Visual Abstract
Abstract
Fibroblast activation protein-α (FAP) is often highly expressed by sarcoma cells and by sarcoma-associated fibroblasts in the tumor microenvironment. This makes it a promising target for imaging and therapy. The level of FAP expression and the diagnostic value of 68Ga-FAP inhibitor (FAPI) PET for sarcoma subtypes are unknown. We assessed the diagnostic performance and accuracy of 68Ga-FAPI PET in various bone and soft-tissue sarcomas. Potential eligibility for FAP-targeted radiopharmaceutical therapy (FAP-RPT) was evaluated. Methods: This prospective observational trial enrolled 200 patients with bone and soft-tissue sarcoma who underwent 68Ga-FAPI PET/CT and 18F-FDG PET/CT (186/200, or 93%) for staging or restaging. The number of lesions detected and the uptake (SUVmax) of the primary tumor, lymph nodes, and visceral and bone metastases were analyzed. The Wilcoxon test was used for semiquantitative assessment. The association of 68Ga-FAPI uptake intensity, histopathologic grade, and FAP expression in sarcoma biopsy samples was analyzed using Spearman r correlation. The impact of 68Ga-FAPI PET on clinical management was investigated using questionnaires before and after PET/CT. Eligibility for FAP-RPT was defined by an SUVmax greater than 10 for all tumor regions. Results: 68Ga-FAPI uptake was heterogeneous among sarcoma subtypes. The 3 sarcoma entities with the highest uptake (mean SUVmax ± SD) were solitary fibrous tumor (24.7 ± 11.9), undifferentiated pleomorphic sarcoma (18.8 ± 13.1), and leiomyosarcoma (15.2 ± 10.2). Uptake of 68Ga-FAPI versus 18F-FDG was significantly higher in low-grade sarcomas (10.4 ± 8.5 vs. 7.0 ± 4.5, P = 0.01) and in potentially malignant intermediate or unpredictable sarcomas without a World Health Organization grade (not applicable [NA]; 22.3 ± 12.5 vs. 8.5 ± 10.0, P = 0.0004), including solitary fibrous tumor. The accuracy, as well as the detection rates, of 68Ga-FAPI was higher than that of 18F-FDG in low-grade sarcomas (accuracy, 92.2 vs. 80.0) and NA sarcomas (accuracy, 96.9 vs. 81.9). 68Ga-FAPI uptake and the histopathologic FAP expression score (n = 89) were moderately correlated (Spearman r = 0.43, P < 0.0002). Of 138 patients, 62 (45%) with metastatic sarcoma were eligible for FAP-RPT. Conclusion: In patients with low-grade and NA sarcomas, 68Ga-FAPI PET demonstrates uptake, detection rates, and accuracy superior to those of 18F-FDG PET. 68Ga-FAPI PET criteria identified eligibility for FAP-RPT in about half of sarcoma patients.
Sarcomas are rare and heterogeneous tumors that develop from the connective tissue of bone and soft tissue. There are more than 150 subtypes, including low-grade or intermediate or unpredictable tumors without a World Health Organization grade (not applicable [NA]). The outcome for patients with metastatic disease remains poor, with a median overall survival period of approximately 12–18 mo (1–3). Fibroblast activation protein-α (FAP) is a type II membrane glycoprotein belonging to the dipeptyl-peptidase family and is present in cancer-associated stromal fibroblasts (4,5). Cancer-associated stromal fibroblasts constitute an essential component of the tumor microenvironment (6–8). With the recent development of radiolabeled FAP inhibitors (FAPIs), these stromal markers have opened up opportunities for molecular imaging and FAP-targeted radiopharmaceutical therapy (FAP-RPT) (9). FAPI compounds have been used for the detection of malignant lesions with high stromal content on high-contrast PET/CT images. In recent years, numerous clinical studies have demonstrated high FAPI uptake in various solid tumors, including sarcomas (10–12). In addition, for several sarcoma subentities, such as myofibroblastic sarcoma, osteosarcoma, and undifferentiated pleomorphic sarcoma (UPS), histogenesis-specific FAP expression has been reported (13). In a previous subgroup analysis, our group proved the high intensity of intratumoral 68Ga-FAPI uptake in sarcoma patients (14). Furthermore, we demonstrated a higher detection rate and reproducibility, as well as a more advanced stage of disease, with 68Ga-FAPI PET than with 18F-FDG PET (14). Accurate staging is of great importance in planning appropriate therapy. In the advanced stage, FAP-RPT has demonstrated signs of efficacy (15–17) and is the subject of a prospective phase II safety and tolerability trial in patients with metastatic solid tumors (18). FAP-RPT has the potential to improve outcomes for many patients for whom approved therapeutic options are scarce or unfulfilling, including patients with advanced sarcomas. However, sarcoma is a basket term for a broad spectrum of distinct molecular subtypes that show heterogeneous uptake intensity, hence the importance of identifying subentities potentially more suitable for FAP-RPT. To address this issue, we assessed the diagnostic performance and accuracy of 68Ga-FAPI PET versus 18F-FDG PET in a large cohort of sarcoma patients. In addition, we investigated the association between 68Ga-FAPI PET uptake intensity and histopathologic expression of FAP and explored the eligibility of certain sarcoma subentities for FAP-RPT.
