Predictive value of subcutaneous uptake on 18F-FDG PET/CT for distant metastasis in breast cancer patients

Introduction: In patients with breast cancer, 18F-FDG accumulation just below or along the skin (subcutaneous uptake, SCU) is occasionally seen in the affected breast. However, its clinical significance remains unknown. In this study, we investigated the predictive value of SCU for distant metastasis in patients with breast cancer.

Methods: We retrospectively reviewed 209 female breast cancer patients who underwent whole-body 18F-FDG PET/CT at our hospital from January 2012 to March 2021 before therapy. SCU was defined as abnormal subcutaneous 18F-FDG accumulation extending over the nipple/areolar in the affected breast apart from the tumor. The presence of SCU was visually assessed. Quantitative assessment was also performed using “subareolar SUV-ratio”, defined as subareolar SUVmax of the affected side divided by that of the contralateral side. Clinical N-stage (cN) was determined based on the extent of nodal involvement estimated on PET/CT, and cN2b-3 was defined as “cN-extensive”. Distant metastasis (M-stage) was assessed comprehensively with at least 6 months follow-up. Histological features were evaluated in surgical or biopsy specimen. Subtypes were classified as ER+/HER2-, HER2+, and TN based on immunohistological reports. The association between M-stage and ten imaging and clinicopathological variables, including age, tumor SUVmax, tumor size, presence of SCU, subareolar SUV-ratio, cN-stage, SUVmax in ipsilateral axillary lymph node (ALN SUVmax), subtype, histological grade, and Ki-67 were examined using Student's T-test, Wilcoxon rank-sum test, the chi-square test, or Fisher's exact test. In the subgroup analysis, patients were stratified into the following three groups based on the clinically determined locoregional stages: cT0-3N0-2a, cT4N0-2a, cT1-4N2b-3. Multivariate logistic analysis was performed to identify independent factors predicting distant metastasis.

Results: In the whole subjects, the median age was 56 (range, 23-85 years). Tumor SUVmax and tumor size were 7.1 (range, 3.6-11.3) and 27 (range, 18-35 mm), respectively. Twenty-five (12%) of 209 patients were cN-extensive, and the median ANL SUVmax was 1.3 (range, 0.8-4.5). The most common histological features were invasive carcinoma of no special type (NST) in 173 patients (83%), histological grade 3 in 96 patients (46%), and ER+/HER2- in 139 patients (67%). Of 209 patients, 25 (12%) patients had distant metastasis. Overall, SCU was observed in 31 (15%) of 209 patients, including 21 (11%) of 184 M0 patients and 10 (40%) of 25 M1 patients. High subareolar SUV-ratio with a cut-off value of 1.2 was seen in 13 (7%) of M0 184 patients and in 18 (72%) of 25 M1 patients. Presence of distant metastasis was significantly associated with older age (mean±SD years, M0 vs. M1: 55±13 vs. 63±15; p = 0.014 ), larger tumor size (the median [interquartile range (IQR)] mm, M0 vs. M1: 25 [17-33] vs. 35 [26-56]; p = 0.0007), presence of SCU (M0 vs. M1: 11% [21/184] vs. 40% [10/25]; p = 0.001), higher subareolar SUV-ratio (the median[IQR], M0 vs. M1: 1.1 [1.0-1.3]vs. 1.3 [1.2-1.4]; p = 0.0048), cN-extensive (M0 vs. M1: 7% [13/184] vs. 48% [12/25]; p < 0.001), and higher ALN SUVmax (the median[IQR], M0 vs. M1: 1.1 [0.8-3.5] vs. 4.7 [1.4-9.4]; p = 0.0004). Distant metastasis was not significantly associated with any of the histological findings. Patients were stratified by three locoregional stages in the subanalysis. In the cT0-3N0-2a group, SCU was observed in 15 (9%) of 170 patients, and was significantly associated with M-stage (M0 vs. M1: 7% [12/161] vs. 33% [3/9]; p = 0.0341). In this group, tumor size was also significantly associated with M-stage (the median [IQR] mm, M0 vs. M1: 24 [16-32] vs. 32 [29-44]; p = 0.0043). Multivariate analysis using SCU and tumor size as variables showed that only SCU was a significant predictor for distant metastasis in the cT0-3N0-2a group (Odds ratio 5.54, 95% confidential interval 1.18 to 25.97, p = 0.047). In the cT4N0-2a group, SCU was observed in 8 (57%) of 14 patients, and had no association with M-stage (M0 vs. M1: 60% [6/10] vs. 50% [2/4]; p > 0.05). In the cT1-4N2b-3 group, SCU was observed in 8 (32%) of 25 patients, and was not significantly associated with M-stage (M0 vs. M1: 23% [3/13] vs. 41% [5/12]; p > 0.05).

Conclusions: Our study suggests that SCU on 18F-FDG PET/CT is a potential M-stage indicator. Distant metastasis is usually less common in cT0-3N0-2a stages than cT4 or cN2b-3 stages; however, the presence of distant metastasis should be carefully evaluated if SCU is present on pretreatment PET/CT in cT0-3N0-2a breast cancers.

• Download figure
• Open in new tab
• Download powerpoint