Acute pancreatitis and computed tomography: Interest of portal venous phase alone in the initial phase

Abstract

Background

There are no guidelines in the literature for the use of a computed tomography (CT) protocol in the initial phase of acute pancreatitis (AP).

Purpose

To evaluate the contribution of single portal venous phase CT compared to triple-phase CT protocol, performed in the initial phase of AP for severity assessment.

Material and Methods

In this retrospective study, a total of 175 patients with acute pancreatitis who underwent initial triple-phase CT protocol (non-contrast, arterial phase, and portal venous phase) between D3 and D7 after the onset of symptoms were included. Analysis of AP severity and complications was independently assessed by two readers using three validated CT severity scores (CTSI, mCTSI, EPIC). All scores were applied to the triple-phase CT protocol and compared to the single portal venous phase. Inter-observer analyses were also performed.

Results

No significant difference whatever the severity score was observed after analysis of the single portal venous phase compared with the triple-phase CT protocol (interstitial edematous pancreatitis: CTSI: 2 vs. 2, mCTSI: 2 vs. 2, EPIC: 1 vs. 1; necrotizing pancreatitis: CTSI: 6 vs. 6, mCTSI: 8 vs. 8, EPIC: 5 vs. 5). Inter-observer agreement was excellent (ICC = 0.96–0.99), whatever the severity score.

Conclusion

A triple-phase CT protocol performed at the initial phase of AP was no better than a single portal venous for assessing the severity of complications and could lead to a 63% reduction in irradiation.

Keywords

Computed tomography single portal venous phase CTSI radiation acute pancreatitis

Introduction

Acute pancreatitis (AP) is a common inflammatory disease of the pancreas, with mortality in the range of 5%–20% in the necrotizing form (1,2).

The revised Atlanta classification published in 2012 (1) distinguishes two forms of acute pancreatitis: interstitial edematous pancreatitis (IEP) and necrotizing pancreatitis (NP), which is classified as intra-pancreatic, peri-pancreatic, or mixed necrosis. Abdominal-pelvic computed tomography (CT) with contrast is the reference imaging technique for detecting pancreatic and/or peripancreatic necrosis and stratifying its severity (3). Necrosis is best visualized when the scan is performed later, at least 72 h after the onset of symptoms, so as not to underestimate the necrosis requiring a second scan at a distance (4,5).

A widespread practice consists of performing a CT scan with a systematic triple-phase CT protocol (non-contrast, arterial phase, and portal venous phase) in the initial phase to assess the complications of acute pancreatitis. The revised Atlanta classification of 2012 (1) does not specify the protocol to be used. An American recommendation of 2013 (6) advocates the single portal venous phase during follow-up, and acquisition at arterial pancreatic phase and/or portal venous phase during the first examination.

However, in the literature or in consensus conferences, the added value of this protocol compared with the single portal venous phase has not been demonstrated. Some studies (7 –10) have investigated the potential contribution of portal venous phase in the CT assessment of AP severity using the CTSI score, but with small numbers and without comparing the single portal venous phase versus the triple-phase CT protocol.

Radiology irradiation is a public health problem, with 3% of human cancers thought to be due to medical ionizing radiation (11). Simplifying this protocol would reduce patient irradiation, especially since, in the event of poor progression, the number of examinations is likely to multiply during follow-up, particularly in young patients.

Therefore, the aim of the present study was to assess the relevance of the standard triple-phase CT protocol compared with the single portal venous phase performed in the initial phase of AP.

Material and Methods

Patient population

This is a retrospective, single-center study approved by our institution's ethics committee. A total of 469 patients hospitalized for AP between September 2019 and March 2021 were identified in the registry of information systems medicalization program in our university hospital.

The inclusion criteria were as follows: patients admitted with AP defined by the revised Atlanta classification (1); and CT scan with a triple-phase protocol (non-contrast, arterial phase, and portal venous phase) at the initial phase (after 72 h and less than D7 from the onset of AP symptoms).

The exclusion criteria were as follows: pregnancy in progress; age <18 years; and no imaging or imaging not performed in our university hospital.

