are presented in Table 1, along with adjusted odds ratios (OR) and 95% confidence intervals (CIs). Rates of bone density testing increased over time, with 14.5% of men who initiated ADT during 2007 and 2008 undergoing testing versus 6% of men who initiated ADT during 2001 and 2002 (OR, 2.29; 95% CI, 1.83-2.85) (Fig. 1). Men aged ≥ 85 years were less likely than men ages 66 to 69 years to undergo bone density testing (OR, 0.76; 95% CI, 0.65-0.89). There was not a statistically significant difference in testing rates between men of other ages compared with men ages 66 to 69 years (Table 1). Black men were less likely than white men to undergo testing (OR, 0.72; 95% CI, 0.61-0.86), and men who were not black or Hispanic had higher rates of testing than white men (OR, 1.39; 95% CI, 1.13-1.71). Men living in areas with higher educational attainment were more likely to undergo bone density testing than those living in areas with the lowest education levels. Testing rates did not vary significantly by area-level median household income (Table 1). Unmarried men were less likely to have bone density testing than married men (OR, 0.82; 95% CI, 0.72-0.93). Bone density testing was more frequent among men who had ≥2 (vs none) comorbid medical conditions. Testing also varied substantially by region, with the highest rates in Los Angeles and Hawaii and the lowest rates in Iowa and Connecticut.
The types of physicians with whom patients had visits also were associated with bone density screening. Men who had a medical oncologist or PCP or both involved in their care were more likely to undergo testing than those who visited a urologist but no medical oncologist or PCP (Table 1), with the highest odds of bone density testing for men who visited a urologist, a medical oncologist, and PCP. Hospitalization during the year after initiation of ADT was not associated with receipt of bone density testing.
In sensitivity analyses among men who received continuous ADT for at least 2 years, 17.7% of men had bone density testing from 6 months before starting ADT through the 2-year period. Associations between bone density testing and study year, age, race/ethnicity, marital status, education, region, and physician specialty were similar to the associations observed from examining 1 year of treatment (results not shown).
We examined bone density testing for a large, population-based cohort of older men in the United States who were diagnosed with locoregional prostate cancer and received ADT for at least 1 year, and we observed that only 10.2% of these men received bone density testing between 6 months before and 12 months after the initiation of ADT. Rates of bone density testing increased over time, although only 14.5% of men who initiated ADT during 2007 and 2008 underwent testing, despite guideline recommendations for testing that were published in 2008.15,16 Black men and elderly men were significantly less likely than other men to undergo bone density testing. Treatment by a medical oncologist or PCP in addition to other providers was associated with a greater likelihood of bone density testing.
Our finding of low rates of bone density testing among prostate cancer survivors who received ADT is consistent with previously reported evidence. Several small, single-institution studies in Veteran's Health Administration settings indicated that 8.7% to 14% of men who received ADT for prostate cancer of various stages underwent bone density testing.17-19 A recent study of Canadian men who received at least 6 months of ADT for prostate cancer reported that the rate of bone density testing during the 2 years after ADT initiation ranged from 0.5 per 100 person-years in 1995 to 18 per 100 person-years in 2008.25 The lower rates of testing in our study are likely related to our assessing bone density testing through 12 months after ADT initiation rather than 24 months, although differences in the US versus Canadian health care systems also may have contributed. Despite the low rates of bone density testing in our study, rates were higher than in the general male population in Medicare, in which the rates were 0.6% in 1999 and 1.7% in 2005, substantially lower than the rates for women during the same periods.26
Our observation that black men had lower rates of bone density testing than white men may have been the result of physicians' awareness that black men generally have higher baseline bone density than white men.27,28 However, despite starting with a higher baseline bone density, black men and white men on ADT lose bone at an equivalent rate.29 Consistent with this, guidelines suggest that all men who receive ADT undergo baseline and subsequent bone density testing to assist in determining whether pharmacologic therapy to increase bone density is necessary.30
Men aged > 85 years also were less likely to undergo bone density testing in the current analysis. This finding is consistent with other evidence suggesting less bone density testing in other settings as patients grow older.31 However, the risk of fracture increases with increasing age. One study indicated that 98.8% of men aged > 80 years met criteria for treating bone loss to prevent fracture based on recommendations using the World Health Organization Fracture Risk Assessment (FRAX) algorithm treatment threshold.32,33 It is possible that we observed lower rates of bone density testing in this group because physicians assumed they should be treated for osteoporosis based on age and receipt of ADT alone, obviating the need for additional radiographic data. We lacked data on oral medications to assess bisphosphonate therapy. An alternate explanation is that physicians recommend less bone density testing for older men because they perceive lower benefits to screening. It is also possible that physicians recommended testing, but patients elected not to undergo testing.
