Recent Advances in Imaging Steroid Hormone Receptors in Breast Cancer

Assessment of ER Status

Chae et al. recently published the largest prospective study comparing ¹⁸F-fluoroestradiol imaging with ER protein expression (5). They aimed to demonstrate agreement between qualitative ¹⁸F-fluoroestradiol PET/CT interpretation and ER immunohistochemistry. The study population consisted of 85 patients with recurrent or metastatic breast cancer (47 ER-positive, 38 ER-negative). ER positivity was defined as an Allred score of at least 3, which reflects the proportion of positive cells and staining intensity. ¹⁸F-fluoroestradiol uptake was interpreted as positive if it was increased compared with background activity or negative if it was equivocal or decreased relative to background. Positive percentage agreement was 76.6% (95% confidence interval [95%CI], 62.0–87.7; P = 0.0018). Negative percentage agreement was 100% (95%CI, 90.8–100; P = 0.00053). Interreader agreement was high at 0.90 (95%CI, 0.78–1.0). The study concluded that ¹⁸F-fluoroestradiol PET/CT is safe and accurate for determining ER status.

In addition to visual assessment, quantitative ¹⁸F-fluoroestradiol analyses for distinguishing ER-positive from ER-negative lesions have been performed. Using an SUV_max of at least 1.5, the positive and negative percentage agreements were 85.1% and 78.9%, respectively, as published by Chae et al. (5). In a metaanalysis of 5 ¹⁸F-fluoroestradiol imaging studies with ER immunohistochemistry results (143 patients), pooled sensitivity and specificity were 82% and 95%, respectively (6). In another study of 13 heavily pretreated patients with metastatic ER-positive breast cancer, Venema et al. identified 1.54 as the optimal ¹⁸F-fluoroestradiol SUV_max cutoff to discriminate ER-positive from ER-negative lesions (7). In a retrospective study of 46 patients with histologically confirmed invasive breast cancer reported by Yang et al., the optimal ¹⁸F-fluoroestradiol threshold was 1.21 for SUV_mean and 1.82 for SUV_max (8). Thus, including previously published data, optimal ¹⁸F-fluoroestradiol SUV_max thresholds for distinguishing between ER-positive and ER-negative lesions ranged from 1.5 to 2.0 (1–4,6).

Recent studies have also confirmed earlier evidence of a correlation between ¹⁸F-fluoroestradiol uptake and ER protein expression. A strong association between ¹⁸F-fluoroestradiol SUV_max and the Allred score (correlation coefficient, 0.83; P < 0.0001) was reported by Chae et al. (5). Similarly, Venema et al. reported a significant correlation between semiquantitative ER measurement using the H-score (0–300) and ¹⁸F-fluoroestradiol SUV_max (R² = 0.78; P = 0.01) (7). Furthermore, Yang et al. found a correlation coefficient of 0.726 (P < 0.001) between ER expression (scored 0, 1+, 2+, or 3+) and ¹⁸F-fluoroestradiol SUV_mean (using a 40% SUV_max threshold), which was slightly higher than the correlation observed using SUV_max (r = 0.691; P < 0.001) (8). Gupta et al. reported a correlation coefficient of 0.767 (95%CI, 0.266–0.942; P = 0.0096) between ¹⁸F-fluoroestradiol SUV_max and ER (scored 15%, 40%, 90%, and 100%) (9). Overall, a strong correlation has been established between imaging and pathology despite different methods for ER protein quantification.

