Contrast-enhanced ultrasound and contrast-enhanced computed tomography for differentiating mass-forming pancreatitis from pancreatic ductal adenocarcinoma: a meta-analysis : Chinese Medical Journal

Secondary Logo

Journal Logo

Meta Analysis

Contrast-enhanced ultrasound and contrast-enhanced computed tomography for differentiating mass-forming pancreatitis from pancreatic ductal adenocarcinoma: a meta-analysis

Yang, Jie1; Huang, Jiayan2; Zhang, Yonggang3; Zeng, Keyu1; Liao, Min1; Jiang, Zhenpeng1; Bao, Wuyongga1; Lu, Qiang1

Editor(s): Guo, Lishao

Author Information
Chinese Medical Journal ():10.1097/CM9.0000000000002300, December 26, 2022. | DOI: 10.1097/CM9.0000000000002300



Mass-forming pancreatitis (MFP) is a specific type of chronic pancreatitis associated with autoimmune responses, alcohol consumption, tobacco use, and a history of biliary tract disease.[1,2] It is challenging to differentiate MFP from pancreatic ductal adenocarcinoma (PDAC) because both conditions may present with clinically recurrent abdominal pain, weight loss, exocrine and endocrine pancreatic insufficiency, and a pancreatic mass on imaging. In some patients, PDAC may develop from chronic pancreatitis or may be accompanied by chronic pancreatitis.[2,3] Considering the different treatment strategies, management strategies, and prognoses for MFP and PDAC, accurate differential diagnosis of MFP and PDAC is of paramount importance to prevent unnecessary surgical resection in patients suffering solely from MFP.

Contrast-enhanced ultrasound (CEUS) is a suitable imaging method for the delineation of the intra-pancreatic macro- and microvasculature, and has been recommended by the guidelines for the characterization of solid pancreatic lesions.[4,5] MFP may display an enhancement pattern on CEUS that is similar to that of normal pancreatic parenchyma.[6] Several studies have also reported the diagnostic performance of CEUS for benign pancreatic lesions and have described substantial variability, i.e., a sensitivity of 67.0% to 94.0% and specificity of 30.0% to 100%.[6-13] Contrast-enhanced computed tomography (CECT) and magnetic resonance imaging have been widely used for the preoperative evaluation and post-operative follow-up of pancreatic masses, especially malignant pancreatic tumors; however, limited information regarding the significance of CEUS for patients with PDAC has been reported.[4,5,14,15] One meta-analysis described the overall diagnostic accuracy of CEUS for differentiating PDAC from other pancreatic lesions.[16] However, to the best of our knowledge, no meta-analysis has been conducted to explore the differential diagnosis of PDAC and MFP using CEUS. Only two studies compared the performance of CEUS and CECT for the differential diagnosis of MFP and PDAC; however, these studies were published before 2015 and included ≤120 samples in total;[7,8] thus, the optimal imaging method in differentiating MFP from PDAC could not be determined. Therefore, in this study, we investigated whether the ability of CEUS to diagnose PDAC was comparable to that of CECT. Considering that ultrasound contrast agents had pure blood perfusion characteristics, and that PDAC always showed poor perfusion in contrast-enhanced examination due to its strong intralesional desmoplastic reaction, we hypothesized that the diagnostic performance of CEUS might be comparable to that of CECT in the differential diagnosis of PDAC and MFP. Since previous studies did not reach reliable conclusions, we performed this meta-analysis to compare the diagnostic performance of CEUS and CECT for differentiating MFP from PDAC, and to compare the two imaging modalities.


Literature search strategy

The PubMed, EMBASE, The Cochrane Library (CENTRAL), China National Knowledge Infrastructure (CNKI), Weipu (VIP), and WanFang databases or other sources were searched to identify articles reporting the performance of CEUS and CECT in the differential diagnosis of MFP and PDAC in humans. The search time was from inception until August 20, 2021. The query terms included “pancreatitis,” “pancreatic neoplasm,” “CECT,” “CEUS,” “diagnosis,” and “differential.” The detailed search terms are listed in Supplementary Table 1, The studies were limited to English and Chinese. The study was reported according to the Meta-analysis of Observational Studies in Epidemiology guidelines [Supplemental Table 1,].[17]

Inclusion and exclusion criteria

The inclusion criteria were as follows: (a) patients with MFP or PDAC; (b) index test: CEUS; (c) comparison test: CECT; (d) outcomes: diagnostic accuracy for differentiating MFP from PDAC; (e) study design: clinical diagnoses; and (f) reference standard: both histological and clinical diagnoses. The exclusion criteria were as follows: (a) studies with duplicated patients or data; (b) studies that lacked sufficient information to create a 2 × 2 diagnostic table or to explain the diagnostic criteria; (c) case reports, reviews, editorials, letters, and conference abstracts; (d) studies not within the field of interest; and (e) studies without human subjects. Two reviewers (Huang J.Y. and Yang J., with 10 and 5 years of experience in abdominal ultrasonic radiology) screened the abstracts and titles of the selected studies and reread the full texts of potentially eligible studies. Both review sessions were completed independently, and studies with definite ineligibility were excluded. Studies about which the reviewers disagreed were discussed, and disagreements were resolved by consensus.

Data extraction

Two reviewers independently extracted the data from the included studies. The following data were extracted from the eligible studies: (a) study characteristics (the first author, year of publication, country in which the articles were published, and study design [retrospective or prospective]); (b) lesion characteristics {numbers of MFPs and PDACs, and pancreatitis subtype (chronic mass-form pancreatitis [CMFP] or autoimmune pancreatitis [AIP])}; (d) CECT techniques (multiphase enhanced images, slice thickness, and number of channels); (e) the number of raters; and (f) the reference criteria for pancreatitis and PDAC (histopathological or clinical criteria).

