Objectives: Graft-versus-host disease after solid-organ transplant is exceedingly rare. Although the precise pathogenetic mechanisms are unknown, a progressive increase in donor chimerism is a requirement for its development. The incidence of mixed donor chimerism and its timeline after simultaneous pancreas-kidney transplant is unknown.
Materials and Methods: After encountering 2 cases of graft-versus-host disease after simultaneous pancreas-kidney transplant at our institution over a period of < 2 years, a collaborative pilot study was conducted by the bone marrow transplant, nephrology, and abdominal transplant surgery teams. We enrolled all consecutive patients undergoing sex-mismatched simultaneous pancreas-kidney transplant over 1 year and longitudinally monitored donor chimerism using fluorescence in situ hybridization for sex chro-mosomes.
Results: We found no evidence for chimerism in our 7 patients. In a comprehensive literature review, we found a total of 25 previously reported cases of graft-versus-host disease after kidney, pancreas, and simultaneous pancreas-kidney transplants. The median onset of graft-versus-host disease was ap-proximately 5 weeks after transplant, with a median of about 2 weeks of delay between first presentation and diagnosis. Skin, gut, and bone marrow were almost equally affected at initial presentation, and fever of unknown origin occurred in more than half of patients. The median survival measured from the first mani-festation of graft-versus-host disease was only 48 days.
Conclusions: Within the limitations related to small sample size, our results argue against an unusually high risk of graft-versus-host disease after simulta-neous pancreas-kidney transplant. Collaboration between solid-organ and stem cell transplant investigators can be fruitful and can improve our understanding of the complications that are shared between the 2 fields.
Key words : Graft-versus-host disease, Immunosuppression, Solid-organ transplant
Acute graft-versus-host disease (aGVHD) is the clinical outcome of alloreactive donor T cells attacking host tissues. Although aGVHD is a frequent com-plication of allogeneic hematopoietic stem cell transplant (allo-HSCT),1 it is rare after solid-organ transplant (SOT)2 and exceedingly rare after blood transfusion.3 The rarity of aGVHD after SOT, first described following a pancreas-spleen trans-plant in 1986,4 is explained, at least partially, by the fact that, compared with HSCT recipients, SOT recipients are less immunosuppressed and can eradicate donor T cells. The incidence of SOT-aGVHD also depends on the transplanted organ. For example, SOT-aGVHD is most common after small bowel (up to 5%)5 and liver (up to 1%) transplants.6,7 This difference has been attributed to the larger number of passenger T cells that enter the host along with the small intestine and liver. The diagnosis is firmly established by persistent donor T-cell chimerism in the bone marrow, blood, or other involved organs. Polymerase chain reaction testing for short tandem repeats is the most sensitive method to detect chimerism.8
After encountering 2 cases of aGVHD after simultaneous pancreas-kidney (SPK) transplant in our institution over a period of < 2 years, the bone marrow transplant, nephrology, and abdominal transplant surgery teams at Washington University School of Medicine (St. Louis, MO, USA) decided to pursue a prospective multidisciplinary effort to investigate the incidence of persistent (or progressive) mixed donor chimerism after SPK transplant. A secondary objective of the study was to determine the early time course of donor chimerism in these transplants. We performed this study because (1) the occurrence of 2 cases of aGVHD after SPK transplant over a relatively short period of time was highly unexpected, (2) the literature on aGVHD following SPK is scarce, and (3) SPK is one of the least common transplants performed both globally and in our center, raising the question of whether unique perioperative immunosuppression or preoperative surveillance may be needed in this particular type of SOT. We limited our sample to sex-mismatched transplants because fluorescence in situ hybridization for sex chromosomes (XY FISH) is a rapid, sensitive, and inexpensive test frequently used for chimerism assessment in the allo-HSCT setting.9 We further extended our work to include a review of the literature on all previously reported cases of aGVHD following pancreas, kidney, and pancreas-kidney transplant procedures.
Materials and Methods
The protocol for this study was approved by our Institutional Review Board, and all patients gave written informed consent before enrollment. In the absence of any prior data on the incidence of aGVHD after SPK transplant and considering that SPK transplant is an infrequent type of SOT, a formal sample size calculation was not performed, and we enrolled all consecutive eligible SPK recipients seen between 2015 and 2016. The inclusion criteria were as follows: (1) patient age > 18 years and (2) sex-mismatched SPK transplant. Patients were excluded if they had a previous history of solid or stem cell transplant, known karyotypic abnormalities, or known primary immunodeficiency disorders.
