Visual Abstract
Abstract
To assess the diagnostic accuracy of 68Ga-labeled fibroblast activation protein inhibitor (FAPI) and 18F-labeled FDG PET for the detection of various tumors, we performed a head-to-head comparison of both imaging modalities across a range of tumor entities as part of our ongoing 68Ga-FAPI PET observational trial. Methods: The study included 115 patients with 8 tumor entities who received imaging with 68Ga-FAPI for tumor staging or restaging between October 2018 and March 2022. Of those, 103 patients received concomitant imaging with 68Ga-FAPI and 18F-FDG PET and had adequate lesion validation for accuracy analysis. Each scan was evaluated for the detection of primary tumor, lymph nodes, and visceral and bone metastases. True or false positivity and negativity to detected lesions was assigned on the basis of histopathology from biopsies or surgical excision, as well as imaging validation. Results: 68Ga-FAPI PET revealed higher accuracy than 18F-FDG PET in the detection of colorectal cancer (n = 14; per-patient, 85.7% vs. 78.6%; per-region, 95.6% vs. 91.1%) and prostate cancer (n = 22; per-patient, 100% vs. 90.9%; per-region, 96.4% vs. 92.7%). 68Ga-FAPI PET and 18F-FDG PET had comparable per-patient accuracy in detecting breast cancer (n = 16, 100% for both) and head and neck cancers (n = 10, 90% for both modalities). 68Ga-FAPI PET had lower per-patient accuracy than 18F-FDG PET in cancers of the bladder (n = 12, 75% vs. 100%) and kidney (n = 10, 80% vs. 90%), as well as lymphoma (n = 9, 88.9% vs. 100%) and myeloma (n = 10, 80% vs. 90%). Conclusion: 68Ga-FAPI PET demonstrated higher diagnostic accuracy than 18F-FDG PET in the diagnosis of colorectal cancer and prostate cancer, as well as comparable diagnostic performance for cancers of the breast and head and neck. Accuracy and impact on management will be further assessed in an ongoing prospective interventional trial (NCT05160051).
Imaging is fundamental in the treatment of malignancies, with varying detection rates depending on the tumor entity and diagnostic modality. PET of cancer cells using 18F-FDG PET acquires additional molecular information useful for the detection of disease recurrence and metastases, response assessment, disease management, and prognostication (1–6). However, drawbacks of 18F-FDG include false-positive findings due to physiologic uptake or inflammatory responses, as well as false-negative findings due to elevated serum blood glucose levels. As such, targeting of cancer cells using alternative radioisotopes has been an area of growing interest.
Cancer-associated fibroblasts, a constituent of the tumor microenvironment, are involved in tumor growth, migration, and progression (7). Fibroblast activation protein (FAP) α is expressed by cancer-associated fibroblasts, a marker associated with protumorigenic functions (8–12) and, therefore, a suitable target for diagnostic and therapeutic purposes. Multiple preclinical and clinical studies have shown the promise of FAP-directed therapies, including radiolabeled FAP inhibitors (FAPIs), which exhibit favorable properties in cancer diagnosis and therapy (13–18). These properties include, but are not limited to, fast imaging times, high contrast in tumor lesions, and no dietary requirements with regard to imaging, as well as acceptable side effects and long tumor retention times with regard to therapy.
Because of the favorable characteristics of this imaging modality, patients were referred for clinical 68Ga-FAPI PET staging, both at initial diagnosis and for reevaluation, and were offered subsequent enrollment in our prospective observational 68Ga-FAPI registry.
In this report, we assess the diagnostic accuracy of 68Ga-FAPI compared with 18F-FDG PET separately for various tumor entities by analyzing sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy on per-patient and per-region bases.
MATERIALS AND METHODS
Study Design and Participants
Until March 2022, adult patients who underwent clinical 68Ga-FAPI PET were offered the possibility to consent to a prospective observational trial for correlation and clinical follow-up of PET findings (NCT04571086). Patients signed a written informed consent form, and evaluation of data was approved by the ethics committee of the University Duisburg–Essen (20-9485-BO and 19-8991-BO). We previously reported data on 68Ga-FAPI PET uptake and accuracy in sarcoma (n = 47 (19)), as well as 68Ga-FAPI PET uptake in mixed cohorts (n = 69 (20), n = 91 (21), and n = 324 (22)). Patients with sarcoma, pancreatic cancer, and pleural mesothelioma have been excluded from this analysis since the results have already been or will be published separately. Moreover, solid tumor entities with fewer than 10 patients per entity for 68Ga-FAPI PET accuracy assessment were excluded from this analysis.
