Role of Standardized Reporting and Novel Imaging Markers in Chronic Pancreatitis

Abstract

Purpose of review

This article reviews recent efforts about standardized imaging features and reporting of chronic pancreatitis (CP) and recently published or ongoing imaging studies which aim to establish novel imaging biomarkers for detection of parenchymal changes seen in CP.

Recent findings

New novel MRI techniques are being developed to increase the diagnostic yield of CP specifically in the early stage. T1 relaxation time, T1 signal intensity ratio, and extracellular volume fraction offer potential advantages over conventional cross-sectional imaging, including simplicity of analysis and more objective interpretation of observations allowing population-based comparisons. In addition, standardized definitions and reporting guidelines for CP based on available evidence and expert consensus have been proposed. These new imaging biomarkers and reporting guidelines are being validated for prognostic/therapeutic assessment of adult patients participating in longitudinal studies of The Consortium for the Study of Chronic Pancreatitis, Diabetes, and Pancreatic Cancer.

Summary

New imaging biomarkers derived from novel MR imaging sequences promise a new chapter for diagnosis and severity assessment of CP; a cross-sectional imaging based diagnostic criteria for CP combining ductal and parenchymal findings. Standardized imaging findings and reporting guidelines of CP would enhance longitudinal assessment of disease severity in clinical trials and improve communication between radiologists and pancreatologists in clinical practice.

INTRODUCTION

Chronic pancreatitis (CP) is a chronic inflammatory condition of the pancreas with clinical manifestations ranging from abdominal pain, acute pancreatitis, exocrine and/or endocrine dysfunction and increased risk for development of pancreatic cancer. The histologic hallmarks of CP include fibrosis, chronic inflammation, and loss of acinar cells (1). Computerized tomography (CT) and magnetic resonance imaging (MRI) are the most common cross-sectional imaging studies performed for the evaluation of CP. The Cambridge classification (2), which was developed for endoscopic retrograde cholangiopancreatography (ERCP) in 1984, has been suggested for translational use in CT and Magnetic Resonance Cholangiopancreatography (MRCP) interpretation by guidelines of the American Pancreatic Association (3). Radiologists attempt to convert pancreatic ductal findings as defined by the Cambridge classification to what they see on cross-sectional images, but this translational approach is resulting in low or moderate agreement in terms of inter-observer and intra-observer variability (4). A major gap to address in current practice is that parenchymal changes that we observe by CT and MRI are not considered in diagnosis or assessment of disease severity. Hence, there is need for new cross-sectional imaging criteria that incorporate changes occurring within the parenchyma in addition to the ductal findings. Furthermore, as therapeutic agents for use in CP emerge from clinical trials, imaging features of CP are needed as biomarkers to assess longitudinal disease severity. Standardized imaging features and reporting standards of CP on cross-sectional imaging studies would facilitate classification of disease severity and longitudinal assessment in clinical trials.

The Consortium for the Study of Chronic Pancreatitis, Diabetes, and Pancreatic Cancer (CPDPC) was established by the National Institute of Diabetes and Digestive and Kidney Diseases and National Cancer Institute in 2015 to undertake collaborative studies on CP, diabetes mellitus and pancreatic adenocarcinoma (5). A recently published consensus by CPDPC investigators identifies, defines and provides metrics for reporting features of CP that will allow a more standardized approach to diagnosis and assessment of severity (6). Magnetic resonance imaging as a non-invasive method for the assessment of pancreatic fibrosis (MINIMAP) is an ongoing CPDPC prospective study in well-phenotyped patients aiming to ascertain the value of potential MR imaging biomarkers in assessment of CP severity (7). These efforts aim to incorporate MRI features of pancreatic parenchyma, such as T1 relaxation time, extracellular volume fraction, tissue enhancement dynamics, diffusion, stiffness, focal or diffuse gland atrophy and artificial intelligence features into a severity scoring system that can be used in clinical trials as well as clinical practice. Considering that parenchymal changes of CP precede ductal irregularities, there would be a significant benefit from developing MRI/MRCP based, robust diagnostic criteria combining ductal and parenchymal findings.

TEXT

This article reviews recently published or ongoing imaging studies which aim to establish novel imaging biomarkers for detection of parenchymal changes seen in CP and recent publication about standardized imaging features and reporting of CP.