MATERIALS AND METHODS
Patient Population
The patient flowchart is illustrated in Figure 1. This is a subgroup analysis of an ongoing 68Ga-FAPI PET observational trial at University Hospital Essen (NCT04571086). Between October 2019 and 2022, 68Ga-FAPI PET was used for the staging or follow-up of sarcomas. In total, 200 bone sarcoma (BS) and soft-tissue sarcoma (STS) patients who underwent 68Ga-FAPI PET were included (31.8% of the cohort). Before enrollment, patients gave written informed consent to undergo 68Ga-FAPI PET for a clinical indication.
Enrollment flowchart. Q = questionnaire.
Image Acquisition and Evaluation
The synthesis and administration of 68Ga-FAPI-04 (n = 14) and 68Ga-FAPI-46 (n = 186) have been described previously (9,19). Patients did not require specific preparation before 68Ga-FAPI PET. Clinical PET/CT was performed craniocaudally on 200 patients: 3 (1.5%) with Biograph mMR, 6 (3%) with Biograph mCT, and 191 (95.5%) with Biograph mCT Vision (Siemens Healthineers). The mean activity ± SD injected intravenously was 120 ± 38.3 MBq for 68Ga-FAPI and 248.6 ± 89.2 MBq for 18F-FDG. The mean acquisition time after injection ± SD was 23.5 ± 19.0 min for 68Ga-FAPI PET and 69.5 ± 15.5 min for 18F-18F FDG PET. A diagnostic CT scan was obtained using a standard protocol (80–100 mA, 120 kV) before PET imaging (20). For each imaging modality, the number of lesions per region and per patient was recorded. Focal tracer uptake higher than the surrounding background and not associated with physiologic uptake was considered suggestive of malignancy. SUVmax was determined for lesions with the highest tracer uptake per region, using Syngo.via software (Siemens Healthineers). All devices had been cross-calibrated to European Association of Nuclear Medicine Research Ltd. accreditation standards. SUVmean was measured in 3 regions normalized according to tumor-to-background ratio (TBR): mediastinal blood pool (center of ascending aorta), liver (unaffected areas of the right lobe), and surrounding normal tissue, including bone or normal soft tissue. The images were read by 2 nuclear medicine physicians or radiologists during a joint reading session. Divergent findings were discussed and resolved by consensus between the readers.
Lesion Validation
Patients underwent histopathologic analysis of biopsy samples and surgical excision. Lesions that were histopathologically validated within 3 mo of a 68Ga-FAPI PET scan were included in the accuracy analysis. When histopathology was unavailable, validation was performed by correlative or follow-up imaging, that is, CT, MRI, or PET.
Immunohistochemistry
Biopsy and surgical specimens were stained with standard hematoxylin and eosin, as well as FAP immunohistochemistry, and evaluated as previously described (14,21). FAP expression is categorized semiquantitatively in the histologic section of the tumor as the percentage of FAP-positive cells. Semiquantitative analysis of FAP expression in stroma and tumor cells is assessed using the following scoring system: 0 is the absence or a low degree of FAP-positive cells (<1%), 1+ is FAP-positive in 1%–10% of cells, 2+ is FAP-positive in 11%–50% of cells, and 3+ is FAP-positive in more than 50% of cells. Pathologists were not informed of PET findings.
