CT Texture Analysis in Breast Cancer Patients Undergoing CT-Guided Bone Biopsy: Correlations With Histopathology

Texture analysis has the potential to deliver quantitative imaging markers. Patients receiving computed tomography (CT)-guided percutaneous bone biopsies could be characterized using texture analysis derived from CT. Especially for breast cancer (BC) patients, it could be crucial to better predict the outcome of the biopsy to better reflect the immunohistochemistry status of the tumor.

The present study examined the relationship between texture features and outcomes in patients with BC receiving CT-guided bone biopsies.

This study is based on a retrospective analysis.

The present study included a total of 66 patients. All patients proceeded to undergo a CT-guided percutaneous bone biopsy, using an 11-gauge coaxial needle. Clinical and imaging characteristics as well as CT texture analysis were included in the analysis. Logistic regression analysis was performed to predict negative biopsy results.

Overall, 33 patients had osteolytic metastases (50%) and 33 had osteoblastic metastases (50%). The overall positivity rate for the biopsy was 75%. The clinical model exhibited a predictive accuracy for a positive biopsy result, as indicated by an area under the curve (AUC) of 0.73 [95% confidence interval (CI) = 0.63-0.83]. Several CT texture features were different between Luminal A and Luminal B cancers; the best discrimination was reached for "WavEnHH_s-3" with a P-value of .002. When comparing triple-negative to non-triple-negative cancers, several CT texture features were different, the best discrimination achieved "S(5,5)SumVarnc" with a P-value of .01. For the Her 2 discrimination, only 3 parameters reached statistical significance, "S(4,-4)SumOfSqs" with a P-value of .01.

The utilization of CT texture features may facilitate a more accurate characterization of bone metastases in patients with BC. There is the potential to predict the immunohistochemical subtype with a high degree of accuracy. The identified parameters may prove useful in clinical decision-making and could help to identify patients at risk of a negative biopsy result.