A Delphi consensus on the management of Spanish patients with osteoporosis at high risk of fracture: OSARIDELPHI study

Osteoporosis is a skeletal disease characterized by decreased bone mass and deterioration of bone microarchitecture, leading to increased fragility and consequent fracture risk [1]. It is the most common metabolic bone disease in Western countries, affecting 25–32% of Spanish women over 50 years and almost 50% over 75 years [2]. In the absence of more recent updates on disease prevalence data, in 2010, it was estimated that in Spain, there were around 1.9 million cases in women and 400,000 cases in men, with an estimated cost of 2.8% of total health expenses [3, 4]. The incidence has increased in the following years, as well as the direct costs of incident fractures. Thus, in 2019, it was calculated that in Spain, there were around 2.9 million cases, of which almost 80% were women, with an estimated cost of 3.8% of total healthcare expending [5, 6].

Fragility fractures are a significant cause of disability, morbidity, and mortality in the population [7]. Hence the importance of prevention and early diagnosis, especially in patients at high risk and very high risk of fracture. A previous fracture due to this pathology, especially recent in time, is considered one of the risk factors identified for fragility fractures [8‐12]. However, patients perceive them to be due to the environment or accidental falls. They often do not consider that they should be screened, observe preventive strategies, or even receive therapy for osteoporosis [13]. Recently, different studies analyzed the risks of imminent fractures in various cohorts in Spain [14, 15]. These studies showed that higher risks occur in women aged ≥ 80 years, and the 10-year risk levels were estimated between 1.8 and 21.5% in women and between 0.7 and 10.8% in men using the QFRACTURE tool [14]. More recently, an estimate of the incidence ratio of subsequent fractures in the 3 years following an initial fracture observed that 3.2% of women patients ≥ 50 years with a previous fracture experienced a new fracture every following year [15]. Despite all these data, underdiagnosis and under-treatment of osteoporosis are quite common, as is the case of vertebral fractures, which are the most frequent [16‐18]. Vertebral fractures are often asymptomatic or mildly painful, and radiographic detection routines are scarce [17].

Optimal management of this patients’ profile has also impacted the quality of life of people with osteoporosis. Due to the presence of comorbidities, they are often polymedicated, usually associated with lower adherence to treatment, in addition to the dosage of the drugs and the side effects of the different therapies [19]. Indeed, patient follow-up program strategies are critical for those at increased risk of fracture to improve long-term treatment persistence and prevent secondary fractures [20‐22].

Consequently, it is relevant to establish consensus among specialists in managing patients with osteoporosis to have effective strategies for prevention and diagnosis. It is also important to adapt the best therapy to each patient profile and to optimize the different lines of treatment over time, from the early stages of the disease. It is essential to consider the current availability of therapies, particularly biological treatments, in patients at high risk of fracture. In this sense, updated guidelines and recommendations from a number of organizations, including European Society for Clinical and Economic Aspects of Osteoporosis (ESCEO) and the International Osteoporosis Foundation (IOF) are available [23‐25]. However, these recommendations need to be adapted locally, taking into account the characteristics of the population and the specific health resources at regional level.

The results of this Delphi study demonstrate a high degree of consensus between experts involved in the management of osteoporosis patients with a high and very high–risk of fracture. This expert panel study provides insights on important topics such as the factors associated with more probability of high and very high–risk of fracture, the diagnosis, and the treatment choice for the specific patient profile.

Overall, according to the Delphi survey, experts agreed on the categorization of the fracture risk due to osteoporosis is determined by the history of previous osteoporotic fractures, a recent major fragility fracture, the family history of hip fracture, the treatment with high-dose glucocorticoids, the 10-year fracture risk probability, and the data of the bone densitometry. The IOF and ESCEO recommended that the risk of fracture should be expressed as absolute risk, i.e., the probability of fracture over a 10-year interval [30]. The assessment strategy to categorize risk was recently improved, including low and high risk and very high risk [23, 24]. Very high risk was defined as a fracture probability above the upper assessment threshold after a FRAX® assessment, including bone mineral density (BMD) if available [23, 24]. The panel of experts of the present Delphi survey also agreed that the implementation of secondary prevention actions is a consequence of this categorization. This result is in line with previous recommendations suggesting that preventive treatment given as soon as possible after fracturing in patients with very high fracture risk would avoid a higher number of new fractures and reduce the attendant morbidity [23, 24].

The panelists recommended follow-up systematically and proactively to the patients at high and very high risk of fracture (agreed in round 1 by 100% of the experts). In addition, experts, besides age and female gender, recommended including in the assessment the following risk factors: personal and parental history of fragility fractures (97.5%); history of treatment with glucocorticoids at doses ≥ 5 mg/day for ≥ 3 months (98.8%); causes of secondary osteoporosis (98.8%); history of falls in the past 3 months (92.5%), radiological imaging tests (92.5%); and BMD (97.5%). Another aspect related to risk assessment and categorization was recently highlighted as a new pivotal point in osteoporosis: FRAX® arithmetically integrating with novel risk factors [25, 31]. Accordingly, it is widely recognized that BMD alone for fracture risk assessment is less sensitive than algorithms, such as FRAX®, which incorporate risk indicators in addition to BMD [31].

