Skip to main content

Multiparameter diagnostic model based on 18F-FDG PET metabolic parameters and clinical variables can differentiate nonmetastatic gallbladder cancer and cholecystitis



To evaluate the diagnostic value of a multiparameter model based on 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) metabolic parameters and clinical variables in differentiating nonmetastatic gallbladder cancer (GBC) from cholecystitis.

Patients and methods

In total, 122 patients (88 GBC nonmetastatic patients and 34 cholecystitis patients) with gallbladder space-occupying lesions who underwent 18F-FDG PET/CT were included. All patients received surgery and pathology, and baseline characteristics and clinical data were also collected. The metabolic parameters of 18F-FDG PET, including SUVmax (maximum standard uptake value), SUVmean (mean standard uptake value), SUVpeak (peak standard uptake value), MTV (metabolic tumour volume), TLG (total lesion glycolysis) and SUVR (tumour-to-normal liver standard uptake value ratio), were evaluated. The differential diagnostic efficacy of each independent parameter and multiparameter combination model was evaluated using the receiver operating characteristic (ROC) curve. The improvement in diagnostic efficacy using a combination of the above multiple parameters was evaluated by integrated discriminatory improvement (IDI), net reclassification improvement (NRI) and bootstrap test. Decision curve analysis (DCA) was used to evaluate clinical efficacy.


The ROC curve showed that SUVR had the highest diagnostic ability among the 18F-FDG PET metabolic parameters (area under the curve [AUC] = 0.698; sensitivity = 0.341; specificity = 0.971; positive predictive value [PPV] = 0.968; negative predictive value [NPV] = 0.363). The combined diagnostic model of cholecystolithiasis, fever, CEA > 5 ng/ml and SUVR showed an AUC of 0.899 (sensitivity = 0.909, specificity = 0.735, PPV = 0.899, NPV = 0.758). The diagnostic efficiency of the model was improved significantly compared with SUVR. The clinical efficacy of the model was confirmed by DCA.


The multiparameter diagnostic model composed of 18F-FDG PET metabolic parameters (SUVR) and clinical variables, including patient signs (fever), medical history (cholecystolithiasis) and laboratory examination (CEA > 5 ng/ml), has good diagnostic efficacy in the differential diagnosis of nonmetastatic GBC and cholecystitis.

Peer Review reports


As a relatively rare malignant tumour, the 5-year survival rate of patients with invasive stage III or IV gallbladder cancer (GBC) is estimated to be less than 5% [1, 2]. Resection is the best treatment for patients with clinically localized gallbladder cancer, providing the only chance for cure [3]. The clinical manifestations of GBC are usually nonspecific symptoms, including abdominal pain, anorexia, weight loss, jaundice, pruritus, and scleral icterus [4]. Therefore, sporadic GBC is most often found in patients who receive evaluation of symptoms related to gallstones or surgery [5]. Similarly, there are no highly sensitive or specific tumour markers for GBC diagnosis in laboratory examination, even though carcinoembryonic antigen (CEA) and carbohydrate antigen 19 − 9 (CA19-9) may be elevated and are often used in the management of GBC patients [6, 7]. Cholecystitis is a common benign disease in the biliary system that is usually caused by gallstones (and less often, biliary sludge) obstructing bile egress from the gallbladder [8]. Patients can experience abdominal pain, tachycardia, fever and other symptoms, accompanied by leukocytosis during the acute attack, and some chronic cholecystitis patients may only have slight signs [9]. Since the symptoms and signs of GBC and some cholecystitis patients are similar, and there is no special laboratory examination to distinguish them, imaging examination is extremely important. However, in some patients with GBC and cholecystitis, traditional imaging examinations (such as ultrasound [US], computed tomography [CT] and magnetic resonance imaging [MRI]) are very difficult to differentiate [10, 11], which may lead to changes in treatment strategies, especially for patients with early resectable GBC. Especially when GBC is diagnosed by traditional imaging examinations, it is very important to evaluate the stage to determine the patient's treatment strategies. However, traditional imaging examination has limitations in GBC staging, and 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/CT (PET/CT) could play a role.

118F-FDG PET/CT has proven its value in the management of patients with GBC; in particular, it may be helpful to detect local lymph node metastasis and distant metastasis [12]. In the current guidelines, 18F-FDG PET/CT is recommended to identify lymph node metastasis, distant metastasis and disease recurrence [13]. When the anatomical imaging is equivocal, 18F-FDG PET/CT can also be considered for diagnosis [14]. However, the accumulation of FDG in inflammatory lesions will also increase [15], which may affect the accuracy of 18F-FDG PET/CT in diagnosing patients with gallbladder-occupying lesions, especially for patients without metastasis. Therefore, in the clinical work of nuclear medicine, the characteristics of solitary gallbladder lesions with high FDG uptake are often confused. The purpose of this study was to first determine whether 18F-FDG PET metabolic parameters can differentiate nonmetastatic GBC from cholecystitis; second, we sought to determine whether we could more effectively differentiate nonmetastatic GBC from cholecystitis based on FDG metabolic parameters and clinical variables.