A total of 175 patients were finally included (Fig. 1).

Fig. 1.

Flow chart of the study.

CT acquisition

All acquisitions were performed on a 256-detector row CT scanner (GE Healthcare, Milwaukee, WI, USA).

Examinations were carried out according to a triple-phase CT protocol including an abdominopelvic acquisition, non-contrast, then an injection of iodinated contrast medium at arterial pancreatic phase at 45–50 s centered on the pancreas, then to the portal venous phase at 70 s (pitch = 1.38; gantry rotation time = 0.6 s; 120 kV; 80–160 mA).

An iodinated contrast medium (Iomeprol, Iomeron 300, or Iopromide, Ultravist 300) was injected at a dose of 1.5 mL/kg, at a flow rate of 3 mL/s.

Image analysis

Images were analyzed using Advantage Windows Console AW 4.7 postprocessing software (GE Healthcare, Milwaukee, WI, USA) with a native slice thickness of 0.625 mm, in abdominal windowing (W 400, L 40). No display filters were used for image analysis with multiplanar reconstructions (MPR).

Two readers, a senior radiologist (R1) and a radiology intern (R2) with 15 and 4 years of imaging experience, respectively, independently analyzed the two protocols, including a minimum delay of 14 days between the analysis of the triple-phase protocol and the single portal venous portal phase and were blinded to clinico-biological data and imaging reports.

Severity score calculation

Each reader was provided with a reading grid containing the CTSI, mCTSI, and EPIC score criteria.

The complication rate detected by triphasic examination compared with single portal venous phase was assessed using the CTSI severity score at the initial phase.

Each score was applied in the analysis by the two assessors.

Radiation dose measurements

The CT dose index (CTDIvol) and dose-length product (DLP) were recorded for each protocol. Our examination protocol provided for separate recordings of CTDIvol and DLP values.

The effective radiation dose (E) was calculated using the following formula: E = DLP * k, where k is a conversion coefficient for the abdomen and pelvis in adults. The dose was 0.015. mSv/(mGy*cm) (12).

Statistical analysis

RStudio software version 1.3. 1093 was used for statistical analysis.

Categorical data were reported as numbers and percentages. The normality test used was the Shapiro–Wilk test. In the case of normal distribution, means and standard deviations were used to describe continuous variables. Otherwise, results were described using the median and interquartile range (IQR).

Means were compared using the Student's t-test for normally distributed data. The sum of the CTSI, mCTSI, and EPIC scores was compared between analyses performed during the portal venous phase and during the triple-phase protocol and by both readers using the Wilcoxon test for non-normally distributed data.

Inter-observer agreement between the two readers for quantitative measurements was assessed using the intraclass correlation coefficient (ICC). The ICC results were interpreted as follows: 0–0.20 = slight agreement; 0.21–0.40 = fair agreement; 0.41–0.60 = moderate agreement; 0.61–0.80 = substantial agreement; 0.81–1.00 = excellent agreement.

Two-tailed P values <0.05 were considered statistically significant for all tests.

Results

Patient population

The characteristics of the general population are reported in Table 1. We included 175 patients (97 men, 78 women; age range = 21–97 years; mean age = 54 ± 18 years). Of the patients, 139/175 (79%) had interstitial edematous pancreatitis and 36/175 (21%) had acute necrotizing pancreatitis according to the Atlanta classification (1).

Table 1.

General characteristics of the study population.

