Osteoporosis Screening in Postmenopausal ... - Wiley Online Library

25 downloads 23728 Views 148KB Size Report
Jan 16, 2014 - Technologists used standard protocols for positioning and analysis of DXA measurements. The quality assurance program is available at ..... compared to cardiovascular disease and breast cancer: the WomenLs.
ORIGINAL ARTICLE

JBMR

Osteoporosis Screening in Postmenopausal Women 50 to 64 Years Old: Comparison of US Preventive Services Task Force Strategy and Two Traditional Strategies in the Women’s Health Initiative Carolyn J Crandall,1 Joseph Larson,2 Margaret L Gourlay,3 Meghan G Donaldson,4 Andrea LaCroix,2 Jane A Cauley,5 Jean Wactawski‐Wende,6 Margery L Gass,7 John A Robbins,8 Nelson B Watts,9 and Kristine E Ensrud10 1

Department of Internal Medicine, University of California at Los Angeles, Los Angeles, CA, USA Fred Hutchinson Cancer Research Center, Seattle, WA, USA 3 Department of Family Medicine, University of North Carolina, Chapel Hill, NC, USA 4 Centre for Clinical Epidemiology and Evaluation, University of British Columbia, Vancouver, Canada 5 Department of Epidemiology, University of Pittsburgh, Pittsburgh, PA, USA 6 Department of Social and Preventive Medicine, State University of New York at Buffalo, Buffalo, NY, USA 7 Consultant, Cleveland Clinic Center for Specialized Women’s Health, Mayfield Heights, OH, USA 8 Center for Healthcare Policy and Research, UC Davis Medical Center, Sacramento, CA 9 Mercy Health Osteoporosis and Bone Health Services, Cincinnati, OH, USA 10 Division of Epidemiology and Community Health, University of Minnesota Medical School, and Minneapolis VA Health Care System, Minneapolis, MN, USA 2

ABSTRACT The US Preventive Services Task Force (USPSTF) recommends osteoporosis screening for women younger than 65 years whose 10‐year predicted risk of major osteoporotic fracture is 9.3%. For identifying screening candidates among women aged 50 to 64 years, it is uncertain how the USPSTF strategy compares with the Osteoporosis Self‐Assessment Tool (OST) and the Simple Calculated Osteoporosis Risk Estimate (SCORE). We examined data (1994 to 2012) from 5165 Women’s Health Initiative participants aged 50 to 64 years. For the USPSTF (Fracture Risk Assessment Tool [FRAX] major fracture risk 9.3% calculated without bone mineral density [BMD]), OST (score 7) strategies, we assessed sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) to discriminate between those with and without femoral neck (FN) T‐score 2.5. Sensitivity, specificity, and AUC for identifying FN T‐score 2.5 were 34.1%, 85.8%, and 0.60 for USPSTF (FRAX); 74.0%, 70.8%, and 0.72 for SCORE; and 79.8%, 66.3%, and 0.73 for OST. The USPSTF strategy identified about one‐third of women aged 50 to 64 years with FN T‐scores 2.5. Among women aged 50 to 64 years, the USPSTF strategy was modestly better than chance alone and inferior to conventional SCORE and OST strategies in discriminating between women with and without FN T‐score 2.5. © 2014 American Society for Bone and Mineral Research. KEY WORDS: OSTEOPOROSIS; FRACTURE; BONE MINERAL DENSITY; FRACTURE RISK ASSESSMENT TOOL; USPSTF, OST, SCORE, FRAX

Introduction

O

ne half of all postmenopausal women will have an osteoporosis‐related fracture during their lifetime.(1) Testing for and treating women with low bone mineral density (BMD) (BMD T‐score 2.5 or less) can decrease the risk for subsequent fractures and fracture‐related morbidity and mortality.(1) In 2011, the United States Preventive Services Task Force (USPSTF) recommended routine screening for osteoporosis for all women

aged 65 years and older and endorsed use of the Fracture Risk Assessment Tool (FRAX) to identify screening candidates among younger postmenopausal women aged 50 to 64 years.(2) FRAX is a Web‐based tool that uses clinical risk factors with and without femoral neck (FN) BMD to estimate 10‐year probability of hip and major osteoporotic (hip, clinical vertebral, humerus, or wrist) fractures. Specifically, the USPSTF recommends BMD testing for women aged 50 to 64 years whose 10‐year predicted risk of major osteoporotic fracture (calculated using the FRAX model