Details of data collection (23,24), imaging and administration of radioligands (20,25,26), imaging analysis, lesion validation, follow-up (27), and statistical analysis are provided in the supplemental materials (available at http://jnm.snmjournals.org).
RESULTS
Patient Characteristics
We identified 133 patients, of whom 115 with adequate lesion validation were included in this analysis. In total, 8 tumor entities and 313 regions were analyzed; patient characteristics (n = 115) are outlined in Table 1. The median age was 63 y (interquartile range, 17 y). The most common tumor entities were prostate cancer (22/115, 19%), head and neck cancers (18/115, 16%), breast cancer (16/115, 14%), colorectal cancer (15/115, 13%), and bladder cancer (12/115, 10%). Most patients (81/115, 70%) underwent 68Ga-FAPI PET imaging for restaging purposes. A total of 103 (90%) patients underwent concomitant imaging via 68Ga-FAPI and 18F-FDG PET and had adequate lesion validation for the accuracy analysis, and this set of patients was included in the composite analysis.
Patient Characteristics (n = 115)
Composite Analysis: Higher Diagnostic Accuracy with 68Ga-FAPI PET Than with 18F-FDG PET
68Ga-FAPI PET showed higher diagnostic accuracy than 18F-FDG PET in the diagnosis of colorectal cancer and prostate cancer as listed in Table 2.
Comparison of Diagnostic Efficacy Between 68Ga-FAPI and 18F-FDG PET (Per-Patient and Per-Region Analysis) for Tumors in Which 68Ga-FAPI Outperformed 18F-FDG PET
At a per-patient level in colorectal cancer, 68Ga-FAPI PET was superior to 18F-FDG PET in accuracy (85.7% vs. 78.6%), sensitivity (90.9% vs. 81.8%), and NPV (66.7% vs. 50%). At a per-region level, 68Ga-FAPI PET was superior to 18F-FDG PET in accuracy (95.6% vs. 91.1%), sensitivity (94.1% vs. 88.2%), and PPV (94.1% vs. 88.2%).
Furthermore, at a per-patient level in prostate cancer, 68Ga-FAPI PET was superior to 18F-FDG PET in accuracy (100% vs. 90.9%) and sensitivity (100% vs. 90.9%). At a per-region level, 68Ga-FAPI PET was superior to 18F-FDG PET in sensitivity (94.3% vs. 88.6%) and NPV (90.9% vs. 83.3%).
Composite Analysis: Comparable Diagnostic Accuracy Between 68Ga-FAPI PET and 18F-FDG PET
68Ga-FAPI PET was comparable to 18F-FDG PET in the diagnosis of breast cancer and head and neck cancers as listed in Table 3.
Comparison of Diagnostic Efficacy Between 68Ga-FAPI and 18F-FDG PET (Per-Patient and Per-Region Analysis) for Tumors in Which 68Ga-FAPI Was Comparable to 18F-FDG PET
At a per-patient level in breast cancer, 68Ga-FAPI PET and 18F-FDG PET showed equal accuracy, sensitivity, specificity, PPV, and NPV (all 100%). At a per-region level, 68Ga-FAPI PET showed accuracy (97.9% vs. 100%) and sensitivity (96.6% vs. 100%) comparable to those of 18F-FDG PET but lower NPV (94.7% vs. 100%).
At a per-patient level in head and neck cancers, 68Ga-FAPI PET and 18F-FDG PET showed equal accuracy (90%), sensitivity (100%), and PPV (90%). At a per-region level, 68Ga-FAPI PET showed accuracy (90.3% vs. 93.6%) and specificity (86.7% for both) comparable to those of 18F-FDG PET but lower sensitivity (93.8% vs. 100%) and NPV (92.9% vs. 100%).
Composite Analysis: Lower Diagnostic Accuracy with 68Ga-FAPI PET Than with 18F-FDG PET
68Ga-FAPI PET showed lower accuracy than 18F-FDG PET in the diagnosis of bladder and kidney cancers, lymphoma, and myeloma as shown in Table 4.