Approximately 5 mL of venous blood were collected and placed in sterile sodium heparin green-top tubes and transported within 6 hours to the cytogenomics and molecular pathology laboratory. The first sample was obtained within 5 days after the transplant and before the patient was discharged from the hospital. A subsequent sample was obtained on day 28 (± 7 days) after transplant in all patients. If any level of donor chimerism was detected at one or both time points, the protocol allowed for additional samples at 8 and 12 weeks posttransplant. All samples were de-identified, and the laboratory was blinded to patient and donor sex. Standard XY FISH on 200 cells per sample was performed for detection of donor chimerism, as previously described.9
Table 1 shows the characteristics of our patients. We performed 13 SPK transplants during the study period; all 7 patients who received a sex-mismatched transplant were enrolled. All HLA mismatches were bidirectional except for one patient whose transplant was mismatched at locus A in the host-versus-graft but not graft-versus-host direction. The median (range) ages of patients and donors were 42 years (36-60 y) and 23 years (19-35), respectively. Uniform preoperative induction and posttransplant pro-phylaxis were administered according to our institutional algorithm. Induction was with rabbit antithymocyte globulin (ATG; Thymoglobulin, Sanofi, Cambridge, MA, USA) and methylprednisolone. Maintenance was with tacrolimus, mycophenolic acid, and prednisone. The target total dose of rabbit ATG was 7 mg/kg over 4 days, which was adjusted for leukopenia or thrombocytopenia. The initial methyl-prednisolone dose was 7 mg/kg to a maximum of 500 mg intravenously at the start of the transplant operation. The target tacrolimus level was 8 to 12 ng/mL for the first month and then 5 to 10 ng/mL thereafter. The initial dose of mycophenolic acid was 720 mg twice daily, which was adjusted for leukopenia or thrombocytopenia. The maintenance dose of prednisone was initially 1 mg/kg for 3 days, which was tapered to 5 mg daily by week 5.
The first chimerism assay was performed on the day of transplant (day 0) in 1 patient, day 1 in 2 patients, day 2 in 1 patient, day 3 in 2 patients, and day 5 in 1 patient. All of these assays showed no donor chimerism, with all 200 cells being of recipient origin. The second chimerism assay was performed on days 22, 24, 27, 28, and 34 in 1, 3, 1, 1, and 1 patient. Similar to the first assay, all 200 cells were of recipient origin in all assays. Because chimerism did not exist in any of our patients at the first 2 time points, repeat assays at future time points were not performed. Two patients experienced a major complication after the transplant (between the first and second chimerism assays): the pancreas allograft had to be removed due to thrombosis in patient 2 and the kidney allograft had to be removed due to torsion in patient 5. Patient 2 showed normal renal function but still continues to be insulin-dependent, whereas patient 5 is off insulin but continues to be treated with hemodialysis. All patients are alive and otherwise have normal allograft function at last follow-up, with no clinical or laboratory manifestations of mixed chimerism or GVHD at any time point.
We found a total of 25 previously reported cases of GVHD following kidney, pancreas, or SPK transplant10-25 (Tables 2 and 3). This list was compiled by a search of Medline using the following key word combinations: “graft-versus-host disease” OR “graft versus host disease” OR “GVHD”; AND “kidney” OR “renal” OR “pancreas” OR “pancreatic”; AND “allograft” OR “allograft” OR “transplant.” This search resulted in 588 unique records, 23 of which were ultimately selected for inclusion in this series. The criteria for selection were as follows: (1) kidney, pancreas, or SPK transplant; (2) development of GVHD; (3) no evidence or suspicion for transfusion-associated GVHD. The full text (when available) of the extracted reports were reviewed, and their citations were considered for potential additional reports.
Fifteen reports (60%) were after 2000, and 10 reports (40%) were after 2010. The transplanted solid organ was kidney in 10 (40%), pancreas in 8 (32%), and SPK in 7 patients (28%). The transplant was from a deceased donor in 13 cases (52%), living donor in 9 cases (36%), mixed (deceased donor pancreas and living donor kidney) in 2 cases (8%), and unknown in 1 case (4%). Recipient sex was male in 12 (48%), female in 5 (20%), and unknown in 8 (32%). Donor sex was male in 7 (28%), female in 6 (24%), and unknown in 12 (48%). The median (range) age of patients was 37 years (14-62 y), and the median (age) of donors was 23 years (16-52 y). The most common indication for transplant was type 1 diabetes in 12 patients (48%), unknown in 9 patients (36%), and other reasons in the remaining 4 patients (16%).
