A Pilot Proof of Concept Evaluation of Sacral BMD Measurement in Women with Gynecologic Cancers

Purpose/Aims

To evaluate a potential DXA approach to sacral BMD measurement using extended field and standard L1-4 scans in women with gynecologic cancer.

Rationale/Background

Few data exist regarding bone status in women with gynecologic cancer despite known bone toxic effects of treatment-induced menopause, chemotherapy, and pelvic radiation. Cancer treatment-induced bone loss almost certainly increases subsequent fracture risk. Pelvic insufficiency fracture is a potentially catastrophic complication occurring in up to 7.8% of women. It is plausible that sacral BMD measurement could identify women at higher risk for this complication. Whether sacral BMD can be measured as part of routine DXA scanning has not been explored.

Methods

Subjects were from a study evaluating BMD change in women treated for gynecologic cancers. Standard clinical spine, hip, forearm and VFA scans, along with an extended length spine scan to include the sacrum, were acquired. Using a GE Lunar iDXA, sacral scans were obtained from the pubic tubercle cranially to the standard spine termination at T12. Sacral regions of interest (ROIs) were placed by outlining the sacrum (ROI 1) then this ROI was divided in half horizontally (ROIs 2 and 3; Figure 1). L1-L4 BMD from standard and extended scans were compared by Pearsons correlation and Bland-Altman analyses. Sacral ROI BMD was correlated by Pearsons with mean total hip, L1-4 and 0.3 radius BMD.

Results

Ten women, mean (SD) age and BMI of 53.7 (11.0) years and 32.9 (9.5) kg/m2 were studied. All subjects underwent hysterectomy with bilateral oophorectomy within 35 (14.9) days of baseline DXA scan. Mean L1-4 BMD was 1.146 (0.177) g/cm2 and lowest T-score -0.3 (1.5). Sacral BMD at ROIs 1, 2 & 3 was 0.808 (0.192), 0.897 (0.170) and 0.771 (0.210) g/cm2 respectively. Extended spine scan L1-4 BMD was highly correlated (r = 0.996) with standard L1-4 spine BMD and demonstrated a low bias, -0.006 g/cm2. Sacral BMD of all ROIs correlated with L1-4 (r = 0.88 – 0.93; p < 0.001) and mean total hip BMD (r = 0.79 – 0.84; p < 0.05), but not 0.3 radius (r = -0.23 to -0.12).

Implications

These data suggest that lumbar spine BMD can be measured using longer scan length DXA, equivalent to standard L1-4 measurements. That sacral BMD corelates with trabecular (spine and hip) but not a cortical sites (0.3 radius) could be expected and may suggest potential utility to monitor BMD change following gyn cancer therapy. Future research will focus on sacral BMD reproducibility and change post treatment.