Introduction

Allopurinol, a structural analogue of hypoxanthine and a xanthine oxidase inhibitor, has been utilized experimentally in the attenuation of warm and cold ischemia and reperfusion injury of various organs since 1971 [1]. Initial studies on hemorrhagic shock in 1969 analyzed the effect that this drug had on the potential loss of functional purine bases during hypoxia-related events [2].

Studies in rats have shown a protective benefit when animals that underwent liver ischemia reperfusion injury were pretreated with allopurinol. These studies have demonstrated that allopurinol’s protective mechanism [3–14] may be wholly or partly due to inhibition of the irreversible breakdown of purine metabolites, thereby facilitating intracellular ATP production, inhibition of the formation of reactive oxygen species (ROS), and protection against mitochondrial membrane damage [3–14].

This paper reviews allopurinol’s role in warm liver ischemia and reperfusion. Additionally, it identifies the ideal dose of allopurinol in liver ischemia and reperfusion and emphasizes the mechanisms of injury and drug protection.

Brief history on the use of allopurinol in hepatic ischemia reperfusion injury

Allopurinol (Fig. 1) is best known for its use in the treatment of gout with hyperuricemia, even though it was discovered in an effort to develop anticancer drugs. In the 1950s, Hitchings noticed that purine antimetabolites had antitumor activity in cultures of transplanted tumors, but their effect was limited because the tumors were refractory to treatment [15]. Developed in 1963 [16], allopurinol was used in conjunction with the antitumor drug 6-mercaptopurine (6-MP) because it was both a substrate and an inhibitor of the enzyme that metabolized 6-MP [15]. Allopurinol was successful in slowing the degradation of 6-MP in human trials and is used today in secondary gout induced by tumors, radiation, or chemotherapy [17–19]. It was incidentally noted that there was a decrease in the uric acid in serum and the urate in urine accompanied by a concomitant increase in xanthine and hypoxanthine in urine of these same original patients [15].

Fig. 1
figure 1

The chemical structure of allopurinol, [1,4]dihydropyrazolo[4,3-d]pyrimidin-7-one (chemical formula C5H4N4O) (mol. wt. 136 g/mol)

Full size image

With the knowledge that by inhibiting xanthine oxidase, allopurinol inhibited the formation of uric acid from xanthine and xanthine from hypoxanthine, the focus of its use was consequently shifted to the treatment of hyperuricemia in gout.

In 1969, Crowell and associates tested allopurinol pretreatment in hemorrhagic shock. It was their hypothesis that xanthine oxidase prevented the irreversible breakdown of purine substrates that could be used for ATP re-synthesis after the ischemic insult [2]. As early as 1971, the effect of allopurinol in ischemic tissues was tested, starting with the ischemic myocardium by DeWall and group [1]. In 1972, Vasko et al. [20] demonstrated the protective effect of allopurinol in renal ischemia. In 1973, Toledo-Pereyra et al. [21] studied the effect of allopurinol in ischemically preserved kidneys for transplant. Others, like Lazarus et al. [22] used allopurinol in hemorrhagic shock to test hepatic function. The first experimental testing of allopurinol pretreatment in liver donors was performed by Toledo-Pereya et al. [23] in 1975. This experiment found increased survival in canine donor liver recipients if pretreated with isoproterenol, allopurinol, and heparin [23].

In 1968, McCord and Fridovich [24] found that reduction of cytochrome c by xanthine oxidase produced ROS. It was not until the 1980s that it was considered that ischemia reperfusion injury (IR) was associated with ROS as a mechanism of injury [25]. Recalling that reactive oxidative species were formed in both reactions catalyzed by xanthine oxidase, theories arose as to allopurinol’s mechanism of protection in IR. Since that time, theories of its usefulness and its mechanism of protection in ischemic and reperfused tissues have continued.

Mechanism of allopurinol in ischemia reperfusion injury

Although a beneficial effect of allopurinol in IR injury has been frequently documented, there has not been a general consensus as to the mechanism by which it exerts its beneficial effect. As a way of explanation, we can begin by endeavoring to understand the pathway and metabolites associated with allopurinol and IR.

