Objectives: Complement activation in organs from deceased donors is associated with allograft injury and acute rejection. Because use of organs from donors after circulatory death is increasing, we characterized relative levels of complement activation in organs from donors after brain death and after circulatory death and examined associations between donor complement factor levels and outcomes after kidney and liver transplant.
Materials and Methods: Serum samples from 65 donors (55 donations after brain death, 10 donations after circulatory death) were analyzed for classical, lectin, alternative, and terminal pathway components by Luminex multiplex assays. Complement factor levels were compared between groups, and associations with posttransplant outcomes were explored.
Results: Serum levels of the downstream complement activation product C5a were similar in organs from donors after circulatory death versus donors after brain death. In organs from donors after circulatory death, complement activation occurred primarily via the alternative pathway; the classical, lectin, and alternative pathways all contributed in organs from donors after brain death. Donor complement levels were not associated with outcomes after kidney transplant. Lower donor complement levels were associated with need for transfusion, reintervention, hospital readmission, and acute rejection after liver transplant.
Conclusions: Complement activation occurs at similar levels in organs donated from donors after circulatory death versus those after brain death. Lower donor complement levels may contribute to adverse outcomes after liver transplant. Further study is warranted to better understand how donor complement activation contributes to posttransplant outcomes.
Key words : C5a complement activation product, Kidney transplant, Liver transplant
Organ transplantation is the gold standard therapy for patients with end-stage organ failure. Despite improvements in patient and graft survival,1,2 inflammation and immune activation remain inevitable consequences of transplant that limit short- and long-term posttransplant outcomes.3-6 Accumulating evidence suggests that both the innate and adaptive immune systems contribute to the immune response to transplantation. Components of these systems act in concert to cause immediate allograft injury and the development of immune memory against the allograft that predisposes to ongoing injury.7-9
The complement cascade is a central component of the innate immune system that is implicated in immediate and delayed allograft injury.9 Complement activation is a well-recognized sequela of ischemia-reperfusion injury, contributing to inflammation and tissue injury that compromise early allograft function.4,5,8,10 The complement system additionally plays a role in antigen presentation and can stimulate the development of adaptive immune responses against allograft tissue, leading to acute and chronic rejection.8,9,11
Complement activation and inflammation beginning in the organ donor prior to transplant may contribute to suboptimal posttransplant outcomes.12-14 It is well-known that brain death triggers a systemic inflammatory response, characterized in part by activation of the complement cascade.15-17 In accordance with the pathophysiology of brain death, complement levels are elevated in serum and tissue samples from donation after brain death (DBD) organ donors and are associated with poor early and late graft function and a heightened incidence of acute rejection compared with levels found in organs from living donors.14,18,19 Although DBD donors provide most of the organs for transplantation, use of donation after circulatory death (DCD) donor organs is increasing.20 Limited evidence suggests that DCD donors exhibit a milder inflammatory phenotype compared with DBD donors, with relatively greater importance of apoptosis and hypoxia-related pathways.21-24 Although several studies of complement activation in deceased donors have included both DBD and DCD donors,13,25 the relative levels of complement activation in these 2 distinct populations remain uncharacterized.
Accordingly, we aimed to characterize relative levels of systemic complement activation in DBD versus DCD donors. We measured serum levels of classical, lectin, alternative, and terminal pathway complement components to determine which of the 3 pathways are predominantly activated in each donor subgroup and explored the associations between levels of complement activation products and perioperative clinical outcomes in kidney and liver transplant recipients.
Materials and Methods
This study was approved by the Duke University Institutional Review Board and by Carolina Donor Services (Durham, NC, USA). Signed written informed consent was received from study participants or authorized legal representatives for the use of samples for research purposes prior to inclusion in this study. None of the donors were prisoners at the time of donation. Donor data were obtained from the United Network for Organ Sharing and a research protocol approved by Carolina Donor Services.