Management Questionnaires
To assess changes in planned treatment management after 68Ga-FAPI PET, referring physicians completed a questionnaire before PET, which was necessary to assess the patient’s existing treatment plan without the contribution of 68Ga-FAPI PET, and a second questionnaire after PET and after reviewing 68Ga-FAPI PET images, which was used to check for implemented change in management.
Statistical Analysis
Descriptive statistics and individual patient data are reported. For continuous data, the mean ± SD SUVmax and TBR were compared and tested for statistical differences using Wilcoxon and Mann–Whitney U tests. The sensitivity, specificity, and accuracy of 68Ga-FAPI PET on a per-region basis for the detection of tumor location, confirmed by histopathology or a composite reference standard, were calculated, along with the corresponding 2-sided 95% CIs. A difference of more than 10% was considered relevant. CIs were determined using the Wilson score method. The association of 68Ga-FAPI uptake intensity, grade, and histopathologic FAP expression was analyzed using Spearman r correlation. All statistical analyses were performed using SPSS software (version 20.0; SPSS Inc.) and GraphPad Prism (version 9.1.1; GraphPad Software).
RESULTS
Patient Characteristics
The clinical characteristics of the study population are summarized in Table 1 and Supplemental Table 1 (supplemental materials are available at http://jnm.snmjournals.org). Between October 2020 and 2022, 200 patients were included, 91 (45%) women and 109 (55%) men. Of the 200 patients, 65 (33%) had BS and 135 (67%) had STS; 141 (70%) cases were high grade, 32 (16%) cases were low grade, and 27 (14%) cases had no World Health Organization grade (NA). Patients underwent clinical 68Ga-FAPI PET imaging for either staging (49/200 [25%]) or follow-up (151/200 [75%]). Fourteen (7%) patients were imaged with 68Ga-FAPI-04, and 186 (93%) were imaged with 68Ga-FAPI-46. All patients imaged with 68Ga-FAPI-46 underwent 18F-FDG PET imaging within 4 wk. No PET-related adverse events were reported.
Patient Characteristics (n = 200)
FAP Expression in Sarcoma Subtypes
Tumor SUVmax and the tumor-to-liver ratio for 68Ga-FAPI versus 18F-FDG in different sarcoma subentities (n = 12) are summarized in Figure 2. We observed heterogeneous tumor uptake of 68Ga-FAPI in our cohort, ranging from an SUVmax of 3.1 in myxoid liposarcoma to an SUVmax of 47.1 in solitary fibrous tumor (SFT). In terms of mean SUVmax ± SD, the 3 sarcoma entities with the highest FAP expression were SFT (24.7 ± 11.9), UPS (18.8 ± 13.1), and leiomyosarcoma (15.2 ± 10.0). By descriptive comparison, the mean SUVmax was higher for 68Ga-FAPI than for 18F-FDG in most sarcoma subentities, with the exception of synovial sarcoma, spindle cell sarcoma, and other BS. According to the Wilcoxon test, SUVmax and the tumor-to-liver ratio of 68Ga-FAPI PET were significantly higher than those of 18F-FDG PET for SFT (mean SUVmax ± SD, 24.7 ± 11.9 vs. 6.8 ± 8.7, P = 0.0005; mean tumor-to-liver ratio ± SD, 22.0 ± 11.9 vs. 4.1 ± 8.9, P = 0.0005) and myxoid liposarcoma (mean SUVmax ± SD, 5.6 ± 2.2 vs. 3.5 ± 2.1, P = 0.03; mean tumor-to-liver ratio ± SD, 1.7 ± 1.9 vs. 0.8 ± 1.7, P = 0.04). Additional information on SUVs and TBR is shown in Supplemental Figures 1 and 2 and in Supplemental Table 2.
Comparison of 68Ga-FAPI and 18F-FDG PET SUVmax (A) and tumor-to-liver ratio (B) for sarcoma subentities. Individual data and mean (bars) are shown. Horizontal dotted lines in A indicate patients with SUVmax greater than 10 and 20. Black dot = 68Ga-FAPI; white dot = 18F-FDG.
