Early elevated serum gamma glutamyl transpeptidase after liver transplantation is associated with better survival

Introduction

Gamma glutamyl transpeptidase (GGT) is a membrane-bound enzyme that is essential for the synthesis of glutathione (GSH), a key antioxidant¹. In clinical practice elevated serum GGT is generally used as an indicator of liver disease, such as biliary obstruction, alcohol consumption, and exposure to certain medical drugs¹. Recently, several epidemiological studies have shown that a higher serum GGT level, even within the normal range, is associated with cardiovascular risk factors such as hypertension, hypertriglyceridemia, obesity, type 2 diabetes mellitus and stroke, as well as certain types of cancer^2–10. In contrast to these studies, we observed that after surgery for ruptured abdominal aortic aneurysm¹¹ or after liver resection¹², GGT is transiently increased in patients who had a good outcome. In these short-term observational studies GGT level was inversely related to other liver laboratory parameters such as aspartate aminotransferase (ALT), alanine aminotransferase (AST) as well as total bilirubin (TBI)^11,12. We observe that in the early postoperative period after a liver transplantation (LT) a transient gradual increase in GGT also occurs. How early and late postoperative changes in serum GGT are related to survival is not known.

Here we present a study in which we assessed the relationship between early (postoperative day seven) elevated GGT levels with early and late survival (i.e. survival within the 90 days and five years post-LT, respectively). We also evaluated the relationship between late (six months postoperatively) elevated GGT levels with late survival.

In addition, we studied the early and late post-LT kinetics of GGT, AST, ALT, and TBI in patients who survived more than 90 days and in patients who did not. Likewise, these kinetics were compared with long-term survival.

Materials and methods

Results

Early post-LT GGT and other liver variables and early mortality

Early postoperative laboratory variables are shown in Figure 1. Postoperatively, GGT levels increased gradually, reaching a maximum at POD 9 and decreased thereafter. Notably, the increase in GGT levels was significantly more pronounced, i.e. deviated more from the normal range, in patients who survived more than 90 days, as compared to those who did not: 297 (178–464) vs. 172 (69–271) U/l, p<0.0001, respectively. This pattern was different from that of postoperative levels of TBI, AST, and ALT. TBI was consistently and significantly lower in patients who survived more than 90 days following LT as compared to those who did. AST and ALT levels increased rapidly until POD 1 and POD 2, respectively, followed by rapid normalization thereafter. Contrary to GGT, the peak levels of AST and ALT were significantly lower in patients who survived more than 90 days following LT as compared to those who did not: AST 659 (326–1267) vs. 1201 (451–1990) U/l, p=0.01, and ALT 527 (280–1080) vs. 1082 (529–2631) U/l, p=0.001, respectively.

Figure 1. Evaluation of early post-LT laboratory variables (i.e. 0–30 POD) in early mortality.

Curves represent patients who survived more than 90 days (gray) and those who did not (black). Median values are shown. * p < 0.05; ** p ≤ 0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.

Thirty patients developed NAS within the 90 days post-LT. Since these patients may present with abnormal high TBI and GGT levels, we repeated the analysis with exclusion of these 30 patients with no significant affect on the GGT levels (graphical representation not shown). Also exclusion of patients who were treated for developing an acute rejection (n=103) and those who underwent LT for cholestatic liver disease (n=155) did not significantly affect the outcomes (graphical representation not shown).

Late post-LT GGT and other liver variables and mortality

Late postoperative laboratory variables are shown in Figure 2. We studied the changes in GGT levels over time in patients who survived more than five years as compared to those who did not survive. Compared to patients who died within five years post-LT, those who survived more than five years had significantly lower GGT at three months post-LT; 95 (42–244) vs. 212 (92–400) U/l, p=0.001, six months post-LT; 70 (31–189) vs. 281 (103–438) U/l, p<0.001, and 1 year post-LT; 57 (25–153) vs. 124 (45–431) U/l, p=0.003, respectively. Notably, late post-LT the GGT levels showed the same patterns as TBI, AST, and ALT, i.e. higher levels in non-survivors.

Figure 2. Evaluation of three, six, and twelve months post-LT laboratory variables in late mortality.