Sample collection and processing
Peripheral blood was collected from healthy donors and deceased organ donors by venipuncture. Blood samples from deceased organ donors were collected in serum separator tubes (Beckton, Dickinson and Company, Franklin Lake, NJ, USA) in the operating room before organ procurement and were stored at 4 °C for less than 24 hours prior to processing. Serum was isolated by centrifugation at room temperature for 10 minutes at 500× gravity (g) to separate cells, followed by transfer to a 15-mL conical tube and subsequent centrifugation at 1200× g for 15 minutes to deplete platelets. Serum samples were aliquoted and stored at -80 °C until use.
Serum complement analysis
Serum concentrations of human complement factors were measured in duplicate using the Luminex platform with MILLIPLEX MAP Human Complement Panel 1 and Panel 2 (Millipore Sigma, Burlington, MA, USA). Assays were performed according to the manufacturer’s recommended protocol and read using a Bio-Plex 200 array reader (Bio-Rad, Hercules, CA, USA). Data were analyzed using Bio-Plex Manager software (Bio-Rad). Although complement component C5a has a particularly short serum half-life,26 the assay manufacturer has reported consistent quantification of C5a above the limit of detection and stable concentrations with repeated thawing and freezing to -80 °C for sample processing and storage, respectively. Although some cross-reactivity with the stable metabolite C5a des arginine has been noted, this is thought to be minimal and has not yet been validated.
Posttransplant clinical outcomes
All patients who underwent kidney or liver transplant with organs from donors who provided samples for the study were included. Pediatric (age <18 years) and multiorgan transplant recipients were included due to the small number of patients in the study.
Variables of interest were abstracted from patient charts and the United Network for Organ Sharing database. Recipient outcomes common to both kidney and liver transplant procedures included reintervention (surgical, endoscopic, radiologic, or biopsy) within 30 days of transplant, intensive care unit readmission within 30 days of discharge from the index hospitalization, hospital readmission within 30 days of discharge from the index hospitalization, mortality within 90 days of transplant, and the occurrence and timing of any biopsy-proven acute rejection episodes. Additional organ-specific perioperative outcomes of interest included delayed graft function (DGF) and discharge with a Foley catheter for kidney transplant recipients and need for blood product transfusion within 72 hours posttransplant, occurrence of posttransplant medical complications (neurologic, cardiovascular, pulmonary, abdominal, renal, infectious), early allograft dysfunction, and primary allograft failure for liver transplant recipients. Delayed graft function was defined as the need for postoperative dialysis within 7 days of transplant. Early allograft dysfunction was defined as aspartate aminotransferase or alanine aminotransferase >2000 U/L within 7 days of transplant, international normalized ratio >1.6 on posttransplant day 7, or total bilirubin >10 mg/dL on posttransplant day 7.27
Characteristics of DBD and DCD donors were compared using Wilcoxon rank-sum tests for continuous variables and chi-square and Fisher’s exact tests for categorical variables. Complement factor levels were compared between DBD, DCD, and healthy donor groups using Kruskal-Wallis with Dunn’s multiple comparison tests with a Bonferroni correction for multiple testing. Spearman correlation coefficients were used to determine associations between C5a levels and levels of all other complement factors in donor serum samples. Since C2, C4, and C4b are common to the classical and lectin pathways, the specific contributions of classical and lectin pathway activation to C5a generation were explored using partial correlation analysis, examining associations between levels of C5a versus C2, C4, and C4b, respectively, adjusting for C1q (independent contribution of the lectin pathway) and mannose-binding lectin (MBL) (independent contribution of the classical pathway).
For each clinical outcome, Wilcoxon rank-sum tests were used to compare levels of complement factors in donor serum samples between recipients with and without the outcome of interest. No adjustment for multiple comparisons was implemented as analyses of complement factor levels and clinical outcomes were primarily exploratory. A 2-sided P value less than .05 was considered statistically significant. All analyses were performed using R version 3.6.1 (Vienna, Austria).
Characteristics of the donors included in this study have been described previously.12 Overall, samples were collected from 65 deceased donors of whom 55 met DBD criteria and 10 met DCD criteria. For comparison, samples were additionally collected from 10 living healthy donors. Demographic characteristics were similar between DBD and DCD groups as shown in Table 1.
