Application of 18F-fluorodeoxyglucose positron emission tomography/computed tomography imaging in recurrent anastomotic tumors after surgery in digestive tract tumors

Abstract

BACKGROUND

This study was to investigate the application value of whole-body dynamic ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) imaging in recurrent anastomotic tumors of digestive tract after gastric and esophageal cancer surgery. Postoperative patients with gastric and esophageal cancer have a high risk of tumor recurrence, and traditional imaging methods have certain limitations in early detection of recurrent tumors. Whole-body dynamic ¹⁸F-FDG PET/CT imaging, due to its high sensitivity and specificity, can provide comprehensive information on tumor metabolic activity, which is expected to improve the early diagnosis rate of postoperative recurrent tumors, and provide an important reference for clinical treatment decision-making.

AIM

To investigate the clinical value of whole-body dynamic ¹⁸F-FDG PET/CT imaging in differentiating anastomotic recurrence and inflammation after the operation of upper digestive tract tumors.

METHODS

A retrospective analysis was performed on 53 patients with upper digestive tract tumors after operation and systemic dynamic ¹⁸F-FDG PET/CT imaging indicating abnormal FDG uptake by anastomosis, including 29 cases of gastric cancer and 24 cases of esophageal cancer. According to the follow-up results of gastroscopy and other imaging examinations before and after PET/CT examination, the patients were divided into an anastomotic recurrence group and anastomotic inflammation group. Patlak multi-parameter analysis software was used to obtain the metabolic rate (MRFDG), volume of distribution maximum (DVmax) of anastomotic lesions, and MRmean and DVmean of normal liver tissue. The lesion/background ratio (LBR) was calculated by dividing the MRFDG and DVmax of the anastomotic lesion by the MRmean and DVmean of the normal liver tissue, respectively, to obtain LBR-MRFDG and LBR-DVmax. An independent sample t test was used for statistical analysis, and a receiver operating characteristic curve was used to analyze the differential diagnostic efficacy of each parameter for anastomotic recurrence and inflammation.

RESULTS

The dynamic ¹⁸F-FDG PET/CT imaging parameters MRFDG, DVmax, LBR-MRFDG, and LBR-DVmax of postoperative anastomotic lesions in gastric cancer and esophageal cancer showed statistically significant differences between the recurrence group and the inflammatory group (P < 0.05). The parameter LBR-MRFDG showed good diagnostic efficacy in differentiating anastomotic inflammation from recurrent lesions. In the gastric cancer group, the area under the curve (AUC) value was 0.935 (0.778, 0.993) when the threshold was 1.83, and in the esophageal cancer group, the AUC value was 1. When 86 is the threshold, the AUC value is 0.927 (0.743, 0.993).

CONCLUSION

Whole-body dynamic ¹⁸F-FDG PET/CT imaging can accurately differentiate the diagnosis of postoperative anastomotic recurrence and inflammation of gastric cancer and esophageal cancer and has the potential to be an effective monitoring method for patients with upper digestive tract tumors after surgical treatment.

Key Words: Gastric cancer; Esophageal cancer; Anastomosis; Dynamic positron emission tomography; Differential diagnosis; Metabolic rate

Core Tip: This study evaluated the efficacy of whole-body dynamic ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) imaging in the diagnosis of recurrent anastomotic tumors of the digestive tract in postoperative patients with gastric and esophageal cancer. This study evaluated the sensitivity and specificity of the whole body dynamic ¹⁸F-FDG PET/CT scan in patients with recurrent tumors by analyzing the image data and combining with clinicopathological data. The accuracy and clinical application value of this imaging technology in early identification of recurrent tumors and its potential role in postoperative patient management are explored, with a view to providing new diagnostic tools and methods for clinical practice.

