Annals of Internal Medicine | 2019
National Institutes of Health Pathways to Prevention Workshop: Research Gaps for Long-Term Drug Therapies for Osteoporotic Fracture Prevention
Abstract
Advances in understanding the biology of osteoporosis have resulted in several medications that have been demonstrated to reduce fracture risk (1, 2). However, access to screening, medication costs, and patient concerns about rare adverse effects of antiresorptive treatments have coincided with decreased prescribing of osteoporosis drug therapy (ODT) and a leveling off in the incidence of osteoporotic hip fractures (35). Osteoporosis is a skeletal disorder that compromises bone strength (6) and increases the likelihood of fractures. Loss of bone mass is associated with aging; hence, osteoporosis primarily affects older people (7). Among U.S. adults older than 50 years, it is estimated that 8 million women and 2 million men have osteoporosis (8), and 27 million women and 16 million men have low bone mass (8). It is projected that by 2025, five fractures will occur for every 100 people older than 65 years, and total U.S. health care costs attributable to fractures will reach $25 billion annually (7). Nonpharmacologic approaches to manage osteoporosis, including adequate calcium and vitamin D intake and physical activity, can positively affect bone mass. Coupled with preventing falls and limiting modifiable risk factors, such as smoking and alcohol use, these measures can help reduce a person s risk for osteoporotic fractures (9). Furthermore, pharmacologic treatments may be prescribed to prevent fractures for people who have very low bone mineral density (BMD) or a prior fragility fracture, and the U.S. Food and Drug Administration has approved antiresorptive treatments that inhibit bone resorption and anabolic treatments that stimulate bone formation. Nevertheless, many people at high fracture risk remain untreated. Less than 20% of women received osteoporosis treatment in the year after diagnosis of an initial fragility fracture, and compliance rates are low (7). On 30 and 31 October 2018, the National Institutes of Health (NIH) convened the Pathways to Prevention (P2P) Workshop: Appropriate Use of Drug Therapies for Osteoporotic Fracture Prevention to assess the available scientific evidence to better understand the clinical benefits and harms of ODTs, with the aim of identifying research opportunities to fill gaps in our understanding of long-term treatment of osteoporosis. The workshop, cosponsored by the NIH Office of Disease Prevention (ODP), the National Institute of Arthritis and Musculoskeletal and Skin Diseases, and the National Institute on Aging, brought together osteoporosis experts and advocates. The ODP commissioned a systematic evidence review on long-term (>3 years) ODT use and fractures, and other researchers were invited to present evidence. Intermediate outcomes, such as BMD, and the more established side effects and adverse outcomes were discussed but were not the major focus. This report focuses on research gaps for long-term (>3 years) ODT use. To complement the commissioned review, this report relied on other systematic reviews and studies presented by speakers. Evidence Regarding Benefits and Harms of Long-Term ODTs Although varied in treatment regimens and outcomes assessment, trials have shown the safe and effective use of ODTs with 3 to 5 years of exposure in reducing the incidence of vertebral fractures in postmenopausal white women (1). Shorter-term ODT exposure was reviewed previously in the context of an earlier systematic review (2). Studies (2, 1012) have shown that some ODTs reduce the incidence of nonvertebral, including hip, fractures. The workshop participants noted the importance of fracture on patient morbidity and survival; however, the trial results presented no data on nonfracture patient outcomes or sequelae of fractures, such as functional status, mobility, hospitalizations, and nursing home placement. There was limited or no evidence on whether patient characteristics would result in different fracture outcomes with these treatments (referred to as effect modification hereafter), because individual studies were underpowered to detect changes in outcomes or not large enough to maintain adequate type I error rates for multiple comparisons. Many of the side effects and possible adverse outcomes associated with ODT are documented elsewhere. The workshop presented rare and serious complications thought to be specific to selected antiresorptive therapies: atypical femoral fractures (AFFs) and osteonecrosis of the jaw (ONJ). These complications have been the subject of advisories from regulators and professional societies. Evidence of the incidence of these complications is limited owing to their rarity and lack of studies designed to systematically identify and ascertain these adverse events. These events have been variably defined; some studies report (13) on subtrochanteric and femoral shaft fractures that may not have had all the radiographic characteristics of AFFs. Evidence