Head-to-Head Comparison between [68Ga]Ga-DOTA.SA.FAPi and [18F]F-FDG PET/CT Imaging in Patients with Breast Cancer

This study aimed to compare the diagnostic performance of [68Ga]Ga-DOTA.SA.FAPi with that of [18F]F-FDG PET/CT in detecting primary and metastatic lesions of breast cancer. [18F]F-FDG and [68Ga]Ga-DOTA.SA.FAPi PET/CT scans of histologically proven breast cancer patients were compared according to patient-based and lesion-based analysis. Forty-seven patients with a mean age of 44.8 ± 9.9 years (range: 31–66 years) were evaluated. A total of 85% of patients had invasive ductal carcinoma, and 15% had invasive lobular carcinoma. The tracer uptake [SULpeak, SULavg, and the median tumor-to-background ratio (TBR)] was significantly higher in [68Ga]Ga-DOTA.SA.FAPi than with [18F]F-FDG PET/CT for lymph nodes, pleural metastases, and liver lesions (p < 0.05). However, for brain metastasis, only the median TBR was significantly higher (p < 0.05) compared to [18F]F-FDG. In patient-based analysis the sensitivity of [68Ga]Ga-DOTA.SA.FAPi PET/CT was higher, but not significant than that of [18F]F-FDG PET/CT in the detection of both primary tumors and metastatic lesions. According to lesion-based analysis, on diagnostic CT, 47 patients had 44 primary tumors, 248 lymph nodes, 15 pleural, 88 liver, and 42 brain metastases. [68Ga]Ga-DOTA.SA.FAPi scan identified more abnormal lesions than [18F]F-FDG in all the primary and metastatic sites with a maximum marked difference in the primary site [88.6% vs. 81.8%; p-0.001], lymph nodes [89.1% vs. 83.8%; p-0.0001], pleural metastases [93.3% vs. 73%; p-0.096] and brain metastasis [100% vs. 59.5%; p-0.0001]. [68Ga]Ga-DOTA.SA.FAPi PET/CT was superior to [18F]F-FDG PET/CT in the imaging of breast cancers.


Introduction
Breast cancer is the most common malignancy in women. With the advent of several advanced imaging and treatment options, death due to breast cancer has declined over time [1]. Among the imaging modalities, [ 18 F]F-FDG PET/CT is the most widely used systemic option for the diagnosis, staging, response assessment, and prognostication of breast cancers. As compared to other conventional imaging modalities, the sensitivity of [ 18 F]F-FDG is remarkable with an estimated sensitivity of 97%. However, it exhibits a rather low specificity of 77% in breast cancer patients due to the high false positive results of uptake in inflammatory lymph nodes, and specifically low detection rates in micrometastases and sclerotic healed bone lesions [2].
Hence, there is an unmet need to evaluate new molecular whole-body imaging techniques based on cell-specific oncological targets from a theranostic point of view in endstage breast cancer patients. Since 2018, imaging fibroblast activation protein (FAP) expression in tumor microenvironment utilizing fibroblast activation protein inhibitors (FAPi) labeled with gallium-68 has gained interest and an increasing number of studies have attempted to explore this modality in various cancers. Huang et al. [3] observed FAP expression in breast cancer but not in normal breast tissue and proved that significant FAP expression was paralleled by increased tumor growth rates in a mouse model of human breast cancer. Their further investigations lead to significant findings that the proteolytic activity of FAP participates in matrix degradation, but other functions of the protein stimulate increased tumor growth.
One of the first investigations found [ 68 Ga]Ga-FAPI-04 PET/CT useful in the detection of breast cancers [4]. Another retrospective study involving 48 breast cancer patients found that [ 68 Ga]Ga-FAPI-04 PET/CT imaging revealed more lesions in all classified regions and had higher uptake values than [ 18 F]F-FDG PET/CT and reported a remarkable improvement in the sensitivity and specificity compared to [ 18 F]F-FDG [5]. At our institute, we use the squaric acid-based FAP inhibitor DOTA.SA.FAPi, which additionally has a theranostic option of treatment with DOTAGA. (SA.FAPi) 2 dimer [6-12]. However, not much has been reported on this molecule in breast cancer. Accordingly, in this study, we compare the diagnostic efficacy of [ 68 Ga]Ga-DOTA.SA.FAPi PET/CT with [ 18 F]F-FDG PET/CT in breast cancer patients.

