FAP expression in alpha cells of Langherhans insulae—implications for FAPI radiopharmaceuticals’ use

Radiopharmaceuticals targeting fibroblast activation protein (FAP) alpha are increasingly studied for diagnostic and therapeutic applications. We discovered FAP expression at immunohistochemistry (IHC) in the alpha cells of the Langerhans insulae of few patients. Therefore, we planned an investigation aimed at describing FAP expression in the pancreas and discussing the implications for radioligand applications. We retrospectively included 40 patients from 2 institutions (20 pts each) according to the following inclusion/exclusion criteria: (i) pathology proven pancreatic ductal adenocarcinoma and neuroendocrine tumors (NET), 10 pts per each group at each center; (ii) and availability of paraffin-embedded tissue; and (iii) clinical-pathological records. We performed IHC analysis and applied a semiquantitative visual scoring system (0, negative staining; 1, present in less than 30%; 2, present in more than 30% of the area). FAP expression was assessed according to histology—NET (n = 20) vs ductal adenocarcinoma (n = 20)—and to previous treatments within the adenocarcinoma group. The local ethics committee approved the study (No. INT 21/16, 28 January 2016). The population consisted of 24 males and 16 females, with a median age of 68 and a range of 14–84 years; 8/20 adenocarcinoma patients received chemotherapy. In all the Langerhans insulae (40/40), pancreatic alpha cells were found to express FAP, with a score of 2. No difference was found among NET (20/20) and adenocarcinoma (20/20), nor according to neoadjuvant chemotherapy in the adenocarcinoma cohort (received or not received). Pancreatic Langerhans islet alpha cells normally express FAP. This is not expected to influence the diagnostic accuracy of FAP-targeting tracers. In the therapeutic setting, our results suggest the need to better elucidate FAPI radioligands’ effects on the Langerhans insulae function.


Introduction
The fibroblast activation protein alpha (FAPα) is a type II transmembrane proteolytic enzyme [1]. FAPα has both dipeptidyl peptidase and a collagenolytic activity; thus, it is capable of degrading gelatine and type I collagen [2]. It has been described to be underexpressed in healthy tissues while it is overexpressed in disease states such as woundhealing granulation tissue and neoplasms [3]. Indeed, FAP positivity has been found in 50-100% of cancer patients. Within tumors, FAP-positive stromal cells generate a protumorigenic microenvironment by producing a variety of This article is part of the Topical Collection on Oncology-General.
* Margarita Kirienko margarita.kirienko@icloud.com structural and regulatory molecules, which contribute to local immunosuppression, extracellular matrix remodeling, and stimulation of angiogenesis. Indeed, a higher FAP expression is associated with higher local invasion and increased risk of nodal metastases. Additionally, in patients affected by FAP-positive cancers, a shorter survival has been reported [4].
In recent years, FAP-targeted diagnostics and therapy research has gained a lot of interest, even in nuclear medicine. By exerting effects on the tumor microenvironment, FAPtargeting approaches (e.g., low molecular weight FAP inhibitors, FAP antibodies and their conjugates, FAP vaccines, and FAP-targeting genetically engineered chimeric antigen receptor T cells-FAP-CAR T) have been tested to disrupt several hallmarks of cancer [4]. In nuclear medicine, small-molecule and peptide-based FAP-targeting radiopharmaceuticals are currently the most promising compounds [5] for both diagnostic (PET and SPECT) and therapeutic purposes [6,7].
Although several ongoing studies are expected to provide the evidence for the introduction of these radiopharmaceuticals into the clinical arena [8], the pathophysiology of FAP expression-in terms of sites and role-has not been extensively elucidated. Several bioactive peptides and structural proteins (e.g., neuropeptide Y, peptide YY, substance P, and B-type natriuretic peptide) have been proposed to be FAP substrates. Moreover, in a few patients, we discovered at immunohistochemistry (IHC) that alpha cells of the Langerhans insulae expressed FAP. Therefore, we planned an investigation aimed at describing FAP expression in the pancreas and discussing its implications for radioligand applications.

Patients
For the present retrospective multicenter study, we applied the following inclusion/exclusion workflow. In 2 Italian institutions (Institution 1-Fondazione IRCCS Istituto Nazionale dei Tumori and Institution 2-IRCCS Humanitas Research Hospital), we randomly selected (i) 10 patients affected by a pathology proven pancreatic neuroendocrine tumor (NET) and 10 patients affected by a pathology proven pancreatic adenocarcinoma (total n = 20 per each institution), (ii) for whom formalin-fixed paraffinembedded healthy tissue, and (iii) clinical-pathological records were available. For all patients, we collected the following data: age, sex, tumor grading, pathological TNM staging, and treatment before surgery.
The local ethics committee approved the study; informed consent was waived due to the retrospective nature of the investigation (No. INT 21/16, 28 January 2016).