INTRODUCTION

At present, a variety of molecular and imaging methods, such as tumor markers, endoscopy, computed tomography (CT), and positron emission tomography (PET)/CT, have been used for the detection of recurrence, staging, and efficacy evaluation of upper gastrointestinal malignancies after treatment[1-3]. However, tumor markers cannot diagnose the recurrence site, ordinary endoscopy cannot detect recurrent lesions outside the lumen, and CT diagnostic ability only depends on the morphological changes and anatomical structure dislocation of the affected organs, so the three factors have obvious limitations in detecting cancer recurrence[4].

PET/CT is an imaging technique that combines functional characterization and anatomic localization and has been widely used in the diagnosis, staging, prognosis assessment, and postoperative treatment monitoring of gastric cancer and esophageal cancer[5-8]. Early studies on static PET/CT using ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) as a tracer reported its high sensitivity, specificity, and accuracy in detecting recurrent gastric cancer and esophageal cancer[9]. Standardized uptake value (SUV) is often used for quantitative analysis in conventional PET static imaging. However, SUV value is a semi-quantitative parameter, and the measurement’s accuracy may be affected by non-specific uptake, such as uptake time, blood glucose level, and population heterogeneity[10]. This makes PET less useful for diagnosing diseases. In addition, it has certain limitations in the detection of infectious or inflammatory lesions and is often difficult to distinguish from neoplastic lesions[11-13].

In recent years, some studies have introduced tracer kinetic model analysis in dynamic PET/CT imaging and emphasized its diagnostic role in tumorigenesis[14]. The study found that dynamic analysis-based PET/CT imaging can improve the visibility of lesions in clinical oncology while also more accurately reflecting post-treatment changes, contributing to the evaluation of treatment response[15]. Conventional static ¹⁸F-FDG PET/CT imaging produces a single SUV image, while dynamic analysis can distinguish between FDG involved in metabolism and free FDG in the tissue, resulting in a metabolic rate (MR) image and a volume distribution (DV) image. Each parameter image conveys a different physiological significance, which can provide more physiological and metabolic information in the pathological tissue. Many earlier studies have also shown that dynamic PET/CT may be useful for telling the difference between benign and malignant lesions[16]. For example, it can help tell the difference between neuroendocrine tumors and physiological uptake of pancreatic unguis, as well as between benign and malignant pulmonary nodules and metastatic and non-metastatic lymph nodes. Compared with the SUVmax of static PET, dynamic 18F-FDG PET with an irreversible uptake rate (Ki) showed better diagnostic efficacy in differentiating subcutaneous and in-situ inflammation from malignancy. However, none of these studies investigated the potential value of the dynamic PET parameters DVmax and its lesion/background ratio (LBR) in differentiating benign and malignant lesions[17].

There are few reports on the differential diagnosis of postoperative anastomotic recurrence and inflammation with dynamic PET/CT in upper gastrointestinal tumors. Therefore, this study intends to analyze the whole-body dynamic ¹⁸F-FDG PET/CT image data of 53 patients with gastric cancer and esophageal cancer after surgery. To evaluate the differential diagnosis performance of dynamic PET metabolic parameters MR, DVmax, and LBRs for the recurrence and inflammation of the upper Xiaohuo channel.

MATERIALS AND METHODS

Data collection

We looked back at the medical records of 53 people who came to our hospital between February 2023 and December 2023 with either stomach cancer or esophageal cancer. These people had whole-body dynamic ¹⁸F-FDG PET/CT imaging, which showed abnormal uptake of FDG by anastomosis. Of these 53 people, 29 had stomach cancer and 24 had esophageal cancer. Inclusion criteria: (1) Patients diagnosed with gastric cancer or esophageal cancer by operation and pathology who underwent whole-body dynamic ¹⁸F-FDG PET/CT examination and had abnormal metabolism of FDG at the anastomosis; (2) The whole-body dynamic ¹⁸F-FDG PET/CT examination was performed at least 1 month after the operation; (3) The surgical pathology showed no cancer involvement at the resection margin; and (4) No antitumor therapy was performed between the postoperative and PET/CT examinations. Exclusion criteria: (1) Combined with other malignant tumors; and (2) Relevant clinical data are missing. The Ethics Committee of our hospital reviewed and approved this study.

