Advances in MRI of Chronic Pancreatitis

Abstract

MRI and MRCP play an essential role in diagnosing CP by imaging pancreatic parenchyma and ducts. Quantitative and semi-quantitative MR imaging offers potential advantages over conventional MR imaging, including simplicity of analysis, quantitative and population-based comparisons, and more direct interpretation of disease progression or response to drug therapy. Using parenchymal imaging techniques may provide quantitative metrics for determining the presence and severity of acinar cell loss and aid in diagnosing CP. Given that the parenchymal changes of CP precede the ductal involvement, there would be a significant benefit from developing a new MRI/MRCP based, more robust diagnostic criteria combining ductal and parenchymal findings.

Introduction

Chronic pancreatitis (CP) is a fibroinflammatory syndrome resulting in chronic abdominal pain, exocrine and endocrine pancreatic insufficiency, reduced quality of life, and shorter life expectancy¹. CP is a low-prevalence disease; however, the incidence and prevalence of this debilitating disease are rising¹. Clinical features of CP are highly variable and include minimal or no symptoms of debilitating pain, repeated episodes of acute pancreatitis, pancreatic exocrine and endocrine insufficiency.

While diagnosing severe CP is often straightforward, establishing a diagnosis at an early stage remains elusive^2–5. Patients often undergo an exhaustive array of costly examinations since a biopsy of the pancreas is not a favored clinical practice as it entails a risk of biopsy-related pancreatitis. MRI and MRCP play an essential role in diagnosing chronic pancreatitis (CP) by imaging pancreatic parenchyma and ducts. MRI/MRCP is more widely used than computed tomography (CT) for early or progressing CP due to its increased sensitivity for pancreatic ductal and gland changes; however, it does not detect parenchymal calcifications seen in advanced CP. Quantitative MR imaging offers potential advantages over conventional qualitative imaging, including simplicity of analysis, quantitative and population-based comparisons, and more direct interpretation of detected changes. These techniques may provide quantitative metrics for determining the presence and severity of acinar cell loss and aid in diagnosing chronic pancreatitis. Given that the parenchymal changes of CP precede the ductal involvement, there would be a significant benefit from developing an MRI/MRCP based, more robust diagnostic criteria combining ductal and parenchymal findings. This review article will discuss advances in pancreatic imaging techniques by MRI.

Discussion

T1-Weighted Signal Intensity Ratio (SIR)

The pancreas is an essential exocrine gland of the digestive system, secreting more than a liter of proteinaceous fluid per day^6,7. This abundance of proteinaceous material in the acinar cells is the reason the normal pancreas exhibits a relatively higher T1W signal intensity than other solid organs^8,9. A recent multicenter study reported the average T1W signal in CP to be 1.11 vs. controls 1.29, as seen in Figure 1¹⁰. This association is supported by studies that included surgical histopathology^8,11–13 and reported a correlation of parenchymal MRI features (T1 SIR, T1 relaxation time, diffusion-weighted imaging, enhancement ratio, and MR elastography) with the degree of fibrosis. In addition to fibrosis, some studies have reported lower T1W signal in patients with exocrine pancreatic dysfunction^14–16. One of these studies included 60 surgical specimens and showed that the T1 score had a higher correlation for pancreatic fibrosis than the Cambridge score¹³. Another study found a significant positive correlation (r=0.70) between pancreatic fluid bicarbonate level and T1 SIR of the pancreas to the spleen. A pancreas-to-splenic SIR threshold of less than 1.2 had a sensitivity of 77% and specificity of 76% for detecting low pancreatic fluid bicarbonate, i.e., pancreatic exocrine dysfunction¹⁵. These results concurred with the previously reported poor sensitivity of the Cambridge classification for detecting early CP¹⁷. They suggested that T1W SIR can be useful for detecting early chronic pancreatitis even when the ductal imaging is normal.

Figure 1.

T1-weighted signal of the normal pancreas. This is an axial fat-suppressed T1W image with no contrast showing the pancreas (arrow) in a 19-year-old female with no known pancreatic disease. The pancreas shows a significantly higher signal than the other solid organs in the upper abdomen. The T1W signal is typically assessed by comparing it to a reference organ, most commonly the spleen. This ratio is called T1 SIR (or T1 Score as the name given by the PROCEED study).

