Human Epidermal Growth Factor Receptor 2–Targeting [<sup>68</sup>Ga]Ga-ABY-025 PET/CT Predicts Early Metabolic Response in Metastatic Breast Cancer
Ali Alhuseinalkhudhur, Henrik Lindman, Per Liss, Tora Sundin, Fredrik Y. Frejd, Johan Hartman, Victor Vishwanath Iyer, Joachim Feldwisch, Mark Lubberink, Caroline Rönnlund, Vladimir Tolmachev, Irina Velikyan, Jens Nørkær Sørensen
Abstract
Imaging using the human epidermal growth factor receptor 2 (HER2)–binding tracer <sup>68</sup>Ga-labeled Z<sub>HER2:2891</sub>-Cys-MMA-DOTA ([<sup>68</sup>Ga]Ga-ABY-025) was shown to reflect HER2 status determined by immunohistochemistry and in situ hybridization in metastatic breast cancer (MBC). This single-center open-label phase II study investigated how [<sup>68</sup>Ga]Ga-ABY-025 uptake corresponds to biopsy results and early treatment response in both primary breast cancer (PBC) planned for neoadjuvant chemotherapy and MBC. <b>Methods:</b> Forty patients with known positive HER2 status were included: 19 with PBC and 21 with MBC (median, 3 previous treatments). [<sup>68</sup>Ga]Ga-ABY-025 PET/CT, [<sup>18</sup>F]F-FDG PET/CT, and core-needle biopsies from targeted lesions were performed at baseline. [<sup>18</sup>F]F-FDG PET/CT was repeated after 2 cycles of therapy to calculate the directional change in tumor lesion glycolysis (Δ-TLG). The largest lesions (up to 5) were evaluated in all 3 scans per patient. SUVs from [<sup>68</sup>Ga]Ga-ABY-025 PET/CT were compared with the biopsied HER2 status and Δ-TLG by receiver operating characteristic analyses. <b>Results:</b> Trial biopsies were HER2-positive in 31 patients, HER2-negative in 6 patients, and borderline HER2-positive in 3 patients. The [<sup>68</sup>Ga]Ga-ABY-025 PET/CT cutoff SUV<sub>max</sub> of 6.0 predicted a Δ-TLG lower than −25% with 86% sensitivity and 67% specificity in soft-tissue lesions (area under the curve, 0.74 [95% CI, 0.67–0.82]; <i>P</i> = 0.01). Compared with the HER2 status, this cutoff resulted in clinically relevant discordant findings in 12 of 40 patients. Metabolic response (Δ-TLG) was more pronounced in PBC (−71% [95% CI, −58% to −83%]; <i>P</i> < 0.0001) than in MBC (−27% [95% CI, −16% to −38%]; <i>P</i> < 0.0001), but [<sup>68</sup>Ga]Ga-ABY-025 SUV<sub>max</sub> was similar in both with a mean SUV<sub>max</sub> of 9.8 (95% CI, 6.3–13.3) and 13.9 (95% CI, 10.5–17.2), respectively (<i>P</i> = 0.10). In multivariate analysis, global Δ-TLG was positively associated with the number of previous treatments (<i>P</i> = 0.0004) and negatively associated with [<sup>68</sup>Ga]Ga-ABY-025 PET/CT SUV<sub>max</sub> (<i>P</i> = 0.018) but not with HER2 status (<i>P</i> = 0.09). <b>Conclusion:</b> [<sup>68</sup>Ga]Ga-ABY-025 PET/CT predicted early metabolic response to HER2-targeted therapy in HER2-positive breast cancer. Metabolic response was attenuated in recurrent disease. [<sup>68</sup>Ga]Ga-ABY-025 PET/CT appears to provide an estimate of the HER2 expression required to induce tumor metabolic remission by targeted therapies and might be useful as an adjunct diagnostic tool.