Received in original form November 6, 2013; revised form January 3, 2014; accepted January 9, 2014. Accepted manuscript online January 16, 2014. Address correspondence to: Carolyn J Crandall, MD, David Geffen School of Medicine, University of California at Los Angeles, UCLA Medicine/GIM, 911 Broxton Avenue, 1st Floor, Los Angeles, CA 90024, USA. E‐mail: [email protected] Additional Supporting Information may be found in the online version of this article. Journal of Bone and Mineral Research, Vol. 29, No. 7, July 2014, pp 1661–1666 DOI: 10.1002/jbmr.2174 © 2014 American Society for Bone and Mineral Research

1661

without BMD) is 9.3% (equivalent to that of a 65‐year‐old white woman with no other FRAX clinical risk factors).(1) Before the advent of FRAX, several tools were available for the prediction of osteoporosis risk, including the Osteoporosis Self‐ Assessment Tool (OST, based on weight and age) and the Simple Calculated Osteoporosis Risk Estimation Tool (SCORE, based on race, rheumatoid arthritis, history of nontraumatic fracture, age, prior estrogen therapy, and weight).(3–6) For identifying osteoporosis by BMD (T‐score 2.5) among postmenopausal women, an OST score cut‐off of 1, 1 T‐score >2.5, T‐score  2.5); 2) the proportion of women with femoral neck T‐score 2.5 who would be identified for screening under each strategy; and 3) the sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) for identifying low BMD (T‐score between 1 and 2.5) and osteoporosis (T‐score  2.5) under each strategy. In secondary analyses, we calculated the AUC of the three tools for identifying of T‐score 2.5 at one or more of the following sites: lumbar spine, total hip, or femoral neck. We also estimated the cut‐off score that would identify 90% of women with femoral neck T‐scores 2.5.

Risk assessment strategies

Materials and Methods Participants The Women’s Health Initiative was conducted at 40 clinical centers nationwide.(7) Eligibility criteria for the clinical trials (WHI‐ CT) and the observational study (WHI‐OS) included being aged 50 to 79 years at baseline, postmenopausal, and free from serious medical conditions.(8,9) The WHI‐CT consisted of randomized controlled trial evaluation of three interventions: a low‐fat eating pattern, menopausal hormone therapy (HT), and calcium and vitamin D supplementation.(9) Details are available at https://cleo.whi.org/about/SitePages/About%20WHI.aspx. All WHI participants were postmenopausal, defined as at least 6 months of amenorrhea for women aged 55 years, and at least 12 months of amenorrhea for women aged 50 to 54 years.(10) At enrollment, WHI‐OS and WHI‐CT participants at three of the 40 clinical centers (Tucson and Phoenix, AZ; Pittsburgh, PA; and Birmingham, AL) underwent hip and anteroposterior lumbar spine BMD testing by dual‐energy X‐ray absorptiometry (DXA, Hologic QDR2000 or QDR4500, Bedford, MA, USA).(11,12) Technologists used standard protocols for positioning and analysis of DXA measurements. The quality assurance program is available at https://biolincc.nhlbi.nih.gov/static/studies/whios/ doc/whi/procedur/bone/1.pdf. Quality assurance included review of lumbar spine and hip phantom scans at each center, use of calibration phantoms across clinical sites, flagging of scans with specific problems, and review of a random sample of all scans.(13) Femoral neck T‐score classification was based on the National Health and Nutrition Examination III normative reference database.(14) Of the 11,488 participants at the three clinical centers that measured BMD, 6294 were aged 50 to 64 years at baseline. The current analysis is based on the 5165 participants aged 50 to

1662

CRANDALL ET AL.

Information regarding osteoporosis risk factors (age, race, rheumatoid arthritis, history of prior fracture, medication use, smoking, alcohol intake, and parental history of hip fracture) was obtained from baseline self‐assessment questionnaires and weight and height measurements. Ten‐year risk of major osteoporotic fracture was calculated for each participant by the WHO Collaborating Centre for Metabolic Bone Disease, using the FRAX tool without BMD (version 3.0).(2,15) Per the USPSTF screening guidelines, we defined participants with FRAX‐ predicted 10‐year risk of major osteoporotic fracture 9.3% to be recommended for BMD testing. The calculation of SCORE and OST scores was based on prior publications (Supplemental Table S1).(3–6) SCORE values vary according to use of HT, but the USPSTF strategy does not account for HT. Thus, we present overall results and results stratified according to baseline current use or randomization to active treatment versus non‐use of HT (oral or transdermal patch).