Whole body dynamic ¹⁸F-FDG PET/CT imaging equipment and image collection

All patients with gastric cancer and esophageal cancer after surgery fasted for at least 6 hours before examination and had fasting blood glucose ≤ 11.1 mmol/L. A Siemens Biograph Vision 450 PET/CT system was used to collect data. In this study, a two-stage scanning protocol for short-time dynamic PET collection was used: CT images of the heart area were collected, and the list-mode dynamic PET collection of a single bed in the same part simultaneously. After resting for about 60 minutes, CT images from the skull base to the middle of the femur were collected, followed by dynamic PET scans of three passes at the same site to capture the late changes of tracers in plasma and tissue.

Image reconstruction

In the TrueD system of the Siemens Biograph Vision workstation, 0-6 minutes single-bed list-mode dynamic PET images and 3 frames of dynamic PET images collected in the next 15 minutes were obtained, and the input function of the proximal descending aorta was calculated. Parameter images of FDG MR (MRFDG) and apparent volume of distribution (DVFDG) were generated using the Patlak reconstruction algorithm. Reconstruction parameters: Patlak recon method, OSEM True X + TOF, 4 iterations, 5 subsets, reconstruction matrix of 220 × 220, scattering correction, no filtering function. Low-dose CT scan parameters for attenuation correction: Tube voltage 120 keV, tube current 30-180 mA, spiral pitch factor 0.8, layer thickness 3 mm.

Image analysis and parameter calculation

Two skilled nuclear medicine doctors who were unaware of the patient’s other clinical details reviewed the ¹⁸F-FDG PET/CT images, and the Siemens image post-processing workstation allowed for a visual comparison of the MRFDG and DV images. The volume of interest (VOI) of lesions was mapped automatically, with 41% of the maximum pixel value as the segmentation threshold. The VOI was mapped in the areas of the lesion that had anastomotic inflammation or recurrence, and the MRFDG and DVmax values were found. The liver VOI was marked out in the healthy tissue of the right lobe with a sphere of fixed diameter and volume of about 3 cm³. The liver’s mean parameters, MRmean and DVmean, were then found. The method of calculating LBRs is to divide the target lesion MRFDG and DVmax by the MRmean and DVmean of the normal liver tissue, respectively, to obtain LBR-MRFDG and LBR-DVmax. The calculation method of LBR is similar to that of previous Ki analysis studies.

Gastroscopic, pathological, clinical, and imaging follow-up

For patients with anastomotic ¹⁸F-FDG uptake found by PET/CT examination, in order to confirm the nature of the lesion, gastroscopic pathology was used before and after PET/CT examination for a definitive diagnosis. For patients who did not undergo gastroscopy, PET/CT examination time was taken as the starting point of follow-up, and recurrence or inflammation was judged as the end point of follow-up. The average follow-up time of these patients was 3.8 ± 0.6 months; the longest follow-up time was 6 months. According to the CT results of multiple follow-up visits and tumor indicators, recurrence or inflammation was determined. During the follow-up period, if the anastomotic area was found to be significantly thicker than before, combined with the increase of tumor markers in the patient, the diagnosis was anastomotic recurrence. If there are no obvious signs of recurrence, the diagnosis is anastomotic inflammatory lesions.

Statistical analysis

The statistical software SPSS 27.0 was used to analyze the data. Quantitative data conforming to a normal distribution were represented by the mean ± SD, and an independent sample t test was used for comparison between groups. For quantitative data that did not conform to the normal distribution, the Wilcox rank sum test was used to compare between groups. We used receiver operating characteristic (ROC) curves to look at how well each parameter could tell the difference between anastomotic inflammation and recurrence. The Jorden index was used to find the best cut-off points for each variable, and the area under the curve (AUC), sensitivity, and specificity were calculated. P < 0.05 was considered statistically significant.