Comparison of Diagnostic Efficacy Between 68Ga-FAPI and 18F-FDG PET (Per-Patient and Per-Region Analysis) for Tumors in Which 68Ga-FAPI Underperformed in Comparison to 18F-FDG PET
At a per-patient level in bladder cancer, 68Ga-FAPI PET showed lower accuracy (75% vs. 100%), sensitivity (72.7% vs. 100%), and NPV (25% vs. 100%) than 18F-FDG PET. At a per-region level, 68Ga-FAPI PET showed lower accuracy (89.2% vs. 94.4%), sensitivity (78.6% vs. 92.3%), and NPV (88% vs. 95.7%) than 18F-FDG PET.
At a per-patient level in kidney cancer, 68Ga-FAPI PET showed sensitivity comparable to that of 18F-FDG PET (87.5% for both) but lower accuracy (80% vs. 90%), specificity (50% vs. 100%), and PPV (87.5% vs. 100%). At a per-region level, 68Ga-FAPI PET showed accuracy (90.3% vs. 93.6%), sensitivity (92.9% for both), and NPV (93.8% vs. 94.1%) comparable to those of 68Ga-FAPI PET but lower specificity (88.2% vs. 94.1%) and PPV (86.7% vs. 92.9%).
At a per-patient level in lymphoma, 68Ga-FAPI PET showed lower accuracy (88.9% vs. 100%), sensitivity (87.5% vs. 100%), and NPV (50% vs. 100%) than 18F-FDG PET. At a per-region level, 68Ga-FAPI PET showed lower accuracy (90% vs. 96.7%), sensitivity (78.6% vs. 100%), and NPV (84.2% vs. 100%) than 18F-FDG PET.
Finally, for myeloma at per-patient and per-region levels, accuracy (80% vs. 90%) and sensitivity (75% vs. 87.5%) were lower with 68Ga-FAPI PET than with 18F-FDG PET.
Histopathology-Only Analysis
In a subgroup of 45 patients and 5 tumor entities, accuracy was assessed by histopathology validation only (Supplemental Table 1). In line with the findings of the composite analysis, 68Ga-FAPI PET demonstrated higher accuracy than 18F-FDG PET for prostate cancer, comparable accuracy for breast cancer and colorectal cancer, and lower accuracy for bladder and kidney cancers.
DISCUSSION
Here, we compare the diagnostic accuracy of 68Ga-FAPI and 18F-FDG PET for various tumors. Tumor validation by a composite reference standard revealed that the diagnostic accuracy of 68Ga-FAPI PET was higher than that of 18F-FDG PET in colorectal cancer and prostate cancer, comparable in breast cancer and head and neck cancer, and lower in bladder and kidney cancers, lymphoma, and myeloma. Histopathology-only analysis revealed that the diagnostic accuracy of 68Ga-FAPI PET was higher than that of 18F-FDG PET in prostate cancer, comparable in breast and colorectal cancers, and lower in bladder and kidney cancers.
For cancers of the abdomen and pelvis, 68Ga-FAPI uptake was low in normal parenchyma, such as bowel (SUVmax range, 0.08–3.56), liver (SUVmax range, 0.47–2.91), and spleen (SUVmax range, 0.64–2.81) (15,28,29). This improves tumor delineation, with absolute and tumor-to-liver uptakes being higher on 68Ga-FAPI PET than on 18F-FDG PET, which may lead to superior diagnostic accuracy (22). This is particularly relevant in abdominal surgery, for example, after which patients are required to take nothing by mouth until bowel recovery. Also, the prevalence of coexisting diabetes (≤15.5% in patients with colon cancer, for instance (30)) poses limitations for molecular imaging with 18F-FDG PET. 68Ga-FAPI PET in such a context has protocol advantages, given that no diet or fasting is required in preparation for imaging, and image acquisition can take place a few minutes after tracer application. 68Ga-FAPI PET, therefore, has the potential to replace 18F-FDG for abdominal staging.
Our findings are corroborated by other studies that have also shown 68Ga-FAPI PET to have diagnostic accuracy superior to that of 18F-FDG PET in breast cancer (31–33) and head and neck cancers (34–36). Moreover, reports have shown that 68Ga-FAPI PET can detect PSMA-negative prostate cancer lesions (37–39), which can aid in the diagnostic process, with potential therapeutic implications.
With regard to lymphoma and myeloma, several studies have shown that 68Ga-FAPI PET is inferior to (or at best, not superior to) 18F-FDG PET (40–43). For example, in comparison to colorectal cancer, lymphoma lesions show lower uptake with 68Ga-FAPI than with 18F-FDG PET (22,41,44), higher background uptake, and, thus, lower tumor-to-background values (e.g., median SUVmax of 7.4 vs. 22.5 and median liver tumor-to-background ratio of 6.4 vs. 10.5 for 68Ga-FAPI vs. 18F-FDG PET, respectively (22)). Taking this a step further, using systematic lesion validation and follow-up, our study revealed 68Ga-FAPI to be less accurate than 18F-FDG PET in lymphoma and myeloma.
