untitled

pissn: 2093-940X, eissn: 2233-4718 Journal of Rheumatic Diseases Vol. 25, No. 4, October, 2018 https://doi.org/10.4078/jrd.2018.25.4.263 Original Article Korean Guideline for the Prevention and Treatment of Glucocorticoid-induced Osteoporosis So Young Park 1, Hyun Sik Gong 2, Kyoung Min Kim 3, Dam Kim 4, Hayoung Kim 5, Chan Hong Jeon 6, Ji Hyeon Ju 7, Shin-Seok Lee 8, Dong Ah Park 9, Yoon-Kyoung Sung 10, Sang Wan Kim 11 1 Division of Endocrinology and Metabolism, Department of Internal Medicine, Kyung Hee University Hospital, Seoul, 2 Department of Orthopedic Surgery, 3 Division of Endocrinology and Metabolism, Department of Internal Medicine, Bundang Hospital, Seoul National University College of Medicine, Seongnam, 4 Division of Rheumatology, Department of Internal Medicine, Yonsei University College of Medicine, Seoul, 5 Division of Endocrinology, Department of Internal Medicine, Sanbon Hospital, Wonkwang University School of Medicine, Gunpo, 6 Division of Rheumatology, Department of Internal Medicine, Soonchunhyang University Bucheon Hospital, Bucheon, 7 Division of Rheumatology, Department of Internal Medicine, Seoul St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, 8 Department of Rheumatology, Chonnam National University Medical School and Hospital, Gwangju, 9 National Evidence-based Healthcare Collaborating Agency, Seoul, 10 Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, Seoul, 11 Division of Endocrinology and Metabolism, Department of Internal Medicine, SMG-SNU Boramae Medical Center, Seoul National University College of Medicine, Seoul, Korea Objective. To develop guidelines and recommendations to prevent and treat glucocorticoid-induced osteoporosis (GIOP) in Korea. Methods. The Korean Society for Bone and Mineral Research and the Korean College of Rheumatology developed this guideline based on Guidance for the Development of Clinical Practice Guidelines version 1.0 established by the National Evidence-Based Healthcare Collaborating Agency. This guideline was developed by adapting previously-published guidelines, and a systematic review and quality assessment were conducted. Results. This guideline applies to adults aged 19 years or older who are using or plan to use glucocorticoids (GCs), but does not include children and adolescents. An initial assessment of fracture risk should be performed within 6 months of initial GC use. Fracture risk should be estimated using FRAX (Fracture Risk Assessment Tool) with adjustments for GC dose, previous osteoporotic fracture history, and bone mineral density (BMD) results. All patients taking more than 2.5 mg/day prednisolone or equivalent for more than 3 months are recommended to take adequate calcium and vitamin D. Patients at moderate to high fracture risk should be treated with additional osteoporosis medication. All patients continuing GC therapy should receive an annual BMD measurement, vertebral X-ray, and fracture risk assessment using FRAX. When a treatment failure is suspected, switching to another drug should be considered. Conclusion. This guideline is intended to provide guidance for clinicians in prevention and treatment of GIOP. (J Rheum Dis 2018;25:263-295) Key Words. Denosumab, Diphosphonates, Glucocorticoids, Osteporosis, Teriparatide INTRODUCTION Glucocorticoids (GCs) are very effective drugs for the treatment of inflammatory diseases and have been widely used in various diseases [1-3]. However, long-term use of GCs has detrimental effects on bone microstructure, leading to a decrease in bone mass and an increase in the risk of fracture [3,4]. It is estimated that fractures occur in 30% 50% of patients receiving long-term GC therapy, but many patients have no symptoms of fracture because of the analgesic effects of GCs [5-7]. There are no definitive data on the number of patients receiving GCs for three months or longer in Korea, but it is estimated to be much higher than the 1% reported in other countries Received:July 27, 2018, Revised:September 1, 2018, Accepted:September 2, 2018 Corresponding to:yoon-kyong Sung http://orcid.org/0000-0001-6691-8939 Department of Rheumatology, Hanyang University Hospital for Rheumatic Diseases, 222-1 Wangsimni-ro, Seongdong-gu, Seoul 04763, Korea. E-mail:sungyk@hanyang.ac.kr Sang Wan Kim http://orcid.org/0000-0001-9561-9110 Division of Endocrinology and Metabolism, Department of Internal Medicine, SMG-SNU Boramae Medical Center, Seoul National University College of Medicine, 20 Boramae-ro 5-gil, Dongjak-gu, Seoul 07061, Korea. E-mail : swkimmd@snu.ac.kr Copyright c 2018 by The Korean College of Rheumatology. All rights reserved. This is a Open Access article, which permits unrestricted non-commerical use, distribution, and reproduction in any medium, provided the original work is properly cited. 263

So Young Park et al. [8,9]. Therefore, glucocorticoid-induced osteoporosis (GIOP) is recognized as the most common cause of secondary osteoporosis (OP) and, because there may be no specific symptoms, active management focused on prevention is needed. The decrease of bone mass caused by GCs occurs in two stages. First, a rapid decrease in bone mass begins within the first 3 6 months of GC use, with a 6% 12% loss of bone mass in the first year of GC use [10]. Second, long-term use of GCs can result in a 3% reduction in bone mass every year [11]. Glucocorticoids affect both cortical bone and trabecular bone, but fractures occur most commonly in the vertebral body, especially in areas rich in trabecular bone, such as the lumbar spine [3,4,12]. The risk of GC-induced fracture is already increased before a significant reduction in bone mass occurs [3]. Therefore, measurements of bone mineral density (BMD) are not sufficient to evaluate the degree of GC-induced bone loss, so it is very important to identify patients with a high risk of fracture to prevent fractures caused by GIOP [13-15]. The absolute risk of an individual fracture is determined by age, gender and other risk factors for OP. Currently, FRAX (Fracture Risk Assessment Tool, https://www. shef.ac.uk/frax/tool.jsp) is a well-known method for assessing the risk of fracture. In addition to the risk factors included in the FRAX, low bone strength at the beginning of GC treatment