To determine the diagnostic accuracy, the numbers of true positives, false positives, false negatives, and true negatives were extracted. Since individual characteristics were used in the case of multiple sets of results, the highest Youden index value was considered for further analyses. The two reviewers independently extracted the data, and all disagreements were reassessed in a consensus meeting.

The quality of the reviewed studies was evaluated using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) criteria.[18] Two reviewers independently carried out the data analysis, evaluated the quality assessment of the studies, and reached a consensus by discussion.

Data synthesis and statistical analysis

To examine the effect of each imaging test on diagnostic accuracy, the results of all the studies were separated and evaluated as separate studies according to the type of modality (CECT vs. CEUS). Statistical analysis was performed on a per-lesion basis. The sensitivity and specificity of the differential diagnosis of MFP and PDAC, and their 95% confidence intervals (CIs), were calculated using the relevant data extracted from each individual study. The sensitivity was defined as the number of patients diagnosed with MFPs on imaging divided by all patients with MFPs, while the specificity was defined as the number of patients presenting PDACs on imaging divided by all patients with PDAC. The diagnostic ability of CEUS and CECT for differentiating MFP from PDAC was compared by Bayesian network meta-analysis.

The threshold effect between the sensitivity and false positive rate was evaluated by Spearman correlation analysis. A correlation coefficient >0.6 was considered to indicate a considerable threshold effect. Cochran's Q test and Higgins’ I2 statistic were used to determine the heterogeneity, and P value <0.1 or an I2 statistic >50% was considered to indicate significant heterogeneity.[19,20] If there was no threshold effect present, a random effects model was used to obtain pooled estimates of the sensitivity, specificity, positive likelihood ratio (+LR), negative likelihood ratio (−LR), and diagnostic odds ratio (DOR). Bivariate and hierarchical summary receiver operating characteristic models were used to provide equivalent summary estimates for sensitivity and specificity. We performed a Bayesian network meta-analysis to indirectly compare the overall diagnostic performance.[21,22]

If significant heterogeneity presented, a meta-regression analysis was conducted to investigate the cause of heterogeneity. The following covariates were used in the meta-regression: (a) patient numbers (patients ≤50 vs. patients >50); (b) publication year (before 2015 vs. after 2015); (c) subtype of pancreatitis (CMFP vs. AIP); (d) CT slice thickness (≤3 mm vs. >3 mm); (e) the use of multiphase enhanced CT (multiphase vs. single-phase); (f) the use of multiple CT channels (≥64 vs. <64 or unclear); (g) image reviewer (single or unclear vs. multiple raters); (h) reference standard for MFP and PDAC, respectively (unique vs. not unique); (i) country of publication (Asia or non-Asian country); and (j) blinded review (clear or unclear). Publication bias was determined by visual assessment of Deeks’ funnel plot, and statistical significance was evaluated by Deeks’ asymmetry test. All statistical analyses were performed using Stata version 15.0 (College Station, USA) software, and P < 0.05 was considered statistically significant.


Literature search

The study selection process is shown in [Figure 1]. A total of 1558 studies were identified after searching the databases. After screening the studies, 26 studies with 2115 pancreatic masses were included.[6-13,23-40] Of the 26 included studies, 18 used only CECT for diagnosis,[23-40] seven used only CEUS for diagnosis,[6,8-13] and only one cohort study simultaneously compared the differential diagnosis between CEUS and CECT.[7] Fan et al[8] did not report the diagnostic standard of CECT for pancreatic lesions; thus, the CECT data were not included for data analysis. The characteristics of the included articles are summarized in [Table 1]. Regarding the subtype of MFP, 11 studies included AIP, and 15 studies included CMFP [seven studies were conducted with CEUS[6-10,12,13] and eight were conducted with CECT.[23,31,33,34,36-40] Of the 19 articles that evaluated CECT, 15 used multiphase contrast-enhanced images,[23,25-37,39,40] three used single/two-phase contrast-enhanced images, and one study did not report the phase. Twelve studies used thin slices (≤3 mm),[7,23,25-28,31-35,40] whereas the others used a slice thickness >3 mm or did not report the slice thickness. Of the 26 studies, 20 studies involved multiple image raters (>2 raters),[8,9,24-29,31,33-35] and the remaining studies involved only one or an unclear number of image raters.

Figure 1:
Flowchart of the studies’ selection.
Table 1 - Characteristics of the 26 eligible included articles.
Reference standard CECT

