The diagnostic value of the multiparameter combination diagnostic model based on 2-[¹⁸F]FDG PET/CT metabolic parameters and clinical variables in distinguishing non-metastatic cholangiocarcinoma from cholangitis

Introduction

Cholangiocarcinoma (CCA) is a malignant tumor originating from the epithelium of the bile duct and peribiliary glands. CCA is the second most prevalent primary liver neoplasm subsequent to hepatocellular carcinoma (HCC) (1). The main anatomical subtypes of CCA include intrahepatic CCA (iCCA; located proximal to the second-order bile duct in the liver parenchyma), perihilar CCA (pCCA; localized between the second-order bile ducts and the insertion of the cystic duct into the common bile duct), and distal CCA (dCCA; located common bile duct below the cystic duct insertion) (2); pCCA and dCCA can also be collectively referred to as ‘extrahepatic CCA’ (eCCA) (3). The only treatment modality with the potential for curative intent in CCA is surgical resection. However, due to its insidious onset, CCA is often diagnosed at an advanced stage, with the feasibility of surgical intervention limited to only approximately one-third of early-stage patients (4). Therefore, accurate diagnosis of early-stage CCA is crucial.

Medical imaging techniques such as ultrasonography (US), computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET)/CT serve as non-invasive diagnostic tools that aid in the identification of CCA, facilitating improved treatment selection and prognosis (5). The 2-[¹⁸F]fluoro-D-glucose {2-[¹⁸F]FDG} PET/CT, capable of evaluating metabolic activity in malignant diseases, has demonstrated superior accuracy compared to CT and MRI in various primary and metastatic tumors, including bile duct cancers (6). To date, few studies have demonstrated that 2-[¹⁸F]FDG PET/CT has clinical value in the staging, evaluation of treatment response, and prediction of prognosis in different anatomical subtypes of CCA (7-12). Nevertheless, inflammatory lesions can also exhibit FDG avidity, potentially resulting in false-positive findings on 2-[¹⁸F]FDG PET/CT scans in cases of cholangitis. The overlap of clinical symptoms and certain laboratory and imaging results complicates the differentiation between benign and malignant lesions. Consequently, this may lead to misdiagnosis of cholangitis as malignant tumors, particularly in patients with localized lesions, resulting in unnecessary treatments. Therefore, the aims of our study were as follows: firstly, to compare the 2-[¹⁸F]FDG PET/CT metabolic parameters of different anatomical subtypes of non-metastatic CCA (including iCCA and eCCA) with those of the corresponding pathological location cholangitis. Secondly, to integrate 2-[¹⁸F]FDG PET/CT metabolic parameters with clinical variables to establish diagnostic models and assess its diagnostic value in distinguishing different anatomical subtypes of non-metastatic CCA from corresponding pathological location cholangitis. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1626/rc).

Methods

Patients

This retrospective cohort study was approved by the Ethics Committee of the People’s Liberation Army General Hospital (No. S2014-052-01). All patients were informed and signed a consent form before undergoing 2-[¹⁸F]FDG PET/CT examination. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Patients with bile duct space-occupying lesions who underwent 2-[¹⁸F]FDG PET/CT examination between January 2013 and August 2023 were retrospectively recruited. The inclusion criteria were as follows: (I) 2-[¹⁸F]FDG PET/CT examination was conducted prior to any treatment; (II) complete images and clinical data were available for all patients; (III) the pathological diagnosis of all patients was confirmed through surgical resection; and (IV) the absence of evident distant metastasis was observed in all patients with non-metastatic CCA (non-metastatic CCA). The exclusion criteria were as follows: (I) the patient was complicated with other malignant tumors; and (II) the PET/CT images had inadequate quality for evaluation. The study recruitment flow chart is shown in Figure 1. After excluding patients with lymphatic or distant metastasis indicated by 2-[¹⁸F]FDG PET/CT (n=474), other types of malignant tumors (n=54), those without pathological results (n=198), and those lacking complete clinical data (n=9), a total of 181 patients were included in this retrospective study (44 patients with cholangitis, 137 patients with non-metastatic CCA). Clinical variables, including medical history, symptoms (fever, abdominal pain, jaundice), laboratory examination [including white blood cell (WBC), lymphocyte (LYM), neutrophil (NEUT), monocyte (MON), carcinoembryonic antigen (CEA), and carbohydrate antigen 19-9 (CA19-9)], and lesion size were collected by review of the medical record system. The precise anatomical location of the patient’s lesion was determined through pathological examination.

Figure 1 The flow chart. CT, computed tomography; eCCA, extrahepatic cholangiocarcinoma; FDG, fluoro-D-glucose; iCCA, intrahepatic cholangiocarcinoma; PET, positron emission tomography.

