To the best of our knowledge, this is the first study to investigate the potential involvement of microarchitectural bone deterioration in individuals affected by FD using TBS. Specifically, our findings demonstrate that subjects with FD display not only a simple reduction in BMD, but also qualitative alterations in bone tissue, indicating broader bone damage; moreover, FD was an independent predictor of low lumbar spine TBS and BMD at bivariate and multivariate analyses.

Bone involvement in patients with FD has long been misrecognized and underestimated and only recently it has gained significant attention. In this regard and in order to possibly mitigate the increased risk of osteoporosis, current consensus recommendations suggest early assessment of vitamin D status starting as early as childhood [2].

Our data confirm that bone is rarely spared in patients with FD, and in our population only 6/14 individuals had a normal bone picture.

However, the most notable feature is that bone involvement appears to be particularly early, given that as many as 3 subjects among those younger than 50 years old (3/7) exhibited a BMD below the expected range for age; on the other hand, the proportion of subjects with osteoporosis and osteopenia was similar between the two groups.

Furthermore, as already reported by some [6, 11] but not all studies [7, 8], bone deterioration seems to be particularly evident at the lumbar spine. Our results are in line with this: indeed, considering the 3 young subjects who were diagnosed with “BMD lower than expected for age”, the sole assessment of the Z-score at the femoral neck or total hip would have led to a misdiagnosis of “BMD within the expected range for age”.

Several possible causes of osteopenia and osteoporosis in FD patients have been proposed. The use of antiepileptic drugs such as carbamazepine is one possible cause of bone loss [8] but no patients in our population were taking them.

Moreover, unlike other studies [7], in our population we did not find a correlation between BMD and renal function. In our cohort, however, eGFR was preserved in all patients, including the woman who had previously undergone renal transplantation, while kidney impairment was mainly represented by persistent albuminuria, present in more than half of the subjects.

Vitamin D insufficiency is highly prevalent among individuals with FD, often due to either limited sun exposure, CKD or intestinal malabsorption [27]. Certainly, heat intolerance is a common complication, frequently attributed to concurrent neuropathy, with an incidence in FD patients ranging up to almost 40% of subjects, potentially elucidating this phenomenon [2, 28]. Similarly, patients with FD may present underlying cardiac disorders that can impair physical exercise and, conversely, facilitate home staying [29]. Finally, the accumulation of Gb3 deposits in the skin has been proposed as a possible interference in vitamin D synthesis [29, 30]. Indeed, compared to controls, subjects with FD were significantly more likely to have 25(OH) vitamin D levels consistent with insufficiency [26], even though two of them were already receiving cholecalciferol supplementation.

FD patients had higher PTH levels than controls, although there were no significant differences in serum calcium, 25(OH) vitamin D and eGFR values; however, it is relevant to note that 11 controls were already taking cholecalciferol supplementation at the time of the evaluation.

Even though statistical significance was not reached in our cohort, it has already been reported in the literature how PTH levels can potentially increase in proteinuria, thus reflecting the progression of tubular damage [31]. Moreover, FD frequently affects the gastrointestinal tract and it has been previously proposed as a potential contributor to the increased risk of malabsorption, which is another well-known cause of secondary hyperparathyroidism [32, 33]. Daily calcium intake can be reliably inferred through 24-hour urinary calcium excretion, but such measurement was not available in our study. Therefore, neither hypo- nor hypercalciuria could be ruled out in our FD population [33, 34].

Despite a higher prevalence of secondary hyperparathyroidism, patients with FD appear to demonstrate greater involvement of trabecular rather than cortical bone. All this considered, it is plausible that the underlying mechanisms of the characteristic bone damage in FD are closely intrinsic to the disease itself. Additionally, no differences in ALP levels were observed between the two groups. Even if data regarding bone-specific alkaline phosphatase (BAP) as well as key markers of bone resorption (such as serum C-terminal telopeptide [CTX]) were not available, this finding may support a rather impaired bone formation in patients affected by FD.

Finally, BMD was negatively correlated with age at all sites except at the lumbar spine: this should not be surprising, as there is ample evidence that spinal degeneration can appear with age, including vertebral deformities and osteophytes, making this measurement less reliable in particular after age 65 [35].

In any case, no data regarding bone damage at the microarchitectural level using TBS in patients with FD were available to date.