Study Country Study design No. analyzed Sub-type of pancreatitis MFP PDAC Image test Reviewer Phase Slice Channel
D’Onofrio et al [6] Italy Retros- 35/138 CMFP His- His- CEUS NA
Fan et al [8] China Retros- 28/36 CMFP His-/clinical His-/clinical CEUS Three
Wang et al [9] China Retros- 25/86 CMFP His-/clinical His- CEUS Two
Xie et al [10] China Retros- 8/42 CMFP His-/clinical His-/clinical CEUS Two
Xu and Feng [12] China Retros- 5/34 CMFP His-/clinical His-/clinical CEUS Two
Li et al [13] China Retros- 9/39 CMFP His-/clinical His-/clinical CEUS Three
Yuan et al [11] China Retros- 2/26 AIP His- His- CEUS Two
Grossjohann et al [7] Denmark Retros- 5/44 CMFP His- His- CEUS/CECT NA Single ≤3 64
Chari et al [24] USA Retros- 48/100 AIP HISORt His-/clinical CECT Single Single >3 NA
Yamada et al [33] Japan Retros- 18/34 CMFP His-/clinical His-/clinical CECT Two Three ≤3 NA
Brimiene et al [23] USA Pros- 58/78 CMFP His- His- CECT NA Multiple ≤3 64
Kawai et al [27] Japan Retros- 55/50 AIP HISORt/ JPC/ KDC His- CECT Two Multiple ≤3 ≥64
Muhi et al [29] Japan Retros- 10/70 AIP JPC His- CECT Three Multiple 5 16
Naitoh et al [30] Japan Retros- 36/60 AIP JPC/rHISORt His-/clinical CECT Single Multiple NA NA
Sun et al [32] China Retros- 19/30 AIP ADCA His- CECT Three Multiple ≤3 ≤64
Zaheer et al [35] USA Retros- 32/32 AIP His-/ICDC His- CECT Three Multiple <3 128
Furuhashi et al [26] Japan Retros- 22/61 AIP ICDC/JPC His- CECT Two Multiple ≤2 64
Yin et al [34] China Retros- 15/20 CMFP His- His- CECT Three Two 2.5 64
Lee et al [28] Korea Retros- 61/122 AIP HISORt/ICDC His- CECT Two Multiple <3 ≥64
Ren et al [31] China Retros- 21/47 CMFP His- His- CECT Two Multiple <3 ≥64
Linning et al [25] China Retros- 45/51 AIP His-/ICDC His- CECT Two Multiple ≤3 256
Luo et al [37] China Retros- 25/31 CMFP His-/clinical His-/clinical CECT Two Multiple 6 16
Lv et al [38] China Retros- 42/59 CMFP His- His- CECT Two Multiple 5 320
Wang et al [39] China Retros- 32/36 CMFP His- His- CECT NA NA 6 64
Zhu and Lu [40] China Retros- 21/21 AIP His- His- CECT Two Multiple 1 320
Guo [36] China Retros- 37/35 CMFP His- His- CECT Three Multiple 5 32
ADCA: The Asian Diagnostic Criteria for Autoimmune Pancreatitis; AIP: Autoimmune pancreatitis; Clinical: Clinical diagnosis; CMFP: Chronic mass-form pancreatitis; CECT: Contrast-enhanced computed tomography; CEUS: Contrast-enhanced ultrasound; His-: Histological diagnosis; HISORt: Histology, imaging, serology, other organ involvement, and response to steroid therapy; ICDC: International Consensus Diagnostic Criteria; JPC: Japanese diagnostic criteria; KDC: Korean diagnostic criteria; MFP: Mass-forming pancreatitis; NA: Not available; Pros-: Prospective; PDAC: Pancreatic ductal adenocarcinoma; Retros-: Retrospective study; rHISORt: Revised histology, imaging, serology, other organ involvement, and response to steroid therapy.

For the reference diagnosis of MFP, ten studies used only histopathology,[6,7,11,23,31,34,36,38-40] and the others used only clinical criteria or both histopathology and clinical criteria. For the reference diagnosis of PDAC, 18 studies used only histopathology,[6,7,9,23-29,31,32,34,35] and eight studies employed both clinical follow-up and histopathology. The detailed imaging criteria for differentiating MFP from PDAC in the 26 included studies are shown in [Supplementary Table 2,].

Study quality of included studies

Based on the results of QUADAS-2, the quality assessment of the 26 included studies was considered good. The domain of flows and timing was the main quality issue because the patients included in 14 of the 26 studies were compared with more than one reference standard. Patient selection and reference standards were pivotal areas of quality concern, which performed well in the current study. Twenty-four studies included consecutive patients, while two studies included random or matched patients.[31,34] The other five studies did not clearly describe the patient selection process.[8,24,28-30] Eight studies reported patients with PDAC confirmed by histopathology or clinical follow-up results, while the other studies diagnosed MFP by various criteria. Thus, the reference standard bias of the included studies was unclear. A composed reference, to some extent, could avoid inappropriate exclusions but also cause a risk of bias in the reference standard. In addition, when we assessed the section of the index test, 11 studies were unclear about the information of the blinded review,[6,7,11,12,23,30,36-40] but had acceptable quality. The detailed assessed results are shown in [Figure 2A and 2B and Supplementary Tables 3A and 3B,].

Figure 2:
Quality assessment of the studies according to QUADAS-2 criteria. (A) The methodological quality of the articles is presented as the proportion of 26 articles with low (ie, high quality), or high risk of bias; and (B) the proportion of articles with low, or high unclear concerns regarding the applicability of each domain. The X axis presents the numbers of the included 26 articles. QUADAS-2: Quality Assessment of Diagnostic Accuracy Studies 2.

Diagnostic performance of CEUS and CECT for differentiating MFP from PDAC

There was no significant threshold effect among the eight studies (562 lesions) that used CEUS (Spearman correlation coefficient = 0.270, P = 0.518). The sensitivities (72%–100%) and specificities (86%–100%) of CEUS and the 95% Cis of each study are presented in [Figure 3A]. The pooled sensitivity and specificity values for CEUS were 82% (95% CI, 73%–88%; I2 = 0.00%) and 95% (95% CI, 90%–97%; I2 = 63.44%), respectively. The pooled +LR, −LR, and DOR values were 15.12 (95% CI, 7.61–30.01), 0.19 (95% CI, 0.13–0.29), and 78.91 (95% CI, 30.94–201.27), respectively [Supplementary Figure 1A,B,], and the SROC value was high (AUC = 0.90; 95% CI, 0.87–0.92; Figure 4A).