Advanced Imaging of Chronic Pancreatitis

Since histologic diagnosis is rarely pursued due to risk of complications, cross-sectional imaging is the most practical clinical tool to diagnose and observe progression of CP. Among cross-sectional imaging studies, CT is readily available and performed more often in the setting of acute or chronic pancreatitis and is the modality of choice for detection of pancreatic calcifications. There have been some technical improvements in CT over the decades, but novel imaging techniques of MRI and MRCP have a greater potential to make a significant impact. It is widely acknowledged that MRI combined with MRCP is more sensitive than CT for detection of ductal and parenchymal changes, especially during early stages of CP (8, 9). T1 signal intensity ratio (T1 SIR), T1 mapping and extracellular volume (ECV) fraction have been shown to be useful quantitative imaging parameters for diagnosis and severity of CP (9, 10) and will be described in detail here. These new MR imaging features are specific to the parenchyma, and none contribute to the Cambridge classification (2), therefore their potential value is not yet captured in any classification system. Some of these advanced cross-sectional imaging techniques are listed in Table 1.

Table 1.

Advanced imaging techniques for chronic pancreatitis.

Previously published by Parakh and Tirkes. Advanced imaging techniques for chronic pancreatitis. Abdom Radiol (NY). 2020 May;45(5):1420-1438.doi:10.1007/s00261-019-02191-0.

Clinical Question	Specific Question	Useful Imaging techniques
Diagnosis of Structural/Functional Changes	Parenchymal	• T1 signal changes • Extracellular volume fraction • Contrast enhanced CT • Diffusion weighted imaging • MR Elastography • Pancreatic fat fraction (by MRI) • Dual-energy CT • 3D volumetry using AI • EUS with elastrography
	Ductal	• MRCP (with/without Secretin)
	Functional	• Secretin-enhanced MRCP • Perfusion CT • Radiomics
Assist in Therapeutic Decision Making	Diagnose obstruction	• MRCP (with/without Secretin) • Dual-energy CT
	Detect fluid collections	• Dual-energy CT (including material density images and virtual unenhanced images)
	Identify vascular complications	• Dual-energy CT (including material density images and virtual unenhanced images)
Differentiate CP from Pancreatic Adenocarcinoma	Determine contrast enhancement pattern and/or metabolic activity	• Dual-energy CT • MRI/MRCP • Perfusion CT • PET-CT • PET-MRI • Radiomics

T1-weighted Signal Intensity Ratio

MRI is commonly performed together with MRCP for evaluation of CP and provides imaging features of the parenchyma. One of the most commonly performed MRI sequences in abdominal imaging is the T1-weighted gradient echo image acquired with fat suppression. Normal pancreas has a shorter T1 relaxation time relative to other intra-abdominal organs due to protein-rich acinar cells (11). As a result, the normal pancreas is relatively hyperintense on unenhanced T1-weighted images. Decreased T1 signal intensity has been shown to correlate with loss of acinar cells which are replaced by fibrosis in CP (12–14) (Figure 1). The signal intensity needs to be quantified by the SIR which is calculated by dividing the average signal intensity of the pancreas with a reference organ of choice (commonly spleen or paraspinal muscle) (15). The value of T1 SIR as a biomarker has been histologically confirmed with degree of parenchymal fibrosis in patients with adenocarcinoma who underwent pancreatectomy (11) and total pancreatectomy with islet auto transplantation (TPIAT) (16). A different study looked into the association of endoscopically collected pancreatic fluid bicarbonate level with T1 SIR in patients with abdominal pain consistent with CP but with normal ductal imaging (Cambridge 0 or 1). There was a strong positive correlation of T1 SIR with the bicarbonate level (r=0.70). Due to this relationship, there was a significant difference in T1 SIR in normal bicarbonate level group as compared to low bicarbonate, (i.e., pancreatic exocrine dysfunction) (8). T1 SIR of 1.2 yielded sensitivity of 77% and specificity of 83% (AUC: 0.89) to diagnose early CP, which was not detected by Cambridge classification.

Figure 1.

T1 signal changes of the pancreas as seen on unenhanced, fat suppressed, T1-weighted gradient echo image of the abdomen. It is easy to notice that T1 signal of the normal pancreas is brighter than the spleen (S) or paraspinal muscle (PSM). Since this is not a quantitate MR image, measured signal intensity does not reflect the true T1 relaxation time, therefore, signal intensity ratio needs to be used which is the ratio of the pancreas to a reference organ, such as the spleen, or paraspinal muscle. (a) Axial image in a patient without CP shows relatively higher signal of the pancreas (thick arrow) compared to the S and PSM. (b) Axial image in a patient with known CP shows a relatively decreased signal intensity of the pancreas (thin arrow) which is similar to the S and PSM. A pancreas to splenic SIR of <1.2 was reported to be 77% sensitive and 76% specific for the diagnosis of CP.