Prospective Evaluation of Chronic Pancreatitis for Epidemiologic and Translational Studies (PROCEED) is an ongoing, longitudinal cohort study of CP in adult participants in the US (NCT# 03099850)¹⁸. It is one of the four major studies conducted by the Consortium for the Study of Chronic Pancreatitis, Diabetes, and Pancreatic Cancer, established by the National Cancer Institute and the National Institute of Diabetes and Digestive and Kidney Diseases^19,20. PROCEED analyzed MRI on 820 of its participants and reported that T1 SIR of the pancreas can be used as an imaging biomarker for staging chronic pancreatitis²¹. PROCEED study called T1 SIR of the pancreas-to-spleen “T1 score” and reported a negative linear correlation between the signal intensity and progression of CP. T1 score appears to be a practical imaging biomarker in evaluating disease severity in clinical research and practice, and longitudinal analysis should be available from the PROCEED study in the near future. PROCEED published a new classification of CP severity named mechanistic stages of chronic pancreatitis (MSCP) based on their clinical history (i.e., abdominal pain suspected of pancreatic origin, AP or RAP), the Cambridge grade on MRCP, and the presence of parenchymal and/or ductal calcifications on the CT scan as seen on Table 1²¹.

Table 1.

Mechanistic stages of chronic pancreatitis (MSCP). Defined by the PROCEED study.

MSCP	PROCEED cohort	Cambridge grade by MRCP	Calcification by CT	Sample size (n)
0	Chronic abdominal pain or indeterminate CP	0	−	56
1	Acute or recurrent AP	0	−	179
2	Chronic abdominal pain or indeterminate CP	1 or 2	−	38
3	Acute or recurrent AP	1 or 2	−	169
4	Definite CP	0, 1, or 2	+	49
5	Definite CP	3 or 4	−	116
6	Definite CP	3 or 4	+	213

Limitations exist in conventional T1-weighted imaging in which the tissue contrast depends on multiple factors, including acquisition parameters, receiver coil geometry sensitivity and signal amplifier gains. Variation in signal intensity is commonly observed through the choice of pulse sequence and manipulation of acquisition parameters (e.g., flip angle, echo time, repetition time, inversion time, etc.).

Quantitative MRI Imaging

T1 mapping may be a more reliable method than traditional T1-weighted images since quantitative data allows ready comparison across longitudinal time points, potentially more accurate interpretation of intensity changes²². With this idea in mind, the Magnetic Resonance Imaging as a Non-invasive Method for the Assessment of Pancreatic Fibrosis (MINIMAP) study evaluated multiple MRI and MRCP features in a multi-institutional, multi-vendor setting to explore potential imaging biomarkers for CP²³. MINIMAP reported that quantitative MR parameters are helpful in the diagnosis of CP²⁴ and generated a multi-parametric score (Q-MRI) that combines three MR parameters (T1, ECV fraction, and fat fraction). Further studies are warranted in this new field using larger study populations and longitudinal follow-ups.

T1 mapping

T1 mapping is a quantitative MR imaging technique that allows us to measure the tissue specific T1 relaxation time of the tissues (Figure 2). Utilizing recently introduced fast 3D pulse sequences, T1 mapping takes less time than other imaging techniques such as T2 mapping, DWI or S-MRCP. Multiple T1 mapping pulse sequences are available either as a product version or prototype sequence under development by the manufacturers. There is no consensus about which T1 mapping pulse sequence is ideal for abdominal imaging. A study compared four different pulse sequences for the imaging of the pancreas: variable flip angle (VFA), modified look-locker inversion recovery (MOLLI), a prototype inversion recovery (IR-SNAPSHOT), and a prototype saturation recovery single-shot acquisition (SASHA)²⁵. The principle of variable flip angle (VFA) pulse sequence is to calculate T1 by acquiring voxel signals at a steady state using multiple flip angles²⁶. Inversion recovery (IR-SNAPSHOT) is based on the relaxation of longitudinal magnetization after an inversion radio frequency (RF) pulse is applied. A series of quick acquisitions are collected at different delay times following the inversion RF pulse, and signals at different delays are fitted using the relaxation model²⁷. MOLLI is a commercially available sequence developed for myocardial imaging and uses a similar but modified IR-SNAPSHOT principle. The series of acquisitions following the inversion RF are segmented and synchronized using an ECG signal so that the data acquisition only occurs during the diastolic period of a cardiac cycle²⁸. SASHA is also similar to IR-SNAPSHOT, except it utilizes a saturation RF pulse instead of an inversion RF pulse²⁹. This study reported that MOLLI, SASHA, and IR-SNAPSHOT provided the highest precision, while VFA had substantial precision. MOLLI and SASHA were designed initially for myocardial imaging and provided only one image in one breath hold time, while IR-SNAPSHOT can acquire three images. The advantage of the VFA sequence is fast 3D acquisition, generating 64 slices in one breath hold. The disadvantage of VFA is the inherent sensitivity to pulsatile flow within the aortic blood. This study concluded that more refinement of pulse sequences is necessary to provide a large spatial coverage in one breath hold together with high precision in abdominal imaging. Further improving these MRI sequences will expand the use of these techniques, probably allowing them to be introduced into clinical practice to enable reliable quantitative diagnosis and follow-up of CP.