Serum complement levels
To compare levels of complement activation between DBD, DCD, and healthy donors, we measured serum levels of classical, lectin, alternative, and terminal pathway complement components (Figure 1). As indicated in Table 2, we found significantly higher levels of C2, C5a, and properdin in both DBD and DCD donors compared with levels found in serum from healthy donors. Serum levels of C1q, C4b, C5, factor I, and factor B were significantly higher in DBD donors than in healthy donors but were similar between DCD donors and healthy donors. Compared with DBD donors, significantly lower levels of C2, C4b, and factor D adipsin were found in DCD donor serum. Conversely, serum levels of C3 were significantly higher among DCD donors and healthy donors compared with DBD donors.
Since production of C5a is common to the classical, lectin, and alternative pathways of complement activation, we examined correlations between levels of upstream complement factors and levels of C5a to determine the relative activation of each pathway in DBD and DCD donors. Among DBD donors, C2, C3b, C4, C4b, factor B, and factor H were positively correlated with C5a (all P < .05). Partial correlation showed that C2 was associated with C5a independent of MBL (C2/C5a corrected for MBL: r = 0.31, P = .02) and independent of C1q (C2/C5a corrected for C1q: r = 0.38, P = .005). C4 was associated with C5a independent of MBL (C4/C5a corrected for MBL: r = 0.27, P = .046) but not independent of C1q (C4/C5a corrected for C1q: r = 0.19, P = .2). Among DCD donors, only factor I, factor B, and factor H were significantly correlated with C5a (factor I/C5a: r = 0.77, P = .014; factor B/C5a: r = 0.76, P = .016; factor H/C5a: r = 0.72, P = .02) (Table 3).
Association between donor complement levels and post-transplant clinical outcomes
A total of 15 kidney and 25 liver transplants were performed using organs from donors who provided samples for this study, including 1 simultaneous heart-kidney, 1 simultaneous pancreas-kidney, and 3 simultaneous liver-kidney transplants. Perioperative complications that occurred among recipients in the study are summarized in Table 4.
The associations between donor complement levels and perioperative complications in liver transplant recipients are summarized in Tables 5 to 8. Among kidney transplant recipients, donor complement levels were similar between those with and without clinical complications. Among liver transplant recipients, donor levels of C2 and MBL were significantly lower for patients who required blood product transfusion within 72 hours posttransplant compared with those who did not (median values were 9.1 vs 23.2 μg/mL for C2, P = .03, and 3.9 vs 10.0 µg/mL for MBL, P = .04) (Figure 2A and 2B; Table 5). Donor C4b levels were also significantly lower for patients who required posttransplant reintervention (median of 17.8 vs 22.3 μg/mL, P = .04) and those with acute rejection within 1 year (median of 12.8 vs 21.1 μg/mL, P = .015) (Figure 2C and 2D; Table 6 and Table 7). Donor levels of C3 and properdin were likewise significantly lower for recipients who were readmitted within 30 days of discharge from the transplant hospitalization (median of 171 vs 262 μg/mL for C3, P = .049, and 19.1 vs 29.5 μg/mL for properdin, P = .043) (Figure 2E and 2F; Table 8). No significant associations were found between donor complement levels and early allograft dysfunction, intensive care readmission within 30 days of discharge, or postoperative medical complications or biopsy-proven acute rejection within 30 days posttransplant.
As we continue to face a critical shortage of organs for transplantation, use of DCD organs has increased in an attempt to maximize the potential donor pool.20 While a growing body of evidence suggests that outcomes after DCD transplant are acceptable,28,29 better understanding of DCD donor physiology is needed to optimize outcomes for recipients of these organs. Activation of the complement system in DBD donors contributes to allograft injury and is associated with poor graft function and acute rejection after kidney transplant.13,14,18,19 To date, the role of the complement system in DCD donors and its contribution to perioperative outcomes in kidney and liver transplant recipients remain undefined and may be complicated by heterogeneity in underlying pathophysiology among DCD donors.