Materials and methods


We collected patients with gallbladder space-occupying lesions who underwent 18F-FDG PET/CT from January 2012 to June 2022. The inclusion criteria were as follows: (1) no history of malignancy or complications with other cancers; (2) nonmetastatic GBC (no evidence of distant metastasis); and (3) GBC and cholecystitis who underwent surgery and were diagnosed by pathology. The exclusion criteria were as follows: (1) low-quality 18F-FDG PET/CT images and (2) no clinical data. The flow chart is shown in Fig. 1. The patient's clinical variables, including medical history, symptoms and signs, and laboratory examination (including WBC [white blood cell], LYM [lymphocyte], NEUT [neutrophil], CEA and CA19-9) were collected through medical records.

Fig. 1
figure 1

Flow chart

Image acquisition

All patients were scanned with 18F-FDG PET/CT (Discovery 710, GE Healthcare, Germany/Biograph 64, Siemens Healthineers, Germany). Before intravenous injection of 18F-FDG (3.5–4.5 MBq/kg), the patient needed to fast for 6 hours, have blood glucose levels < 11.1 mmol/L, and rest in a quiet waiting room for at least 20 minutes. After 60 minutes of injection, images were collected from the skull base to the upper femur in free-breathing mode. The parameters of PET were 3-dimensional mode, 2-2.5 min/bed (30% overlap), 4–5 beds/person, three iterations, 21 subsets, and Gaussian filter half-height width = 4.0 mm. The low-dose CT (LDCT) parameters included voltage = 120–140 kV, current = 100 mAs, rotation = 0.8, layer thickness = 3–5 mm, and pitch = 1 [16]. The images were reconstructed with CT attenuation correction (AC) using the ordered subset expectation maximization algorithm (OSEM).

Image analysis

Two experienced nuclear medical doctors (LC and WGY) both read all cases separately and reached a consensus on the commercial workstation (Advantage workstation 4.6, GE HealthCare) when the patient's clinical data were not clear. Gallbladder regions with abnormal 18F-FDG uptake on PET were defined as lesions. The two-dimensional region of interest (ROI) was manually delineated according to the boundary of tumour lesions on each horizontal axis CT image to form a three-dimensional volume of interest (VOI). There are some essential differences between the two PET/CT systems in machine design and scintillation detection, which may confuse the SUVmax measurement results, at least to some extent [17]. To solve this issue, we retrospectively calculated the SUVmean of the liver parenchyma in 126 patients for whom the original PET/CT images were available (GE Discovery 710, n = 100; Siemens Biograph 64, n = 26) [18]. To measure normal liver parenchyma activity, 3 nonoverlapping spherical 1-cm3-sized VOIs were drawn in the normal liver on axial PET images. There were no significant differences in terms of SUVmean-liver among the 2 PET/CT scanners (GE Discovery 710, 2.36 ± 0.32 vs. Siemens Biograph 64, 2.34 ± 0.34, P = 0.860).

The parameters of PET/CT included SUVmax (maximum standard uptake value), SUVmean (mean standard uptake value), SUVpeak (peak standard uptake value), MTV (metabolic tumour volume), TLG (total lesion glycolysis, SUVmean×MTV), and SUVR (tumour-to-normal liver standard uptake value ratio, SUVmax of the tumour/SUVmean of the normal liver parenchyma). MTV was measured from attenuation-corrected 18F-FDG-PET images by two nuclear medicine physicians (WGY and LC) with 5 and 10 years of experience, respectively, in making these measurements. The 40% threshold of SUVmax in the lesion was used to calculate the MTV [19].

Statistical analysis

The statistical analysis was performed using commercially available software (IBM SPSS Statistics 24, IBM, Armonk, NY; and R software program, version 4.0.2, Bell Laboratories, USA). Quantitative data are described as the means ± SD (standard deviation) for continuous variables, and qualitative data are described as the numbers of cases and percentages [n (%)] for categorical variables. Student’s t test, the Mann‒Whitney test and the chi-squared test were used to compare 18F-FDG PET metabolic parameters, baseline characteristics and clinical variables between nonmetastatic GBC and cholecystitis. The area under the receiver operating characteristic (ROC) curve was calculated to assess the predictive value of 18F-FDG PET metabolic parameters. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated. The data of P < 0.01 for comparison of clinical parameters between nonmetastatic GBC and cholecystitis and 18F-FDG PET metabolic parameters with the highest AUC were selected for inclusion in the diagnostic model. Multivariate logistic regression analysis was used to construct a diagnostic model for distinguishing GBC from cholecystitis. The bootstrap test, integrated discriminatory improvement (IDI) and net reclassification improvement (NRI) were calculated for comparison of the diagnostic model and 18F-FDG PET metabolic parameters with the highest area under the curve (AUC). The bootstrap test was performed with the pROC package, and IDI and NRI were performed with the PredictABEL package. Decision curve analyses (DCA) evaluated the clinical utility and accuracy of the 18F-FDG PET metabolic parameter and model by calculating the net benefits for a range of threshold probabilities in metabolic parameters with the highest AUC [20]. DCA was performed with the rmda package. P < 0.05 was considered statistically significant.