Figure 3:
Coupled forest plots of the sensitivity and specificity for the differential diagnosis of MFP from PDAC on CEUS (A) and CECT (B). CECT: Contrast-enhanced computed tomography; CEUS: Contrast-enhanced ultrasound; MFP: Mass-forming pancreatitis; PDAC: Pancreatic ductal adenocarcinoma.
Figure 4:
SROC of the sensitivity and specificity for the differential diagnosis of MFP from PDAC on CEUS (A) and CECT (B). CECT: Contrast-enhanced computed tomography; CEUS: Contrast-enhanced ultrasound; SROC: Summary receiver operating characteristic; MFP: Mass-forming pancreatitis; PDAC: Pancreatic ductal adenocarcinoma.

For the 19 CECT studies, no significant threshold effect was found (Spearman correlation coefficient = 0.007, P = 0.076). The pooled sensitivity and specificity for MFP and the 95% CI of each study are shown in Figure 3B. The pooled sensitivity and specificity were 81% (95% CI, 75–85%; I2 = 66.37%) and 94% (95% CI, 90–96%; I2 = 74.87%), respectively. The pooled +LR, −LR, and DOR values were 12.91 (95% CI, 7.86–21.20), 0.21 (95% CI, 0.16–0.27), and 62.53 (95% CI, 34.45–113.51), respectively [Supplementary Figure 1C,D,], and the SROC AUC was 0.92 (95% CI, 0.90–0.94; Figure 4B). A summary of the data about the two modalities reported in the 26 studies is shown in [Table 2].

Table 2 - The diagnostic performance of CEUS and CECT for differentiating MFP from PDAC.
Number of patients

Study No. of patients TP FP FN TN Sensitivity (%, 95% CI) Specificity (%, 95% CI)
 D’Onofrio et al [6] 173 31 3 4 135 89 (73–97) 98 (94–100)
 Grossjohann et al [7] 49 4 6 1 38 80 (28–99) 86 (73–95)
 Fan et al [8] 64 23 3 5 33 82 (63–94) 92 (78–98)
 Wang et al [9] 111 18 9 7 77 72 (51–88) 90 (81–95)
 Xu and Feng [12] 39 5 4 0 30 100 (48–100) 88 (73–97)
 Yuan et al [11] 28 2 2 0 24 100 (16–100) 92 (75–99)
 Li et al [13] 48 7 0 2 39 78 (40–97) 100 (91–100)
 Xie et al [10] 50 6 0 2 42 75 (35–97) 100 (92–100)
Higgins I 2 for heterogeneity 0.00 63.44
Meta-analytic summary estimate using the bivariate model 82 (73–88) 95 (90–97)
 Chari et al [24] 148 25 0 23 100 52 (37–67) 100 (96–100)
 Grossjohann et al [7] 49 2 3 3 41 40 (5–85) 93 (81–99)
 Yamada et al [33] 52 15 2 3 32 83 (59–96) 94 (80–99)
 Brimiene et al [23] 136 52 19 6 69 90 (79–96) 78 (68–86)
 Muhi et al [29] 81 11 1 0 69 100 (72–100) 99 (92–100)
 Kawai et al [27] 105 37 5 18 45 67 (53–79) 90 (78–97)
 Naitoh et al [30] 96 30 0 6 60 83 (67–94) 100 (94–100)
 Sun et al [32] 49 17 0 2 30 89 (48–84) 97 (85–100)
 Zaheer et al [35] 64 22 6 10 26 69 (50–84) 81 (64–93)
 Furuhashi et al [26] 83 19 1 4 60 83 (61–95) 98 (91–100)
 Yin et al [34] 35 14 2 1 18 93 (68–100) 90 (68–99)
 Lee et al [28] 183 49 10 12 112 80 (68–89) 92 (85–96)
 Ren et al [31] 68 18 7 3 44 86 (64–97) 85 (72–94)
 Linning et al [25] 96 33 6 12 45 73 (58–85) 88 (76–96)
 Wang et al [39] 68 25 4 7 32 78 (60–91) 89 (74–97)
 Zhu and Lu [40] 42 17 2 4 19 81 (58–95) 90 (70–99)
 Luo et al [37] 56 20 7 5 24 80 (59–93) 77 (59–90)
 Guo [36] 90 40 2 5 43 89 (76–96) 96 (85–99)
 Lv et al [38] 101 37 3 5 56 88 (74–96) 95 (86–99)
Higgins I 2 for heterogeneity 66.37 74.87
Meta-analytic summary estimate using the bivariate model 81 (75–85) 94 (90–96)
The value of I2: 0% to 25%, heterogeneity might not be important; 25% to 50%, may represent low heterogeneity; 50% to 75%, may represent moderate heterogeneity; 75% to 100%, high heterogeneity.CI: Confidence index; CECT: Contrast-enhanced computed tomography; CEUS: Contrast-enhanced ultrasound; FP: False positive; FN: False negative; MFP: Mass-form pancreatitis; PDAC: Pancreatic ductal adenocarcinoma; TP: True positive; TN: True negative.

Based on the result of adjusted indirect comparison, the results showed that CEUS and CECT had comparable overall diagnostic accuracy for the differentiation of MFP from PDAC (relative DOR [RDOR] = 1.26, 95% CI [0.42–3.83], P > 0.05; Figure 5).

Figure 5:
The overall diagnostic performance of CEUS and CECT for differentiating MFP from PDAC by using network meta-regression analysis. AUC: Area under the SROC curve; CECT: Contrast-enhanced computed tomography; CEUS: Contrast-enhanced ultrasound; DOR: Diagnostic odds ratio; LR: likelihood ratio; MFP: Mass-forming pancreatitis; PDAC: Pancreatic ductal adenocarcinoma.