Image acquisition

All patients underwent 2-[¹⁸F]FDG PET/CT scanning. The scanners used were Discovery 710 (GE Healthcare, Solingen, Germany), Biograph 64 (Siemens Healthineers, Erlangen, Germany), and uMI 510 (United Imaging Healthcare, Shanghai, China). Patients were prohibited from strenuous exercise for 24 hours and fasted for at least 6 hours before the examination to ensure that their blood glucose levels remained below 11.1 mmol/L. A free-breathing scan from the skull base to the upper femur was performed 60 minutes after the administration of intravenous 2-[¹⁸F]FDG (3.5–4.5 MBq/kg, the radiochemical purity was >95%). The CT acquisition protocol was configured with the following settings: tube voltage ranging from 120 to 140 kV, tube current between 100 and 200 mAs, gantry rotation time of 0.8 seconds, slice thickness of 3–5 mm, and a pitch factor of 0.9–1. For PET imaging, standard acquisition parameters were applied, including three-dimensional (3D) acquisition mode with 2–2.5 minutes per bed position (30% overlap), 4–5 bed positions per patient, reconstruction using 3 iterations and 21 subsets, and post-reconstruction smoothing with a 4.0 mm full-width at half-maximum Gaussian filter. The images were reconstructed with CT attenuation correction (AC) using the ordered subset expectation maximization (OSEM) algorithm.

Image analysis

The images were centrally analyzed by two experienced nuclear medicine physicians (J.Z. and C.L.), who were blinded to the clinical and pathological information. In cases where there was controversy surrounding the diagnosis, input was sought from a third nuclear medicine physician (G.W.) for further discussion. Lesions were defined as liver or bile duct with abnormal 2-[¹⁸F]FDG uptake on PET or abnormal density on CT. The two-dimensional (2D) region of interest (ROI) was manually delineated according to the boundary of the lesion on each horizontal axis CT image to form the 3D volume of interest (VOI). The three PET/CT systems exhibit fundamental variations in both scanner architecture and scintillator detection methodology, potentially introducing systematic variations in maximum standardized uptake value (SUVmax) quantification. To address this issue, liver parenchyma mean standardized uptake value (SUVmean) was retrospectively calculated from available original PET/CT images of 181 patients (GE Discovery 710, n=86; Siemens Biograph 64, n=57; United Imaging uMI 510, n=38). To measure activity in the normal liver parenchyma, three nonoverlapping spherical 1-cm³-sized VOIs were drawn in the normal liver parenchyma on axial PET images. There was no significant difference in liver parenchyma SUVmean among the three PET/CT scanners (GE Discovery VCT, 2.62±0.40 vs. Siemens Biograph 64, 2.51±0.36 vs. United Imaging uMI 510 2.55±0.41, respectively; F=1.548, P=0.215, variance analysis) (13).

The 2-[¹⁸F]FDG PET/CT metabolic parameters include SUVmax, SUVmean, peak standardized uptake value (SUVpeak), metabolic tumor volume (MTV), total lesion glycolysis (TLG, SUVmean × MTV), and tumor-to-normal liver standardized uptake value ratio (SUVR), SUVmax of the tumor/SUVmean of the normal liver parenchyma). Two nuclear medicine physicians separately performed semi-automatic segmentation of the 3D VOIs of the lesions using a post-processing workstation (Advantage Workstation 4.6, GE HealthCare), and calculated MTV and TLG with a threshold of 40% of SUVmax.

Statistical analysis

The statistical software SPSS 24.0 (IBM Corp., Armonk, NY, USA) and R4.0.2 (R software program, version 4.0.2; Bell Laboratories, Windsor, WI, USA) were used for statistical analysis. The significance level of all statistical tests was set at P=0.05. Clinical variables and 2-[¹⁸F]FDG PET/CT metabolic parameters were statistically described. Quantitative data were presented as mean ± standard deviation (SD) or median and range for continuous variables and compared with the use of Student’s t-test or the Mann-Whitney U test. However, qualitative data were described as the numbers of cases and percentage [n (%)] of categorical variables and compared by Chi-squared test. To assess the predictive performance of 2-[¹⁸F]FDG PET/CT metabolic parameters, the area under the receiver operating characteristic (ROC) curve (AUC) was computed. Additionally, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were determined. The clinical variables with P<0.01 were screened by univariate logistic regression analysis. Then, these variables and the metabolic parameters with the largest AUC were included in the multivariate logistic regression analysis to construct the diagnostic models for CCA and cholangitis. To compare the performance of diagnostic models and metabolic parameters with the highest AUC, the integrated discrimination improvement (IDI) and net reclassification improvement (NRI) were computed using the PredictABEL package. Additionally, decision curve analysis (DCA) was conducted with the rmda package to evaluate the clinical utility and accuracy of the diagnostic model. Net benefit was calculated across a range of threshold probabilities for both the top-performing metabolic parameter (based on AUC) and the final diagnostic model.