The number of graft-versus-host direction antigen mismatches (HLA-A, -B, and -DR) was 0 in 1 (4%), 2 in 2 (8%), 3 in 5 (20%), and 5 in 5 patients (20%), with unreported results in 12 patients (48%). The median (range) time from transplant to the first presentation of GVHD was 37 days (7 days to 16 months), and the median (range) time from the first presentation of GVHD to diagnosis was 13 days (2 days to 18 months). The earliest organ showing clinical manifestation of GVHD was the gut in 7 patients (28%), skin in 5 patients (20%), and bone marrow in 5 patients (20%). The first involved organ was unreported in 7 patients (28%), and more than one organ was involved in 1 patient (4%). During the course of the disease, the most frequently affected organ by GVHD was the gut (17 cases, 68%), followed closely by skin (15 cases, 60%), bone marrow (14 cases, 56%), and liver (13 cases, 52%). Fever of unknown origin was a common clinical presentation (14 cases, 56%).
Of 20 cases with a reported positive diagnostic modality, a positive blood or marrow XY FISH, blood or marrow polymerase chain reaction, tissue XY FISH, and tissue polymerase chain reaction was the basis for diagnosis in 5 cases (25%), 2 cases (10%), 7 cases (35%), and 3 cases (15%), respectively. This included 5 patients (25%) who had more than one positive diagnostic modality. The diagnosis was established based only on morphology of biopsied specimens in 8 patients (40%). Treatment was conceptually classified as escalation versus de-escalation of immunosuppression. These approaches were used in 15 patients (60%) and 3 patients (12%), respectively, and the treatment was not reported in the remaining 7 patients (28%).
Survival data were available for 17 reports. The median survival measured from the first mani-festation of GVHD was only 48 days, with a 6-month survival of 21% (Figure 1). Of the 10 patients with reported cause of death, 5 died with sepsis and 5 with multiorgan failure. The small sample size prohibited an analysis of survival predictors.
Although major advances have occurred in our understanding of aGVHD pathogenesis after allo-HSCT, our understanding of the events (and their mechanisms) resulting in SOT-aGVHD remains poor. It is presumed that pretransplant defects (genetic or acquired) in the immune system of SOT-aGVHD patients make them unable to eliminate donor T cells, ultimately resulting in T-cell-mediated tissue injury and aGVHD. The use of ATG as a component of preoperative induction may cause such a profound T-cell-depleting effect on the recipient that promotes expansion of donor T cells. Although donor T cells are inevitably affected by ATG (due to its long half-life), they may be less severely affected than recipient T cells, which are also influenced by other com-ponents of induction with shorter half-lives. Com-paring the incidence of SOT-GVHD in transplants using versus not using ATG would be potentially informative. It has been suggested, with some empiric evidence, that persistent microchimerism is common after SOT26,27 and may improve transplant outcomes by decreasing rejection rates. Unopposed activation of donor T cells, on the other hand, can cause tissue injury and GVHD. Importantly, although most transplant centers perform HLA-A, -B, and -DR typing on patients and donors, testing for HLA-C, -DP, and -DQ may offer additional information about graft survival and perhaps the risk of GVHD. The direction of the mismatch at each locus may also be important in GVHD. Although a mismatch in the graft-versus-host direction may increase the risk of GVHD, a mismatch in the opposite direction would be expected to help patient T cells eradicate donor T cells.
In this study, we focused on donor chimerism after SPK transplants because of the unexpected occurrence of aGVHD in 2 cases at our center over a short period of time. With the limits posed by the infrequent nature of SPK transplants, a larger study was not feasible; hence, we enrolled all eligible patients over 1 year. Although no definitive conclusions can be drawn from this small sample, our results showing no detectable chimerism argue against an unusually high risk of SOT-aGVHD after SPK transplants. In addition, with undetectable donor chimerism shortly after the transplant and 4 weeks after the transplant, the emergence of significant chimerism at a later time point would be unlikely. However, given the lower sensitivity of FISH compared with polymerase chain reaction-based chimerism assays, we cannot rule out the possibility of lower levels of microchimerism below our detection threshold.