Binding specificity is necessary for establishing the analytic validity of ¹⁸F-fluoroestradiol imaging. Evidence inferring that ¹⁸F-fluoroestradiol specifically binds to ER, similarly to endogenous estradiol, has been provided through clinical studies correlating ¹⁸F-fluoroestradiol uptake with ER protein expression, in vitro receptor ligand binding assays, and receptor blocking studies using coinjected nonradiolabeled estradiol. To confirm the binding specificity of ¹⁸F-fluoroestradiol, Salem et al. used—in engineered breast cancer cell lines—a mutagenesis approach that prevents ER from binding estradiol (10). Through substitution of glycine 521 for arginine in the ER ligand-binding pocket, ¹⁸F-fluoroestradiol binding in cells and uptake in tumor xenografts was completely abolished. These results suggest that ¹⁸F-fluoroestradiol retention in ER-positive breast cancer is strictly dependent on an intact ligand-binding pocket with high specificity.

Given the importance of the ER ligand binding domain for ¹⁸F-fluoroestradiol binding, frequently identified hot-spot mutations within this region of ER in patients with metastatic breast cancer raised questions regarding a potential effect on the diagnostic accuracy of ¹⁸F-fluoroestradiol imaging. These mutations in the ER gene (ESR1) are biologically significant since they exhibit constitutively active ER transcriptional function with partial resistance to endocrine therapy and reduced binding affinity to estradiol, and they correlate with reduced patient survival (11). Kumar et al. created breast cancer cell lines expressing wild-type or mutant ER to test whether activating ESR1 mutations impact the accuracy of ¹⁸F-fluoroestradiol PET imaging for identifying ER-positive lesions (12). The study showed that ESR1 mutations occurring at tyrosine 537 are not detrimental to ¹⁸F-fluoroestradiol binding or uptake in cells and tumor xenografts. These results support ¹⁸F-fluoroestradiol imaging as a valid option for ER-positive breast cancer with these ESR1 mutations.

Additional phase II data regarding the diagnostic accuracy of ¹⁸F-fluoroestradiol imaging have been published by Paquette et al. (13). The performance of a newer ER-targeted radiopharmaceutical, 4-fluoro-11β-methoxy-16α-¹⁸F-fluoroestradiol, or ¹⁸F-4FMfluoroestradiol, was compared with ¹⁸F-fluoroestradiol in 31 patients with ER-positive breast cancer. Decreased background uptake on PET/CT and improved tumor contrast were observed for ¹⁸F-4FMfluoroestradiol compared with ¹⁸F-fluoroestradiol, which may improve diagnostic confidence and reduce false-negative results.

Imaging ER Heterogeneity in Metastatic Disease

Site-to-site ER heterogeneity across metastases is evident with ¹⁸F-fluoroestradiol imaging. Nienhuis et al. retrospectively analyzed ¹⁸F-fluoroestradiol PET/CT scans for 91 patients (1,617 lesions) with metastatic ER-positive breast cancer (14). They found that 36% of patients had both ¹⁸F-fluoroestradiol–positive and ¹⁸F-fluoroestradiol–negative lesions, which is within the previously published range of 28%–45% (1–4). Aside from the interesting biologic phenomena this finding may represent, such as clonal evolution and influence of the metastatic tumor microenvironment, it has potentially important clinical implications. Since targeting the site of biopsy to confirm metastatic disease in a patient with multiple lesions is based on lesion size and on safety or accessibility, different treatment approaches may be pursued with differing outcomes in patients with site-to-site heterogeneity in ER expression.

Image-Guided Dosing of ER Antagonists

¹⁸F-fluoroestradiol imaging allows quantification of receptor binding during ER antagonist treatment. This approach can be used in dose escalation trials to maximize treatment benefits while minimizing side effects. Using a preclinical ER-positive breast cancer model, Heidari et al. demonstrated reduced tumor uptake of ¹⁸F-fluoroestradiol correlating with decreased ER expression in a dose-dependent manner with fulvestrant, thus tracking early efficacy of ER blockade and degradation (15). This approach was investigated by van Kruchten et al. in a prospective study of 16 patients with metastatic ER-positive breast cancer (16). They tested whether the current standard dosing of fulvestrant optimally blocks tumor ER availability by performing ¹⁸F-fluoroestradiol PET/CT at baseline and during treatment. Incomplete blockade of ER was observed in 38% of patients and was associated with early disease progression. Thus, therapeutic monitoring via ¹⁸F-fluoroestradiol imaging could reveal ineffective ER antagonist dosing and potentially allow for dose adjustment or switching to an alternative drug.

