Raman spectroscopic platform for transcutaneous monitoring of bone quality

Osteoporosis (OP) is a global health concern with enormous socioeconomic burdens. Alarmingly, rates of osteoporosis screening using standard dual-energy x-ray absorptiometry (DXA) are abysmal, which contributes to high rates of preventable fragility fractures. Our long-term objective is to develop an accessible screening method based on Raman spectroscopy (RS) for early, accurate identification of at-risk patients during visits to primary care providers or community clinics. We posit that this would increase referrals for a gold standard DXA diagnosis and earlier interventions to reduce the incidence of preventable fragility fractures. We previously demonstrated the reliability and sensitivity of RS to detect biochemical changes associated with bone diseases in various mouse models and reported robust correlations of Raman spectral features with whole bone strength and fracture toughness. We also developed instrumentation and sophisticated algorithms to subtract optical contributions from overlying soft tissue to enable reliable diagnostically sensitive transcutaneous RS (tRS) of murine bone on intact limbs. As we pivot to scale up our instrumentation to make diagnostic measurements in humans, we are addressing three significant challenges. First, the ability to make reliable transcutaneous bone measurements is hampered by the signal from the thick layers of soft tissues that overlay the bone. Therefore, we identified the phalanges and metacarpals in the hand as anatomical sites suitable for tRS measurements, which can be accomplished by adjusting illumination source-detector offsets and establishing spectral libraries of various tissues. Second, there are no currently demonstrable associations between RS of peripheral bones of the hand and BMD at the clinically relevant sites of fragility fractures such as the wrist, hip, and spine. Third, there is currently no evidence that RS of bone can be safely and reliably acquired in living subjects for bone health diagnosis. In Aim 1, we will adapt our group’s novel spectral unmixing algorithm, SOLD, to apply it to a recently acquired dataset of human cadaver hand tRS measurements. The adaptation will account for spatial heterogeneities in Raman spectral responses at the midshaft versus the epiphyseal regions. In Aim 2 we will demonstrate diagnostic associations between tRS measurements in the phalanges and metacarpals of cadaver hands with clinically relevant wrist and hip DXA T-scores and wrist fracture risk. The cadaver hands will be obtained from donors with different sexes, ages, races, BMI, and DXA BMD and T-scores. In Aim 3, we will launch a pilot in vivo tRS study on 50 volunteers to improve methodology, identify differences from cadaver data, and perform fracture risk estimations directly from the tRS data. The proposed studies represent essential steps to demonstrating proof-of-concept of transcutaneous Raman spectroscopy as a clinically relevant diagnostic and prescreening tool for osteoporosis. Successful completion of the project will lay the foundation for future clinical studies.