The majority of transplants are derived from donors who suffered from brain injury. There is evidence that brain death causes inflammatory changes in the donor. To define the impact of brain death, we evaluated the gene expression of cytokines in human brain dead and ideal living donors and compared these data to organ function following transplantation.
Hepatic tissues from brain dead (n = 32) and living donors (n = 26) were collected at the time of donor laparotomy. Additional biopsies were performed before organ preservation, at the time of transplantation and one hour after reperfusion. Cytokines were assessed by real-time reverse transcriptase–polymerase chain reaction (RT–PCR) and cytometric bead array. Additionally, immunohistological analysis of tissue specimens was performed. Inflammatory cytokines including IL-6, IL-10, TNF-α, TGF-β and MIP-1α were significantly higher in brain dead donors immediately after laparotomy compared to living donors. Cellular infiltrates significantly increased in parallel to the soluble cytokines IL-6 and IL-10. Enhanced immune activation in brain dead donors was reflected by a deteriorated I/R injury proven by elevated alanin-amino-transferase (ALT), aspartat-amino-transferase (AST) and bilirubin levels, increased rates of acute rejection and primary nonfunction. Based on our clinical data, we demonstrate that brain death and the events that precede it are associated with a significant upregulation of inflammatory cytokines and lead to a worse ischemia/reperfusion injury after transplantation.
The brain dead (BD) organ donor is the main source for solid organ transplantation. Clinical studies have shown that allografts from unrelated living donors (LD) have superior graft function and survival compared to allografts from deceased donors (1–3). Experimentally, donor brain death reduces allograft survival of rat livers undergoing prolonged ischemia compared to living donor controls (4). In this context, unspecific inflammatory events as a consequence of brain death, prolonged ischemia and additional recipient and donor related risk factors seem to be important (5–9). The unphysiological status of brain death is associated with extensive physiological and hormonal alterations influencing organ quality prior to graft recovery. BD is a catastrophic physiological event, defined as an irreversible injury of cerebrum, cerebellum and brain stem and is associated with severe hemodynamic changes, coagulopathies, pulmonary changes, hypothermia and electrolyte imbalances (10,11). The hemodynamic instability is well recognized as ‘autonomic storm', an initial period of excessive parasympathetic activity, which is immediately followed by a sympathetic activation with high plasma levels of catecholamines, extreme arterial hypertension and tachycardia. Sympathetic activities subside thereafter and are followed by a severe reduction in sympathetic outflow with impairment of cardiac ino- and chronotropy (12,13). During these phases potential donor grafts undergo a transient period of ischemia. Consequently, these pathways result in structural and functional changes in somatic organs prior to engraftment (11,14–16). In donor livers the central catastrophe leads to cellular infiltrates and an increased rate of apoptosis (3,17). Experimental data demonstrated the upregulation of cytokines, adhesion molecules, endothelial antigens and increased leukocyte infiltration in all organs suitable for transplantation which was accompanied by a compromised organ function after transplantation (18–21). Therefore grafts derived from brain dead donors lead to a triggering of an accelerated inflammatory response with rapid mononuclear cell infiltrates and increased acute rejection rates (21,22). It has been postulated that predominantly Th1 cytokines are responsible for the enhanced immune response after brain death (23–25). Overall, donor brain death can cause and promote organ injury which can alter the immunological and inflammatory status of the graft before and after transplantation (26,27). This may lead to an increased sensitivity of the organ towards preservation injury and as a consequence to an increased rate of primary nonfunction and acute rejections following transplantation. While this correlation has already been shown experimentally, there has been no clinical trial so far supporting this correlation after liver transplantation. We determined the kinetics and expression patterns of intragraft and serum cytokines after BD and compared those to findings in living donor livers. Consequences of immune activation in grafts from brain dead donors were followed during the transplant procedure and the early posttransplant course.
Material and Methods
Patients and biopsy material
This study was approved by the ethics committee of the Universitätsmedizin Charité, Berlin, Germany. From 2000 through 2002, liver biopsy specimens were obtained during organ recovery from two groups of liver transplant donors: deceased (brain dead, BD, n = 32) and living donor (LD, n = 26, Table 1). The first biopsy was taken immediately after laparotomy (time point 1). A second biopsy was collected immediately prior to organ preservation (time point 2), a third after completion of the vascular anastomosis before reperfusion (time point 3) and a fourth specimen was obtained one hour after reperfusion (time point 4, Figure. 1).
Table 1. Donor factors and pretransplant clinical parameters for deceased (BD) and living donors (LD) (*p < 0.05)
BD (n = 32)
LD (n = 26)
aDefined as infection of the donor requiring antibiotic treatment.
Donor gender (% M:F)
Donor age (years ± SD)
37 ± 18
42 ± 11
ICU (hours ± SD)
72 ± 14*
aDonor infection (yes/no)
Cardiac arrest (yes/no)
Hypotensive period (yes/no)
Ischemia time (min ± SD)
592 ± 151*
95 ± 15
Sample preparation, RNA extraction and cDNA synthesis
Liver tissue biopsy samples were removed at the indicated points, immediately snap-frozen in liquid nitrogen and stored at −80°C until processing. Before analysis, thawing tissues were transferred in 700 μL guanidinium isothiocyanate solution and homogenized using ultrasound. RNA extraction was performed applying the Stratagene Mini Kit according to the manufacturers instruction (Stratagene, Amsterdam, The Netherlands). Samples were tested for genomic DNA contamination and if tested positive excluded from the study. cDNA synthesis and real-time-RT-PCR of nonamplified mRNA were performed as recently described (28). The HPRT housekeeping gene (hypoxanthin-guanin-phosphoribosyl-transferase) was used as an internal standard in the comparative threshold cycle method (2-ΔCt). All primers and probes were designed using Primer Express software (Applied Biosystems, Darmstadt, Germany).
mRNA expression analysis of cytokines
Gene expression was quantified using an ABI PRISM 5700 Sequence Detection System (Applied Biosystems). The amplification took place in a two-step PCR (40 cycles; 15 seconds denaturation step [95°C] and 1 minute annealing/extension step [60°C]). Measurement of the following transcripts was investigated: IL-4, IL-6, IL-10, IFN-γ, TNF-α, TGF-β and MIP-1α. Furthermore we defined the transcription rate of HO-1, CD3 and CD25. The mean Ct values for the housekeeping gene and the genes of interest were calculated from double determinations and samples were considered negative if the Ct values exceeded 40 cycles.
Cytometric bead array system
Blood samples were obtained before explantation immediately after BD. Samples were collected in ethylene-diamine-tetra-acetic acid (EDTA) containing tubes and stored at –70°C until processing. Blood samples were analyzed for the concentration of soluble cytokines including IL-6, IL-10 and TNF-α by Cytometric Bead Array System (CBA) from BD (Becton Dickinson, Heidelberg, Germany) according to the company´s instruction.
Histology and Immunohistology
Liver tissue was fixed in 10% buffered formalin. Paraffin sections were stained with hematoxylin and eosin and assessed by light microscopy. For immunohistology segments were immediately snap frozen in liquid nitrogen. Monoclonal antibodies (mAb), obtained from Serotec (Wiesbaden, Germany), were directed against CD3+ T cells, CD4+ helper cells and CD25+ cells, Serotec, Wiesbaden, Germany). Additional antibodies were directed against MHC class II antigens (OX-3), CD41 (activated and resting platelets) and neutrophil elastase (NP 57, Pharmingen, Ltd, San Diego, CA). After specific mAb staining, sections were stained with rabbit anti-mouse IgG, following mouse anti-alkaline phosphatase complex; counterstaining was performed with hematoxylin. The percentage of CD3+ and CD25+ areas was evaluated by image analysis and morphometric point counting (IPLab software, Scananalytics, Richmond, VA) in 10 livers/group after organ recovery. MHC class II expression within these fields was assessed semiquantitatively. Occasional positive staining of sinusoidal endothelium and Kupffer cells with negative hepatocyte staining was classified as grade 1, positive hepatocyte staining with multiple sinusoids with positive endothelium and positive Kupffer cells was classified as grade 2. Rejection episodes were proven by histology based on standard histological evaluation and hepatitis C (HCV) reinfections were additionally diagnosed by the quantitative assessment of HCV-RNA applying RT-PCR.
Measurement of bilirubin, ALT, AST, serum lactate and INR was performed routinely at the Institute of Clinical Chemistry of the Virchow Clinic, Universitätsmedizin Charité, Berlin.
Statistical significance was ascertained using Student's t test, log-rank-sum test and Mann-Whitney test. The level of significance was P = 0.05 (two-sided) and results are expressed as mean ± SD. Multiple comparisons of cytokine expression levels at various time points between donor groups were analyzed by two-way ANOVA. Bonferroni post-hoc test was applied for individual time point analysis and differences within the groups.
There were no statistical differences between brain dead and living donors regarding age, donor gender, infections, hypotension and perioperative cardiac arrests (Table 1). The ischemia time experienced by organs from brain dead donors as well as the time spent at the intensive care unit was significantly longer in the brain dead donor group (Table 1, p = 0.049). All deceased donors receiving steroids during admission or during intensive care were excluded from the study. In the living donor group exclusively right liver lobes were transplanted.
There were no statistical differences between recipients of organs from brain dead and living donors regarding the age, donor gender and diagnosis (Table 2). The recipients status displayed by assessing the meld score did not reveal significant differences between the groups (meld score average 12.5 BD vs. 13.8 LD, p = ns). The immunosuppressive regimen in both groups was based on Tacrolimus or Cyclosporin A in addition with Mycophenolate Mofetil and steroids. The induction therapy was performed with Basiliximab in all patients. There were no significant differences in the immunosuppressive regimen between both groups, especially the dosage of steroids did not differ significantly (p = ns). The time for vascular anastomosis, defined as warm ischemia time was significantly longer in the living donor group (BD 38 ± 8 min vs. LD 56 ± 4 min, p = 0.0009).
Table 2. Recipient characteristics for deceased (BD) and living donors (LD)
BD (n = 32)
LD (n = 26)
ESLD = end-stage liver disease.
Recipient gender (% M:F)
Recipient age (years ± SD)
38 ± 14
43 ± 13
Budd Chiari syndrome
Primary biliary cirrhosis
Primary sclerosing cholangitis
Cytokine mRNA expression in serial liver biopsies
The analysis revealed that mRNA expression levels of all investigated cytokines were up-regulated in livers of brain dead donors before organ recovery compared to living donors. Especially IL-6 demonstrated an induction (5.6 fold) in brain dead livers before recovery compared to living donor livers and increased over time in both groups (p = 0.0394; Figure 2A). Furthermore mRNA expression of IL-10 illustrated significant differences between both investigated groups (p < 0.0001) and time (p = 0.0202) and was significantly higher expressed in brain dead livers immediately after laparotomy and before organ preservation (p < 0.0001 and p = 0.0010, respectively; Figure 2B). The cytokines including TNF-α, TGF-β and MIP-1α were at all time points significantly increased in brain dead livers versus living donor livers (p < 0.05) (Figures 2C to 2E). Although IL-4 and IFN-γ were consistently higher in brain dead livers than in living donor livers, the mRNA expression displayed a statistically significance only at time point 1 (IL-4: p = 0.0168, IFN-γ: p = 0.0221, Figures 2F, G). Induction of HO-1 mRNA in brain dead livers was only detected immediately after laparotomy (p = 0.0208, Figure. 2H). In contrast, the transcription rate of HO-1 displayed increased expression in living donor livers at the remaining time points compared to brain dead donors (p = ns).
Histology and Immunohistology
As a direct evidence of increased lymphocyte infiltration, we could detect significantly higher levels of CD3 mRNA expression (p = 0.035) in livers from deceased donors compared to living donor livers at all investigated time points (Figure 3A). Furthermore, livers from deceased donors showed an elevated expression of CD25. In particular, after laparotomy and before organ preservation, the CD25 expression was higher in brain dead than in living donor livers (p = 0.0483 and p = 0.0120, respectively, Figure 3B). These findings were confirmed by immunohistological assessments revealing significantly higher levels of CD3+ lymphocytes (BD 2.2%± 0.5% vs. LD 0.7 ± 0.2%, p = 0.0008) and CD25 + cells (BD 4.5%±1.0% versus LD 1.1%± 0.3%, p = 0.0009) prior to transplantation. Stainings for neutrophil elastase showed only slight differences before transplantation (BD 2.8 ± 0.9% vs. LD 2.2 ± 0.6%, p = ns), but increased levels were observed in the brain dead donor group 1 h after transplantation (BD 8.9 ± 1.5% vs. LD 2.4 ± 0.7%, p = 0.0004). CD41 positive deposits of platelets were more frequent one hour after transplantation in liver tissues from brain dead donors compared to living donors. Before transplantation no differences were obvious (p = ns).
In addition MHC class II expression assessed in a semiquantitive manner was significantly higher in organs from brain dead donors (MHC class II: 10/12 sinusoidal endothelium and Kupffer cells positive [grade 2] in brain dead donors versus 2/12 sinusoidal endothelium positive [grade 1] in living donors). Positive hepatocyte staining for MHC class II was observed in brain dead donor organs versus negative staining in living donor organs (p = 0.004).
Concentration of soluble cytokines in brain dead and living donor
Serum concentrations of IL-6 and IL-10 were significantly higher in brain dead donors compared to living donors (p = 0.0001 and p = 0.004, respectively) at all investigated time points. In contrast, serum TNF-α demonstrated only a slightly increase in brain dead donors compared to sera of living donors (p = ns, Table 3, Figures 4A to 4C).
Table 3. Serum concentration of soluble cytokines in deceased (BD) and living donors (LD) before organ recovery (*p < 0.05, **p < 0.01, ***p < 0.001)
BD (n = 32)
LD (n = 26)
BD = deceased donor; LD = living donor.
490 ± 78
3 ± 0.8**
29 ± 10
4 ± 1.7***
4.3 ± 0.6
3.7 ± 1
Initial graft function and rejection rates
To investigate whether the immune activation within the graft due to brain death might have an impact on allograft outcome, we calculated AST and ALT values for 100g liver tissue. The results illustrated that both serum parameters were significantly higher for the deceased donor organ group compared to living donors at the 1st and 3rd postoperative day (AST: p = 0.0001 and p = 0.049; ALT: p = 0.0001 and p = 0.008, Table 4, Figures 5A, B). Total bilirubin as a marker for liver function was significantly higher in the deceased donor organ group at the 10th postoperative day (p = 0.049, Table 4, Figure 5C). The INR assessed at day 3 was in average 1.45 ± 0.2 in the living donor group versus 1.34 ± 0.3 in the brain dead donor group (p = 0.035), the serum lactate did not differ significantly between the groups (p = ns, data not shown). Organs from brain dead donors experienced a significant higher rate of biopsy proven acute rejections compared to ideal living donor organs within 6 and 24 months after transplantation (BD 25% vs. LD 15% at 6 months, p = 0.040, BD 38% vs. LD 28% at 24 months, p = 0.041, Figure 5D). Albeit not significant there was an obvious higher number of primary nonfunction in the BD group leading to a higher rate of retransplantations (BD 10% vs. LD 1.6%; p = ns, Figure 5D). The graft to recipient body weight ratio in the brain dead donor group was in average 1.2 ± 0.3% with a range between 0.89% and 1.65% and in the living donor group in average 2.2 ± 0.3% with a range between 1.5% and 2.8%. Therefore no ‘small for size syndrome' as defined in the literature was observed in the living donor group (29–31).
AST = aspartat-amino-transferase; ALT = alanin-amino-transferase; BD = deceased donor; LD = living donor; POD = postoperative day; results are expressed as mean ± SD.
34 ± 5
7.6 ± 1
2.9 ± 0.5
16 ± 1 ***
7 ± 1 *
3 ± 0.5
34 ± 6
18 ± 1
7 ± 0.5 **
19 ± 1 ***
12 ± 1 **
10 ± 0.5
Total Bilirubin (mg/dL)
4.6 ± 0.74
2.9 ± 0.46 *
The deleterious effects of brain death on donor organs provoke and sustain inflammatory changes that have a negative impact on the outcome after liver transplantation. It has been shown in various experimental brain death models that the lethal central injury leads to a worse organ function which is accompanied by an inflammatory response at the time of organ recovery and transplantation (19,32,33). This process was associated with an increased release and expression of intragraft cytokines (18). Our clinical study confirms previous experimental reports and shows that there is indeed a profound difference in the immunological status between living and brain dead donors. The data indicate the existence of an intense donor response reaction in brain dead subjects. Based on the neuronal and hormonal communication between the central nervous system and peripheral organs, it is likely that alterations in the prospective grafts are related to severe brain injury and associated with somatic reactions before brain death occurs and is verified. This hypothesis is in accordance with previous experimental findings showing a significant upregulation of serum IL-6 in patients suffering from stroke or severe brain injury (34). Furthermore the transcription rate of IL-6, IL-10, TNF-α, TGF-β and MIP-1α in liver tissue of deceased donors was significantly increased, confirming experimental studies where this upregulation resulted in an accelerated rejection of kidney, liver and heart grafts (19,22). These observations confirm recent publications stating that a profound change in cellular regulatory mechanisms is observed after donor brain death. This is followed by alterations in the trancriptional profile of livers after central injury (5).
The importance of an increased immune activation has been demonstrated by several studies which describe the pathway and interaction between cytokines, endothelial antigen upregulation and the subsequent expression of adhesion molecules in human donor organs (27,35,36). In the present study the highest levels of inflammatory cytokines were found in livers failing after transplantation. The upregulation for the majority of investigated cytokines was most obvious after laparotomy of the brain dead donor indicating the downstream summation of deleterious effects of various origins. Besides the immune activation effects mediated by brain death, vascular instability, hypoxemia, the application of catecholamines, for example, contribute to the observed results as well. Surprisingly, further injury as additional ischemia during organ storage did not increase the transcription rate of the measured mediators suggesting that the circumstances before organ recovery lead to an exhaustive immunological activation based on changes in cellular regulatory mechanisms. This is in accordance with data from the United Network for Organ Sharing (UNOS) data base showing that limited ischemia to a certain extent (<24 h) plays a minor role in damaging the graft and does not lead to significant worse results after transplantation. In contrast in living donors even minor additional injury is associated with a slow and steady increase of cytokine production indicating an intact response towards an ongoing damage.
Immunohistological stainings of hepatic tissue revealed upregulated infiltrates of CD3+ lymphocytes in human deceased donors (37,38). This observation is confirmed by experimental studies illustrating an increased rate of CD3+ lymphocytes and macrophages in livers and kidneys (17,39). The observed leukocyte infiltration is mainly promoted by cytokine induced expression of adhesion molecules (e.g. ICAM-1) (20,40), promoting in turn, inflammatory cytokines as IL-1, IL-6, IFN-γ and TNF-α (18,20,33,41). In our study, CD3 antigen as a specific marker for T-lymphocytes indicated a significant infiltration of immunocompetent cells in organs from brain dead donors. Surprisingly the highest expression of cytokines and cellular infiltrates peaks at the time of BD and organ recovery and not after reperfusion. Furthermore higher levels of CD25+ cells were present in tissues of brain dead donor livers before organ recovery, explaining partly the increased number of primary nonfunction after transplantation of organs from deceased donors. These observations have particular relevance regarding the fact that in living donors surgical manipulation, perfusion and other injuries lead to a slow and moderate step by step activation, which however, never reached the intensity of immune activation observed in brain dead donors. Interestingly the brain has the capacity to reduce the peripheral inflammatory response. The function and importance of central anti-inflammatory substances as melanocyte stimulating hormone, thyreotropine releasing hormone and cholinergic pathways is well documented (42,43). On the other side centrally released and produced cytokines can enhance and trigger peripheral inflammation, neutrophil infiltration and edema (44); a finding frequently observed in brain dead organ donors. It is likely that these effects are triggered by a couple of alterations surrounding brain death and not solemnly by the brain death itself.
