High-resolution peripheral quantitative computed tomography: research or clinical practice?

Abstract

High-resolution peripheral quantitative CT (HR-pQCT) is a low-dose three-dimensional imaging technique, originally developed for in vivo assessment of bone microarchitecture at the distal radius and tibia in osteoporosis. HR-pQCT has the ability to discriminate trabecular and cortical bone compartments, providing densitometric and structural parameters. At present, HR-pQCT is mostly used in research settings, despite evidence showing that it may be a valuable tool in osteoporosis and other diseases. This review summarizes the main applications of HR-pQCT and addresses the limitations that currently prevent its integration into routine clinical practice. In particular, the focus is on the use of HR-pQCT in primary and secondary osteoporosis, chronic kidney disease (CKD), endocrine disorders affecting bone, and rare diseases. A section on novel potential applications of HR-pQCT is also present, including assessment of rheumatic diseases, knee osteoarthritis, distal radius/scaphoid fractures, vascular calcifications, effect of medications, and skeletal muscle. The reviewed literature seems to suggest that a more widespread implementation of HR-pQCT in clinical practice would offer notable opportunities. For instance, HR-pQCT can improve the prediction of incident fractures beyond areal bone mineral density provided by dual-energy X-ray absorptiometry. In addition, HR-pQCT may be used for the monitoring of anti-osteoporotic therapy or for the assessment of mineral and bone disorder associated with CKD. Nevertheless, several obstacles currently prevent a broader use of HR-pQCT and would need to be targeted, such as the small number of installed machines worldwide, the uncertain cost-effectiveness, the need for improved reproducibility, and the limited availability of reference normative data sets.

Key messages

• HR-pQCT is a three-dimensional imaging technique designed for the evaluation of bone density and microstructure at peripheral sites (especially at the distal radius and tibia). It is characterized by low dose (3–5 μSv of effective dose per scan) and high resolution (61–82 μm).

• HR-pQCT can be used for the study of primary and secondary osteoporosis; it has been very recently demonstrated that HR-pQCT measures improve the prediction of fragility fractures beyond aBMD by DXA and FRAX algorithm.

• Being a low-dose technique, HR-pQCT is suitable for monitoring effects of medications over time and/or individual trajectories of bone deterioration.

• Being a high-resolution technique, HR-pQCT is finding several novel applications in the study of bone, skeletal muscle, and vascular calcifications at peripheral sites.

• HR-pQCT is currently employed mostly in research settings, due to the limited availability of the machine, cost-effectiveness still largely to be determined, and need for validated and accessible normative data sets for HR-pQCT parameters. Tackling these obstacles will result in greater use of HR-pQCT in the near future, with the potential to improve patient outcomes.

Introduction

High-resolution peripheral quantitative CT (HR-pQCT) is a three-dimensional imaging technique originally designed for in vivo assessment of bone microarchitecture at the distal tibia and distal radius in osteoporosis (Figure 1). ¹ HR-pQCT allows to differentiate between bone trabecular and cortical compartments (Figure 2), providing density and structure measurements (Figure 3). ² Standard quantitative parameters that can be determined with HR-pQCT at the distal tibia and radius are listed in Table 1, with the respective abbreviations used throughout this article.Starting from HR-pQCT images, it is possible to calculate estimates of bone stiffness and failure load using finite element analysis (FEA), a widely applied numerical method that can partition in small pieces and then solve complex problems of structural mechanics.

Figure 1.

HR-pQCT standard image acquisition process at the distal radius (a) and distal tibia (b), shown for the XCT2 scanner. a1, b1: patient positioning aided by dedicated limb casts. a2, b2: on a frontal scout radiograph, the operator is asked to place a reference line on epiphyseal contours of the radius (a2) or tibia (b2) (green continuous lines); regions of interest of around 10 mm for HR-pQCT imaging (greed dashed lines) are automatically identified proximally to the reference line (at a fixed or limb-length-relative distance). a3, b3: an example of HR-pQCT reconstructed axial slices displaying micro-structural bone features. HR-pQCT, high-resolution peripheral quantitative CT.

Figure 2.

HR-pQCT standard image processing shown for a XCT2 scan of the distal radius (analogous steps apply for a scan of the distal tibia). (a) The operator, aided by algorithms available on the XCT2 workstation, contours on each slice the periosteal bone contour. (b) The contoured image is binarized to obtain a segmented image. (c) Cortical (c1) and trabecular (c2) compartments are extracted from the segmented image. HR-pQCT, high-resolution peripheral quantitative CT.

Figure 3.

HR-pQCT sample report for a distal radius XCT2 scan. A 3D reconstruction is shown for visualization purposes (a). Results of regional quantitative analyses are shown for geometry (b), density (c) and microstructure (d) of trabecular and cortical compartments. HR-pQCT, high-resolution peripheral quantitative CT.

Table 1.

