Opportunistic assessment of bone mineral density in cystic fibrosis patients using ultra-low dose thoracic CT

  • Sahil Shet
  • , Catherine Henry
  • , Mairead O’Donnell
  • , Eid Kakish
  • , Muhammad Ghauri
  • , Patrick O’Regan
  • , Kevin Deasy
  • , Hisham Ibrahim
  • , Michael Maher
  • , Barry Plant
  • , David J. Ryan

Research output: Contribution to journalArticlepeer-review

Abstract

Summary: In this study, we used routine ultra-low dose computed tomography scans of patients with cystic fibrosis to predict bone mineral density (BMD). A strong correlation was found between the attenuation of trabecular bone in thoracic vertebrae and the BMD in the proximal femur and lumbar spine as measured on DEXA. Purpose: Osteoporosis is a serious global health concern with millions of people affected worldwide. A particularly vulnerable cohort in developing osteoporosis are patients with cystic fibrosis (CF). Bone mineral density (BMD) is typically measured with dual energy X-ray absorptiometry (DEXA) scanning; however, this comes at a cost to the healthcare system and an exposure to ionising radiation. In our institution, patients with cystic fibrosis undergo routine ultra-low dose computed tomography (ULDCT) for monitoring of disease progression. The aim of this study was to assess the validity of estimating BMD using data derived from ULDCT scans. Methods: Adult CF patients were included if they had undergone a routine ULDCT scan within 12 months of a DEXA scan. Additionally, 100 non-CF patients with non-contrast standard dose CT scans were selected to act as the control group. Trabecular bone density (T-BD) at T4, T7 and T10 was measured on PACS in Hounsfield units (HU) and compared to DEXA scan results and a formula developed to the predict BMD. Results: Fifty-two female and 62 male patients were included with mean ages of 34.4 and 35.1 respectively. Moderately strong correlation was found between the T-BD and BMD of both the lumbar spine (r = 0.629, p < 0.001) and proximal femur (r = 0.649, p < 0.001). Receiver operator characteristic (ROC) curve analysis found a sensitivity and specificity of 0.700 and 0.714 respectively at predicting osteoporosis at T-BD of 193.33 HU or below. Conclusion: T-BD measured on ULDCT may be a valuable tool in the early identification of CF patients at risk of osteoporosis.

Original languageEnglish
Article number91
JournalArchives of Osteoporosis
Volume20
Issue number1
DOIs
Publication statusPublished - Dec 2025

Keywords

  • Bone mineral density
  • Cystic fibrosis
  • Morphomics

Fingerprint

Dive into the research topics of 'Opportunistic assessment of bone mineral density in cystic fibrosis patients using ultra-low dose thoracic CT'. Together they form a unique fingerprint.

Cite this