and the rate of bone loss during treatment are suggested as risk factors for GC-induced fractures, the latter being determined by the dose and duration of GC treatment [16]. In a study of individual absolute fracture probability, patients receiving prednisolone at a dose greater than 30 mg/day (cumulative dose >5 g/year) showed significantly increased risk of vertebral and femoral fractures [17]. However, if GC treatment is terminated, BMD gradually increases and fracture risk decreases. Therefore, patients could benefit from continuous risk assessment with an emphasis on the appropriate duration of GC use [9,18,19]. As evidence of fracture risk in patients using GCs accumulates, drugs that effectively prevent fractures have been developed. However, many primary care physicians and specialists fail to recognize the severity of GIOP or determine which patients are at greatest risk for GIOP. Therefore, many patients still do not receive treatment to prevent fractures. To address this problem, guidelines for GIOP prevention and treatment have recently been developed by several countries. Notably, the guideline of the American College of Rheumatology (ACR) were revised in 2017 based on the latest evidence and applied to clinical practice [16]. In Korea, it is necessary to provide standardized clinical practical guideline (CPG) for the primary prevention and treatment of GIOP to all clinicians, to ensure that Korean patients who plan to use or use GCs receive the appropriate services for fracture prevention. The Korean Society for Bone and Mineral Research (KSBMR) and the Korean College of Rheumatology (KCR) have mutually developed guideline for the treatment of GIOP. Because of limited domestic data, this guideline was developed by adapting previously-published guidelines. MATERIALS AND METHODS This guideline was developed for adults over the age of 19 who plan to use or currently use GCs. Pediatric populations and people with a glomerular filtration rate of <30 ml/minute were excluded. A development committee and a working committee were organized to develop guideline for the treatment of GIOP. These committees were composed of multi-disciplinary and multi-institutional organizations and included endocrinologists, rheumatologists, an orthopedist, and a methodologist (Supplementary Appendix 1). A systematic literature review was conducted, and guidelines were selected and adapted from the existing literature. The completed guideline should be revised within five years, and earlier revisions may be required if a new drug is approved for GIOP or if the evidence changes significantly. Framework for GIOP guideline development Methods were based on guidance for the development of CPGs version 1.0 by the National Evidence-based Healthcare Collaborating Agency (NECA) (Supplementary Appendix 2). The process of developing this guideline included three major stages [20]: planning, development, and finalization. Each stage was divided into individual steps, for a total of 12 steps. The planning stage consisted of selecting topics (Step 1), assembling the development committee (Step 2), reviewing previously published guidelines (Step 3), establishing the development plan (Step 4), and selecting key questions (Step 5). The development stage consisted of searching for, evaluating, and synthesizing evidence (Steps 6 8), making recommendations and determining the grades of recommendations (Step 9), and consensus building (Step 10). The finalization stage consisted of external reviews and 264 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea publications (Steps 11 and 12). Selection of key questions To select the key questions (KQs) to be addressed by the GIOP guideline, a working committee consisting of a total of nine members first reviewed six guidelines developed by the United States, France, Spain, Japan, Brazil, and the International OP Foundation-European Calcified Tissue Society (IOF-ECTS). From these six guidelines, the committee selected 14 topics. After reviewing these topics, the development committee considered domestic circumstances and clinical significance to select the most relevant KQs. A final list of seven KQs was chosen, which included the patient population (P), the intervention (I), Table 1. Key questions KQ1. Is non-pharmacological treatment beneficial for the preventing and treating GIOP in adults taking glucocorticoids? KQ2. Which pharmacological treatments are effective for preventing and treating GIOP in adults <40 years of age? KQ2-1. Is calcium and vitamin D supplementation effective in the prevention and treatment of GIOP in adults <40 years of age? KQ2-2. Is bisphosphonate effective in the prevention and treatment of GIOP in adults <40 years of old? KQ2-3. Is teriparatide effective in the prevention and treatment of GIOP in adults <40 years of old? KQ2-4. Is denosumab effective in the prevention and treatment of GIOP in adults <40 years of old? KQ3. Which pharmacological treatments are effective for preventing and treating GIOP in adults 40 years of age? KQ3-1. Is calcium and vitamin D supplementation effective in the prevention and treatment of GIOP in adults 40 years of age? KQ3-2. Is bisphosphonate effective in the prevention and treatment of GIOP in adults 40 years of old? KQ3-3. Is teriparatide effective in the prevention and treatment of GIOP in adults 40 years of old? KQ3-4. Is denosumab effective in the prevention and treatment of GIOP in adults 40 years of old? KQ3-5. Are selective estrogen receptor modulators effective in the prevention and treatment of GIOP in postmenopausal women? KQ4. Is it safe to use OP medication in women planning to have a pregnancy? KQ5. How should response to treatment be monitored in patients with GIOP using physical measurements, imaging, and biochemical methods? KQ6. Should discontinuation of OP medication be considered if the fracture risk is reassessed to be low during GIOP treatment? KQ7. How should initial treatment failure be defined for GIOP? GIOP: glucocorticoid-induced osteoporosis, GC: glucocorticoid, OP: osteoporosis. Figure 1. Flowchart of the systematic search of literature and selection process for the development of glucocorticoid-induced osteoporosis by adaptation. NGC: National Guideline Clearinghouse, G-I-N: Guidelines International Network, KoMGI: Korean Medical Guideline Information Center, GC: glucocorticoid. www.jrd.or.kr 265

