If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Human Structure and Repair, Ghent University Faculty of Medicine, Ghent, BelgiumDivision of General, HPB and Transplant Surgery, Dept. of Clinical Medicine and Surgery, Federico II University Hospital, Naples, Italy
This study aimed to analyze the predictive value of Hepatobiliary scintigraphy (HBS) for posthepatectomy liver failure (PHLF) after major liver resection with a comparison to assessment of liver volume in a multicenter cohort.
Methods
Patients who underwent liver resection after HBS were included from six centers. Remnant liver volume was calculated from CT images. PHLF was scored and graded according to the grade B/C ISGLS criteria.
Results
In 547 patients PHLF incidence was 10% (56/547) and 90-day mortality rate 8% (42/547). Overall predictive value of remnant liver function was 0.66 (0.58–0.74) and similar to that of remnant volume (0.63 (0.72). For biliary tumors, a function cut-off of 2.7%/min/m2 and 30% volume cut-off resulted in a PHLF rate 12% and 13%, respectively. While an 8.5%/min (4.5%/min/m2) function cut-off resulted in 7% PHLF for those with a function above the cutoff while a 40% volume cutoff still resulted in 14% PHLF rate. In the multivariable analyses for PHLF, liver function was predictive but liver volume was not.
Conclusion
The current study shows that preoperative liver function assessment using HBS is at least as predictive for PHLF as liver volume assessment, and likely has several advantages, particularly in the high-risk sub-group of biliary tumors.
Introduction
Posthepatectomy liver failure (PHLF) remains the single most important complication after major liver resection and is the main driver of postoperative mortality.
Prospective evaluation of the International Study Group for Liver Surgery definition of post hepatectomy liver failure after liver resection: an international multicentre study.
Regardless of the definition, there is no effective treatment available and the mortality associated with PHLF is high, which is why prevention of PHLF is key.
Relevance of the ISGLS definition of posthepatectomy liver failure in early prediction of poor outcome after liver resection: study on 680 hepatectomies.
Preoperative assessment of the future liver remnant (FLR) can estimate the risk of PHLF after surgery. The FLR can be quantified through various methods, of which volume assessment remains the current standard.
But liver volume remains a surrogate for liver function. This may work quite well under the assumptions of homogenous distribution and normal total liver function. Quantitative liver function testing, for instance hepatobiliary scintigraphy (HBS) with technetium labeled mebrofenin has been suggested to improve the predictive value for remnant liver function.
The advantage of HBS over other functional tests such as indocyanine green (ICG) clearance and LIMAX (13C-methacetin breath) metabolic tests is that it does not only measure total function but also shows its distribution. This is especially relevant in patients with deformed anatomy, after portal vein embolization or other augmentative procedures, underlying liver or biliary disease and after chemotherapy. Current evidence on HBS comes from single center series and does not definitively support widespread implementation of this technique.
This study aimed to analyze the predictive value of hepatobiliary scintigraphy for PHLF after major liver resection with a comparison to assessment of remnant liver volume in a multicenter design.
Methods
All patients who underwent major liver resection (defined as the resection of at least three Couinaud liver segments)
after assessment of liver function with mebrofenin-labeled HBS between 2000 and 2020 were included. Six experienced liver surgery centers included all consecutive cases. For Amsterdam UMC the indication for HBS was a planned right hepatectomy for any indication, any major resection for perihilar cholangiocarcinoma, or any major liver resection for any indication that was considered to have a high risk of PHLF. In Ghent and Bologna, HBS is performed before every major liver resection. In Birmingham, Lille and Antwerp, HBS is performed for every major liver resection that is considered to have a high risk of PHLF. The study protocol was reviewed by the medical ethics committee of the Amsterdam UMC and there was no need for ethical approval or individual informed consent according to Dutch law. All data was handled anonymously.
Hepatobiliary scintigraphy was performed according to the standard protocol developed at the Amsterdam UMC and applied with minor local adaptations in each center.
Only patients who underwent SPECT in order to quantify regional liver function were included. All preoperatively calculated liver functions were adjusted to be in accordance with the actually performed surgical resection. FLR function was expressed as %/min/m2. All patients for whom no FLR volume calculation was available were excluded. Liver volume was measured on CT images and FLR volume was expressed as the percentage of total liver volume. Alternatively standardized FLR (sFLR) volume was calculated according to the formula of Vauthey et al.