HR-pQCT parameters

Name [unit]	Abbreviation
Density parameters
Total volumetric bone mineral density [mgHA/cm3]	Tt.vBMD
Cortical volumetric bone mineral density [mgHA/cm3]	Ct.vBMD
Trabecular volumetric bone mineral density [mgHA/cm3]	Tb.vBMD
Structure parameters
Cortical area [mm2]	Ct.Ar
Cortical thickness [mm]	Ct.Th
Cortical porosity [1]	Ct.Po
Cortical pore diameter [mm]	Ct.Po.Dm
Trabecular area [mm2]	Tb.Ar
Trabecular thickness [mm]	Tb.Th
Trabecular number [1 /mm]	Tb.N
Trabecular separation [mm]	Tb.Sp
Trabecular bone volume to total volume or trabecular bone volume fraction [1]	Tb.BV/TV
Standard deviation of 1/Tb.N (inhomogeneity of trabecular network) [mm]	Tb.1/N.SD

HR-pQCT, High-resolution peripheral quantitative CT.

Note: the parameters found in this table and the respective abbreviations are derived from the report provided by the XtremeCT II (XCT2) machine.

Currently, HR-pQCT is produced by a single manufacturer (SCANCO Medical AG, Switzerland). There are two generations of HR-pQCT machines available, with notable differences between the two. ³ The first generation of HR-pQCT (XtremeCT/XCT) was introduced in 2004, while the second (XtremeCT II/XCT2) was introduced in 2014. As compared to XCT, XCT2 is characterized by reduced scanning time (2 vs 3 min), slightly increased but still low effective dose per scan (5 vs 3 μSv), improved resolution (61 vs 82 µm), and larger selectable region of interest. ³ In addition, XCT2 allows to measure trabecular microarchitecture and Ct.Th directly, whereas a derived (indirect) approach is used in XCT. ⁴ XCT2 also permits the study of bone microstructure at the knee. ⁵

Since its introduction, the applications of HR-pQCT have broadened remarkably. Despite its valuable potential, HR-pQCT is currently employed mostly in the research setting and is not yet integrated into routine practice. This review summarizes the most relevant applications of HR-pQCT and addresses the potential advantages and limitations to its implementation in current clinical practice.

Primary osteoporosis

Osteoporosis is a systemic skeletal disease characterized by low bone mass, bone microarchitectural deterioration, and bone fragility, leading to an increased fracture risk. Because osteoporosis is clinically silent until a fracture occurs, it is important to identify affected patients and estimate their fracture risk to begin an appropriate therapy.

Osteoporosis is currently diagnosed based on World Health Organization (WHO) criteria, which consider the T-score based on areal bone mineral density (aBMD) by dual-energy X-ray absorptiometry (DXA). aBMD is inversely related to fracture risk. ^6,7 However, there are several limitations linked to the use of aBMD to predict fractures and diagnose osteoporosis. An important number of fragility fractures occurs in individuals with an aBMD not in the osteoporotic range, ⁸ suggesting that other factors beyond aBMD play a role in defining fracture risk. In addition, DXA does not allow to evaluate trabecular and cortical bone compartments separately and does not provide any insight into bone microarchitecture, which is an integral part in the definition of osteoporosis. These limitations could be overcome by the use of HR-pQCT.

Prediction of fracture risk in primary osteoporosis

Several individual studies addressed the association of HR-pQCT parameters with fractures in post-menopausal females and males, such as the OFELY, ⁹ MrOS, ¹⁰ CaMos, ¹¹ and STRAMBO ¹² studies. In general, they yielded positive results, but their population size was limited or the relative number of incident fractures was small. To examine more comprehensively the association between bone microarchitecture determined by first-generation HR-pQCT and incident fractures, 7254 individuals (males and females) from 8 cohorts of different countries were included in a prospective study by the Bone Microarchitecture International Consortium (BoMIC), ¹³ with a total number of 765 incident fractures. Most participants had femoral neck T-scores in the osteopenic or normal range. Results confirmed that HR-pQCT parameters can predict incident fractures independently of femoral neck aBMD determined by DXA and Fracture Risk Assessment Tool (FRAX) score. The hazard ratio (HR) for the association between femoral neck aBMD (per SD decrease) and incident fracture was 1.57 (95% confidence interval [CI]: 1.42–1.74). By contrast, the HR (per SD decrease) for failure load in Newton estimated by μFEA (per SD decrease) at the distal tibia was 2.40 (1.98–2.91) and that at the distal radius was 2.13 (1.77–2.56). The HRs for other parameters ranged from 1.12 (1.03–1.23) for Ct.Po at the tibia and 1.58 (1.45–1.72) for Tb.vBMD at the radius. After adjustment for femoral neck aBMD or FRAX, the HRs were attenuated but remained significant for most parameters, meaning that HR-pQCT indices can improve prediction of fracture beyond femoral neck aBMD or FRAX alone.

There is now compelling evidence for the ability of HR-pQCT measures to predict fractures in primary osteoporosis from two recent systematic reviews and meta-analyses. ^4,14 Mikolajewicz et al ¹⁴ included 40 studies in their systematic review and meta-analysis, mostly conducted in Europe or North America. Except for one study using XCT2, all studies used XCT. Four studies were prospective. Despite moderate to high heterogeneity among data sets, the meta-analysis demonstrated that all examined radial and tibial HR-pQCT parameters (except Ct.Po) were significantly different in fractured subjects compared to non-fractured controls. In addition, data from retrospective and prospective studies were consistent, indicating that HR-pQCT can effectively predict fractures. The authors also examined whether fracture-associated differences exceeded the least significant change (LSC) threshold to exclude precision errors. They concluded that only some of the measures were significantly larger than the LSC (Tt.vBMD; Tb.vBMD; Ct.Th and Ct.vBMD at the tibia only), prompting the need for improved reproducibility.Cheung et al ⁴ included 25 studies (of which two using XCT2) in their systematic review and 14 studies in their meta-analysis. The authors confirmed that almost all HR-pQCT parameters (including Ct.Po but only at the tibia) can predict incident fractures and/or major osteoporotic fractures. The best predictors were Ct.vBMD, Tb.Th, and bone stiffness.

HR-pQCT can also be used to formulate or test hypotheses about the microstructural determinants of bone fragility, as well as to suggest new metrics or classifiers. For example, three bone microarchitecture phenotypes associated with different levels of osteoporotic fracture risk were identified using data from the BoMIC cohort. ¹⁵ In addition, a recent study ¹⁶ suggested that heterogeneous microarchitectural deterioration leading to the formation of void spaces may play a role in the determination of bone fragility, on top of low bone mineral density.

To conclude, there is a good deal of evidence supporting the use of HR-pQCT to gain insight into bone microstructure in osteoporosis and improve prediction of fragility fractures beyond aBMD by DXA. At present, HR-pQCT cannot be used for the diagnosis of osteopenia/osteoporosis, because this is made by WHO criteria according to DXA results and because reference normative data are still being collected. In addition, current guidelines for the diagnosis and management of osteoporosis, including the International Society for Clinical Densitometry (ISCD) 2019 official positions, ^17,18 do not delineate precisely the role that HR-pQCT should play in routine practice. However, HR-pQCT may be used in selected patients with normal or osteopenic aBMD to identify those at highest risk of fracture, who could benefit from treatment. In addition, being HR-pQCT a low-dose technique, it may be utilized for the follow-up of patients deemed at increased risk of fracture or affected by diseases that influence bone metabolism, enabling the monitoring of individual trajectories of bone deterioration, which may guide personalized treatment.

Monitoring of anti-osteoporotic therapy

HR-pQCT also offers the possibility to monitor the effects of anti-osteoporotic therapy on bone microstructure. Differing responses were reported with the use of anti-remodeling agents as compared to anabolic agents in a recent review, ¹⁹ with anabolic agents increasing Ct.Po and decreasing Ct.vBMD but preserving (or slightly improving) estimated bone strength, and anti-remodeling agents slightly increasing vBMD without resulting in an increase in estimated bone strength. In most studies included, greater changes were appreciated at the distal tibia compared to the radius, possibly due to its weight-bearing nature.

Several HR-pQCT studies, including some randomized controlled trials (RCTs), were published examining the effects of bisphosphonates (alendronate, risedronate, ibandronate), ^20–22 denosumab, ²³ and teriparatide, ^24–26 mostly in post-menopausal females. Tsai et al ^27,28 compared the response to teriparatide, denosumab, and their combination in post-menopausal females. Two recent RCTs examined the effects of denosumab or alendronate compared to placebo ²⁹ and teriparatide (daily or weekly) compared to bisphosphonates, ³⁰ showing that teriparatide and denosumab, but also bisphosphonates, may improve failure load.

In general, available published studies reported small but statistically significant changes in HR-pQCT parameters with the use of anti-osteoporotic therapy, often with an improvement of a few percentage points. ² Most studies included a limited number of subjects, with considerable heterogeneity and limited follow-up duration. HR-pQCT thus appears to be a promising tool for the assessment of the response to anti-osteoporotic drugs, being the only clinically available scanner that permits the simultaneous assessment of bone microstructure and bone density, in both the cortical and trabecular compartment. However, the expected improvement in HR-pQCT-measured outcomes is modest and the relationship between observed changes and fractures in treated patients is still unknown. More evidence from larger studies is needed to define the role of HR-pQCT in the monitoring of anti-osteoporotic treatment.

Chronic kidney disease

Mineral and bone disorder (MBD) is a systemic condition that encompasses mineral, skeletal, and calcific cardiovascular abnormalities, arising as a complication of chronic kidney disease (CKD). The term renal osteodystrophy is used to refer to the pathologic alterations in bone morphology seen in CKD, which can be quantified by bone biopsy. ³¹ From the skeletal point of view, CKD-MBD includes a spectrum of conditions, ranging from high-turnover bone disease associated with secondary hyperparathyroidism (osteitis fibrosa) to low-turnover bone disease (adynamic bone disease or osteomalacia).