	Interstitial edematous pancreatitis (n = 139, 79%)	Necrotizing pancreatitis (n = 36, 21%)	Total patients (n = 175)
Male	76 (55)	21 (58)	97 (55)
Age (years)	54 ± 15	54 ± 14	54 ± 15
Etiology
Biliary	48 (35)	9 (25)	57 (33)
Alcohol	31 (22)	18 (50)	49 (28)
Indeterminate	27 (19)	4 (10)	31 (18)
Post ERCP	7 (5)	1 (3)	8 (4)
Postoperative	3 (2)	1 (3)	4 (2)
Drug induced	7 (5)	1 (3)	8 (4)
Post COVID-19	3 (2)	0 (0)	3 (2)
Pancreatic malignant tumor	2 (2)	1 (3)	3 (2)
Other causes (food poisoning, hypertriglyceridemia, IPMN, CDHP, autoimmune)	11 (8)	1 (3)	12 (7)
Pancreatic and extra-pancreatic analyses
CTSI score
0	49 (35)	0 (0)	49 (28)
1	10 (7)	0 (0)	10 (6)
2	46 (33)	0 (0)	46 (26)
3	18 (13)	0 (0)	18 (10)
4	16 (12)	6 (16)	22 (13)
5	0 (0)	3 (8)	3 (2)
6	0 (0)	13 (37)	13 (7)
7	0 (0)	0 (0)	0 (0)
8	0 (0)	3 (8)	3 (2)
9	0 (0)	0 (0)	0 (0)
10	0 (0)	11 (31)	11 (6)
Necrosis
Necrotic parenchyma
<30	0 (0)	15 (42)	15 (9)
30–50	0 (0)	2 (6)	2 (11)
>50	0 (0)	11 (30)	11 (6)
ANC	0 (0)	8 (22)	8 (5)
Other complications
Pleural effusion	19 (14)	27 (75)	46 (26)
Ascites	41 (30)	27 (75)	68 (39)
Venous thrombosis	1 (1)	16 (44)	17 (10)
Pseudoaneurysm	0 (0)	0 (0)	0 (0)
Infected pancreatic necrosis	0 (0)	5 (14)	5 (3)
Death	0 (0)	5 (14)	5 (3)

Values are given as n (%) or mean ± SD.

ANC, acute necrotic collection; CDHP, cystic dystrophy of heterotopic pancreas; ERCP, endoscopic retrograde cholangiopancreatography; IPMN, intraductal papillary mucinous neoplasm.

The main causes of AP were lithiasis (n = 57, 33%) and alcohol (n = 49, 28%).

CT findings

A CT analysis of R1's pancreatic and extra-pancreatic CT complications was carried out (Table 1).

CTSI scores of 0 and 2 were most common in patients with IEP. CTSI scores of 6 and 10 were most common in patients with NP.

The most frequent extra-pancreatic signs were pleural effusions (14% and 75% in IEP and NP, respectively) and ascites (30% and 75% in IEP and NP, respectively). The most frequent vascular complication was venous thrombosis (1% and 44% in IEP and NP, respectively). No pseudoaneurysm was detected in the observed population.

Comparison of CT severity score

The comparative results for the sum of the CTSI, mCTSI, and EPIC scores as a function of CT protocols (single portal venous phase vs. triple-phase protocol) for each reader and for both readers are presented in Tables 2–4.

Table 2.

Comparison of three different CT severity scores for IEP and NP of reader 1.

	IEP (n = 139, 79%)	NP (n = 36, 21%)	Total patients (n = 175)
CTSI score	2 (0–4.5)	6 (4–10)	2 (0–6)
mCTSI	2 (0–6)	8 (6–10)	2 (0–7)
EPIC score	1 (0–3)	5 (3.8–6.3)	1 (0–5)

Values are given as median (IQR).

CT, computed tomography; IEP, interstitial edematous pancreatitis; NP, necrotizing pancreatitis.

Table 3.

Comparison of three different CT severity scores for IEP and NP of reader 2.

	IEP (n = 139, 79%)	NP (n = 36, 21%)	Total patients (n = 175)
CTSI score	2 (0–5)	6 (3–10)	2 (0–6)
mCTSI score	2 (0–6)	8 (6–10)	2 (0–7)
EPIC score	1 (0–3)	5 (4–6)	2 (0–6)

Values are given as median (IQR).

CT, computed tomography; IEP, interstitial edematous pancreatitis; NP, necrotizing pancreatitis.

Table 4.

Comparison of three scanographic severity scores for IEP and NP by the two readers.