An ongoing prospective clinical trial at our department (NCT05160051) is exploring the diagnostic accuracy of 68Ga-FAPI-46 PET and its effect on patient management and interreader reproducibility for different tumor entities. An interim analysis of findings has shown that 68Ga-FAPI PET is associated with a lower rate of false-positive findings, especially in lymph node assessments (44).
With high tumor and low organ uptakes (22), as well as diagnostic accuracy across various tumor entities, 68Ga-FAPI PET has a role as a gatekeeper for FAP-directed radioligand therapy. Feasibility of FAP radioligand therapy has been reported for 90Y- and 153Sm-labeled compounds in breast (13) and ovarian (45) cancer, as well as sarcomas and pancreatic cancers (17,46). 177Lu-labeled compounds have also been used in multiple advanced and refractory tumors, including thyroid cancer (16,47–49). In patients with intense FAP expression on 68Ga-FAPI PET, 90Y-FAPI-46 radioligand therapy led to disease control in about one third of patients with initially progressive sarcomas, pancreatic cancer, and other cancers (50), and the novel dimeric 177Lu-labeled FAPI radioligand (177Lu-DOTAGA.(SA.FAPi)2) led to disease control in almost half the patients with radioiodine-refractory differentiated thyroid cancer who had progressed on tyrosine kinase inhibitors (49). FAPI imaging therefore has the potential to enhance drug development with targeted clinical applications.
One notable example of a FAP-targeting drug that has shown clinical promise is talabostat, which has demonstrated tumor control in 21% of patients with colorectal cancer (51). Moreover, targeting FAP with chimeric antigen receptor T cells has shown promise in preclinical studies and case reports (52,53), and there is potential for combination with cancer vaccines or immune checkpoint inhibitors (such as PD-1 inhibitors), which would lead to further blockade of immunosuppressive factors (52). Another promising approach is using cancer vaccines that successfully target FAP, particularly the genome of stromal fibroblasts (54). As such, future drug development and its potential clinical applications may be enhanced through 68Ga-FAPI imaging, which aids in selecting patients whose tumors exhibit high 68Ga-FAPI uptake and who would potentially benefit from FAP-directed therapy. This theranostic approach also has the potential to improve clinical trial design.
Our study is limited by its retrospective design and the small number of patients included per tumor entity. Histopathology was not available for all patients, as tissue sampling is not routinely performed, and biopsy of metastatic lesions may be difficult because they may be small or remote. Thus, most lesion follow-up was based on correlative or follow-up imaging with known intrinsic limitations. Despite these limitations, the study provided valuable systematic information on the diagnostic efficacy of 68Ga-FAPI PET from an ongoing registry study evaluating 68Ga-FAPI and 18F-FDG PET, using a composite reference standard with adequate follow-up time (≤∼6 mo) and across a wide range of tumor entities, thereby adding to the growing pool of theranostic data.
CONCLUSION
When compared with 18F-FDG PET, 68Ga-FAPI PET demonstrated higher accuracy in the diagnosis of colorectal cancer and prostate cancer, as well as comparable diagnostic performance for cancers of the breast and head and neck. 68Ga-FAPI has the potential for improved staging or theranostic screening, particularly for these tumor entities.
