Current status of contemporary diagnostic radiotracers in the management of breast cancer: first steps toward theranostic applications

The clinical relevance and solid prognostic value of overexpression of the HER2 receptor were discovered almost four decades ago [18]. Pivotal adjuvant trials such as HERA and BCIRG-006 in patients with HER2-overexpressing breast cancer paved the way for numerous studies with HER2-targeted drugs in different cancer patient populations [19]. Currently, anti-HER2 therapies are approved using various types of drugs (monoclonal antibodies, tyrosine kinase inhibitors and antibody drug conjugates [ADC]) in various tumor types with overexpression, gene amplifications and mutations of the HER2 receptor.

For breast cancer, HER2-targeted drugs are indicated for treating HER2-positive breast cancer, i.e., cancer with a high degree of HER2 expression. Moreover, the HER2-targeted ADC trastuzumab deruxtecan (TDXd) has recently been approved for breast cancer patients with tumors that are formally not HER2-positive but so-called ‘HER2-low’ [20]. In the DESTINY-Breast04 study, an improvement in overall survival by TDXd versus physicians’ choice of chemotherapy was found in patients with HER2-low metastatic breast cancer [21‐23].

Heterogeneity in HER2 status between and within metastases and receptor conversion over time has been reported in several breast cancer patient cohorts [9]. Reliable assessment of HER2 status and variations in formal HER2-positive lesions have become even more relevant after the identification of the clinical relevance of the HER2-low subgroup, which is estimated to constitute ~ 50% of all breast cancer patients [20].

Several radiopharmaceuticals targeting HER2 have been developed and tested in breast cancer patients, both tracers for imaging with PET and with single-photon emission tomography (SPECT) [24]. Due to the lower image quality and lower resolution, SPECT scans are nowadays sparsely used. PET imaging tracers used to identify HER2-positive lesions include radiolabeled monoclonal antibodies ([⁸⁹Zr]Zr-trastuzumab, [⁸⁹Zr]Zr-pertuzumab [25, 26]) and smaller HER2-specific affinity proteins ([⁶⁸Ga]Ga-ABY-025, [⁶⁸Ga]Ga-HER2-nanobody and [⁶⁸Ga]Ga-DOTA-F(ab’)2-trastuzumab) [27, 28].

Table 1 summarizes the different tracers that have been studied up to now, ⁸⁹Zr-trastuzumab being the one that has been evaluated most and with the most important results to date.

Up to now, various HER2-targeting tracers have been tested in clinical trials in breast cancer patients. All of those are small to moderately sized patient populations, and the prime focus of these studies has been to investigate safety, feasibility and correlation with HER2 status on a tumor biopsy. The ZEPHIR trial is the only one that has investigated the therapy-predictive role of [⁸⁹Zr]Zr-trastuzumab for treatment with the ADC trastuzumab emtansine [29]. In this prospective multicenter trial of 56 patients with HER2-positive mBC, a combination of pretreatment HER2 imaging and early FDG-PET/CT was found to accurately predict morphological treatment response, leading to the conclusion that targeted HER2 imaging could be of great value both for the understanding of tumor heterogeneity and function as an aid in the selection of patients that would benefit from targeted treatment. While this may certainly be the case for trastuzumab emtansine, the much stronger bystander effect of recent ADCs such as TDXd may cause difficulties in using HER2-targeted imaging for the prediction of treatment response, a potential issue which should be evaluated further in future trials.

Of note, one study with [⁸⁹Zr]Zr-trastuzumab [47] and one with [⁸⁹Zr]Zr-pertuzumab [26] showed a number of false-positive PET results for HER2 status. It is with the current knowledge of HER2-low tumors unknown whether these metastases nowadays might have classified as HER2-low tumors.