Baseline characteristics and clinical variables

A total of 122 patients were included in the study, including 88 nonmetastatic GBC patients and 34 cholecystitis patients. Table 1 shows the baseline and clinical characteristics between nonmetastatic GBC and cholecystitis patients. Regarding baseline characteristics, a significant difference was observed in age (65.06 ± 8.34 vs. 59.06 ± 12.25, P = 0.019), sex (male:female: 45.5%:54.5% vs. 76.5%:23.5%, P = 0.002), smoking history (13 [14.8%] vs. 16 [47.1%], P = 0.001) and drinking history (7 [8.0%] vs. 8 (23.5%), P = 0.029) between nonmetastatic GBC and cholecystitis patients. According to clinical variables, in the relevant medical history, the proportion of patients with cholecystitis who were diagnosed with gallbladder polyps (3 [3.4%] vs. 5 [14.7%], P = 0.038) and cholecystolithiasis (4 [4.5%] vs. 0 [0%], P = 0.575) was not significantly different compared with patients with nonmetastatic GBC. In patients' clinical symptoms, there were more patients with fever in cholecystitis than in nonmetastatic GBC (1 [1.1%] vs. 7 [20.6%], P = 0.001). There was no significant difference in jaundice (14 [15.9%] vs. 9 [26.5%], P = 0.202), abdominal pain (41 [46.6%] vs. 21 [61.8%], P = 0.160) or abdominal mass (0 [0.0%] vs. 0 [0.0%], P = 0.202) between the two groups. In the laboratory examination, the value of CEA (21.73 ± 69.50 vs. 2.32 ± 1.52, P = 0.001) and the proportion of CEA > 5 ng/ml (32 [36.4%] vs. 2 [5.9%], P = 0.001) in nonmetastatic GBC were higher than those in cholecystitis. The values of WBC (6.96 ± 3.09 vs. 6.25 ± 2.30, P = 0.249), LYM (0.27 ± 0.10 vs. 0.27 ± 0.10, P = 0.924), NEUT (0.62 ± 0.12 vs. 0.63 ± 0.11, P = 0.714), CA19-9 (937.86 ± 2942.96 vs. 1009.54 ± 3562.21, P = 0.910) and the proportion of CA19-9 > 37 U/mL (48 [54.5%] vs. 16 [47.1%], P = 0.545) were not significantly different.

Table 1 Baseline characteristics and clinical variables between nonmetastatic gallbladder cancer and cholecystitis

The pathologic results demonstrated that 78 patients had adenocarcinoma (89%, including 5 patients with malignant transformation of GBC, 1 squamous cell carcinoma (1%), 2 adenosquamous carcinomas (2%) and 7 others (8%, 2 sarcomatoid carcinomas; 3 undifferentiated carcinomas and 2 neuroendocrine carcinomas); 17 had acute cholecystitis (50%); 13 had chronic cholecystitis (38%); and 4 had xanthogranulomatous cholecystitis (12%) (Table 1).

Comparison of 18F-FDG PET metabolic parameters between nonmetastatic GBC and cholecystitis

Compared with patients with cholecystitis, the metabolic parameters of 18F-FDG PET in patients with nonmetastatic GBC, including SUVmax (11.62 ± 6.85 vs. 7.62 ± 3.58, P = 0.002), SUVmean (6.69 ± 4.15 vs. 4.26 ± 2.04, P = 0.002), SUVpeak (9.55 ± 5.90 vs. 5.92 ± 2.66, P = 0.001) and SUVR (5.07 ± 3.04 vs. 3.17 ± 1.55, P = 0.001), were significantly higher, but there were no differences in TLG (205.76 ± 269.23 vs. 101.67 ± 100.24, P = 0.063) or MTV (27.39 ± 29.51 vs. 23.44 ± 19.77, P = 0.828). 18F-FDG PET metabolic parameters between nonmetastatic GBC and cholecystitis patients are summarized in Table 2.

Table 2 Comparison of 18F-FDG PET metabolic parameters between nonmetastatic gallbladder cancer and cholecystitis

The differential diagnostic performance of 18F-FDG PET metabolic parameters and clinical variables in nonmetastatic GBC and cholecystitis

The diagnostic performance is demonstrated in Table 3. The ROC curve showed that SUVR had the highest diagnostic ability among the 18F-FDG PET metabolic parameters; the cut-off was 5.9, and the AUC was 0.698 (95% confidence interval [CI]: 0.599–0.796). The results showed that the sensitivity was 0.341 (95% CI: 0.245–0.451), the specificity was 0.971 (95% CI: 0.829–0.998), the PPV was 0.968 (95% CI: 0.815–0.998), and the NPV was 0.363 (95% CI: 0.266–0.471). For clinical variables, its diagnostic ability is relatively low (Table 3). Meanwhile, according to the difference in clinical characteristics between patients with nonmetastatic GBC and cholecystitis, we constructed a diagnostic model based on multivariate logistic regression analysis, and the parameters included fever, cholecystolithiasis, CEA > 5 ng/ml and SUVR. The model showed that the AUC was 0.899 (95% CI: 0.840–0.958), the sensitivity was 0.909 (95% CI: 0.824–0.957), the specificity was 0.735 (95% CI: 0.553–0.865), and the PPV and NPV were 0.899 (95% CI: 0.812–0.950) and 0.758 (95% CI: 0.574–0.883), respectively.