Previously published by Parakh and Tirkes. Advanced imaging techniques for chronic pancreatitis. Abdom Radiol (NY). 2020 May;45(5):1420–1438. doi: 10.1007/s00261-019-02191-0.

T1 Mapping

T1 mapping is a quantitative MR imaging technique that allows measurement of tissue specific T1 relaxation time and the unit of measurement is in milliseconds. The MINIMAP study is a prospective, multi-institutional, multi-vendor study currently looking into validation of the normal range of pancreas T1 relation times in the control population, as well as the spectrum of relaxation times throughout the course of CP (7). Following recent technical developments to decrease imaging time, newer protocols are able to produce T1 maps in a single breath hold (17) and eligible to enter clinical practice. T1 mapping offers a great potential advantage over conventional T1-weighted (gradient echo) imaging, which is “quantitative analysis”, allowing population-based comparisons across different MRI scanners and imaging techniques provided by different vendors. Assessment is done directly by measuring the relaxation time of the pancreas alone as seen in Figure 2, which cannot be done in conventional MRI. It should be noted that T1 relaxation time is specific to the MR signal strength therefore different numbers are used for 1.5 versus 3.0 Tesla.

Figure 2.

T1 mapping is a quantitative MR image and provides the true T1 relaxation of the pancreas reported in milliseconds. This is an axial grayscale T1 map obtained at a 3.0 Tesla MR scanner in a patient with no known CP. The T1 relaxation time in the pancreatic neck (P) measures 701 msec. T1 map can also be presented in a color scale format and the intensity of the pancreatic signal can be visually assessed.

Previously published by Parakh and Tirkes. Advanced imaging techniques for chronic pancreatitis. Abdom Radiol (NY). 2020 May;45(5):1420–1438. doi:10.1007/s00261-019-02191-0.

Several T1 mapping techniques have been assessed and compared by a recent study as a potential application in abdominal imaging (17). The advantage of the 3-dimensional T1 mapping techniques (such as variable flip angle gradient echo) have been acknowledged for pancreatic imaging due to higher spatial resolution provided within one breath hold (17). Using these 3-dimensional T1 mapping techniques, normal pancreatic parenchyma has been reported to be median T1 of 654 ms at 1.5 Tesla and 717 ms at 3.0 Tesla (18). A different retrospective study looked into suspected CP patients and reported threshold of 900 msec (at 3.0 Tesla) to be 80% sensitive and 69% specific for mild CP (9).

Extracellular Volume Imaging

ECV is a quantitative MR biomarker that calculates volume of extracellular matrix in any solid organ tissue. Changes to the extracellular matrix is being used as a biomarker for a variety of pathologies such as liver cirrhosis and myocardial fibrosis. The MINIMAP study is validating the ECV as a potential biomarker for fibrosis in CP patients (7, 9, 10). ECV of the pancreas is expected to increase in patients with CP secondary to increased tissue fibrosis, namely, collagen (19, 20) and proteoglycan content (21). This radiomics method measures the T1 relaxation time of the pancreas before and after gadolinium administration. Since gadolinium is an extracellular MRI contrast agent, enhancement fraction in the extracellular tissue relative to the plasma can be calculated and then depicted pixel by pixel forming an image called ECV map (Figure 3) (10).

Figure 3.

Extracellular volume (ECV) fraction technique utilizes T1 maps obtained before and after MR contrast enhancement. Following post-processing of two time point T1 relaxation times, the final image depicts calculated ECV fraction in each pixel, reported as percentage, either in grayscale or color map. A preliminary study showed that ECV threshold of >0.27 had 92% sensitivity and 77% specificity for CP. (P=pancreas; S=spleen).

Previously published by Parakh and Tirkes. Advanced imaging techniques for chronic pancreatitis. Abdom Radiol (NY). 2020 May;45(5):1420–1438. doi: 10.1007/s00261-019-02191-0.

MR Elastography (MRE)

MRE of the liver has been validated as a non-invasive imaging biomarker for the degree of hepatic fibrosis. While MRE of the liver is currently being used in clinical practice, MRE of the pancreas is still under development and has been technically more challenging due to the smaller size and deep location of the gland (22). In a pilot study of 20 healthy volunteers, pancreas MRE demonstrated promising and reproducible stiffness measurements (23). In a limited study of thirteen CP and pancreatic adenocarcinoma patients, stiffness measured by MRE was suggested as a potential indicator of disease (24). The MINIMAP study is also looking into validation of the feasibility of MRE for CP (7).