Results

Baseline characteristics and clinical variables

The baseline characteristics and clinical variables of all patients are comprehensively outlined in Table 1. The baseline characteristics of patients with non-metastatic iCCA and intrahepatic cholangitis, as well as non-metastatic eCCA and extrahepatic cholangitis, respectively, did not exhibit any significant disparities.

In the differential diagnosis of intrahepatic lesions, there was a statistically significant difference in the proportion of history of cholangiolithiasis between non-metastatic iCCA and intrahepatic cholangitis [4 (8.3%) vs. 7 (46.7%), P=0.002], but there was no statistically significant difference in clinical presentation. In the laboratory examination, the value of CEA [3.67 (1.85–11.20) vs. 1.85 (1.15–3.27), P=0.006], the proportion of CEA >5 ng/mL [20 (41.7%) vs. 1 (6.7%), P=0.012], CA19-9 [201.40 (32.34–1,911.00) vs. 14.88 (5.44–57.55), P<0.001], and the proportion of CA19-9>37 U/mL [35 (72.9%) vs. 4 (26.7%), P=0.001] in patients with iCCA were much higher than those in intrahepatic cholangitis. In addition, the values of WBC (6.82±2.11 vs. 5.57±1.83, P=0.044), LYM (0.26±0.08 vs. 0.30±0.08, P=0.015), NEUT (0.66±0.08 vs. 0.59±0.08, P=0.005), and lesion size (37.65±22.79 vs. 52.25±24.68, P=0.046) were also statistically different between patients with and non-metastatic iCCA and intrahepatic cholangitis.

In the differential diagnosis of extrahepatic lesions, there was a statistically significant difference in the proportion of history of cholangiolithiasis [0 (0.0%) vs. 10 (34.5%), P<0.001] and the proportion of fever [9 (10.1%) vs. 9 (31.0%), P=0.016] between non-metastatic eCCA and extrahepatic cholangitis. In the laboratory examination, the value of CEA [3.30 (2.03–4.82) vs. 2.06 (1.34–3.23), P=0.002], CA19-9 [170.6 (34.68–459.00) vs. 24.90 (15.21–118.04), P<0.001] and the proportion of CA19-9 >37 U/mL [63 (70.8%) vs. 12 (41.4%), P=0.002] in patients with eCCA were much higher than they were in those with extrahepatic cholangitis. In addition, the value of LYM (0.26±0.08 vs. 0.30±0.08, P=0.015) was also statistically different between patients with non-metastatic eCCA and extrahepatic cholangitis.

Comparison of 2-[¹⁸F]FDG PET/CT metabolic parameters between non-metastatic CCA and cholangitis

Table 2 presents a comparison of 2-[¹⁸F]FDG PET/CT metabolic parameters between patients with non-metastatic iCCA and intrahepatic cholangitis, as well as non-metastatic eCCA and extrahepatic cholangitis.

In the differential diagnosis of non-metastatic iCCA and intrahepatic cholangitis, significant differences were observed between the two groups for SUVmax [8.42 (5.76–11.66) vs. 6.15 (4.57–8.02), P=0.027], SUVmean [5.03 (3.49–6.59) vs. 3.32 (2.57–4.24), P=0.015], SUVpeak [7.43 (4.87–10.03) vs. 5.06 (3.20–6.25), P=0.005], and TLG [205.98 (86.71–531.07) vs. 120.61 (63.51–165.53), P=0.021]. MTV [43.82 (19.06–81.64) vs. 24.71 (16.99–56.63), P=0.151] and SUVR [3.44 (2.61–4.86) vs. 2.93 (1.90–3.42), P=0.066] were not significantly different between non-metastatic iCCA and intrahepatic cholangitis.

In the differential diagnosis of non-metastatic eCCA and extrahepatic cholangitis, significant differences were observed between the two groups for SUVmax [4.31 (3.12–6.63) vs. 3.27 (2.62–4.26), P=0.022], SUVmean [2.42 (2.00–3.87) vs. 2.01 (1.82–2.50), P=0.018], SUVpeak [3.61 (2.61–5.21) vs. 2.61 (2.03–3.45), P=0.012], and MTV [14.91 (9.18–19.36) vs. 21.89 (12.18–38.58), P=0.007]. TLG [34.80 (22.13–56.04) vs. 43.56 (22.38–90.01), P=0.107] and SUVR [1.57 (1.16–2.57) vs. 1.39 (1.09–1.84), P=0.354] were not significantly different between non-metastatic eCCA and extrahepatic cholangitis.