Based on our previous results (15), intense FAP expression, defined by an SUVmax of at least 10 in each tumor region, was deemed sufficient for FAP-RPT, as shown in Figure 2A. These PET criteria were met in 62 of 138 (45%) patients with metastatic disease: 16 of 20 with SFT, 3 of 9 with UPS, 7 of 11 with leiomyosarcoma, 5 of 10 with osteosarcoma, 3 of 8 with undifferentiated liposarcoma, 13 of 27 with other STS, 4 of 8 with spindle cell sarcoma, 5 of 13 with chondrosarcoma, 3 of 11 with other BS, 2 of 8 with Ewing sarcoma, and 1 of 4 with synovial sarcoma. FAP expression was highly intense (SUVmax of 20 or higher in all regions, as shown in Fig. 2A) in 25 of 138 (18%) patients: 10 of 20 with SFT, 3 of 9 with UPS, 1 of 11 with leiomyosarcoma, 1 of 10 with osteosarcoma, 1 of 8 with undifferentiated liposarcoma, 5 of 27 with other STS, 2 of 8 with spindle cell sarcoma, 1 of 13 with chondrosarcoma, and 1 of 11 with other BS. A complete list of subentities included in the other BS and other STS groups is given in Supplemental Table 3.
68Ga-FAPI versus 18F-FDG uptake was assessed separately for high-grade, NA, and low-grade sarcomas (Fig. 3). Uptake of 68Ga-FAPI versus 18F-FDG was significantly higher in low-grade sarcomas (10.36 ± 8.5 vs. 7.0 ± 4.5, P = 0.01) and NA sarcomas (22.3 ± 12.5 vs. 8.5 ± 10, P = 0.0004), particularly SFT. An example patient is shown in Figure 4.
Comparison of 68Ga-FAPI and 18F-FDG PET SUVmax (A) and tumor-to-liver ratio (B) separated into high-grade, NA, and low-grade groups. Individual data and mean (bars) are shown. Black dot = 68Ga-FAPI; white dot = 18F-FDG.
62-y-old patient with metastatic SFT. Higher 68Ga-FAPI uptake (A and B) than 18F-FDG uptake (C and D) is shown in images of primary tumor in right pelvis (SUVmax of 25.1 in B vs. 9.0 in C) and multiple pelvic bone metastases (right sacrum, SUVmax of 23.0 in B vs. 2.2 in C). Shown are maximum-intensity projection PET images (A and D), axial PET images (B and C, top), and axial CT images (B and C, bottom).
Detection Efficacy
Detection efficiency is given in Table 2 for primary tumors, lymph nodes, and distant metastases (lung, muscle, viscera [organ], liver, and bone). The detection efficacy of 68Ga-FAPI PET was superior to that of 18F-FDG PET for distant metastases in NA (100% vs. 67%) and low-grade (95% vs. 81%) sarcomas.
Detection Efficacy on Per-Region Basis in High-Grade (n = 141), NA (n = 27), and Low-Grade (n = 32) Groups
Overall, 68Ga-FAPI PET versus 18F-FDG PET detected 1,181 (95%) versus 1,023 (85%) lesions. 68Ga-FAPI PET outperformed 18F-FDG PET in detecting primary tumors (144 [100%] vs. 124 [86%]) and distant metastases (945 [97%] vs. 797 [83%]).
Accuracy
The accuracy of per-region analysis is summarized in Table 3. In total, 142 lesions were histologically validated (110 [77%] primary tumors, 7 [5%] lymph nodes, 22 [15%] visceral metastases, and 3 [2%] bone metastases). In addition, 1,056 lesions were validated by correlative or follow-up imaging (34 [3%] primary tumors, 97 [9%] lymph nodes, 659 [63%] visceral metastases, and 266 [25%] bone metastases). In patients with high-grade sarcomas, sensitivity (96% vs. 94%), specificity (86% vs. 68%), and accuracy (95% vs. 92%) were higher for 68Ga-FAPI than for 18F-FDG. The same was true for patients with NA sarcomas (sensitivity, 96% vs. 83%; specificity, 80% vs. 67%; and accuracy, 95% vs. 82%) and patients with low-grade sarcomas (sensitivity, 93% vs. 85%; specificity, 89% vs. 44%; and accuracy, 92% vs. 80%). Relevant improvement, defined as a difference of 10% or more, was observed with 68Ga-FAPI PET in the specificity of detection of high-grade sarcomas and for all 3 accuracy measures for NA and low-grade sarcomas.
