Can Molecular Imaging Find a Path to Navigate Evolving Breast Cancer Treatments?

In this issue of Clinical Cancer Research, Iqbal and colleagues report their findings testing [¹⁸F]fluoroestradiol (FES) PET/CT to measure regional estrogen receptor (ER) expression and drug occupancy in patients treated with rintodestrant, an oral selective estrogen receptor degrader (SERD; ref. 1). Endocrine therapy of breast cancer is one of the oldest and most successful approaches to molecularly target cancer treatment. Now, four decades after the development of effective antiestrogens and the ability to measure steroid receptor expression in tumor tissue, the landscape of both treatment and diagnosis continues to evolve. New drugs and drug combinations continue to improve treatment outcomes, and advances in genomic assays can identify ER mutation variants with distinct biologic behavior that requires specific treatment choices (2). Our improved understanding of the biology and pharmacology of ER+ breast cancer calls for new and more sophisticated diagnostics to guide drug selection. The Netherlands’ molecular imaging team studied a cohort of 16 patients using FES and [¹⁸F]fluorodeoxyglucose (FDG) PET/CT prior to therapy, and obtained FES PET/CT during the first 4 weeks of treatment, and at the time of progression. The investigators found that all lesions that were FES-positive pre-therapy were FES-negative after 4 weeks of rintodestrant. Quite strikingly, they found that pre-therapy within-patient lesion-to-lesion heterogeneity, defined as of FDG-positive lesions that were not also positive by FES PET—i.e., lesions with active tumor that that did not exhibit ER expression by FES PET—were associated with poor progression-free survival (PFS).

FES, a radiopharmaceutical estradiol analog binds specifically to the ER and can thus detect ER-expressing breast cancer. Its FDA's approval as a PET agent for diagnosing ER-expressing breast cancer was supported by developmental and translational research over 40 years (3), and its accuracy has been confirmed by 2 well powered prospective trials that used IHC assay of biopsy material as the reference standard (4, 5). Unlike biopsy, however, FES can measure ER expression across the full burden of disease (6) and, as an approach that works by tracer radioligand binding, serial FES PET can assess the degree of receptor occupancy by ER blocking agents such as SERDs (7–9). Prior studies supported the ability of FES PET to predict response to endocrine therapy, showing that the presence of disease sites with quantitatively low or qualitatively absent FES uptake portended poor treatment outcomes (7–9). In the current paper by Iqbal and colleagues, the authors noted a similar pattern for rintodestrant, finding poor PFS for patients with 25% or greater disease sites with qualitatively absent FES. An important component of this analysis is the use of paired FDG and FES scans to identify active metastases that lack the ability to bind estrogen but which have aberrant glucose metabolism. Prior studies have also used this same approach and found that using FDG PET—an independent and reliable way to localize metastatic disease sites using the same imaging modality—is an important companion to FES PET for identifying heterogeneous ER expression (10). The current CCR paper's introduction of a practical and clinical practical measure of ER expression heterogeneity to predict a lack of response to SERDS, is an important outcome and innovation of this study.

The ability to determine ER binding is especially relevant for the use of SERDs to treat metastatic breast cancer harboring ESR1 mutations. For ER mutations that lead to constitutive activation even in the absence of hormone binding, the receptor binding pocket provides an opportunity to target agents designed to lead to ER degradation (i.e., SERDs) to interrupt the tumor growth signal mediated by these mutated ER molecules (11, 2). FES PET provides not only a way to assure that ER binding is present in the tumor, but also provides a means to assess ER occupancy by pharmacologic doses of SERDs. Effective SERD treatment should lead to either blockade of the ER binding or loss of ER, which in either case leads to a loss of FES retention. As such, the disappearance of FES after the initiation of a SERD indicates adequate dosing and is predictive of response (12, 13). Serial FES PET demonstrated the variability of ER blockade by fulvestrant at its initial post-approval suggested dose (14), and has also been used to guide dosing in early phase trials of other SERDs, where complete loss of FES uptake on treatment indicates adequate dosing (7–9). This same approach was also used in the Iqbal paper and confirmed complete FES blockade at the recommended doses of rintodestrant. FES was also used to guide dosing in early studies of the SERD, elacestrant, now FDA-approved for treating metastatic breast cancers harboring ESR1 mutations (15).

The authors also provide early data on an issue of clinical importance, the rate at which tumor ER expression and estrogen binding recover after discontinuing the SERD. This has important implications for the use of FES PET in patients who progress on a SERD, where imaging could assess whether ER expression has changed and inform the choice of subsequent treatment. However, imaging too soon after SERD discontinuation could lead to a falsely negative FES PET if the SERD and/or its effect on ER expression is still present. This consideration led the FDA to recommend a conservative 28-week washout period after discontinuing fulvestrant prior to considering FES PET, making it highly impractical to use this test to help make clinical decisions on treatment post-SERD. In the Iqbal study, the authors showed that all patients undergoing FES PET at progression at least 5 drug clearance half-lives after discontinuation (5 days for rintodestrant) recovered FES uptake at all sites of disease seen pre-therapy, but imaging at earlier times did not. These data provide important guidance on using FES PET to reassess breast cancer ER expression after SERD discontinuation, suggesting that the shorter half-life of oral SERDs compared with fulvestrant can likely make post-SERD FES PET clinically practical.

One surprising result of the study is the finding that quantitatively higher FDG uptake pre-therapy portended longer PFS in patients treated with rintodestrant, while patients with low pre-therapy FDG had short PFS. This is a somewhat paradoxical result that is counter studies showing that higher FDG uptake is associated with a more aggressive metastatic phenotype (16, 17) and a prior study using both FDG and FES in metastatic breast treated with endocrine therapy (10) that showed lower uptake was associated with longer PFS. Further study is needed to understand whether these conflicting findings are related to a population selected for ESR1-mutant breast cancer and the use of SERDs in the new study, or simply random differences related to patient selection in small studies.

Overall, the Iqbal study makes an important contribution towards defining the clinical utility of FES PET for ER variants and SERDs, adding to the body of work indicating the value of this new clinical tool for guiding the treatment of ER-expressing breast cancer.

H.M. Linden reports grants and personal fees from GE HealthCare, Sanofi, and Eli Lilly and grants from Zymeworks, Veru Inc., and Zeno Therapeutics outside the submitted work. D.A. Mankoff reports personal fees from GE HealthCare outside the submitted work.

H.M. Linden is supported by NIH research funding FHCC core grant P30CA015704. D.A. Mankoff is supported by the NIH Abramson Cancer Center support grant P30CA016520.