During novel drug development, serial ¹⁸F-fluoroestradiol imaging can be useful for selecting the optimal dosage to achieve ER suppression while minimizing side effects. GDC-0810, a novel ER antagonist and degrader, was investigated in a phase I trial involving 30 patients with ER-positive metastatic breast cancer (17). The optimal dosage for phase II evaluation was determined by measuring the reduction in ¹⁸F-fluoroestradiol uptake during treatment, reflecting ER occupancy and downregulation. As a substudy of a therapeutic trial of Z-endoxifen (an active metabolite of tamoxifen), ¹⁸F-fluoroestradiol PET/CT was performed at baseline and 1–5 d after treatment initiation in 8 patients with a variety of ER-positive solid tumors (18). The study found a 33.6% decrease (P = 0.0078) in the average SUV_max as early as 1 d after Z-endoxifen administration. Thus, ¹⁸F-fluoroestradiol imaging could serve as a useful pharmacodynamic marker for evaluating ER antagonist therapies.

Predicting Endocrine Therapy Response

Another clinically relevant application of ¹⁸F-fluoroestradiol imaging is predicting response to endocrine therapy. In the neoadjuvant setting, 2 recent trials have incorporated baseline ¹⁸F-fluoroestradiol PET/CT in the study design. The Neo-ALL-IN trial tested the aromatase inhibitor, letrozole, in combination with a human epidermal growth factor receptor 2 (HER2)–targeted drug, lapatinib, in postmenopausal patients with stage II–III, ER-positive, HER2-positive breast cancer administered before surgery (19). Baseline ¹⁸F-fluoroestradiol PET/CT was performed for 24 patients. An SUV_max of greater than 5.5 significantly correlated with clinical overall response rate (P = 0.007). For the NEOCENT trial, postmenopausal patients with ER-rich primary invasive breast cancer were randomized to either neoadjuvant chemotherapy (5-fluorouracil, epirubicin, and cyclophosphamide) or neoadjuvant endocrine therapy (letrozole) before surgery (20). Baseline ¹⁸F-fluoroestradiol PET/CT was performed for 26 patients (13 randomized to chemotherapy, 13 randomized to endocrine therapy). Two patients in the chemotherapy cohort had visually negative ¹⁸F-fluoroestradiol imaging, and both achieved a pathologic complete response. None of the 10 evaluable patients in the chemotherapy cohort with ¹⁸F-fluoroestradiol–avid tumors achieved a pathologic complete response. None of the 5 patients in the endocrine therapy cohort with an SUV_max of less than 7.3 were pathologic responders, whereas 5 of the 7 patients in the chemotherapy group with an SUV_max of less than 7.3 were pathologic responders (P = 0.03). These data imply that neoadjuvant chemotherapy could be superior to endocrine therapy in tumors with low ¹⁸F-fluoroestradiol uptake, indicating nonfunctional ER. Together, these studies suggest that ¹⁸F-fluoroestradiol imaging could be helpful for neoadjuvant therapy selection.

In the metastatic setting, a study by van Kruchten et al. assessed the predictive value of ¹⁸F-fluoroestradiol imaging for response to estradiol treatment, a therapeutic approach for extensively pretreated patients (21). Patients with acquired endocrine-resistant metastatic ER-positive breast cancer and progression after at least 2 lines of therapy underwent baseline ¹⁸F-fluoroestradiol PET/CT followed by estradiol treatment. For 15 evaluable patients, the positive predictive value was 60%, the negative predictive value was 80%, and the area under the curve was 0.62 using a median SUV_max cutoff of 1.5 for discriminating clinical benefit from progressive disease. The study highlighted the potential for ¹⁸F-fluoroestradiol imaging to identify patients with acquired endocrine resistance who are unlikely to benefit from estradiol therapy indicated by poor ¹⁸F-fluoroestradiol uptake. This study adds to landmark studies from Washington University and the University of Washington, which demonstrated a predictive value of ¹⁸F-fluoroestradiol imaging for endocrine therapy benefit (1–4).

