Post hepatectomy liver failure (PHLF) is one of the most severe complications after liver re-section. Preoperative evaluation of liver function is complicated and imprecise. The volume and function needed for each individual patient is unknown and the methods used for evaluation are uncertain. Preoperative MRI with Gadolinium may give dynamic information regarding liver function correlating with postoperative liver failure. A retrospective analysis will be performed regarding this topic.

Post hepatectomy liver failure (PHLF) is one of the most severe complications after liver re-section. Preoperative evaluation of liver function is complicated and imprecise. The volume and function needed for each individual patient is unknown and the methods used for evaluation are uncertain. Arbitrary volume limits have been postulated but still there are patients dying of PHLF with volumes well above these limits.

On multivariate analyses, age ≥70 years, pre-operative chemotherapy, steatosis, resection of >3 segments, vascular reconstruction and intraoperative blood loss >300 mL significantly increased the risk of PHLF. Combining ISGLS grades B and C groups resulted in a high sensitivity for predicting mortality compared to the 50-50 rule and Peak bilirubin >7 mg/dL.

For patients with low volume of the future liver remnant (FLR) there are techniques to in-crease the volume prior to resection. Portal vein embolization (PVE) is an established meth-od, whereas associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) is relatively new. Augmentation of the volume of the FLR is indicated for patients with FLR less than 20%, provided no intrinsic liver disease exists, 30% after chemotherapy and 40% for patients with compromised liver function, for example liver cirrhosis.

Both methods, PVE and ALPPS, have high risks of morbidity and mortality, especially post hepatectomy liver failure (PHLF), why other alternatives are evaluated. Liver venous depri-vation (LVD), adds liver vein/s occlusion with a plug/plugs when PVE is performed or after PVE. Preliminary data indicate that the growth rate with this percutaneous method is equivalent to ALPPS with function following in parallel. Still all three methods add one extra treatment stage to increase liver volume of the FLR but still suffers the risk of PHLF. Furthermore, a large group of patients is treated with one stage hepatectomy where the volume and function also may be limited.

Liver volume estimation has been the main method to ensure safe resections, but as many studies have explored functional tests with the goal of increasing safety. The most used ones are Indocyanine green (ICG), maximum liver function capacity (LiMax) , Child Pugh (CP) score and Hepatobiliary scintigraphy (HBS). Still, none of these, alone or in combination have been able to completely ameliorate this dreadful complication. In addition, the cut-off levels for resection for each of these methods are arbitrary, mainly because PHLF is relative-ly rare as are major liver resections. Furthermore, all the tests are time consuming, costly and demanding for the patients.

The development of liver gadoxetate (=Primovist) MRI has given hope that this method may provide similar information as that of hepatobiliary scintigraphy (HBS), but with a higher resolution of liver function on a segmental level and at the same time enable liver volume estimation of the FLR as well as tumour burden and anatomy. In addition, MRI may also add quantitative information regarding steatosis, fibrosis (MRE) and iron content, factors previ-ously not included in any test used to estimate liver function.

So, with one modality most of the information gathered with all the other methods could possibly be incorporated into one.

Data regarding the increase in volume, as well as function, during the first week after PVE/ALPPS/LVD is sparse. Previous studies have shown that after PVE the increase in func-tion in the non-embolized lobe is larger than the increase in volume. The opposite has been shown for patients operated with ALPPS where the increase in volume was larger than the increase in function.

HBS has become common in several centers to estimate the liver function prior to major hepatectomy, including TSH and ALPPS. One disadvantage is the relatively poor spa-tial resolution which therefore necessitates another radiological study to assess the tumor burden in the liver. Dynamic gadoxetate MRI has been shown to be comparable with HBS to assess the liver function, with a very strong correlation between the two methods. Also, in PVE patients, it strongly indicates the risk of PHLF after resection when there is no in-creased enhancement in the FLR after 2 weeks.

Liver MRI is increasingly used prior to liver resection for analysis of liver tumour burden and anatomy. The long-term goal of this project is to find simple functional measures that can be obtained from the standard clinical MRI used today in everyday clinical practice. In this way we would ultimately have a method to both evaluate segmental function (gadoxetate), fat and iron storage, fibrosis, volume, tumour burden and anatomy.

Study aim The overall aim of this retrospective study is to evaluate if a preoperative MRI with liver specific contrast agent (gadoxetate, aka Primovist) add segmental functional information and if this information can be used to predict PHLF more accurate than with FLR calculation alone. Thus, the result of the functional analysis will be used to identify factors in the MRI examination that correlates to the risk of PHLF. Based on the gadoxetate MRI data, a pro-spective clinical trial will be designed afterwards to validate the results of this retrospective study. The final goal is to establish a pre-operative workup based on a Primovist MRI proto-col, providing a robust and predictive functional and volumetric measure in all preoperative examinations. This might contribute to a more accurate risk assessment of patients sched-uled for hepatcetomy, thus reducing both post-operative, PHLF related morbidity and mortality.

Inclusion Criteria:

1. Patients who have had a Primovist MRI within 8 weeks before hemihepatectomy or extended hemi-hepatectomy +/- bile duct anastomosis independent of diagnosis.

2. Patients with cirrhosis who have had a Primovist MRI within 8 weeks before resection of more than one liver segment

3. Patients who had liver volume augmentation with a pre-operative Primovist MRI less than 2 weeks before resection

4. Patients who have had a Primovist MRI within 6 weeks be-fore any liver resection where PHLF or death within 90 days occurred.

   -

Exclusion Criteria:

1. <18 years of age
2. Resection was not performed -