Immunohistochemistry
We performed an immunohistochemical analysis of the pancreas using FAP monoclonal antibody similarly to Mona et al. [9]. Briefly, we cut 2.5/3 micron-thick sections from paraffin blocks; then we dried, de-waxed, rehydrated, and unmasked them with Dako PT-link, EnVision™ FLEX Target Retrieval Solution (High pH-96 °C-15 min). Rabbit monoclonal anti-FAP alpha (Clone EPR20021-Abcam-diluition 1:250) was incubated with a commercially available detection kit (EnVi-sion™ FLEX + , Dako, Agilent) in an automated immunostainer (Dako Autostainer Link 48-Agilent). An experienced surgical pathologist (M.M.) confirmed the histological diagnoses and performed the IHC analysis using a semiquantitative visual scoring system (0, negative staining; 1, present in less than 30% of the area; 2, present in more than 30% of the area).

Statistical analysis
Frequency tables and descriptive statistics were used to summarize baseline study cohort data and to assess FAP expression according to histology (NET vs ductal adenocarcinoma) and previous treatments.

Results
Patient characteristics are summarized in Table 1.

FAP expression
In all the specimens (40/40), FAP expression was demonstrated in the human endocrine pancreas. Particularly, the pancreatic alpha cells in all the patients (40/40) expressed FAP with the score of 2 ( Fig. 1).
Additionally, in all specimens, FAP expression was found in fibrous tissue close to pancreatic primary tumors and in large pancreatic ducts. Intense FAP expression was detected in hyperplastic Langerhans insulae. Specifically, in 2/20 NET patients with hyperplastic Langerhans insulae, FAP expression resulted to involve the whole insulae (Fig. 2).
In one patient affected by adenocarcinoma, intense FAP expression was found in the fibrotic tissue, while epithelial pancreatic cells resulted negative (Fig. 3).

FAP expression according to histotype and chemotherapy
In pancreatic alpha cells, no differences in FAP expression score were reported among NET and adenocarcinoma patients (Fig. 4A).
Eight out of 20 patients affected by pancreatic adenocarcinoma received neoadjuvant chemotherapy (Table 1). We did not observe differences in terms of location and intensity score between the two groups (Fig. 4B).

