Answers are grounded in this paper and cite their sources.
Accumulative evidence during the last decade has established an association between haemophilia (A and B) and low bone mineral density (BMD), both in adults and children 1-3. Despite the heterogeneity among studies, regarding sample size, population and definition criteria of low BMD, haemophilia leads to bone loss due to a variety of mechanisms. In brief, repeated bleeding episodes and concomitant arthropathy from childhood (‘haemophilic arthropathy’), lead to impaired mobility and avoidance of regular (mostly weight-bearing) exercise, which compromises optimal acquisition of peak bone mass 1-4. Other factors also contribute to bone loss to a various extent, such as hepatitis C (HCV) and human immunodeficiency virus (HIV) infections, level of physical activity, 25-hydroxy-vitamin D concentrations and the degree of haemophilic arthropathy 1, 5-7. Nonetheless, what matters most is whether these patients are indeed at increased risk of fracture, taking into account the increased morbidity and the difficulties in their fracture management. The purpose of this editorial is to focus on fracture risk in patients with haemophilia (PWH), regarding prevalence, potential pathogenetic mechanisms and future perspectives. Despite the evidence on the association between haemophilia and low BMD, data on fractures are not adequate to draw firm conclusions and the exact magnitude of the problem is not known. No prospective study exists at the moment on this concept. Data mostly emerge from cross-sectional or retrospective studies primarily designed to report prevalence and risk factors for low BMD in this population. Nevertheless, the prevalence of fractures in these studies ranges from 4% to 37% 1, 5-10. A recent retrospective comparative study including 316 patients with haemophilia A (mean age 20 years) and 66 with haemophilia B (mean age 26 years) showed a significantly higher incidence of fractures in PWH compared with the general population (24.8 cases per 1000 patient-years vs. 9.6 fractures per 1000 patient-years) yielding a relative risk (RR) of 10.7 [95% confidence interval (CI): 8.2–10.4] 9. Others have reported an even higher estimated RR for fracture (21.4–24) 5. For older populations (i.e. around 50 years), it can be speculated that fracture risk is also higher in PWH compared with that of the general population. Indeed, using data from the general (Greek) population, the prevalence of hip or clinical vertebral fracture in men of 50–54 years old is about 0.1% 11, which contrasts to the rates of 16–18% seen in PWH of the same age and region 1. In children with haemophilia, conflicting data exist. In general, fractures occur as a result of major trauma due to the childhood (mainly sport) activities. Of course, since haemophilia is diagnosed, these intense activities are significantly reduced due to the fear of repeated bleeding and fractures. This was common in the past, when prophylactic treatment was not available. However, current management promotes physical activity and children are encouraged to take part in sport. It is not known if children with haemophilia are indeed at increased risk of fracture, either pathological (after a trivial trauma) or under intense conditions. In a cohort of 500 PWH, 22 fractures were recorded. Four of them had occurred in children with severe haemophilia and all were the consequence of a trivial fall, with three localized at the femur. However, there was no control group in this study 8. A Finnish study did not find any difference in fracture rates between children with severe haemophilia and healthy age-matched controls, although BMD was lower and urinary calcium excretion was higher in PWH 12. The average age at the time of sustaining a fracture is 28–30 years in PWH 1, 8, 13. Most common fracture sites in PWH are the lower limbs and, in particular, femur (mainly shaft and less frequently supracondylar and neck sites) and, to a lesser extent, tibia 5, 8. In other studies, upper limbs (wrist, elbow, radius, fibula) were the most common sites 1, 14. A recent retrospective study showed a shift in fracture prevalence from the lower to the upper limbs, as the most common skeletal site, during the last decade (63% compared with 24–30% during the period before 2002). The mean age was 32 years showing also a decrease during the last decade (from 35 to 26 years) 15. It is interesting that in most cases, fractures in PWH occur after minor trauma, such as a fall from a standing position, which are not common in the general population of these ages 1, 8. Interestingly, this was also reported for femoral neck fractures 13. Stress fractures (i.e. occurring in the long bones of lower limbs after repetitive low-grade forces) have also been reported in PWH, due to overuse. They are mostly seen in athletes, such as long distance runners. However, their exact incidence in PWH is not known 16. Bone mineral density in PWH having sustained fractures is usually low 1, 5, 8. In one study, Z scores in lumbar spine ranged from −2.5 to −4.9, whereas Z scores in femoral neck ranged from −1.4 to −3.8 8. Z scores are used for men younger than 50 years and values lower than −2 are considered as ‘below the expected range for age’ 1. Low lumbar spine BMD seems to be prevalent in PWH with a fracture history, although vertebral fractures are uncommon 1, 5, reflecting perhaps deficits in trabecular bone in these patients. Interestingly, a recent study showed that both low trabecular and cortical BMD exist in PWH 17. These microarchitectural deficits translate into a relatively low biomechanical bone strength and may in part account for the increased fracture risk in these patients, irrespectively of BMD 