Magnetic resonance imaging of inflammatory pseudotumor of the liver: a 2021 systematic literature update and series presentation

Abstract

Purpose: Inflammatory pseudotumors of the liver (IPTL) are not exceptional benign lesions with various etiologies, histology, and imaging appearances. The incomplete knowledge of this pathology and the wide polymorphism sometimes resembling malignancy often induce long and expensive diagnostic flow, biopsy and occasionally unnecessary surgery. We propose a systematic revision of MRI literature data (2000-2021) with some narrative inserts and 10 new complete MRI cases, with the aim of organizing the data about IPTL and identifying some typical features able to improve its diagnosis from imaging.

Methods: We performed a systematic revision of literature from 2000 to 2021 to obtain MRI features, epidemiological, and clinical data of IPTL. The basic online search algorithm on the PubMed database was "(pseudotumor) AND (liver) AND (imaging)." Quality assessment was performed using both scales by Moola for case report studies and by Munn for cross-sectional studies reporting prevalence data. A case-based retrospective study by collecting patients diagnosed with IPTL from three different university hospitals from 2015 to 2021 was done as well. Only cases with MR examinations complete with T1/T2/contrast-enhanced T1/Diffusion-Weighted (W) images and pathology-proven IPTL were selected.

Results: After screening/selection 38 articles were included for a total of 114 patients. In our experience we selected 10 cases for a total of 16 IPTLs; 8 out of 10 patients underwent at least 1 MRI follow-up. Some reproducible and rather typical imaging findings for IPTL were found. The targetoid aspect of IPTL is very frequent in our experience (75% on T1W, 44% on T2W, 81% on contrast-enhanced T1W (at least one phase), 100% on Diffusion-W images) but is also recurrent in the literature (6% on T1W, 31% on T2W, 51% on CE-T1W (at least one phase), 18% on Diffusion-W images, and 67% on hepatobiliary phase). In our experience, Apparent Diffusion Coefficient map values were always equal to or higher than those of the surrounding parenchyma, and at MRI follow-up, nodule/s disappeared at first/second control, in six patients, while in the remaining 2, lesions persisted with tendency to dehydration.

Conclusion: A targetoid-like aspect of a focal liver lesion must raise diagnostic suspicion, especially if IgG4-positive plasma is detected. MRI follow-up mainly shows the disappearance of the lesion or its reduction with dehydration.

Keywords: Focal liver lesion; IgG4-related disease; Inflammatory pseudotumor of the liver; Magnetic resonance imaging; Targetoid aspect.

Fig. 1

Flow diagram illustrating the selection of papers extracted from literature

Fig. 2

Flow diagram illustrating the selection of cases from our databases

Fig. 3

Inflammatory pseudotumor of the liver of the hepatic segment 7 in patient with history of breast cancer, 9 years ago. On MR, two-layered concentric “targetoid” appearance with the hypointense central core is seen on arterial phase (a) and gradient echo T1W out of phase (d). Target appearance is maintained on T2W images (e), with hyperintense central core, high b-value (750 mm²/s) DWI (g), and ADC map (h) with low and high central core signal intensity, respectively. On portal (b), equilibrium (c), and hepatobiliary phase (f), the lesion shows signal hypointensity. A transcutaneous biopsy was performed. Histological examination (i, hematoxylin–eosin, original magnification × 100) shows a mixture of spindle-shaped cells (myofibroblasts and fibroblasts) and inflammatory cells (predominantly plasma cells and lymphocytes with scattered neutrophils and eosinophils). The findings are indicative of inflammatory pseudotumor

Fig. 4

Rapid regression of incidental inflammatory pseudotumor of the liver (IPTL) in a 67-year-old Caucasian woman. MR was performed after US examination for fever and abdominal pain in March 2012, detecting a large hypoechoic lesion of the liver (not shown). MR confirmed the lesion with inhomogeneous signal hyperintensity on fat sat T2W (a) and high b-value DW (b) images, hypointensity on hepatobiliary phase (c). With the histological diagnosis of IPTL (mixed inflammatory infiltrate and spindle cells on needle biopsy whip), corticosteroid therapy was undertaken at the beginning of April 2012. This allowed a subtotal regression of the lesion as early as May 2012, as shown on T2W (d), DW (e), and hepatobiliary phase (f). On the same sequences (g–i), a complete regression is seen on MR follow-up in September 2013

Fig. 5

Histologically proven inflammatory pseudotumor of the liver (IPTL) in the left hepatic lobe during MR follow-up in patient with focal nodular hyperplasia (partially visible in front of IPTL). In 2017, at the onset, IPTL shows high hydration in T2W (a), high b-value DW (b) images, and in the ADC map (c). Progressive inhomogeneous enhancement at peripheral starting is shown in the portal (d) and equilibrium (e) phases. After 24 months, size reduction of the lesion, T2 (f) signal hypointensity, spin mobile lessening on DW images (g) and ADC map (h), and hypovascularization in the portal (i) and equilibrium (j) phases are observed. On histological examination the main cell population is macrophages, intermixed with other chronic inflammatory cells, like plasma cells and lymphocytes. Some scattered hepatocytes are sequestered in the inflammatory infiltrate. S100 and CD1a stains: negative. Endothelial markers: negative. The findings indicate as the more likely the diagnosis of inflammatory pseudotumor

Fig. 6

IgG4-related inflammatory pseudotumor of the liver of a 58-year-old Caucasian male patient. MR T2W (a) and T2 SPAIR (b) images show a three-layered concentric targetoid aspect lesion with hyperintense core. CA administration confirms the targetoid aspect of the lesion on arterial (d) and portal (e) phase. On unenhanced T1W (c) and hepatobiliary phase (f), the lesion appears hypointense. Pathologic analysis of needle biopsy whip shows inflammatory infiltrates with polyclonal cells, myofibroblastic-fused cells, eosinophilic granulocytes, and band of fibrosclerotic tissue (g, hematoxylin–eosin stain, original magnification × 100). On IgG4 immunostaining (not shown) IgG4-positive plasma cells > 20/HPF. On Arginasi 1 coloration (h, × 40) infiltrates inflammatory cells (median area) with residual biliary duct, surrounded by normal hepatic parenchyma (upper and lower areas, brown colored) are shown

Fig. 7

MR patterns of inflammatory pseudotumors of the liver found both in literature and our experience