RESULTS

Clinical data characteristics of patients

A total of 53 patients with gastric or esophageal cancer who underwent whole-body dynamic ¹⁸F-FDG PET/CT imaging and showed abnormal uptake of FDG by anastomosis were selected. There were 29 cases of gastric cancer, mean age (60.03 ± 10.56). There were 26 males and 3 females. 24 patients with esophageal cancer, mean age (62.04 ± 9.43) years old; there were 20 males and 4 females. The interval between patients’ operation and dynamic ¹⁸F-FDG PET/CT examination ranged from 3 to 56 months, with an average of 23.5 months.

Surgical method

Total gastrectomy was performed in 5 cases, proximal subtotal gastrectomy in 8 cases, and distal subtotal gastrectomy in 16 cases. There were 5 cases of esophagojejunal anastomosis, 8 cases of esophagogastric anastomosis, and 16 cases of gastrojejunal anastomosis. Radical esophagectomy was performed in all patients with esophageal cancer, including 6 cases of cervical anastomosis and 18 cases of intrathoracic anastomosis.

Analysis of pathological examination results

In gastric cancer patients: (1) Location of the lesions: 3 cases in the stomach body, 10 cases in the cardia, 16 cases in the gastric antrum pyloric region; and (2) Histological types: 25 adenocarcinomas (22 poorly differentiated, 3 moderately differentiated), 2 sig-ring cell carcinomas, 1 adeno-squamous carcinoma, and 1 undifferentiated carcinoma. Esophageal cancer patients: (1) The location of the lesions: 6 cases in the upper thoracic segment, 17 cases in the middle thoracic segment, and 1 case in the lower thoracic segment; and (2) Histological types: 5 cases were poorly differentiated squamous cell carcinoma, 16 cases were moderately differentiated squamous cell carcinoma, and 3 cases were highly differentiated squamous cell carcinoma.

Lesion analysis in the inflammation group and the recurrence group

Following up with gastroscopy and other imaging examinations, 16 cases of anastomotic inflammatory lesions and 13 cases of anastomotic recurrence were confirmed. There were 13 cases of anastomotic inflammatory lesions and 11 cases of anastomotic recurrence. In gastric cancer and esophageal cancer, there was no significant difference in age or weight between the recurrent anastomosis group and the inflammatory group (P > 0.05). There was a statistical difference in metabolic tumor volume between the two groups in gastric cancer (P = 0.044), but there was no statistical difference in tumor metabolic volume between the two groups in esophageal cancer (P > 0.05) (Table 1).

Table 1 Characteristics of anastomotic inflammation group and recurrent group in patients with gastric cancer and esophageal cancer after operation.

Characteristics	Anastomotic		Z	P value
	Inflammation	Recurrence
Gastric cancer (n)	16	13	-	-
Gender (male/female) (n)	15/1	11/2	-	-
Age (mean ± SD, year)	59.56 ± 10.91	60.62 ± 10.52	-0.439	0.661
Body weight (mean ± SD, kg)	55.44 ± 10.35	59.85 ± 9.10	-1.187	0.235
ΜΤV (mean ± SD, cm3)	7.10 ± 3.95	10.71 ± 4.71	-2.017	0.044
Clinical staging (I, II, III, IV) (n)	3, 5, 8, 0	1, 6, 4, 2	-	-
Esophageal cancer (n)	13	11	-	-
Gender (male/female) (n)	11/2	9/2	-	-
Age (mean ± SD, year)	62.38 ± 10.39	61.64 ± 8.65	-0.174	0.862
Body weight (mean ± SD, kg)	56.00 ± 9.08	60.45 ± 5.91	-1.219	0.223
ΜΤV (mean ± SD, cm3)	6.27 ± 3.44	7.75 ± 3.56	-1.13	0.258
Clinical staging (I, II, III, IV) (n)	1, 4, 7, 1	2, 5, 3, 1	-	-

MTV: Metabolic tumor volume.