Several clinical trials with HER2-PET imaging are ongoing, aimed at establishing the role of HER2-PET imaging as a biomarker for treatment selection. In the Affibody-3 trial (NCT03655353), the Affibody molecule based on [⁶⁸Ga]Ga-ABY-025 is used for non-invasive quantification of HER2 expression in patients with primary and metastatic breast cancer, with the primary aim to study the correlation between the HER2 expression measured by [⁶⁸Ga]Ga-ABY-025 PET and standard histopathology. A metabolic response (based on findings from [¹⁸F]F-fluordesoxyglucose [FDG]-PET) activity after anti-HER2 treatment is included, and in an interim analysis of 40 patients presented at the San Antonio Breast Cancer Conference 2022, it was found that tracer uptake could predict metabolic response to treatment better than conventional IHC and that the tracer may be useful as an adjunct diagnostic tool. [⁶⁸Ga]Ga-ABY-025 is now being investigated for its ability to detect HER2-low metastatic breast cancer at our institution, where patients with HER2-low metastatic breast cancer will undergo one HER2-PET followed by tumor biopsies (NCT05619016). Furthermore, the closely related HER2-specific tracer [¹⁸F]F-GE-226 is being investigated in the HERPET study (NCT03827317). The trial determines the uptake in tumors and healthy tissues of [¹⁸F]F-GE-226 to compare the difference between patients with HER2-positive and HER2-negative lesions. An extended trial has been initiated that will further investigate basic properties of the tracer, here called [¹⁸F]F-GEH121224, to guide further clinical development.

Another type of tracer molecule is the single variable domain of a heavy-chain (VHH) antibody specific for HER2. After a successful initial study, two continued studies have now been initiated with [⁶⁸Ga]Ga-NOTA-Anti-HER2 VHH1. In the VUBAR study (NCT03924466), image-based HER2 quantification repeatability will be investigated, and one cohort will undergo [⁶⁸Ga]Ga-NOTA-Anti-HER2 VHH1 PET/CT before and after start of neoadjuvant treatment to study potential for added value of HER2 imaging in the neoadjuvant setting. The second study is an evaluation of uptake of the tracer in brain metastases in breast cancer patients (NCT03331601), where a change in uptake in brain lesions in response to treatment will be assessed.

Imaging of HER2 was pioneered using the HER2-specific antibody trastuzumab, and trials are still investigating scientific questions using this tracer, e.g., to define which patients are likely to respond to targeted HER2 agents using [⁸⁹Zr]Zr-trastuzumab (NCT03321045) or exploring a different PET radionuclide [⁶⁴Cu]Cu-DOTA-trastuzumab (NCT05376878). Also, other antibodies are investigated, such as the site-specifically labeled [⁸⁹Zr]Zr-ss-pertuzumab (NCT04692831).

Currently, two checkpoint inhibitors are approved for breast cancer patients and used in clinical practice. The European Medicines Agency (EMA) and Federal Drug Administration (FDA) approved the programmed death receptor 1 (PD-1) antibody pembrolizumab for the treatment of patients with early triple-negative breast cancer (TNBC) who are receiving neoadjuvant systemic therapy and for patients with advanced TNBC whose tumors express PD-L1 (graded as combined positive score, CPS ≥ 10) by means of the PD-L1 IHC 22C3 pharmDx test. Atezolizumab, a PD-L1 antibody, is approved by the EMA for treating patients with advanced or metastatic TNBC with PD-L1 expression in at least 1% of immune cells with the SP142 Ventana antibody.

Extensive research efforts have searched for the optimal therapy-predictive biomarker for checkpoint inhibitors in solid tumors. Up to now, the only approved biomarker is PD-L1 expression on a tumor biopsy according to pre-defined criteria mentioned above [48]. However, this method also has a suboptimal performance, as concluded in a systematic review and meta-analysis across all solid tumors [49]. For TNBC, several alternative candidate biomarkers have been proposed but have yet to be validated thoroughly enough to be implemented in clinical practice and replace PD-L1 IHC on a tumor biopsy [50]. It has been unequivocally described that PD-L1 expression can change over time from primary to metastatic breast cancer and have varying levels of expression in metastases in different organs [8]. This further limits the therapy-predictive role of PD-L1 IHC for the benefit of checkpoint inhibitors for patients with TNBC. Interestingly, PD-L1 status in the primary tumor has no therapy-predictive role for adding pembrolizumab to a chemotherapy backbone in the neoadjuvant treatment setting, as was observed in the Keynote-522 trial [4].

