In this retrospective study, we demonstrated that preoperative serum ALP level could predict 3-year mortality of patients after kidney transplantation. The incidence of 1-year and 3-year mortality after kidney transplantation was the highest in the third tertile of ALP, whereas the incidence of 1-month and 6-month mortality among tertiles showed no significant differences. Moreover, the third tertile of ALP was demonstrated to be an independent predictor for developing 3-year mortality after kidney transplantation.
ALP is a hydrolase enzyme that dephosphorylates various molecules expressed in the bone, liver, placenta, and kidneys. Serum ALP is a biochemical marker of bone turnover or liver disease. There are four ALP isozymes (tissue-nonspecific isozyme, intestinal-type, placental-type, and placental-like). Among these isozymes, tissue-nonspecific ALP is the most abundant isoform and is involved in skeletal mineralization 13. Since mineral metabolism is commonly altered in patients with CKD, ALP is frequently used to monitor bone metabolism associated with renal insufficiency 8.
Disorders related to calcium, phosphorous, and parathyroid hormone (PTH) are commonly observed in patients with CKD. Along with renal osteodystrophy, secondary hyperparathyroidism is common 14, and it may be associated with abnormal mineral metabolism leading to vascular calcification and poor prognosis 15–17. Serum ALP levels have been demonstrated to be associated with mortality in patients with renal failure 11,12,18. Serum ALP ≥ 120 IU/L was reported to be associated with mortality among patients undergoing hemodialysis 18. The highest ALP quartile was associated with higher risk of mortality in patients undergoing peritoneal dialysis 12. Although the association of abnormal mineral metabolism and CKD have been well established and the prognostic role of ALP in these patients has been demonstrated, the prognostic role of ALP in patients after kidney transplantation has not been investigated before. Further, bone turnover and abnormal bone mineralization showed changes even after kidney transplantation 19,20. Thus, the strength of our study is that we investigated the prognostic role of ALP in patients undergoing kidney transplantation, rather than patients with CKD.
The mechanisms responsible for the association between elevated ALP levels and mortality after kidney transplantation are unclear. The first possible explanation is that ALP is a marker of high-turnover bone disease. Elevated serum PTH induces bone resorption and this can be manifested by elevated bone ALP levels 21. A previous study demonstrated that mortality prediction by ALP was likely due to renal osteodystrophy 22. Further, secondary hyperparathyroidism was associated with increased mortality in patients with CKD 23. From these results, it can be inferred that ALP levels are increased due to renal osteodystrophy and secondary hyperparathyroidism in patients with CKD, and it can be associated with increased mortality. Furthermore, ALP levels have been demonstrated to be related with vascular calcification. A previous in vitro study revealed that vascular damage induces expression of tissue-nonspecific ALP and that ALP can promote calcification by hydrolyzing inorganic pyrophosphates 24. An association between ALP and vascular calcification was also demonstrated in previous clinical studies 1,25. Another possibility is the association of ALP and systemic inflammation. Elevated ALP levels were shown to be related to elevated C-reactive protein levels, indicative of systemic inflammation 26. Further, inflammatory stimuli could lead to cellular responses, which could increase ALP expression 13. Along these lines, a previous study reported an association between higher serum ALP level and infection-related mortality in patients undergoing peritoneal dialysis 27.
The optimal cut-off value we found through ROC curve analysis was 71 IU/L, which was similar to the cut-off value of the highest ALP tertile (> 72 IU/L), which was considerably lower than the value obtained in previous studies investigating the association between ALP and mortality in patients undergoing dialysis. ALP ≥ 120 IU/L was reported to be associated with mortality in patients undergoing haemodialysis 18. Another study conducted in patients undergoing peritoneal dialysis showed that the highest tertile of ALP (> 155 IU/L) was associated with an infection-related mortality 27. A previous study on the association between pre-transplant ALP levels and mortality reported that ALP > 120 IU/L was associated with an increased risk of mortality 28. One notable difference is that the incidence of graft failure was significantly different according to the ALP level in the previous study, but not in our study. As calcium-phosphate-PTH homeostasis was found to be a predictor of graft rejection in previous studies 29,30, there may be some correlation between elevated ALP, graft rejection, and increased mortality, which were not addressed in this study. This is because the cut-off value obtained in our study was not high enough to predict graft failure although it could predict mortality. In addition, since there is literature evidence demonstrating an increase in the ALP value itself rather than based on the specific cut-off value as a risk factor for mortality 12, the lower cut-off value obtained in the current study seems plausible.
Notably, ALP could predict long-term (1-year and 3-year) mortality rather than short-term mortality (1-month and 6-month) in this study. Considering that ALP levels are indicative of high-turnover bone disease, the cause of this prediction of long-term mortality may be a reflection of the severity of pre-existing disorders of bone and mineral metabolism. Because there are post-transplant bone disorders, which indicates pathologic processes occurring after transplantation that are superimposed on pre-existing disorders of bone and mineral metabolism secondary to kidney failure and/or diabetes mellitus 20. Further, there have been several reports investigating postoperative changes in bone histomorphometry 19,31 supporting this hypothesis, which is beyond the scope of this study.
The strengths of our study are that we found the prognostic value of ALP in patients undergoing kidney transplantation rather than patients with CKD and this was not reported before. Further, we included a relatively long-term follow-up period and found that ALP levels might predict long-term mortality other than short-term mortality, and we identified a relatively lower cut-off value, which was within normal limits.
The current study has several limitations. First, possible confounding factors such as liver disease, inflammatory status, and comorbidities were not adjusted. However, since ALP levels are affected by various factors, it is worthwhile to divide patients based on the ALP tertile and compare the characteristics accordingly. Second, there is a possibility that the recruited data are insufficient, resulting in a selection bias due to the retrospective nature of the study.
In conclusion, preoperative serum ALP level was an independent predictor of 3-year mortality after kidney transplantation. This retrospective study also revealed that ALP levels can predict long-term mortality rather than short-term mortality after kidney transplantation, which was not investigated before. Our results suggest that ALP level has a prognostic impact on mortality of patients after kidney transplantation, and that the possible mechanisms for this might be associated with high-turnover bone disease, vascular calcification, and inflammation.