Meta-regression analysis

The heterogeneity of pooled sensitivity on CEUS for MFP was acceptable; however, moderate heterogeneity (I2 = 63.44%) in the pooled specificity of CEUS was observed. The meta-regression analysis results are shown in [Supplementary Table 4,]. The results indicated that all factors were not strongly associated with the specificity of CEUS for differentiating MFP from PDAC (all P ≥ 0.05). Furthermore, no publication bias was found for the CEUS studies (t = −0.55, P > |t| = 0.60; Supplementary Figure 2A,

The results of the meta-regression analysis of CECT are summarized in [Table 3]. The publication year, subtype of pancreatitis, CT slice thickness (meets the ≤3 mm), CT channels, blinded review, and single reference criteria for MFP were significant factors associated with the study heterogeneity of the sensitivity on CECT (all P ≤ 0.02). In addition, CT slice thickness (≤3 mm), CT channels, the presence of more than two reviewers, and the use of a single reference criteria for PDAC were significantly associated with the study heterogeneity for the specificity of CECT (all P ≤ 0.02). No remarkable publication bias was found in the CECT studies (t = 0.12, P > |t| = 0.91; Supplementary Figure 2B,

Table 3 - Results of the meta-regression analysis on the CECT studies.
Covariate Subgroup Sensitivity (%, 95% CI) P values Specificity (%, 95% CI) P values
Country of publication Asia (n = 15) 83 (78–88) 0.43 94 (91–97) 0.18
Non-Asia (n = 4) 70 (56–83) 92 (85–100)
Patients’ numbers ≤ 50 (n = 4) 83 (71–95) 0.17 95 (89–100) 0.35
> 50 (n = 15) 80 (75–84) 92 (89–96)
Publication year Before 2015 (n = 13) 80 (75–86) <0.01 93 (90–97) 0.17
Year 2015 and after (n = 6) 82 (74–90) 91 (84–98)
Sub-type of pancreatitis CMFP (n = 9) 85 (80–91) <0.01 90 (84–96) 0.16
AIP (n = 10) 76 (69–83) 96 (93–99)
CT slice thickness ≤3 mm (n = 12) 80 (74–86) <0.01 92 (92–92) <0.01
>3 mm (n = 7) 81 (74–89) 96 (93–99)
CT multi-phase Yes (n = 16) 83 (78–87) 0.76 93 (90–97) 0.05
No (n = 3) 61 (46–77) 96 (92–100)
Multiple CT channel Yes (n = 13) 81 (75–87) 0.01 91 (87–96) <0.01
No (n = 6) 81 (72–90) 97 (94–100)
Image reviewer Single reviewer (n = 5) 75 (64–86) 0.12 95 (92–98) 0.02
Multiple reviewers (n = 14) 83 (77–88) 93 (89–97)
Blinded to read Clear (n = 11) 77 (70–84) <0.01 95 (91–98) 0.06
Unclear (n = 8) 84 (78–90) 92 (87–98)
Reference criteria for MFP Single criteria (n = 11) 83 (77–89) 0.02 95 (91–98) 0.05
Combine ≥2 criteria (n = 8) 77 (70–85) 93 (88–98)
Reference criteria for PDAC Histologic criteria (n = 15) 82 (77–87) 0.18 93 (89–96) 0.02
Histologic and clinical criteria (n = 4) 74 (62–86) 97 (93–100)
AIP: Autoimmune pancreatitis; CMFP: Chronic mass-form pancreatitis; CI: Confidence index; CECT: Contrast-enhanced computed tomography; MFP: Mass-form pancreatitis; PDAC: Pancreatic ductal adenocarcinoma.


Patients with MFP or PDAC present similar clinical symptoms but require different treatment approaches, and have different survival outcomes. Therefore, it is necessary for clinicians to differentiate between MFP and PDAC. The current systematic review found that the summary diagnostic accuracy of CEUS was comparable to that of CECT for the differentiation of MFP from PDAC (RDOR = 1.26, 95% CI [0.42–3.83], P > 0.05).

The diagnostic sensitivity of CEUS for MFP was consistent with two recent studies that reported that CEUS had a higher sensitivity than CECT (80% vs. 40%, P = 0.003; and 82% vs. 68%, P = 0.253, respectively).[7,8] Due to its ability to provide real-time continuous visualization of the blood perfusion of the pancreas and its masses, CEUS has been increasingly used for the diagnosis of pancreatic lesions.[4,5,14] Based on this unique imaging characteristic, MFP on CEUS showed slight iso/hyperenhancement of contrast compared to the adjacent normal pancreatic parenchyma.[5,8,9] In contrast to CEUS, CECT could not be used to visualize transient enhancement features due to the fixed time point and the fixed slice thickness. Furthermore, it is not recommended that patients with chronic pancreatitis repeatedly undergo CT examination within a short period of time. Of note, CEUS showed the specific enhancement characteristics of MFPs in patients with the disease. Therefore, CEUS could play a key role in evaluating treatment efficiency.

D’Onofrion et al[5,41] reported that for PDAC, the diagnostic sensitivity of CEUS was comparable to that of CECT (86.47% vs. 83.58%). Therefore, if conventional ultrasound detects a suspected pancreatic mass, CEUS might be recommended to identify the characteristics of PDAC. This meta-analysis also found that CEUS had a slightly higher overall sensitivity than CECT for differentiating PDAC from MFP (95% vs. 94%), which was inconsistent with the results of the most recent studies.[7,8] The various definitions of PDAC in CECT image features might explain the low summary sensitivity in CECT reports.[24,25,35,37] On the other hand, PDAC showed hypoenhancement in all phases on CEUS due to its abundant fibrosis and desmoplasia.[5] The results of the meta-analysis suggest that the imaging characteristics of PDAC on CECT require a relatively unified standard and that CEUS and CECT are mutually complementary for suspected focal pancreatic malignant lesions.