As of today, only one other study [36] has evaluated bone quality in patients with several disorders of inborn errors of metabolism (including FD) using the high-resolution peripheral quantitative computed tomography (HR-pQCT) technique. This is an in vivo bone imaging modality that conducts scans of the distal limbs (radius and tibia) and allows for assessment of volumetric bone density and microarchitecture of cortical and trabecular compartments.

In their work, Sidhu et al. [36] reported a tendency towards cortical and trabecular alteration, although less pronounced compared to other genetic conditions primarily affecting bone; however, the analysis was not focused on FD and it was not clear if patients were evaluated upon diagnosis or they had previously been treated with ERT.

Indeed, our study showed that these patients have significantly lower TBS values than controls and that FD itself is an independent risk factor at multivariate analysis associated with bone microarchitecture deterioration.

Due to its accessibility and simplicity, TBS has become one of the most commonly employed non-invasive methods for assessing bone quality in recent years [37]. Moreover, unlike lumbar spine BMD, TBS is significantly less affected by age-related degenerative changes [38, 39], rendering it a suitable tool for the evaluation of trabecular bone microarchitecture even in the elderly population.

Previous studies in postmenopausal women have already shown that TBS is able to predict the occurrence of fragility fractures regardless of BMD [40, 41]. Furthermore, the evaluation of TBS assumes particular importance in instances of secondary osteoporosis, in which increased fracture risk is often associated with changes in microarchitecture and bone quality rather than depending on reduced BMD [17].

At the bivariate analysis TBS was inversely correlated with PTH and ACR in patients with FD. Data about microarchitecture bone alteration in CKD have already been reported [42, 43], but mainly in individuals with reduced eGFR, while data about moderately and severely increased albuminuria and/or proteinuria are scanty. Anyway, TBS values in patients with FD were shown to be significantly reduced compared to controls even after removing from the analysis patients with CKD, thus limiting the involvement of this comorbidity.

At the multivariate analysis, indeed, TBS was shown to be positively correlated with 25(OH) vitamin D levels and lumbar spine BMD and negatively correlated with BMI. This was not an unexpected finding, as a positive correlation with both 25(OH) vitamin D levels and BMD had already been shown [44, 45]. Likewise, inverse relationship between TBS and BMI has already been reported in the literature [46]: in this regard obesity may have a negative impact on bone quality potentially due to chronic low-grade inflammation [47].

Despite patient matching, a nearly significant trend of lower BMI values in patients with FD compared to controls was observed in our cohort. As previously reported, in fact, FD often involves the gastrointestinal tract and can lead to reduced BMI as a consequence of increased risk of malnutrition [32]. Considering the small sample size of our study, it is even possible that the higher BMI evident in the control group may have partly “reduced” the evidence of degraded TBS in subjects with FD, given the correlation just mentioned [44, 46].

Of note, no correlation between lyso-Gb3 levels at diagnosis and either TBS or BMD was observed, in contrast to the findings of a recent Japanese study which suggested a possible correlation with BMD at the lumbar and femoral neck sites, at least in males [11]. On the other hand, the evidence regarding lyso-Gb3 as a measure of clinical severity and as an indicator of the degree of target organ involvement with prognostic significance, still requires further investigation [48]. Therefore, our present data do not support the role of lyso-Gb3 in this context, at least regarding bone health status.

Our study presents some limitations. Firstly, the population was rather small, and no fractures were detected in patients with FD: this considered, it was not possible to evaluate the potential correlation between TBS and BMD values and the fracture risk, nor to perform sub analyses in subjects stratified for BMD categories. In addition, data on major markers of bone formation (such as BAP) as well as markers of bone resorption (such as CTX) were not available, although no differences were evidenced regarding ALP levels.

Finally, only data at baseline at the time of FD diagnosis were evaluated, so any potential beneficial effect of ERT or chaperone treatment on TBS and long-term bone quality, although intriguing, remains yet fully to be demonstrated. In fact, the paper by Nose and colleagues which suggested an improvement in BMD correlated with lyso-Gb3 reduction following the initiation of ERT [11] is severely limited in its actual clinical meaning, considering the very small sample of patients. On the other hand, the recent study by Aitken et al. showed that subjects with FD treated with medical therapy had worse bone density trajectories compared to those who were not receiving any specific treatment. This data suggests that Fabry-specific therapies may not be able to reverse all disease mechanisms, such as those responsible for the underlying bone damage [12].