Protective genes as HO-1 are also upregulated in organs from brain dead donors at the time of organ recovery. However, HO-1 reacts highly sensitive to different stimuli associated with oxidative stress, for example, ischemia, hypotension, or reperfusion (45–47). While HO-1 was increased in livers derived from brain dead donors at the time of BD, with subsequently decreased concentrations in follow-up biopsies, HO-1 in tissues originating from living donors was higher expressed immediately before reperfusion and one hour after restoring the blood flow. The sublethal injury to hepatocytes induced by central injury brain death leads to an inability of cells to express further HO-1 as cytoprotective enzyme, again a consequence of death associated organ failure from active changes in regulatory mechanisms. This concept of cell exhaustion and the inability to sustain the cellular integrity is supported by recent findings showing significant higher rates of apoptosis in livers from brain dead donors (4).
The serum concentrations of inflammatory and anti-inflammatory cytokines as presented by IL-6 and IL-10 suggest that brain injury leads to an unspecific upregulation of protective cytokines in parallel to pro-inflammatory mediators. Different investigators demonstrated significantly higher IL-6 and IL-10 levels after central injury in neurological patients compared to healthy controls, suggesting that deleterious effects in brain death donors start before the final diagnosis of brain death (34,48). This leads, beside the obvious inflammatory response which is well known as ‘systemic inflammatory response syndrome (SIRS)' in intensive care patients, to a systemic partial immunosuppressive reaction resulting in an increased susceptibility to infections in patients with central injury (49). Nevertheless circulating inflammatory cytokines are likely to induce endothelial activation and subsequent systemic inflammatory responses in the future graft. This theory is supported by the observed upregulation of T-cell and macrophage-associated products in organs of anaesthetised living rats after cross-circulation with brain dead rodents, showing that soluble transferred cytokines can lead to an activation in organs from healthy living donors (18).
In the early postoperative course liver enzymes were significantly higher in recipients of organs from brain dead donors compared to ideal living donors, indicating a deteriorated ischemia/reperfusion injury. This may be not of importance for ideal donor organs but can be at least a significant factor for the transplantation of fatty, old or marginal organs leading to a higher rate of initial organ nonfunction. As described in our study a higher rate of primary nonfunction was observed even for organs from so-called ideal brain dead donors outside the criteria for marginal donor organs. It is of outmost interest that in organs failing after transplantation the levels of IL-6 and MIP-1 alpha were highest. Further investigations with a higher number of primary nonfunctions is required to identify reliable markers for initial nonfunction after liver transplantation. Experimental data confirmed these clinical findings and showed a worse I/R injury for various organs leading afterwards to a higher rate of acute rejection (20). Additionally the production and excretion of bilirubin, as a marker for liver function, was despite a reduced transplanted organ mass in the living donor group significantly better compared to the brain dead donor group where the transplanted functional organ mass was in average 50% higher. This fact strongly suggests a significant better organ function of living donor livers that was obvious up to 10 d postoperative. Consequently, PNF and early acute rejection rates were lower in recipients of grafts from living donors. With the exception of the initial nonfunction these effects indeed could be of minor importance for the short-term outcome, but could lead to a higher rate of chronic alterations in transplanted livers over the long-term. This is supported by the rarification of bile ducts and the onset of ischemic type biliary lesions, which is known to be lower in living donor livers. Acute and chronic alterations due to damaging effects of the risk factor brain death have been described for various organs, as for heart, kidney, lung and pancreas (50–56). Beside the effects of cytokine upregulation, hypotension with hypoperfusion as an integral part of brain death may have a major effect on the organ function after transplantation (57). These phenomenons have been described for all solid organs available for transplantation (18,19,58,59), depending on their susceptibility to ischemic injury.
In agreement with experimental results our data underline the importance of the risk factor brain death after human liver transplantation. The central injury is a multifactorial event leading to a significantly deteriorated I/R injury with an increased rate of acute rejections and poorer initial graft function compared to living donors. The transcriptional profile is altered and tends to induce a strong inflammatory response after transplantation. However, whether these findings have a consequence on the long-term function after human liver transplantation still needs to be proven. On the other hand experimental data show beneficial effects in the long-term after the treatment of organs from brain dead donors and the reduction of cytokine levels before organ recovery (60). An improved outcome after deceased donor liver transplantation especially in regard to marginal donor organs will rely on efficient donor management to attenuate the insults related to brain death (61).
This work was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG/Pr 578/2-1/2-2). We thank Annelie Dernier for excellent technological assistance.