Of the 19 CT reviewed studies, Chari et al[24] reported high specificity (100%) but low sensitivity (52%) for CECT when differentiating MFP from PDAC. The imaging criteria of AIP were diffuse pancreatic enlargement without a low-density mass, pancreatic duct cutoff, or distal pancreatic atrophy, which were inconsistent with the typical imaging features of pancreatitis.[2] Grossiohann et al[7] reported a low sensitivity of CECT (40%) and a high sensitivity of CEUS (80%) based on the characteristics of CMFP with isoechogenic lesions in 49 patients; these results were consistent with the current study and indicated that CEUS was a useful tool for differentiating pancreatitis from PDAC.[4] Both studies reported the diagnostic performance of each CECT feature; however, the evaluated data varied according to the feature analyzed. Therefore, further studies combining multiple imaging features should be conducted to further elucidate the diagnostic performance of these imaging tests.

The meta-regression analysis failed to elucidate the factors that caused the moderate heterogeneity in overall specificity among the CEUS studies (all P > 0.05). To examine the cause of the considerable heterogeneity in the CECT studies, meta-regression was performed, and the results indicated that certain CT parameters (thin-slice and multiple CT channels) affected the heterogeneity in both the sensitivity and specificity of CECT (P ≤ 0.01). Overall, the use of advanced equipment for CECT could yield more information about pancreatic lesions, which might also explain why the studies published after 2015 showed higher overall sensitivity than those published before 2015 (82% vs. 80%, P < 0.01). Although the diagnosis of CMFP had higher sensitivity than that of AIP by CECT, the combination of imaging features and IgG4 serum levels might improve the diagnostic accuracy.[42] Interestingly, we found that a single rater had higher specificity than multiple raters and that blinded reviews showed lower sensitivity than those with unclear information (all P < 0.05). Some CECT-related studies did not provide detailed information about the radiological image analyses, which might have influenced the diagnostic results.[7,23,30,36,38-40] Therefore, it was suggested that in later related studies, the standards for analyses should be included. The included articles had various reference standards for the target patients. Interestingly, the meta-regression results showed that a single reference standard for MFP or PDAC had significantly higher diagnostic performance (all P = 0.020). These results suggested that using a single gold standard for the included patients was beneficial for improving the diagnostic heterogeneity.

This study has some limitations. First, most of the original studies were originally designed to examine the performance of only one of the two modalities, and thus, the effect of the characteristics of different patients could not be avoided. Although we performed a Bayesian network meta-analysis to indirectly compare the overall diagnostic performance of the two modalities, more original studies should be designed to directly compare the two diagnostic tests. Second, substantial study heterogeneity was found in terms of both the overall sensitivity and specificity of CECT/CEUS. Therefore, we used a meta-regression analysis to explore the reason for this study heterogeneity, and the results suggested that advanced imaging modality and a standardized imaging analysis and reporting system would help to improve the diagnostic heterogeneity. Third, AIP was a unique type of chronic pancreatitis related to an autoimmune mechanism,[43] and the meta-analysis failed to compare the performance of CEUS and CECT in the differential diagnosis of AIP and PDAC because of the limited number of studies.[11] Additional studies are needed to evaluate the diagnostic accuracy of CEUS in the differentiation of AIP from PDAC. Finally, the study excluded certain types of literature, such as letters, conference abstracts, and unpublished data, which might raise concerns over publication bias. However, the data from these trials could not be accurately extracted for further meta-analysis.

In conclusion, both CEUS and CECT have outstanding sensitivity for the diagnosis of PDAC; however, they have suboptimal sensitivity for the diagnosis of MFP. This results of this work indicated that the overall diagnostic performance of CEUS for pancreatic tumors is similar to that of CECT. Taking into account the characteristics of the two imaging modalities and the patient's condition, CEUS and CECT should be considered mutually complementary diagnostic tools for suspected focal pancreatic lesions.


Author thank Prof. Jinhui Tian from Lanzhou University for his statistical assistance in this meta-analysis.


The study was supported by grants from the Science and Technology Department of Sichuan Province (No. 2018FZ0044) and the National Natural Science Foundation of China (No. 81571697).