	Portal venous phase		P	Triple-phase protocol		P
	IEP (n = 139)	NP (n = 36)	P	IEP (n = 139)	NP (n = 36)	P
CTSI score	2 (0–5)	6 (3–10)	<0.001	2 (0–5)	6 (3–10)	<0.001
mCTSI score	2 (0–6)	8 (6–10)	<0.001	2 (0–6)	8 (6–10)	<0.001
EPIC score	1 (0–3)	5 (4–6)	<0.001	1 (0–3)	5 (4–6)	<0.001

Values are given as median (IQR).

CT, computed tomography; IEP, interstitial edematous pancreatitis; NP, necrotizing pancreatitis.

The average sums of the CTSI, mCTSI, and EPIC scores for both evaluators in the single portal venous phase were: 2 for IEP vs. 6 for NP using the CTSI score; 2 for IEP vs. 8 for NP using the mCTSI score; 1 for IEP vs. 5 for NP using the EPIC score. These values were similar in the triple-phase CT protocol. No significant differences of scores were found between the single portal venous phase and the triple-phase protocol for both readers and in both types of pancreatitis (Table 4, Fig. 2).

Fig. 2.

Comparison of three computed tomography severity scores assessed by the two readers. Columns represent median values.

Scores were significantly higher in NP than in IEP for both readers, irrespective of the score used and the CT protocol (Table 4, in Figs. 3–5).

Fig. 3.

Necrotizing pancreatitis. Triple-phase computed tomography protocol (non-contrast, arterial phase, and portal venous phase): (A) non-contrast: ill-defined infiltration of peripancreatic fat (arrow); (B) arterial phase: non-enhanced focal areas (star) associated with pancreatic necrosis and peripancreatic fat infiltration; (C) portal venous phase (dotted arrow).

Fig. 4.

Splenic thrombosis. (A) Arterial phase: splenic vein difficult to identify (white arrow). (B) Portal venous phase: splenic vein thrombosis only visible at portal venous phase (dotted arrow).

Fig. 5.

Interstitial edematous pancreatitis of biliary origin with a CTSI score of 2, a CTSI m score of 2. (A) CT non-contrast: infiltration of peripancreatic fat (arrow). (B) CT arterial phase (star). (C) CT portal venous phase (dotted arrow). CT, computed tomography.

Inter-observer agreement

The two CT protocols showed excellent agreement for interstitial edematous and necrotizing pancreatitis (the ICC value was 0.97 during the single portal venous phase, identical during the triple-phase protocol with the CTSI score for IEP, and 0.99 for NP). Detailed ICC results are presented in Table 5.

Table 5.

ICCs between the two readers in IEP and NP according to the two scanographic protocols and scanographic severity scores.

	IEP				NP
	Single portal venous phase	P	Triple-phase protocol	P	Single portal venous phase	P	Triple-phase protocol	P
CTSI score	0.971	<0.001	0.971	<0.001	0.989	<0.001	0.989	<0.001
mCTSI score	0.988	<0.001	0.988	<0.001	0.958	<0.001	0.958	<0.001
EPIC score	0.986	<0.001	0.986	<0.001	0.984	<0.001	0.984	<0.001

CT, computed tomography; ICC, intraclass correlation; IEP, interstitial edematous pancreatitis; NP, necrotizing pancreatitis.

Radiation dose

Median CTDIvol and DLP calculated for the single portal venous phase were significantly lower than for the triple-phase protocol (DLP: 689 mGy*cm [IQR ± 377] vs. 1879 mGy*cm [IQR ± 1167]; CTDI vol: 14 mGy [IQR ± 7] vs. 38 mGy [IQR ± 22]; P < 0.001).

Median effective doses calculated for the single portal venous phase and the triple-phase protocol were 10 mSv (IQR ± 5) and 28 mSv (IQR ± 17), respectively (P < 0.001) (Table 6).

Table 6.

Comparison of DLP, CTDIvol, and effective dose according to the single portal venous phase or the triple-phase protocol.

	Single portal venous phase (n = 175)	Triple-phase protocol (n = 175)	Ratio (single portal venous phase /triple-phase protocol) (%)	Potential saving (%)	P
DLP	689 (312–1066)	1879 (712–3046)	37	−63	<0.001
CTDIvol	14 (7–21)	38 (16–60)	37	−63	<0.001
Effective dose	10 (5–15)	28 (11–45)	37	−63	<0.001

Values are given as median (IQR).

