Breast Cancer 18F-ISO-1 Uptake as a Marker of Proliferation Status

Abstract

The σ2 receptor is a potential in vivo target for measuring proliferative status in cancer. The feasibility of using N-(4-(6,7-dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)butyl)-2-(2-18F-fluoroethoxy)-5-methylbenzamide (18F-ISO-1) to image solid tumors in lymphoma, breast cancer, and head and neck cancer has been previously established. Here, we report the results of the first dedicated clinical trial of 18F-ISO-1 in women with primary breast cancer. Our study objective was to determine whether 18F-ISO-1 PET could provide an in vivo measure of tumor proliferative status, and we hypothesized that uptake would correlate with a tissue-based assay of proliferation, namely Ki-67 expression. Methods: Twenty-eight women with 29 primary invasive breast cancers were prospectively enrolled in a clinical trial (NCT02284919) between March 2015 and January 2017. Each received an injection of 278-527 MBq of 18F-ISO-1 and then underwent PET/CT imaging of the breasts 50-55 min later. In vivo uptake of 18F-ISO-1 was quantitated by SUVmax and distribution volume ratios and was compared with ex vivo immunohistochemistry for Ki-67. Wilcoxon rank-sum tests assessed uptake differences across Ki-67 thresholds, and Spearman correlation tested associations between uptake and Ki-67. Results: Tumor SUVmax (median, 2.0 g/mL; range, 1.3-3.3 g/mL), partial-volume-corrected SUVmax, and SUV ratios were tested against Ki-67. Tumors stratified into the high-Ki-67 (≥20%) group had SUVmax greater than the low-Ki-67 (<20%) group (P = 0.02). SUVmax exhibited a positive correlation with Ki-67 across all breast cancer subtypes (ρ = 0.46, P = 0.01, n = 29). Partial-volume-corrected SUVmax was positively correlated with Ki-67 for invasive ductal carcinoma (ρ = 0.51, P = 0.02, n = 21). Tumor-to-normal-tissue ratios and tumor distribution volume ratio did not correlate with Ki-67 (P > 0.05). Conclusion:18F-ISO-1 uptake in breast cancer modestly correlates with an in vitro assay of proliferation.

Keywords: 18F-ISO-1; TMEM-97, proliferation; breast cancer; σ-2.

© 2020 by the Society of Nuclear Medicine and Molecular Imaging.

Figures

FIGURE 1.

Tumor with low proliferative status. A 42-y-old woman with ER-positive/human epidermal growth factor receptor 2–negative primary breast cancer. (A) Tomosynthesis mediolateral oblique projection demonstrates irregular mass with spiculated margins and associated calcifications (arrows). (B) Axial contrast-enhanced T1-weighted subtraction image demonstrates irregular mass in medial breast with heterogeneous enhancement (arrows). (C) Ki-67 staining demonstrates low percentage of actively dividing cells (11%) (×20). (D) Axial 18F-ISO-1 image demonstrates no qualitative uptake in medial breast (arrow, SUVmax of 1.5 g/mL). (E) Corresponding 18F-ISO-1 PET/CT demonstrates biopsy clip marking site of malignancy (arrow). PET and PET/CT images are scaled to 0–5 g/mL SUV, −160 to +240 HU.

FIGURE 2.

Tumor with high proliferative status. A 40-y-old woman with triple-negative breast cancer. (A) Mammographic craniocaudal projection demonstrates high-density irregular mass (arrow) with overlying palpable marker. (B) Axial contrast-enhanced T1-weighted image demonstrates that mass (arrow) is irregular with heterogeneous enhancement, with central signal dropout from biopsy marker. (C) Ki-67 staining demonstrates high percentage of actively dividing cells (74%) (×20). (D) Axial 18F-ISO-1 image demonstrates qualitative uptake at site of malignancy (arrow; SUVmax of 2.6 g/mL) (E) Corresponding CT image demonstrates irregular mass (arrow). PET and CT images are scaled to 0–5 g/mL SUV, −160 to +240 HU.

FIGURE 3.

Plot of SUVmax in groups stratified by Ki-67 below or above 20. (A) SUVmax shows significant difference between patient tumors stratified by low (n = 15) and high (n = 14) Ki-67 values in all 29 tumors. (B) SUVmax stratified by low (n = 8) and high (n = 13) Ki-67 values restricted to IDC (n = 21) show significant differences based on Ki-67 threshold. Center line of each distribution indicates median value; error bars show 95% confidence interval of median. *P < 0.05.

FIGURE 4.

Scatterplots of SUVmax vs. Ki-67 for all 29 tumors. Spearman tests found significant correlations with Ki-67 (ρ = 0.46, P = 0.01). Solid linear regression trend line and dashed 95% confidence intervals are included for reference. TN = triple-negative.