A few radiotracers representing targets of checkpoint inhibitors, or their downstream intracellular effects, have been studied [51] in the first-in-human (FIH) study with [⁸⁹Zr]Zr-atezolizumab in 25 patients (NCT02453984), including four patients with mTNBC [39]. This study showed that the tracer was feasible and safe. The results indicated clear expression heterogeneity within and between lesions and a promising role as a predictive marker for response to treatment with PD-L1 antibodies. For the patients with mTNBC in this study, the maximum standardized uptake value (SUV_max) was clearly over the SUV_mean in the blood pool measured over the aorta, with varying levels of SUV_max in different organs. Higher SUV_max was related to a better antitumor response, according to RECIST. To the best of our knowledge, this is the only published study on non-invasive imaging with a PD-L1 tracer in patients with breast cancer.

Several preclinical studies have reported warranting results, e.g., with the anti-PD-L1-B11 clone antibody coupled to zirconium-89 [52], [⁸⁹Zr]Zr-Df-bintrafusp-alfa and [⁸⁹Zr]Zr-avelumab [53‐55].

PD-L1 IHC on a tumor biopsy (either from the primary tumor or a metastatic lesion) is the only approved biomarker for adding checkpoint inhibitors in mTNBC. There is, up to now, minimal evidence from clinical trials in patients that reveal the therapy-predictive and tumor biological value of molecular imaging with PD-L1 targeted tracers in patients.

Several other radiolabeled PD-L1 and PD-1 antibodies are currently tested in clinical trials, for example, ⁸⁹Zr-pembrolizumab (active trials: NCT02760225, NCT03065764), [⁸⁹Zr]Zr-durvalumab (NCT03610061, NCT03829007) and [⁸⁹Zr]Zr-avelumab (NCT03514719). We have not identified other clinical trials in patients with mTNBC or other biological subtypes of breast cancer where the role of non-invasive PD-L1 PET imaging is investigated.

Currently, only one trial in breast cancer patients with this radiotracer has been registered at clinicaltrials.gov, one in patients with metastatic lobular breast cancer treated with carboplatin and atezolizumab (NCT04222426). This study was recently terminated after including one patient in the main trial, and a parallel biomarker imaging substudy was closed.

At our institution, we are initiating a clinical trial for patients with metastatic or irresectable TNBC who undergo a baseline [⁸⁹Zr]Zr-atezolizumab PET/CT and a tumor biopsy for PD-L1 IHC before starting with first-line systemic therapy. In this trial (NCT05742269), patients with PD-L1-positive tumors according to PD-L1 PET and/or IHC will receive atezolizumab with a chemotherapy backbone (carboplatin and nab-paclitaxel). The primary endpoint of this PD-L1 PET trial is the statistical agreement between PD-L1 assessed by IHC and PET by estimating a kappa coefficient.

Anti-hormonal drugs have been a cornerstone of systemic breast cancer therapies for decades. Around 70% of breast cancers express ER. ER expression is a highly reliable and systematically used predictive biomarker for treatment with antihormonal therapy [10]. Different antihormonal drugs are available, the largest groups being selective estrogen receptor modulators (SERM) and steroidal/non-steroidal aromatase inhibitors. Changes in ER expression over time, heterogeneity within and between tumors and ER mutations are known complicating factors [9, 56], underscoring the vital need to obtain representative and actual proof of ER status before antihormonal therapies are initiated.

A few ER-targeted imaging tracers have been studied, most of them with the FDA-approved 16α-[¹⁸F]F-fluoro-17β-estradiol ([¹⁸F]F-FES) tracer [57]. Other ER-targeting tracer studies include 4-fluoro-11β-methoxy-16α-[¹⁸F]F-fluoroestradiol ([¹⁸F]F-4FMFES) [44] exhibiting a sensitivity of 95% (89 to 97), a specificity of 80% (66 to 89), a positive predictive value of 93% (87 to 96) and a negative predictive value of 85% (72 to 92) in predicting ER IHC.

The use of [¹⁸F]F-FES-PET is supported by the strongest evidence in targeted nuclear tracers so far, with a meta-analysis supporting the reliability of tracer uptake to predict ER status using IHC on a tumor biopsy. Several limitations have so far impeded a more general use of FES-PET to assess ER status in a patient with metastatic breast cancer. The most apparent reason is the need for increased availability of the tracer and local image acquisition experience and protocols, even when PET camera facilities are present. The washout of prior ER antagonists is a special consideration for [¹⁸F]F-FES-PET imaging. Because [¹⁸F]F-FES-PET measures the regional binding of estrogens to the estrogen receptor, exposure to SERDs will block ER and thereby prevent tracer accumulation leading to a [¹⁸F]F-FES-negative lesion. For this reason, a 6-week washout is indicated for tamoxifen and fulvestrant. This phenomenon is not present in prior treatment with aromatase inhibitors.