Accuracy of Detection of Sarcoma on Per-Region Basis in High-Grade (n = 141), NA (n = 27), and Low-Grade (n = 32) Groups
Change in Therapeutic Management
Changes in therapeutic management are presented in Supplemental Table 4. For 168 of 200 (84%) patients, questionnaires completed and returned before and after imaging were available. The management implemented was assessed by reviewing the clinical files. Therapeutic changes based on 68Ga-FAPI PET results were documented in 33 of 168 (20%) patients: 20 (61%) patients changed from active surveillance to chemotherapy, 6 (18%) patients changed from isolated limb perfusion to surgery, 3 (9%) patients changed from a biopsy to surgery, 1 (3%) patient changed from a biopsy to chemotherapy, 1 (3%) patient changed from surgery to chemotherapy, 1 (3%) patient underwent resection plan adjustment, and 1 (3%) patient changed from therapy to active surveillance. Moreover, of the 62 patients with metastatic disease and an SUVmax greater than 10, 17 (27%) patients were deemed eligible and underwent at least 1 cycle of FAP-RPT. A patient flowchart is presented in Supplement Figure 3.
PET Versus Immunohistochemistry Target Expression
The association between 68Ga-FAPI PET uptake intensity and FAP immunohistochemistry score is shown in Figure 5 and Supplemental Table 5. Of 89 samples, 30 (34%) samples demonstrated no FAP expression on immunohistochemistry (score 0), and 59 samples had scores 1–3. A moderate positive correlation (Spearman r = 0.43, P = 0.0002) was found between SUVmax and histopathologic FAP expression. Higher uptake values (mean SUVmax ± SD) were observed on lesions with FAP score 3 (22.7 ± 14.2) than on those with FAP score 0 (11.4 ± 7.0).
Association between 68Ga-FAPI PET uptake intensity (SUVmax) and FAP immunohistochemistry score (n = 89). Individual data and mean (bars) are shown. Immunohistochemistry scoring: 0 = no expression (<1%), 1 = 1%–10%, 2 = 11%–49%, + = ≥50% FAP-positive cells. Comparison of SUVmax with established immunohistochemistry scoring system showed moderate linear relationship (Spearman r = 0.43, P = 0.0002).
DISCUSSION
In recent years, FAP has been identified as a promising theranostic target for various cancers, including sarcomas (14,15,22,23). We analyzed 68Ga-FAPI PET images of 200 patients with 13 subentities of sarcoma. Our study revealed the heterogeneous tumor uptake intensity of FAP, with a mean SUVmax ± SD ranging from 5.6 ± 2.2 in myxoid liposarcoma to 24.7 ± 11.9 in SFT. In addition, we report that diagnostic performance of 68Ga-FAPI PET is superior to that of 18F-FDG PET in patients with low-grade and NA sarcomas.