DISCLOSURE
Rainer Hamacher is supported by the Clinician Scientist Program of the University Medicine Essen Clinician Scientist Academy (UMEA) sponsored by the faculty of medicine and Deutsche Forschungsgemeinschaft (DFG) and has received travel grants from Lilly, Novartis, and Pharma Mar, as well as fees from Lilly and Pharma Mar. Lukas Kessler is a consultant for AAA and BTG and received fees from Sanofi. Kim Pabst is a consultant for Novartis, has received a Junior Clinician Scientist Stipend from the University Medicine Essen Clinician Scientist Academy (UMEA) sponsored by the Faculty of Medicine at the University of Duisburg–Essen and the Deutsche Forschungsgemeinschaft (DFG), and has received research funding from Bayer outside the submitted work. Stefan Kasper reports personal fees and grants from AstraZeneca, Merck Serone, Merck Sharpe & Dohme, Amgen, Bristol Myers Squibb, Roche, Lilly, Servier, Incyte, and SanofiAventis outside the submitted work. Claudia Kesch has received consultant fees from Apogepha, research funding from AAA/Novartis and Curie Therapeutics, and compensation for travel from Janssen R&D, Amgen, and Bayer. Bastian von Tresckow is an advisor or consultant for Allogene, BMS/Celgene, Cerus, Incyte, IQVIA, Gilead Kite, Lilly, Miltenyi, Novartis, Noscendo, Pentixapharm, Roche, Amgen, Pfizer, Takeda, Merck Sharp & Dohme, and Gilead Kite; has received honoraria from AstraZeneca, BMS, Incyte, Lilly, Novartis, Roche Pharma AG, Takeda, and Merck Sharp & Dohme; reports research funding from Novartis (to the institution), Merck Sharp & Dohme (to the institution), and Takeda (to the institution); and reports travel support from AbbVie, AstraZeneca, Gilead Kite, Lilly, Merck Sharp & Dohme, Pierre Fabre, Roche, Takeda, and Novartis. Hubertus Hautzel reports research funding and travel support from PARI GmbH outside the submitted work. Martin Glas reports honoraria from Roche, Novartis, UCB, Abbvie, Daiichi Sankyo, Novocure, Bayer, Janssen-Cilag, Medac, Merck, and Kyowa Kirin; travel support from Novocure and Medac; and a research grant from Novocure. Ken Herrmann reports personal fees from Bayer, SIRTEX, Adacap, Curium, Endocyte, IPSEN, Siemens Healthineers, GE Healthcare, Amgen, Novartis, ymabs, Aktis, Oncology, and Pharma15, as well as personal and other fees from SOFIE Biosciences, nonfinancial support from ABX, and grants and personal fees from BTG, all of which are outside the submitted work. Boris Hadaschik declares grants to the institution from Novartis, BMS, and the German Research Foundation; consulting fees from ABX, Amgen, AstraZeneca, Bayer, BMS, Janssen, Lightpoint Medical, Novartis, and Pfizer; payment for lectures from Janssen and Monrol; support for travel or meeting attendance from AstraZeneca, Bayer, and Janssen; and participation on data safety monitoring boards for Janssen, all outside the submitted work. Philipp Harter reports honoraria from Amgen, AstraZeneca, GSK, Roche, Sotio, Stryker, Zai Lab, MSD, Clovis, Eisai, Mersana, and Exscientia. He is on the advisory board for Astra Zeneca, Roche, GSK, Clovis, Immunogen, MSD, Miltenyi, Novartis, and Eisai. He has received research funding (to the institution) from AstraZeneca, Roche, GSK, Genmab, DFG, European Union, DKH, Immunogen, Seagen, Clovis, and Novartis. Jens Siveke received honoraria as a consultant or for continuing medical education presentations from AstraZeneca, Bayer, Bristol-Myers Squibb, Eisbach Bio, Immunocore, Novartis, Roche/Genentech, and Servier; his institution receives research funding from Bristol-Myers Squibb, Celgene, Eisbach Bio, and Roche/Genentech; and he holds ownership and serves on the board of directors of Pharma15, all outside the submitted work. Wolfgang P. Fendler reports fees from SOFIE Biosciences (research funding), Janssen (consultant, speaker), Calyx (consultant, image review), Bayer (consultant, speaker, research funding), Novartis (speaker, consultant), Telix (speaker), GE Healthcare (speaker), Eczacıbaşı Monrol (speaker), Abx (speaker), and Amgen (speaker) outside the submitted work. No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: How does 68Ga-FAPI compare with 18F-FDG PET in the diagnosis of various malignancies?
PERTINENT FINDINGS: We compared the diagnostic accuracy of 68Ga-FAPI and 18F-FDG PET for the detection of various tumors. Tumor validation by a composite reference standard revealed higher diagnostic accuracy for 68Ga-FAPI PET in colorectal and prostate cancers, comparable diagnostic performance for cancers of the breast and head and neck, and lower diagnostic accuracy for bladder and kidney cancers, lymphoma, and myeloma.
IMPLICATIONS FOR PATIENT CARE: 68Ga-FAPI PET is particularly suited for the diagnosis of colorectal cancer, prostate cancer, and cancers of the breast and head and neck. 68Ga-FAPI PET offers theranostic screening and has the potential for more precise staging and management of patients who have these entities than is possible with 18F-FDG PET.
Footnotes
Published online Feb. 8, 2024.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication September 8, 2023.
- Accepted for publication December 20, 2023.