In the recently published meta-analysis presenting the results of the IMPACT-trial, the authors present a flowchart for the work-up of patients with metastatic breast cancer, including the role of [¹⁸F]F-FES-PET in guiding treatment decisions [43].

An unresolved challenge in establishing the role of [¹⁸F]F-FES-PET in clinical practice is the need for more evidence regarding the therapy-predictive role of ER status and how the imaging results can guide systemic therapies. Due to disease heterogeneity, it seems rational to integrate a quantitative component in reviewing an [¹⁸F]F-FES-PET scan to describe the overall ER status of the metastases found on CT.

Based on the current evidence, FES-PET can be preferentially used in clinical practice when a biopsy is not feasible or not wanted, for instance, at a later disease stage. Nevertheless, a tumor biopsy will remain the cornerstone of the work-up in metastatic breast cancer since [¹⁸F]F-FES-PET alone will not be able to inform the clinician about additional tumor biological factors such as histology and/or HER2 status.

Currently, around 12 trials incorporating imaging with [¹⁸F]F-FES are ongoing (search in Clinicaltrials.gov, accessed November 21, 2022). Several interesting research questions that are being evaluated in these trials are, among others, the establishment of a pharmacokinetic model and validation of quantitative parameters for clinical practice (NCT05088785); the therapy-predictive value of FES-PET for the efficacy of first-line antihormonal treatment for ER + metastatic breast cancer (NCT02398773); and the combined assessment of FDG-PET and FES-PET in the management of both early and metastatic breast cancer (NCT04692103). In the SONIA trial, which investigates the optimal timing of adding CDK4/6 inhibitors to antihormonal drugs, an imaging substudy SONImage is performed where FES-PET is done at baseline prior to start of systemic treatment (NCT04125277).

The nuclear enzyme PARP1 is central in sensing DNA damage and facilitating repair. Tumors with BRCA1/2 mutations are highly dependent on PARP1 as an alternative DNA repair mechanism. PARPis generate synthetic lethality in tumors with BRCA mutations, resulting in cell cycle arrest and apoptosis [58]. PARPis have proven clinical efficacy for breast cancer management and are approved for use in early and metastatic disease settings for patients with germline mutations in the BRCA1/2 genes. In the OlympiA trial, the efficacy of 1-year treatment of olaparib versus placebo in the adjuvant setting was investigated in patients with early HER2-negative breast cancer who had received (neo)adjuvant chemotherapy, surgery, antihormonal and radiation therapy if indicated. In the olaparib group, overall survival was significantly improved compared to the placebo, with a four-year overall survival of 89.8% in the intervention group versus 86.4% in the placebo group (Δ 3.4%, 95% CI − 0.1% to 6.8%) [5, 59]. For patients with germline BRCA1/2 mutations with metastatic HER2-negative breast cancer, treatment with the PARPis olaparib or talazoparib is associated with more prolonged progression-free survival compared to regular chemotherapeutic treatments [60, 61].

There is a strong biological rationale to assume that the benefit from PARPis is likely not to be confined to patients with germline BRCA1/2 mutations but that this may exist even in those with other defects of homologous recombination (HRD) [62].

Theoretically, PARP1 overexpression should be the best and direct predictive biomarker for PARP1 inhibitors. To date, PARP expression and BRCA status in tumors can be assessed by IHC or genetic sequencing on biopsy samples. However, the results of ICH of PARP1 expression have demonstrated mixed results, suggesting inconsistency of staining procedures [63] with lack of a validated staining protocol that could be applied in clinics. In addition to the time-consuming pathological procedures, such analyses' results depend on the representativity and quality of tumor biopsy samples. Therefore, PET imaging of PARP expression is a candidate predictive biomarker to select patients that could benefit from PARPis. In addition, radiolabeled PARPis could also be used to characterize dynamic changes in tumoral PARP expression during treatment with PARPis or DNA-damaging agents.

Several radiopharmaceuticals targeting PARP have been developed and tested in phase 1 studies, assessed with either intra-operative optical imaging or PET imaging [64]. The first PARP-tracer studied is [¹⁸F]F-fluorthanatrace (FTT), which was evaluated in a FIH study in eight patients with various solid malignancies [65].