Table 3 Differential diagnostic efficiency of 18F-FDG PET metabolic parameters and clinical variables and model between nonmetastatic gallbladder cancer and cholecystitis

The model is shown below.

$$y=\frac1{{}_{1+\mathrm e}-\left(0.49\times SUVR-4.10\times Chelecystolithiasis-3.28\times Fever+4.03\times CEA>5\mathrm{ng}/\mathrm{ml}-0.42\right)}$$

The diagnostic efficiencies of the 18F-FDG PET parameters and model are shown in Fig. 2 and Table 3.

Fig. 2
figure 2

The ROC curves of 18F-FDG PET metabolic parameters (A) and model (B). The areas under the ROC curves for the ability to differentiate nonmetastatic gallbladder cancer from cholecystitis were 0.677 for SUVmax, 0.684 for SUVmean, 0.690 for SUVpeak, 0.698 for SUVR, and 0.899 for the model

The addition of cholecystolithiasis, fever and CEA > 5 ng/ml to SUVR allowed a significant reclassification with IDI = 0.369 (95% CI: 0.246–0.491, P < 0.001) and categorical NRI = 0.426 (95% CI: 0.160–0.692, P < 0.001) compared to SUVR alone. According to the bootstrap test, compared with SUVR alone, the combination of cholecystolithiasis, fever, CEA > 5 ng/ml and SUVR had a statistically significant improvement in ROC (D = 0.435; boot: n = 2000; boot: stratified = 1; P < 0.001) (Table 4).

Table 4 Comparison of the SUVR and model to with Bootstrap test, IDI and NRI

Clinical application

The DCA for the SUVR and the model are presented in Fig. 3. DCA showed that the model had a higher overall net benefit than SUVR across the entire range of risk thresholds.

Fig. 3
figure 3

Decision curve analyses (DCA) showed that, regardless of the threshold probability of a doctor or a patient, using the combined model (cholecystolithiasis plus fever plus CEA > 5 ng/ml plus SUVR) in our study to differentiate nonmetastatic gallbladder cancer from cholecystitis was more valuable than using SUVR alone. The x-axis represents the threshold probability, and the y-axis represents the net benefit


Our research found that 18F-FDG PET metabolic parameters have certain value in differentiating nonmetastatic GBC and cholecystitis. Furthermore, diagnostic models based on 18F-FDG PET metabolic parameters (SUVR) and clinical variables, including patient signs (fever), medical history (cholecystolithiasis) and laboratory examination (CEA > 5 ng/ml), can effectively help differentiate nonmetastatic GBC and cholecystitis.

Surgical resection is the only possible cure for GBC [21]. Patients with T1a stage tumours can be treated by simple cholecystectomy [22]. After T1a stage, GBC patients who are likely to have tumours removed need radical surgery, including extended cholecystectomy, resection of adjacent liver parenchyma to ensure negative surgical margins, removal of any organs that may be involved in the tumour, and removal of periportal and hepatoduodenal ligaments [21]. Importantly, if only cholecystectomy is carried out for GBC, many patients will have residual disease, and R0 resection is an important prognostic factor [23]. In particular, only a few patients will be suspected of gallbladder cancer before surgery [24], and the others will only be diagnosed after cholecystectomy on histopathological study of the gallbladder specimen [25]. The likelihood of finding residual disease in patients re-explored after incidental discovery of GBC after cholecystectomy has been reported as 38%, 57%, and 77% in patients with T1b, T2, and T3 tumours, respectively [26]. For patients with cholecystitis, the treatment mainly depends on the type of cholecystitis. The main treatment methods for cholecystitis include supportive care and surgical treatment, while surgical treatment is mainly applied to laparoscopic cholecystectomy [21]. Expectant management is the treatment of choice for patients with asymptomatic cholelithiasis [5]. Therefore, judging GBC and cholecystitis before surgery can help patients choose the right treatment. However, the GBC patient's performance is nonspecific, and some of the patient's signs and laboratory examination may coincide with those of cholecystitis patients. The conventional imaging findings of cholecystitis may overlap with those of gallbladder carcinoma [11].