Bars represent patients who survived more than five years (gray) and those who did not (black). Median values are shown. *p<0.05; **p≤0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.

Kaplan-Meier survival analysis

We also studied the clinical relevance of early versus late elevated GGT levels with Kaplan-Meier survival analysis. Figure 3 plots the overall 90-day and 5-year overall survival for the study cohort based on GGT-tertiles. A high GGT level at POD 7 was significantly associated with better early survival following LT (Figure 3A). The overall 90-day survival was 98% for high GGT (≥ 351 U/l), compared to 94% for intermediate GGT levels (188 and 350 U/l), and 87% for the low GGT (≤ 187 U/l), at POD 7 (Figure 3A). Similarly, five-year overall survival was 86%, 83%, and 73% for high, intermediate, and low GGT at POD 7 (p=0.003; Figure 3B), respectively. Remarkably, the differences in five-year survival mainly developed during the first three months post-LT with almost no difference in survival curves thereafter (Figure 3B). In sharp contrast with early GGT, a high GGT level six months post-LT was associated with lower five-year survival (Figure 3C). The overall survival within five years was 71% for elevated GGT (> 163 U/l), compared to 91% for intermediate GGT levels (44 and 163 U/l), and 93% for the low GGT (< 43 U/l), p<0.001.

Figure 3. Kaplan–Meier analysis showing the opposite relationships of early and late GGT levels with early and late mortality.

Panels A and B represent survival analysis post-LT in relation to GGT tertiles at POD 7, within 90 days and five years respectively (p=0.003). Panel C demonstrates five-year survival in relation to GGT tertiles at six months post-LT (p<0.001). Curves represent low (thick black), intermediate (thin black), and high GGT (gray).

Discussion

In this study, we evaluated the changes in GGT over time following liver transplantation and the clinical relevance of these changes for early and late survival. We found that a transiently elevated GGT early after LT was associated with increased survival rates within the first 90 days. In contrast, late elevation of GGT was associated with decreased five-year survival following LT. Although the early GGT elevations was also associated with five-year survival, this difference mainly developed during the first 90 days post-LT.

This peculiar effect was not observed for TBI, AST, and ALT since higher levels for these parameters at POD 7 and six months were associated with increased mortality at both 90 days and five years after LT.

To our knowledge, this is the first study showing the short and long term kinetics of GGT and the clinical relevance of an early elevated serum GGT in LT recipients. Previously, we have reported improved outcome in patients with significantly increased levels of GGT in the early post-operative period following liver resection¹² and surgical repair of a ruptured abdominal aortic aneurysm¹¹. However, those studies were not designed to address changes in GGT progression over time.

While we acknowledge that association does not necessarily indicate causation, these data support the hypothesis that high GGT in an early post-LT setting may be a marker of some protective process.

Although the precise mechanism responsible for an elevated serum GGT early after LT is yet to be determined, experimental studies have demonstrated that cellular GGT modulates crucial redox-sensitive functions, such as antioxidant and antitoxic defenses, cellular proliferation, and apoptotic balance¹³. Cellular GGT is a key enzyme in the gamma-glutamyl cycle resulting in production of intracellular GSH^14–16, an important antioxidant agent that protects the cells against reactive oxygen species (ROS)^17–19. GSH has been shown to protect the liver against ischemia reperfusion injury in animal models^16,20,21. Hepatic ischemia can cause elevation of serum GGT with peak blood levels within 20 and 30 hours after restoration of hepatic arterial blood flow^18,22. Reperfusion is associated with a surge of ROS, which may overwhelm host natural antioxidant defenses²¹. The oxidative stress from the ROS formed after reperfusion may lead to increased cellular death by damaging membrane lipids through peroxidation, disrupting normal enzymatic activities, and diminishing mitochondrial oxidative metabolism²³. Cardin and colleagues²⁴ studied oxidative stress in patients with chronic hepatitis C virus infection. Surprisingly, the authors observed an association between GGT and 8-hydroxydeoxyguanosine (8-OHdG), a marker of oxidative DNA damage. Patients who had a high level of 8-OHdG had significantly higher GGT levels but normal ALT levels²⁴.