So Young Park et al. the comparator (C), and the outcome of the intervention (O) (PICO, Table 1). Literature search Two members of the working committee performed systematic literature searches, using the databases PubMed, OVID-EMBASE, KoreaMed, KMbase, National Guideline Clearinghouse, Guidelines International Network, and Korean Medical Guideline Information Center (KoMGI). The researchers identified a total of 309 potentially relevant articles published since 2010, excluding duplicates. By reviewing titles and abstracts, the researchers narrowed this list to the 27 most relevant articles. From these 27 articles, seven of previously-published guidelines were identified: 1) guidelines including PICO that are consistent with KQs; 2) evidence-based guidelines, which are defined by a clear link between the recommendation and the supporting evidence, including systematic literature searches; 3) guidelines for peer review; and 4) guidelines published in Korean or English (Figure 1, Supplementary Appendix 3 and 4). Final selection process for guidelines Seven guidelines were chosen based on the systematic literature review, selection criteria, and exclusion criteria (Table 2) [21-27]. A quality assessment was performed on these seven guidelines using the Appraisal of Guidelines for Research and Evaluation II (AGREE II). In this comprehensive assessment, the scope and objectives of the guidelines, strictness of development, participation of stakeholders, clarity of expression, applicability, and editorial independence were assessed [28]. The quality assessment of guidelines using AGREE II was conducted by three members of the working committee as per rec- Table 2. The guidelines selected for 1st AGREE evaluation No. Title Country Institute Year Reference 1 American College of Rheumatology 2010 recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis USA 2 A framework for the development of guidelines International for the management of glucocorticoid-induced osteoporosis 3 2010 clinical practice guidelines for the diagnosis Canada and management of osteoporosis in Canada: summary 4 2014 update of recommendations on the prevention France and treatment of glucocorticoid-induced osteoporosis 5 Guidelines for the prevention and treatment of glucocorticoid-induced osteoporosis 6 Guidelines on the management and treatment of glucocorticoid-induced osteoporosis of the Japanese Society for Bone and Mineral Research: 2014 update 7 Guidelines for the diagnosis, prevention and management of osteoporosis Brazil Japan Italy American College of Rheumatology Joint IOF-ECTS GIO Guidelines Working Group Scientific Advisory Council of Osteoporosis Canada Bone Section of the French Society for Rheumatology (SFR) and Osteoporosis Research and Information Group (GRIO) Committee for Osteoporosis and Bone Metabolic Disorders of the Brazilian Society of Rheumatology; Brazilian Medical Association; Brazilian Association of Physical Medicine and Rehabilitation Japanese Society for Bone and Mineral Research The Italian Society for Osteoporosis, Mineral Metabolism and Bone Diseases (SIOMMMS) 2010 21 2012 22 2010 23 2014 24 2012 25 2014 26 2016 27 AGREE: Appraisal of Guidelines for Research and Evaluation, IOF-ECTS: International Osteoporosis Foundation-European Calcified Tissue Society, GIO: glucocorticoid-induced osteoporosis. 266 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea ommendations, and the items with given divergent scores by the three researchers were reviewed and agreed upon. The strictness of development was specifically considered in the process of selecting guidelines. Four guidelines published in the United States (2010), IOF-ECTS (2012), Canada (2010) and France (2014) were selected [21-24]. During the selection of guidelines, the National Osteoporosis Guideline Group (NOGG) and the ACR published new guidelines for GIOP treatment in 2017 [16,29]. Therefore, the 2010 ACR guideline was replaced by the 2017 ACR guideline (Table 3), and the quality assessment was reevaluated for the five guidelines. The 2017 ACR guidelines, which ranked highly in items of "comprehensive evaluation" and "strictness of development", were chosen and adapted (Supplementary Appendix 5). The characteristics of final five guidelines were summarized in Supplementary Appendix 6. Writing process for guidelines The working committee reviewed the recommendations and evidence for the final five guidelines, and then summarized the primary recommendations for KQs (Supplementary Appendix 7). The acceptability and applicability of the recommendations of final five guidelines to the key question was assessed (Supplementary Appendix 8). After collecting the opinions of all committee members, the final recommendations were completed. If there was a lack of evidence or a need for clinical interpretation, the consensus process proceeded among the members of the working committee. The strength of evidence was divided into five levels (Table 4), and recommendation grade was assessed considering the level of evidence and clinical effects, patient satisfaction, quality of life, harmful reactions, and unnecessary use of resources (Table 5). The final recommendation grade was decided on the principle of agreement of more than 80%, along with the consent of the working committee. Selection of eligible patients and fracture risk assessment The development committee and working committee decided to apply the 2017 ACR guideline for the fracture risk classification and assessment, because relatively little evidence was available beyond these guidelines. Tables and figures were used with permission from the original author. Table 4. The level of evidence Level Expression Type of evidence I High Further research is very unlikely to change our confidence in the estimate of effect (systematic overview of meta-analysis of randomized controlled trials) II Moderate Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate (not applicalble by I, prospective observation study, case-control study) III Low Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate (retrospective observation study, case-control study) IV Very low We are very uncertain about the estimate EC Expert consensus Evidence from expert committees without explicit critical scientific analysis or that based on physiology, basic research or first principles Table 3. The additional guidelines for 2nd AGREE evaluation No. Title Country Institute Year Reference 1 2017 American College of Rheumatology Guideline for the Prevention and Treatment of Glucocorticoid-Induced Osteoporosis 2 UK clinical guideline for the prevention and treatment of osteoporosis USA UK American College of Rheumatology National Osteoporosis Guideline Group (NOGG) 2017 16 2012 29 AGREE: Appraisal of Guidelines for Research and Evaluation. www.jrd.or.kr 267