18F-FDG PET/CT has been proven to be a very accurate noninvasive tool for evaluating primary tumours in GBC patients [27] and can be used for staging patients with GBC before treatment [12] and evaluating residual lesions after surgery [28]. Therefore, 18F-FDG PET/CT is a more accurate noninvasive examination for patients suspected of GBC and metastasis in clinical practice and patients suspected of recurrence after treatment. When 18F-FDG PET/CT shows obvious distant metastasis, the diagnosis of gallbladder cancer will be clearer. However, sometimes FDG uptake occurs only in gallbladder lesions. Because benign gallbladder lesions, such as cholecystitis [29], adenomyomatosis [30] and gallbladder polyps [31], also have FDG uptake, when this occurs, the diagnosis is often controversial. The uptake of adenomyomatosis and gallbladder polyps is often lower than that of the liver parenchyma; thus, it is easier to differentiate from GBC [31, 32]. The accumulation of FDG in inflammatory tissues may lead to a false-positive diagnosis of malignant tumours [33]; therefore, it is challenging to judge the characteristics of gallbladder space-occupying lesions with positive FDG uptake, especially for independent gallbladder lesions. Previous studies have disputed the ability of 18F-FDG PET in the differential diagnosis of benign and malignant gallbladder lesions [34,35,36,37]. Many factors can affect the FDG uptake of gallbladder lesions, such as small lesions, which may lead to false-negatives [34]. Therefore, the patients included in our study were all patients with positive FDG uptake and no distant metastasis found on 18F-FDG PET/CT, and all of them received surgical pathology of gallbladder lesions. Our research results showed that, compared with patients with cholecystitis, patients with nonmetastatic GBC have higher FDG uptake (Fig. 4), and SUVR has the best differential diagnostic ability among all metabolic parameters of 18F-FDG PET. However, there are still problems in the differential diagnosis of cholecystitis and nonmetastatic GBC using only metabolic parameters. The metabolic parameters show good specificity (range: 0.912–0.971) and are not sensitive enough (range: 0.341–0.398), which may lead to false-negatives.

Fig. 4
figure 4

Images A1 and A2 show a 77-year-old woman with medium-differentiated gallbladder adenocarcinoma (yellow arrow). During the physical examination one month prior, the patient was found to have a space-occupying gallbladder, without jaundice, fever, abdominal pain or abdominal mass. The patient had no history of cholecystolithiasis, gallbladder polyps or cholecystitis. The CEA was 6.59, and the CA19-9 was 9.22. The lesion showed that SUVR was 5.5 (A1-MIP image of 18F-FDG PET/CT; A2-First line: axial PET, CT, and PET and CT fusion images; Second line: coronal PET, CT, and PET and CT fusion images; Third line: sagittal PET, CT, and PET and CT fusion images). Images B1 and B2 show a 39-year-old man with acute cholecystitis with chronic inflammation of the gallbladder mucosa (red arrow). The patient had no obvious cause of abdominal pain with fever for 1 month. The patient had a medical history of cholecystolithiasis. The CEA was 6.59, and the CA19-9 was 9.22. The lesion showed an SUVR of 2.1 (A1-MIP image of 18F-FDG PET/CT; A2-First line: axial PET, CT, and PET and CT fusion images; Second line: coronal PET, CT, and PET and CT fusion images; Third line: sagittal PET, CT, and PET and CT fusion images)

Cholecystolithiasis is the most important inducing factor of cholecystitis. Blocking the bile duct may induce acute cholecystitis, while recurrent cholecystitis may lead to chronic cholecystitis [9, 38]. Gallstones are an important risk factor for GBC [39], but only one in every 200 gallstone patients has GBC (incidence rate is 0.5%) [4]. Cholecystitis is an inflammatory disease, and there is a possibility of fever in patients. GBC often presents with vague abdominal complaints and systemic signs of anorexia and weight loss [38]. Meanwhile, although the application of serum CEA in the diagnosis of gallbladder cancer is relatively limited, compared with non-GBC, the CEA in gallbladder cancer patients will be higher, with a higher proportion higher than 5 ng/ml [40]. In our study, due to the differences in the signs, medical history and laboratory examination of patients with GBC and cholecystitis, as well as the diagnostic ability of metabolic parameters of 18F-FDG PET, we combined clinical variables, including cholecystolithiasis, fever, CEA > 5 ng/ml, and SUVR, to form a differential diagnostic model. Our research results showed that the model has good sensitivity (0.909) and specificity (0.735) in differential diagnosis and has better diagnostic ability than the single application of 18F-FDG PET metabolic parameters (SUVR). DCA suggested that the differential diagnosis of nonmetastatic GBC and cholecystitis using the model in our study was more valuable than using SUVR alone, regardless of the physician's or patient's threshold probability. To ensure the reliability of the results, all patients were confirmed by pathology. When a patient is diagnosed with cholecystitis, surgery is not the first choice [38], but if there is an incorrect diagnosis, it is likely to delay treatment. Therefore, our diagnosis model may avoid the above question. This model can help nuclear medicine doctors have a more accurate diagnosis when they encounter independent gallbladder lesions with high FDG uptake and, to a certain extent, help patients choose treatment.

