CT value ≥41.5 HU and well-defined boundaries are excellent diagnostic variables for diagnosing plasma cell tumors in the appendicular skeleton of patients with multiple myeloma using whole-body low-dose CT.
Key Findings
Results
A CT value ≥41.5 HU demonstrated high diagnostic accuracy for differentiating plasma cell tumors from red bone marrow lesions in the appendicular skeleton.
Sensitivity of CT value ≥41.5 HU as a diagnostic criterion for PCT was 94.2%
Specificity of CT value ≥41.5 HU was 91.8%
Overall diagnostic accuracy of CT value ≥41.5 HU was 93.5%
This threshold was derived from receiver operating characteristic curve analysis of variables with statistically significant intergroup differences (P < .05)
Results
A well-defined lesion boundary served as an independent diagnostic parameter for identifying plasma cell tumors.
Sensitivity of well-defined boundary as a diagnostic parameter was 88.9%
Specificity of well-defined boundary was 93.6%
Diagnostic accuracy of well-defined boundary alone was 88.7%
This morphological feature was identified through intergroup comparison using chi-square test
Results
The combined criterion of CT value ≥41.5 HU plus well-defined boundaries yielded high specificity but slightly lower sensitivity compared to CT value alone.
Combined criterion sensitivity was 86.4%
Combined criterion specificity was 96.5%
Combined diagnostic accuracy was 91.6%
The combined variable showed higher specificity (96.5%) than either criterion alone but lower sensitivity than CT value alone (94.2%)
Methods
The study population consisted of 234 patients with multiple myeloma whose medullary lesions in the appendicular skeleton were categorized into plasma cell tumor and red bone marrow groups.
Total of 234 patients met inclusion criteria, with one lesion evaluated per patient
Mean patient age was 59.5 ± 8.4 years; 128 patients were male
Lesions were categorized as plasma cell tumor (PCT; n = 154) or red bone marrow (n = 80)
Data were collected retrospectively from December 2020 to February 2023
Intergroup comparisons used independent-samples t test, Mann-Whitney U test, or chi-square test as appropriate
Methods
Variables showing statistically significant intergroup differences were selected for receiver operating characteristic curve analysis to evaluate diagnostic accuracy.
Inclusion threshold for ROC analysis was P < .05 on intergroup comparison
Both quantitative (CT value in Hounsfield units) and qualitative (lesion boundary definition) variables were identified as significant
Whole-body low-dose CT (WB-LDCT) was used as the imaging modality for all evaluations
Both clinical and radiological data were incorporated into the analysis
What This Means
This research suggests that two specific features visible on whole-body low-dose CT scans can reliably distinguish between cancerous deposits (plasma cell tumors) and normal red bone marrow in the arm and leg bones of patients with multiple myeloma, a type of blood cancer. The study examined CT scans and clinical data from 234 myeloma patients, categorizing 154 lesions as cancerous and 80 as normal red marrow. The researchers found that if a lesion measured 41.5 Hounsfield units or higher on the CT scan — a measure of tissue density — and had a clearly defined border, it was very likely to be a plasma cell tumor rather than benign red marrow.
Using a CT density of 41.5 HU or above as a standalone criterion correctly identified cancer in about 94% of cases (sensitivity) while correctly ruling it out in nearly 92% of cases (specificity), giving an overall accuracy of 93.5%. A clearly defined lesion boundary alone was slightly less sensitive but similarly specific. When both criteria were applied together, specificity rose to 96.5%, meaning false positives were rare, though sensitivity dropped slightly to 86.4%.
This research matters because accurately distinguishing malignant from benign bone marrow lesions is critical for staging multiple myeloma and guiding treatment decisions. Whole-body low-dose CT is already widely used in myeloma care, and this study suggests that relatively straightforward measurements from these existing scans — lesion density and boundary sharpness — can serve as reliable, practical diagnostic tools without the need for additional invasive procedures.
Zhang K, Wang S, Shi B, Wang J, Wang X, Lu W, et al.. (2026). Using WB-LDCT to differentiate between benign and malignant medullary lesions in the appendicular skeleton of patients with plasma cell diseases.. Medicine. https://doi.org/10.1097/MD.0000000000050461