Patients
A total of 47 breast cancer patients were included in the study and underwent [ 18 F]F-FDG and [ 68 Ga]Ga-DOTA.SA.FAPi PET/CT scans. The mean age of the patients was 44.8 ± 9.9 years (range: 31-66 years). A total of 85% had invasive ductal carcinoma, and 15% had invasive lobular carcinoma. Twelve patients had the triple-negative disease, 17 had HER2-positive, and 18 had HER2-negative disease. A total of 95.5% of patients had distant metastases, and in three patients the disease was locally advanced. Forty patients underwent surgery and the remaining did not. The majority (80.8%, n = 38) of the patients underwent chemotherapy and five among them underwent two lines of chemotherapy. Local RT, hormonal treatment, immunotherapy, tyrosine kinase inhibitors, and palliative RT for distant metastases were administered to 23.6%, 17.0%, 12.7%, 10.6%, and 8.5% of patients, respectively. Scans were conducted in 39 patients for re-staging after anti-cancer treatments and in eight patients for initial staging/workup.

Patient-Based Detection Rate Analysis
The primary tumor detection rates for [ 68  In one patient ( Figure 1) with invasive ductal carcinoma (ER-/PR-/HER2+), on imaging, [ 18 F]F-FDG PET/CT (A; red arrow) showed uptake in the pectoralis major and minor muscle suggesting inflammation. The same was not observed on [ 68 Ga]Ga-DOTA.SA.FAPi PET/CT (B). A well-defined cystic lesion with peripheral FDG uptake is noted in the right anterior chest wall at the post-operative site (right modified radical mastectomy) (yellow arrow) but no uptake was observed on [ 68 Ga]Ga-DOTA.SA.FAPi PET scan (B). In this patient FAPi could detect all the lesions corresponding to CT, but [ 18 F]F-FDG detected false positive findings in the primary site and increased muscle uptake. Another example of FDG uptake in fibroadenoma was [ 68 Ga]Ga-DOTA.SA.FAPi negative is depicted in Figure 2.   showed diffuse non-specific radiotracer uptake in the entire axial skeleton, which is likely reactive (FP) (Figure 3).

Comparison of Uptake and TBRs in Tumor Lesions
The mean SULpeak and average values of lymph node metastases, liver metastases, and pleural metastases were significantly higher on [ 68 Ga]Ga-DOTA.SA.FAPi than [ 18 F]F-FDG imaging (p < 0.05). However, SULpeak values of the primary breast tumors, lung metastases, brain metastases, and bone metastases did not exhibit a statistically significant difference between the two imaging techniques (p > 0.05) (Table 1).
Interestingly, there was no significant difference in the SULpeak and average values in brain metastases, but remarkably high TBR values were observed on   (Figures 4 and 5).