Current multicenter trials aim to validate the predictive accuracy for endocrine therapy response in the metastatic setting. The ECOG/ACRIN EAI142 trial (NCT02398773) in the United States is designed to determine the negative predictive value of ¹⁸F-fluoroestradiol PET/CT for clinical benefit at 6 mo in 99 patients with ER-positive metastatic breast cancer treated with first-line endocrine therapy. Ongoing European multicenter studies include the TRANSCAN ET-FES randomized trial (EUDRACT 2013-000-287-29) and the IMPACT-MBC trial (NCT01957332) in The Netherlands.

Prognostic Significance of ¹⁸F-Fluoroestradiol Imaging

Additional prognostic information may be gained from both ¹⁸F-fluoroestradiol and ¹⁸F-FDG imaging. Kurland et al. investigated ¹⁸F-fluoroestradiol and ¹⁸F-FDG as combined predictors of progression-free-survival in 84 evaluable patients with ER-positive breast cancer undergoing endocrine therapy for primary, recurrent, or metastatic disease (22). The study identified 3 groups with distinct outcomes. Metabolically indolent disease (low ¹⁸F-FDG uptake) corresponded with the longest median progression-free-survival (26.1 mo; 95%CI, 11.2–49.7). Metabolically aggressive but strongly ER-positive disease (¹⁸F-FDG high; ¹⁸F-fluoroestradiol high) corresponded with intermediate progression-free-survival (7.9 mo; 95%CI, 5.6–11.8). Metabolically aggressive and weakly ER-positive disease (¹⁸F-FDG high; ¹⁸F-fluoroestradiol low) corresponded with the shortest progression-free-survival (3.3 mo; 95%CI, 1.4—not evaluable). Thus, prognostic information provided by ¹⁸F-fluoroestradiol imaging may be most impactful for patients with metabolically aggressive disease.

Future of ER Imaging

Robust evidence exists to support the clinical use of ¹⁸F-fluoroestradiol imaging to determine ER status in metastatic breast cancer for endocrine therapy selection. ¹⁸F-fluoroestradiol is currently clinically approved for this indication only in France. In 2019, Zionexa submitted a new-drug application for ¹⁸F-fluoroestradiol (EstroTep) to the U.S. Food and Drug Administration. After approval, important next steps for clinical ¹⁸F-fluoroestradiol imaging include development of practice guidelines detailing patient preparation, scan parameters, expected physiologic distribution, tracer uptake quantification, scan interpretation, and standardized reporting (23).

Another critical step will be establishing payer coverage. In a simulation study, ¹⁸F-fluoroestradiol PET/CT appears to be cost effective as an initial imaging test for diagnosing ER-positive metastatic breast cancer by decreasing the number of biopsies and false-negative results compared with conventional work-up (24). Additional cost-effectiveness studies are needed.

Future of PR Imaging

Research continues into optimizing ¹⁸F-FFNP production methods and developing novel agents for PR imaging. A new radiosynthesis technique has been reported to reduce the time required for ¹⁸F-FFNP synthesis, as would be advantageous for high-volume facilities (29). Two new ¹⁸F-labeled ethisterone derivatives, ¹⁸F-EAEF and ¹⁸F-FPTT, have been developed and tested in preclinical models (30,31). Additional preclinical and phase I studies will be needed to evaluate potential advantages of these agents over ¹⁸F-FFNP. As PR function in breast cancer continues to be revealed, ¹⁸F-FFNP imaging may be useful for identifying PR-rich targets for combined ER and PR antagonists.