Xanthine dehydrogenase (XDH) is an oxidizing agent of hypoxanthine to xanthine and xanthine to uric acid. In the creation of these reduced molecules under normal conditions, nicotinamide adenine dinucleotide (NAD+) is used as an electron acceptor. In ischemia, intracellular calcium homeostasis is disturbed and influx of calcium into the cell triggers activation of many enzymes including a protease that changes XDH such that it can no longer reduce NAD+ and is called xanthine oxidase (XO). XO instead transfers the electrons to molecular oxygen creating ROS [26]. ROS are generally accepted as harmful to a cell’s homeostasis, and their insult can eventually lead to cell death. This is especially relevant in relation to IR injury, since the liver has the highest concentration of XDH in the body [27] (Fig. 2).

Fig. 2
figure 2

Xanthine oxidase inhibition, free radicals (ROS), and liver ischemia

Full size image

As we analyze the mechanistic approach of allopurinol in liver IR, several theories are readily evident. The first theory can be represented by the initial work of Crowell and associates in 1969 [2]. They hypothesized that allopurinol inhibited the irreversible breakdown of hypoxanthine during ischemia to be later used for ATP re-synthesis. The evidence that dogs with induced hemorrhage had increased survival when pretreated with allopurinol indirectly supported this theory [2]. Since then, additional formal experiments testing this theory have been performed. Several researchers have found that allopurinol pretreatment before hepatic IR insult resulted in increased hypoxanthine accumulation [14, 28], decreased xanthine [10, 29], increased adenosine/xanthine ratio [7], increased total adenine nucleotide concentration during ischemia [10], decreased total purine catabolites [10], and a resultant increase in energy phosphates after reperfusion [7, 14, 29]. In some cases, increases in hypoxanthine did not translate into increased ATP synthesis after reperfusion [4, 28]. Additionally, when hypoxanthine was added to allopurinol pretreatment, no additive benefit was found, suggesting that purine catabolite conservation may not be allopurinol’s only mechanism in IR injury [6].

Using the knowledge that XO produces ROS, the second theory suggests that allopurinol was protective in liver IR by inhibiting the formation of ROS [6]. This was demonstrated by the addition of XO pre-ischemically. The additional increased damage suggests it as a primary mechanism in IR injury [6]. In a study that tested preconditioning with allopurinol compared to free radical scavengers, the protective benefit was greatest with allopurinol, indicating that XO and XDH are a primary source of ROS during IR injury [13]. Interestingly, the addition of free radical scavengers pre-ischemically is protective to IR damage, implicating ROS production as the mechanism of IR injury [30, 31].

Other findings describes a different protective effect. Conversion from XDH to XO occurs in ischemia, but the exact timing has been shown to range from 10 min [32] to 2 h [33, 34]. Since it may take 2 h for the enzyme to convert from XDH to XO, and ROS appear long before 2 h, it is difficult to support the role of XO as the primary source of ROS. However, if the conversion occurred in the first 10 min as suggested by Della Corte’s studies, this could account for ROS formation by XO as a means of IR injury [32].

Another argument that might not be in accordance with the inhibition of XO as the reason for the protective effect is that allopurinol in very high concentrations is needed to prevent IR injury, concentrations much higher than those needed to completely inhibit the enzyme in vivo [35]. Explanations for this finding are two-fold. First, rats have higher concentrations of XO and XDH in blood and liver than humans [36]. Second, as suggested by Godin et al. [36], allopurinol’s mechanism of protection may not be singularly related to the ability to inhibit the formation of ROS by XO. Godin et al. [36] found that in rabbit ischemic myocardium, which has practically undetectable XO activity, allopurinol has a protective effect in IR injury. This research considers that the protective mechanism of allopurinol may not be totally related to the inhibition of ROS generation by XO.

The third theory indicates allopurinol’s beneficial effect as the preservation of mitochondrial function by protecting mitochondrial membrane integrity [7, 37]. Support for this theory has been demonstrated by the decreased mitochondrial swelling and cell death [7] and decreased lipid peroxidation [10] with allopurinol pretreatment. It is also suggested that allopurinol has scavenging properties of its own, either by hydroxyl radical [38] or as an electron transfer agent [39] (Fig. 3).