Conflicts of interest



1. Huang SW, Mao EQ, Wang HS, Zhao B, Chen Y, Qu HP, et al. Clinical characteristics of 5375 cases of acute pancreatitis from a single Chinese center, 1996-2015. Chin Med J 2019;132:1233–1236. doi: 10.1097/CM9.0000000000000208.
2. Beyer G, Habtezion A, Werner J, Lerch MM, Mayerle J. Chronic pancreatitis. Lancet 2020;396:499–512. doi: 10.1016/S0140-6736(20)31318-0.
3. Schima W, Böhm G, Rösch CS, Klaus A, Függer R, Kopf H. Mass-forming pancreatitis versus pancreatic ductal adenocarcinoma: CT and MR imaging for differentiation. Cancer Imaging 2020;20:52–63. doi: 10.1186/s40644-020-00324-z.
4. Sidhu PS, Cantisani V, Dietrich CF, Gilja OH, Saftoiu A, Bartels E, et al. The EFSUMB guidelines and recommendations for the clinical practice of contrast-enhanced ultrasound (CEUS) in non-hepatic applications: update 2017 (short version). Ultraschall Med 2018;39:154–180. doi: 10.1055/s-0044-101254.
5. D’Onofrio M, de Sio I, Mirk P, Vidili G, Bertolotto M, Cantisani V, et al. SIUMB recommendations for focal pancreatic lesions. J Ultrasound 2020;23:599–606. doi: 10.1007/s40477-020-00522-2.
6. D’Onofrio M, Zamboni G, Tognolini A, Malago R, Faccioli N, Frulloni L, et al. Mass-forming pancreatitis: value of contrast-enhanced ultrasonography. World J Gastroenterol 2006;12:4181–4184. doi: 10.3748/wjg.v12.i26.4181.
7. Grossjohann HS, Rappeport ED, Jensen C, Svendsen LB, Hillings⊘ JG, Hansen CP, et al. Usefulness of contrast-enhanced transabdominal ultrasound for tumor classification and tumor staging in the pancreatic head. Scand J Gastroenterol 2010;45:917–924. doi: 10.3109/00365521003702718.
8. Fan Z, Li Y, Yan K, Wu W, Yin S, Yang W, et al. Application of contrast-enhanced ultrasound in the diagnosis of solid pancreatic lesions - a comparison of conventional ultrasound and contrast-enhanced CT. Eur J Radiol 2013;82:1385–1390. doi: 10.1016/j.ejrad.2013.04.016.
9. Wang Y, Yan K, Fan Z, Ding K, Yin S, Dai Y, et al. Clinical value of contrast-enhanced ultrasound enhancement patterns for differentiating focal pancreatitis from pancreatic carcinoma: a comparison study with conventional ultrasound. J Ultrasound Med 2018;37:551–559. doi: 10.1002/jum.14363.
10. Xie X, Xu E, Xu H, Xu Z, Liu G, Zheng Y, et al. Role of contrast-enhanced ultrasound in the differentiation of solid focal lesions of pancreas. Acta Acad Med Sin 2008;30:35–39. doi: 10.3321/j.issn:1000-503X.2008.01.008.
11. Yuan H, Ding H, Liu L, Wen J, Cao J, Wang W. Value of contrast-enhanced ultrasound in solid neoplasms of pancreas. Shanghai Med Imaging 2010;19:102–104.
12. Xu H, Feng L. Diagnostic value of real-time microbubbles ultrasonography in pancreatic mass lesions. Ultrasound Imaging 2011;19:650–654. doi: 10.3969/j.issn.1005-5185.2011.09.003.
13. Li X, Zhang R, Xiuli Z, Zhang W, Zhang C. Value of contrast- enhanced ultrasound in differential diagnosis of solid lesions of pancreas (SLP). J Qihar Univ Med 2013;34:545–546. doi: 10.3969/j.issn.1002-1256.2013.04.053.
14. Gardner TB, Adler DG, Forsmark CE, Sauer BG, Taylor JR, Whitcomb DC. ACG clinical guideline: chronic pancreatitis. Am J Gastroenterol 2020;115:322–339. doi: 10.14309/ajg.0000000000000535.
15. Liu SL, Li S, Guo YT, Zhou YP, Zhang ZD, Li S, et al. Establishment and application of an artificial intelligence diagnosis system for pancreatic cancer with a faster region-based convolutional neural network. Chin Med J 2019;132:2795–2803. doi: 10.1097/CM9.0000000000000544.
16. D’Onofrio M, Biagioli E, Gerardi C, Canestrini S, Rulli E, Crosara S, et al. Diagnostic performance of contrast-enhanced ultrasound (CEUS) and contrast-enhanced endoscopic ultrasound (ECEUS) for the differentiation of pancreatic lesions: a systematic review and meta-analysis. Ultraschall Med 2014;35:515–521. doi: 10.1055/s-0034-1385068.
17. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–2012. doi: 10.1001/jama.283.15.2008.
18. Whiting PF, Rutjes AWS, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529–536. doi: 10.7326/0003-4819-155-8-201110180-00009.
19. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–560. doi: 10.1136/bmj.327.7414.557.
20. Kim KW, Lee J, Choi SH, Huh J, Park SH. Systematic review and meta-analysis of studies evaluating diagnostic test accuracy: a practical review for clinical researchers-part I. General guidance and tips. Korean J Radiol 2015;16:1175–1187. doi: 10.3348/kjr.2015.16.6.1175.
21. Lumley T. Network meta-analysis for indirect treatment comparisons. Stat Med 2002;21:2313–2324. doi: 10.1002/sim.1201.
22. Wu J, Ge L, Zhang J, Li J, Lin X, Tian J. Comparation of accuracy of different diagnostic tests: an introduction of network meta-analysis methods. Chin J Evid Based Med 2017;17:987–992. doi: 10.7507/1672-2531.201706041.
23. Brimienė V, Brimas G, Strupas K. Differential diagnosis between chronic pancreatitis and pancreatic cancer: a prospective study of 156 patients. Medicina 2011;47:154–162.
