Predicting Radiation Esophagitis using 18F-FAPI-04 PET/CT in Patients with LA-ESCC Treated with Concurrent Chemoradiotherapy

Abstract


Introduction
Esophageal cancer is one of the most common malignant tumors of the digestive system globally, ranking seventh in terms of incidence and sixth in mortality overall in 2020 [1,2].Squamous cell carcinoma is the main histological type of esophageal cancer among patients in central and Southeast Asia [3], and de nitive concurrent chemoradiotherapy (CCRT) is the standard of care currently available for unresectable locally advanced esophageal cancer (LA-ESCC) [4][5][6].However, radiation esophagitis (RE) is a common adverse reaction in esophageal cancer patients treated with radiotherapy (RT) [7].The typical clinical features of acute RE include dysphagia, odynophagia and substernal pain, and this condition usually develops 2-3 weeks after the initiation of radiotherapy [8].RE is usually self-limited; however, studies have found that about 18% of patients receiving CCRT will develop RE with severity of grade 3 or higher [9,10].Grade 3 RE is commonly accompanied by many complications, such as ulcers, perforation, and even the formation of tracheoesophageal stula [11,12].These complications can negatively affect patients' quality of life and have a signi cant adverse impact on long-term survival [13].
Therefore, a means for the prediction or early identi cation of patients who may develop RE, and particularly grade 3 RE, is needed to minimize the risk of acute and long-term esophageal injury.
Fibroblast activation protein (FAP) is a member of the dipeptidyl peptidase (DPP) 4 protein family and has both endopeptidase and DPP activities [14,15].Its expression has been shown to increase signi cantly during tissue modeling and wound healing as well as in diseases such as arthritis, atherosclerosis and different cancers [16][17][18].Previous studies have shown that positron emission tomography (PET)/computed tomography (CT) imaging with a tracer targeting FAP, 68 Ga-DOTA-FAP inhibitor (FAPI)-04, offers superior diagnostic e cacy in patients with various types of cancer [19].We previously performed a pilot clinical study in which [ 18 F]AlF-NOTA-FAPI-04 (denoted as 18 F-FAPI-04), a novel tracer, was proven to be safe and to offer high speci city for FAP imaging [20].In one case report, a patient with esophagitis showed increased 68 Ga-FAPI uptake at the site of esophageal thickening [21].
However, the potential value of 18 F-FAPI-04 PET/CT imaging for identifying the development of RE has not been established in the literature.
The aim of the present study was to assess whether 18 F-FAPI-04 PET/CT can predict the development and severity of RE in patients with LA-ESCC treated with CCRT, and to explore the prediction parameters for RE, especially grade 3 RE.

Patient Cohort
This study was part of an ongoing prospective clinical study (SDZLEC2021-112-02) that received ethical approval from the a liated Cancer Hospital of Shandong First Medical University.This manuscript mainly reports a secondary analysis of this prospective trial.Participants were consecutively recruited from June 2021 to February 2022, and 16 of the 30 patients included herein were also included in a previous study [22].

Imaging Protocol
Participants were scanned within 1 week before RT (pre-RT) and after delivery of approximately 40 Gy (during-RT).The during-RT scan was performed once approximately 40 Gy of the total prescribed dose had been delivered, with the intent that a dose to this threshold would provide control of microscopic disease but still leave a reasonable amount of treatment remaining to alter the RT plan to include an additional RT boost.The total RT dose ranged from 50.4-69.4Gy, and intensity-modulated RT was delivered to all patients with X-ray (6 MV).RT was given according to the conventionally fractionated regimen of 1.8-2.0Gy/fraction for 5 days per week.
18 F-FAPI-04 was synthesized as described previously [23].Fasting and blood glucose measurement were not needed before scanning.After intravenous injection of 18 F-FAPI-04 (4.81 MBq/kg), patients rested for about 60 min before scanning was performed with an integrated in-line PET/CT system (GEMINI TF Big Bore; Philips Healthcare, Cleveland, OH, USA).Whole-body CT scans were obtained using a low-dose protocol (300 mAs, 120 kV, 512×512 matrix, rotation time of 1.0 s, and pitch index of 0.688; reconstruction with a soft-tissue kernel to a slice thickness of 2 mm) for attenuation correction.PET data were acquired in three-dimensional mode using a 200×200 matrix with an imaging time of 1 min per bed position.During image acquisition, the patients continued normal shallow breathing.Body-ctac-SB.
Lstcln, BioGraph 3D iterative reconstruction software with time-of-ight correction was used for attenuation and correction of PET and CT images.

