Prognostic value of trabecular bone score for major osteoporotic fractures in patients with radiographic axial spondyloarthritis: a 3-year prospective cohort study

Osteoporosis (OP) is characterized by reduced bone mineral density (BMD) and impaired bone microarchitecture [1]. This increases the risk of fractures, which can occur even without significant force during everyday activities.

The dual-energy X-Ray absorptiometry (DXA) scan has been considered the gold standard in OP diagnostics for many years. The scan assesses the risk of osteoporotic fractures based on areal BMD measurement. Under certain circumstances, the discriminatory power of DXA testing is limited, for example in diabetes mellitus (DM), chronic kidney disease related metabolic bone disease (CKD-MBD), glucocorticoid (GCS) therapy or in radiographic axial spondyloarthritis (r-axSpA), formerly known as ankylosing spondylitis (AS) [2‐5].

In r-axSpA, chronic inflammation leads to proliferation of bone tissue, resulting in the formation of syndesmophytes on the spine and osseous bridging of the sacroiliac joints [6, 7]. While bone mass around the vertebrae increases, the inner cancellous bone can be severely impaired. The total spine BMD as measured by DXA may appear normal or even increased, while the biomechanical properties are significantly reduced [4]. Extending the DXA examination to include a scan of the femoral neck is not optimal in this instance. The highest incidence of r-axSpA is in the third decade of life, and due to slower bone turnover, BMD changes within the femoral neck are usually only observable at an older age [8]. Quantitative computer tomography could be a better diagnostic option distinguishing between cortical and trabecular bone, but this technique is used less frequently and needs substantially higher doses of ionizing radiation, which is relevant in relatively young patients and multiple follow-up examinations.

The prevalence of vertebral fractures in patients with r-axSpA varies across different studies. Some reports indicate rates of up to20-30% [9‐11], which may be due to specific characteristics of the study populations. Nevertheless, it can be assumed that the proportion of patients with vertebral fractures is around 10–20% [4, 12‐14]. This is a significant number compared to the general population, especially when considering the average age of patients with r-axSpA.

For these reasons, risk fracture assessment in r-axSpA patients should be supplemented by an evaluation of bone microarchitecture. DXA scans can be used to determine the trabecular bone score (TBS).

TBS is a numerical method in which the assessment of bone microarchitecture is based on the same raw data used to calculate BMD [1]. TBS is the grey scale distribution across the individual pixels that make up the image of the spine. TBS correlates with quantitative computer tomography which provides similar information about cancellous bone [15].

The association between TBS and fracture risk in patients with r-axSpA is now well established [4, 9]. Nevertheless, there is still a lack of clear recommendations on the clinical use of TBS in the assessment and management of OP in r-axSpA, particularly in patients with normal or osteopenic BMD values. The current International Society for Clinical Densitometry (ISCD) and the International Osteoporosis Foundation (IOF) 2023 guidelines do not provide clear recommendations regarding the management of bone health in patients with r-axSpA [16, 17].

In the case of OP, patients must be monitored over longer periods of time in order to assess the benefits of TBS in evaluating fracture risk, and larger study populations are required. In patients with r-axSpA, the challenge lies not only in the duration of follow-up but also in recruiting a sufficiently large cohort. It is noteworthy that in the Manitoba BMD Registry study, which included data from over 65,000 individuals, only 188 patients with r-axSpA were identified [9, 18]. This highlights the challenges associated with both the recruitment of adequate sample sizes and the long-term follow-up required for meaningful clinical research in this population.

This study focuses specifically on patients with r-axSpA and not on axial spondyloarthritis in general. In routine clinical practice, the diagnosis of osteoporosis and assessment of fracture risk in this population remain particularly challenging. Evaluating the usefulness of TBS in the diagnostics of OP in patients with r-axSpA is therefore an important issue.

The objective of this study was to evaluate the clinical utility of the TBS, derived from DXA measurement of the lumbar spine, in assessing fracture risk in patients with r-axSpA over a follow-up period of three years.

The study was devised as a prospective cohort study of r-axSpA patients. All patients enrolled in the study signed an informed consent to participate in accordance with the Declaration of Helsinki. The study was approved by the bioethics commission of the Collegium Medicum in Bydgoszcz, Poland (approval number: KB 771/2018; date of approval: 20/NOV/2018).

67 (18 F/49 M) patients were enrolled into the study as reported earlier [4]. 63 patients (16 F/47 M) from the initial cohort finished the study and were analyzed here. Patients were divided into two groups according to their baseline TBS value: Group 1 (TBS >1.31, low risk) and Group 2 (TBS ≤ 1.31, intermediate & high risk). The TBS cut-off of 1.31 was chosen based on the previous publication. Table 1 shows basic demographic data of the patient group that finished the study.

The results presented in Table 1 indicate a significant difference in age and the prevalence of MOF between the groups. The mean age was higher in Group 2, which also included all patients diagnosed with MOF at baseline (n = 10, 100%). It is worth noting that no statistically significant differences were found between the groups in terms of mean disease duration or BMD - despite the significant differences in both age and the proportion of patients with osteoporotic fractures - as well as BMI, the presence of syndesmophytes and ankylosis and the number of biologic-naïve patients.

During the 3-year follow-up, new MOFs occurred in 7 patients. Among them, 1 patient (14%) belonged to Group 1 (TBS > 1.31) and sustained a vertebral body fracture, while 6 patients (86%) were in Group 2 (TBS ≤ 1.31) − 5 experienced new vertebral body fractures and 1 suffered a wrist fracture.

Accordingly, Table 2 presents baseline variables of clinical relevance and their impact on the risk of MOF over the 3-year follow-up period.