For patients who underwent preoperative portal vein embolization the liver function and volume measurements after embolization were used in the analyses.
Liver failure was defined and graded according to the International Study Group of Liver Surgery (ISGLS), and only grades B and C were considered to be of clinical relevance.
Biliary leakage and hemorrhage were also defined and graded according to the respective ISGLS criteria, and only grades B and C were included as clinically relevant. Ascites was defined as drainage of more than 200 cc/24 h after postoperative day 3 with an operative drain or abdominal distention with ascites seen on ultrasound. Grade B ascites was defined as the need for invasive treatment without general anesthesia or more than 1000 cc/24 h after postoperative day 3 in patients with an operative drain. Grade C was defined as ascites that required treatment under general anesthesia or treatment in the intensive care.
All categorical data were presented as numbers with percentages and differences were tested using chi-square of Fisher's exact tests where appropriate. All continuous data were presented as medians with inter-quartile-ranges (IQR) and differences were tested using Mann–Whitney tests. Predictive value of variables was tested using receiver operating characteristic analyses and area under the curve values were shown with 95% confidence intervals (95%CI). Multivariable analysis was performed using binary logistic regression with all variables with a P value below 0.100 at univariable analyses included. All statistical analyses were performed using SPSS (Version 26.0, IBM, Chicago, IL) and figures were generated using Graphpad Prism (Version 9, Graphpad Inc., La Jolla, CA).
Results
In the study period 581 patients underwent preoperative hepatobiliary scintigraphy before liver resection. For 37 patients preoperative liver volume was not available and were excluded. A total of 547 i patients were included in the analyses. Baseline characteristics are shown in Table 1. The reason for liver surgery were colorectal liver metastases in 174 patients (32%) and perihilar cholangiocarcinoma in 146 patients (27%). Right liver resection was performed in 335 patients (61%) of which 125 patients underwent extended right liver resection. Biliary reconstruction was performed in 200 patients (37%).
Postoperative outcomes are shown in Table 2. Liver failure grade B or C was diagnosed in 56 patients (10%) and 90-day mortality rate was 8% (42/547). Liver failure was associated with a 64% mortality rate, compared to 1% in patients who did not suffer liver failure (36/56 versus 7/488, P < 0.001).
In the overall cohort, the predictive value of FLR function, FLR volume, and sFLR volume for liver failure was similar (Fig. 1a). When using a 2.7%/min/m2 FLR liver function cutoff, the liver failure rate was 21% (24/112) in those below the cut-off, and 7% (32/435) in those above the cutoff (P < 0.001). When using 30% as a FLR volume cutoff, liver failure rate was 22% (21/97) in those below the cutoff, and 8% (35/450) in those above the cut-off (P < 0.001). The FLR volume and function for each individual patient in relation to liver failure is shown in Fig. 2a.
Figure 1(a) ROC curve analysis for liver failure in the entire cohort. (b) Volume and function in each individual patient. (c) ROC curve analysis for liver failure for non-biliary tumors. (d) ROC curve analysis for liver failure in biliary tumors.
Figure 2(a) Distribution of FLR volume and FLR function share across different types of liver resections. (b) Distribution of FLR volume and FLR function across different types of liver resections.
Abiding by the 2.7%/min/m2 FLR function and 30% FLR volume cut-off, 385 patients met both cutoffs (89%) with a 6% liver failure rate (24/385). Forty-seven (9%) did not meet both cutoffs with a 28% liver failure rate (13/47). FLR volume was sufficient, but function insufficient in 65 patients (12%), and 17% suffered liver failure (11/65). FLR volume was insufficient with sufficient function in the remaining 50 patients (9%) with a 16% liver failure rate (8/50). The predictive value of all volume and function parameters are shown in Table 3.
Table 3Sensitivity, specificity, positive predictive value and negative predictive value of commonly used liver volume and function parameters
Risk factors for failure were identified using uni- and multivariable analyses, which are shown in Table 4. While FLR and FLR function were both associated with liver failure at univariable analyses, only FLR function was identified as risk factor for liver failure at multivariable analysis. For patients with biliary tumors this did not reach statistical significance at multivariable analysis (P = 0.058), however, the odds ratio was 2.20.