The differential diagnostic efficiency of 2-[¹⁸F]FDG PET/CT metabolic parameters and models in non-metastatic CCA and cholangitis

Table 3 displays the differential diagnostic efficiency of 2-[¹⁸F]FDG PET/CT metabolic parameters and models. The ROC curve analysis demonstrated that SUVpeak [cut-off value =6.5, AUC =0.740, 95% confidence interval (CI): 0.603–0.878] had the highest differential diagnostic ability in non-metastatic iCCA and intrahepatic cholangitis among 2-[¹⁸F]FDG PET/CT metabolic parameters. The SUVpeak had sensitivity of 0.625 (95% CI: 0.473–0.757), specificity of 0.867 (95% CI: 0.584–0.977), PPV of 0.938 (95% CI: 0.778–0.989), and NPV of 0.419 (95% CI: 0.270–0.504). Meanwhile, according to the differences in clinical variables between patients with non-metastatic iCCA and intrahepatic cholangitis, we constructed a diagnostic model based on multivariate logistic regression analysis, including cholangiolithiasis, CA19-9 >37 ng/mL, and SUVpeak. Compared with the 2-[¹⁸F]FDG PET/CT metabolic parameter SUVpeak alone, the diagnostic performance of model was significantly improved with an AUC of 0.853 (95% CI: 0.744–0.962) (Figure 2A). The model 1 displayed the sensitivity of 0.833 (95% CI: 0.692–0.920), specificity of 0.800 (95% CI: 0.514–0.947), PPV of 0.930 (95% CI: 0.799–0.982), and NPV of 0.600 (95% CI: 0.364–0.800). Overall, the model with combined cholangiolithiasis, CA19-9 >37 ng/mL, and SUVpeak significantly improved reclassification compared to SUVpeak alone, as demonstrated by the IDI and categorical NRI in Table 4 [IDI =0.238 (95% CI: 0.109–0.367, P<0.001), categorical NRI =0.292 (95% CI: 0.027–0.557, P=0.031)].

Figure 2 The ROC curves of 2-[¹⁸F]FDG PET/CT metabolic parameters and the models between non-metastatic iCCA and intrahepatic cholangitis (A), and non-metastatic eCCA and extrahepatic cholangitis (B), respectively. The AUC for the ability to differentiate non-metastatic iCCA and from intrahepatic cholangitis for SUVpeak was 0.740, and that for the model (cholangiolithiasis plus CA19-9 >37 ng/mL plus SUVpeak) was 0.853. The AUC for the ability to differentiate non-metastatic eCCA and from extrahepatic cholangitis for MTV was 0.669, and model (cholangiolithiasis plus fever plus CA19-9 >37 ng/mL plus MTV) was 0.863. AUC, area under the ROC curve; CA19-9, carbohydrate antigen 19-9; CT, computed tomography; eCCA, extrahepatic cholangiocarcinoma; FDG, fluoro-D-glucose; MTV, metabolic tumor volume; iCCA, intrahepatic cholangiocarcinoma; PET, positron emission tomography; ROC, receiver operating characteristic; SUVpeak, peak standardized uptake value.

The model is presented as follows:

y=11+e−(−2.33×cholangiolithiasis+1.95×CA19−9>37ng/mL+0.27×SUVpeak−0.97)[1]

The ROC curve analysis demonstrated that MTV (cut-off value =18.1, AUC =0.669, 95% CI: 0.551–0.787) had the highest differential diagnostic ability in non-metastatic eCCA and extrahepatic cholangitis among 2-[¹⁸F]FDG PET/CT metabolic parameters. The MTV had sensitivity of 0.655 (95% CI: 0.457–0.814), specificity of 0.663 (95% CI: 0.552–0.759), PPV of 0.396 (95% CI: 0.261–0.547), and NPV of 0.851 (95% CI: 0.738–0.922). Meanwhile, according to the differences in clinical variables between patients with non-metastatic eCCA and extrahepatic cholangitis, we constructed a diagnostic model based on multivariate logistic regression analysis, including cholangiolithiasis, CA19-9 >37 ng/mL, fever, and MTV. Compared with the 2-[¹⁸F]FDG PET/CT metabolic parameter MTV alone, the diagnostic performance of the model was significantly improved with an AUC of 0.863 (95% CI: 0.785–0.941) (Figure 2B). The model displayed a sensitivity of 0.884 (95% CI: 0.792–0.940), specificity of 0.724 (95% CI: 0.525–0.866), PPV of 0.905 (95% CI: 0.816–0.955), and NPV of 0.677 (95% CI: 0.485–0.827). Overall, the model with combined cholangiolithiasis, fever, CA19-9 >37 ng/mL, and MTV significantly improved reclassification compared to MTV alone, as demonstrated by the IDI and categorical NRI in Table 4 [IDI =0.302 (95% CI: 0.171–0.434, P<0.001), categorical NRI =0.344 (95% CI: 0.164–0.525, P<0.001)].