In this study, we found that serum levels of the downstream complement activation product C5a were significantly higher in DCD donors compared with living healthy controls. Serum samples from DCD donors demonstrated similar levels of C5a versus levels shown in DBD donors, suggesting that activation of the complement cascade occurs to similar degrees in both DCD and DBD donors.
Although DBD donor organ physiology is characterized in large part by a profound systemic inflammatory response to brain death,17 current understanding suggests that DCD donor organ physiology is characterized primarily by cell death and response to hypoxia that may result from warm ischemia that is unique to DCD organs.21-23,30 Indeed, several studies of transcriptional profiles in DBD and DCD organs have identified upregulation of hypoxia response genes in DCD organs with comparatively lower levels of proinflammatory gene transcription compared with DBD organs.21-23,30 These findings are further supported by studies demonstrating elevated levels of proinflammatory cytokines in serum samples from DBD versus DCD donors and demonstrating increased leukocyte infiltration in prereperfusion renal allograft biopsies from DBD versus DCD and living donor kidneys.24,31 As a central component of the innate immune response to injury, the complement cascade is activated by both brain death and ischemia.16,32 Our findings of comparable complement activation in DCD versus DBD donors suggests that, in the pretransplant setting, brain death and ischemia are both similarly potent activators of the complement cascade.
Because the physiologic basis for complement activation differs between DCD and DBD donors, it may reasonably be expected that the dominant pathways of complement activation differ between these groups. In studies of unstratified cohorts of deceased donors, all 3 of the classical, lectin, and alternative pathways have been shown to contribute to generation of the terminal membrane attack complex.13,14 Consistent with these findings, levels of C5a in serum from DBD donors in our study were positively correlated with classical/lectin (C2, C4, C4b) and alternative (factor B, factor H) pathway factors, suggesting that all 3 pathways contributed to C5a generation in this group. Furthermore, C2 was associated with C5a independent of MBL and C1q, suggesting specific contributions from each of the classical and lectin pathways in DBD donors. Notably, current understanding suggests that C1q exhibits many anti-inflammatory and immune regulatory functions beyond the complement cascade; nonetheless, it continues to be recognized as a key initiator of the classical pathway, supporting our use of C1q as a means to separate the contribution of the classical pathway from that of the lectin pathway in DBD donors.33
In contrast, levels of C5a in serum from DCD donors were only associated with factors B, H, and I, suggesting that the alternative pathway was the primary contributor to complement activation in this group. Despite statistical significance, levels of upstream complement components were weakly correlated with levels of C5a in DBD donors (all r < 0.4), whereas factors B, H, and I were strongly correlated with C5a in DCD donors (all r > 0.7). These findings may further emphasize the dominant contribution of the alternative pathway to complement activation in DCD donors, compared with a unified process involving the classical, lectin, and alternative pathways in DBD donors. Whereas damage- and pathogen-associated molecular patterns exposed during brain death may activate the classical pathway in DBD donors,16,17 ischemic injury, which factors more prominently in DCD donor physiology, may preferentially activate the alternative pathway.14 Furthermore, in the absence of the potent immunogenic stimuli associated with brain death, sole upregulation of the constitutively active alternative pathway in DCD donors may additionally explain the lower levels of upstream complement factors including C2 and C4b in this group. Improved understanding of the mechanisms of complement activation in DBD versus DCD donors may inform efforts to target inflammatory pathways in deceased donors to optimize allograft quality prior to organ donation.
Importantly, our findings of complement activation via the alternative pathway in DCD donors may provide new understanding of the heightened occurrence of early posttransplant complications in recipients of organs from these donors. In particular, rates of DGF are significantly higher among recipients of DCD versus DBD donor kidneys, which may be attributed to the more severe ischemic insult associated with DCD organ donation.30,34,35 However, in a rodent model of kidney transplant, Yu and colleagues found that treatment with a selective inhibitor of the alternative complement pathway improved posttransplant renal function and graft survival in treated versus untreated rodents.32 In light of these findings, our study suggests that complement activation via the alternative pathway in DCD donors may contribute to the heightened occurrence of DGF in recipients of these donor kidneys. However, further studies are needed to elucidate the relationship between ischemia, alternative pathway activation, and DGF in the clinical setting.
