Comparison of 68Ga-DOTA-FAPI and 18FDG PET/CT Imaging Modalities in the Detection of Liver Metastases in Patients with Gastrointestinal System Cancer

Purpose We aimed to compare the diagnostic performance of PET/CT imaging performed with 68Ga-DOTA-FAPI and 18FDG in detection of liver metastases in patients with gastrointestinal system (GIS) cancer. Methods A total of 31 patients who underwent 68 Ga-DOTA-FAPI and 18 F-FDG PET/CT examinations and diagnosed with GIS cancer (15 colorectal, 9 pancreas, 4 stomach and 3 other cancers) were included in the study. The presence of liver metastasis was decided based on histopathologic diagnosis, PET/CT, other radiologic examinations or tumor biomarker ndings, and both PET/CT imaging ndings were compared on the patient and lesion basis. Of the 31 patients, 28 were found as true positive with 68 Ga-DOTA-FAPI-PET/CT and 17 with 18 FDG-PET/CT. Of the 98 metastatic liver lesions determined according to our diagnostic criteria, 92 were found as true positive lesions with 68 Ga-DOTA-FAPI-PET/CT and 65 with 18 FDG-PET/CT. There was a statistically signicant difference between both imaging modalities in the patient and lesion based comparisons (p<0.05). When semiquantitative values (SUVmax, mlr) obtained from the lesions were compared between the two imaging methods, mlr values showed statistically signicant difference in all tumor subgroups (p<0.05). It of GIS


Introduction
Liver is one of the most critical organs for metastasis in gastrointestinal system cancers and liver metastasis may develop especially in about 70% of patients with adenocarcinomas [1]. The presence and prevalence of liver metastases is of great importance for treatment options and survival [2].
Radiological imaging modalities (USG, CT, MRI) and 18 Fluorodeoxyglucose-positron emission tomography/computed tomography ( 18 FDG-PET/CT) methods are commonly used in daily practice in detection of liver metastases and treatment planning [3,4]. 18 FDG that re ects tumor glucose metabolism is the most widely used PET radiopharmaceutical worldwide. On 18 FDG-PET/CT imaging, false negative or positive ndings may be seen depending on various factors such as the histopathological type, differentiation status, size and glucose metabolic activity of the tumor, or some benign pathologies, infective in ammatory processes, physiological radiopharmaceutical uptake.
Therefore, studies are underway to develop novel radiopharmaceuticals that are more speci c to the tumor and have a high target/background ratio.
Recently, a novel radiopharmaceutical developed on the basis of broblast activation protein (FAP) has been increasingly used in tumor diagnosis and staging.
Fibroblast activation protein (FAP) is a membrane glycoprotein highly expressed by cancer-associated broblasts (CAFs) seen in cancer types such as colorectal, pancreatic, ovarian and hepatocellular cancers with a strong desmoplastic reaction in which approximately 90% of gross tumor mass is characterized by stromal content [5][6][7].
Due to the relative speci city of FAP to tumor cells, FAP-speci c inhibitors (FAPI) have been produced rstly as anticancer drugs and then as tumor-targeted radiopharmaceuticals. [8,9].
Previous studies have reported that FAPI compound combined with 68 Ga-DOTA is very practical for diagnosis because of its in vivo positive pharmacokinetic properties and high tumor-background activity in different tumor types on PET imaging [10][11][12].
Especially due to physiological 18 FDG uptake in liver; metastases with low FDG uptake or small-size metastases may not be detected on 18 FDG-PET/CT. In this context, it is seen that there are several studies in the literature comparing various radiological methods (CT, MRI) and 18 FDG-PET/CT in detection of liver metastases of GIS cancers [13,14].
Recently, studies comparing 18 FDG-PET/CT and 68 Ga-DOTA-FAPI-PET/CT in various cancer types and primary liver tumors have also been published [15][16][17]. These studies have reported that GIS physiological radiopharmaceutical uptake and liver ground activity were low in 68 Ga-DOTA-FAPI-PET/CT. However, looking at the relevant literature, there is no study directly comparing these two PET radiopharmaceuticals ( 68 Ga-DOTA-FAPI and 18 FDG) in detecting liver metastases of GIS cancers. Therefore, in this study we aimed to compare these two PET radiopharmaceuticals in the detection of liver metastases in patients with GIS cancers and to show which PET imaging method would be more useful.