Dynamic PET parameter analysis

The dynamic PET parameters and LBR of anastomotic inflammation and recurrent lesions in postoperative patients with gastric and esophageal cancer are shown in Table 2. In the two types of patients after surgery for upper gastrointestinal tumors, the mean values of MRFDG and LBR-MRFDG in the anastomotic recurrence group were higher than those in the anastomotic inflammation group, while the mean values of DVmax and LBR-DVmax were lower than those in the anastomotic inflammation group. There were significant differences in metabolic parameters (MRFDG, DVmax, LBR-MRFDG, and LBR-DVmax) between the anastomotic recurrence group and the inflammation group (P < 0.05).

Table 2 Parameter analysis of fluorodeoxyglucose uptake in anastomotic inflammation and recurrent lesions.

Parameters	Anastomotic inflammation (mean ± SD)	Anastomotic recurrence (mean ± SD)	t	P value	95%CI
Gastric cancer
MRFDG (μmol/min/mL)	0.10 ± 0.03	0.19 ± 0.08	-3.819	0.002	-0.139 to -0.039
DVmax (%)	145.74 ± 47.11	94.71 ± 32.17	3.321	0.003	19.504 to 82.546
LBR-MRFDG	1.60 ± 0.42	2.93 ± 1.05	-4.446	< 0.001	-2.026 to -0.714
LBR-DVmax	2.73 ± 1.19	1.47 ± 0.67	3.627	0.001	0.547 to 1.990
Esophageal cancer
MRFDG (μmol/min/mL)	0.10 ± 0.03	0.19 ± 0.07	-3.742	0.002	-0.137 to -0.037
DVmax (%)	144.46 ± 37.99	96.53 ± 23.60	3.628	0.001	20.532 to 75.335
LBR-MRFDG	1.60 ± 0.45	3.14 ± 1.14	-4.188	0.001	-2.331 to -0.741
LBR-DVmax	2.06 ± 0.70	1.10 ± 0.42	3.963	< 0.001	0.458 to 1.462

MRFDG: Fluorodeoxyglucose metabolic rate; LBR: Lesion/background ratio; DV: Volume distribution; CI: Confidence interval.

ROC curve analysis of dynamic PET parameters and LBR

The diagnostic efficiency of dynamic PET parameters and LBRs is shown in Table 3. In patients with gastric cancer and esophageal cancer after surgery, the AUC values of MRFDG, LBR-MRFDG, DVmax, and LBR-DVmax in differential diagnosis of anastomotic inflammation and recurrence ranged from 0.808 to 0.935, with P values less than 0.05. All the parameters showed better sensitivity and specificity. Among them, the threshold values of MRFDG and DVmax in the gastric cancer group and the esophageal cancer group are similar, which are 0.11 and 0.13, 112% and 117%. Also, LBR-MRFDG was the best at telling the difference between anastomotic inflammation and recurrent lesions in patients with both types of upper gastrointestinal tumors who had surgery. In the gastric cancer group, LBR-MRFDG: AUC was 0.935 to 1.83 is the threshold, and its sensitivity is 92.31%; the specificity was 81.25%; in the esophageal cancer group, LBR-MRFDG: AUC was 0.927 to 1.86 is the threshold, and its sensitivity is 90.91%; the degree of differentiation is 76.92%. The ROC curves of dynamic PET parameters in the gastric cancer group and the esophageal cancer group are shown in Figure 1.

Figure 1 Receiver operating characteristic curve of dynamic positron emission tomography parameters in gastric cancer and esophageal cancer. A: Receiver operating characteristic curve of dynamic positron emission tomography parameters in gastric cancer; B: Receiver operating characteristic curve of dynamic positron emission tomography parameters in esophageal cancer. MR: Metabolic rate; LBR: Lesion/background ratio; ROC: Receiver operating characteristic; DV: Volume distribution.

Table 3 Dynamic positron emission tomography/computed tomography parameters and diagnostic efficacy of lesion/background ratio.