This study had some limitations. First, to our knowledge, this is the first study to apply 18F-FDG PET metabolic parameters to the differential diagnosis of nonmetastatic GBC and cholecystitis, and more cases were included compared with other articles that used 18F-FDG PET metabolic parameters to differentiate benign and malignant gallbladder. This was a retrospective cohort study, and compared with patients with nonmetastatic GBC, the number of patients with cholecystitis was significantly lower. This may be because some patients did not undergo surgery after clinical diagnosis of cholecystitis and could not obtain pathological results. These factors may lead to statistical bias of the diagnostic model. Second, although no patients with nonmetastatic GBC received tumour-related treatment, some patients with cholecystitis and nonmetastatic GBC received antibiotics after cholecystitis-related symptoms. This may have led to an inaccurate comparison of inflammatory indicators (such as WBC, LYM, and NEUT) in the laboratory examination. However, antibiotic treatment is necessary in some patients with cholecystitis, such as acute cholecystitis [41]. We did not include inflammatory indicators in the diagnostic model, which helps avoid inaccuracies. Third, SUV is influenced by many factors [42], which may lead to nonreproducibility of the model constructed with metabolic parameters. Because 18F-FDG PET/CT of patients comes from different machines, the measurement of metabolic parameters may be different due to machine parameters and 18F-FDG injection dose. Therefore, we corrected this issue through SUVR to minimize the result bias. Fourth, there were no FDG-negative patients among our cohort. Because these patients were not clinically diagnosed with nonmetastatic GBC, no surgery was performed to obtain pathology. Fifth, some important parameters (such as T stage) were not included in our study due to insufficient medical records. In the future, we will conduct prospective research with a larger sample size and incorporate more variables to improve the diagnostic ability, stability and repeatability of the model.


In general, 18F-FDG PET metabolic parameters are still defective in differentiating nonmetastatic GBC from cholecystitis, but the multiparameter diagnostic model composed of 18F-FDG PET metabolic parameters (SUVR) and clinical variables, including patient signs (fever), medical history (cholecystolithiasis) and laboratory examination (CEA > 5 ng/ml), has good diagnostic efficacy in the differential diagnosis of nonmetastatic GBC and cholecystitis. Our results may help judge the characteristics of lesions when solitary high uptake gallbladder lesions are found on 18F-FDG PET and provide a more accurate and reliable evaluation for the differential diagnosis of preoperative gallbladder disease, which can avoid unnecessary treatment and surgery.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.



18F-fluorodeoxyglucose positron emission tomography


Attenuation correction


Carcinoembryonic antigen


Decision curve analysis


Gallbladder cancer


Integrated discriminatory improvement




Metabolic tumor volume




Negative predictive value


Net reclassification improvement


Ordered subset expectation maximization algorithm


Positive predictive value


Maximum standard uptake value


Mean standard uptake value


Peak standard uptake value


Tumor-to-normal liver standard uptake value ratio


Total lesion glycolysis


Volume of interest


White blood cell


  1. Hickman L, Contreras C. Gallbladder Cancer: Diagnosis, Surgical Management, and Adjuvant Therapies. Surg Clin North Am. 2019;99(2):337–55.

    Article  Google Scholar 

  2. Lee NK, Kim S, Moon JI, Shin N, Kim DU, Seo HI, Kim HS, Han GJ, Kim JY, Lee JW. Diffusion-weighted magnetic resonance imaging of gallbladder adenocarcinoma: analysis with emphasis on histologic grade. Clin Imaging. 2016;40(3):345–51.

    Article  Google Scholar 

  3. Ramos-Font C, Gomez-Rio M, Rodriguez-Fernandez A, Jimenez-Heffernan A, Sanchez Sanchez R, Llamas-Elvira JM. Ability of FDG-PET/CT in the detection of gallbladder cancer. J Surg Oncol. 2014;109(3):218–24.

    Article  Google Scholar 

  4. Lau CSM, Zywot A, Mahendraraj K, Chamberlain RS. Gallbladder Carcinoma in the United States: A Population Based Clinical Outcomes Study Involving 22,343 Patients from the Surveillance, Epidemiology, and End Result Database (1973–2013). HPB Surg. 2017;2017:1532835.

    Article  Google Scholar 

  5. Gunasekaran G, Bekki Y, Lourdusamy V, Schwartz M. Surgical Treatments of Hepatobiliary Cancers. Hepatology. 2021;73(Suppl 1):128–36.

    Article  Google Scholar 

  6. Gamboa AC, Maithel SK. The Landmark Series: Gallbladder Cancer. Ann Surg Oncol. 2020;27(8):2846–58.

    Article  Google Scholar 

  7. Wang YF, Feng FL, Zhao XH, Ye ZX, Zeng HP, Li Z, Jiang XQ, Peng ZH. Combined detection tumor markers for diagnosis and prognosis of gallbladder cancer. World J Gastroenterol. 2014;20(14):4085–92.

    Article  Google Scholar 

  8. Alemi F, Seiser N, Ayloo S. Gallstone Disease: Cholecystitis, Mirizzi Syndrome, Bouveret Syndrome, Gallstone Ileus. Surg Clin North Am. 2019;99(2):231–44.

    Article  Google Scholar 

  9. Knab LM, Boller AM, Mahvi DM. Cholecystitis. Surg Clin North Am. 2014;94(2):455–70.

    Article  Google Scholar 

  10. Akkurt G, Birben B, Coban S, Akgul O, Kulacoglu S, Doganay M. Xanthogranulomatous Cholecystitis and Gallbladder Cancer: Two Diseases with Difficult Differential Diagnoses. Turk J Gastroenterol. 2021;32(8):694–701.

    Article  Google Scholar 

  11. Yu MH, Kim YJ, Park HS, Jung SI. Benign gallbladder diseases: Imaging techniques and tips for differentiating with malignant gallbladder diseases. World J Gastroenterol. 2020;26(22):2967–86.