Patients
Institutional board approval was received for this study. A total of 31 patients with 12 being female and 19 male, and aged between 41-85 years old were included in this study. The patients consisted of those newly diagnosed or followed-up with GIS cancer. 68 GaDOTA-FAPI-PET/CT and 18 FDG-PET/CT were performed to them for staging or re-staging in our clinic. There was at least a two-week interval between two scans and no treatment or interventional procedure was performed in this time.
Patients with primary liver malignancy or another malignancy and those who were using steroids, cirrhosis, pregnant and under 18 years of age were excluded from the study.
Patient preparation and PET/CT imaging protocols 18 FDG-PET/CT Following fasting for 12 hours, blood glucose level of the 12 patients was analyzed and those with a blood glucose level < 150 mg/dL were injected with 5 MBq/kg intravenous (iv) 18 FDG.
Sixty minutes after the injection, PET/CT scan was performed with GE Discovery IQ; BGO, 5 Ring, 16 Slice combined PET/CT device (General Electric Company, Milwaukee, Wisconsin, USA) in our clinic. PET/CT scan was performed as whole body imaging from the vertex to the upper part of the femur. 68 GaDOTA-FAPI-PET/CT 68 Ga was obtained from a 68 Ge/ 68 Ga radionuclide generator, combined with DOTA-FAPI compound in the modular laboratory in our center and 68 Ga-DOTA-FAPI was synthesized (Modular lab-Easy, Eckert & Ziegler). At the end of the 60-minute preparation process, 60-70 % radiochemical yield was obtained. Control by TLC, HPLC revealed that the radiochemical purity was more than 98%. The synthesized 68 Ga-DOTA-FAPI was injected to the patient as iv at a dose of 2-3MBq/kg.

Evaluation of the Images and Patient-Lesion Analysis
PET/CT images were evaluated independently by 3 experienced nuclear medicine specialists on the patient and lesion basis.
The diagnosis was con rmed histopathologically only in 5 of 31 patients, while the decision was made according to imaging methods (PET/CT and other radiological methods) and tumor biomarker ndings (CA19-9, CEA, AFP, CA125) in 26 patients. While the lesion positivity was considered as "making a decision in favour of metastatic lesion nding in the liver in PET/CT images at least by 2 nuclear medicine specialists" and "elevation of tumor biomarkers in addition to liver metastasis ndings with radiological imaging methods", the other situations were considered as lesion negative.
The radiopharmaceutic uptake areas higher than liver parenchyma were quanti ed with SUVmax and mlr (metastatic lesion/liver uptake ratio) and accepted as positive lesion (liver metastasis) nding in also both PET/CT imaging methods. SUVmax and mlr values of the lesions that were considered positive according to our diagnostic criteria, but no positive ndings were detected in PET/CT images, were calculated by re-evaluation of the images.
Imaging ndings obtained from both PET/CT methods were compared on the patient and lesion basis.

Statistical Analysis
Radiological, histopathological and biochemical data of the patients were obtained from the computer registry system (electronic health records) of our hospital, and PET/CT images from the image archive in our clinic.
Data were statistically analyzed with SPSS 22.0 statistical analysis software.
Age, gender and sensitivity were expressed with mean, standard deviation and percentage descriptive statistics.
Normally distributed quantitative variables were expressed as mean ± standard deviation (SD) and non-normally distributed variables with [IQR, interquartile range].  Patient based comparison of the both imaging modalities is given in Table 1.  Lesion based comparison of the both imaging modalities is given in Table 2.   Table 3.