Patients with CKD-MBD are at increased risk of fracture: this risk generally becomes clinically relevant from CKD G3 (grading according to Kidney Disease Improving Global Outcomes—KDIGO categories, ³² increases as CKD progresses, and is especially high in patients with CKD G5 or undergoing dialysis. ³³ Current KDIGO guidelines suggest the determination of BMD with DXA in patients with CKD G3a-5 and evidence of CKD-MBD to evaluate their fracture risk, and eventually the execution of a bone biopsy to define the type of renal osteodystrophy. ³² These diagnostic tests are important because they have an impact on treatment, which is different in high-turnover compared to low-turnover status. KDIGO guidelines do not provide specific recommendations on the use of HR-pQCT in CKD-MBD, but this technique is increasingly used in the research setting to gain insight into bone microstructure in CKD patients.

An initial study with HR-pQCT showed that trabecular microarchitecture starts to be impaired early in CKD (stage II–IV according to 2003 Kidney Disease Outcomes Quality Initiative—KDOQI recommendations) in both genders, with CKD patients exhibiting Tb.vBMD, Tb.N, and Tb.Sp values in between those of healthy controls and osteopenic controls. ³⁴ CKD was later associated with a rapid decline in cortical bone microarchitecture in a prospective HR-pQCT study, ³⁵ reporting a significant annual decrease in Ct.Ar (- 1.7%), Ct.vBMD (−1.2%), and Ct.Th (−1.4%) and a significant annual increase in Ct.Po (+ 4.2%) at the distal radius in CKD patients with eGFR <90 ml/min/1.73 m². Interestingly, cortical deterioration could be predicted by average serum levels of parathyroid hormone (PTH). ³⁵

HR-pQCT parameters were recently found to be significantly different in patients with KDOQI Stage 4 CKD compared to Stage 3 CKD (lower Tt.vBMD, Tb.vBMD, Ct.vBMD, Tb.BV/TV, Tb.N, Tb.Th, Ct.Th; more heterogeneous trabecular network in terms of Tb.Sp SD), in line with the progressive nature of metabolic bone disease in CKD. ³⁶ These alterations were associated with modifications in body composition and hormonal levels. ³⁶

Results from HR-pQCT were compared with those from iliac crest bone biopsy in patients with CKD in a small study, ³⁷ showing significant agreement in several parameters, particularly cortical. Ct.vBMD at the radius was higher in patients with low-turnover compared to high-turnover status, meaning that HR-pQCT could be useful to discriminate the turnover status in CKD-MBD, especially in combination with serum biomarkers. This finding was confirmed in a later study on patients with KDOQI Stages 4–5 CKD (including patients on dialysis), in which distal radius Tt.vBMD and structural parameters were negatively associated with bone turnover. ³⁸

HR-pQCT was also used to evaluate specifically hemodialyzed patients, showing that hemodialysis is associated with a marked decrease in Ct.Th, Ct.Ar, and Ct.vBMD, and with a significant reduction in trabecular parameters, at both the radius and tibia. ³⁹ A recent meta-analysis confirmed that HR-pQCT parameters are significantly altered in dialyzed patients compared to controls, including Tt.vBMD at the distal radius and tibia and Ct.vBMD at the distal radius. ⁴⁰ The ability of HR-pQCT to categorize CKD-MBD into high- vs low-turnover status was also demonstrated specifically in patients on hemodialysis, in comparison with bone biopsy. ⁴¹

Overall, it can be said that HR-pQCT offers a valuable opportunity to study metabolic bone disease in CKD across its various stages, although larger, prospective studies are needed before this technique becomes routinely applied in clinical practice. In the future, it may be used to define the type of renal osteodystrophy in alternative to bone biopsy, which is invasive but currently remains the gold-standard. HR-pQCT may be combined with serum biomarkers to improve the differentiation between high and low-turnover status. HR-pQCT may also be useful in the management of CKD for its ability to predict fracture risk. Microarchitecture was shown to be deteriorated irrespective of femoral neck BMD category (normal, osteopenic, osteoporotic) in patients with CKD, and to be an independent predictor of bone failure load. ⁴²

Secondary osteoporosis and endocrine disorders affecting bone

Diabetes mellitus

HR-pQCT can be used for the study of bone microstructure in Type 1 and 2 diabetes mellitus (T1DM/T2DM). In adolescents with T1DM, Tb.Th at the tibia and estimated trabecular loading at the distal radius and tibia were shown to be significantly lower compared to controls. ⁴³ These changes occurred before alterations in DXA became apparent and were possibly related to poor glycemic control. ⁴³ Bone microarchitecture and strength in T1DM might also be influenced by the presence of microvascular complications, such as diabetic peripheral neuropathy, at least at the distal tibia. ⁴⁴ A recent meta-analysis showed that T1DM is specifically associated with adverse trabecular characteristics, in contrast to T2DM. ⁴⁵

T2DM is associated with an increased fracture risk despite normal or increased aBMD, suggesting a significant role for compromised microarchitecture in determining fracture risk. ² In general, HR-pQCT studies in patients with T2DM provided highly variable results, possibly due to their heterogeneity. ² Nevertheless, several studies reported an impairment in bone microarchitecture in T2DM patients compared to controls. ^46–48 Changes in HR-pQCT parameters might be related to glycemic control and presence of microvascular disease. Shanbhogue et al ⁴⁹ found that HR-pQCT parameters differed significantly in T2DM patients with microvascular disease compared to controls, but not in T2DM patients without microvascular disease; while de Waard et al ⁵⁰ highlighted differences in HR-pQCT parameters in T2DM patients depending on their glycated hemoglobin (HbA1c) levels, with HbA1c > 7% being associated with lower Ct.vBMD and Ct.Th and higher Ct.Po and Tb.N at the radius, and higher Tb.N and lower Tb.Th at the tibia.

The reported large variability in results of HR-pQCT studies may discourage implementation of HR-pQCT imaging in this setting. However, recent evidence supports the inclusion of HR-pQCT in diagnostic strategies aimed at refining the detection of bone fragility in T2DM. These may focus specifically on impaired HR-pQCT-derived cortical bone properties, especially cortical porosity. ^45,51 A recent ex vivo study reported that in human tibial cortical bone from individuals with T2DM, high cortical porosity was associated with larger HR-pQCT-derived cortical pore diameter and altered bone material composition. ⁵²

Glucocorticoid-induced osteoporosis

HR-pQCT was applied in a few studies to the assessment of glucocorticoid-induced osteoporosis. Sutter et al. ⁵³ reported that post-menopausal females treated with oralglucocorticoids for > 3 months had abnormal cortical and trabecular vBMD and microarchitecture at the distal radius and tibia compared to controls, despite similar aBMD by DXA. HR-pQCT parameters were also shown to predict vertebral fractures in patients under long-term steroid treatment. ⁵⁴

Primary hyperparathyroidism

HR-pQCT allowed to identify impaired bone density and microstructure in both trabecular and cortical compartments in patients with primary hyperparathyroidism (PHPT). ^55,56 In particular, Wang et al ⁵⁶ reported lower cortical and trabecular vBMD, thinner cortices, and more widely spaced trabeculae in PHPT patients compared to controls. Of note, parathyroidectomy in PHPT patients was shown to improve bone microstructure and estimated bone strength at both the distal radius and tibia. ⁵⁷

Hypoparathyroidism

Hypoparathyroidism is an uncommon endocrine disorder associated with a profound decrease in bone remodeling due to reduced levels of parathyroid hormone (PTH). ⁵⁸ Affected subjects were shown to have BMD values by DXA greater than controls at all skeletal sites. ⁵⁸ HR-pQCT was also applied to the study of bone changes in hypoparathyroidism, showing an increase in Ct.vBMD in both females and males at the radius and tibia, and a decrease in Ct. Po in females at both sites and in young males at the tibia, with respect to controls. ⁵⁹ While advancing age decreased ultimate stress and failure load at the radius and tibia, longer duration of hypoparathyroidism increased these indices. ⁵⁹ A recent study showed that both trabecular and cortical bone compartments are affected by hypoparathyroidism, and that changes in bone microstructure might be influenced by the etiology of the disorder, especially at the trabecular level, with higher Tb.N in non-surgical compared to post-surgical hypoparathyroidism. ⁶⁰ HR-pQCT was also used in one study to monitor the effects of long-term recombinant human PTH treatment in hypoparathyroidism. ⁶¹

Systemic lupus erythematosus

HR-pQCT was used in several studies to address the changes in bone density and microstructure in patients affected by systemic lupus erythematosus (SLE): a chronic autoimmune disorder associated with osteoporosis, osteopenia, and increased fracture risk. ⁶² It was demonstrated that SLE per se contributes to deterioration in bone density, structure, and strength assessed by HR-pQCT and that SLE seems to affect predominantly cortical bone, both in Chinese female patients treated with glucocorticoids and in those not receiving glucocorticoids, with lower Ct.vBMD and Ct.Th, and higher Ct.Po compared to controls at the distal radius. ^63,64 However, a cross-sectional study conducted in Caucasian female patients showed that trabecular bone may also be affected by SLE: in particular, SLE patients had lower Tb.N and estimated failure load with higher Tb.Sp at the radius, and lower Tt.vBMD, Ct.Ar, and Ct.Th, with higher Tb.Ar at the tibia. ⁶⁵ HR-pQCT also showed to be valuable in monitoring longitudinal changes in bone density and microstructure in SLE patients treated with glucocorticoids or vitamin D supplementation. ^66,67