We found a total of 25 reports of GVHD after kidney, pancreas, or SPK transplant. Although the small sample size was a major limiting factor for detailed analysis, a number of important obser-vations were made. (1) The median onset of GVHD was approximately 5 weeks after transplant; however, GVHD can occur as early as 1 week and as late as several months after transplant. (2) The diagnosis tends to be established relatively late, with a median interval between first presentation and diagnosis of approximately 2 weeks. Given the aggressive and often recalcitrant nature of SOT-GVHD, early diagnosis is crucial for successful treatment, highlighting the value of high clinical vigilance. Suspecting the diagnosis is particularly challenging because of the nonspecific symptoms of GVHD, which may overlap with those of drug toxicity and infections after SOT. (3) Skin, gut, and bone marrow are almost equally affected at initial presentation, although liver involvement occurs as frequently during the course of the disease. (4) Fever of unknown origin is a common clinical presentation, occurring in more than half of patients. Fever is probably caused by the cytokine storm arising from T-cell alloreactivity and tissue injury. (5) Although a variety of therapies have been attempted, there are currently no hints for superior efficacy of one versus another, and most patients fail to achieve a durable response and die shortly after diagnosis.
Another knowledge gap in SOT-GVHD is ap-propriate treatment(s). Although our progress over several decades in treatment of GVHD following allo-HSCT has been poor and mainly limited to systemic corticosteroids, advances in treatment of SOT-aGVHD have been even more primitive. This unfortunate situation is due to a combination of the rarity and complexity of SOT-GVHD. Escalation and de-escalation of immunosuppression (attempting to suppress alloreactive T cells and augment the host immunity, respectively) have both been implemented in previous cases of SOT-GVHD.28 However, mortality remains high. There is an unmet need to better understand the pathophysiology of SOT-GVHD and to improve patient outcomes. Consi-dering the unfeasibility of prospective trials, mathematical modeling was used recently to capture several puzzling features of SOT-aGVHD.29 These included (1) stable mixed donor chimerism in some cases,30-32 (2) full donor chimerism in rare cases, (3) fluctuations in the level of chimerism in some cases,30 and (4) beneficial effects of escalating immunosup-pression in some cases and detrimental effects in others.28,32 There may be opportunities to apply observations and discoveries in the field of allo-HSCT to SOT-GVHD, as interdisciplinary efforts such as the one in the present study might remove some of the barriers to progress in SOT-GVHD.
Volume : 16
Issue : 3
Pages : 307 - 313
DOI : 10.6002/ect.2016.0299
From 1Bone Marrow Transplantation and Leukemia Section, Division
of Oncology, 2Division of Nephrology, 3Cytogenomics and
Molecular Pathology Laboratory, Division of Laboratory and Genomic Medicine,
Department of Pathology and Immunology, and 4Department of Surgery,
Section of Abdominal Transplantation, Washington University School of Medicine,
St. Louis, Missouri, USA
Acknowledgements: The authors have no conflicts of interest to declare. The study was partially supported by intramural funding from the Cytogenomics and Molecular Pathology Laboratory, Division of Laboratory and Genomic Medicine, Department of Pathology and Immunology, Washington University School of Medicine, St Louis, MO. AR, DCB, and AFC developed the concept and designed the study. AR collected the data, performed the analysis, and wrote the first draft of the manuscript. JRW, IEA, DCB, and AFC critically reviewed the results and the manuscript. The authors thank Sandra Crocker for assistance with sample processing.
Current address of A. Rashidi: Division of Hematology, Oncology, and Transplantation, University of Minnesota, Department of Medicine, Minneapolis, MN 55455.
Corresponding author: Armin Rashidi, Division of Hematology, Oncology, and Transplantation, University of Minnesota, Department of Medicine, 14-100 PWB, MMC480, 420 Delaware Street SE, Minneapolis, MN 55455, USA
Phone: +1 612 301 1894
Table 1. Patient Characteristics
Table 2. Previous Cases of Graft-Versus-Host Disease after Kidney, Pancreas, or Pancreas-Kidney Transplant
Table 3. More Details on Cases Shown in Table 2
Figure 1. Survival Measured from Diagnosis of Graft-Versus-Host Disease