Fig. 3
figure 3

Xanthine oxidase inhibition and preservation of mitochondrial function

Full size image

Even though none of these arguments can confirm or refute any one mechanism for allopurinol’s role in IR injury, the evidence remains that allopurinol appears to be protective for IR injury.

Allopurinol in liver ischemia-reperfusion injury

Allopurinol appears to be hepato-protective for IR injury as measured by markers of liver injury and function, histopathology, and markers of oxidative stress as reviewed below (Table 1). This work will review studies from all animal models, but primarily the rat as it is the predominant model in this field of research. This presents a challenge in generalizing these findings and predicting results in humans as rodents have higher levels of XO activity [27]. However, using rat models has some advantages, such as the low expense involved, extensive past research in this area for comparison, and avoiding the many ethical hurdles of other models. Since many studies do not utilize the same tests to measure liver injury, this work will systematically review those results in groups by the mechanism of injury measured.

Table 1 Allopurinol (ALO) in liver ischemia reperfusion injury experiments
Full size table

Allopurinol protection in liver IR as measured by markers of liver injury

Liver injury can be measured with traditional markers of tissue injury including alanine aminotransferase (ALT), aspartate aminotransferase (AST), which are both specific to the liver, and lactate dehydrogenase (LDH), which is not. Measured after reperfusion, along different time intervals, allopurinol has been shown to attenuate the IR induced rise in both transaminases [7–10, 13, 37, 40–43] and LDH [9, 40]. There are a few studies that have shown allopurinol has no effect on transaminases [5, 44] and that LDH [42, 45] increases. In most cases, if the attenuation in transaminase increases was not seen immediately, it occurred at least 5 h after reperfusion had elapsed [10, 37, 41, 42]. However when the period was extended to 24 h, no effect was observed by Metzger et al. [5]. The majority of these studies have shown that allopurinol has a protective effect on liver tissue injury based on transaminases and LDH.

Allopurinol protection in liver IR injury as measured by markers of liver metabolic response

Liver injury can also be measured indirectly by detecting the liver’s ability to perform other metabolic responses, for example, the ability to conjugate bilirubin [13]. Allopurinol’s effect on this is divided into no effect or a significant increase in conjugated bilirubin [29]. Bile flow has been shown in experiments to cease after 10 min of ischemia [11]. However, measuring the return of bile flow in reperfusion can determine the liver’s ability to secrete bile into ducts. This was measured by cannulating the efferent bile duct [11, 37] and by measuring alkaline phosphatase [13]. In these experiments, it was shown that allopurinol pretreatment significantly restored bile flow in reperfusion of IR injury [11, 13, 37].

It is worth mentioning that early studies on the effect of allopurinol on hepatic injury after reperfusion were measured in terms of RNA synthesis in dogs with induced hemorrhage [4]. It was found that pretreated animals had improved RNA synthesis after hemorrhage compared to controls [4].

Allopurinol protection in liver IR injury as measured by histopathology

Measuring cellular damages histologically can be undertaken in many different ways. Allopurinol has been shown to have a protective effect on hepatic congestion [40, 41], fatty degeneration [40], necrosis [8, 13, 40], apoptosis [8, 9], lesions in central lobule [9], mitochondrial swelling by inhibiting mitochondrial pore opening [7, 37], and intracellular [4] and extracellular water accumulation [40] as a measurement of membrane damage. However, one experiment showed no protective benefit of allopurinol pretreatment on lipid droplets in cytoplasm, glycogen granule disappearance, loss of normal endoplasmic reticulum architecture, and disappearance of outer nuclear membrane [12]. The tissue samples in this study were taken after 45 min of ischemia but before reperfusion. In all of the other histological studies where significant findings were produced with allopurinol pretreatment, samples were taken after reperfusion began. These findings are good evidence that damage from oxygen species occurs after reperfusion begins. Most studies show allopurinol improves liver response to IR injury histopathologically (Fig. 2).