Radiomics Analysis

Extraction of quantitative features from radiological images to decode underlying tissue biology, histopathology, local hypoxia and genetics for potentially improving clinical decision-making is called radiomics. An in-depth explanation of different radiomic parameters is beyond the scope of this review (25). Radiomic features, including texture analysis, a first order radiomic extraction have been shown to determine patient prognosis, predict response to chemotherapy and determine resectability in pancreatic adenocarcinoma (26, 27). Significantly different radiomic features were recently demonstrated in autoimmune pancreatitis compared to pancreatic adenocarcinoma on PET/CT images (28). Another pilot study postulated the feasibility of radiomics to predict diabetes with an AUC ranging from 0.61 to 0.69 (29). Before these techniques are considered for clinical implementation, standardization of technique, impact of image reconstructions and large scale, prospective, multi-institutional validation is needed. Potential areas of further work include differentiating chronic mass forming pancreatitis from pancreatic adenocarcinoma, quantifying exocrine and/or endocrine dysfunction and predicting likelihood of progression from acute pancreatitis to CP. Most useful CP-related artificial intelligence applications would be in screening and serial monitoring with automated pancreatic volumetry, quantification of calcification burden, and prediction of secondary disease states, such as diabetes and pancreatic adenocarcinoma.

Pancreatic Fat Fraction

Another commonly observed imaging finding in CP patients is higher pancreatic fat fraction. In a recent study performed in midwestern USA, patients with CP showed higher visceral fat volume and pancreatic fat fraction (30). Reported direct positive correlation of pancreatic fat fraction with visceral adipose tissue volume (r = 0.54) was suggestive of pancreatic steatosis being part of the metabolic syndrome. As such, type 2 diabetes mellitus was also associated with higher pancreatic fat fraction (30). While the mean pancreatic fat fraction in this US study was found to be 15%, a much lower fraction (4%) was reported in a German study (31). More prospective studies are needed to investigate the normal pancreatic fraction in different populations and association of pancreatic steatosis with diabetes and acute/chronic pancreatitis.

Standardized Reporting in Chronic Pancreatitis

It is intuitive that precise and standardized communication between the radiologist and clinician promotes better patient care and accuracy of longitudinal monitoring. A formalized assessment of CP imaging features would enhance radiologists’ ability to establish the diagnosis (perhaps at an earlier stage) and decrease the need for more invasive diagnostic procedures. It is with this concept in mind that the CPDPC investigators proposed standardized definitions and reporting guidelines based on available evidence and expert consensus (6) (Table 2). These imaging features were co-authored by the pancreas radiologists, surgeons, and gastroenterologists from many institutions in the US. This multidisciplinary approach represents the most desirable combination that provides high-level experientially derived evidence. The group set out two goals: first, to minimize imaging interpretative variability by offering metrics for reporting features of CP and second, to facilitate classification of disease severity and longitudinal assessment in clinical trials. These recommendations are expected to promote the imaging reports that comprehensively focus on key parameters that aid in therapeutic decision making. This publication also includes recommendations on how to perform the dedicated CT or MRI examinations which is essential for accurate assessment. With carefully chosen high quality images, the authors have demonstrated unique findings at CT (presence and number of calcifications), followed by common findings at both CT and MRI (distribution; parenchymal thickness; main pancreatic duct narrowing, caliber, and contour) and finally MRI or MRCP findings (side-branch ectasia with and with secretin enhancement, degree of duodenal filling following secretin stimulation, T1 signal intensity ratio, and parenchymal enhancement following gadolinium administration) for patients suspected of having CP. These imaging features are currently being validated for prognostic/therapeutic assessment of patients participating in longitudinal CPDPC studies, potentially establishing CT, MRI or MRCP as a biomarker for CP progression (7, 32). Future use of standardized metrics in well-controlled clinical trials will help to validate these techniques and potentially allow for clinical adoption. Individual radiologists may choose to present this information in a format that is suitable to their reporting style and the needs of their referring physicians.

Table 2.

Guidelines for reporting imaging findings proposed by the CPDPC investigators.

Previously published by Tirkes et al. Radiology. 2019 Jan;290(1):207-215. doi: 10.1148/radiol.2018181353. Epub 2018 Oct 16.