Thus, a transient increase in GGT level post-LT may reflect the host compensatory mechanism against oxidative stress and toxic metabolites generated by hypoxia, reperfusion, and surgical stress²¹. Therefore, an increased GGT early after LT may reflect the ability of the host to initiate an appropriate systemic response.

Elevated serum GGT in the early post-LT period might also be a marker of better liver regeneration. Eisenbach and colleagues²⁵, showed that an early increase in serum GGT after LT was associated with a better outcome and the authors reasoned that this rise could be due to liver regeneration. Although the liver might regenerate to some extent after LT, there is no conclusive evidence to support this hypothesis. As we mentioned earlier, we observed a transient increase in serum GGT levels in patients who survived a surgical repair for a ruptured abdominal aortic aneurysm¹¹. In the latter group, it is less likely that significant liver regeneration occurs.

Furthermore, elevated serum GGT early post-LT may also be an indicator of better early secretory function of the liver. GGT in part comes from the surface of the bile ducts and is released from the anchor, which attaches this enzyme to the cell surface, by bile acids in bile. Bile acid/phospholipid ratio in bile has shown to be elevated after LT²⁶. Thus elevated serum GGT immediately after transplantation may be an indicator or an active bile flow. Hence, it may protect hepatocytes from cytotoxic bile acids.

Contrary to early elevated GGT, but in line with published literature^2–10, we observed that a late (i.e. six months post-LT) elevated GGT was significantly associated with decreased survival within the 5-years following LT. Although the finding that normal GGT levels in the late post-operative period is predictive of good outcomes is obvious and intuitive, the contrasting influences of GGT levels between early and late post-LT periods on survival may be compatible with the physiologic function of intracellular GGT. Notably, at three months, six months, and one year post-LT, the relative increase in serum GGT was two to four times higher in patients who did not survive for more than five years compared to those who did survive. This proportion was much higher than that of AST, ALT and TBI, which might imply that an elevated serum GGT is not only a marker of harm to the liver but it could be seen as a systemic response to harmful environmental factors. Indeed, in two studies, Lee and colleagues^27,28 postulated that serum GGT in the general population might be a marker of increased exposure to environmental stress, internal xenobiotics, or other unknown factors that cause oxidative stress in the long run.

To avoid possible bias we performed our analysis after excluding obstructive mechanisms such as NAS, cholestatic disorders, and acute graft rejection early postoperatively. Exclusion of these cases did not change our results significantly, suggesting that the elevation in serum GGT early post-LT is independent of obstructive disorders.

A practical clinical consequence of our findings may be that care providers in hospitals should realize that an abnormally high GGT early post-LT is not a cause for alarm or specific diagnostic procedures. In fact, a GGT activity four to five times above the normal range during the second post-operative week might even be considered beneficial.

We acknowledge some considerable limitations in this study. First, due to the retrospective design of the study we can only identify association rather than causation. This could only be established by specific (intervention) studies that measure the interaction between serum GGT and ROS markers post-LT. However, there is a strong correlation between GGT and oxidative DNA damage in cirrhotic patients²⁴. Next, we cannot entirely exclude that liver regeneration plays a role in the early post-LT rise of GGT levels since GSH and to some extent GGT are mainly produced by the liver¹. Hence a high serum GGT in patients who survived more than 90 days can also be a reflection of a well-functioning graft in LT patients. It will be important to understand the relationship of serum GGT and cellular GGT in the period immediately after surgery. Besides, cumulative evidence suggests that there is a relationship between the induction and release of ROS and ischemia reperfusion injury after other types of abdominal surgery^18,21,22.

In conclusion, an elevated GGT level early after LT was associated with a better short-term outcome. However, chronically elevated GGT was associated with poor long-term outcome in the outpatient setting after LT. This peculiar switch in the prognostic meaning of GGT may result from the superposition of several mechanisms. Apparently, higher expression of GGT reflects some protective process in the acute phase but reflects chronic damage over the long term.

Data availability

figshare: Serum gamma glutamyl transpeptidase levels post-liver transplantation and survival data, http://dx.doi.org/10.6084/m9.figshare.900343²⁹