Figure 2.

This is an axial T1 map obtained without IV contrast in a 36-year-old male with no pancreas disease. Each pixel of this quantitative MR image depicts the T1 relation time (ms) of any tissue, including the pancreas (arrows).

T1 mapping has been investigated in quantifying short-term and mid-term autoimmune pancreatitis (AIP) responses to corticosteroid treatment³⁰. The T1 relaxation time is reported to be prolonged in AIP. However, after four weeks of corticosteroid therapy, T1 relaxation time was shortened significantly, further shortening towards normalization in 12 weeks. In AIP patients with elevated serum IgG4 at baseline, T1 relaxation time demonstrated a significant positive correlation with serum IgG4 level; in patients with normal serum IgG4, T1 relaxation time shortening preceded or was in accordance with symptom relief, suggesting a promising role of T1-mapping as a treatment outcome measure^30,31.

Challenges associated with conventional sequences also exist for quantitative MRI techniques: motion, spatial resolution, field non-uniformity, magnetization transfer, and partial volume³².

Extracellular Volume Imaging (ECV)

ECV is a radiomics method that exploits changes to the extracellular matrix. Utilizing gadolinium’s tissue and blood plasma concentration, the ECV technique dichotomizes the tissues into intra- and interstitial spaces. ECV fraction calculates the interstitial fraction, which increases following replacement of the normal cells with fibrosis. ECV fraction can be either measured as a value or depicted as pixels on an image (Figure 3). In the case of CP, ECV interrogates increased interstitial space seen with tissue fibrosis. T1 relaxation times obtained from the pancreas and the blood pool (typically aortic lumen) in unenhanced and post-contrast equilibrium phases are entered into this formula to calculate the ECV fraction:

$$ECV\mathit{\text{pancreas}}=\frac{(1-\text{hematocrit})\times ∆\text{R}1\mathit{\text{pancreas}}}{∆\text{R}1\mathit{\text{blood}}}$$

where ΔR1_pancreas and ΔR1_blood are defined as the change of 1/T1 relaxation rate in pancreas and blood pool relaxivity before and after contrast administration²³.

Figure 3.

Extracellular volume (ECV) fraction map. This is an axial color scale ECV image of the abdomen at the level of the pancreas (arrows) in a patient with no pancreas disease.

A recent study on patients with no pancreatic disease reported median T1 on 1.5T as 654 ms, median T1 on 3T as 717 ms, median ECV on 1.5T as 0.28 and median ECV on 3T as 0.25³³. T1 (r =0.24) had a mild positive correlation with rising age. Despite the well-known differences in T1 relaxation times between 1.5T and 3T, the ECV fraction does not show variations when using different magnet strengths³⁴.

Alterations of tissue T1 have been observed in the presence of various pathologic conditions in the abdomen, including CP^35,36. A retrospective study performed using 3T showed that T1 >950 ms had 64% sensitivity and 88% specificity, while ECV >0.27 had 92% sensitivity and 77% specificity³⁶ for the diagnosis of CP. Combining ECV and T1 yielded a sensitivity of 85% and specificity of 92% (AUC: 0.94)³⁶. Another study using 3T showed the T1 >900 ms threshold to be 80% sensitive and 69% specific for diagnosing mild CP (AUC: 0.81)³⁵.