Parameters	Sensitivity (%)	Specificity (%)	Threshold value	Youden index	AUC (95%CI)
Gastric cancer
MRFDG	84.62	81.25	0.11	0.659	0.904 (0.736-0.981)
DVmax	76.92	75	112%	0.519	0.808 (0.619-0.929)
LBR-MRFDG	92.31	81.25	1.83	0.736	0.935 (0.778-0.993)
LBR-DVmax	84.62	75	1.77	0.596	0.837 (0.653-0.947)
Esophageal cancer
MRFDG	81.82	84.62	0.13	0.664	0.881 (0.684-0.976)
DVmax	90.91	76.92	117%	0.678	0.853 (0.650-0.963)
LBR-MRFDG	90.91	76.92	1.86	0.678	0.927 (0.743-0.993)
LBR-DVmax	81.82	84.62	1.38	0.664	0.888 (0.693-0.979)

MRFDG: Fluorodeoxyglucose metabolic rate; LBR: Lesion/background ratio; ROC: Receiver operating characteristic; DV: Volume distribution; AUC: Area under the curve; CI: Confidence interval.

PET/CT image analysis of anastomotic recurrence or inflammation

In this study cohort, the whole-body dynamic ¹⁸F-FDG PET/CT images of recurrent anastomotic lesions often showed anastomotic wall thickening, most of which had unclear boundaries with surrounding tissues. FDG metabolism showed high uptake on MR images, while free FDG uptake was low on DV images, as shown in Figure 2. In the whole-body dynamic ¹⁸F-FDG PET/CT images of anastomotic inflammatory lesions, most anastomotic walls did not significantly thicken. FDG metabolism showed a slightly higher uptake on MR images, and free FDG showed a high uptake on DV images, as shown in Figure 3.

Figure 2 A patient with gastric signet-ring cell carcinoma underwent ¹⁸F-fluorodeoxyglucose-positron emission tomography/computed tomography for initial staging. A: Computed tomography; B: ¹⁸F-fluorodeoxyglucose-FAPI-positron emission tomography/computed tomography; C: ⁶⁸Ga-positron emission tomography/computed tomography. FDG: Fluorodeoxyglucose. Arrow: Tumor location.

Figure 3 Subcutaneous and bone metastases of gastric cancer were identified by ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography. FDG: Fluorodeoxyglucose. Arrow: Tumor location.

DISCUSSION

After operation for upper digestive tract tumors, the anastomosis is prone to inflammation due to the large residual stomach, long-term stimulation of food and digestive fluid, and infection by pathogenic bacteria[18]. In conventional static ¹⁸F-FDG PET/CT imaging, anastomotic inflammation is often characterized by high FDG uptake, which is difficult to distinguish from recurrent anastomotic lesions[19]. However, it is necessary to accurately identify these two types of lesions in clinical practice because it is very important for the postoperative restaging of patients and the choice of follow-up treatment[20]. Some scholars have proposed that calculating the changes in SUVmax value in the early and delayed ¹⁸F-FDG PET/CT imaging of the anastomotic site can help distinguish tumor recurrence and anastomotic inflammation and avoid false-positive PET[21-23]. However, its optimal ∆SUVmax% value has not been deeply studied, and it still has limitations. Some foreign studies have reported that dynamic PET/CT imaging has potential value in differentiating benign and malignant lesions. Therefore, the purpose of this study was to investigate the differential diagnosis of postoperative anastomotic recurrence and inflammation with dynamic ¹⁸F-FDG PET/CT in patients with gastric and esophageal cancer[24].