Numerous previous studies have demonstrated the usefulness of 18F-FDG PET imaging for high-grade sarcomas (24–26). However, sarcomas are highly heterogeneous in terms of aggressiveness and tumor origin. Consequently, imaging these tumors with 18F-FDG PET, as currently indicated for follow-up (4,27), is often challenging and does not appear to be a viable universal imaging method. In an analysis of 21 tumor entities, Hirmas et al. (28) reported that 68Ga-FAPI versus 18F-FDG had higher absolute uptake and TBR, as well as better tumor detection, in sarcomas and pancreatic cancers. Concordant with this observation, we demonstrated that the mean absolute uptake and TBR of 68Ga-FAPI were higher than those of 18F-FDG in all sarcoma subentities except synovial sarcoma, spindle cell sarcoma, and other BS. In a recent prospective study of 45 STS patients, low-grade STS had significantly higher FAP uptake, whereas high-grade STS had significantly higher 18F-FDG uptake (29). We also found significantly higher 68Ga-FAPI versus 18F-FDG uptake in low-grade and NA sarcomas. Here, SFT demonstrated high FAP expression, almost twice the average level for all sarcomas. In addition, higher tumor uptake of 68Ga-FAPI translated into a higher per-region detection rate and higher accuracy of 68Ga-FAPI than of 18F-FDG in NA and low-grade sarcomas. 68Ga-FAPI PET led to a change in therapeutic management in around 20% of patients. In around a third of these patients, 68Ga-FAPI PET led from active surveillance to systemic treatment. A small subgroup switched from locoregional to systemic therapy, and a single patient switched from systemic therapy to active surveillance. Most of our cohort were patients with advanced metastatic disease who had already undergone extensive imaging, so 68Ga-FAPI PET only moderately affected clinical decision-making. Nevertheless, we believe that the impact on clinical management will increase if 68Ga-FAPI-PET is performed at earlier stages of the disease. The better tumor detection and specificity of FAPI versus current imaging standards, especially for NA and low-grade sarcomas, could be pivotal to implement staging (i.e., M0 vs. M1) and hence affect therapy planning adjustment (i.e., curative vs. palliative). Moreover, it could implement the assessment of disease extent before local therapies (i.e., target tumor volume before external beam radiotherapy).
Immunohistochemistry analysis was performed on 89 patients. A high level of FAP expression in tumor stroma has been reported previously (6,10,30,31). In our study, immunohistochemistry confirmed the presence of the FAP target in tumor lesions and showed a moderate positive correlation, with a higher FAP score associated with higher 68Ga-FAPI PET uptake.
Because of their origin in soft tissue, most sarcomas intrinsically express FAP on the surface of tumor cells and surrounding fibroblasts (8,13,32), which may make this tumor entity particularly suitable for 68Ga-FAPI PET and FAP-RPT (14,28,33,34). Metastatic sarcoma has a poor prognosis, with an overall 5-y survival rate of 15% (35). Treatment options for this metastatic disease are scarce and unfulfilling. FAP-positive cells play a vital role in remodeling the tumor microenvironment. Therefore, FAP is increasingly considered a potential pantumoral target for the design of tumor-targeting drugs, which explains why several in vitro and vivo studies are ongoing.
The development of immunomodulatory therapies based on oncolytic viruses is playing an increasingly important role in the treatment of solid tumors, involving both direct cell lysis and immunogenic cell death. In this context, oncolytic viruses armed with an FAP-targeting bispecific T-cell engager have been designed to target infiltrating lymphocytes toward cancer-associated stromal fibroblasts, thereby enhancing viral propagation and T-cell–mediated cytotoxicity against tumor stroma to improve therapeutic activity (36). FAP-targeting bispecific T-cell engager activators, which costimulate T cells and improve tumor cell destruction in FAP-expressing tumors, are the subject of several phase I studies in patients with advanced solid tumors, with preliminary results demonstrating tolerability and safety (37,38), as well as signs of response (39).
Moreover, when conjugated with doxorubicin, FAP has been used to generate chemotherapeutic prodrugs, activated only in the tumor microenvironment, to selectively release anticancer agents and improve the targeting effect of these cytotoxic agents, thus reducing their systemic side effects (40). FAP represents a promising target for other potential treatments, such as immunotherapy (41,42); FAP-targeted chimeric antigen receptor–T-cell therapy, which is being investigated in 2 phase I clinical trials in patients with malignant pleural mesothelioma (43); nectin-4–positive advanced solid malignancies (44); and RPT.