Statistical analysis Using chi‐square tests, we compared the proportion of participants who would be identified for BMD testing using the three strategies (USPSTF FRAX 9.3%, OST score 7). Next, within each T‐score category (T‐score 1, 1> T‐score >2.5, T‐score 2.5), we used chi‐square tests to compare the proportion of participants who would be recommended for BMD testing by each of the strategies. We determined the sensitivity, specificity, positive predictive value ([number of participants with T‐score in the interval of interest who would be recommended for BMD testing/total number recommended for BMD testing] 100) and AUC curves of the strategies in discriminating participants with femoral neck T‐score 1 (ie, a normal T‐score) from those with T‐score 2.5 Journal of Bone and Mineral Research

and 1> T‐score >2.5. Because the scores of the three risk strategies were correlated with each other within the same women, we calculated differences in AUCs of the three tools for 10,000 bootstrap samples. We stratified the AUC results according to age categories chosen a priori: 50 to 54 years, 55 to 59 years, and 60 to 64 years. For each of the three scores, we constructed receiver operating characteristic (ROC) curves for the identification of femoral neck T‐scores 2.5. Finally, for each of the three screening strategies, we calculated the thresholds that would correspond to sensitivities in the range of 80% to 99% for the detection of femoral neck T‐scores 2.5, along with the associated specificities and AUC values. Our primary analyses focused on participants who were nonusers of menopausal hormone therapy (n ¼ 2857). In supplemental analyses, we stratified our results according to use versus non‐use of menopausal hormone therapy. Analyses were performed using SAS for Windows Version 9.2 (SAS Institute, Cary, NC, USA).

Results Study participant characteristics Seventy‐two percent of participants were white, 17% were black, and 8% were Hispanic (Table 1). At baseline, approximately one‐ third had body mass index 30 kg/m2 and 9.5% were current smokers. Average age was 57.7 years (median 58, interquartile range 54 to 61). Five percent of the analytic sample had femoral neck T‐scores 2.5; 46% had 1> femoral neck T‐score 2.5. Mean 10‐year predicted major osteoporotic fracture risk was 6.6% (median 5.8%, range 0.74% to 47.7%). Compared with analytic sample participants, excluded participants were less likely to be black (9.4% verus 17.3%); distributions of age, smoking, body mass index (BMI), alcohol use, and diabetes were not significantly different between groups (data not shown).

Comparisons of the three risk‐assessment strategies The USPSTF strategy identified 15.2% of all participants aged 50 to 64 for BMD screening, compared with 31.5% under the SCORE strategy and 36.0% under the OST strategy (chi‐square p < 0.001, Fig. 1). Among women with T‐score 2.5 who were not using menopausal hormone therapy, the USPSTF strategy identified 33.3% for BMD testing, compared with 74.1% using SCORE and 79.3% using OST (chi‐square p < 0.001, Table 2, Supplemental Fig. S2). The proportion of women with 1 > T‐score >2.5 who were identified for testing under the three strategies were 17.5% for USPSTF, 42.2% for SCORE, and 48.9% for OST (chi‐square p < 0.001). Stratifying for baseline HT use did not notably alter these results. Results were similar among participants taking menopausal hormone therapy (Supplemental Table S2). Among participants not using menopausal hormone therapy, of the three strategies, the USPSTF strategy had the lowest sensitivity (34.1%) for identifying femoral neck T‐score 2.5, but the highest specificity (85.8%) (Table 3). Pairwise comparisons of the AUCs for the three strategies in identifying women with T‐score 2.5 revealed significantly higher AUC for both SCORE and OST compared with the USPSTF strategy (p < 0.01, data not shown). The AUC for OST was not statistically significantly different from that of SCORE. Among women with femoral neck T‐score 2.5, the positive predictive value under each of the three strategies was similar, approximately 11% in the

Journal of Bone and Mineral Research

Table 1. Selected Baseline Characteristics of the 5165 Study Participants No. (%) of participants Age (years) 50–54 1371 (26.5) 55–59 1744 (33.8) 60–64 2050 (39.7) Body mass index (kg/m2) Missing 21 (0.4) 5 captured 90.3% of participants with femoral neck T‐score 2.5. An OST score of 2 captured 89.9% of participants with femoral neck T‐score 2.5. The pattern of lower AUC for the USPSTF strategy compared with SCORE and OST was especially pronounced among women aged 50 to 54 years and 55 to 59 years (Supplemental Table S4). The AUC for identifying T‐score 2.5 at any skeletal site (lumbar spine, femoral neck, total hip) was lower for the USPSTF strategy than for the OST or SCORE strategies (Supplemental Table S5). For all three tools, the AUC for identification of femoral neck

OSTEOPOROSIS SCREENING IN POSTMENOPAUSAL WOMEN

1663

Discussion

Fig. 1. Proportion of women aged 50 to 64 years who would be identified for BMD testing according to each of the three strategies (n ¼ 5165). Proportions are unadjusted. The USPSTF, SCORE, and OST strategies significantly differed from each other in identifying participants for BMD testing (all pairwise chi‐square p < 0.001 for all participants group).