Table 4Uni- and multivariable analysis for liver failure
Liver failure rate was high in the biliary tumor group with an incidence of 17% (36/207), compared to 6% patients with other diagnoses (20/340, P < 0.001). In the non-biliary tumor group the predictive value for liver failure was shown in Fig. 1c. In contrast to the entire cohort, the AUC value of sFLR volume was 0.78 (0.69–0.87), 0.70 (0.57–0.80) for FLR volume, and 0.69 (0.57–0.80) for FLR function. Using the 2.7%/min/m2 FLR function cut-off, liver failure rate was 12% for those below the cutoff (8/66) and 4% for those above the cut-off (12/274). Using the 30% FLR volume cut-off, liver failure rate was 13% for those below the cut-off (7/55) and 5% for those above it (13/285). Using a 30% sFLR volume cut-off, liver failure rate was 16% for those below the cut-off (11/68) and 3% for those above it (9/271). The predictive value of the parameters are summarized in Table 3.
The predictive value for patients with biliary tumors is shown in Fig. 1d, with values similar to that in the overall cohort. The 2.7%/min/m2 FLR function and 30% FLR volume cut-offs both resulted in a liver failure rate over 10% in patients meeting the cut-offs. For the 2.7%/min/m2 function cut-off, liver failure rate was 35% for those below it (16/46) and 12% for those above it (20/161). For the 30% volume cut-off, liver failure rate was 33% for those below it (14/42) and 13% for those above it (22/165). 23 patients were below both cut-offs and liver failure was 39% (9/23). Above both cut-off were 142 patients with 11% liver failure (15/142). Insufficient function and sufficient volume resulted in 30% liver failure (7/23) and sufficient function but insufficient volume in 26% liver failure (5/19).
A 40% FLR volume cut-off results in 22% liver failure for those below it (19/85) and 14% for those above it (17/122). A previously proposed 8.5%/min (unknot corrected for bodysurface area (BSA)) FLR function cut-off results in a 25% liver failure rate for those below it (30/119) and 7% for those above it (6/88). The 8.5%/min cut-off approximates 4.5%/min/m2 when the 8.5 is divided by the median BSA value. For patients with a biliary tumor, FLR volume above 40% and FLR function above 8.5%/min, liver failure rate was 6% (5/82). Those with FLR volume below 40% and function below 8.5%/min, liver failure rate was 23% (18/79). For patients with volume below 40%, but function above 8.5%/min, liver failure occurred in 17% (1/6). For patients with volume above 40%, but function below 8.5%/min, liver failure rate was 30% (12/40). The predictive value of the volume and function parameters are summarized in Table 4.
The distribution of FLR liver volume in relation to FLR function share (in %) was shown in Fig. 1b. The discrepancies between these parameters was low for left hemihepatectomies while the mismatch was greater for (extended) right hemihepatectomy. This difference was less pronounced when comparing FLR volume to absolute FLR function (in %/min/m2) (Fig. 2b).
Discussion
This series is the largest to date to report on HBS compared to volume to predict PHLF. A total of 547 patients who underwent major liver resection after assessment of both liver volume and function were included in the analysis. The most important finding is that, in the multivariable analyses for PHLF, liver function was predictive of PHLF but liver volume was not. In patients with biliary tumors the previously proposed 2.7%/min/m2 liver function and 30% liver volume cutoffs for the FLR resulted in a PHLF rate 12% and 13%, respectively. While a 8.5%/min (4.5%/min/m2) function cut-off resulted in 7% PHLF for those with a function above the cutoff, a 40% volume cutoff still resulted in 14% PHLF rate.
The additional value of mebrofenin-labeled hepatobiliary scintigraphy in the prediction of PHLF was shown in a prospective trial in 2010.
In this trial of 55 patients, the FLR function of 2.7%/min/m2 resulted in a positive predictive value of 57% and negative predictive value of 98%. These excellent predictive values should be interpreted in comparison to liver volume, a 30% volume cut-off for normal and 40% for compromised livers resulted in a positive predictive value of 44% and a negative predictive value of 95%. Therefore the additional predictive value of HBS was limited and the implementation in clinical practice has been slow. Nevertheless, subsequent studies have illustrated a reduction of PHLF and mortality following implementation of HBS.