RPT is capable of delivering radiation to FAP- and stroma-rich tumor lesions while limiting damage to surrounding tissue. This new therapeutic approach has been widely applied to metastatic neuroendocrine tumors and prostate cancers, improving quality of life and overall survival (45,46). Several FAP ligands are being investigated in preclinical and clinical settings as theranostic agents. In a head-to-head comparison, 177Lu-labeled FAP ligands were evaluated in vitro in cell lines with low and high human FAP expression and in mice bearing low and high FAP-expressing models. The 177Lu-FAPI-46 dimer presented higher uptake and longer tumor retention than those of the monomer, whereas the tumor–to–critical organ values were in favor of cyclic peptide FAP-2286 (47). In a first-in-human dosimetry study, 177Lu-FAP-2286 showed longer tumor retention than a small FAPI tracer, such as FAPI-02/04, and the doses absorbed by the whole body, bone marrow, and kidney were comparable to those of other radiopharmaceuticals previously reported to be effective, namely, 177Lu-DOTATATE and 177Lu-PSMA-617 (48). The results of the studies available so far ultimately indicate that dimerization of FAPI small molecules and the cyclic peptide are 2 promising strategies for enhancing the tumor radiation dose.
Various radionuclides are taken into consideration for labeling. If on one side, the β-particle energy of 90Y is higher than that of 177Lu, then on the other side, the longer range of 90Y-β could increase the risk of bone marrow and renal toxicity. Because of the high and precise energy delivery to the tumor per unit of radioactivity, α-emitters, such as 225Ac, could also be potential candidates, as reported in a proof-of-concept study (49). FAP-RPT with 90Y-FAPI and 177Lu-FAPI has been documented in several case reports and case series for the treatment of various tumor entities (17,50–52). Our group has previously reported favorable safety and evidence of the efficacy of FAP-RPT in a mixed cohort of patients mainly with metastatic sarcomas (15). Furthermore, FAP-RPT is undergoing a prospective phase II safety and tolerability trial in patients with advanced solid tumors (18), with preliminary results showing no significant toxicity and some signs of early efficacy (53). In accordance with therapeutic criteria (45,54), intense FAP expression, defined by an SUVmax of at least 10 for all tumor lesions, indicated eligibility for FAP-RPT. Based on these criteria, more than half of our patients could be eligible for FAP-RPT. Several subentities of sarcoma, including SFT, UPS, and leiomyosarcoma, demonstrated 68Ga-FAPI uptake that ranged up to highly intense (SUVmax > 20), indicating favorable target expression for FAP-RPT. Because of the heterogeneous expression of the target, 68Ga-FAPI PET could become a tool for determining eligibility for FAP-RPT and identifying subentities of sarcoma likely to benefit from this therapeutic approach.
Several limitations were identified. We found a moderate correlation between 68Ga-FAPI uptake by PET and target expression by immunohistochemistry. Thus, SUVmax may not be representative of the entire tumor lesion, which may underestimate the intralesional heterogeneity of FAP expression. Moreover, some patients did not undergo 18F-FDG PET, potentially leading to a selection bias.
In this analysis, we focus on the diagnostic accuracy of 68Ga-FAPI PET. This study does not include mandatory follow-up. The absence of follow-up data may have led to bias. FAP-RPT eligibility was in line with previously published criteria (15). However, these criteria have not yet been validated on the basis of oncologic outcomes.
CONCLUSION
68Ga-FAPI PET demonstrates tumor uptake, detection rate, and accuracy superior to that of 18F-FDG PET in patients with low-grade and NA sarcomas. Tumor uptake for 68Ga-FAPI PET correlated moderately with FAP expression for immunohistochemistry. 68Ga-FAPI PET criteria identified eligibility for FAP-RPT in about half of sarcoma patients, especially those with SFT, UPS, and leiomyosarcoma.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Does the diagnostic performance of 68Ga-FAPI PET in BS and STS vary according to grade of disease and subentities, and if so, which subentities are more likely to be good candidates for FAP-RPT?
PERTINENT FINDINGS: We observed diagnostic performance and accuracy of 68Ga-FAPI superior to that of 18F-FDG in intermediate and low-grade sarcomas. The subentities that consistently show intense FAPI uptake (SUVmax > 20), namely, SFT, UPS, and leiomyosarcomas, are more likely to benefit from this therapeutic approach.
IMPLICATIONS FOR PATIENT CARE: 68Ga-FAPI PET is a diagnostic tool for low-grade and NA sarcomas and allows the determination of eligibility for FAP-RPT.
Footnotes
↵* Contributed equally to this work.
Published online May 9, 2024.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication December 15, 2023.
- Revision received March 15, 2024.