    Article  CAS  Google Scholar 

  12. Goel S, Aggarwal A, Iqbal A, Gupta M, Rao A, Singh S. 18-FDG PET-CT should be included in preoperative staging of gall bladder cancer. Eur J Surg Oncol. 2020;46(9):1711–6.

    Article  Google Scholar 

  13. Vogel A, Bridgewater J, Edeline J, Kelley RK, Klümpen HJ, Malka D, Primrose JN, Rimassa L, Stenzinger A, Valle JW, et al. Biliary tract cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol. 2023;34(2):127-40.

  14. Moradi F, Iagaru A. The Role of Positron Emission Tomography in Pancreatic Cancer and Gallbladder Cancer. Semin Nucl Med. 2020;50(5):434–46.

    Article  Google Scholar 

  15. Shreve PD. Focal fluorine-18 fluorodeoxyglucose accumulation in inflammatory pancreatic disease. Eur J Nucl Med. 1998;25(3):259–64.

    Article  CAS  Google Scholar 

  16. Wang G, Zhang W, Chen J, Luan X, Wang Z, Wang Y, Xu X, Yao S, Guan Z, Tian J, et al. Pretreatment Metabolic Parameters Measured by (18)F-FDG PET to Predict the Pathological Treatment Response of HCC Patients Treated With PD-1 Inhibitors and Lenvatinib as a Conversion Therapy in BCLC Stage C. Front Oncol. 2022;12:884372.

    Article  Google Scholar 

  17. Armstrong IS, Thomson KE, Rowley LM, McGowan DR. Harmonizing standardized uptake value recovery between two PET/CT systems from different manufacturers when using resolution modelling and time-of-flight. Nucl Med Commun. 2017;38(7):650–5.

    Article  Google Scholar 

  18. Hsieh CE, Cheng NM, Chou WC, Venkatesulu BP, Chou YC, Liao CT, Yen TC, Lin CY. Pretreatment Primary Tumor and Nodal SUVmax Values on 18F-FDG PET/CT Images Predict Prognosis in Patients With Salivary Gland Carcinoma. Clin Nucl Med. 2018;43(12):869–79.

    Article  Google Scholar 

  19. Arslan E, Aksoy T, Dursun N, Gursu RU, Sevinc MM, Cermik TF. The Role of 18F-FDG PET/CT in staging of gallbladder carcinomas. Turk J Gastroenterol. 2020;31(2):105–12.

    Article  Google Scholar 

  20. Zhang Z, Rousson V, Lee WC, Ferdynus C, Chen M, Qian X, Guo Y, written on behalf of AMEB-DCTCG. Decision curve analysis: a technical note. Ann Transl Med. 2018;6(15):308.

    Article  Google Scholar 

  21. Ganeshan D, Kambadakone A, Nikolaidis P, Subbiah V, Subbiah IM, Devine C. Current update on gallbladder carcinoma. Abdom Radiol (NY). 2021;46(6):2474–89.

    Article  Google Scholar 

  22. Fuks D, Regimbeau JM, Le Treut YP, Bachellier P, Raventos A, Pruvot FR, Chiche L, Farges O. Incidental gallbladder cancer by the AFC-GBC-2009 Study Group. World J Surg. 2011;35(8):1887–97.

    Article  Google Scholar 

  23. Butte JM, Kingham TP, Gonen M, D'Angelica MI, Allen PJ, Fong Y, DeMatteo RP, Jarnagin WR. Residual disease predicts outcomes after definitive resection for incidental gallbladder cancer. J Am Coll Surg. 2014;219(3):416–29.

    Article  Google Scholar 

  24. Varshney S, Butturini G, Gupta R. Incidental carcinoma of the gallbladder. Eur J Surg Oncol. 2002;28(1):4–10.

    Article  CAS  Google Scholar 

  25. Choi KS, Choi SB, Park P, Kim WB, Choi SY. Clinical characteristics of incidental or unsuspected gallbladder cancers diagnosed during or after cholecystectomy: a systematic review and meta-analysis. World J Gastroenterol. 2015;21(4):1315–23.

    Article  Google Scholar 

  26. Pawlik TM, Gleisner AL, Vigano L, Kooby DA, Bauer TW, Frilling A, Adams RB, Staley CA, Trindade EN, Schulick RD, et al. Incidence of finding residual disease for incidental gallbladder carcinoma: implications for re-resection. J Gastrointest Surg. 2007;11(11):1478–86. discussion 1486 – 1477.

    Article  Google Scholar 

  27. Annunziata S, Pizzuto DA, Caldarella C, Galiandro F, Sadeghi R, Treglia G. Diagnostic accuracy of fluorine-18-fluorodeoxyglucose positron emission tomography in gallbladder cancer: A meta-analysis. World J Gastroenterol. 2015;21(40):11481–8.

    Article  CAS  Google Scholar 

  28. Butte JM, Redondo F, Waugh E, Meneses M, Pruzzo R, Parada H, Amaral H, De La Fuente HA. The role of PET-CT in patients with incidental gallbladder cancer. HPB (Oxford). 2009;11(7):585–91.

