Five Lumbar-Type Vertebrae, a Bridging Osteophyte, and a Hip That Is Only Borderline
A 76-year-old woman (163 cm, 71 kg) is referred for her first DXA after fracturing her distal radius in a fall from standing height. Her history includes a lumbar laminectomy at L4–L5 in 2016 and a 22-degree degenerative lumbar scoliosis. She takes calcium and has never had osteoporosis treatment. On the spine image the software labels five lumbar-type vertebrae, L1–L5. L1, L2, and L3 look structurally normal. L4 carries a large bridging osteophyte with a sclerotic superior endplate. L5 lies partly below the iliac crest line. Lumbar BMD values are L1 0.790 g/cm², L2 0.762, L3 0.744, L4 1.090, L5 1.120. The software prints the L1–L4 mean as T −1.9. The hip scan gives femoral neck 0.590 g/cm² (T −2.4) and total hip 0.690 (T −2.2). The referring clinician asks whether the patient has osteoporosis and whether the spine result can be used.
1. Which lumbar vertebrae should be included in the reported spine result, and how do you justify each inclusion or exclusion?
Read the worked answer
Work down the image vertebra by vertebra, not down the list of numbers.
- L4: exclude. The bridging osteophyte and sclerotic endplate are focal structural change. Such changes add mineral to the measurement and falsely raise BMD, so the T-score comes out falsely high (here, far higher than the rest of the spine). The reason for exclusion is the structural change seen on the image, documented on the report - not the fact that the number is inconveniently high.
- L5: exclude. It sits partly below the iliac crest, so the region of interest is not a valid lumbar measurement.
- L1–L3: include. Three structurally unremarkable vertebrae remain, which satisfies the requirement of at least two evaluable vertebrae for a diagnostic spine T-score.
The governing principle: exclusions must be pre-specified and image-based. Excluding a vertebra because it is the highest or the lowest value is illegitimate; here the justification would stand even if L4 had measured low. Report the region as "L1–L3," not "L1–L4."
2. The software-generated report already lists L1–L4 with T −1.9. What do you do with that value?
Read the worked answer
That mean is contaminated and must not stand as the primary spine result. Averaging L1–L4 lets two non-evaluable vertebrae dilute three valid ones: the mean comes out around 0.847 g/cm² (T −1.9) instead of about 0.765 g/cm² (T −2.6) for L1–L3. The direction of the error matters - it makes the patient look better than she is, so the risk of undertreatment, not overtreatment, is the problem.
Practical handling:
- Recalculate and report the L1–L3 mean and its T-score as the valid spine value.
- Keep the individual vertebra values visible so the reader can audit the decision.
- State explicitly, in the body of the report, which vertebrae were excluded and why ("L4 excluded - bridging osteophyte/endplate sclerosis; L5 excluded - below iliac crest").
- If the L1–L4 mean appears anywhere in the output, label it as not valid for diagnosis. Do not simply delete the data.
- Note the scoliosis as a technical limitation; mild-to-moderate scoliosis does not by itself invalidate properly placed regions of interest.
3. Combining the valid spine and hip data, what classification should be reported and how should it be worded?
Read the worked answer
Classification uses the lowest T-score among valid sites, not the average of sites and not the site that happens to be first on the page.
- Valid spine (L1–L3): T ≈ −2.6
- Femoral neck: T −2.4
- Total hip: T −2.2
Lowest valid T-score = −2.6, which is at or below −2.5, so the WHO densitometric category is osteoporosis. The distal radius fragility fracture is consistent with that classification, and the clinician should be told the spine value is based on L1–L3 only.
Wording points that matter for scoring and for the chart: name the region used, state that L4 and L5 were excluded and why, and avoid presenting the L1–L4 mean as if it were equivalent. Keep the technologist's scope in view: the report classifies BMD; whether to start pharmacotherapy is the clinician's decision and should not be asserted in the DXA report.