17. However, a study in children with haemophilia did not show differences in trabecular or cortical volumetric BMD compared with the age-specific reference values, after adjustment for patients’ shorter body height 14. It can be speculated that the trabecular and cortical bone deficits may occur later in life, perhaps under the influence of other co-factors. It seems that BMD is not the sole factor contributing to increased fracture risk in PWH. In one study, there was no difference in the prevalence of fractures between those with low and normal BMD 1, reflecting that other factors, except for BMD, may account for increased fracture risk. The number of target joints and the level of physical activity are positively associated with fractures 1. Fracture risk also increases with age and the presence of inhibitors 9. Conflicting data exist with respect to the association of HCV infection and fracture risk 1, 9. One study found that children with haemophilia had decreased muscle area in the forearm, as compared with their age-specific reference values 14. Although the term ‘sarcopenia’ was used without a strict definition, this study is of outmost importance because if it is confirmed in adult studies, it may further elucidate the spectrum of fracture risk in PWH 18. Moreover, it should be underscored that the same factors accounting for low BMD (physical inactivity, vitamin D deficiency, haemophilic arthropathy) also predispose PWH to increased risk for falls and, therefore, to increased fracture risk, creating a vicious cycle 19, 20. One-third of PWH sustains a fall in the past year and 42% of them more than once (with an average of four falls) 19. Risk factors predisposing to falls are the living conditions, physical inactivity, urinary incontinence, history of a previous fall, orthopaedic status, pain with daily activities, avoidance of pain-related exercise, depression and HIV infection 19, 20. It is not clear if haemophilia per se contributes to increased fracture risk or via its associated complications. The severity of haemophilia seems to be positively associated with fractures 1, 9. In one study, the risk of fracture in patients with mild to moderate disease was significantly lower compared with those with severe disease (RR: 0.56, CI: 0.32–0.97) 9. Thus, a direct role of haemophilia on fracture risk cannot be excluded. This is also supported by data from recent experimental studies, showing that factor VIII knockout mice demonstrate significantly decreased bone mass and bone strength, without the presence of haemarthroses 21. The mechanism leading to this skeletal impairment is increased rate of bone resorption 22. It is evident that what matters most is if an individual is indeed at increased risk of fracture and not just if she/he has simply low BMD (osteoporosis, osteopenia or Z score below than the expected range for age). On this concept, a more comprehensive approach should be developed when the clinician evaluates the individual's fracture risk. Independent risk factors, such as age, body mass index, smoking, alcohol, corticosteroids, rheumatoid arthritis, history of previous fracture and parental hip fracture, causes of secondary osteoporosis, should be taken into account. These factors have been incorporated in a mathematical model, the ‘FRAX’ tool, which estimates the 10-year risk for major osteoporotic and hip fractures 23. Specific conditions are included under the term ‘secondary osteoporosis’, such as hypogonadism, type 1 diabetes mellitus, osteogenesis imperfecta, untreated long-standing hyperthyroidism, chronic malnutrition/malabsorption and chronic liver disease, since these have been associated with increased fracture risk irrespectively of BMD 23. Under these circumstances, current evidence is not robust for haemophilia to be incorporated into the list of risk factors included in the FRAX tool, either as a separate factor or, at least, as a cause of ‘secondary osteoporosis’. However, FRAX can be used in adults with haemophilia older than 40 years for better assessing fracture risk and deciding whether to commence pharmacological therapy or not. It can be empirically proposed that the calculated 10-year risk would be higher in PWH. One study exists at the moment that has used the FRAX tool in PWH. It showed that the calculated 10-year risk for major osteoporotic and hip fractures in PWH with low BMD was 3.2 ± 1.0% and 0.9 ± 0.8% respectively, whereas in those with normal BMD, it was significantly lower (2.2 ± 0.3% and 0.2 ± 0.1% respectively). The relatively low 10-year risk may be mainly attributed to the patient's younger age (45 years) 24. Although data are not conclusive, it seems that PWH may be at increased risk of fracture, especially those with severe disease. It seems that haemophilic arthropathy, physical activity, age and the presence of inhibitors contribute to increased fracture risk. A direct role of FVIII deficiency has also been proposed. Furthermore, some PWH may be at increased risk of falls which also predisposes to fractures and creates a vicious cycle. The FRAX model can be used for estimating fracture risk in PWH older than 40 years. However, current data cannot support the incorporation of haemophilia in the FRAX model (under the term ‘secondary osteoporosis’). Thus, what we need is prospective, multicentre trials, evaluating fracture incidence in PWH compared with age-matched healthy male controls in order to investigate whether an independent association between haemophilia and fracture risk exists. The authors stated that they had no interests which might be perceived as posing a conflict or bias.
No takes yet. Share an insight, caveat, or question.
Anagnostis et al. (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: