Clinical Study to Predict the Risk of Bone Fractures With the POROUS Ultrasound Device

A Prospective, Single-cohort, Multicentre Clinical Investigation to Evaluate the Performance of POROUS R3C Ultrasound Device for Fracture Risk Prediction in Middle-aged and Elderly Men and Women

Osteoporosis is a widespread medical condition among older people. It causes the bones to weaken and become more likely to break. Osteoporosis and bone fracture risk are currently evaluated by looking at clinical risk factors and measuring bone mineral density (BMD). The lower the BMD is, the higher the risk of osteoporotic fractures in the future. However, most bone fractures occur in people who do not have very low BMD values. This means that osteoporosis and fracture risk are often not diagnosed. Many of these non-diagnosed patients would benefit from treatment to reduce the probability of bone fractures.

An X-ray device called DXA is the main tool used to diagnose osteoporosis and fracture risk clinically. DXA measures two-dimensional BMD in the hip and spine. The POROUS ultrasound device measures various properties of the outer layer of the bone in the lower leg. It has several advantages over DXA: (1) its image resolution is higher and three-dimensional; (2) it can detect bone changes without radiation; (3) it can detect these bone changes early and how they change over time.

For this clinical study, we will recruit men and women over 55 years old. Most will have clinical risk factors, such as background diseases, for developing osteoporosis. The study is anticipated to last 4 years.

Our major research questions are:

• Can the POROUS ultrasound device predict fracture risk?
• How does its performance compare to DXA?
• What is the safety of the new device?

The participants will:

• answer questions about their medical history.
• be measured for height and weight, and take a physical test.
• be examined for the presence of 'silent' fractures in the spine.
• be examined at the beginning and end of the study with the two devices, DXA and POROUS.
• be called by telephone every six months and asked if they suffer from new bone fractures, take any medication that might affect their bones, or if their health status has changed.

The participants will be monitored for 3 years.

Background and purpose:

Currently, osteoporosis and fracture risk are indirectly evaluated via the assessment of risk factors and bone mineral density (BMD) measurement. Although BMD is currently the most important indicator for osteoporosis-associated bone fractures, most of those fractures occur in persons who do not show pathologically reduced BMD value. Therefore, osteoporosis is one of the most frequently underdiagnosed common diseases. Established guidelines for the diagnosis of osteoporosis recommend the assessment of fracture risk factors and the T-Score, which is derived from the measurement of areal bone mineral density (aBMD) by means of DXA at major fracture sites, i.e., spine and proximal femur. DXA is regarded as the "gold standard" well-established methodology to determine aBMD for diagnostic purposes. Epidemiological data emphasize the urgency of developing diagnostic tools that can improve fracture risk prediction so that patients can be treated with the appropriate anti-osteoporotic therapies. Current guidelines for diagnosis and treatment lead to treatment gaps. It is estimated that at least 80% of males and 77% of females who would benefit from osteoporosis treatment are neither diagnosed nor treated in Germany.

Device description:

The POROUS R3C ultrasound device enables a non-invasive, non-ionizing quantitative detection of microstructural bone changes. As opposed to diagnosis based on a combination of clinical risk factors and a relative decrease of BMD, the novel device enables detecting pathological changes of bone microstructure at an earlier timepoint as well as monitoring such changes in a longitudinal manner. In the course of this clinical investigation, data will be collected to establish relevant ultrasound-based physical biomarkers for the prediction of fracture risk.

Study design:

This is a single-cohort, multicenter, prospective, age- and sex-stratified study in participants > 55 years of age. In this study, Baseline data will be collected to establish a corrected standardized gradient of fracture risk using the POROUS R3C ultrasound device and test its performance in predicting fracture risk. Further, the performance of the POROUS R3C ultrasound device in the analysis of cortical bone properties and discrimination of prevalent fractures will be assessed. Participants will be enrolled into different groups based on their age (consisting of five-year bands), sex (males and females), and risk status for hip and vertebral fractures (i.e., high risk of ≥ 2-fold and low risk of < 2-fold increased risk compared to the general population of the same age and sex).

Two measurements with each device: investigational device (POROUS R3C ultrasound device at the midshaft tibia), and comparator (DXA of the lumbar spine and proximal femur), are scheduled per participant:

• Measurement 1: At Baseline
• Measurement 2: At the End-of-Study (EoS) visit.

Study Part 1:

In Part 1, information on prevalent fractures will be used to establish a corrected standardized gradient of fracture risk using the POROUS R3C ultrasound device. In other words, Part 1 aims to establish the discriminative performance and a standardized gradient of fracture risk based on prevalent fractures. In addition, the discriminative performance of the POROUS R3C ultrasound device and standard-of-care DXA will be compared based on prevalent fractures.

Study Part 2:

In Part 2, information on incident fractures will be used to establish a corrected standardized gradient of fracture risk using the POROUS R3C ultrasound device. It will be developed to demonstrate the predictive performance of the derived fracture risk. In other words, Part 2 aims to establish a standardized gradient of fracture risk based on incident fractures. In addition, the predictive performance of the POROUS R3C ultrasound device and standard-of-care DXA will be compared based on incident fractures.

The collection of data obtained by DXA and the POROUS R3C device at the EoS visit is used to monitor changes in the bone state in comparison to the respective data obtained at Baseline. However, only measurement data obtained by the POROUS R3C device at Baseline is used to develop the POROUS-Score model and the standardized gradient of fracture risk (for Part 1 and Part 2).

Primary objectives:

• Part 1 -To establish a corrected standardized gradient of fracture risk using the POROUS R3C ultrasound device based on prevalent fractures.
• Part 2 - To establish a corrected standardized gradient of fracture risk using the POROUS R3C ultrasound device based on incident fractures and to demonstrate the predictive performance of the derived fracture risk.

Secondary objectives:

• Part 1 - To compare the discriminative performance of the POROUS R3C ultrasound device and standard-of-care DXA based on prevalent fractures.
• Part 2 - To assess the safety of the POROUS R3C ultrasound device by monitoring adverse events affecting participants or the healthcare professionals using the device.
• Part 2 - To compare the predictive performance of the POROUS R3C ultrasound device and standard-of-care DXA based on incident fractures.

Exploratory objectives:

• Part 1 - To assess the association of various ultrasound parameters measured by the POROUS R3C ultrasound device with specific clinical risk factors/indicators for vertebral and hip fractures (as outlined by the DVO) and prevalent fractures.
• Part 1 - To demonstrate the discriminative performance of the POROUS R3C ultrasound device based on subgroups of prevalent fractures, e.g., hip, vertebral, and major osteoporotic fractures.
• Part 1 - To establish reference data for developing age-adjusted POROUS Z-scores using the POROUS R3C ultrasound device based on prevalent fractures.
• Part 2 - To assess the association of various ultrasound parameters measured by the POROUS R3C ultrasound device with specific clinical risk factors/indicators for vertebral and hip fractures (as outlined by the DVO) and incident fractures.
• Part 2 - To demonstrate the predictive performance of the POROUS R3C ultrasound device based on subgroups of incident fractures, e.g., hip, vertebral, and major osteoporotic fractures.
• Part 2 - To establish reference data for developing age-adjusted POROUS Z-scores using the POROUS R3C ultrasound device based on incident fractures.
• Part 2 - To explore the treatment effect of anti-osteoporotic medication
• Part 2 - To explore the treatment effect of drugs known to influence bone metabolism.

Participants:

A total of 1,600 female and male participants (> 55 years of age) will be included in the investigation. This population is planned to include 1,120 participants with ≥ 2-fold increased age- and sex-adjusted risk for hip and vertebral fractures and 480 participants with a < 2-fold increased age- and sex-adjusted risk for hip and vertebral fractures.

Stratification of participants, fracture risk < 2-fold increased risk:

• Age group 56-60, males/females = 40 per group, total = 80.
• Age group 61-65, males/females = 40 per group, total = 80.
• Age group 66-70, males/females = 40 per group, total = 80.
• Age group 71-75, males/females = 40 per group, total = 80.
• Age group 76-80, males/females = 40 per group, total = 80.
• Age group 81-85, males/females = 40 per group, total = 80.

Stratification of participants, fracture risk ≥ 2-fold increased risk:

• Age group 66-70, females = 100 per group.
• Age group 71-75, males = 100 per group, females = 160 per group, total = 260.
• Age group 76-80, males/females = 180 per group, total = 360.
• Age group 81-85, males/females = 200 per group, total = 400.

The expected number of incident fractures* ~140 (at 24 months), ~210 (at 36 months).

*The assumed number of fractures per analysis group (DXA and POROUS R3C) at Month 24 and Month 36, respectively, is based on age- and sex-matched incidence rates for fractures (Rupp et al., Deutsches Ärzteblatt International, 2021), further updated by data on the incidence of vertebral fractures (Fink et al., JBMR, 2005).

Risk calculation for hip and vertebral fractures:

At Screening, assessment of clinical risk factors and application of risk calculating scheme as outlined in the Dachverband Osteologie (DVO, tri-national umbrella association of German, Austrian, and Swiss medical and scientific professional societies in the field of bone diseases) osteoporosis guideline will determine whether a participant is at increased risk (≥ 2-fold compared to the general population of the same age and sex) for hip and vertebral fractures. Please note that the calculation of the pertinent risk as performed in this study does not include any DXA-based BMD measurements and T-Scores. No DXA BMD measurement or DXA-based vertebral fracture assessment (VFA) will be conducted at Screening. Noteworthy, DXA-based BMD measurement results and T-Scores will be included in the different analyses of the study endpoints but not for the risk calculation at Screening as described above. Screened and eligible individuals will be enrolled in the investigation until the necessary sample size for his/her corresponding group (based on age, sex, and risk status) has been reached (see Table below). To avoid over-recruiting, once the necessary sample size for one group has been reached, no further individuals with matching age band, sex, and risk status will be enrolled in the study.

Duration of the study:

The planned overall clinical investigation is expected to take 48 months, including an enrolment period of approximately 12 months and a clinical investigation period of 36 months. Per participant, the total time of participation in the investigation will take 36 months. 20 months (+/-1 month) after the first participant has been included in the investigation, a checkpoint assessment of the number of fracture events so far recorded is planned. The aim is to assess whether the number of fractures projected to be reached after each participant will have been followed-up for 24 months would be sufficient for a test of the primary endpoint with sufficient statistical power. If this is the case, the clinical investigation period will be shortened to 24 months (correspondingly, the overall duration will be shortened to 36 months). If, at the checkpoint assessment, the number of fractures is insufficient, the follow-up period will remain as planned and the EoS visit will take place at 36 months, amounting to the originally planned overall duration of 48 months.

Duration of participation:

The total time of participation in the investigation will take 36 months per participant, with the option to be shortened to an investigational period of 24 months based on the results of a checkpoint assessment of reported fractures scheduled at Month 20 (+/-1 month).

Fracture risk prediction model:

The diagnostic value of physical biomarkers, which are derived from Baseline measurements with the POROUS R3C ultrasound device, will be assessed, and selected physical biomarkers will be integrated into a model, resulting in a composite POROUS-Score, for fracture risk prediction. In Part 1, the model will be developed using data on prevalent fractures, while in Part 2, the model will be developed using data on incident fractures. The resulting POROUS-Scores are, therefore, different in Part 1 and Part 2. In Part 1, it is termed POROUS-Score(Prev), whereas in Part 2, it is termed POROUS-Score. Additionally, anthropometric data (age, sex, and BMI) will be evaluated and selected for adding predictive strength to the prediction model.

Part 1:

Prevalent fractures will be recorded, and all relevant variables, including ultrasound parameter values and anthropometric information, will be used to perform partial least squares - discriminant analysis (PLS-DA) followed by subwindow permutation analysis using a Monte-Carlo approach. Only variables that have statistically significant discriminative power will be used in the next step, where a new fracture discrimination model will be created using PLS-DA analysis. Thereafter, the performance of the model will be investigated using receiver operating characteristics (ROC) analysis. More precisely, logistic discriminant analysis will be performed. Area under the curve (AUC) values and their confidence intervals will be computed from ROC curves for each model. Internal validation of the model will be performed by using cross-validation followed by bootstrap analysis. Odds ratios will be determined for both the final composite POROUS-Score(Prev) (generated from the internally validated model) and DXA T-Score. Standardized odds ratios (sOR) will be calculated, i.e., the fold-increase of the relative fracture risk per standard deviation decrease of the respective score. The discriminative ability of the POROUS model for prevalent fractures is demonstrated if the lower limit of the 90% confidence interval of the corrected sOR is higher than 1.

Part 2:

Incident fractures will be recorded, and all relevant variables, including ultrasound parameter values and anthropometric information, will be used to perform PLSDA followed by sub-window permutation analysis using a Monte-Carlo approach. Only variables that have statistically significant predictive power will be used in the next step, where a multivariate Cox-proportional Hazard model will be performed to calculate the hazard ratio (HR). Then, the model performance will be investigated using ROC analysis. Internal validation of the model will be done by running cross-validation followed by Bootstrap analysis. The standardized relative risk (sRR) will be calculated, i.e., the fold-increase of the relative fracture risk per standard deviation decrease of the respective score. The sRR will be determined for both the final composite POROUS-Score (generated from the internally validated model) and DXA T-Score. The predictive ability of the POROUS model for incident fractures is demonstrated if the lower limit of the 90% confidence interval of the corrected sRR is higher than 1.

Clinical procedures:

The following clinical procedures are performed during the investigation:

Screening

• Informed consent
• Inclusion/exclusion criteria
• Calculation of individual risk for hip and vertebral fractures as per modified version of the DVO risk calculation scheme (excluding DXA-based BMD measurement and T-Score)
• Enrolment into corresponding age-sex-risk group.

Baseline (Visit 2, 0-14 days after Screening)

• Scanning with the POROUS R3C ultrasound device at the central anteromedial tibia region
• Scanning with DXA (aBMD of the lumbar spine and proximal femur).
• Fracture anamnesis, including DXA-based VFA. Alternatively, VFA may be replaced by projectional radiography if VFA is not available.
• Assessment of clinical risk indicators (as per DVO guideline)
• Assessment of concomitant medication (initiation of anti-osteoporotic medication as well as oral glucocorticoids, proton pump inhibitors, aromatase inhibitors, hormone ablation therapy/antiandrogens in males).
• Recording of adverse events.

Follow-up period Interim visits (phone survey) - Visits 3-7 at Month 6, 12, 18, 24, 30 ± 14 days

• Assessment of incident fractures (to be validated by requesting medical documents - X-ray images, physician's findings)
• Concomitant medication assessment
• Recording of adverse events

Early-termination (ET) visit/phone survey Participants who withdraw early (i.e., before undergoing the EoS visit) will be offered an ET clinical visit, including

• VFA (or projectional radiography of the spine, respectively, if VFA is not available)
• Anamnesis of incident fractures since the last visit/phone survey
• Concomitant medication assessment (see above)
• Recording of adverse events. (Alternatively, the visit can be conducted as a phone survey if the time gap since the last DXA/VFA is < 6 months or the participant is not available for a clinical visit).

EoS visit - Visit 8 at Month 36 ± 14 days

• Scanning with the POROUS R3C ultrasound device at the central anteromedial tibia region
• Scanning with DXA (aBMD of the lumbar spine and proximal femur).
• Assessment of incident fractures including VFA (or projectional radiography of the spine)
• Re-assessment of fracture risk factors/indicators as per DVO guideline and recalculation of risk for hip and vertebral fractures
• Concomitant medication assessment
• Recording of adverse events.

Subgroup analysis:

General subgroups for analyses are defined based on age, sex, and risk (for hip and vertebral fractures). Stratification groups are presented in the table on "Number of participants to be included in the investigation" included above. In addition, subgroup analyses are done to monitor treatment effects, as described in the exploratory endpoint of Part 2.

Inclusion Criteria:

• Female or male individuals aged 56 to and including 85 years.
• Written informed consent has been obtained.

Assessment of risk factors for hip and vertebral fractures:

To avoid over- and under-recruiting with regard to the required sample size of participants with ≥ 2-fold increased age-and sex-adjusted risk for hip and vertebral fractures and participants with < 2-fold increased age- and sex-adjusted risk, clinical risk factors necessary for the calculation of the risk for hip and vertebral fractures (based on the risk calculation scheme outlined in the DVO osteoporosis guideline) are assessed at Screening.

Vertebral fractures:

• Vertebral fracture(s) during the last year
• Vertebral fracture(s) > 12 months ago
• Number of vertebral fractures
• Maximal severity of vertebral fractures, according to Genant

Hip fractures and other fractures:

• Hip fracture during the last year
• Hip fracture > 12 months ago
• Humerus fracture
• Pelvic fracture
• Wrist (radius distal) fracture

General risk factors:

• Mother or father with hip fracture, if the participant is under 75 years of age
• Significant alcohol consumption (50 g/day or more)
• Smoking (currently > 10 cigarettes/day)
• Chronic-obstructive pulmonary disease (COPD)
• Body Mass Index (BMI) ≤ 20

Medication:

• Opioids
• Proton pump inhibitors > 3 months
• Oral glucocorticoids > 3 months (prednisone equivalent in mg/d)

Fall-associated risk factors/geriatrics:

• Number of falls within the last year
• Chronic hyponatremia
• Depression/antidepressants
• Anticonvulsants in epilepsy
• Immobility (being dependent on a walking aid)
• Alzheimer's disease/dementia
• Parkinson's disease
• Multiple sclerosis
• Stroke
• Time up and Go Test > 12 seconds in participants ≥ 70 years of age

Endocrinology:

• Diabetes mellitus Type I
• Diabetes mellitus Type II (including time since onset)
• Primary hyperparathyroidism
• Thyroid-stimulating hormone (TSH) suppression (if yes, including TSH level)

Other diseases/medications:

• Chronic heart failure
• Monoclonal gammopathy of unclear significance (MGUS)
• Chronic kidney disease (CKD) stages 3a, 3b, 4

Rheumatology:

• Rheumatoid arthritis
• Axial spondyloarthritis

Exclusion Criteria:

• Presence of diseases that rule out valid measurements with the DXA and/or POROUS R3C devices (e.g., fractures or metal implants in the examined bones, paralysis of the lower extremities, severe bone abnormalities).
• Inability to undergo the investigations required by the Clinical Investigation Plan (CIP) or cognitive limitations that preclude understanding of the Participant Information Sheet and the Informed Consent Document.
• Previous medical procedures involving exposure to a cumulative dose of ionising radiation deemed by the Investigator to exceed usual limits within standard of care.
• Pregnancy and breastfeeding
• Enrolment in any other interventional clinical study (current or during the last three months)
• Individual is in custody by order of an authority or a court of law.
• Close affiliation with an investigational site, e.g. employed at investigational site, close relative of an investigator, dependent person (e.g. student of the investigational site).

Further, individuals who are being or have been treated within the indicated period prior to the beginning of the study with any of the following antiresorptive therapies are excluded from the clinical investigation:

• Bisphosphonates (due to residual effects of bisphosphonates after discontinuation):

  • Intravenous (IV) zoledronate within the last 3 years.
  • Oral alendronate within the last year, if (continuous) treatment duration before was > 1 year.
  • Oral risedronate within the last year, if (continuous) treatment duration before was > 1 year.
  • Ibandronate (IV or oral) within the last year, if (continuous) treatment duration before was > 1 year.

• Denosumab within the last 3 years

• Hormone replacement therapy (HRT) including combination therapy or oestrogen alone in postmenopausal women within the last 6 months.

• Raloxifene within the last 6 months.

Individuals who are being or have ever been treated with any of the following anabolic therapies are excluded from the clinical investigation:

• Teriparatide
• Romosozumab
• Abaloparatide.