A prospective non-randomized clinical trial of 30 participants with early and locally advanced breast cancer studied the correlation of [¹⁸F]F-FTT uptake in different breast cancer subtypes [66]. The SUV from these patients ranged from 2.6 to 11.3 g/mL, independent of cancer subtypes and germline BRCA 1/2 mutation status. The SUV varied greatly in patients with BRCA1/2 pathogenic variants, and the range overlapped values from patients without BRCA 1/2. In this preliminary study, patients with mutations in BRCA1/2 genes showed lower levels of tracer uptake than patients who retained loss of heterozygosity. These results showed that the level of radioligand binding varied considerably across and within investigated breast cancer subtypes, including germline or tumor mutations in BRCA-related genes.

In a follow-up study in patients with mBC receiving PARPi therapy, [¹⁸F]F-FTT uptake in known sites of disease was blocked by PARPi treatment [45]. Drug PARP1 occupancy was also measured by autoradiography radioligand-binding studies of cancer tissue samples using [¹²⁵I]I-KX1 with and without pharmacologic levels of olaparib. [¹²⁵I]I-KX1 binding was suppressed by greater than 80% by olaparib and matched PARPi blockade measured at pre- and post-PARPi PET. Despite the small sample size (four patients; age range, 41–71 years; median age, 52 years; all women; stage III or IV breast cancer), this study demonstrates the potential of [¹⁸F]F-FTT PET to non-invasively quantify PARP1 expression and provides early evidence of using this modality to assess PARPi drug-target engagement, indicating its potential as a biomarker for treatment with PARPis.

Other phase I trials have indicated the safety and feasibility of visualizing tumors/quantifying PARP expression by means of [¹⁸F]F-PARPi in a range of tumor types (excluding breast cancer), for example, in patients with ovarian cancer [67] and head-and-neck cancer [68]. In a preclinical model, the tracer [¹⁸F]F-olaparib showed successful uptake on PET imaging in mice with xenografts overexpressing PARP1 [69].

The results of such studies could be used to identify rational therapy selection with PARPis to maximize the therapeutic benefits while minimizing exposure to toxicities in patients who would not respond, apart from the currently used selection based on germline mutations in BRCA 1/2. Radiolabeled PARPis could also be used to characterize dynamic changes in tumoral PARP expression during treatment with PARPis or DNA-damaging agents, thereby enhancing tumor biological understanding and providing a rationale for the combination of PARPis with other drugs.

For several decades, palliative treatment with bone-seeking isotopes such as radium-223, strontium-90 and samarium-153 ethylenediamine tetra-methylene phosphonic acid (EDTMP) has been used to alleviate symptoms from painful bone metastases [70]. Due to potent analgesics and other palliative treatment options, such as external beam radiotherapy, these are sparsely used nowadays and no longer routinely incorporated into clinical practice guidelines [10, 46].

The concept of targeted delivery of radioisotopes coupled with tumor-specific antibodies is now increasingly being explored in breast cancer management. This approach is in its very beginning, and very few studies in humans are published. Targets of interest include HER2 and new tumor-specific targets, as discussed below.

Up to now, one phase I study in six healthy volunteers and three patients with HER2-positive metastatic breast cancer has been reported with Iodine131 coupled to a HER2 antibody fragment: [¹³¹I]I -GMIB-anti-HER2-VHH1 [71]. No drug-related adverse events were observed in any of the nine subjects, and tracer uptake was noted in metastatic lesions in the breast cancer cohort. An expanded cohort with a phase I/II dose escalation study with the same product is ongoing in patients with metastatic HER2-expressing breast, gastric and gastroesophageal cancer (NCT04467515); the primary outcome is the therapeutic efficacy of the experimental drug. Other HER2-targeted therapeutic isotopes that are currently trialed in early phase clinical trials include a thorium-227-coupled HER2 antibody BAY2701439 for patients with HER2-expressing breast and gastroesophageal cancers that have progressed on earlier HER2-targeted treatment lines (NCT04147819).

Another alpha-therapy antibody is in a phase 1/2 study of [²²⁵Ac]Ac-FPI-1434. The antibody targets insulin growth factor 1 receptor (IGF1R) and is currently being tested in patients with tumors that express IGR1R, including breast cancer (NCT03746431).