CT Findings
Pancreatic Calcifications	- Count both intra-ductal and parenchymal calcifications and note the size of calcification  - Document the location if present;   ○ In the head, body or tail  - Size   ○ Punctate calcification is < 3 mm   ○ Coarse calcification is ≥ 3mm  - Number and size of calcifications are separated into three groups in terms of increasing severity   ○ < 7 punctate foci   ○ 7 – 49 punctate foci; < 7 coarse foci   ○ Innumerable (≥50) punctate foci; ≥7 coarse foci
CT/MRI Common Findings
Pancreas Thickness	- Measure the thickness of the pancreatic parenchyma in the region of the pancreatic body; perpendicular to the longitudinal axis of the parenchyma at the level of the lateral margin of the adjacent vertebral body or upstream to pancreatic ductal calculus/stricture.  - Avoid including the splenic vein and artery in the measurement. If possible, measure in a plane which doesn’t include a significantly dilated pancreatic duct  - If pancreatic duct is dilated and in the measurement plane, subtract the caliber of the duct from the pancreatic thickness  - On MRI, measure the thickness of the parenchyma on an axial T1 or T2 weighted image that best shows the anterior and posterior margins.
Distribution of Findings	- Report the involvement of the pancreas with abnormal findings (e.g., contour irregularity, atrophy, and calcifications) as:   ○ Normal   ○ ≤ 30% of entire pancreas   ○ 30 – 70% of entire pancreas   ○ ≥ 70% of entire pancreas
Pancreatic Duct Narrowing	- Document the location if present; in the head/neck, body or tail
Pancreatic Duct Caliber	- Measure maximum diameter of pancreatic duct on any location
Pancreatic Duct Contour	- Report the pancreatic duct contour as;   ○ Smooth: no contour irregularity   ○ Mild irregularity: subtle subjective contour irregularity without significant narrowing (except for the dominant obstructing stricture) of the duct   ○ Moderate to markedly irregular: distinct irregularity of the contour demonstrating focal areas of narrowing and dilation
Pancreatic Operations	- Resection performed (select one or more segments)   ○ In the Head / Body / Tail  - Drainage performed   ○ Pancreaticojejunostomy   ○ Cystgastrostomy   ○ Cystjejunostomy   ○ Cystduodenostomy   ○ Biliary bypass (e.g., choledochojejunostomy or hepaticojejunostomy)   ○ Other
MRI and MRCP Findings
Number of Ectatic Branch Ducts	- Document number of ectatic side-branch ducts   ○ No side-branch ectasia   ○ <3 side-branch ectasia   ○ ≥ 3 side-branch ectasia
PD compliance after secretin	- Main pancreatic duct diameter   ○ increases with secretin   ○ does not increase with secretin
Duodenal Filling after secretin	- Report the duodenal filling on MRCP after secretin   ○ Fluid filling up to the proximal jejunum   ○ Fluid filling duodenum beyond the genu inferius (beyond the second portion) but not reaching the jejunum   ○ Fluid fills the second portion of the duodenum, but not beyond genu inferius   ○ Fluid limited to the duodenal bulb   ○ No response
T1 Signal Intensity Ratio	- Document the T1-weighted signal intensity ratio using the unenhanced axial gradient echo images with fat suppression   ○ Pancreas to Spleen   ○ Pancreas to Paraspinal muscle   ○ Pancreas to Liver  - Special attention should be given to drawing a region of interest in a homogenous region of the parenchyma avoiding volume averaging from peri-pancreatic fat, vessels, and PD
Parenchymal Enhancement Ratio	- Document the enhancement ratio of the pancreas in the most homogenous section of the parenchyma   ○ Arterial-to-venous enhancement ratio

CONCLUSION

While MRCP is used for detecting ductal abnormalities, the potential value of MRI is being explored to augment the diagnostic accuracy by providing information about parenchymal changes related to CP. Ongoing prospective imaging studies within the CPDPC aim to validate these parenchymal findings and possibility generate a more robust severity scoring system combining ductal and parenchymal findings utilizing MRI and MRCP. In addition, CPDPC investigators published a consensus statement about imaging features and reporting guidelines that identifies, defines, and provides metrics for reporting features of CP. These recommendations are expected to facilitate classification of disease severity in clinical practice and longitudinal assessment in clinical trials.

Key points.

• Novel MR imaging biomarkers offer potential advantages over conventional imaging by using quantitative analysis which would potentially allow more objective observation of disease progression, response to therapy and population-based comparisons.

• There may be a significant benefit from new MRI/MRCP based, more robust diagnostic/severity scoring criteria combining ductal and parenchymal findings seen in CP.

• Standardized qualitative and quantitative reporting of imaging findings promotes reproducibility of diagnosis and improves communication between radiologists and pancreatologists.