The results of this study showed that among the two types of patients with upper digestive tract tumors, the mean value of MR in the anastomotic recurrence group was higher than that in the inflammatory lesion group, while the mean value of DVmax was lower than that in the inflammatory lesion group, with statistical significance (P < 0.05). In the study of dynamic ¹⁸F-FDG PET imaging in differentiating malignant tumor and inflammation in subcutaneous and in situ models of NSCLC mice[25]. There was a statistically significant difference in Ki values between tumor and inflammation in both in-situ and subcutaneous lesions. Ki in tumor lesions was significantly higher than that in infectious or inflammatory lesions (P = 0.002)[26]. ¹⁸F-FDG PET/CT could show the difference between benign and malignant pulmonary nodules, they also found that Ki was much higher in malignant nodules than in benign nodules[27]. The reason may be that malignant lesions and inflammatory lesions have different levels of FDG uptake. ¹⁸F-FDG PET/CT parametric imaging can distinguish between phosphorylated FDG involved in metabolism and reversible free FDG in blood and tissue and reconstruct the corresponding MR image and DV image accordingly[28-30]. Therefore, parameter MR and DVmax values differ in malignant and inflammatory lesions. Previous research has shown that the main reason why FDG is taken up so much in both malignant and inflammatory lesions is that they both have higher levels of glucose transporter-1 and glucose transporter-3 expression[31-33]. The level of expression of glucose transporter-1 is much higher in tumors than in inflammation, which makes the FDG uptake higher in both types of lesions. This makes ¹⁸F-FDG PET/CT parametric imaging a great advantage in visual evaluation and quantitative analysis.

In addition, this study found that the AUC, sensitivity, and specificity of the two metabolic parameters, MR and DVmax, were better in the differential diagnosis of anastomotic recurrence and inflammation. In the AUC analysis of benign and malignant lung lesions, values of MRFDG and DVFDG had a good ability to distinguish. At the same time, dynamic FDG-PET imaging used to distinguish metastatic and non-metastatic lymph nodes in lung cancer found that Ki also showed high specificity (91.8%)[34]. At the same time, we found that the threshold values of the two metabolic parameters were similar in the gastric cancer group and the esophageal cancer group, which were 0.11, 0.13, and 112%, 117%. The sensitivity and specificity of the two groups were observed by taking the average of the threshold values, and the parameters MR and DVmax showed better performance. Using 0.12 as the MR threshold, the sensitivity of the gastric cancer group was 69.23%, with a specificity of. 87.5%, a sensitivity of 81.82%, and a specificity of 76.92%. To 114.5% was the DVmax threshold, and the sensitivity of the gastric cancer group was 76.92%, with a specificity of 68.75%, a sensitivity of 72.73%, and a specificity of 76.92%. So, we thought that if the sample size was big enough, dynamic PET metabolic parameters might be able to find the best threshold for telling the difference between upper gastrointestinal tumors that have come back after surgery and those that are inflamed. Few previous studies have evaluated the diagnostic performance of LBR[35]. There was a significant difference in LBR between the recurrent anastomosis group and the inflammation group (P < 0.05). Compared with MR parameters, LBR-MR showed the best diagnostic performance in both gastric and esophageal cancer patients. In their study of whole-body dynamic PET/CT uptake of pancreatic unguis and differential diagnosis of neuroendocrine tumors. Compared with MR alone, LBR has slightly higher diagnostic performance. LBR-Ki is the best, and the AUC is 0.990 with a sensitivity of 95.7%, a specificity of 94.4%, and an accuracy of 95.1%[36]. However, their study did not evaluate the diagnostic performance of LBR-DVmax. In our research results, compared with DVmax alone, the specificity of LBR-DVmax improved to 84.62%. So, when we look at the metabolic parameters of dynamic PET/CT imaging, we can’t forget about the diagnostic performance of LBRs. This could lead to a new way to diagnose diseases and figure out how likely they are to get worse[37].

Major limitations of this study: Small sample size and no group observation of recurrent anastomotic lesions from different pathologic subtypes; it is only used to diagnose recurrent and inflammatory lesions of high metabolism anastomosis. Few of the target lesions analyzed in the cohort lacked histopathological confirmation, but we addressed this by collecting additional clinical data on the lesions.

CONCLUSION

In summary, whole-body dynamic ¹⁸F-FDG PET/CT imaging can accurately differentiate the diagnosis of postoperative anastomotic recurrence and inflammation of upper digestive tract tumors and has the potential to become an effective means of monitoring upper digestive tract tumors after surgical treatment.