The model is presented as follows:

y=11+e−(−22.68×cholangiolithiasis−1.57×fever+1.51×CA19−9>37ng/mL−0.04×MTV−1.76)[2]

Discussion

Our study demonstrated that 2-[¹⁸F]FDG PET/CT metabolic parameters can help to distinguish between non-metastatic iCCA and intrahepatic cholangitis, and non-metastatic eCCA, and extrahepatic cholangitis, respectively. Meanwhile, the diagnostic model {including 2-[¹⁸F]FDG PET/CT metabolic parameters (SUVpeak), clinical presentation (fever), medical history (cholangiolithiasis), and laboratory examination (CA19-9 >37 ng/mL)} can effectively help differentiate non-metastatic iCCA and cholangitis, and the diagnostic model {including 2-[¹⁸F]FDG PET/CT metabolic parameters (MTV), clinical presentation (fever), medical history (cholangitis), and laboratory examination (CA19-9 >37 ng/mL)} can effectively help differentiate non-metastatic eCCA and extrahepatic cholangitis.

CCA accounts for approximately 2–3% of all malignant tumors in the gastrointestinal system. Among them, pCCA represents about 50–60% of all CCA cases, eCCA represents 20–30%, and iCCA is the second most prevalent primary liver cancer after HCC, comprising 10% of all primary liver cancers (14,15). Surgery is a potential curative option for CCA, but only a few patients (~25%) with early-stage CCA are able to undergo surgery (16). Pathological examination is an indispensable prerequisite for the diagnosis of CCA and can offer invaluable insights for the clinical management of patients (17), but some patients are recommended not to undergo pathological examination due to concerns regarding tumor size, location, and the risk of peritoneal seeding (18). Therefore, imaging examinations, laboratory examination, and patient clinical presentation will play a significant role in the diagnosis of CCA patients. However, when a patient is found to have a bile duct space-occupying lesion, the diagnostic challenge of distinguishing between CCA and cholangitis is exacerbated by the difficulty in obtaining pathological results and the potential overlap of clinical presentation, imaging findings, and laboratory examination results. Importantly, the treatment of cholangitis often involves medication or drainage procedures [including endoscopic retrograde cholangiopancreatography (ERCP), percutaneous transitional cholangiography (PTC), endoscopic ultrasound-guided biliary drainage (EUS-BD)], whereas surgery is the last treatment option for patients with cholangitis (19). Therefore, the accurate differentiation of early-stage CCA from cholangitis through non-invasive diagnostic methods is crucial in order to prevent patients from undergoing unnecessary treatment.

The utility of 2-[¹⁸F]FDG PET/CT in CCA at different anatomical locations is variable. For example, iCCA has typical high metabolic characteristics, which can be manifested as large circular masses, whereas eCCA usually has low activity and occasional branching (20). Meanwhile, the FDG accumulation at the site of inflammatory lesions (21) may lead to the occurrence of false-positives on 2-[¹⁸F]FDG PET/CT when diagnosing CCA. Therefore, in cases where patients with bile duct space-occupying lesions undergo 2-[¹⁸F]FDG PET/CT examination and exhibit positive FDG uptake lesions, particularly when only bile duct lesions are present, it becomes challenging to ascertain the characteristic of these lesions. Previous studies on 2-[¹⁸F]FDG PET/CT in CCA have mostly focused on staging, evaluation of treatment response, and prediction of prognosis (7-11). According to our knowledge, there is currently a lack of research on the differential diagnosis between CCA and cholangitis. Our study summarized the clinical variables and 2-[¹⁸F]FDG PET/CT metabolic parameters of non-metastatic CCA patients and compared them with those of cholangitis patients. Considering the possible differences in FDG uptake between iCCA and eCCA, we compared the 2-[¹⁸F]FDG PET/CT metabolic parameters of these two types of non-metastatic CCA and corresponding anatomical positions of cholangitis. In the differential diagnosis of non-metastatic iCCA and intrahepatic cholangitis, SUVpeak remains the 2-[¹⁸F]FDG PET/CT metabolic parameter with the best diagnostic ability (AUC =0.740, 95% CI: 0.603–0.878). The SUVpeak is determined by averaging the pixel values on a 1.0 cc spherical nucleus centered on the high uptake portion of the tumor (22), and studies have shown that SUVpeak is more resistant to statistical fluctuations than SUVmax (23,24) (Figure 3). However, the use of MTV demonstrated superior diagnostic capability in distinguishing between non-metastatic eCCA and extrahepatic cholangitis (AUC =0.669, 95% CI: 0.551–0.787). The volumetric parameters of MTV represent the volume of lesions exhibiting active FDG uptake (25). It is worth noting that the MTV of extrahepatic cholangitis is higher than that of eCCA, potentially attributed to the broader cumulative range observed in cases of extrahepatic cholangitis when compared to eCCA (Figure 4). Extrahepatic cholangitis is primarily characterized by diffuse inflammatory reactions, whereas eCCA manifests as focal tumor proliferation. Although necrosis is rare in cholangitis, its extensive involvement of the bile duct wall results in a larger MTV. In contrast, eCCA exhibits high metabolic activity but confined growth, leading to localized high SUVmax yet a smaller MTV. Another potential factor contributing to the lower MTV in eCCA is the 40% SUVmax threshold used in this study for MTV segmentation, which may only include the core hypermetabolic region of the tumor, whereas the lower threshold in cholangitis allows more surrounding inflammatory tissue to be incorporated.