Obesity

Body weight was shown to influence HR-pQCT parameters in studies conducted in patients with obesity. ^68–71 Obesity was associated with higher vBMD, thicker and denser cortices, higher Tb.N, and lower Ct.Po (in older patients at the tibia) compared to normal weight. ⁷⁰ However, the differences observed in bone density, architecture, and strength were shown to be out of proportion to the excess of body mass index (BMI) in other studies ^68,71 ; bone adaptations in morbid obesity may indeed be inadequate with respect to the increased mechanical demands. ⁶⁹ Moreover, HR-pQCT measurements appear to be susceptible to artifacts derived from increased adiposity in soft tissue overlying bone, albeit theoretically less than those by DXA. ^72,73 This should be considered, especially in studies involving individuals with differing body composition or in those assessing longitudinal changes in body weight. ^72,73

Rare diseases

HR-pQCT has been employed for the study of osteogenesis imperfecta (OI): a genetic disease including different clinical phenotypes of variable severity, characterized by bone fragility and increased fracture risk. DXA is often used in this setting, but aBMD is not necessarily decreased ⁷⁴ and does not correlate well with fracture risk in OI. In addition, kyphoscoliotic deformities in patients with OI can affect the execution of DXA at the lumbar spine. HR-pQCT may thus be a precious tool for the evaluation of OI. Kocijan et al ⁷⁴ showed that adult patients with OI have lower Tb.vBMD and impaired trabecular microstructural parameters in comparison to healthy controls, especially in type III-IV as opposed to type I. Hald et al ⁷⁵ reported significant differences in HR-pQCT parameters between type I and type IV OI in adults, which were correlated with molecular abnormalities; in addition, HR-pQCT was able to distinguish type I from type IV OI better than DXA. HR-pQCT was also explored in children with OI. ⁷⁶ In summary, available evidence suggests that the use of HR-pQCT is feasible in patients with OI and may improve prediction of fracture risk and aid treatment planning.HR-pQCT may also be used in other rare diseases, such as X-linked hypophosphatemia, ⁷⁷ cystic fibrosis, ⁷⁸ inborn errors of metabolism, ⁷⁹ and Marfan syndrome. ⁸⁰

Further applications

Rheumatoid arthritis

HR-pQCT has recently gained interest in rheumatology for the evaluation of rheumatoid arthritis (RA), psoriatic arthritis, ⁸¹ and even ankylosing spondylitis. ⁸² HR-pQCT may be used to assess joint space narrowing and detect erosions in hand joints in patients with RA, especially at the second and third metacarpophalangeal (MCP) joints. ^83,84 This is routinely done with conventional radiographs, which however do not allow tridimensional visualization of joints. HR-pQCT has been validated for the determination of joint space width ⁸⁵ and has demonstrated reliability in the detection of erosions and in their monitoring over time. ⁸⁶ HR-pQCT is capable of detecting small cortical interruptions, which may precede frank erosions and are not visible using conventional radiographs. ⁸⁷ Because of its high sensitivity, HR-pQCT has also been proposed to monitor the effects of anti-rheumatic treatment. ^88,89

Knee osteoarthritis

Using second-generation HR-pQCT, it is possible to scan the knee, allowing the study of subchondral bone microstructure in patients with osteoarthritis (OA). Bhatla et al ⁹⁰ showed a significant increase in subchondral bone plate thickness in knees with prior anterior cruciate ligament reconstruction compared to contralateral or control knees, which may be associated with earlier development of OA. Shiraishi et al ⁹¹ analyzed alterations in subchondral trabecular bone microstructure in patients with medial knee OA, reporting significantly higher Tb.vBMD, Tb.BV/TV, and Tb.Th in the medial tibial plateau compared to the lateral plateau. In addition, subchondral bone microstructure at the anterior region of the medial plateau correlated well with Kellgren-Lawrence score (used to grade the severity of OA) and lower limb alignment. ⁹¹ Despite holding promise for the evaluation of knee OA, further studies are needed to define the precise role of HR-pQCT in this setting. Of note, patient positioning for knee scanning is uncomfortable and may not be tolerated by elderly subjects.

Distal radius and scaphoid fractures

The healing process of fractures affecting the distal radius or scaphoid can be studied with HR-pQCT, with considerable advantage compared to conventional radiographs ^92,93 and relevant practical implications: bone stiffness estimated by μFEA from HR-pQCT has been explored to guide the duration of cast immobilization following distal radius fractures. ⁹⁴ HR-pQCT may even improve the detection of suspected scaphoid fractures, which remain challenging to diagnose based on conventional radiographs. ⁹⁵ However, these specific applications of HR-pQCT remain mostly limited to research studies and are unlikely to become widespread in clinical practice soon, given that supporting evidence is still being collected and that there is restricted availability of HR-pQCT machines worldwide (except possibly in few reference centeres).