Diffusion-Weighted Imaging (DWI)

Diffusion-weighted MRI measures the restriction of free water molecules in the gland. The more fibrosis there is, the more likely there will be less diffusion of water molecules, measured as apparent diffusion coefficient (ADC). The ADC is expected to be higher in patients with no pancreas disease than chronic pancreatitis with fibrosis Figure 4. Exploiting this idea, diffusion MRI was used after IV secretin stimulation, and enhancing the sensitivity to depict subtle abnormalities in diffusion restriction has been reported³⁷. However, the MINIMAP study, which was performed in a multi-institutional and multi-vendor setting, showed no difference in ADC of the pancreas between the control, suspected and definite CP participants²⁴.

Figure 4.

Diffusion-weighted imaging of the pancreas. This is an apparent diffusion coefficient (ADC) map of the pancreas (arrows) in a 19-year-old patient with no pancreas disease. ADC is a quantitative map, and measurements reflect the diffusion properties of the tissues. The ADC is expected to be lower in patients with tissue fibrosis.

MR Elastography (MRE)

MRE is a reliable marker of hepatic fibrosis in patients with chronic liver disease. While there are no controlled data evaluating this technique in patients with CP, there is room for optimism as recent data demonstrated the feasibility of using MRE to determine pancreatic stiffness in healthy volunteers³⁸. Reproducible stiffness measurements were noted throughout the pancreas, with imaging parameters and equipment different than those used for liver imaging. Preliminary data suggest that pancreatic MRE can provide promising and reproducible stiffness measurements throughout the pancreas, potentially allowing for assessing pancreatic fibrosis³⁹.

Correlation of MRI Features with Histopathology

The two most common histologic features of CP are the loss of acinar tissue (atrophy) and fibrosis. The fibrosis may surround the lobules (perilobular or interlobular fibrosis) or extend into the lobules of acinar tissue (intralobular fibrosis)⁴⁰. Chronic inflammatory infiltration may be present in the developmental stages, but this feature is highly variable and disappears late during CP. Chronic pancreatitis can be a patchy or localized process with regional involvement. This feature is best understood by considering the mechanisms of pathogenesis, particularly the necrosis-fibrosis hypothesis, which posits that CP develops as a result of multiple episodes of AP with necrosis and scarring. This process may be patchy initially, progressing to a diffuse pattern after multiple episodes. This is commonly considered to be the mechanism in alcoholic CP, para-duodenal CP, and likely hereditary pancreatitis⁴¹.

Several studies reported that MRI parenchymal signal changes reflect the histopathological changes seen in CP^42–44. T1 SIR (T1 Score), T1 relaxation time, ECV fraction, DWI, arterio-venous enhancement ratio (AVR), and MR elastography have been shown to correlate with CP^8,11–13. One of the best examples of these studies analyzed 60 surgical specimens and reported that the MRI parameters correlated with pancreatic fibrosis better than the Cambridge score¹³. Some studies reported that MRI features can show changes in the parenchyma in participants with suspected CP while the Cambridge score remains normal or equivocal. In a study comparing 24 participants with suspected CP to 20 controls, AVR was significantly lower in the suspected CP cohort⁴⁵. In another study with 29 suspected CP participants and 22 controls, T1 SIR was significantly lower in those with exocrine dysfunction, as determined by an endoscopic pancreatic function test¹⁵. These studies point out the potential of parenchymal changes in detecting CP before the appearance of ductal morphologic changes. These suggestions are probably valid, considering that the ductal system is only 4% of the pancreas, while the remainder of the gland comprises acinar cells, the extracellular matrix, and islet cells⁴⁶.

Revised Classification System

Cambridge classification, designed for ERCP, has been used for over four decades⁴⁷. This classification system is based on ERCP findings; however, it has been adapted to MRCP and has remained the de facto diagnostic standard, primarily due to the familiarity of referring physicians. The role of MRI/MRCP in diagnosing CP has been acknowledged by recent guidelines of the American Pancreatic Association, and a modified Cambridge classification for MR/MRCP and CT has been proposed (Table 2)⁴⁸.

Table 2.

Cambridge classification adapted for findings seen on MRCP, CT, and US. American Pancreatic Association Practice Guidelines, 2014.