So Young Park et al. Table 5. The grade of recommendations Grade Expression Type of recommendations A Recommended It is recommended that the intervention be conducted with sufficient evidence of the desired effect B C I Conditionally recommended Not recommended No recommendation The evidence for the desired effect of the intervention is between moderate and high grounds It is recommended that intervention (inspection) be provided selectively, or to be conducted to a specific individual at the discretion of the expert There is sufficient evidence on the undesirable effects of the arbitration, and it is not recommended (not recommended) There is insufficient evidence that the intervention is effective or not, and further study of the effect is needed The degree of confidence in the effectiveness of the intervention is so low that it is judged that the recommendation rating itself is not meaningful RESULTS Fracture risk classification It is important to classify patients according to fracture risk before deciding whether to use fracture-prevention treatments. To predict fracture risk, we decided to use FRAX, which is widely used in many guidelines for GIOP, including the 2017 ACR guidelines. In addition to FRAX, we also considered using history of previous fracture, BMD, and GC dose as predictors. Patients aged 50 years have been the focus of many studies of OP. However, FRAX applies only to patients 40 years old, so we divided our recommendations by adults 40 years for whom FRAX is applicable, and those under 40 years old who are not included in FRAX. Fracture risk was classified as high, moderate, and low (Table 6). For example, the high risk category included patients 40 years of age with previous osteoporotic fracture, BMD T-score 2.5 in men who are 50 years of age or postmenopausal women, FRAX-calculated 10-year major osteoporotic fracture risk 20%, or FRAX-calculated 10-year hip fracture risk 3%. The risk of fracture is correlated with prednisolone dose. For patients taking prednisolone at a dose >7.5 mg/day, FRAX calculates an increase in fracture risk of 15% for major osteoporotic fracture and 20% for hip fracture risk [30]. Table 6. Fracture risk categories in glucocorticoid treated patients Fracture risk Adults 40 years of age Adults <40 years of age High fracture risk Prior osteoporotic fracture(s) Prior osteoporotic fracture(s) Hip or spine bone mineral density T-score 2.5 in men age 50 years and postmenopausal women FRAX (GC-adjusted) 10-year risk of major osteoporotic fracture* 20% FRAX (GC-adjusted) 10-year risk of hip fracture 3% Moderate fracture risk FRAX (GC-adjusted) 10-year risk of major osteoporotic fracture* 10% 19% FRAX (GC-adjusted) 10-year risk of hip fracture >1% and <3% Hip or spine bone mineral density Z score < 3 or rapid bone loss ( 10% at the hip or spine over 1 year) and Continuing GC treatment at 7.5 mg/day for 6 months Low fracture risk FRAX (GC-adjusted) 10-year risk of major osteoporotic fracture* <10% None of above risk factors other than GC treatment FRAX (GC-adjusted) 10-year risk of hip fracture 1% FRAX: https://www.shef.ac.uk/frax/tool.jsp. GC-adjusted: increase the risk generated with FRAX by 1.15 for major osteoporotic fracture and 1.2 for hip fracture if glucocorticoid (GC) treatment is> 7.5 mg/day (e.g., if hip fracture risk is 2.0%, increase to 2.4%). *Major osteoporotic fracture includes fractures of the spine (clinical), hip, wrist, or humerus. Adapted from Buckley et al. Arthritis Care Res (Hoboken) 2017;69:1095-110 [16], with permission of the American College of Rheumatology. 268 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea Fracture risk assessment 1) Initial fracture risk assessment The initial fracture risk assessment should be performed as early as possible in patients with long-term GC treatment. It is appropriate to assess the fracture risk within 6 months of the initiation of long-term GC treatment, and to consider the calculated risks when selecting the specific GC treatment (Figure 2). The most important factors in the initial assessment are the dose, duration, and method of administration of GCs, and history of previous fracture, fall, and frailty. The current nutritional status, weight loss, and the possibility of secondary OP, including thyroid disease, need to be evaluated carefully. In particular, it is necessary to investigate the variables included in FRAX (history of previous fracture, comorbidities, smoking history, alcohol consumption, family history of fracture) in patients 40 years of age. Adjusting FRAX for GC dose and BMD testing are also necessary. If a patient <40 years of age has risk factors for fracture (previous osteoporotic fracture, malnutrition, thyroid disease, weight loss, secondary hyperparathyroidism, hypogonadism, family history of femoral fracture, smoking, alcohol consumption, etc.), it is important to conduct a BMD early in treatment to assess fracture risk further. 