Unmarried men and men living in areas with lower educational attainment were less likely than other men to undergo bone density testing. It is generally recognized that married men's health behaviors are influenced by their spouses, and unmarried men with prostate cancer have poorer overall survival.34 Unmarried men may not receive additional encouragement to access health care resources in general. Less bone density testing in men living in areas with lower educational attainment may reflect the challenges of communicating the benefits of testing to men who have less education or lower health literacy. Additional resources may be needed to improve communication with such populations about the benefits of testing.
Treatment by a medical oncologist and/or a PCP versus a urologist without either of these providers was associated with higher rates of bone density testing. Medical oncologists and PCPs may be more attuned to issues like osteoporosis prevention than urologists. Alternatively, this finding may reflect differences in the patients who seek care from multiple providers, who may differ from individuals who receive care from a single provider. These men may have more time available for additional physician visits and testing and may acquire more knowledge about the risks of treatments because of care from an interdisciplinary team. In addition, guidelines recommending routine use of bone density testing before and during treatment with ADT are published by the National Comprehensive Cancer Network and the American College of Physicians and may be used more commonly by medical oncologists and PCPs than by urologists. The American Society of Oncology has not developed guidelines addressing bone density testing for men with prostate cancer.
Men with more comorbidities were more likely to undergo bone density testing than those who had no comorbidities. Similar to men who have multiple practitioners involved in their cancer care, men who have more comorbidities may have more opportunities for identifying a need for testing.
It is noteworthy that some of the factors we investigated were not associated with receipt of bone density testing in our study, including tumor grade, primary treatment, area-level income, and urban/rural status, although we observed large regional variations. Unlike an analysis of bone density testing among men who received treatment in a Veteran's Health Administration hospital, Hispanic ethnicity was not associated with bone density testing in our study.17
Evidence about the adverse effects of ADT on skeletal health has been available for almost a decade; and, since 2008, bone density testing has been recommended by the National Comprehensive Cancer Network.15,30 A recent modeling study suggested that bone density testing to guide treatment with bisphosphonates in men who are receiving ADT for localized prostate cancer is a cost-effective approach to this aspect of survivorship care.35 Thus, as efforts to improve the delivery of cost-effective preventive care increase, measuring and incentivizing the use of bone density testing for this population may be an effective strategy.
Our study has some limitations. First, our study period started before evidence about ADT-associated bone loss was widely available and ended in 2009. Thus, low rates of bone density testing in the early years of this analysis may be explained by limited knowledge about the effects of ADT on bone. However, the rates of testing remained low throughout the study period, even after guidelines were published. It is possible that practitioners were unaware of the guidelines or disagreed with them because they were based on trials that examined bone mineral density, which is a surrogate endpoint, rather than fracture prevention. Second, we lacked information about physicians' recommendations for testing and could only determine whether testing was received. Third, we studied only men aged > 65 years who were living in SEER areas and were enrolled in fee-for-service Medicare; testing patterns may differ among younger men or men in other parts of the United States. In addition, we did not assess bisphosphonate use, because information on oral medication was not available before 2006 and was available only for men enrolled in Medicare Part D after 2006. We also did not examine visits with radiation oncologists, who may prescribe adjuvant ADT. Finally, the current study was exploratory, and our statistical analysis did not adjust P values for multiple comparisons.
In conclusion, the current results indicate that few prostate cancer survivors who receive long-term ADT undergo bone density testing, and several key populations, including African Americans and older men, have considerably lower rates of bone density screening. Additional efforts are needed to increase screening for treatment-associated osteoporosis to prevent fractures in these men.
We thank Yang Xu, MS, for expert programming assistance and Garrett Kirk for assistance with article preparation.
This study was funded by the Prostate Cancer Foundation.