Allopurinol protection in liver IR injury as measured by markers of oxidative stress

Oxidative stress is a direct measure of tissue injury as it is known that it will greatly injure or ultimately destroy cells. Intracellular ROS can be measured directly by liver peroxide levels, or indirectly by malondialdehyde (MDA), a byproduct of lipid peroxidation. Several studies show that allopurinol pretreatment significantly decreased liver peroxidation [7, 8, 10, 12, 37, 41]. However, other studies dispute these findings, indicating other extracellular sources, such as Kupffer cells and polymorphonuclear cells, as the primary means of ROS production [46–48].

Reducing hydrogen peroxide by glutathione peroxidases is a mechanism of protecting cells from oxidative stress. In this process, glutathione (GSH) is oxidized to glutathione disulfide (GSSG). Measuring GSSG concentration in bile has been shown in animal models to be a sensitive index of oxidant stress in vivo [49, 50], but the GSSG/GSH ratio is a more sensitive index of intrahepatic oxidative stress [50]. GSH can be used as a marker of membrane damage based on leakage into the extracellular space [50]. Results of experimental models after hepatic IR injury disagreed, suggesting allopurinol pretreatment decreased oxidant stress [7, 8, 11, 12, 37] or did not have an effect on oxidant stress [5, 44, 51] as measured by intracellular GSH stores, biliary GSH, and GSSG efflux, and decreases in mitochondrial glutamate dehydrogenase.

Oxidative stress causes a halt in cellular aerobic respiration and hence depletion of energy stores. Therefore, oxidative stress can be measured indirectly by changes in cellular ATP during hepatic IR injury, or by measuring byproducts of anaerobic respiration including lactic acid. Allopurinol pretreatment attenuated the rise in lactic acid formation [40], a sign that hepatic energy stores were more plentiful and anaerobic respiration was less necessary. Allopurinol also showed a beneficial effect in hepatic IR injury by increasing ATP stores after reperfusion, increasing total adenine nucleotide concentration, and decreasing purine catabolites compared to controls. Accordingly, these studies show that allopurinol improves resynthesis of ATP during reperfusion [7, 10, 14, 29, 52]. One study showed a detrimental effect; allopurinol pretreated animals had decreased ATP compared to controls [45]. In this study the dose of allopurinol was minimal at 1.5 mg/kg in a rat model. It has been shown by Jeon et al. [10] that the hepatoprotective dose in rats is 50 mg/kg. These data suggest that part of allopurinol’s beneficial mechanism is increased aerobic respiration and a reciprocal decrease in anaerobic response.

A newer field of study is exploring allopurinol’s effect on molecular signaling pathways within the liver. Since it is known that ROS can activate inflammation, some have investigated if allopurinol would have an inhibitory effect on the inflammatory pathways by decreasing ROS production. Liver production of TNF-α, an inflammatory cytokine, as measured by ELISA, was shown to be decreased with allopurinol pretreatment [8]. Allopurinol pretreatment in lobar liver ischemia reduced NF-kappa B activation [53]. NF-kappa B is a quick responder to inflammatory stimulus, activating the transcription of genes that are pro-inflammatory, cell proliferative, and anti-apoptotic [53]. Both of these studies suggest that ROS produced by XO may have an impact on NF-kappa activation. High mobility group box 1 (HMGB1) is a nuclear factor released extracellularly as an early mediator of IR injury. In liver IR, it is released as early as 1 h after reperfusion, and inhibition of HMGB1 significantly decreased liver damage after IR [54]. Science is sure to benefit as molecular biology explores allopurinol’s effect in cell signaling during IR injury.

Human study of allopurinol pretreatment in ischemia reperfusion injury

In the period of 1992–1994, Vriens and group [51] prospectively followed a group of 16 human patients with IR injury to the liver induced by liver resection of five or fewer colorectal metastases confined to the liver. Half of the patients were pretreated with 10 mg/kg of allopurinol given p.o. the day before the operation with two additional doses of 5 mg/kg just before the operation and just before reperfusion. Ischemia to the liver was induced for a minimum of 30 min. Measurements of liver injury were determined in terms of MDA, GSH, GSSG, vitamin C, coagulation studies, ALT/AST, LDH, and albumin.