2) Reassessment of fracture risk If GCs are used continuously, it is necessary to reassess the risk of fracture every 12 months (Figure 3). For adults 40 years of age who continue GC treatment and are not treated with OP medications beyond calcium and vitamin D, FRAX and BMD should be performed every one to three years. FRAX and BMD are recommended every year if the initial GC dose is prednisolone 30 mg/day, if the cumulative dose is greater than 5 g in the previous year, or if osteoporotic fractures have occurred. For adults 40 years of age who are at high risk of fracture (initial GC dose of prednisolone 30 mg/day, cumulative dose >5 g/year, osteoporotic fracture occurring 12 months after beginning OP medications, poor medication adherence or absorption, or other significant OP risk factors) who are taking OP medications with GC treatment, BMD testing should be completed every two to three years and is recommended as early as possible. BMD testing should be performed at intervals as soon as Figure 2. Initial fracture risk assessment. A clinical fracture risk assessment includes obtaining a history with the details of glucocorticoid (GC) use (dose, duration, pattern of use), an evaluation for falls, fractures, frailty, and other osteoporosis (OP) risk factors (malnutrition, significant weight loss or low body weight, hypogonadism, secondary hyperparathyroidism, thyroid disease, family history of hip fracture, history of alcohol use [at 3 units/day] or smoking) and other clinical comorbidities, and a physical examination including measurement of weight and height (without shoes), testing of muscle strength, and assessment for other clinical findings of undiagnosed fracture (i.e., spinal tenderness, deformity, and reduced space between lower ribs and upper pelvis) as appropriate given the patient s age. The risk of major osteoporotic fracture calculated with the FRAX tool (https://www. shef.ac.uk/frax/tool.jsp) should be increased by 1.15, and the risk of hip fracture by 1.2, if the prednisone dose is >7.5 mg/day (e.g., if the calculated hip fracture risk is 2.0%, increase to 2.4%). It is recognized that in some cases, bone mineral density (BMD) testing may not be available. Adapted from Buckley et al. Arthritis Care Res (Hoboken) 2017;69:1095-110 [16], with permission of the American College of Rheumatology. www.jrd.or.kr 269

So Young Park et al. Figure 3. Reassessment of fracture risk. A clinical fracture risk reassessment includes obtaining a history with the details of glucocorticoid (GC) use (dose, duration, pattern of use), an evaluation for falls, fractures, frailty, and other osteoporosis (OP) risk factors (malnutrition, significant weight loss or low body weight, hypogonadism, secondary hyperparathyroidism, thyroid disease, family history of hip fracture, history of alcohol use [at 3 units/day] or smoking) and other clinical comorbidities, and a physical examination including measurement of weight and height (without shoes), testing of muscle strength, and assessment for other clinical findings of undiagnosed fracture (i.e., spinal tenderness, deformity, and reduced space between lower ribs and upper pelvis) as appropriate given the patient s age. Very highdose GC treatment was defined as treatment with prednisone 30 mg/day and a cumulative dose of >5 g in the past year. Reliability of FRAX (https://www.shef.ac.uk/frax/tool.jsp) after OP treatment is debated, but FRAX calculation can be repeated in adults age 40 years who have not received treatment. It is recognized that in some cases, bone mineral density (BMD) testing may not be available. Adapted from Buckley et al. Arthritis Care Res (Hoboken) 2017;69:1095-110 [16], with permission of the American College of Rheumatology. possible, even after discontinuation of OP medication. For adults <40 years of age who have moderate or high risk and are receiving continuous GC treatment, BMD testing should be conducted every 2 3 years. Treatment and follow-up of GIOP 1) KQ1: Is non-pharmacological treatment beneficial for the prevention and treatment of GIOP in adults taking glucocorticoids? Non-pharmacological treatments such as exercise, good nutrition, smoking cessation, and avoiding alcohol abuse are recommended for all adults taking glucocorticoids. Because there is insufficient evidence for the effects of these treatments in GIOP, it is recommended that treatment be based on data from these treatments in postmenopausal OP patients [III/B]. Because there are limited data on effects of non-pharmacological treatments in the prevention and treatment of GIOP, it is recommended that treatment be based on the established non-pharmacological treatment of postmenopausal OP patients. Although the effects of these lifestyle modifications on fracture risk have not been established for patients with GIOP, non-pharmacological treatments such as weight-bearing exercise, good nutrition, smoking cessation, and avoiding alcohol abuse is recommended for all adults taking GCs [III/B]. 2) KQ2: Which pharmacological treatments are effective for prevention and treatment of GIOP in adults <40 years of age? There are few randomized controlled trials comparing the effects of drugs on prevention of bone mass loss and 270 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea fracture for this age group. This is because adults <40 years of age have relatively high BMD and fewer fractures than postmenopausal women. However, it is reported that long-term use of GCs in premenopausal women <40 years of age may cause changes in bone structure and weaken bone strength [31,32]. Initial pharmacologic treatment for adults <40 years of age is summarized in Figure 4. (1) KQ2-1: Is calcium and vitamin D supplementation effective in the prevention and treatment of GIOP in adults <40 years of age? 1 Calcium and vitamin D are recommended for all adults taking prednisone 2.5 mg/day for 3 months [II/B]. 2 Sufficient calcium (1,000 1,200 mg) and vitamin D (800 IU) intake are recommended, and adequate vitamin D concentrations ( 20 ng/ml) should be maintained [Expert consensus/b]. 3 The use of supplements should be considered when dietary intake of calcium and vitamin D is insufficient [Expert consensus/b]. 4 The minimum adequate dosage should be determined, because high doses of calcium and vitamin D supplementation may increase the risk of gastrointestinal side effects and renal stones [Expert consensus/b]. Figure 4. Initial pharmacologic treatment for adults. Recommended doses of calcium and vitamin D are 1,000 1,200 mg/day and 600 800 IU/day (serum level 20 ng/ml), respectively. Lifestyle modifications include a balanced diet, maintaining weight in the recommended range, smoking cessation, regular weight-bearing and resistance training exercise, and limiting alcohol intake to 1 2 alcoholic beverages/day. Very high-dose glucocorticoid (GC) treatment was defined as treatment with prednisone 30 mg/day and a cumulative dose of >5 g in the past year. The risk of major osteoporotic fracture calculated with the FRAX tool (https://www.shef.ac.uk/frax/tool.jsp) should be increased by 1.15, and the risk of hip fracture by 1.2, if the prednisone dose is >7.5 mg/day (e.g., if the calculated hip fracture risk is 2.0%, increase to 2.4%). It is recognized that in some cases, bone mineral density (BMD) testing may not be available. SERM: selective estrogen receptor modulator. Adapted from Buckley et al. Arthritis Care Res (Hoboken) 2017;69:1095-110 [16], with permission of the American College of Rheumatology. www.jrd.or.kr 271