24. Chari ST, Takahashi N, Levy MJ, Smyrk TC, Clain JE, Pearson RK, et al. A diagnostic strategy to distinguish autoimmune pancreatitis from pancreatic cancer. Clin Gastroenterol Hepatol 2009;7:1097–1103. doi: 10.1016/j.cgh.2009.04.020.
25. Linning E, Yan X, Zhifeng W, Li L, Na Z, Hao Y, et al. Differentiation of focal-type autoimmune pancreatitis from pancreatic ductal adenocarcinoma using radiomics based on multiphasic computed tomography. J Comput Assist Tomogr 2020;44:511–518. doi: 10.1097/RCT.0000000000001049.
26. Furuhashi N, Suzuki K, Sakurai Y, Ikeda M, Kawai Y, Naganawa S. Differentiation of focal-type autoimmune pancreatitis from pancreatic carcinoma: assessment by multiphase contrast-enhanced CT. Eur Radiol 2015;25:1366–1374. doi: 10.1007/s00330-014-3512-3.
27. Kawai Y, Suzuki K, Itoh S, Takada A, Mori Y, Naganawa S. Autoimmune pancreatitis: assessment of the enhanced duct sign on multiphase contrast-enhanced computed tomography. Eur J Radiol 2012;81:3055–3060. doi: 10.1016/j.ejrad.2012.04.023.
28. Lee S, Kim JH, Kim SY, Byun JH, Kim HJ, Kim MH, et al. Comparison of diagnostic performance between CT and MRI in differentiating non-diffuse-type autoimmune pancreatitis from pancreatic ductal adenocarcinoma. Eur Radiol 2018;28:5267–5274. doi: 10.1007/s00330-018-5565-1.
29. Muhi A, Ichikawa T, Motosugi U, Sou H, Sano K, Tsukamoto T, et al. Mass-forming autoimmune pancreatitis and pancreatic carcinoma: differential diagnosis on the basis of computed tomography and magnetic resonance cholangiopancreatography, and diffusion-weighted imaging findings. J Magn Reson Imaging 2012;35:827–836. doi: 10.1002/jmri.22881.
30. Naitoh I, Nakazawa T, Hayashi K, Okumura F, Miyabe K, Shimizu S, et al. Clinical differences between mass-forming autoimmune pancreatitis and pancreatic cancer. Scand J Gastroenterol 2012;47:607–613. doi: 10.3109/00365521.2012.667147.
31. Ren S, Chen X, Cui W, Guo K, Chen R, Zhang H, et al. Differentiation of chronic mass-forming pancreatitis from pancreatic ductal adenocarcinoma using contrast-enhanced computed tomography. Cancer Manag Res 2019;11:7857–7866. doi: 10.2147/CMAR.S217033.
32. Sun GF, Zuo CJ, Shao CW, Wang JH, Zhang J. Focal autoimmune pancreatitis: radiological characteristics help to distinguish from pancreatic cancer. World J Gastroenterol 2013;19:3634–3641. doi: 10.3748/wjg.v19.i23.3634.
33. Yamada Y, Mori H, Matsumoto S, Kiyosue H, Hori Y, Hongo N. Pancreatic adenocarcinoma versus chronic pancreatitis: differentiation with triple-phase helical CT. Abdom Imaging 2010;35:163–171. doi: 10.1007/s00261-009-9579-7.
34. Yin Q, Zou X, Zai X, Wu Z, Wu Q, Jiang X, et al. Pancreatic ductal adenocarcinoma and chronic mass-forming pancreatitis: differentiation with dual-energy MDCT in spectral imaging mode. Eur J Radiol 2015;84:2470–2476. doi: 10.1016/j.ejrad.2015.09.023.
35. Zaheer A, Singh VK, Akshintala VS, Kawamoto S, Tsai SD, Gage KL, et al. Differentiating autoimmune pancreatitis from pancreatic adenocarcinoma using dual-phase computed tomography. J Comput Assist Tomogr 2014;38:146–152. doi: 10.1097/RCT.0b013e3182a9a431.
36. Guo J. Differential diagnosis of mass pancreatitis and pancreatic cancer with enhanced CT scan. Med Equip 2018;31:30–31. doi: 10.3969/j.issn.1002-2376.2018.04.020.
37. Luo Y, Yu T, Wang Y, Qiu Y, Yu B, Duan Y, et al. Differential diagnosis between pancreatic cancer and chronic pancreatitis by enhanced scanning of both CT and MRI. Chin J Magn Reson Imaging 2011;2:42–46. doi: 10.3969/j.issn.1674-8034.2011.01.011.
38. Lv Q, Wang P-J, Shao Z-H, Ni J, Wang G, CT. enhancement scanning and diffusion-weighted magnetic resonance imaging for differential diagnosisbetween chronic mass-forming pancreatitis and pancreatic carcinoma. Acad J Second Mil Med Univ 2013;34:974–979. doi: 10.3724/SP.J.1008.2013.00974.
39. Wang X, Gu D, Sang J, Zhuang C. The value of CT enhanced scan and Mr DWI sequence in the differential diagnosis of mass pancreatitis and pancreatic cancer. Imaging Res Med Appl 2018;2:142–143. doi: 10.3969/j.issn.2096-3807.2018.12.090.
40. Zhu X, Lu JP. Differential diagnosis of autoimmune pancreatitis and pancreatic cancer in small imaging field of vision pancreatic artery CT enhanced angiography. Mil Med J Southeast China 2018;20:506–510. doi: 10.3969/j.issn.1672-271X.2018.05.013.
41. D’Onofrio M, Crosara S, Signorini M, De Robertis R, Canestrini S, Principe F, et al. Comparison between CT and CEUS in the diagnosis of pancreatic adenocarcinoma. Ultraschall Med 2013;34:377–381. doi: 10.1055/s-0032-1325324.
42. Khandelwal A, Inoue D, Takahashi N. Autoimmune pancreatitis: an update. Abdom Radiol 2020;45:1359–1370. doi: 10.1007/s00261-019-02275-x.
43. Yoshikawa T, Watanabe T, Kamata K, Hara A, Minaga K, Kudo M. Intestinal dysbiosis and autoimmune pancreatitis. Front Immunol 2021;12:621532. doi: 10.3389/fimmu.2021.621532.

Pancreatitis; Pancreatic neoplasms; Ultrasonography; Tomography, X-ray computed; Meta-analysis

Supplemental Digital Content

Copyright © 2022 The Chinese Medical Association, produced by Wolters Kluwer, Inc. under the CC-BY-NC-ND license.