CTDI, computed tomography volume-weighted dose index; DLP, dose-length product.

Discussion

The main finding of our study is the absence of any significant difference in the CTSI scanographic severity score after analysis of the single portal venous phase compared with the triple-phase protocol; this is also found in the other two scanographic severity scores (mCTSI and EPIC) (Figs. 2 and 5).

These results are in line with the observations of Avanesov et al. (10), who retrospectively included 102 patients, comparing the single portal venous phase with a dual-phase protocol comprising arterial phase and portal venous phase in the initial phase of AP. In their study, no significant differences were found in the various severity scores, with results close to those of our study for IEP and NP.

The inter-observer agreement in that study, whatever the score, was also excellent, apart from the ICC value for IEP, which was in the range of 0.77–0.79 compared with 0.97 in our analysis. Some authors have reported inter-observer agreement of the initial AP CTSI score as moderate (13,14) or even poor, with an ICC estimated at 0.34 (15).

Our good results may be explained by the establishment of a “consensus” on definitions between readers before the start of the study, notably on the distinction between peripancreatic fat infiltration and a small collection (16), and the difference between a pure fluid collection and a necrotic collection (17) (any acute peripancreatic fluid collection) with density <15 HU and acute necrotic collection with high density >15 HU (13,18).

In a similar study to Avanesov et al. (10), Kwon et al. (7) and Winchmann et al. (9) in 52 and 66 patients with AP, respectively, comparing only the CTSI and mCTSI severity scores, without taking into account the EPIC score, obtained equivalent results while considerably reducing irradiation. In the study by Avanesov et al. of 102 patients, a 36% reduction in irradiation was achieved (10). In our study, performing the portal venous phase alone made it possible to evaluate the scanographic severity of AP. Non-contrast and arterial phase acquisitions proved of little value. This would reduce the radiation dose by 63%.

In our population, the venous thromboses commonly found were perfectly detected in approximately 10% of patients with AP, most often during NP (up to 44%), compared with 17% in a meta-analysis of 44 studies (19). In our work, all venous complications were sufficiently detected during the portal venous phase (Fig. 4).

Non-contrast acquisition is used primarily to detect lithiasis. In our study, over 80% of patients underwent ultrasonography, with better detection of lithiasis than CT, but only 59% of lithiasis detected by ultrasonography was detected by CT. This is an argument in favor of abandoning the non-contrast acquisition.

The potential contribution of perfusion and spectral imaging is currently being evaluated. Yadav et al. (20) have shown that perfusion CT within the first three days of severe AP is a reliable predictor of parenchymal necrosis. Dual-energy CT imaging (21) enabled better differentiation of peripancreatic collections or detection of lithiasis (22).

The present study has some limitations. These are related to recruitment, as we included patients from a single center and retrospectively. In our study, no cases of pseudoaneurysm were detected due to a later development, although this is the arterial complication with an estimated incidence in the range of 0.01–0.2% (23). This may be explained by the rarity of this complication (24), and the delay in detection, which occurs 3–5 weeks after the onset of AP. Cases of early onset have nevertheless been reported a few days after the onset of pancreatitis (25). In cases of suspected acute hemorrhage or deglobulation, arterial phase is indicated. After D7, we remain cautious about performing an arterial phase, which remains indicated, as well as in cases of suspected acute hemorrhage.

In conclusion, the contribution of the single portal venous phase compared with the triple-phase protocol carried out in the initial phase between D3 and D7 of the onset of AP symptoms shows no significant difference for severity assessment. The average radiation dose could be reduced by 63% with this simplified protocol in the initial phase of AP.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Wilfried Ansel-Wallois

Roger Bouzerar

References

Banks

Bollen

Dervenis

, et al. Classification of acute pancreatitis 2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013;62:102–111.

Whitcomb

. Acute pancreatitis. N Engl J Med 2006;9:2142-2150.

Thoeni

. The revised Atlanta classification of acute pancreatitis: its importance for the radiologist and its effect on treatment. Radiology 2012;262:751–764.