So Young Park et al. Glucocorticoids reduce intestinal absorption and renal reabsorption of calcium, and increase calcium excretion into the urine [33]. Therefore, calcium and vitamin D are recommended for adults <40 years of age who are taking prednisolone 2.5 mg for 3 months [II/B]. However, calcium alone is not effective in the prevention and treatment of GIOP [34-36], it is more effective to take calcium and vitamin D together. Patients receiving vitamin D3 (cholecalciferol) [37,38] or activated vitamin D (calcitriol, alphacalcidol) [39,40] in combination with calcium showed an inhibitory effect on bone loss compared to calcium alone or no treatment. Adequate intake of calcium (1,000 1,200 mg) and vitamin D (800 IU) are recommended to maintain vitamin D concentration ( 20 ng/ml) [expert consensus/b]. The use of supplements may be considered if the intake of calcium and vitamin D through meals is insufficient [expert consensus/b]. (2) KQ2-2: Is bisphosphonate effective in the prevention and treatment of GIOP in adults <40 years of age? 1 Oral bisphosphonates are recommended for adults <40 years of age with moderate to high risk of fracture [II/A]. 2 If oral bisphosphonates are not appropriate intravenous bisphosphonate is recommended [II/A]. Bisphosphonates are recommended for women without childbearing potential and men <40 years of age at moderate to high fracture risk [II/A]. Alendronate was effective in prevention of bone loss and fractures in premenopausal women with high-dose GC therapy [41]. Subgroup analysis of adults <40 years of age has been done in several previously published studies on GIOP treatment. These studies have shown that bisphosphonates such as alendronate [41-46], risedronate [47-49], zoledronic acid [50] increased BMD and decreased fracture risk compared to placebo or calcium/vitamin D. Oral bisphosphonates, which have a relatively short half-life, are preferred. If oral bisphosphonates are not appropriate, intravenous bisphosphonate is recommended. (3) KQ2-3: Is teriparatide effective in the prevention and treatment of GIOP in adults <40 years old? Teriparatide is recommended for adults <40 years old with moderate to high risk of fracture [II/A]. Teriparatide is recommended for women without childbearing potential and men <40 years of age at moderate to high fracture risk [II/A]. Among patients taking prednisolone 5 mg/day for 3 months, teriparatide treatment led to significantly higher spine BMD than alendronate treatment in both postmenopausal and premenopausal women [51]. (4) KQ2-4: Is denosumab effective in the prevention and treatment of GIOP in adults <40 years of age? Denosumab is recommended for adults <40 years of old with moderate to high risk of fracture [II/A]. Denosumab is recommended for women without childbearing potential and men <40 years of age at moderate to high fracture risk [II/A]. Recently, in a randomized, double-blind, comparative study of denosumab versus risedronate in patients 19 years of age taking prednisolone 7.5 mg/day for 3 months, denosumab significantly increased spine and hip BMD compared to risedronate [52]. 3) KQ3: Which pharmacological treatments are effective for the prevention and treatment of GIOP in adults 40 years of age? Initial pharmacologic treatment for adults 40 years of age is summarized in Figure 4. (1) KQ3-1: Is calcium and vitamin D supplementation effective in the prevention and treatment of GIOP in adults 40 years of age? 1 Calcium and vitamin D are recommended for all adults taking prednisolone 2.5 mg/day for 3 months [II/B]. 2 Sufficient calcium (1,000 1,200 mg) and vitamin D (800 IU) intakes are recommended, and adequate vitamin D concentrations ( 20 ng/ml) should be maintained [Expert consensus/b]. 3 The use of supplements should be considered when the intake of calcium and vitamin D through meals is insufficient [Expert consensus/b]. 4 The adequate dosage should be considered, because high doses of calcium and vitamin D supplementation may increase the risk of gastrointestinal side effects and renal stones [Expert consensus/b]. Recommendations for calcium and vitamin D supplementation in adults 40 years of age are the same as for adults <40 years of age. Calcium and vitamin D are recommended for adults 40 years of age who are taking prednisolone 2.5 mg for 3 months [II/B]. Sufficient calcium (1,000 1,200 mg) and vitamin D (800 IU) intake are recommended, and adequate vitamin D concentrations ( 20 ng/ml) should be maintained [expert con- 272 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea sensus/b]. The use of supplements should be considered when the intake of calcium and vitamin D through meals is insufficient [expert consensus/b]. The adequate dosage should be considered, because high doses of calcium and vitamin D supplementation may increase the risk of gastrointestinal side effects and renal stones [53] [expert consensus/b]. There are concerns about increased risk of cardiovascular disease due to calcium intake, but a recent meta-analysis reported that calcium intake did not increase cardiovascular outcomes and mortality, regardless of the combination of vitamin D agents [54]. (2) KQ3-2: Is bisphosphonate effective in the prevention and treatment of GIOP in adults 40 years of age? 1 Oral bisphosphonates are recommended for adults 40 years of age with moderate to high risk of fracture [I/A]. 2 If oral bisphosphonates are not appropriate, intravenous bisphosphonate is recommended [I/A]. 