Discussion
Our study demonstrated that pancreatic alpha cells normally express FAP. Indeed, the alpha cells within the Langerhans islets in all patients (40/40) resulted FAP positive at immunohistochemistry. Additionally, no difference in FAP expression was found according to different pancreatic cancer types (NET vs adenocarcinoma) nor according to neoadjuvant chemotherapy in the adenocarcinoma cohort (received or not received).
Our results are in line with data from Busek et al. [10], who demonstrated the co-expression of FAP and dipeptidyl peptidase-IV (DPP-IV) in pancreatic alpha cells in adult humans, potentially implying modulation of the paracrine signaling in the human Langerhans islets.
The implications for FAPI-PET/CT imaging are expected to be limited since Langerhans insulae are definitely smaller (mean islet diameter of 108.92 μm (± 6.27 μm) [11] than the resolution limit of a PET scanner (2.36 mm FWHM) [12]. According to biodistribution studies, Mona et al. found  [14]. Figure 5 shows no significantly increased uptake in the pancreas as compared to the background in the rectal cancer patient scanned at the Department of Nuclear Medicine, Union Hospital, Tongji Medical College, and Huazhong University of Science and Technology in Wuhan, China. Consequently, alpha cells' uptake does not affect the detection of primary pancreas neoplasms. According to the recent paper from Hirmas et al., the mean primary exocrine pancreatic cancer tumor-to-background ratio was prominent and significantly higher for [ 68 Ga]Ga-FAPI than [ 18 F]FDG (14.7 vs. 3.0, p < 0.001) [15]. Indeed, FAP expression by cancer lesions confirms the opportunity to detect and stage malignant lesions, both NET and adenocarcinoma lesions.
In NET patients, FAPI uptake on PET/CT has been demonstrated by initial investigations [16,17]. Kratochwil et al. reported NET lesions to display an average SUVmax ranging between 6 and 12 [17]. A case of a NET patient, scanned by our group at the Department of Nuclear Medicine and Minnan PET Center, Xiamen University, is illustrated in Fig. 6. Moreover, Kreppel suggested FAPI PET/CT-derived parameters for patient risk stratification. Indeed, they found the ratio between volumes of liver metastases positive on [ 68 Ga]Ga-FAPI and [ 68 Ga]Ga-DOTA-TOC scans to be significantly and strongly correlated with Ki-67 (rho = 0.808, p < 0.01) [18].
In pancreatic ductal carcinoma, Kratochwil et al. reported that the average SUV max ranged between 6 and 12, while the tumor-to-background ratios were more than 3 [17]. Hirmas et al. found a mean primary tumor and metastatic lesions SUV max to be 13.2 and 9.4, respectively [15]. These high uptake values resulted in high image contrast and excellent tumor delineation. Moreover, in the study by Röhrich et al., the authors found [ 68 Ga]Ga-FAPI-4 and [ 68 Ga]Ga-FAPI-46 PET/CT changed TNM staging in 10/19 patients compared to the contrast-enhanced CT [19].
The implications for FAPI-targeted treatments on glucose metabolism have not been elucidated yet. The initial reports in a variety of cancer types have not investigated nor reported increased glucose levels after radioligand treatment. Ballal et al. in a cohort of 15 radioiodine-refractory differentiated thyroid cancer patients found that none of the patients' experienced grade 3 or 4 hematological, renal, or liver toxicity after [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 treatment [20]. In  thrombocytopenia and anemia were the most prevalent [21].
As for dosimetry assessments, a recent study found the kidney and colon to be dose-limiting organs with [ 177 Lu] Lu-DOTA.SA.FAPi. On the other hand, the highest estimated absorbed radiation dose by [ 177 Lu]Lu-DOTAGA. (SA.FAPi) 2 was observed in the colon, followed by the gall bladder, pancreas, and kidneys [22]. The dosimetry assessments in the study by Fendler et al. focused on bone marrow, liver, lung, and kidneys [21]. In view of the peculiar Langerhans insulae anatomy, future micro dosimetry studies might elucidate the absorbed dose to the endocrine component of the pancreas and establish which radionuclide holds the best properties for FAPI radioligand treatments.
FAP is known to be expressed in non-malignant conditions, particularly those that involve tissue remodeling [3]. A wide range of diseases have been found to be positive on [ 68 Ga]Ga-FAPI PET/CT. These also include non-oncological lesions of the pancreas (e.g., pancreatic pseudocysts, sites of prior pancreatitis, and foci of IgG 4-related disease) [23]. In our cohort, indeed, we found FAP expression at the fibrotic tissue within the pancreas (Fig. 3). Moreover, FAPI-PET/CT positive findings comprised infections [24][25][26], heart infarction [27,28], Crohn's disease [29], Erdheim-Chester disease [30], inflammatory arthritis [31,32], thyroiditis [33][34][35], idiopathic retroperitoneal fibrosis [36], chronic cholecystitis, degenerative osteophyte [37], and vertebral body fracture [38]. This great number of non-cancer-positive sites may represent a challenge for FAP-targeting imaging interpretation and might introduce some limitations in using FAPI radioligand therapy. In this scenario, FAP-targeting imaging should be used with a double purpose: first, to demonstrate the treatment target, and second, to assess non-malignant FAPexpressing conditions that may influence treatment eligibility.
We acknowledge some limitations in our study. Firstly, the retrospective design; however, multicenter data availability provided robustness to the results. Some of the included patients received different chemotherapy regiments before surgery which might have influenced the results; however, immunohistochemistry demonstrated score 2 expression of FAP in both treatment naïve and treated groups, suggesting expression is not significantly related to treatment. We have not performed PET/CT imaging since it was out of the scope of the present study.
In conclusion, alpha cells in the pancreatic Langerhans islets normally express FAP, in both chemotherapy-treated and treatment naïve patients. In view of promising anticancer activity using FAP-targeting radioligands, as reported in recent literature, these radiopharmaceuticals are expected to be increasingly used and early translated to clinics. Our results suggest the need to better elucidate FAPI radioligand therapy effects on the Langerhans insulae cells' function and establish the limit absorbed dose.
Author contribution Margarita Kirienko, Giovanni Centonze, Ettore Seregni, and Massimo Milione contributed to the study's conception and design. Material preparation, data collection, and analysis were performed by Giovanni Centonze, Giovanna Sabella, Mauro Sollai, Martina Sollini, and Luigi Terracciano. PET/CT clinical cases were curated by Xiaoli Lan and Haojun Chen. The first draft of the manuscript was written by Margarita Kirienko, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.