Figure 3 iCCA lesions exhibit higher FDG uptake (SUVpeak) compared to cholangitis lesions. (A) A 62-year-old woman was diagnosed with moderately-poorly differentiated iCCA (yellow arrows). The patient was found to have a mass in the left lobe of the liver near the hepatic hilum during ultrasound examination 1 month prior, with no obvious clinical presentation. The patient had no history of cholecystolithiasis. The CEA was 1.22 ng/mL and the CA19-9 was 48.89 U/mL. The lesion showed that SUVpeak was 8.5. (B) A 50-year-old woman was diagnosed with intrahepatic cholangitis (blue arrows). The patient was admitted for examination 1 week due to unexplained abdominal pain. The patient had a medical history of cholecystolithiasis. The CEA was 1.73 ng/mL and the CA19-9 was 21.67 U/mL. The lesion showed that the SUVpeak was 3.2. CA19-9, carbohydrate antigen 19-9; CEA, carcinoembryonic antigen; FDG, fluoro-D-glucose; iCCA, intrahepatic cholangiocarcinoma; SUVpeak, peak standardized uptake value.

Figure 4 eCCA exhibits lower MTV compared to extrahepatic cholangitis lesions despite its malignant nature. (A) Image A showed a 66-year-old man with moderately-poorly differentiated eCCA (yellow arrows). The patient was admitted for examination due to unexplained abdominal pain for over a month, without fever. The patient had no history of cholecystolithiasis. The CEA was 5.00 ng/mL and the CA19-9 was 101.7 U/mL. The lesion showed that MTV was 10.1. (B) A 65-year-old male was diagnosed with hilar extrahepatic cholangitis (blue arrows). The patient was admitted due to unexplained fever with jaundice for 1 week. The patient had no medical history of cholecystolithiasis. The CEA was 1.33 ng/mL and the CA19-9 was 15.99 U/mL. The lesion showed that MTV was 83.4. CA19-9, carbohydrate antigen 19-9; CEA, carcinoembryonic antigen; eCCA, extrahepatic cholangiocarcinoma; MTV, metabolic tumor volume.

The 2-[¹⁸F]FDG PET/CT metabolic parameters showed a moderate accuracy in distinguishing between non-metastatic iCCA and intrahepatic cholangitis, and non-metastatic eCCA and extrahepatic cholangitis, respectively. Relying solely on a single parameter may not achieve higher diagnostic efficacy. Therefore, we constructed models based on clinical variables and 2-[¹⁸F]FDG PET/CT metabolic parameters that differ between non-metastatic iCCA and intrahepatic cholangitis, and non-metastatic eCCA and extrahepatic cholangitis, respectively. We initially integrated SUVpeak with the patient’s medical history of cholangiolithiasis and laboratory examination results indicating CA19-9 levels exceeding 37 ng/mL to develop a comprehensive model. By comparing it to the sole use of SUVpeak, we observed significant enhancements in both the differential diagnostic capability for non-metastatic iCCA and intrahepatic cholangitis, as confirmed by improvements in IDI and NRI. Meanwhile, we also constructed a diagnostic model {including 2-[¹⁸F]FDG PET/CT metabolic parameters (MTV), clinical presentation (fever), medical history (cholangitis), and laboratory examination (CA19-9 >37 ng/mL)} for distinguishing non-metastatic eCCA from extrahepatic cholangitis, and compared to 2-[¹⁸F]FDG PET/CT metabolic parameter (MTV) alone, the differential diagnostic efficiency was significantly improved. The development of diagnostic models can potentially compensate for the limitations inherent in routine PET/CT reports, thereby enhancing the accuracy of diagnosing non-metastatic CCA.