3 There is no evidence of increased side effects such as atypical femoral fractures or osteonecrosis of the jaw caused by bisphosphonates in patients with GIOP. However, when planning long-term bisphosphonates use in patients with GIOP, the risk-benefit ratio should be considered [Expert consensus/b]. Oral bisphosphonates are recommended for adults 40 years of age with moderate to high risk of fracture. If oral bisphosphonates are not appropriate, intravenous bisphosphonate is recommended [I/A]. Alendronate and risedronate can be used as oral bisphosphonates, and the effects of alendronate [42-46,55,56] and risedronate [47-49,57-59] on the prevention and treatment of GIOP have been demonstrated through several studies. Ibandronate, an oral bisphosphonate, has been shown to increase spine and hip BMD and decrease bone turnover markers compared to placebo in postmenopausal women taking GCs for rheumatic diseases [60]. However, evidence for the use of ibandronate for GIOP prevention is still insufficient. Zoledronic acid was superior to risedronate as an intravenous bisphosphonate for the prevention and treatment of GIOP [50,61]. Bisphosphonates are associated with osteonecrosis of the jaw [62-64] or atypical femoral fracture [65-70], but there is no evidence of increased side effects caused by bisphosphonates in patients with GIOP. However, when planning long-term bisphosphonate use in patients with GIOP, the risk-benefit ratio should be considered [expert consensus/b]. (3) KQ3-3: Is teriparatide effective in the prevention and treatment of GIOP in adults 40 years of age? Teriparatide is recommended for adults 40 years of age with moderate to high risk of fracture [I/A]. Teriparatide is recommended for adults 40 years of age with moderate-to-high risk of fracture [I/A]. When comparing the use of teriparatide with that of estrogen for 12 months in postmenopausal women with GIOP, spine and hip BMD were significantly increased in the teriparatide group, and there was no difference in forearm BMD between the two groups [71]. In addition, the use of teriparatide for 18 months or 36 months led to a significant increase in the spine and hip BMD, and reduction of vertebral fracture risk, compared with the use of alendronate. However, there was no difference in the reduction effect of non-vertebral fracture risk between the two groups [51,72,73]. (4) KQ3-4: Is denosumab effective in the prevention and treatment of GIOP in adults 40 years of age? Denosumab is recommended for adults 40 years of age with moderate to high risk of fracture [I/A]. Denosumab is recommended for adults 40 years of age with moderate to high risk of fracture [I/A]. For patients with rheumatoid arthritis taking methotrexate and prednisolone 2.5 mg/day for 3 months, the effects of denosumab on BMD and bone turnover were compared with placebo. Denosumab treatment increased spine and hip BMD and reduced bone turnover markers for 12 months [74]. Recently, in a randomized, double-blind, comparative study of denosumab and risedronate in patients 19 years of age taking prednisolone 7.5 mg/day for 3 months, denosumab significantly increased spine and femoral BMD compared to risedronate [52]. (5) KQ3-5: Is SERM effective in the prevention and treatment of GIOP in postmenopausal women? In the treatment of GIOP, selective estrogen receptor modulating agents have not been sufficiently proven to be effective in the prevention of fractures. However, if bisphosphonates, teriparatide, and denosumab are not available to postmenopausal women with moderate to high risk of fracture, SERM should be considered [II/B]. In the treatment of GIOP, selective estrogen receptor modulating agent (SERM) has not been sufficiently proven to be effective in the prevention of fractures. However, if bisphosphonates, teriparatide, and denosu- www.jrd.or.kr 273

So Young Park et al. mab are not available in postmenopausal women with moderate to high risk of fracture, SERM should be considered [II/B]. For postmenopausal women taking prednisolone 10 mg/day for 6 months, 12 months of raloxifene treatment significantly increased lumbar spine and hip BMD and decreased bone turnover markers [75]. 4) KQ4: Is it safe to use OP medications in women planning to have a pregnancy? 1 Oral bisphosphonates are preferred for women with moderate to high risk of fracture who are planning to have a pregnancy [Expert consensus/c]. 2 Teriparatide is considered in women with moderate to high risk of fracture who are planning to have a pregnancy [Expert consensus/c]. 3 Because of lack of evidence of fetal safety, intravenous bisphosphonate and denosumab are not recommended [Expert consensus/c]. 4 OP medications except calcium and vitamin D are not recommended for use during pregnancy [Expert consensus/c]. Women who plan to become pregnant should be cautious in the use of OP medication [expert consensus/c]. Oral bisphosphonates are preferred, and teriparatide is also considered for women with moderate to high risk of fracture. However, intravenous bisphosphonate and denosumab are not recommended because of lack of evidence of fetal safety. OP medications except calcium and vitamin D are not recommended for use during pregnancy. When bisphosphonates were used in pregnant rats in vivo, abnormal ossification and calcification of the offspring skeletal system were found at birth [76]. However, in pregnant women exposed to bisphosphonates before or during pregnancy, there were no differences in pregnancy outcomes or birth defects when compared to unexposed pregnant women [77,78]. 5) KQ5: How should response to treatment be monitored in patients with GIOP using physical measurements, imaging, and biochemical methods? 1 Annual BMD and fracture risk assessment using FRAX are recommended in patients taking continuous GCs. The simple spine x-ray examination is recommended to evaluate vertebral fractures [Expert consensus/b]. 2 Regular follow-up is recommended to assess compliance with OP medication [Expert consensus/b]. 