Accurate response assessment of breast cancer bone metastasis using computed tomography and new response criteria

Abstract

568 Background: The lack of accurate criteria for assessing the response of bone metastases from breast cancer has led to bone disease being considered unmeasurable. Criteria from the UICC and the WHO based on radiography (XR) and skeletal scintigraphy (SS) are not satisfactory for clinical use. Criteria for computed tomography (CT) and magnetic resonance imaging (MR) have yet to be validated. We recently proposed a set of CT- and MR-inclusive criteria (MDA) to assess bone response (JCO 22:2942, 2004). Here we retrospectively reviewed image sets to evaluate the accuracy of MDA criteria (vs UICC, WHO) and the accuracy of each imaging modality for assessing bone response. Methods: XR, SS, CT, and MR images from patients with breast cancer and bone-only metastases were reviewed by musculoskeletal radiologists at baseline and after treatment (early: 2–6 mo, late: 11–13 mo). Responses were characterized as complete or partial, or stable or progressive disease (CR, PR, SD or PD) according to the 3 sets of criteria (MDA, WHO, UICC). Results from each imaging modality were compared with clinical evidence of response (symptoms, tumor markers, imaging). Results: We reviewed 180 image sets from 47 patients and analyzed images from 41 patients. At the early phase, we found that CT was more sensitive (77%; 95% CI 54–99; P=0.03) than SS (36%; 95% CI 11–61), and MDA criteria were more sensitive (87%; 95% CI 69–100; P=0.02) than WHO (40%; 95% CI 15–65) for detecting response. Specificity was no different among modalities or criteria. Those with CR or PR by MDA criteria early in treatment may have had better overall survival than nonresponders (SD+PD) (HR 0.4; 95% CI 0.17–1.14; P=0.09), with no difference by WHO criteria (HR 1.0; 95% CI 0.38–2.56; P=0.98). We could not examine XR, MR, or UICC criteria because too few XR and MR images were available. Conclusions: CT is more sensitive than SS for showing clinical response of bone lesions from breast cancer. The MDA criteria is more likely to be appropriate than WHO criteria for assessing bone metastases from breast cancer. Prospective studies are warranted to determine whether use of the CT- and MR-inclusive MDA criteria improves the ability to detect and quantify responses in bone metastases. No significant financial relationships to disclose.