Purpose: To investigate parameters based on fluorine 18 fluorodeoxyglucose (FDG) positron emission tomographic (PET) imaging that are best correlated with pathologic complete response (PCR) in human epidermal growth factor receptor type 2 (HER2)-positive cancer and triple-negative breast cancer (TNBC) and with partial or complete response in ER-positive/HER2-negative breast cancer.
Materials and methods: This study was approved by institutional review board with waivers of informed written consent and included consecutive patients treated by neoadjuvant chemotherapy. Five PET examination-derived parameters were tested: standard uptake value (SUV) maximum (SUV(max)), peak (SUV(peak)), and mean (SUV(mean)), metabolically active tumor volume, and total lesion glycolysis (TLG). Absolute values at baseline PET, at PET imaging after two cycles of chemotherapy, and variation (ie, change) were measured. Correlations with pathologic response (Wilcoxon rank-sum test) and predictive power assessed (area under the curve [AUC] on the basis of receiver operating characteristic analysis) were examined.
Results: Included were 169 consecutive patients (mean age, 50 years). PCR was more frequent in HER2-positive tumors (16 of 33 patients [48.5%]) and TNBCs (20 of 54 patients [37%]) than in ER positive/HER2-negative tumors (four of 82 [4.9%]) (P < .001). Among patients with ER-positive/HER2-negative cancers, 33 patients had partial response. In TNBC, best association with PCR was obtained with change in SUV(max) (AUC, 0.86) or change in TLG (AUC, 0.88). In HER2-positive phenotype, absolute SUV(max) (or SUV(peak)) values at PET imaging after two cycles of chemotherapy (AUC for each cycle, 0.93) were better correlated with PCR than change in SUV(max) (AUC, 0.78; P = .11) or change in TLG (AUC, 0.62; P = .005). Regarding ER-positive/HER2-negative cancers, change in SUV(max) or change in TLG (AUC, 0.75) were parameters best correlated with partial or complete response. Baseline SUV(max) was higher in lymph nodes than in primary tumor in 31 patients. Findings were similar considering the site with highest FDG uptake.
Conclusion: Quantitative indexes of tumor glucose use that are best correlated with pathologic response vary by phenotype: change in SUV(max) or TLG are most adequate for TNBCs and ER-positive/ HER2-negative cancers and absolute SUV(max) after two cycles of chemotherapy for HER2-positive breast cancers.
© RSNA, 2015.