Our study had some limitations. Firstly, our study, being a single-center retrospective analysis, is limited by its relatively small sample size, particularly in terms of patients diagnosed with cholangitis. However, the diagnostic accuracy was ensured in our study as all patients were diagnosed through postoperative pathology. Secondly, due to the lack of detailed clinical data for some preoperative patients, only partial and complete data were analyzed, such as total bilirubin, direct bilirubin, and alkaline phosphatase, carbohydrate antigen 125 (CA125), circulating tumor DNA (ctDNA), and circulating tumor cells (CTCs) (4). Given the inconsistent selection of other preoperative imaging modalities (such as contrast-enhanced CT and MRI) and the limited spatial resolution of CT in PET/CT for these patients, relevant parameters such as lesion morphology were not incorporated into the study. This may have led to bias in model construction and may affect the diagnostic performance of the model. Thirdly, the measurement of 2-[¹⁸F]FDG PET/CT metabolic parameters (SUV) may be influenced by many factors (26), which may affect the diagnostic efficacy of 2-[¹⁸F]FDG PET/CT metabolic parameters and the reproducibility of models constructed using metabolic parameters in differential diagnosis. Furthermore, the discrimination model developed in this study has not undergone external validation, thereby limiting the ability to accurately assess its applicability. In the future, prospective, larger sample size, and multi-center studies are necessary to validate the reliability and reproducibility of the model. Meanwhile, the addition of PET imaging omics related parameters can further improve the accuracy, stability, and repeatability of the model.

Conclusions

Our study found that 2-[¹⁸F]FDG PET/CT metabolic parameters can help distinguish between non-metastatic iCCA and intrahepatic cholangitis (SUVpeak), and non-metastatic eCCA and extrahepatic cholangitis (MTV), respectively. The diagnostic model {including 2-[¹⁸F]FDG PET/CT metabolic parameters (SUVpeak), clinical presentation (fever), medical history (cholangiolithiasis), and laboratory examination (CA19-9 >37 ng/mL)} can effectively help differentiate non-metastatic iCCA and cholangitis; meanwhile, the diagnostic model {including 2-[¹⁸F]FDG PET/CT metabolic parameters (MTV), clinical presentation (fever), medical history (cholangitis), and laboratory examination (CA19-9 >37 ng/mL)} can effectively help differentiate non-metastatic eCCA and extrahepatic cholangitis. Our study can effectively and accurately aid in the identification of patients with non-metastatic CCA and cholangitis, thereby mitigating the risk of incorrect diagnoses that may lead to unnecessary treatments.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1626/rc

