Elevation of Mac‐2 binding protein glycosylation isomer after hepatectomy is associated with post‐hepatectomy liver failure, total Pringle time, and renal dysfunction

Abstract Background Mac‐2 binding protein glycosylation isomer (M2BPGi) is a novel serum glycomarker used to assess liver fibrosis. However, it has been reported that M2BPGi is likely to reflect other factors not limited to liver fibrosis. Methods We retrospectively analyzed 79 patients with liver tumors who underwent liver resection. M2BPGi was measured within 1 week before operation and almost 1 month after operation. We introduced a value termed the “ΔM2BPGi ratio” (=M2BPGiafter operation/M2BPGibefore operation), and analyzed factors that influenced the ΔM2BPGi ratio. Results The median value of the ΔM2BPGi ratio was 1.28 (range, 0.36‐5.68). In 64 patients (81.0%), the cutoff index values of M2BPGi were elevated approximately 1 month after operation, especially in patients who experienced post‐hepatectomy liver failure (PHLF). Multiple linear regression showed total Pringle time, PHLF grade ≥B, and preoperative value of creatinine were significant predictors of the ΔM2BPGi ratio. The mean values of the ΔM2BPGi ratio were 1.37 ± 0.07, 1.52 ± 0.22, and 2.94 ± 0.30 for PHLF grade 0, grade A, and grade B, respectively, resulting in statistically significant differences by the Kruskal‐Wallis test (P = 0.022). Conclusions Total Pringle time, PHLF grade ≥B, and preoperative creatinine significantly influenced the elevation of M2BPGi almost 1 month after liver resection. This study strongly affirms the previous suggestion that M2BPGi is likely to reflect other factors not limited to liver fibrosis.


| INTRODUC TI ON
Mac-2 binding protein glycosylation isomer (M2BPGi) is a novel serum glycomarker used to assess liver fibrosis in patients infected with chronic hepatitis C 1-3 or B virus 4,5 as well as in patients with primary biliary cirrhosis, 6,7 biliary atresia, 8 autoimmune hepatitis, 9 and non-alcoholic fatty liver disease. [10][11][12] Although M2BPGi level correlates with liver fibrosis, M2BPGi levels immediately decrease following the eradication of hepatitis C virus (HCV) by direct-acting antiviral therapy, whereas the rapid improvement of liver fibrosis beyond 24 weeks post-treatment is unlikely. 13 Furthermore, M2BPGi was increased in patients with acute liver injury and decreased after their recovery. 14 These observations suggest that M2BPGi reflects liver fibrosis and other factors such as liver inflammation, liver damage, and hepatocyte regeneration.
In this study, we measured serum M2BPGi levels in patients who underwent liver resection for liver tumors before and approximately 1 month after operation. Our aims were to analyze which factors influence perioperative changes in the M2BPGi value, and to evaluate the biological features of M2BPGi.

| Patients
Between December 2016 and December 2018, 120 consecutive patients underwent liver resection for liver tumor at Hiroshima Red Cross Hospital & Atomic Bomb Survivors Hospital. Of these, patients who underwent open microwave ablation without liver resection (four) and patients with a lack of pre-or postoperative M2BPGi test results (33) were excluded from this study. Four patients who underwent postoperative M2BPGi testing over 1 month after operation were also excluded for data accuracy. Consequently, 79 patients were eligible for this study.
A preoperative serum sample was collected at the time of admission for liver resection within 1 week before surgery, and a postoperative sample was collected at the outpatient clinic approximately 1 month after liver resection. M2BPGi values were measured at SRL (Tokyo, Japan), and reported as the cutoff index (COI). We introduced a value termed the "ΔM2BPGi ratio" (=M2BPGi after operation /M2BPGi before operation ) for perioperative changes in the M2BPGi value. We also analyzed predictors of the ΔM2BPGi ratio. Other serologic fibrosis markers, 7S domain of type IV collagen (4COL7S) and hyaluronic acid were also measured at SRL.
Post-hepatectomy liver failure (PHLF) was diagnosed on the basis of the International Study Group of Liver Surgery definition. 15 Briefly, elevated prothrombin time -international normalization ratio and concomitant hyperbilirubinemia on or after postoperative day 5 was defined as PHLF. The severity of PHLF was graded as follows: Grade A, PHLF that required no change in a patient's clinical management; Grade B, PHLF that required a deviation from the regular course but did not require invasive therapy; and Grade C, PHLF that required invasive treatment including intensive care.
The study protocol was approved by our institutional ethics committee.

