Native T1 mapping for assessment of the perilesional zone in metastases and benign lesions of the liver

Adjacent to hepatic metastases, liver parenchyma is often histopathologically altered even if its visual appearance on native magnetic resonance (MR) images is blunt. Yet, relaxation properties in MR imaging may show structural changes prior to visual alteration, and therefore, the aim of this study was to investigate whether T1 relaxation times in the perilesional zone differ between metastases and benign lesions. A total of 113 patients referred for MRI were included prospectively. Images were assessed for metastases, solid benign lesions and cysts, and regions-of-interest were drawn on T1 maps including the focal lesion and a close (inner perilesional zone = IPZ) and a larger perilesional zone (outer perilesional zone = OPZ). Simple ratios between these zones, as well as a gradient ratio between the IPZ and the entire perilesional zone (EPZ) were calculated. Within the collective, 44 patients had lesions of one or two entities. For metastases, the simple ratio between IPZ and OPZ as well as the mean EPZ gradient was significantly higher than for both solid benign lesions and cysts. Lesion size was not a significant covariate. We conclude, that native T1 properties of the perilesional zones differ significantly between malignant and both solid and cystic benign lesions.

It is well known that liver parenchyma adjacent to hepatic metastases is often altered due to edema, vascular or bile duct proliferations 1,2 , as well as due to malignant infiltration in some cases. Whereas appearance on native images is often blunt, an imaging correlate for this can sometimes be seen as a so-called rim enhancement 3 . Due to the benign nature, perilesional alterations should not be as pronounced around non-malignant hepatic lesions, and especially around cysts, which should-apart from a possibly compressing effect in very large types-not have any effect on perilesional relaxation times. Therefore, to quantify differences in the perilesional zone could be of value for confident diagnostic decision making.
In general, a native approach to quantify the perilesional zone would be preferable to a contrast-enhanced approach as recent research [4][5][6] as well as special contraindications in selected patients make a reduced use of MR contrast agents desirable but also as the contrast-enhanced technique can be influenced by other aspects, such as size and vascularization of the lesion 7 , or by the type and amount of the contrast agent. Whereas vascular proliferation or sinusoidal congestion are probably best to be detected under appliance of contrast agents, most of the alterations in the perilesional zone found in histopathology, i.e. edema and fibrosis, can also be detected by native imaging techniques.
For tissue characterisation in cardiac imaging, native T1 mapping has shown its value allowing for precise quantification at rather short imaging times 8,9 . In liver imaging, mapping techniques have mainly been evaluated using the hepatocyte specific contrast agent gadoxetate disodium [10][11][12] whereas native T1 mapping techniques are not yet well established yet. Therefore, the study was set up to evaluate whether the on native imaging visually inconspicuous liver parenchyma adjacent to a focal lesion shows altered T1 relaxation times compared to the more peripherally located liver parenchyma and, if so, whether there is a difference in terms of this gradient between metastases and solid benign lesions as well as cysts.
Imaging protocol. Images were acquired on a clinical 1.5 T MR scanner (Avanto; Siemens Healthineers, Erlangen, Germany) with a 16-channel body-phased-array coil. All included patients underwent standard liver MR imaging using the hepatocyte specific contrast agent gadoxetate disodium (Gd-EOB-DTPA, Primovist) which includes an axial T1-weighted spin echo sequence, an axial fat-saturated T2-weighted turbo spin echo sequence acquired with a 2D navigator for abdominal imaging (2D Prospective Acquisition CorrEction, PACE), an axial T1-weighted dual echo sequence, and axial T1 GRE (gradient recalled echo) sequences for dynamic imaging before and 15, 55 s and 2, 5, 10 and 20 min after contrast agent administration and a coronally orientated T1 GRE sequence for the hepatobiliary phase at least 20 min after contrast agent administration 13 .
Study sequences. Apart from the clinical routine image protocol, patients received native steady-state precession readout single-shot Modified Look-Locker inversion recovery (MOLLI) sequences in the axial plane.
T1 maps were calculated automatically on a pixel-by-pixel basis, and displayed on a 12-bit lookup table with a visible color-coded map, on which the signal intensity of each of the pixels reflects their absolute T1 value.
The imaging parameters for the study sequence are shown in Table 1.
Image analysis. All imaging sequences were analysed on standard workstations (Centricity PACS, Radiology RA1000, General Electrics). Region-of-interest (ROI) placement was done manually by a reader blinded to the diagnosis of the focal lesion according to the following approach: first, a ROI entirely encircling the focal lesion (or cyst) was placed on the T1 map, adding-if possible-a circumferent narrow safety distance of 1-3 mm. Then, in a circumferent distance in the adjacent liver parenchyma of 5 mm, a larger ROI (zone 1) was added, entirely encircling the first ROI under avoidance of larger vessels, other focal lesions, and extrahepatic structures. The third and largest ROI (zone 2) www.nature.com/scientificreports/ was drawn including the two other ones under adherence to the same requirements. T1 relaxation times as well as the area of the ROIs (in mm 2 ) were recorded. A sketch of the approach is given in Fig. 1. Additionally, Figs. 2, 3, 4 show examples of ROI placement in a patient with metastases, in a patient with cysts, and in a patient with benign liver lesions.
Relaxation times of the inner perilesional zone (IPZ), as well as of the outer perilesional zone (OPZ) were calculated according to the following formula.  www.nature.com/scientificreports/ Additionally, for comparison, to minimize the impact of possible partial volume effects in the inner zone caused by the lesion, the relaxation time of the entire peripheral zone, including the inner perilesional zone was calculated.
The simple ratio (R1) between the IPZ and the OPZ was calculated as follows: The EPZ gradient (R2), as a measure for the relation between the inner perilesional zone (IPZ) and the entire perilesional zone (EPZ), was calculated as follows: In order to test the robustness of the measurement technique, a random sample of 10% of test patients (n = 12 lesions in 12 different patients) was repeated by the same reader after an appropriate time interval.
Statistical analysis. All statistical analyses were performed using SPSS Version 25 (IBM Corporation, New York, USA). Means and standard deviation of T1 relaxation times as well as areas of measured ROIs and secondary values discussed above were calculated for metastases, benign lesions and cysts. Comparisons of the three lesion types were performed with generalised estimating equations (GEE) using the identity link function with  www.nature.com/scientificreports/ robust variance estimate. Since the number of lesions considered for each patient ranged from 1 to 7, data are not independent. GEE models account for within-subject correlations arising when outcomes are measured on the same patient more than once 14 . For each outcome, the three group comparisons with metastases vs. benign lesions, metastases vs. cysts and benign lesions vs. cysts were performed. Lesion type was defined as a factor and lesion area was defined as a covariate. In order to avoid inflation of the α-level due to multiple testing, the global null hypothesis was tested, that all three lesion types have the same mean first, and the three group comparisons were only considered, if the global test was significant. The two-sided level of significance was α = 0.05. Intrareader reliability was tested using intra-class correlation coefficients ICC 15 . Receiver operating curve (ROC) analyses were performed to determine the capacity to distinguish between malign and benign (both solid and cystic) lesions.

Results
All in all, 44 patients within the collective had focal liver lesions of one or two entities. 28 patients showed only metastases on the T1 maps, 2 patients had metastases and cysts, 5 patients had benign lesions, one had a benign lesion and a cyst, and 8 patients had only cysts. The hepatic metastases displayed on the T1 maps in 30 patients were due to cholangiocellular carcinoma (n = 2), neuroendocrine carcinoma (n = 8), colonic cancer (n = 6), breast cancer (n = 5), squamous cell carcinoma (n = 2), angiosarcoma (n = 1), melanoma (n = 1), extragonadal tumour (n = 1), renal cell carcinoma (n = 1), and pancreatic tumour (n = 3). Thereof, several patients showed numerous lesions, leading to an overall number of 83 measureable metastases. 6 patients had one or several hepatic cysts, which led to 26 cysts displayed on the T1 maps. Benign solid hepatic lesions were present in 5 patients, which led to 17 lesions displayed on the T1 maps, as several patients had multiple lesions. These were hemangioma (n = 4), adenoma (n = 5), and focal nodular hyperplasia (n = 8). The size of the ROI of the benign lesions was 489.5 mm 2 (181-1,567 mm 2 , SD 350.9), of zone 1 906.6 mm 2 (415-2,036 mm 2 , SD 468.7), and of zone 2 1,389.2 mm 2 (637-2,705 mm 2 , SD 538.0). Thus, the area of the IPZ was 417.1 mm 2 , and of the OPZ 482.6 mm 2 . The EPZ measured 899.7 mm 2 , accordingly.

Metastases. Mean relaxation
Comparison. For metastases, the simple ratio between the inner and the outer peripheral zone was 1.07 (0.98-1.37, SD 0.07). For benign lesions the simple ratio was 1.03 (0.97-1.18, SD 0.05) and for cysts, it was 1.02 (0.92-1.09, SD 0.05). The difference was statistically significant for metastases vs. benign lesions (p = 0.012) and metastases vs. cysts (p = 0.001) but not for benign lesions vs. cysts (p = 0.606). For all comparisons, lesion size was not a statistically significant covariate (p > 0.05).
The EPZ gradient, as a measure for the relation between the inner and the entire perilesional zone was 3.8 for metastases (− 1.4 to 23.0, SD 4.1). For cysts, the EPZ gradient was 0.8 (− 5.4 to 5.1, SD 2.8), and for benign lesions, it was 0.9 (− 7.1 to 5.8, SD 3.3). Again, the difference was statistically significant for metastases vs. benign lesion (p = 0.001) and cysts (p < 0.001) but not for benign lesions vs. cysts (p = 0.887). Lesion size was not a significant cofactor in any of the comparisons.
Intra-reader reliability as a measure of robustness of the measurement technique was excellent with a mean ICC of 0.976 (range 0.952-0.998, p < 0.001).
Mean relaxation times and ratios are given in Table 2. Additionally, a graphical representation of the findings regarding the EPZ gradient is given as Fig. 5.

Discussion
In the present study we evaluated whether visually blunt liver parenchyma adjacent to a focal lesion shows altered T1 relaxation times compared to the more peripherally located liver parenchyma, and, if so, whether this gradient differs between metastases, benign solid lesions and simple cysts, which we estimated to be acceptable as a negative control. Our results show that there is a gradient in the parenchyma surrounding focal lesions with Scientific RepoRtS | (2020) 10:12889 | https://doi.org/10.1038/s41598-020-69819-w www.nature.com/scientificreports/ higher T1 values directly adjacent and lower T1 values more peripherally, and that this gradient is significantly higher adjacent to metastases compared to benign solid lesions. Characterisation of different compositions of parenchyma by T1 and T2 mapping techniques is well established in cardiac imaging 16 . Whereas T2 relaxation time is mainly influenced by water, and thereby primarily reflects edema, T1 relaxation time is said to reflect changes in water content as well as in the local molecular environment 17 . In acute myocarditis, T1 mapping has even been found to be superior to T2 weighted CMR in detecting myocardial edema in the context of acute myocarditis 8 , even though other studies admit, that there is still no clear advantage of one mapping sequence over the other 18 .
Until now, the peritumoral zone in the liver has primarily been assessed by contrast-enhanced approaches, as peripheral rim enhancement of lesions on early phase post-contrast images was described as a characteristic finding of metastatic lesions 2,7,19,20 . As an underlying etiology malignant or inflammatory cell infiltration, desmoplastic reaction, surrounding parenchymal compression, and vascular as well as bile duct proliferation were discussed, which could be proven with histologic specimens in the experimental setting 19 , as well as in patients 2,20 . Apart from primarily vascular alterations leading to the hyperintense rim, i.e. peritumoral congestion, probably due to a bloodpool effect 2 , the majority of the other cellular and pericellular alterations named above and described histologically, probably also lead to alterations in native relaxation time.
In either way, the most obvious reason for the more prominently altered T1 relaxation time around metastases could be, that usually malignant lesions cause a surrounding edema, either by congestion, or inflammation 2,19 . Additionally, as T1 relaxation time is directly related to the diffusion coefficient of protons 21 , another reason for a gradient in T1 relaxation time surrounding focal lesions might be the mass effect, going along with flattening of parenchyma, which would also explain the slight gradient around benign lesions including cysts in some cases, although lesion size was not a significant covariate in the analyses we performed.
Approaches to classify focal lesions without contrast-agent are desirable not only due to the recent publications on safety issues concerning MR contrast agents 5,6,22 . But also for well-known risks including allergic reactions. The approach presented here might be a first step towards a new methodology, especially as more sophisticated computational approaches such as texture analysis show initial positive results for textural changes before their visual appearance, irrespective of the question whether the textural changes are a result of the metastatic cells themselves or rather a reflection of reactive changes in the surrounding liver parenchyma 23 .

Limitations
Despite careful planning and execution, there are some limitations to our study that need to be discussed. Firstly, we had a rather small number of cysts and benign lesions in our study cohort, as patients only underwent exams for exclusion and assessment of malignant lesions, meaning that all benign lesions included represent mere incidental findings. Therefore, the study sample size is not large enough to provide a definite cut-off value for differentiation between benign and metastatic lesions, which also means that descriptors of diagnostic accuracy, such as sensitivity and specificity, could not be calculated. Another limitation is that histology cannot be given as Table 2. Mean relaxation times of the different focal liver lesions and their calculated ratios.  Figure 5. Box-and-whisker plot depicting the IPZ/EPZ gradient. The whiskers indicate variability outside the upper and lower quartiles, the individual points demonstrate outliers.