Image Analysis
The attenuation-corrected PET, CT, and fused PET/CT images, which were displayed as coronal, sagittal, and transaxial slices, were viewed and analyzed on a Nuclear Medicine Information System (Beijing Mozi Healthcare Ltd, Beijing, China).All 18 F-FAPI PET/CT images were reviewed independently by two experienced nuclear medicine physicians (Z.F. and Y. Y., with 26 and 5 years of nuclear oncology experience, respectively).
Multiple planes for the same patient were superimposed via the MIM system to obtain a series of parameters: primary gross tumor volume (GTV, cc), RT dose, maximal esophageal dose, mean esophageal dose, volume of esophagus receiving ≥ 50 Gy (V50), and volume of esophagus receiving ≥ 60 Gy (V60).PET/CT data for all patients were transmitted to the MIM system.The GTV and esophagus (from cricoid to gastroesophageal junction) were contoured on the rst PET/CT, and this area was fused with the second PET/CT.We analyzed the esophageal area delineated after excluding the region within 5 mm of the GTV to reduce confounding 18 F-FAPI-04 PET/CT changes related to tumor response, as shown in Fig. 1.
18 F-FAPI-04 PET/CT parameters were generated by the MIM system.Regions of interest (ROIs) were normalized to the injected dose per kilogram of body weight to derive standardized uptake values (SUVs), which were calculated as: [measured activity concentration (Bq/mL) × body weight (g)]/injected activity (Bq).Normalized SUVs were used to represent FAPI activity in each ROI to improve reproducibility.The ratio of the maximum SUV of a ROI to the mean SUV for the pulmonary aorta was calculated and denoted as the tumor-to-background ratio (TBR blood ).The change in TBR blood from pre-RT to during-RT was denoted as ΔTBR blood .For controversial lesions, discussion among the imaging experts was carried out with consideration of results from other imaging modalities proceeded until a nal consensus was reached.

Statistical Analysis
Statistical analysis was performed using IBM SPSS Statistics for Windows version 25.0 (IBM, Armonk, NY, USA).Descriptive statistics were used to summarize the demographics and disease characteristics.
The Mann-Whitney U test was used to test the associations of 18 F-FAPI-04 PET/CT, clinical, and dosimetric parameters with the development of any grade of RE and grade 3 RE.Logistic regression analyses were performed to identify which of these parameters could predict development of any grade of RE (grade > 1) or speci cally grade 3 RE.Spearman's rank correlation coe cients were calculated to assess the relationships between parameters.Receiver operating characteristic curve analysis was used to determine the threshold values and accuracy of the parameters for toxicity prediction.All tests were two-sided, and P < 0.05 was considered statistically signi cant.

Patient Characteristics
From June 2021 to March 2022, 30 LA-ESCC patients (22 men, 18 women; median age: 66.5 years [interquartile range: 56-71 years]) were enrolled.The study ow diagram is presented in Fig. 2. All 30 patients were treated with CCRT, with a median RT dose of 59.9 Gy (interquartile range: 54-60 Gy) in fractions of 1.8 or 2.0 Gy.The speci c chemotherapy regimens followed are listed in Supplemental Table 1.Overall, 21 of 30 (70%) patients developed RE, and 6 of 30 (20%) patients developed grade 3 RE according to the Radiation Therapy Oncology Group (RTOG) criteria (Table 1).Figure 3 shows representative PET/CT imaging results for a patient without RE, and Fig. 4 provides representative PET/CT imaging results for a patient with grade 3 RE (grade 0).Patients who developed RE had signi cantly higher TBR blood (during-RT) (P = 0.003) and ΔTBR blood (P = 0.002) values than those who did not develop RE (Table 3).Additionally, patients who experienced grade 3 RE also had signi cantly higher TBR blood (during-RT) (P = 0.003) and ΔTBR blood (P = 0.003) values than those who developed RE rated lower than grade 3 (Table 2).predicted any grade RE (Fig. 5), and with higher cut-off values, TBR blood (during-RT) (AUC = 0.912; cut-off = 6.61) and ΔTBR blood (AUC = 0.922; cut-off = 4.21) also signi cantly predicted grade 3 RE (Fig. 5).

Correlations between Biomarkers
The correlations between 18 F-FAPI-04 PET/CT biomarkers, ECOG performance score, and dosimetric parameters are presented in Fig. 6.Except for a signi cant correlation between TBR blood (during-RT) and ΔTBR blood (r = 0.984; P < 0.01), no correlations were found between these variables (Fig. 6).
Associations between 18 F-FAPI-04 PET/CT Parameters, Clinical Variables, Dosimetric Parameters, and Development of Radiation Esophagitis According to univariate logistic regression analyses, V50 (P = 0.108), TBR blood (during-RT) (P = 0.020) and ΔTBR blood (P = 0.019) were independently associated with the development of RE (Table 3).Additionally, V50 (P = 0.146), TBR blood (during-RT) (P = 0.022) and ΔTBR blood (P = 0.022) were independently associated with the development of grade 3 RE (Table 3).Because of the signi cant positive correlation between TBR blood (during-RT) and ΔTBR blood , we only included ΔTBR blood and V50 in the subsequent multivariate analysis (Table 3), which showed that ΔTBR blood was signi cantly associated with the development of RE (P = 0.021) as well as the development of grade 3 RE speci cally (P = 0.038; Table 3).