| Statistical analysis
All statistical analyses were performed using SAS software (JMP 13.0.1; SAS Institute Inc., Cary, NC, USA). Continuous variables were expressed as the means ± standard deviations or medians with ranges, and compared using Student's t-test. Categorical variables were compared using the Chi-square test or Fisher exact test as required. The difference among PHLF grades and between paired stages was compared using the Kruskal-Wallis test and the Wilcoxon rank-sum test, respectively. In multiple linear regression, predictors were selected through a stepwise procedure using the minimum Bayesian information criterion (BIC) method among the factors selected in single linear regression.
The median value of the ΔM2BPGi ratio was 1.28 (range 0.36-5.68).
In addition, it should be noted that we had one patient whose

| Factors that influence the ΔM2BPGi ratio
Univariate and multivariate analyses of perioperative parameters to predict the value of the ΔM2BPGi ratio were performed. A simple linear regression was calculated to predict the ΔM2BPGi ratio based on each factor as shown in Table 2 We evaluated the association between ΔM2BPGi ratio and total Pringle time in more detail (Figure 3). The total Pringle time was weakly correlated to ΔM2BPGi ratio (R 2 = 0.23, P < 0.001).

| Relationship between PHLF and serologic fibrosis markers
The mean values of the ΔM2BPGi ratio were 1.37 ± 0.07,

| D ISCUSS I ON
The present study is the first to investigate predictive factors for the elevation of M2BPGi after liver resection. Contrary to our expectations, the M2BPGi values were elevated in 80% of all cases even approximately 1 month after liver resection. Furthermore, the factors that influenced the elevation of M2BPGi after liver resection were preoperative creatinine, PHLF grade ≥B and total Pringle time. These observations confirm the previous suggestion that M2BPGi is likely to reflect other factors not limited to liver fibrosis. 13 Bekki et al 16  In this study, the ΔM2BPGi ratio increased with advancing PHLF grade, which confirmed the previous study results. Liver resection with a longer total Pringle time causes greater ischemia-reperfusion injury in the liver.
Thus, prolonged total Pringle time and PHLF were likely to be associated with the activation of HSCs and secretion of M2BPGi, which continued for at least 1 month after liver resection. Furthermore, a recent study reported that total Pringle time was associated with PHLF ( Figure 5). 22 The current study also showed that the elevation of the postoperative value of M2BPGi was associated with preoperative creatinine. Few reports have referred to the relationship between M2BPGi tumor necrosis factor-α. 26 Thus, patients with renal failure might have a tendency to increase M2BPGi expression. In addition, renal failure was also reported to be a risk factor for PHLF ( Figure 5). 27 We had one patient whose M2BPGi markedly dropped after liver resection. Given her background liver was F0 fibrosis, her preoper- Why different serologic fibrosis markers exhibit different behaviors after liver resection is unclear. In our study cohort, the Δ4COL7S ratio exhibited a similar tendency to the ΔM2BPGi ratio regarding its relationship with PHLF.
Type IV collagen is found primarily in the basal lamina, although the basal lamina comprised of type IV collagen appeared around the liver sinusoids after liver injury. 28 Serum levels of 4COL7S, which has greater sensitivity and specificity for the detection of cirrhosis than type IV collagen, was increased abruptly following acute liver injury. 29 It was also reported that type IV collagen accumulated in HSCs during acute viral hepatitis. 27  Thus, ΔM2BPGi ratio might be associated with long-term outcomes of patients who underwent liver resection.
A major limitation of this study was its observational design, which was exploratory in nature. The small subject number of the study and its single-center-based design limited the power of the explanation of results. Furthermore, the number of days after liver resection to postoperative serum sampling differed by individual.
To minimize the influence of these factors, we excluded patients whose M2BPGi were measured more than 1 month after liver resection. We also confirmed that the ΔM2BPGi ratio was less affected by the number of days between operation and serum sampling by simple linear regression (Table 2). Finally, we did not determine the timely elevation of M2BPGi values during PHLF or liver resection because of the study design. Therefore, we should interpret the results of this study as predictors for the elevation of M2BPGi value at 1 month after liver resection, and not for any other time points.
In conclusion, preoperative creatinine, PHLF grade ≥B and total Pringle time were associated with the elevation of M2BPGi 1 month after liver resection. These observations emphasize the previous suggestion that M2BPGi is likely to reflect liver damage, liver inflammation, and renal function not limited to liver fibrosis.

ACK N OWLED G M ENT
We thank Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.