Discussion
As is known, liver metastases may often develop in GIS cancers, and the presence of metastasis of paramount importance in terms of treatment options and prognosis. As in many tumor groups, PET/CT is a globally accepted hybrid molecular imaging method in the diagnosis, staging and restaging of GIS cancers. 18 FDG, a glucose analog, is the most commonly used radiopharmaceutical in PET imaging owing to its physical features and practicality in many tumor groups. There are numerous studies comparing 18 FDG-PET/CT with radiological imaging modalities especially such as MRI in primary and metastatic liver tumors, and these studies have reported some limitations about 18 FDG-PET/CT [13,14].
The more speci c radiopharmaceuticals are also needed in liver tumors as in most tumor subgroups in PET imaging. Recently, it is seen that the studies have been conducted to detect various cancers and metastases with 68 GaDOTA-FAPI which is a novel PET radiopharmaceutical [15][16][17]. To our knowledge, our study is the rst in the literature comparing 68 GaDOTA-FAPI-PET/CT and 18 FDG-PET/CT methods in detection of liver metastases of GIS cancers. Because of the 68 GaDOTA-FAPI has been introduced as a new radiopharmaceutical in our clinic, this is a pilot study. On the other hand, although the number of the patients is low especially for tumor subgroups, data presented with total numbers of patients and lesions are su cient for statistical signi cance.
According to our patient based results; a higher rate of cases was detected with 68 GaDOTA-FAPI-PET/CT compared to those found with 18 FDG-PET/CT modality and the difference was statistically signi cant. We believe that lower radiopharmaceutic uptake of liver parenchyma in 68 GaDOTA-FAPI-PET/CT is one of the most important factors for this result. Although it was not exactly similar to our study in terms of patient groups and study design, in a study by Chen et al. [16] comparing both modalities, 68 GaDOTA-FAPI-PET/CT was found to be superior in detection of the cases and lesions, similarly.
In addition to the advantage provided by 68 GaDOTA-FAPI in terms of radiopharmaceutic uptake of liver parenchyma, previous studies have stated that 68 GaDOTA-FAPI lacks physiological gastric and intestines radiopharmaceutic uptakes that are seen with 18 FDG, and thus primary focus and peritoneal-mesenteric-omental metastases can be better detected in GIS cancers [15,16,18].
We believe that 68 GaDOTA-FAPI-PET/CT may be preferred instead of 18 FDG-PET/CT in GIS cancers in the near future, due to its' superiorities stated in the mentioned studies and its success in detecting liver metastases as demonstrated in our study.
While a signi cantly higher number of lesions were detected with 68 GaDOTA-FAPI-PET/CT compared to 18 FDG-PET/CT in total lesions number, no statistically signi cant difference was found for lesion numbers on each-one patient.
Although histopathological con rmation and distinguishing of the differentiation status could not be made in all lesions, as stated before this situtation can be explained with better detection of the lesions because of low liver parenchymal radiopharmaceutical activity. Furthermore, in addition to the high radiopharmaceutic uptake of liver parenchyma in 18 FDG-PET/CT imaging, especially small-sized lesions might have not been distinguished due to partial volume effect. Other conditions that may explain this results are other processes that may cause false positive and negative ndings in 18 FDG-PET/CT imaging.
On the other hand, some studies have reported that processes related to in ammation may also cause false positive ndings in 68 GaDOTA-FAPI-PET/CT imaging [18][19][20]. In the considering of the possible causes stated in these studies, cirrhotic patients were excluded from our study in order to minimize the processes that could resulted in in ammation and brosis in the liver parenchyma. And also, the decision for positive-negative lesions on the both PET/CT imaging was made taken into account mlr values in addition of SUVmax values to optimize in ammation that could be resulted from hepatosteatosis. In our study, it was found that false positivity in 1 patient and false negativity in 1 patient on 68 GaDOTA-FAPI-PET/CT.
Unfortunately, since majority of the metastatic lesions could not be con rmed histopathologically and studies in the literature on this subjects have just started to be published, we could not made a clear interpretation about the other possible mechanisms that might have caused false positive-negative results in 68 GaDOTA-FAPI-PET/CT.
Therefore, further studies are needed on this subject in which the lesions are con rmed histopathologically.
In the current study, SUVmax values of the lesions were also found to be in 68 Ga-FAPI-PET-CT higher than 18 FDG-PET/CT, and these results are consistent with the results of Chen et al. [18]. But, no statistically signi cant difference was found between the two PET/CT imaging modalities in terms of the SUVmax of the lesions in tumor subgroups.
On the other hand, statistically signi cant difference was found in favor of 68 Ga-FAPI-PET-CT for mlr values in all subgroups. For further studies that will compare both PET/CT imaging methods, we think that mlr would provide a signi cant contribution for the detection of liver lesions, and will be useful especially in the cases where making a decision based on SUVmax values is challenging.
The main limitations in our study can be listed as; the number of cases was relatively small especially in tumor subgroups in spite of the su cient total number of lesions, it was not designed as a prospective randomized study, and not all lesions were histopathologically con rmed.
In conclusion; despite the main limiting factors, our study showed that 68 GaDOTA-FAPI-PET/CT is a superior imaging modality over 18 FDG-PET/CT in detection of liver metastases of GIS cancers, and in current conditions it may be a complementary method for 18 FDG-PET/CT. We believe that 68 GaDOTA-FAPI-PET/CT will be a more practical method in the diagnosis, staging and restaging of GIS cancers because of both its success in detection of liver metastases and the advantages that it provides in visualization of the primary lesions and intra-abdominal metastases, and will take place in diagnosis and follow-up algorithm in line with the results of further studies to be conducted in this eld.

Declarations
Con ict of interest: All authors declare that they have no con ict of interest.
Ethical approval: Institutional board approval was received for this study.
Funding: There is no funding.

Authors' contributions
Transaxial fusion and MIP images of 68Ga-DOTA-FAPI-PET/CT (upper row) and 18FDG-PET/CT [lower row) of an operated patient diagnosed with rectum cancer. 68GaDOTA-FAPI-PET/CT images of the patient show multiple focal increased radiopharmaceutical uptake evaluated in favor of metastases in the liver, while no marked pathologic radiopharmaceutical uptake is observed in the liver on 18FDG-PET/CT images.