from trials is nonexistent or limited on AFF and ONJ for antiresorptive ODTs (14, 15), and these adverse effects have not been associated with anabolic ODTs. The best available data, despite their limitations, come from observational studies and postmarketing surveillance. This work suggests that the age-adjusted incidence rates for AFFs were 1.78 per 100000 person-years (95% CI, 1.5 to 2.0 per 100000 person-years) with bisphosphonate exposure of 0.1 to 1.9 years, and increased to 113.1 per 100000 person-years (CI, 69.3 to 156.8 per 100000 persons per year) with bisphosphonate exposure of 8 to 9.9 years (16). For ONJ, the incidence is 1 to 69 per 100000 patient-years with oral bisphosphonates and 0 to 90 per 100000 patient-years with intravenous bisphosphonates. These incidence rates are marginally higher than those of the general patient population (1, 2, 13, 17). The harms of estrogen treatment and estrogen plus progestin treatment should be considered (strokes, invasive breast cancer, pulmonary embolism, and dementia) because they exceed benefits in most postmenopausal women. Effect modification suggested short-term use of estrogen to reduce the risk for fractures remains in consideration for those with postmenopausal symptoms or younger women who have had a hysterectomy. The American College of Obstetricians and Gynecologists and the American College of Physicians have listed several contraindications to the use of estrogens for osteoporosis (11, 18). Several trials have examined the safety and effectiveness of anabolic agents. However, they were considered out of scope for the systematic review because they are limited to a 2-year lifetime exposure and have a boxed warning based on animal toxicology studies showing an increased risk for osteosarcoma with long-term, high-dose administration. Current Gaps in Knowledge Notably, in the case of both benefits and harms, trials provide evidence mainly for white postmenopausal women, whereas other populations (for example, men, spectrum of race and ethnicities, residents in facilities, and people with advanced and multiple comorbid conditions) were absent or underrepresented. Evidence for ODTs is lacking for people who meet neither BMD nor fracture criteria for osteoporosis but are at high risk owing to other health, genetic, or medication use factors. Given that most trial participants did not represent the true potential patient population, estimates on benefits and harms may differ in actual practice. Few trials extended beyond 5 years, but several observational studies provided limited evidence on potential benefits and harms from longer-term use (1, 10). Evidence is lacking for nonfracture patient outcomes and fracture sequelae that patients may prioritize when making treatment decisions. Gaps exist in how to use information on bone biomarkers and other patient characteristics, such as concurrent medication use, that may modify the effects of ODT on the risk for fractures and their sequelae. Inability to rigorously estimate effect modification for subpopulations was in part because of limitations in trials designed primarily for regulatory approval. Although there have been many participants in ODT trials, analyses that pool patient-level data are limited to initial work on bone turnover markers and BMD. This approach could be potentially useful for addressing some of the gaps in knowledge of effect modification by increasing power to detect differences in outcomes by a given marker or BMD value. Drug Holidays Uncertainty regarding the long-term effects of ODTs has heightened interest in periods of medication discontinuation, or drug holidays, as a means of minimizing potential harms. Drug holidays for bisphosphonates are of primary interest, because evidence suggests that accumulation of bisphosphonates in bone may impede normal remodeling and repair and potentially predispose to ONJ and AFF. Similar concerns exist regarding risk for ONJ and AFF with denosumab therapy. Bisphosphonate Drug Holidays Evidence gaps limit evaluation of potential harms and benefits of drug holidays for antiresorptive medications. The evidence comes from a limited number of efficacy trials and their extensions, and findings are mixed with regard to risk for incident fractures associated with placebo versus continued bisphosphonate use. Discontinuation of alendronate for 5 years was associated with greater risk for incident vertebral fracture but not nonvertebral or hip fracture (19). Similarly, discontinuation of zoledronate was associated in one study with approximately 50% greater risk for vertebral fracture (16). For both alendronate and zoledronate, discontinuation was associated with small (1% to 3%) but statistically significant greater decreases in BMD compared with continued use (1921). Evidence is insufficient from trials to determine whether bisphosphonate drug holidays reduce risk for rare events of ONJ, AFF, or other adverse events. Observational studies suggest that risk for incident AFF may decrease after initiation of drug holidays (22, 23). In additio