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1626/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective cohort study was approved by the Ethics Committee of the People’s Liberation Army General Hospital (No. S2014-052-01). All patients were informed and signed a consent form before the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Oliveira IS, Kilcoyne A, Everett JM, Mino-Kenudson M, Harisinghani MG, Ganesan K. Cholangiocarcinoma: classification, diagnosis, staging, imaging features, and management. Abdom Radiol (NY) 2017;42:1637-49. [Crossref] [PubMed]
2. Cardinale V. Classifications and misclassification in cholangiocarcinoma. Liver Int 2019;39:260-2. [Crossref] [PubMed]
3. Khan AS, Dageforde LA. Cholangiocarcinoma. Surg Clin North Am 2019;99:315-35. [Crossref] [PubMed]
4. Ohaegbulam KC, Koethe Y, Fung A, Mayo SC, Grossberg AJ, Chen EY, Sharzehi K, Kardosh A, Farsad K, Rocha FG, Thomas CR, Nabavizadeh N. The multidisciplinary management of cholangiocarcinoma. Cancer 2023;129:184-214. [Crossref] [PubMed]
5. Zeng N, Sun JX, Liu CQ, Xu JZ, An Y, Xu MY, Zhang SH, Zhong XY, Ma SY, He HD, Wang SG, Xia QD. Knowledge mapping of application of image-guided surgery in prostate cancer: a bibliometric analysis (2013-2023). Int J Surg 2024;110:2992-3007. [Crossref] [PubMed]
6. Kim NH, Lee SR, Kim YH, Kim HJ. Diagnostic Performance and Prognostic Relevance of FDG Positron Emission Tomography/Computed Tomography for Patients with Extrahepatic Cholangiocarcinoma. Korean J Radiol 2020;21:1355-66. [Crossref] [PubMed]
7. Lee SW, Kim HJ, Park JH, Park DI, Cho YK, Sohn CI, Jeon WK, Kim BI. Clinical usefulness of 18F-FDG PET-CT for patients with gallbladder cancer and cholangiocarcinoma. J Gastroenterol 2010;45:560-6. [Crossref] [PubMed]
8. Albazaz R, Patel CN, Chowdhury FU, Scarsbrook AF. Clinical impact of FDG PET-CT on management decisions for patients with primary biliary tumours. Insights Imaging 2013;4:691-700. [Crossref] [PubMed]
9. Kitamura K, Hatano E, Higashi T, Seo S, Nakamoto Y, Narita M, Taura K, Yasuchika K, Nitta T, Yamanaka K, Doi R, Ikai I, Uemoto S. Prognostic value of (18)F-fluorodeoxyglucose positron emission tomography in patients with extrahepatic bile duct cancer. J Hepatobiliary Pancreat Sci 2011;18:39-46. [Crossref] [PubMed]
10. Yuminaga Y, Rothe C, Kam J, Beattie K, Arianayagam M, Bui C, Canagasingham B, Ferguson R, Khadra M, Ko R, Le K, Nguyen D, Varol C, Winter M. (68)Ga-PSMA PET/CT versus CT and bone scan for investigation of PSA failure post radical prostatectomy. Asian J Urol 2021;8:170-5. [Crossref] [PubMed]
11. Moon CM, Bang S, Chung JB. The role of (18)F-fluorodeoxyglucose positron emission tomography in the diagnosis, staging, and follow-up of cholangiocarcinoma. Surg Oncol 2011;20:e10-7. [Crossref] [PubMed]
12. Yang T, Yang Y, Hu S, Xie Y, Shen Z, Zhang Y. (18)F-FAPI-42 PET/CT for preoperatively identifying intrahepatic cholangiocarcinoma and hepatocellular carcinoma. Quant Imaging Med Surg 2025;15:741-51. [Crossref] [PubMed]
13. Wang G, Du L, Lu X, Chen S, Bi X, Zhang M, Luan X, Xu X, Zhang Z, Xu B, Yang J. Multiparameter diagnostic model based on (18)F-FDG positron emission tomography and serological examination for differentiating focal autoimmune pancreatitis from pancreatic ductal adenocarcinoma. Quant Imaging Med Surg 2023;13:5653-63. [Crossref] [PubMed]
14. Ilyas SI, Gores GJ. Pathogenesis, diagnosis, and management of cholangiocarcinoma. Gastroenterology 2013;145:1215-29. [Crossref] [PubMed]
15. Nakeeb A, Pitt HA, Sohn TA, Coleman J, Abrams RA, Piantadosi S, Hruban RH, Lillemoe KD, Yeo CJ, Cameron JL. Cholangiocarcinoma. A spectrum of intrahepatic, perihilar, and distal tumors. Ann Surg 1996;224:463-73; discussion 473-5. [Crossref] [PubMed]
16. van Vugt JLA, Gaspersz MP, Coelen RJS, Vugts J, Labeur TA, de Jonge J, Polak WG, Busch ORC, Besselink MG, IJzermans JNM, Nio CY, van Gulik TM, Willemssen FEJA, Groot Koerkamp B. The prognostic value of portal vein and hepatic artery involvement in patients with perihilar cholangiocarcinoma. HPB (Oxford) 2018;20:83-92. [Crossref] [PubMed]
17. Eloubeidi MA, Chen VK, Jhala NC, Eltoum IE, Jhala D, Chhieng DC, Syed SA, Vickers SM, Mel Wilcox C. Endoscopic ultrasound-guided fine needle aspiration biopsy of suspected cholangiocarcinoma. Clin Gastroenterol Hepatol 2004;2:209-13. [Crossref] [PubMed]
18. Macias RIR, Kornek M, Rodrigues PM, Paiva NA, Castro RE, Urban S, Pereira SP, Cadamuro M, Rupp C, Loosen SH, Luedde T, Banales JM. Diagnostic and prognostic biomarkers in cholangiocarcinoma. Liver Int 2019;39:108-22. [Crossref] [PubMed]
19. Sulzer JK, Ocuin LM. Cholangitis: Causes, Diagnosis, and Management. Surg Clin North Am 2019;99:175-84. [Crossref] [PubMed]
20. Koppula BR, Fine GC, Salem AE, Covington MF, Wiggins RH, Hoffman JM, Morton KA. PET-CT in Clinical Adult Oncology: III. Gastrointestinal Malignancies. Cancers (Basel) 2022;14:2668. [Crossref] [PubMed]
21. Rahman WT, Wale DJ, Viglianti BL, Townsend DM, Manganaro MS, Gross MD, Wong KK, Rubello D. The impact of infection and inflammation in oncologic (18)F-FDG PET/CT imaging. Biomed Pharmacother 2019;117:109168. [Crossref] [PubMed]
22. Shepherd T, Teras M, Beichel RR, Boellaard R, Bruynooghe M, Dicken V, Gooding MJ, Julyan PJ, Lee JA, Lefèvre S, Mix M, Naranjo V, Wu X, Zaidi H, Zeng Z, Minn H. Comparative study with new accuracy metrics for target volume contouring in PET image guided radiation therapy. IEEE Trans Med Imaging 2012;31:2006-24. [Crossref] [PubMed]
23. Lodge MA, Chaudhry MA, Wahl RL. Noise considerations for PET quantification using maximum and peak standardized uptake value. J Nucl Med 2012;53:1041-7. [Crossref] [PubMed]
24. Vanderhoek M, Perlman SB, Jeraj R. Impact of the definition of peak standardized uptake value on quantification of treatment response. J Nucl Med 2012;53:4-11. [Crossref] [PubMed]
25. Hirata K, Tamaki N. Quantitative FDG PET Assessment for Oncology Therapy. Cancers (Basel) 2021;13:869. [Crossref] [PubMed]
26. 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;36:200-2. [Crossref] [PubMed]