Despite evidence of an association between donor complement activation and suboptimal outcomes after kidney transplant,13,14,18,19 the relationship between donor complement activation and outcomes after liver transplant remains unclear. In this study, we found that donor complement levels were significantly lower for liver transplant recipients who required posttransplant blood product transfusion, reintervention, and hospital readmission and for those who had an episode of acute rejection within 1 year. In understanding these findings, it is worthwhile to note that the liver synthesizes the majority of the body’s soluble complement factors, which may complicate the relationship between systemic complement levels and liver-related adverse outcomes.36 Indeed, in a study of complement levels in acute liver failure, Melgaco and colleagues found that serum levels of C3a were significantly lower in samples from patients with acute liver failure compared with healthy individuals.37 Conversely, liver tissue samples from patients with acute liver failure demonstrated increased staining for complement factors, including C3a, C5a, and the membrane attack complex,37,38 suggesting that complement activation in the setting of poor liver function may be reflected primarily by deposition in liver tissue. In liver transplant, positive staining for C4d in liver biopsies may help differentiate acute rejection from posttransplant viral hepatitis reactivation,39,40 supporting an association between complement deposition in the liver allograft and immunologic posttransplant complications. Likewise, evidence from a rodent model of liver transplant suggests that the presence of terminal complement factors in donor livers at the time of transplant may increase the severity of ischemia-reperfusion injury, contributing to cell death, inflammation, and poor graft function regardless of recipient complement status.41 Our findings may be understood in the context of these studies, which suggest that, whereas complement levels were lower in donor serum samples for recipients who experienced clinical complications, complement deposition may have been present in corresponding donor liver tissue. Unfortunately, donor liver samples were unavailable in our study; future work should make specific provisions to obtain donor biopsy samples to test this hypothesis.
In light of the associations between donor complement activation and posttransplant complications, pharmacologic inhibition of the complement cascade in deceased donors prior to transplant may represent a promising therapeutic avenue by which to improve posttransplant outcomes. Recipient treatment with complement-targeted therapeutics, including eculizumab, an anti-C5 monoclonal antibody, and C1 esterase inhibitor (C1-INH), has previously been explored as a means to prevent ischemia-reperfusion injury and DGF after kidney transplant.42-44 In 2 randomized controlled trials, neither eculizumab nor C1-INH reduced rates of DGF in treated versus untreated recipients; however, patients who received C1-INH and developed DGF required fewer hemodialysis sessions after the first posttransplant week, suggesting that although C1-INH administration in transplant recipients may not prevent DGF, it may reduce DGF duration, facilitate renal recovery, and portend sustained benefits in graft function over time.42-44 More recently, Danobeitia and colleagues found that in a nonhuman primate model of kidney transplant, recipients of kidneys from donors treated with C1-INH experienced significantly lower rates of DGF and demonstrated superior posttransplant renal function compared with recipients of untreated donor kidneys.45 These results suggest that although complement inhibition in transplant recipients may improve renal function over time, complement inhibition in deceased donors may not only improve renal function, but may additionally prevent DGF from occurring altogether. Furthermore, in contrast to eculizumab, C1-INH acts upstream on the classical pathway, blocking the generation of complement activation products, including C3a and C4b, mediators that were associated with posttransplant complications in our study.46 Although complement therapeutics remain an area of active investigation, our findings suggest that agents such as C1-INH that inhibit upstream complement activation and consequent generation of complement activation products may offer a novel means by which to optimize allograft function and quality in the deceased donor prior to transplant.