Within 24 h after reperfusion, MDA, GSH/GSSG, and vitamin C were not significantly different between control and allopurinol groups. PT, PTT, fibrinogen, and factor V were not significantly different between groups and in most cases were not elevated. Albumin, AST, ALT, and LDH were measured up to 10 days postoperatively, and no significant differences were seen between the two groups, although trends were identified in each. An almost double increase of AST in the control group occurred from post-operative day 1 through 4 compared to the allopurinol pretreated group. The ALT levels of the treated group decreased drastically on post-operative day 2 compared to a decrease on day 3 for the control group. LDH levels were twice as high in the control group compared to the allopurinol treated group [51].

The authors concluded that liver tissue must be more resistant to IR injury since no significant elevation in markers that measure oxidative damage was found in the control group. This model therefore was not an appropriate one to test the effect of allopurinol in liver IR.

There are several issues that need to be critically considered prior to defining the role of allopurinol in human liver ischemic injury. One, it is related to the model utilized; in the case of Vriens et al. [51], the time of liver ischemia was variable and at times suboptimal at 30 min. Ideal times for assessing ischemia should be 60 min or more to allow for enough oxidative damage for the allopurinol to exert its antioxidant effect. Two, the amount of allopurinol used was extremely low at 10 mg/kg orally and 5 mg/kg twice intravenously. The recommended dose is 50 mg/kg for the treatment of experimental livers subjected to warm ischemic injury. However, caution should be exercised when allopurinol is given to humans in spite of the fact that no single side effect has been observed in animals treated at this dose. Third, the time of administration, although more variable, should remain at around 1 h before ischemia.

For a successful clinical trial, liver resection with the characteristics mentioned above would be the best suitable model. In human liver transplantation, the best model would be on livers obtained from cardiac death donors with approximately 20 min of warm time or those recovered from living liver donors with ischemic injury present. It is unknown in this last circumstance what the time of ischemia should be. At any event, in short, the dose and time of administration are essential when considering the preferred model for the use of allopurinol in human ischemic injured livers.

Ideal dose and timing of allopurinol pretreatment for liver IR injury

Tables 2, 3, and 4 address the effect of the dose of allopurinol given to rats undergoing liver IR. In order to compare the various doses utilized in the literature, common means of measuring liver injury were chosen. Transaminase studies were selected due to their frequent use by experienced researchers and their acceptance as reliable indicators. Other means of studying liver IR injury, such as GSH, have not yielded significant results due to lack of data.

Table 2 Effect of dose of allopurinol in liver IR on liver function tests (ALT and AST)
Full size table
Table 3 Chi-squared analysis of the allopurinol dose versus effect on liver IR in liver function tests
Full size table
Table 4 Effect of timing of administration of allopurinol on liver IR in liver function tests (ALT and AST)
Full size table

Table 3 shows the effect of allopurinol dose on liver function tests in liver IR. When the chi-squared test was analyzed with only the 50 mg/kg group, the p values yielded a significant difference at <0.05. From these studies, allopurinol at 50 mg/kg was significantly more likely to result in a high probability of success in liver IR as judged by a significant decrease in liver transaminases.

Table 4 indicates the role of timing of allopurinol administration in rats with liver IR injury as measured by transaminases. Even though there was a clear demonstration of a relationship between timing of administration and the beneficial effect, these studies showed the extensive window of administration in which allopurinol might still be beneficial. There is not a clear explanation for the physiological response as not enough studies included a specific time period of administration (Table 4).

Conclusion

The mechanism by which allopurinol exerts a protective effect on liver IR injury is still under debate, but strong evidence exists to support that allopurinol aids in resynthesis of ATP by inhibiting the breakdown of its catabolites, inhibiting the formation of ROS, and preventing mitochondrial membrane damage, therefore decreasing anaerobic respiration.

In many animal studies of liver IR injury, allopurinol has been shown to decrease damage, increase functional response after IR injury, and decrease oxidative stress. A study in humans cannot be evaluated at this point because of the surgical model utilized. Recommendations for improving the model are indicated.

Based on the results of liver function tests, experimental data indicated that 50 mg/kg of allopurinol was the ideal dose to achieve a consistent protective effect in animals with liver IR. The timing of administration of allopurinol was not uniform enough, but successful outcomes often centered around 1 h prior to ischemia. Finally, evaluation of further work on this compound will permit establishing more definitive conclusions in this area.