As shown in Table 2, a prior MOF at baseline was the strongest predictor of incident fractures, and the presence of syndesmophytes and ankylosis also appreciably increased risk. A baseline TBS ≤ 1.31 was associated with more than a four-fold rise in fracture risk, but this did not reach statistical significance - most likely because the study recorded only a small number of events during the 3-year follow-up.

PPV, NPV, sensitivity, and specificity were calculated to assess the predictive value of baseline TBS and BMD in identifying patients at risk of MOF during a 3-year follow-up.

Diagnosis of osteoporosis in patients with r-axSpA proves difficult, particularly in routine clinical practice. Assessing fracture risk based solely on DXA BMD in the lumbar spine has little diagnostic value, as has been well documented in a number of studies [4, 9, 22‐24]. BMD evaluation in the proximal femur boasts a higher diagnostic value, but is still limited due to the younger age of most r-axSpA patients [25, 26]. There is a clinical need for an alternative method to assess fracture risk. From a clinical standpoint, a wide availability of such method is paramount. This precludes high-resolution peripheral quantitative computed tomography (HR-pQCT) from being a first-choice method. At this point in time, TBS might be the optimal choice, being performed through additional software on DXA machines, and available at a large number of facilities worldwide.

To the best of our knowledge, this study represents one of the few comprehensive prospective evaluations of TBS in patients with r-axSpA, allowing comparison with previously published retrospective data. Throughout the 3-year observation of the studied group (Table 1), 7 patients, 6 with TBS ≤ 1.31 and 1 with TBS > 1.31, suffered MOF.

As shown in Tables 2 and 3, our findings suggest that TBS may provide more clinically meaningful information about fracture risk than BMD alone in patients with r-axSpA. In terms of relative risk, it was the second most important factor after a prior MOF, although statistical significance was not achieved. It could be argued that the lack of significance was due to the small size of the study group and the relatively short observation period, both of which resulted in a low number of fracture events from a statistical point of view. Unfortunately, limited sample size remains a common limitation of single-center studies.

However, from a clinical perspective, the number of individuals who sustained fractures is by no means small. The fact that 6 (86%) of them had a TBS ≤ 1.31 suggests that low TBS values are associated with an increased risk of fractures and may have significant prognostic value in patients with r-axSpA. It should be noted that when fractures were assessed on the basis of BMD, the ratio was reversed - only one patient (14%) had a T-score < −1.0. The advantage of TBS over BMD is also evident in the PPV, NPV and sensitivity values (Table 3). The sensitivity of TBS in detecting fractures is comparable to results of retrospective studies [4, 9, 27, 28]. Our findings suggest that TBS may be used not only to identify patients with MOF, but also - perhaps even more importantly - to identify individuals who do not require intensive monitoring or osteoporosis specific treatment. This is supported by the high NPV (Table 3). From a clinical perspective, this means that patients with r-axSpA and a normal TBS have a low risk of fractures. The NPV was higher for TBS than for BMD, indicating that TBS provides additional clinical value in this population. At the same time, the low PPV does not diminish the usefulness of TBS, as the occurrence of MOF is inherently a random event. Thanks to its high NPV, TBS enables more accurate identification of patients who are actually at risk and therefore most likely to benefit from OP treatment. This targeted approach supports more effective clinical decision-making and may ultimately lead to improved patient outcomes and more efficient use of healthcare resources.

Although TBS may have considerable predictive value in assessing fracture risk in patients with r-axSpA and the software is relatively widely available, its use in routine clinical practice may still be limited at present. Neither the latest IOF nor ISCD guidelines clearly outline how to take advantage of TBS during routine visits [16, 17]. Furthermore, from a clinical perspective, Romosozumab is currentky the only treatment that has been shown to significantly improve TBS [29, 30]. Our findings suggest that the NPV of TBS could be considered as a supportive parameter when initiating OP diagnostics in patients with r-axSpA - helping to identify individuals with a low fracture propability who may not require immediate drug intervention.

The results of the study suggest that the risk of suffering a MOF is strongly related to previous MOFs (Table 2), a high-risk factor that applies not only to r-axSpA patients but also to the general population.

Taking into account the risk factors listed in Table 2, the presence of syndesmophytes and ankylosis appears to be associated with an increased risk of fracture. Patient age of at least 50 years may also be linked to a higher fracture risk. However, the analysis did not show a significant effect of disease duration on fracture risk. From a clinical perspective, this is not surprising - disease activity may be more relevant than duration alone, especially when patients achieve sustained remission through biological therapy [28, 31]. The use of biologic disease-modifying antirheumatic drugs (bDMARDs), particularly tumor necrosis factor alpha (TNF-α) and interleukin-17 (IL-17) inhibitors, often allows for rapid achievement of low disease activity or even sustained remission. This, in turn, helps to halt radiographic progression and maintain high BMD as well as normal bone microarchitecture [32‐34]. In our cohort, more than half of the patients received biologic treatment (Table 1), which may have reduced the expected fracture incidence during the follow-up period.

The issue of osteoporosis and its consequences in r-axSpA patients is highly complex. Progressive disability may also significantly increase the risk of fracture, particularly due to falls in the home [35]. Therefore, achieving and maintaining low disease activity or remission, preventing progression of syndesmophytes, and improving physical capabilities to reduce risk of falling are important factors in diminishing fracture risk.

Recent studies have shown that patients with r-axSpA tend to have lower TBS values than the general population and that reduced TBS is associated with an increased risk of vertebral fractures [4, 36, 37]. Current evidence also indicates that treatment with romosozumab improves not only BMD but also TBS [38]. This raises the question of whether TBS can help identify patients at increased risk of fractures due to impaired bone microarchitecture. Our findings suggest that this may indeed be the case in patients with r-axSpA.