There are several limitations to our study that warrant discussion. First, we did not have access to DCD-specific data, such as duration of pretransplant warm ischemia, which may influence complement activation and posttransplant outcomes. In addition, as previously noted, donor tissue samples were not available, which limited our ability to comment on the association between local donor complement levels and posttransplant outcomes in both kidney and liver transplant recipients. Although future studies should examine the association between donor tissue complement levels and posttransplant outcomes, additional consideration of the association between tissue complement levels and evidence of tissue injury, including apoptosis or necrosis, is needed to understand how complement activation in the deceased donor may affect allograft quality prior to transplant. Finally, given the relatively small number of transplant recipients and perioperative complications in our study, we may have been underpowered to detect differences in donor complement levels between patients who did and did not have complications of interest. Additionally, multiorgan transplant recipients may be more likely to have complex posttransplant courses with a higher rate of clinical complications, which may have contributed to the complications that were seen in our study. Further studies examining a larger cohort of deceased donors and corresponding isolated kidney or liver transplant recipients may help clarify these associations, providing evidence to direct management decisions such as the use of complement-targeted therapeutics to optimize posttransplant outcomes.15
To our knowledge, this is the first study to define the relative activity of the complement system in DCD versus DBD donors. We have demonstrated that the complement cascade in DCD donors is activated to a similar extent compared with that shown in DBD donors and occurs primarily via the alternative pathway. While our study also provides preliminary evidence to support an association between decreased systemic donor complement activation and perioperative complications after liver transplant, further investigation is warranted to better understand the relative contributions of systemic and local donor complement activation to outcomes after liver transplant, as well as to elucidate specific associations between DCD donor complement activation and posttransplant outcomes in the clinical setting.
Volume : 19
Issue : 7
Pages : 635 - 644
DOI : 10.6002/ect.2020.0425
From the 1School of Medicine, Duke University, Durham, North Carolina; the 2Duke University, Durham, North Carolina; the 3Department of Surgery, Duke University Medical Center, Durham, North Carolina; and the 4Department of Surgery, Cedars-Sinai Medical Center, Los Angeles, California, USA
Acknowledgements: The authors have no conflicts of interest to report. This work was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number TL1TR002555 (SEH) and a grant from the Duke Department of Surgery (JP, TVB). Luminex assays were performed under the management of Dr. Andrew N. Macintyre and direction of Dr. Gregory D. Sempowski in the Immunology Unit of the Duke Regional Biocontainment Laboratory, which received partial support for construction from the National Institute of Allergy and Infectious Diseases (UC6-AI058607). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or Duke Department of Surgery.
Author contributions: Conception and design of the study were performed by TV Brennan, AS Barbas, and J Pollara; data collection was performed by SE Halpern and J Pollara; research was performed by RW Edwards and J Pollara; analyses and interpretation of data were performed by SE Halpern, CK Rush, RW Edwards, AS Barbas, and J Pollara; all authors contributed to drafting and critically revising the article and final approval of the version to be published.
Corresponding author: Justin Pollara, Duke University Medical Center, 915 South LaSalle Street Extension, SORF Building Room 116, Durham, NC 27710, USA
Phone: +1 919 684 4043
Table 1. Deceased Donor Characteristics
Figure 1. Serum Complement Levels by Donor Type
Table 2. Complement Levels in Donor Serum Samples
Table 3. Correlation Between Levels of Upstream Complement Factors and C5a
Table 4. Perioperative Complications Among Kidney and Liver Transplant Recipientsa
Table 5. Donor Complement Levels Among Liver Transplant Recipients Who Did Versus Did Not Require Blood Product Transfusion Within 72 Hours Posttransplant
Table 6. Donor Complement Levels Among Liver Transplant Recipients Who Did Versus Did Not Require Reintervention Within 30 Days of Transplanta
Table 7. Donor Complement Levels Among Liver Transplant Recipients Who Did Versus Did Not Have an Episode of Biopsy-Proven Acute Rejection Within 1 Year of Transplant
Table 8. Donor Complement Levels Among Liver Transplant Recipients Who Did Versus Did Not Require Hospital Readmission Within 30 Days of Discharge From the Index Hospitalization
Figure 2. Donor Complement Factors Associated With Clinical Complications in Liver Transplant Recipients