Comparison of Uptake and TBRs in Tumor Lesions
The mean SULpeak and average values of lymph node metastases, liver metastases, and pleural metastases were significantly higher on [ 68 Ga]Ga-DOTA.SA.FAPi than [ 18 F]F-FDG imaging (p < 0.05). However, SULpeak values of the primary breast tumors, lung metastases, brain metastases, and bone metastases did not exhibit a statistically significant difference between the two imaging techniques (p > 0.05) (Table 1).
Interestingly, there was no significant difference in the SULpeak and average values in brain metastases, but remarkably high TBR values were observed on [ 68 Ga]Ga-DOTA.SA.FAPi compared to that of [ 18 F]F-FDG PET. [ 68 Ga]Ga-DOTA.SA.FAPi imaging showed 10 brain metastases in one patient, while [ 18 F]F-FDG showed only four lesions due to high background activity (Figure 1).
In addition to brain metastases, [ 68 Ga]Ga-DOTA.SA.FAPi imaging yielded significantly higher TBRs in breast lesions as well as hepatic, bone, and lung metastases com- In addition to brain metastases, [ 68 Ga]Ga-DOTA.SA.FAPi imaging yielded significantly higher TBRs in breast lesions as well as hepatic, bone, and lung metastases compared to [ 18 F]F-FDG (p < 0.05) ( Table 1). The median TBR ratio was favorable for [ 68 Ga]Ga-DOTA.SA.FAPi over [ 18 F]F-FDG in the primary and all metastatic lesions but statistically in case of lymph node, pleural metastases, liver and brain metastases (p < 0.05).
This retrospective study was conducted to compare the diagnostic value of [ 18  Tchou et al. [20] evaluated the expression of FAP in a panel of 52 human breast tumor samples using a combination of immunohistochemistry analyses. They observed an abundant expression of FAP in >70% of tumor samples even without a significant correlation with clinicopathologic factors. Tchou et al. [20] additionally investigated five fresh human breast tumor specimens and demonstrated that some of these FAP+ CD45+ cells were CD11b + CD14 + MHC-II+ indicating that they were likely tumor-associated macrophages (TAMs). Significant observation emerged that human breast TAMs also expressed FAP. These results strengthened the fact that future FAP-directed therapy may have dual therapeutic benefits targeting both stromal mesenchymal cells and immune cells such as TAMs. On the contrary, interestingly, a study by Ariga et al. [21] suggested that FAP expression in breast cancer was inversely correlated with breast cancer prognosis.
Similar to Komek et al. [4], in our study [ 68 Ga]Ga-DOTA.SA.FAPi detected more primary tumor lesions, and higher TBR than [ 18 F]F-FDG, but unlike Komek et al. [4], there was no difference in the uptake values. In one of our cases, avid FDG uptake was observed in a patient with fibroadenoma in the right breast but did not show FAP expression. Another patient developed post-operative seroma development at the surgical site of the right breast region. While Regarding the bone metastases, due to the high disease burden, it was impractical to compare each lesion; however, the visual analysis did not reveal any difference in the detection rate between the two scans except for two patients who had diffuse nonspecific uptake on [ 18 F]F-FDG scan. Low activity uptake in the bone is an inflammatory sequel and the findings are consistent with findings from other studies in various cancers. The superiority of [ 68 Ga]Ga-DOTA.SA.FAPi is straightforward with a clear advantage over [ 18 F]F-FDG PET/CT scans in the detection of brain metastases. Although this pattern is expected because of the FDG distribution in the brain, it represents a unique field for FAPi.

Limitations
Our study had some limitations. CECT scans were not acquired in all patients. Gold standard histopathology validation was performed in only the locoregional tumor lesions for all patients as it was impractical to assess biopsy from each lesion.

Future Prospects
More work in a larger patient population is needed to dissect the role of FAP in various histopathologic variants and hormone-receptor status within the tumor microenvironment and explore its role as a potentially targetable molecule in breast cancer treatment.

Materials and Methods
This retrospective study was approved by the institute ethics committee of All India Institute of Medical Sciences, New Delhi, India. Patients were enrolled between April 2020 and August 2022. The study was conducted in the Department of Nuclear Medicine in collaboration with the Department of Medical Oncology at All India Institute of Medical Sciences and the Department of Chemistry, Johannes Gutenberg University, Mainz, Germany, which provided the DOTA.SA.FAPi molecule. Radiolabeling of [ 68 Ga]Ga-DOTA.SA.FAPi was conducted as detailed in our previous publications [9][10][11][12]. In short, a mixture of 10 nmol of DOTA.SA.FAPi precursor, 1 mL ammonium acetate buffer pH 4, and [ 68 Ga]Gachloride solution were heated at 95 • C for 10 min. After heating, the radiolabeled product was eluted through a Sep-Pak C18 Plus Light cartridge with 50% ethanol followed by 10 mL of normal saline. Quality control with sodium citrate buffer pH4 yielded > 95% radiochemical purity.