The experts consistently agreed that treatment with bone-forming drugs should be considered in the very high fracture risk category. A sequential change to antiresorptive medications should be assessed after 1 or 2 years (97.5%). In addition, panelists recommended that the monitoring criteria for the different available alternatives of treatment should be defined by the improvement in bone densitometry (87.5%) and bone remodeling markers (78.8%), the absence of new fractures (93.8%), and the decrease in the risk of further fractures (95%). Regarding long-term treatment, panelists agreed that the duration of treatment with oral bisphosphonates, intravenous bisphosphonates, teriparatide, and denosumab is currently well defined at 5 (71.8%), 3 (69.2%), 2 (94.9%), and 10 years (71.8%).

Although osteoporosis is a well-recognized problem with a choice of widely available treatments, a large treatment gap exists [32]. Some treatments for osteoporosis (i.e., oral bisphosphonates, menopausal hormonal therapy, and selective estrogen receptor modulators) have suboptimal efficacy due to the difficulties in meeting treatment goals with such therapies in the highest fracture-risk patients [33]. The treatment stratification according to baseline fracture risk permits targeting the most effective treatments for this patient profile [34]. Specifically, guidance thresholds leading to distinguish high and very high fracture risk have optimized the use of anabolic agents [32, 34]. In patients at the highest fracture risk, treatment initiation with an anabolic (bone-forming) agent, such as teriparatide, followed by an antiresorptive to maintain the gains in bone mineral density, appears now a highly appropriate strategy to achieve a rapid and sustained reduction in fracture risk [8, 30]. In this sense, the availability of romosozumab and abaloparatide would be after the completion of the present Delphi study (conducted between July 2021 and March 2022). This recommendation has strong evidential support from recently published studies comparing anabolic vs. antiresorptive therapies [35‐38]. These studies demonstrated a more rapid and significant fracture risk reduction with the former, compared with oral antiresorptive treatments alone [35‐38]. However, these benefits must be maintained by following the anabolic with an antiresorptive drug [36, 39, 40]. Indeed, the evidence suggests that the treatment sequence is important, such that an anabolic agent given before an antiresorptive agent is more effective than the opposite sequence [41, 42]. This model suggests the need for physicians to be able to identify the patients who would most benefit from anabolic therapy [32].

The decision on when to stop antiresorptive treatment in a patient who has received a prior anabolic agent is complex. Recent data showed that discontinuation of up to a year might be acceptable in the case of previously alendronate- and zoledronate-treated patients [43]. In contrast, in the other agents (risedronate, ibandronate, raloxifene, teriparatide, denosumab, and romosozumab), the bone loss at the femoral neck and total hip is higher, indicating the importance of continuation of these antiresorptive agents [43]. Considering all these data, the consensus is that prolonged antiresorptive therapy will be necessary after anabolic treatment if the patient remains at high or very high fracture risk. However, if a patient is no longer at high or very high fracture risk, it may be possible to stop treatment for a maximum of 2 years, except for denosumab, due to the risk of rebound vertebral fractures [43].

This Delphi study also highlighted that experts agreed on the importance of the existence of fracture-liaison service (FLS) units in managing patients with high and very high fracture risk. According to experts, these FLS units should be composed of professionals who cover the different aspects crucial in managing this patient profile and with varying levels of specialization. However, there are differences in consensus regarding the current use of FLS (71.8% of the panelists stated that FLS units currently manage this patient profile) and what should happen (97.4% of the experts agreed that an FLS unit should monitor these patients). Notably, implementing FLS has increasing evidence since these units can improve access to better management and treatment to reduce future fractures [44‐46]. A recent systematic review and meta-analysis evaluated the clinical impact of FLS implementation based on the results of several studies encompassing 48,045 patients [46]. The results suggested that FLS significantly improves the rates of dual-energy X-ray absorptiometry (DXA) scanning and antiresorptive therapy prescription and reduces new fracture rates [46]. Regarding possible controversies, a lack of consensus was observed on the current clinical application of bone remodeling markers in the early diagnosis of the disease (20.5%) and the prediction of fracture risk and bone mass loss (42.3%). The lack of consensus was also reflected in the simultaneous combination of a bone-forming drug with an antiresorptive treatment (50%). These results at the time of the Delphi survey reflect the absence of solid recommendations regarding using bone remodeling markers as a current diagnostic tool and of scientific evidence concerning the simultaneous combination of available therapies.

One of the limitations of the present study was a greater representation of specialists in rheumatology or traumatology among the participating experts compared to the rest of the specialties responsible for the management of patients with osteoporosis. Probably, the fact that there was no participation of primary care physicians explains the high percentage of experts in whose centers there were multidisciplinary units and FLS. In addition, it should be taken into account that the Delphi methodology itself means that the fact that experts express a high degree of agreement does not directly imply that a recommendation is necessarily effective. The results of the study represent the starting point for the development of recommendation documents and management guidelines. In summary, the results of this Delphi study suggested that, although the general lines of recommendations and suggestions are in line with the management recommendations established by, among others, the IOF or the ESCEO, an adaptation to the specific characteristics of the Spanish population and available health resources is needed. Specifically, regional differences can be observed in the implementation of FLS units, which may determine an adaptation of the recommendations to the Spanish reality.

The expert consensus recommendations derived from this Delphi panel study may provide support and guidance on clinical decision-making regarding osteoporosis management in specific clinical patients’ profiles, specifically in real-world patients with high or very high fracture risk. Additionally, this consensus analysis may encourage discussion on controversial issues addressed in the consensus statements.