3 There is insufficient evidence for biochemical monitoring of treatment response in patients with GIOP [Expert consensus/i]. Annual BMD and fracture risk assessment using FRAX are recommended in patients taking continuous GCs [expert consensus/b]. BMD is measured at the lumbar spine and hip by dual energy X-ray absorptiometry (DXA). A simple spine X-ray is recommended to evaluate vertebral fractures radiographically [expert consensus/b]. Regular follow-up is recommended to assess compliance with OP medications [expert consensus/b]. There is insufficient evidence for biochemical monitoring of treatment response in patients with GIOP [expert consensus/i]. 6) KQ6: Should discontinuation of OP medication be considered if the fracture risk is reassessed to be low during GIOP treatment? When glucocorticoid treatment is discontinued in patients with GIOP and low fracture risk, discontinuation of OP medication may be considered. Discontinuation of OP medication should be individualized, taking into consideration the risk-benefit ratio [Expert consensus/b]. When GC treatment is discontinued in patients with GIOP and the result of fracture risk reassessment is low (low fracture risk), discontinuation of OP medication may be considered [expert consensus/b]. Discontinuation of medication should be individualized, considering the risk-benefit ratio [expert consensus/b]. For adults 40 years of age, when glucocorticoid treatment was discontinued, discontinuation of OP medication was considered if the patient had a follow-up BMD T-score > 2.5, a 10-year risk of major osteoporotic fracture <10%, or a 10-year risk of hip fracture <1% after FRAX adjustment for GCs dose. For adults <40 years of age, when glucocorticoid treatment was discontinued, discontinuation of 274 J Rheum Dis Vol. 25, No. 4, October, 2018

Glucocorticoid-induced Osteoporosis Guideline in Korea OP medication was considered if there were no risk factors such as low BMD (Z score < 3.0), history of previous fracture, and low body weight. When fracture risk is moderate to high, OP medication should be continued even if GCs are discontinued. 7) KQ7: How should initial treatment failure be defined for GIOP? 1 For adults 40 years of age, initial treatment failure is defined as follows: Osteoporotic fractures occur more than 2 times after initiation of oral bisphosphonate treatment or osteoporotic fractures occur or there is significant BMD reduction ( 10 %/year) after 12 months of treatment initiation. Switching to another OP medication is recommended [Expert consensus/b]. 2 When fracture risk reassessment is moderate to high after five years of oral bisphosphonate treatment, active OP treatment is recommended [Expert consensus/b]. If osteoporotic fractures occur more than 2 times after initiation of oral bisphosphonate treatment or osteoporotic fractures occur after 12 months of treatment initiation, or if there is significant BMD reduction ( 10%/year) at follow-up, this is defined as initial treatment failure. Switching to another OP medication is recommended [expert consensus/b]. When fracture risk reassessment is moderate to high after oral bisphosphonate treatment for 5 years in adults 40 years of age, active OP treatment is recommended [expert consensus/b]. In this case, bisphosphonates could be used continuously without a drug holiday, and medications could be changed to intravenous bisphosphonate or other OP medications depending on the patient's drug compliance and treatment response. DISCUSSION The aim of developing this guideline was to promote effective treatment of GIOP by presenting standardized recommendations for prevention and treatment of GIOP for all clinicians treating patients who use or plan to use GCs. The primary implication of this guideline is that all clinicians treating patients with GCs should be aware of the risk of GIOP, identify those at high risk of fracture, and provide appropriate treatment. While OP treatment is currently given primarily to patients whose decreased bone mass was solely determined by BMD, this new guideline recommends that fracture risk should be assessed generally so that patients with moderate to high risk of fracture could receive OP treatment. Considering the pathophysiological characteristics of GIOP, it is clinically important to prevent the reduction of bone mass by glucocorticoids. To develop this guideline, the KSBMR and the KCR jointly formed a development committee and a working committee with experts in the field. These committees used systematic literature searches and adapted previously-published guidelines, following guidelines for the development of CPGs by NECA. This new guideline includes recommendations for the assessment and monitoring of fracture risk, as well as the treatment and prevention of fractures during the period of GC administration for adults 19 years of age. This guideline recommends the use of oral bisphosphonates as first line therapy for GIOP in adults 40 years of age [I/A]. Other guidelines for OP primarily recommend the use of intravenous bisphosphonate, teriparatide, and denosumab for high-risk groups such as GIOP [79]. However, the major published clinical studies show that oral bisphosphonates have the highest evidence level when considering fracture reduction, safety, and cost. This guideline has several limitations and many further studies are needed. First, since domestic clinical studies on GIOP have been scarce, this guideline was developed by adapting guidelines published in other countries. Unfortunately, the guidelines published in other countries also suffer from limited clinical and epidemiological data on GIOP-induced fractures. Clinical trials assessing fracture as a primary outcome in patients taking GCs are especially needed. Notably, both GIOP and non-giop clinical trials show similar effects of OP medications on relative fracture risk. Therefore, results of non-giop clinical trials could be generalized to patients with GIOP. However, these estimates are not accurate in estimating clinical benefits from practical treatment. Second, this guideline accepts treatment criteria of FRAX, used in the United States. According to FRAX, OP treatment should be considered if the risk of major osteoporotic fractures over 10 years is >10% or hip fracture risk over 10 years is >1%. If these treatment criteria are calculated in reverse by FRAX, we can estimate the range of BMD corresponding to treatment targets according to age. For example, FRAX would recommend treatment in the following www.jrd.or.kr 275