Cambridge Classification	MRCP/ERCP findings	US/CT/MR findings
0 Normal	Normal	no abnormal signs
I Equivocal CP	< 3 dilated side branches	one of the following: dilated main pancreatic duct (2 – 4 mm) slight gland enlargement heterogeneous parenchyma small cavities (<10 mm) irregular ducts focal pancreatitis increased echogenicity of the main duct wall irregular head/body contour
II Mild CP	Three or more dilated side branches	Two or more of the following: dilated main duct (2–4 mm) gland enlargement heterogeneous parenchyma small cavities (<10 mm) irregular ducts focal acute pancreatitis increased echogenicity of the main duct wall irregular head/body contour
III Moderate CP	>3 dilated side branches and dilated main duct	Same as above
IV Severe CP	all above and one or more of large cavity >10 mm intra-ductal filling defects duct obstruction (stricture) duct dilatation or irregularity	Above changes and one or more of the: large cavity >10 mm gland enlargement intra-ductal filling defects/calculi duct obstruction/stricture/ or gross irregularity

The value of parenchymal MRI features has been reported by retrospective studies and touted by expert panel reviews and consensus statements^42–44. Ductal imaging reflects periductal fibrosis but cannot capture other histopathological triad elements, including parenchymal fibrosis, loss of acinar tissue, and parenchymal fat deposition⁴¹. Over the years, several limitations of the Cambridge classification have been reported. There is significant variability in the interpretation of Cambridge classification by MRCP⁴⁹, resulting in moderate interobserver variation among expert radiologists^50,51. Recent studies showed that MRCP findings alone provide a limited evaluation of CP since MRCP images do not reflect the changes in the pancreatic parenchyma^46,52,53. The value of parenchymal MRI features has been reported by retrospective studies and touted by expert panel reviews and consensus statements^42–44. To this date, radiologists have not established a classification for CP based on MRI/MRCP, CT or US-specific features. New criteria for CP, incorporating ductal and parenchymal MRI features, are needed to improve the clinical practice.

Reporting Guidelines for Chronic Pancreatitis

New cross-sectional imaging definitions and standardized reporting recommendations for CP would allow for a more standardized approach to diagnosis and assessment of disease severity for clinical trials. As therapeutic drugs emerge for clinical trials, definitions of disease and severity are needed for longitudinal markers and therapeutic targets. Precise and standardized communication between the radiologist and clinicians is essential for the quality and safety of clinical practice. With this idea in mind, the CPDPC proposed standardized definitions and reporting recommendations based on available evidence and expert consensus. CT, MRI or MRCP data points described in this manuscript can potentially be imaging biomarkers for disease progression⁴³. The guidelines included definitions for pancreatic calcifications, thickness, T1W signal, arterio-venous enhancement ratio, several ductal features and distribution of findings. Future use of these standardized metrics in well-controlled clinical trials will help to validate them and potentially allow for clinical adoption.

Summary

Several new MRI parameters of the pancreas have significant potential to be practical, accurate and cost-efficient imaging biomarkers of CP. These biomarkers can be useful to exclude chronic pancreatitis with high certainty, to reliably rule in early-stage disease, to help predict disease progression in these patients, and to monitor their effects to slow or reverse disease progression.

Key Points.

• Parenchymal MRI features can objectively identify different histopathologic changes seen with chronic pancreatitis.

• T1-weighted MR imaging can be a biomarker for decreased pancreatic exocrine function and chronic pancreatitis.

• Pancreatic T1 relaxation time, ECV fraction, and fat signal fraction are higher in patients with CP.

• Standardized definitions and reporting of chronic pancreatitis on imaging studies will facilitate the classification of disease severity and longitudinal assessment in clinical trials.

Clinical Care Points.

• MRI and MRCP play an essential role in diagnosing CP by imaging pancreatic parenchyma and ducts. Quantitative and semi-quantitative MR imaging offers potential advantages over conventional MR imaging, including simplicity of analysis, quantitative and population-based comparisons, and more direct interpretation of disease progression or response to drug therapy.

• Using parenchymal imaging techniques may provide quantitative metrics for determining the presence and severity of acinar cell loss and aid in diagnosing CP.

• Given that the parenchymal changes of CP precede the ductal involvement, there would be a significant benefit from developing a new MRI/MRCP based, more robust diagnostic criteria combining ductal and parenchymal findings.