Sports medicine

HR-pQCT is emerging as a valuable tool also in sports medicine. It has been explored to determine how bone microarchitecture adapts to physical activity ⁹⁶ and its use may help in the prevention of stress fractures in athletes. ⁹⁷

Effect of medications

As previously stated, HR-pQCT has been proposed to monitor the efficacy of pharmacologic treatment in various diseases, including osteoporosis, RA, and SLE. However, it can also be used to gain insight into the effect of medications that are known to impact bone metabolism, such as aromatase inhibitors (AIs) and androgen deprivation therapy (ADT). Both treatments appear to be associated with a deterioration in bone structure that affects the trabecular as well as the cortical compartment. ^98–100 Given the plethora of parameters measured, HR-pQCT can add invaluable information to define the pathophysiological mechanisms that determine bone fragility in patients treated with ADT or AIs.

Vascular calcifications

HR-pQCT permits the assessment of vascular calcifications in the lower leg, which may be quantified using a specific algorithm. ¹⁰¹ Interestingly, vascular calcifications and osteoporosis may share common etiopathogenic mechanisms. Paccou et al ¹⁰² showed that the presence of lower leg arterial calcifications in females correlated with bone microstructure abnormalities detected by HR-pQCT.

Skeletal muscle

Skeletal muscle is known to be critical for bone quality. Given the intimate relationship existing between skeletal muscle and bone, a complete assessment of musculoskeletal health must include the evaluation of skeletal muscle. ¹⁰³ Several different imaging modalities can be used to study skeletal muscle, including DXA, ultrasound, MRI, CT, and peripheral quantitative CT (pQCT). HR-pQCT was recently proposed for the study of skeletal muscle at the tibial midshaft (66% proximal tibia), thanks to the possibility to assess more proximal sites with second-generation scanners. ¹⁰³ HR-pQCT could achieve good precision, given its higher resolution compared to pQCT, and measures of muscle density were shown to be comparable between HR-pQCT and pQCT. ¹⁰³ In addition, HR-pQCT has the potential to obtain simultaneous measurements of bone and muscle, providing insight into bone–muscle interactions.Erlandson et al ¹⁰⁴ examined the feasibility of using first-generation HR-pQCT to estimate the properties of myotendinous tissue at the distal tibia, showing a moderate correlation with mid-leg muscle density obtained from pQCT.While holding promise for the evaluation of skeletal muscle in research settings, development of standardized protocols for this specific application of HR-pQCT is still ongoing. ¹⁰⁵ In addition, published evidence is mostly limited to feasibility studies. Further research is needed to explore the role of HR-pQCT in the imaging of skeletal muscle and related pathologies, such as sarcopenia.

Current limitations to the use of HR-pQCT in clinical practice

Some concerns remain preventing a broader use of HR-pQCT:

• The number of HR-pQCT machines installed worldwide is limited (approximately 100 as of mid-2022) and they are largely found at research centes.

• The reproducibility of HR-pQCT parameters seems to be suboptimal. ¹⁴ Reproducibility is influenced by various factors, including scanner resolution, motion artifacts, and operator skills. XCT2 has reduced scanning time and improved resolution compared to XCT and calculates trabecular parameters and Ct.Th directly. This is likely to translate into better precision and reproducibility per se but may be complemented by initiatives of open and reproducible science aimed at sharing and cross-validating quantitative analysis methods over large cohorts. As a notable example, Zhao et al ¹⁰⁶ recently developed an open-source image analysis tool for the detection and measurement of cortical interruptions and bone erosions in rheumatoid arthritis.

• HR-pQCT is quite expensive and its cost-effectiveness is still largely to be determined. A first study ¹⁰⁷ concluded that identifying osteopenic females with severe microstructural deterioration using HR-pQCT and treating them with zoledronate was cost-effective compared to standard care. This suggests that HR-pQCT may be especially useful to guide management in osteopenic females, who do not routinely receive anti-osteoporotic therapy. Further studies are absolutely needed to characterize the cost-effectiveness of HR-pQCT.

• Given the technical differences existing between first- and second-generation machines, measures obtained with XCT are not necessarily interchangeable with those obtained with XCT2. Agarwal et al ³ reported good agreement for most of the volumetric BMD, trabecular, and cortical measurements, except for Ct.Po at the radius and Tb.N and Tb.Th at both the radius and tibia. Therefore, XCT2 measures seem to be essentially comparable to XCT1 measures, but some caution is needed, especially for parameters that are more dependent on resolution like Tb.N, Tb.Th, or Ct.Po. Interestingly, Manske et al ¹⁰⁸ found that XCT2 measures could be estimated from XCT data by cross-calibration. Extensive cross-calibration of machines and use of standard metrics are key to the clinical implementation of HR-pQCT and will become even more important if scanners from different producers are marketed.