    Article  Google Scholar 

  29. Gupta V, Vishnu KS, Yadav TD, Sakaray YR, Irrinki S, Mittal BR, Kalra N, Vaiphei K. Radio-pathological Correlation of 18F-FDG PET in Characterizing Gallbladder Wall Thickening. J Gastrointest Cancer. 2019;50(4):901–6.

    Article  CAS  Google Scholar 

  30. Suzuki K, Watada S, Yoko M, Nakahara T, Kumamoto Y. Successful diagnosis of gallbladder carcinoma coexisting with adenomyomatosis by 18F-FDG-PET–report of a case. J Gastrointest Cancer. 2011;42(4):252–6.

    Article  Google Scholar 

  31. Lee J, Yun M, Kim KS, Lee JD, Kim CK. Risk stratification of gallbladder polyps (1–2 cm) for surgical intervention with 18F-FDG PET/CT. J Nucl Med. 2012;53(3):353–8.

    Article  CAS  Google Scholar 

  32. Bonatti M, Vezzali N, Lombardo F, Ferro F, Zamboni G, Tauber M, Bonatti G. Gallbladder adenomyomatosis: imaging findings, tricks and pitfalls. Insights Imaging. 2017;8(2):243–53.

    Article  Google Scholar 

  33. Shreve PD, Anzai Y, Wahl RL. Pitfalls in oncologic diagnosis with FDG PET imaging: physiologic and benign variants. Radiographics. 1999;19(1):61–77. quiz 150–151.

    Article  CAS  Google Scholar 

  34. Koh T, Taniguchi H, Yamaguchi A, Kunishima S, Yamagishi H. Differential diagnosis of gallbladder cancer using positron emission tomography with fluorine-18-labeled fluoro-deoxyglucose (FDG-PET). J Surg Oncol. 2003;84(2):74–81.

    Article  Google Scholar 

  35. Oe A, Kawabe J, Torii K, Kawamura E, Higashiyama S, Kotani J, Hayashi T, Kurooka H, Tsumoto C, Kubo S, et al. Distinguishing benign from malignant gallbladder wall thickening using FDG-PET. Ann Nucl Med. 2006;20(10):699–703.

    Article  Google Scholar 

  36. Bo X, Chen E, Wang J, Nan L, Xin Y, Wang C, Lu Q, Rao S, Pang L, Li M, et al. Diagnostic accuracy of imaging modalities in differentiating xanthogranulomatous cholecystitis from gallbladder cancer. Ann Transl Med. 2019;7(22):627.

    Article  CAS  Google Scholar 

  37. Pericleous S, Doran SLF, Wotherspoon A, Terlizzo M, Riddell A, Brown G, Shur J, Chua S, Hujairi N, Middleton N, et al. The Diagnostic Accuracy of (18) F-FGD-PET/CT for Cancer of the Gallbladder: A Retrospective Study. World J Nucl Med. 2022;21(2):112–9.

    Article  Google Scholar 

  38. Lam R, Zakko A, Petrov JC, Kumar P, Duffy AJ, Muniraj T. Gallbladder Disorders: A Comprehensive Review. Dis Mon. 2021;67(7):101130.

    Article  Google Scholar 

  39. Stinton LM, Shaffer EA. Epidemiology of gallbladder disease: cholelithiasis and cancer. Gut Liver. 2012;6(2):172–87.

    Article  Google Scholar 

  40. Kang JS, Hong SY, Han Y, Sohn HJ, Lee M, Kang YH, Kim HS, Kim H, Kwon W, Jang JY. Limits of serum carcinoembryonic antigen and carbohydrate antigen 19 – 9 as the diagnosis of gallbladder cancer. Ann Surg Treat Res. 2021;101(5):266–73.

    Article  Google Scholar 

  41. Gallaher JR, Charles A. Acute Cholecystitis: A Review. JAMA. 2022;327(10):965–75.

    Article  CAS  Google Scholar 

  42. Laffon E, Cazeau AL, Monet A, de Clermont H, Fernandez P, Marthan R, Ducassou D. The effect of renal failure on 18F-FDG uptake: a theoretic assessment. J Nucl Med Technol. 2008;36(4):200–2.

    Article  Google Scholar 

Download references


Not applicable.


The authors state that this work has not received any funding.

Author information

Authors and Affiliations



Conception and design: C Li, XH Luan, GY Wang and BX Xu; Data collation: C Li, XH Luan, SX Chen, Y Pan, JF Zhang and Y Han; Statistical Analysis: C Li, XH Luan, X Bi and GY Wang; Article writing: C Li, XH Luan, X Bi, SX Chen and XD Xu; Article revision: GY Wang and BX Xu. Final approval of the manuscript: All authors.

Corresponding authors

Correspondence to Guanyun Wang or Baixuan Xu.

Ethics declarations

Ethics approval and consent to participate

This retrospective cohort study was approved by Chinese PLA General Hospital ethical review boards. All patients were informed and signed consent form before 18F-FDG PET/CT. It was performed in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, C., Luan, X., Bi, X. et al. Multiparameter diagnostic model based on 18F-FDG PET metabolic parameters and clinical variables can differentiate nonmetastatic gallbladder cancer and cholecystitis. BMC Cancer 23, 119 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: