One of the most recent and interesting field of diagnostic imaging is diffusion-weighted MR imaging (DW-MRI). Various studies evaluated the application of DW-MRI to diffuse liver disease and focal liver lesions providing controversial results, probably due to the difficult reproducibility of the apparent diffusion coefficient (ADC) measurements. It is conceivable that a wide inter/intra-individual variability actually exists in the apparent diffusion coefficient (ADC)-values, and that each apparent diffusion coefficient (ADC)-value presents an higher reliability in measuring the temporal changes of water diffusion within the same individual (longitudinal-evaluation), than in characterizing tissues between different patients (transverse-evaluation). For these reasons, some previous studies assessed the application of DW-MRI in predicting the chemotherapy (CHT) outcome in liver metastases. The rationale of these studies was the overt biochemical changes shown by the neoplastic cells after CHT and the sensitivity of DW-MRI in the identification of such changes. The same authors noticed that the metastatic lesions with the lowest ADC-values present also the best outcome after CHT. Moreover, these studies suggest that it could be possible to assess if each single patient will respond (R) or not (NR) to the CHT through liver DW-MRI performed from 3 days to 3 weeks after the beginning of CHT.

In R patients an increase of the mean diameter of the neoplastic cells should be present from the first week of CHT with a consequent reduction of the ADC-value. Then, the ADC-value should increase after 2-3 weeks due to the reduction of the cellular volume and to the increase of the membrane permeability. In NR patients, these biochemical changes should not be present with a consequent constant ADC-value before/after CHT. Our multicentre study will allow the recruitment of a higher number of patients vs. previous studies. Moreover, presently the outcome of CHT in liver metastases is evaluated only after 2-3 cycles of CHT according to the RECIST (Response-Evaluation-Criteria-In-Solid-Tumors) and then merely on a dimensional basis. Anyway, frequently liver metastases in R patients, can not show any significant dimensional change, while they present several biochemical/metabolic changes, not included in the RECIST, but probably detectable by DW-MRI.

The principal objective of the study is the early assessment of CHT outcome in liver metastases and advanced hepatocellular carcinoma (HCC). Patients with liver metastasis will be stratified as R and NR on the basis of the contrast-enhanced CT examination performed 20 days after the beginning of the second cycle of CHT; patients with HCC undergoing therapy with Sorafenib will be stratified as R and NR on the basis of the contrast-enhanced CT examination performed 90 days after the beginning of the therapy. The ADC-values obtained before and after the first CHT cycle will be compared between each patient category to assess the grade of agreement between the dimensional and functional parameters. Moreover, the investigators' aim is to assess whether those liver metastases or HCCs presenting a lower ADC-value before CHT are actually more responsive to CHT in comparison to liver metastases and HCCs presenting a higher ADC-value.

Secondary objectives: to propose some additional functional criteria to the RECIST 1.1 and mRECIST (for HCC) criteria based on water diffusion and biochemical changes of the neoplastic cells. As a further objective it is possible to hypothesize a different response to CHT of the different tumor histotypes detectable from the different ADC changes induced by CHT.

Patient population

Inclusion criteria:

• of age, compliant, patients enrolled for CHT, without major contraindications to the MR examination;
• non-confluent liver metastases, from every primary carcinoma histotype biopsy/surgical-proven, without intralesional necrosis/calcification involving >30% of their volume;
• multiple confluent hepatocellular carcinomas, histotype biopsy/surgical-proven in prevision of treatment with Sorafenib;
• at least one marker lesion allowing reproducible ADC measurements, i.e. placed at the level of the lower right liver segments;
• detection/enrolment by contrast-enhanced CT before CHT that allow to define the lesion size or the gross parenchymal involvement (if HCC).

Each patient will sign an informed consent, after the procedure will be completely explained.

For the metastasis: three diameter of each marker lesion will be measured, and the mean/minimal/maximal ADC±standard deviation will be quantified by region-of-interests (ROIs) placed within the lesion avoiding lesion margins and the necrotic/intratumoral calcification areas. All measurements will be repeated for three times even at the level of the adjacent liver parenchyma (within 3 cm from the lesion margins, keeping a ROI diameter >2 cm). Consequently, the absolute values (s/mm2) of ADC, and the ADC percentages vs. the adjacent liver parenchyma measured at the different times will be compared.

For HCC: three diameter of gross parenchymal involvement will be measured, and the mean/minimal/maximal ADC±standard deviation will be quantified by large region-of-interests (ROIs) placed within the liver lobe containing the involvement. All measurements will be repeated for three times even at the level of the adjacent normal liver parenchyma (within 3 cm from the lesion margins, keeping a ROI diameter >2 cm). Consequently, the absolute values (s/mm2) of ADC, and the ADC percentages vs. the adjacent liver parenchyma measured at the different times will be compared.

Imaging

For metastasis: patients will be scanned by DW-MRI and contrast-enhanced CT before the beginning of CHT (Time 0). The time between the initial MRI and contrast-enhanced CT should not be superior to one week. MRI examination will be repeated within one week (Time 1) and 20 days from the beginning of the first cycle of CHT (Time 2), and 20 days from the beginning of the second cycle of CHT (Time 3).

For HCC: patients will be scanned by DW-MRI and contrast-enhanced CT before the beginning of CHT (Time 0). The time between the initial MRI and contrast-enhanced CT should not be superior to one week. MRI examination will be repeated after 30 days (Time 1), 60 days and 90 days (Time 3) from the beginning of the CHT.

Contrast-enhanced CT will be performed contemporarily or within one week after the last MRI examination.

Contrast-enhanced CT examination will be performed according to an established protocol by using a 16/64-row equipment according to the centre involved, contrast bolus-track technology, slice- thickness reconstruction of 3 mm, before/after ev. injection of iodinated contrast agent (3 mL/s), during arterial/portal phase.

All MR examinations will be carried out using the following 1.5-T units:

• Gyroscan ACS NT Intera Release 12 (Philips, Eindhoven, The Netherlands) (Trieste and Florence), gradient strength, 30 mT/m; slew rate, 120 T/m/s; six-channel phased array multicoil;
• Magnetom Avanto (Siemens, Erlangen, Germany) (Treviso and Napoli), gradient strength, 45 mT/m; slew rate, 200 T/m/s; 2 phased-array coils with 18 elements. The different MR equipments employed by the different centers will be calibrated by a dedicated phantom.

The phased array multicoil will be adequately positioned to cover the upper abdomen of the subject lying in a supine position, the arms extended over the head to avoid artifacts. Patients, fasting from 4 hours, will be instructed to maintain a constant respiration depth, even with the possibility to use exogenous oxygen delivery to avoid deep respiration. All acquisitions will be obtained by single-shot sequence to obtain immediately/automatically the ADC-values.

The protocol included the following acquisitions:

1. T2-weighted half-Fourier single-shot turbo spin-echo (HASTE) free-breath sequence; transverse/coronal plane; TR/TE, 810/80 ms; echo-train length, 69; slice number, 40; slice thickness, 5 mm; intersection gap, 10%; field of view, 300-420 mm; effective matrix size, 256 x 165; number signal averages (NSA), 1; total acquisition time, 2-3 min;
2. T1-weighted 2D gradient echo in/out phase breath-hold sequence; transverse plane; TR/TE, 231-121/ 14.6-2.3 ms; slice thickness, 5 mm; slice number, 24; intersection gap,10%; flip angle, 80°; sense factor, 1.5; field of view, 300-420 mm; effective matrix size, 256 x 165; NSA, 1; total acquisition time, 18 s;
3. D-weighted echo-planar imaging (EPI) single-shot free-breath sequences will be acquired on transverse plane with variable EPI factor and the following parameters. Fat suppression will be obtained by spectral pre-saturation inversion recovery. Isotropic motion probing gradients will be applied for each DwI acquisition and for each b-value will be obtained images and corresponding ADC map.