• A standardized protocol for image acquisition and data analysis as well as a consensus on terminology related to HR-pQCT are needed to facilitate research and implementation into clinical practice. Guidelines addressing this subject have been recently published by a joint working group between the International Osteoporosis Foundation, American Society of Bone and Mineral Research, and European Calcified Tissue Society. ¹⁰⁵

• Current guidelines in the field of osteoporosis, specifically International Society for Clinical Densitometry (ISCD) 2019 adult official positions, ¹⁷ provide some general recommendations on the use of non-central DXA devices, including pQCT. They state that T-scores from measurements other than DXA at the femur neck, total femur, lumbar spine, or one-third radius cannot be used to diagnose osteoporosis according to WHO diagnostic criteria, as they are not equivalent to T-scores derived by DXA. However, they suggest that a sufficiently high fracture probability as assessed by pQCT of the radius, in conjunction with clinical risk factors, can be used to initiate pharmacologic treatment if central DXA cannot be performed. In addition, they legitimate the use of pQCT to monitor age-related changes in BMD.International Society for Clinical Densitometry (ISCD) 2019 pediatric official positions ¹⁸ report that HR-pQCT and pQCT are primarily research techniques but can be used clinically in children where appropriate reference data and expertise are available. It should be considered that while the latest ISCD guidelines were published in 2019, significant evidence and knowledge on HR-pQCT has been gained over the past few years. Therefore, it would be desirable to include ad hoc recommendations focused on the preferred indications of HR-pQCT in new position statements by international societies, in particular to better define the categories of patients who would benefit the most from this technique on top of standard care with DXA in routine practice. This would favor an evidence-based implementation of HR-pQCT in the clinical assessment of patients affected by osteoporosis or at risk of fracture.

• There is a paucity of validated normative data on HR-pQCT parameters. This is crucial for a broader clinical use of HR-pQCT throughout all its possible applications. Reference data sets are needed to compare measures obtained in patients with values reported in a standard healthy population and determine whether they represent pathologic findings. In addition, HR-pQCT parameters vary depending on age, gender, and ethnicity, as well as body weight. ² This translates into the necessity of having accessible reference data from several representative populations, across different ages, in both males and females.Some reference data sets are available for first-generation HR-pQCT in adult Brazilian females [n = 450, 20–85 years] ¹⁰⁹ and males [n = 340, 20–92 years] ¹¹⁰ ; in adult Caucasian females and males, both European [n = 499, 20–80 years] ¹¹¹ and Canadian [n = 866, 16–98 years] ¹¹² ; and in Chinese females and males [n = 1072, 20–79 years]. ¹¹³ For second-generation HR-pQCT, there are normative data sets for adult Chinese females and males [n = 863, 20–80 years] ¹¹⁴ ; adult Caucasian females and males, Canadian [n = 1236, 18–90 years], ¹¹⁵ American [n = 1184, 18–80 years], ¹¹⁶ and European [n = 87, 20–39 years] ¹¹⁷ ; and adult Japanese females [n = 61, 31–70 years]. ¹¹⁸ An open-source application is available online to display individual data on normative graphs ¹¹⁵ and there are calculators enabling the determination of Z-scores and T-scores for each patient according to reference data sets. ¹¹⁶ With the use of age-related reference curves, it is indeed possible to determine T-scores for HR-pQCT measures. However, these are not equivalent to T-scores by DXA and their role in clinical practice is still undefined. Some normative data have been published also for children, adolescents, and young adults. Some of the largest data sets were obtained with first-generation HR-pQCT in adolescents and young adults of Caucasian ethnicity [n = 251, 16–29 years], ¹¹⁹ and in a cohort of subjects of mixed Caucasian and Asian origin [n = 349, 10–21 years]. ¹²⁰ A recent systematic review and meta-analysis ¹²¹ analyzed available literature to define normative data in this age range, pointing out the need for standardization in the acquisition of parameters especially in children, adolescents, and young adults. There are indeed two protocols that can be used for the selection of the region of interest (fixed vs relative offset) in growing subjects, which can yield different results. Recent guidelines state that a relative offset is preferred in children and adolescents, although further work is required to achieve a consensus on standardized protocols for pediatric studies. ¹⁰⁵ To summarize, there are some normative data sets available in literature for HR-pQCT parameters. However, they only include subjects from few selected populations and some ethnicities are underrepresented, such as Africans or Hispanics. Undoubtedly, these data sets need to be expanded in the future, requiring effort and collaboration of the research community. In addition, it would be desirable for these reference data to be included directly in the HR-pQCT software.

Conclusion

Thanks to its high spatial resolution, non-invasiveness, low dose, and ability to discriminate trabecular and cortical bone compartments, HR-pQCT has proven to be a valuable tool for the evaluation of primary and secondary osteoporosis, but also for several other potential applications, as discussed in this review. At present, its use is vastly restricted to the research setting since several obstacles prevent its integration into clinical practice. However, with a joint effort of the HR-pQCT community, these limitations may be overcome in the near future, paving the way for a broader implementation of HR-pQCT, with the potential to improve patient outcomes across multiple different pathologies.