T‐score 2.5 was higher than the AUC for identification of T‐score 2.5 at any site. Table 4 displays the sensitivities, specificities, positive predictive values, and AUC values for alternative thresholds in identifying femoral neck T‐score 2.5. The identification of 90% of participants with T‐score 2.5 corresponded to an OST score of 5, and a FRAX score of 4.11. For the range of thresholds that corresponded to sensitivities 90% in detection of femoral neck T‐scores 2.5, we found AUC values greater than 0.70 for SCORE scores greater than >5 and for OST thresholds less than 3. A specificity of 70% was not found for any of the strategies at thresholds that had sensitivities of 80%.

Current osteoporosis screening guidelines are based mostly on studies of women aged 65 years and older. In contrast, there are limited data regarding optimal osteoporosis screening strategies for younger postmenopausal women. In this study of women aged 50 to 64 years, under the USPSTF (FRAX‐based) strategy, only 34.1% of women with T‐score 2.5 would be recommended for BMD testing, compared with 74.0% with SCORE and 79.8% with OST. The positive predictive values of the three strategies for identifying women with femoral neck T‐score 2.5 was similar, approximately 11%. The ability of the strategy to discriminate between women with and without densitometric osteoporosis was significantly lower for USPSTF (AUC 0.60) than for SCORE (AUC 0.72) or OST (AUC 0.73). In contrast, specificity of the USPSTF strategy was higher than SCORE and OST. To our knowledge, prior studies have not compared the current USPSTF (FRAX) strategy to that of SCORE and OST among US women aged 50 to 64 years. Although it did not examine the USPSTF strategy, one prior study of OST and SCORE among women aged 45 to 64 years found that the tools had similar AUC (0.77 for OST and 0.76 for SCORE) for identifying women with T‐score 2.5,(5) as was the case in the current study. The FRAX, OST, and SCORE thresholds that would be required to identify 90% of 50‐ to 64‐year‐olds with femoral neck T‐score 2.5 are different from the cut‐off scores traditionally recommended as screening thresholds. The alternative cut points for OST, SCORE, and FRAX that would have resulted in identification of 80% of women with femoral neck T‐score 2.5 corresponded to specificities less than 70%. Our results have potential clinical implications. The objective of BMD screening is to identify postmenopausal women with T‐scores 2.5 because pharmacologic treatment to prevent fractures has been demonstrated to be effective in this group. (The efficacy of pharmacologic therapy in women at high fracture risk, but without T‐score 2.5 or less or existing vertebral fractures, is uncertain.) Therefore, the ability of the USPSTF strategy to detect BMD T‐score 2.5 is of great clinical importance. Our results suggest that the USPSTF FRAX threshold

Table 2. Proportion of Women Aged 50 to 64 Years Who Would Be Identified for BMD Testing by the Three Methods According to Femoral Neck T‐Score Categorya Nonusers of menopausal hormone therapy (n ¼ 2857) T‐score –2.5

USPSTF (FRAX 9.3%)

SCORE (SCORE score >7)

OST (OST score 7) OST (2.5 who have not experienced fractures. In women of this age group, a simple model (OST) based on weight and age discriminated between women with and without osteoporosis as well as the more complex USPSTF approach. Because the goal of

Journal of Bone and Mineral Research

osteoporosis screening is to identify postmenopausal women with BMD T‐score 2.5 for pharmacologic therapy, these results could have substantial implications for osteoporosis screening of younger postmenopausal women in clinical practice.

Disclosures AL serves on an Amgen Scientific Methodology Advisory Committee for safety monitoring of Prolia. NW is stockholder and director of OsteoDynamics. He has received honoraria for lectures from the following companies in the past year: Amgen, Lilly, Novartis, and Warner Chilcott. He has received consulting fees from the following companies in the past year: Abbott, Amgen, Baxter, Bristol‐Myers Squibb, Imagepace, Johnson & Johnson, Lilly, Medpace, Merck, Novo Nordisk, and Pfizer/Wyeth. Through his university, he has received research support from the following companies: Amgen, Merck, and NPS. All other authors state that they have no conflicts of interest.