Patient Selection
Breast cancer patients with locally advanced or metastatic stage disease who underwent both [ 68 Ga]Ga-DOTA.SA.FAPi PET/CT and [ 18 F]F-FDG PET/CT scans within a 15-day time frame were included in the retrospective study. The inclusion criteria of the patients were; (i) patients > 18 years of age, (ii) breast cancer confirmed by histopathological evaluation, and (iii) patients who did not receive radiotherapy or chemotherapy within six weeks. Patients with a known inflammatory condition, dual malignancies, pregnant patients, and those unwilling to undergo two PET/CT scans were excluded from the study. Histopathological (HPE) evaluation was performed in all patients from the primary and suspicious lymph nodes. Immunohistochemistry analysis included evaluation of estrogen receptor (ER), progesterone receptor (PR), Ki-67, and HER2/neu. In patients with additional lesions on any scan, the HPE was evaluated in the corresponding lesion.

PET/CT Acquisition
According to the eligibility criteria, 47 patients were included in the study. Patients fasted at least 6 h before the [ 18  Ga]Ga-DOTA.SA.FAPi were injected. Scans were acquired on a 128-slice GE Discovery 710 × 128 Slice PET/CT Scanner with a 40-mm detector at a 0.35-s rotation speed after 60 min of intravenous injection of both radiotracers. For all acquisitions, the patient was positioned in a supine position. The scans involved an initial scout image to define the field of view, followed by a CT and a PET scan. The CT scan involved a diagnostic dose CT with 300 to 380 mAs, 120 kVp, slice thickness 3.75 mm, and pitch 0.6. Spot views were acquired if required in brain metastases patients with a slice thickness of 1.25 mm on CT at 120 kVp, 300-380 mAs, and a pitch of 0.6. Images were processed and analyzed on a GE Xeleris workstation. Image acquisition and analysis for [ 18

Data Interpretation
To assess the diagnostic ability of [ 68 Ga]Ga-DOTA.SA.FAPi, a patient-based analysis was conducted in both primary and all metastatic lesions. Lesion-based analysis was feasible in the primary, lymph node, liver metastases, and brain metastases.
[ 18 F]F-FDG and [ 68 Ga]Ga-DOTA.SA.FAPi PET/CT scans were independently reviewed and processed by two nuclear medicine physicians with more than 15 years of experience in interpreting PET/CT data. They were blinded to the clinical history and HPE status of the patients. Any differences in the interpretation were discussed and settled with mutual agreement.

Data Analysis and Processing
The qualitative analysis involved visual judgment of radiotracer uptake which was validated by the morphological findings on diagnostic CT which was considered as the reference standard. For quantitative comparisons between the radiotracers, a 3D autocontour ROI at a 40% threshold of SULpeak was carefully drawn around the site of [ 18 F]F-FDG respective [ 68 Ga]Ga-DOTA.SA.FAPi expressing lesions on transaxial images. The ROIs were presented as standardized uptake value (SUV) corrected for lean body mass: SULpeak and SULavg to quantitatively compare the uptake in the lesions between the radiotracers. The SUV values (peak, average, median, and range) were recorded for each site for both [ 18 F]F-FDG and [ 68 Ga]Ga-DOTA.SA.FAPi. The uptakes in the lesions on both scans were compared with the morphological features/characteristic on the CT counterpart. Tumor-to-background ratio (TBR) was calculated by dividing the SULpeak of the primary tumor/metastases with corresponding background SULpeak values. PET/CT images were found to be positive for malignancy at diagnostic CT/histological examination/clinical or radiological follow-up.

Statistical Analysis
Continuous variables were presented in terms of mean, median, standard deviation (SD), range, and interquartile range (IQR). [

Conclusions
In