The liver function value of 2.7%/min/m2 has not been truly validated to date. Since its introduction in 2010, this value has been the standard for liver function at the Amsterdam UMC and consequently there are very few patients undergoing surgery with a liver function value below 2.7%/min/m2 and therefore no statistical power to allow any further validation. Recently, several centers have implemented HBS in their preoperative work up and found similar results. A study on 38 patients reported that a FLR function of at least 1.9%/min/m2 could prevent PHLF grade B/C and a FLR function of 2.8% could prevent any grade PHLF.
The role of hepatobiliary scintigraphy combined with spect/ct in predicting severity of liver failure before major hepatectomy: a single-center pilot study.
Another larger study on 88 patients found a function of 2.3%/min/m2 effectively predicted PHLF with a positive predictive value of 92% and a negative predictive value of 97%.
Future remnant liver function estimated by combining liver volumetry on magnetic resonance imaging with total liver function on (99m)Tc-mebrofenin hepatobiliary scintigraphy: can this tool predict post-hepatectomy liver failure?.
Although these were not formal prospective validation studies, the similar results underscore the validity of the proposed liver function cut-off. The range from 1.9 to 2.8 indicates that it is probably too simplistic to use an absolute dichromate “cut-off value” as PHLF is often multifactorial.
The predictive value of liver volume might especially be insufficient in high-risk patients with compromised liver parenchyma. For instance, prolonged preoperative chemotherapy of more than 6 cycles is known to be associated with chemotherapy associated liver injury. This injury in turn increases morbidity and the risk of PHLF after major liver resection.
Extended preoperative chemotherapy does not improve pathologic response and increases postoperative liver insufficiency after hepatic resection for colorectal liver metastases.
Liver volume assessment alone does not provide insight in the chemotherapy induces liver injury, but a recent study showed that HBS can detect patients with compromised liver function secondary to chemotherapy associated liver-injury.
The impact of modern chemotherapy and chemotherapy-associated liver injuries (CALI) on liver function: value of 99mTc-labelled-mebrofenin SPECT-hepatobiliary scintigraphy.
Therefore HBS might be indicated in patients scheduled for major liver resection after extensive chemotherapy. Screening for underlying liver disease of damage using HBS might be able to prevent unnecessary liver biopsies.
Other high risk patients includes those with biliary tumors, who often suffer from obstructive cholestasis which is known to impair liver function. Although biliary drainage is common, major liver resection in patients with biliary tumors is still associated with high rates of PHLF and mortality.
Postoperative liver failure risk score: identifying patients with resectable perihilar cholangiocarcinoma who can benefit from portal vein embolization.
A recent study showed that the variation in anatomical liver volume and functional distribution is especially divergent in patients with primary tumors.
A previous study reported on the additional predictive value of HBS for patients with perihilar cholangiocarcinoma. The current data showed liver volume measures and the general 2.7%/min/m2 are not sufficient to prevent PHLF. This study confirmed that in order to perform a safe major liver resection, a higher FLR function of 8.5%/min (4.5%/min/m2) should be present. The importance of this finding is illustrated by the high PHLF and mortality rates in patients with biliary tumors that exceed the rates seen in other patient groups.
The final high-risk liver resection group that illustrates the additional predictive value of HBS is that of two-stage resections, in particular associating liver partition with portal vein ligation for staged hepatectomy (ALPPS). The initial introduction of ALPPS came with high mortality and PHLF rates, which might be attributable to the lack in functional increase in spite of the rapid volume growth associated with ALPPS.
Hepatobiliary scintigraphy to evaluate liver function in associating liver partition and portal vein ligation for staged hepatectomy: liver volume overestimates liver function.
Hepatobiliary scintigraphy to evaluate liver function in associating liver partition and portal vein ligation for staged hepatectomy: liver volume overestimates liver function.