The investigators presently define as R those patients who show a reduction of the liver metastasis or HCC diameter ≥30% on contrast-enhanced CT at three weeks after the beginning of the second CHT cycle; if not, it was considered NR. Changes in tumor size after treatment were calculated by using the formula % Vend =(VB -Vend)/VBx100, where VB was lesion size before treatment (maximum transverse diameter) and Vend was lesion size 20 days after the second administration.

On the basis of the dimensional reduction of liver metastases and parenchymal involvement (for HCC) on contrast-enhanced CT on Time 3 vs. Time 0 scan, each patient will be classified as R or NR according to the RECIST and mRECIST criteria. Afterwards, on the basis of the ADC-values measured during the different MR examinations, the inter/intra-individual ADC-values will be compared to the results of contrast-enhanced CT to assess the relation between reduction of the liver metastasis diameter and:

• increase of the ADC-value on Time 3 (after the end of CHT);
• reduction of the ADC-value on Time 1 (very early assessment);
• increase of the ADC-value on Time 2 (early assessment); and to assess whether the lesions with the highest pre-treatment ADC-value present also the highest dimensional reduction and the highest ADC-value at the end of CHT; In each center each evaluation will be performed three times by two blinded observers (all trained how to place the ROI by an inter-centre conference) to assess the reproducibility of all measurements. The observers who will assess the MR images will be different from the observers assessing the CT images and will not be aware about the size changes after CHT. The investigators will perform data mathematical fitting on multi-b DW-MRI data sets to calculate the true diffusion and the perfusion fraction.

Statistical analysis will employ linear regression analysis to assess the association between the ADC-value changes and the CHT outcome.

The transferability of the results of the present study to the clinical practice will be possible after the achievement of the primary objective, corresponding to the possibility of stratifying patients as R and NR to CHT through DW-MRI just after one week of treatment and/or with the pre-treatment ADC assessment. Nowadays, it is necessary to wait 20-30 days after the end of the second or third CHT cycle to know the individual outcome of CHT. This determines 2-3 months of ineffective therapy with consequent avoidable pain for the patients related to drug administration, and unusable costs. The possibility to know the response of each patient to CHT well in advance will allow avoiding vain drug administration to patients who could attempt a different treatment or drug combinations reducing treatment costs.

The clinical transferability of the present study will be performed after the achievement of one or both the secondary objectives, corresponding to the identification of functional criteria based on water diffusion and biochemical features of the neoplastic cells which can be proposed as additional or alternative criteria to the RECIST 1.1 and mRECIST, and to the identification of a different response to CHT of different tumor histotypes. If only one of these objectives will be reached this study will achieve an important result, allowing a more correct assessment of individual response to CHT.

Inclusion criteria:

• of age, compliant, patients enrolled for CHT, without major contraindications to the MR examination;
• non-confluent liver metastases, from every primary carcinoma histotype biopsy/surgical-proven, without intralesional necrosis/calcification involving >30% of their volume;
• at least one marker lesion allowing reproducible ADC measurements, i.e. placed at the level of the lower right liver segments;
• multiple confluent hepatocellular carcinomas, histotype biopsy/surgical-proven in prevision of treatment with Sorafenib;
• detection/enrolment by contrast-enhanced CT before CHT that allow to define the lesion size or the gross parenchymal involvement (if HCC)

Each patient will sign an informed consent, after the procedure will be completely explained.

For the metastasis: Three diameter of each marker lesion will be measured, and the mean/minimal/maximal ADC±standard deviation will be quantified by region-of-interests (ROIs) placed within the lesion avoiding lesion margins and the necrotic/intratumoral calcification areas.

For the hepatocarcinoma: Three diameter of gross parenchymal involvement will be measured, and the mean/minimal/maximal ADC±standard deviation will be quantified by large region-of-interests (ROIs) placed within the the lobe containing the involvement.

All measurements will be repeated for three times even at the level of the adjacent liver parenchyma (within 3 cm from the lesion margins, keeping a ROI diameter >2 cm). Consequently, the absolute values (s/mm2) of ADC, and the ADC percentages vs. the adjacent liver parenchyma measured at the different times will be compared.

Patients with liver metastases in chemotherapy and with hepatocarcinoma in therapy with Sorafenib