Balthazar

. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002;223:603–613.

Bollen

Besselink

MGH

van Santvoort

, et al. Toward an update of the Atlanta classification on acute pancreatitis: review of new and abandoned terms. Pancreas 2007;35:107–113.

Working Group IAP/APA Acute Pancreatitis Guidelines.

IAP/AP A evidence-based guidelines for the management of acute pancreatitis. Pancreatology 2013;13:e1–e15.

Kwon

Park

Kim

, et al. Multidetector row computed tomography of acute pancreatitis: utility of single portal phase CT scan in short-term follow up. Eur J Radiol 2012;81:1728–1734.

Noda

Goshima

Fujimoto

, et al. Utility of the portal venous phase for diagnosing pancreatic necrosis in acute pancreatitis using the CT severity index. Abdom Radiol 2018;43:3035–3042.

Wichmann

Majenka

Beeres

, et al. Single- portal-phase low-tube-voltage dual-energy CT for short-term follow-up of acute pancreatitis: evaluation of CT severity index, interobserver agreement and radiation dose. Eur Radiol 2014;24:2927–2935.

10.

Avanesov

Weinrich

Kraus

, et al. MDCT Of acute pancreatitis: intraindividual comparison of single-phase versus dual-phase MDCT for initial assessment of acute pancreatitis using different CT scoring systems. Eur J Radiol 2016;85:2014–2022.

11.

Sodickson

Baeyens

Andriole

, et al. Recurrent CT, cumulative radiation exposure, and associated radiation-induced cancer risks from CT of adults. Radiology 2009;251:175–184.

12.

European Commission, Directorate-general for Energy and Transport, Health Protection Agency, Centre for Radiation C and EH (Great European guidance on estimating population doses from medical x-ray procedures. Luxembourg: Publications Office, 2008.

13.

Balthazar

. Staging of acute pancreatitis. Radiol Clin N Am 2002;40:1199–1209.

14.

Sternby

Verdonk

Aguilar

, et al. Significant inter-observer variation in the diagnosis of extrapancreatic necrosis and type of pancreatic collections in acute pancreatitis – an international multicenter evaluation of the revised Atlanta classification. Pancreatology 2016;16:791–797.

15.

Badat

Millet

Corno

, et al. Revised Atlanta classification for CT pancreatic and peripancreatic collections in the first month of acute pancreatitis: interobserver agreement. Eur Radiol 2019;29:2302–2310.

16.

Bollen

. Acute pancreatitis: international classification and nomenclature. Clin Radiol 2016;71:121–133.

17.

Morgan

Baron

Smith

, et al. Pancreatic fluid collections prior to intervention: evaluation with MR imaging compared with CT and US. Radiology 1997;203:773–778.

18.

Chevalier

Baudin

Coco

, et al. Chapter 31: acute pancreatitis. In: Medical imaging of the abdomen. Paris: Lavoisier, 2010:484–500.

19.

White

Baum

Buranasiri

. Aneurysms secondary to pancreatitis. AJR Am J Roentgenol 1976;127:393–396.

20.

Yadav

Sharma

Kandasamy

, et al.

Perfusion CT: can it predict the development of pancreatic necrosis in early stage of severe acute pancreatitis?

Abdom Imaging 2015;40:488–499.

21.

Fung

Svystun

Fouladi

, et al. CT imaging, classification, and complications of acute pancreatitis. Abdom Radiol 2020;45:1243–1252.

22.

Murray

Darras

Walstra

, et al. Dual-energy CT in evaluation of the acute abdomen. RadioGraphics 2019;39:264–286.

23.

Hoilat GJ, Mathew G, Ahmad H. Pancreatic pseudoanevrysm. Treasure Island (FL): StatPearls Publishing, 2020.

24.

Mortele

Wiesner

Intriere

, et al. A modified CT severity index for evaluating acute pancreatitis: improved correlation with patient outcome. AJR Am J Roentgenol 2004;5:1261–1265.

25.

Mendelson

Anderson

Marshall

, et al. Vascular complications of pancreatitis. ANZ J Surg 2005;75:1073–1079.