Considering the mismatch in liver volume and function increase in ALPPS, HBS might be of upmost important when planning the second stage of ALPPS. The mismatch in liver volume and function was also evident in formal hepatectomies in this study, especially in resections other than left liver resections when comparing FLR volume to FLR function share. However, since FLR function share is a relative value and dependent on total liver function, the mismatch in volume and function is evident in all resection type subgroups when comparing FLR volume to FLR function. These findings again highlight the relevance of functional testing in patients at risk of PHLF.
Aside from the added predictive value of HBS in high-risk resections, the multivariable analysis for liver failure illustrates the relevance of functional assessment. While liver volume was not a predictor for PHLF, liver function was. Corrected for relevant variables, the use of liver function is therefore likely able to identify patients at high-risk of liver failure in the absence of other relevant predictors of PHLF.
This study has several limitations, mostly related to the retrospective study design and unavoidable selection bias, for instance, patients who were not operated but did undergo liver function assessment were not included in this study. Several studies have already shown the high risk of liver failure associated with low liver function measurements, therefore the number of patients with low liver function in this study is low. This is likely the main reason why the predictive value of HBS seen in the first study from 2010 was not reached. The limited number of PHLF in the non-biliary group (20/340) compared to the biliary (36/207) may not be sufficient to demonstrate a clear additional value of function in the non-biliary group. Complications subject patients to an increased risk of liver failure. There was insufficient data on the timing of complications to include this in the analysis. This is the first multicenter series on HBS before major liver resection, however, the evidence presented does not completely justify its implementation in every center that performs liver surgery. A prospective trial, preferably randomized trial is needed to demonstrate with high level of certainty that HBS should be the gold standard, but comes with several practical challenges. A randomized prospective trial would rquire thousands of included patients which is clearly not feasible. Centers that routinely use HBS will probably be very reluctant to participate in such a trial due to their positive experiences with HBS and may believe it to be unethical to perform only volume assessment of the FLR. Perhaps a large series such as the current study is the highest level of evidence we can practically achieve.
The current study shows that preoperative liver function assessment using HBS is at least as predictive for PHLF as liver volume assessment, and likely has several advantages. The absence of liver volume as predictor in the multivariable analysis when liver function is included suggests HBS can aid to identify patients at unexpected high risk of liver failure, or at least improve the predictive value when using both liver function and liver volume assessment In patients with biliary tumor that are at high risk of liver failure, liver function assessment more effectively stratified patients according to the risk of PHLF. While the 2.7%/min/m2 seems appropriate for most patients, the risk of PHLF in patients with biliary tumors can potentially be reduced when a liver function cut-off of 8.5%/min (4.5%/min/m2) is used. For these patients with biliary tumors, liver volume is of value in the preoperative work up, but liver function assessment might be able to help to improve the outcomes.
Disclosures
The authors have nothing to disclose.
Funding
The authors report no funding.
Conflict of interest
None to declare.
References
van Mierlo K.M.
Schaap F.G.
Dejong C.H.
Olde Damink S.W.
Liver resection for cancer: new developments in prediction, prevention and management of postresectional liver failure.
Prospective evaluation of the International Study Group for Liver Surgery definition of post hepatectomy liver failure after liver resection: an international multicentre study.
Relevance of the ISGLS definition of posthepatectomy liver failure in early prediction of poor outcome after liver resection: study on 680 hepatectomies.
The role of hepatobiliary scintigraphy combined with spect/ct in predicting severity of liver failure before major hepatectomy: a single-center pilot study.
Future remnant liver function estimated by combining liver volumetry on magnetic resonance imaging with total liver function on (99m)Tc-mebrofenin hepatobiliary scintigraphy: can this tool predict post-hepatectomy liver failure?.
Extended preoperative chemotherapy does not improve pathologic response and increases postoperative liver insufficiency after hepatic resection for colorectal liver metastases.
The impact of modern chemotherapy and chemotherapy-associated liver injuries (CALI) on liver function: value of 99mTc-labelled-mebrofenin SPECT-hepatobiliary scintigraphy.
Postoperative liver failure risk score: identifying patients with resectable perihilar cholangiocarcinoma who can benefit from portal vein embolization.
Hepatobiliary scintigraphy to evaluate liver function in associating liver partition and portal vein ligation for staged hepatectomy: liver volume overestimates liver function.
01680-X&id=gr1.jpg){kind=link}
01680-X&id=gr2.jpg){kind=link}