Acknowledgments We thank the following: program office (National Heart, Lung, and Blood Institute, Bethesda, MD): Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, and Nancy Geller. Clinical coordinating center (Fred Hutchinson Cancer Research Center, Seattle, WA): Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg. Investigators and academic centers: (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA) JoAnn E Manson; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V Howard; (Stanford Prevention Research Center, Stanford, CA) Marcia L Stefanick; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Arizona, Tucson/ Phoenix, AZ) Cynthia A Thomson; (University at Buffalo, Buffalo, NY) Jean Wactawski‐Wende; (University of Florida, Gainesville/ Jacksonville, FL) Marian Limacher; (University of Iowa, Iowa City/ Davenport, IA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (Wake Forest University School of Medicine, Winston‐Salem, NC) Sally Shumaker.

OSTEOPOROSIS SCREENING IN POSTMENOPAUSAL WOMEN

1665

The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, US Department of Health and Human Services through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201 100002C, HHSN268201100003C, HHSN268201100004C, and HHSN271201100004C. The sponsor had no role in the design, analysis, writing, or review of this manuscript. CC received support from the Jonsson Comprehensive Cancer Center at the University of California, Los Angeles. Authors’ roles: study concept and design: CC; acquisition of data: AL, JC, JW‐W, and JR; analysis and interpretation of data: CC, JL, MG, MD, AL, JC, JW‐W, MG, JR, NW, and KE; drafting of the manuscript: CC; critical revision of the manuscript for important intellectual content: CC, JL, MG, MD, AL, JC, JW‐W, MG, JR, NW, and KE; statistical expertise: JL; obtained funding: AL, JC, JW‐W, and JR. JL had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

6. Lydick E, Cook K, Turpin J, Melton M, Stine R, Byrnes C. Development and validation of a simple questionnaire to facilitate identification of women likely to have low bone density. Am J Manag Care. 1998;4(1):37–48. 7. Langer RD, White E, Lewis CE, Kotchen JM, Hendrix SL, Trevisan M. The Women’s Health Initiative Observational Study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol. 2003;13(9 Suppl):S107–21. 8. Cauley JA, Wampler NS, Barnhart JM, et al. Incidence of fractures compared to cardiovascular disease and breast cancer: the Women’s Health Initiative Observational Study. Osteoporos Int. 2008;19(12): 1717–23. 9. Design of the Women’s Health Initiative clinical trial and observational study. The Women’s Health Initiative Study Group. Control Clin Trials. 1998;19(1):61–109. 10. Hays J, Hunt JR, Hubbell FA, et al. The Women’s Health Initiative recruitment methods and results. Ann Epidemiol. 2003;13(9 Suppl): S18–77.

References

3. Cadarette SM, McIsaac WJ, Hawker GA, et al. The validity of decision rules for selecting women with primary osteoporosis for bone mineral density testing. Osteoporos Int. 2004;15(5):361–6.

12. LaCroix AZ, Beck TJ, Cauley JA, et al. Hip structural geometry and incidence of hip fracture in postmenopausal women: what does it add to conventional bone mineral density? Osteoporos Int. 2010;21(6):919–29. 13. Chen Z, Arendell L, Aickin M, Cauley J, Lewis CE, Chlebowski R. Hip bone density predicts breast cancer risk independently of Gail score: results from the Women’s Health Initiative. Cancer. 2008;113(5): 907–15. 14. Looker AC, Wahner HW, Dunn WL, et al. Proximal femur bone mineral levels of US adults. Osteoporos Int. 1995;5(5):389–409.

4. Geusens P, Hochberg MC, van der Voort DJ, et al. Performance of risk indices for identifying low bone density in postmenopausal women. Mayo Clin Proc. 2002;77(7):629–37.

15. Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E. FRAX and the assessment of fracture probability in men and women from the UK. Osteoporos Int. 2008;19(4):385–97.

5. Gourlay ML, Miller WC, Richy F, Garrett JM, Hanson LC, Reginster JY. Performance of osteoporosis risk assessment tools in postmenopausal women aged 45–64 years. Osteoporos Int. 2005;16(8):921–7.

16. National Osteoporosis Foundation. Clinician’s guide to prevention and treatment of osteoporosis. Washington, DC: National Osteoporosis Foundation; 2010.

1. Screening for osteoporosis: U.S. preventive services task force recommendation statement. Ann Intern Med. 2011;154(5):356–64. 2. World Health Organization Collaborating Centre for Metabolic Bone Diseases UoS, UK. FRAX WHO Fracture Risk Assessment Tool, vol. 2012. Sheffield, UK: University of Sheffield; 2009.

1666

CRANDALL ET AL.

11. Beck TJ, Petit MA, Wu G, LeBoff MS, Cauley JA, Chen Z. Does obesity really make the femur stronger? BMD, geometry, and fracture incidence in the Women’s Health Initiative‐observational study. J Bone Miner Res. 2009;24(8):1369–79.

Journal of Bone and Mineral Research