Discussion
This prospective study demonstrated the rst time that the ΔTBR blood on 18 F-FAPI-04 PET/CT can be used as an independent predictor of RE, especially grade 3 RE, in patients with LA-ESCC.Compared with the value pre-RT, the TBR blood was increased signi cantly during RT in patients who developed RE.The prediction of RE, and RE of grade 3 speci cally, can improve the accuracy of prognosis prediction and guide treatment planning for LA-ESCC patients.
The results of this study showed that 18 F-FAPI-04 PET/CT imaging and speci cally the TBR blood and ΔTBR blood determined via this imaging modality, can be used to predict RE of any grade.While no studies investigating the use of FAPI-based imaging for RE prediction were found in the literature, a few published studies have explored the correlation between 18 F-FDG PET/CT parameters and RE.Our previous research found that 18 F-FDG uptake is signi cantly increased in esophagus during RT and that this increase may predict the occurrence of RE later in the course of treatment [24].However, the second time point adopted in this study is 45Gy, earlier than the previous 40Gy.In addition, our AUC is also higher than the previous study.Similarly, Mehmood et al. reported a signi cant increase in 18 F-FDG uptake in patients who developed RE during chemoradiotherapy [25].Therefore, we hypothesized that 18 F-FAPI-04 uptake might also have the potential to predict RE.This hypothesis was con rmed in the present study, and we also found radiation dose and T stage were correlated with RE.In contrast, Dzul et al. reported that the mean esophageal dose was the dosimetric parameter most correlated with grade 2 RE [26], and in their study, Mehmood et al. found that both V50 and V60 were predictors of the development of RE [25].
Furthermore, the present study also demonstrated that the TBR blood (during-RT), ΔTBR blood and V50 on 18 F-FAPI-04 PET/CT correlated with grade 3 RE.RE of this severity is known to potentially prolong treatment times and negatively affect overall survival [27].We observed a signi cant increase in the TBR blood (during-RT) for patients who developed grade 3 RE compared with that in patients who developed RE of a lower grade 3. Similarly, Mehmood et al. reported signi cantly higher 18 F-FDG uptake in patients with grade 3 RE at weeks 2 and 7 of RT compared with uptake values in patients with RE lower than grade 3 [25].These results suggest the possibility of screening high-risk patients in advance in order to carry out early interventions to reduce the incidence of RE.
The acute effects of RT on the esophagus consist of symptoms of substemal burning along with pain on swallowing, which occur approximately 2 weeks after initiation of a conventional RT course (after administration of approximately 20 Gy), and higher grade RE typically occurs in the late period of treatment [12,28].In the present study, the second 18 F-FAPI-04 PET/CT scan was conducted after patients had received a total RT dose of 40 Gy.Therefore, the imaging parameters evaluated in this study showed greater value for the prediction of grade 3 RE.
The main limitations of the present study include its single-center design and relatively small sample size.Further large-scale, multi-center clinical studies are needed to con rm our ndings before their clinical application.Furthermore, in this study, primary tumor regions were excluded to reduce confounding changes on 18 F-FAPI-04 PET/CT associated with tumor response.This approach may have excluded the area receiving the highest radiation dose, but this may also have resulted in underestimation of the examined parameters.Our analysis of the maximum SUV of primary tumors may have helped to reduce the impact of this limitation.Overall, further prospective trials are required to con rm the role of 18 F-FAPI-04 PET/CT imaging for predicting RT toxicity prediction in patients with LA-ESCC.

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download.

Figure 2 Study
Figure 2

Table 1
shown in Table2, the patient groups with or without RE showed no differences in age, sex, Eastern Cooperative Oncology Group (ECOG) performance score, N stage, primary GTV, RT dose, maximal esophageal dose, mean esophageal dose, V50, and V60.T stage was signi cant correlated with RE (P = 0.047), and V50 was signi cantly increased in association with grade 3 RE (P = 0.021). As

Table 3
Predictive Ability of Biomarkers for Any Grade of RE (Grade ≥ 1) and Grade 3 on Univariate and Eastern Cooperative Oncology Group; V50: volume of esophagus receiving ≥ 50 Gy; V60: volume of esophagus receiving ≥ 60 Gy; RT: radiotherapy Receiver-operating characteristic curves were generated to evaluate the predictive accuracy of 18 F-FAPI-04 PET/CT parameters for identifying any grade RE and grade 3 RE (Fig.5).High TBR blood (during-RT) (area under the curve [AUC] = 0.902; cut-off = 1.53) and ΔTBR blood (AUC = 0.911; cut-off = 4.19) signi cantly ΔTBR blood may be a useful imaging parameter for identifying ESCC patients at increased risk for the development of RE of any grade, especially grade 3 RE.Such predictive ability could provide an opportunity to consider earlier aggressive interventions for at-risk patients and facilitate investigations into whether such interventions can alter the course of toxicity.
DeclarationsAuthor contribution All authors contributed to the study conception and design.Material preparation, data collection and analysis were performed by Xiao Han, Mingquan Zhang, Zhengshuai Mu, Kailin Qiao, Jing Jia.Zheng Fu and Jiazhong Ren were responsible for reviewing all 18 F-FAPI PET/CT images.The rst Figures Page 17/21