The Very Low T-Score That Isn't What It Seems
A 71-year-old woman is referred for DXA after a fall from standing height that caused a distal radius fracture. She has no prior DXA. Her height is 158 cm, weight 51 kg, BMI 20.4. Lumbar spine L1–L4 BMD is 0.612 g/cm², T-score −4.1. Left femoral neck BMD is 0.742 g/cm², T-score −1.6; total hip T-score −1.4. The spine image shows dense sclerotic endplates at L3 and L4 and a grading note of mild degenerative change; there is no compression fracture, no aortic calcification, and no prior vertebroplasty. Technologist flagged the spine scan as technically acceptable. The ordering clinician has already written 'severe osteoporosis, start anabolic therapy' on the referral form.
1. Which site should drive the densitometric classification here, and why?
Read the worked answer
The hip should drive classification, not the spine. The spine is the most common site for falsely elevated BMD (degenerative change, endplate sclerosis, facet arthropathy, aortic calcification), but it can also be falsely lowered by focal artifacts. More importantly, the ISCD principle is to use the lowest T-score among valid sites for diagnosis - but a site must first be judged valid and interpretable. Here the spine–hip discordance is large (3.5 T-score units between L1–L4 and femoral neck). A discrepancy of this magnitude is a red flag for spinal artifact rather than a true biological difference.
Practical steps:
- Re-examine the spine image and individual vertebrae, not just the mean. If L3 and L4 are clearly affected by sclerosis, exclude them and report L1–L2 separately.
- If only one vertebra remains evaluable, a single-vertebra spine value is not sufficient to diagnose - use the hip.
- Do not accept 'technically acceptable' as synonymous with 'interpretable.' The technologist judged acquisition quality; the interpreter judges diagnostic validity.
With femoral neck −1.6 and total hip −1.4, the patient is osteopenic by DXA criteria at the hip. The distal radius fracture is a low-trauma fracture, which is itself an independent indication for treatment regardless of T-score. So the clinical decision to treat may be correct, but the label 'severe osteoporosis' based on a T-score of −4.1 at an artifact-affected spine is not defensible.
2. The clinician insists on the spine T-score because 'the guideline says use the lowest site.' How do you respond, and what do you put in the report?
Read the worked answer
The 'lowest site' rule applies to valid, evaluable sites. It is not a license to report an artifact as bone density. The report should:
- State the diagnosis from the hip: femoral neck T-score −1.6, total hip −1.4.
- Explicitly document the spine discordance and the reason the spine is not being used for classification (endplate sclerosis at L3–L4; large site discrepancy out of proportion to expected age-related change).
- Report whatever spine data are defensible (e.g., L1–L2 if clean) with a note that it is not used for diagnostic classification.
- Note that the low-trauma distal radius fracture confers high fracture risk independent of T-score, so treatment is reasonable on clinical grounds.
- Recommend that if the spine value is going to drive a major therapeutic decision (e.g., anabolic vs. antiresorptive), consider repeat imaging or an alternative modality (lateral vertebral fracture assessment, or trabecular bone score with appropriate caveats).
Key teaching point: the interpreter owns the diagnosis, not the referring clinician and not the automated lowest-T-score printout.