Major Histocompatibility Complex Class II Antigen Expression in Rejecting Cardiac Allografts
Detection Using In Vivo Imaging With Radiolabeled Monoclonal Antibody
Background Increased expression of major histocompatibility complex class II (MHC-II) antigen occurs during cardiac allograft rejection. We tested the hypotheses that (1) radiolabeled antibody to MHC-II antigen allows detection of cardiac allograft rejection using nuclear imaging techniques and (2) uptake of radiolabeled antibody to MHC-II antigen correlates with severity of rejection.
Methods and Results Thirteen beagles with cervical cardiac allografts were studied for 64±23 days by use of myocardial biopsy and in vivo imaging. Uptake of radiolabeled (131I [n=2], 123I [n=1], or 111In [n=10]) antibody to MHC-II increased over baseline in 7 animals that developed histological evidence of progressively worsening allograft rejection (group A), from 72.2±46.1 to 176.8±102.0 counts/pixel/mCi (P<.009). In 4 beagles without progressively worsening allograft rejection (group B), uptake was unchanged during follow-up (74.4±43.8 and 60.2±37.4 counts/pixel/mCi; P=NS). In animals studied with 111In-labeled antibody, uptake increased from 102.9±23.1 at baseline to 233.2±82.7 counts/pixel/mCi at follow-up in group A animals (P=.036), with no significant change in group B (91.1±34.9 and 75.9±24.9 counts/pixel/mCi; P=NS). Uptake of 111In-labeled antibody was 107.5±35.7, 135.9±70.8, and 307.8±90.1 counts/pixel/mCi in biopsy samples showing evidence of mild, moderate, and severe rejection, respectively (P=.001). Biopsy samples showing mild, moderate, and intense MHC-II expression antibody uptake had uptakes of 92.6±36.3, 158.5±54.7, and 307.8±90.1 counts/pixel/mCi, respectively (P=.00004).
Conclusions Radiolabeled monoclonal antibodies to MHC-II antigen can detect cardiac allograft rejection in this large mammal model of cardiac allograft transplantation, and this technique may have a potential role in the detection of rejection in patients after cardiac transplantation.
Survival after cardiac transplantation has consistently improved during the last decade, with current 1- and 5-year survival rates of 80% to 90% and 60% to 70%, respectively.1 This has led to the increasing use of cardiac transplantation in the management of patients with end-stage cardiac disease. Despite the availability of improved immunosuppressive agents, one of the most important clinical problems in the posttransplant period is detection and management of rejection. At present, detection and quantification of cardiac allograft rejection require serial endomyocardial biopsies to obtain tissue for histological analysis.2 Typically, endomyocardial biopsies are performed weekly for the first 4 to 8 weeks after transplantation and progressively reduced in frequency over the next few months to 3 to 4 biopsies per year.3 These invasive procedures are costly and prone to sampling errors.4 In addition, histological interpretation is subjective and, to a certain extent, arbitrary.
Rejection is associated with upregulation of the adhesion receptors belonging to the immunoglobulin superfamily, including the major histocompatibility complex (MHC) class II antigens.5 6 The increased expression of MHC class II antigens during rejection is seen in the graft endothelium and on infiltrating mononuclear cells. Isobe et al,7 using a murine model of cardiac transplantation, were the first to show that it was feasible to detect cardiac rejection with 111In-labeled monoclonal antibodies. One limitation of this model is that unlike larger mammals, including humans, rejection in mice is primarily provoked by MHC class I antigens, not class II antigens, and spontaneous long-term tolerance has been reported in murine graft models in the presence of class II mismatching.8 For this reason, we believed it was important to evaluate this technique further using an animal model in which the pathophysiology of graft rejection more closely paralleled that found in patients. In addition, we wanted to apply the technique of immunoscintigraphy to the serial evaluation of cardiac allografts for longer periods of time, allowing evaluation of the cardiac allograft under conditions of varying immunosuppression and degrees of rejection.
Therefore, we used a canine model of cardiac transplantation to test the hypotheses that (1) abnormal expression of MHC class II antigens can be detected noninvasively in rejecting allografts with the use of radiolabeled monoclonal antibodies and (2) detection of abnormal expression of MHC class II antigens correlates with the histological severity of rejection in cardiac allografts.
The surgical technique involved the transplantation of the cardiac allograft into the recipient’s neck.9 The surgical anatomy of this nonworking cardiac allograft is illustrated schematically in Fig 1⇓. Thirteen conditioned beagles (weighing 10 to 15 kg each) received heterotopic cardiac allografts from outbred puppy donors (weighing 2 to 4 kg each). Donor and recipient were premedicated with acepromazine maleate (0.1 to 0.2 mg/kg) and oxymorphone (0.05 mg/kg). Anesthesia was induced with intravenous thiopental sodium (5 to 15 mg/kg) and maintained with isoflurane (0.5% to 2.5%), which was delivered through a volume-controlled ventilator. Muscle paralysis was provided with pancuronium (0.02 to 0.06 mg/kg IV). The donor heart was harvested through a median sternotomy. After preliminary dissection and heparinization (5000 U), the heart was arrested with cold potassium cardioplegic solution. After explantation, the heart was prepared for implantation by ligation of the pulmonary veins and inferior and superior vena cavae. The recipients’ hearts were prepared by isolating and exposing the left common carotid artery and external jugular vein and by creating a small cervical skin pocket. After this, a median sternotomy and preliminary dissection were performed, followed by intravenous heparinization. The graft was then implanted into the recipient’s neck, with the donor ascending aorta anastomosed end to side with the recipient’s common carotid artery and the donor pulmonary artery anastomosed end to end with the recipient’s external jugular vein. After reperfusion, hearts were defibrillated and implanted into the previously prepared cervical pocket. Cyclosporin A was given starting 2 days before transplantation at a dose of 20 mg/kg and was resumed on the first postoperative day. Prednisolone (2 mg/kg) was also given 2 days before transplantation and restarted on the following day. Varying degrees of graft rejection were obtained by gradual reductions in the prednisolone and cyclosporine dosages using serum cyclosporine levels and clinical indicators. All animals used in the study received humane care in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences (NIH publication No. 85-23, revised 1985).
Rejection was evaluated by serial percutaneous transmyocardial biopsies. Two to three tissue samples were obtained at each time point. Some were fixed in formalin and stained with hematoxylin and eosin for microscopic evaluation and the remainder were snap-frozen after embedding in OCT compound (Ames). After the final image acquisition, the animals were killed and their hearts excised and processed for subsequent histological analysis. Representative sections of the right and left ventricles were selected and stained with hematoxylin and eosin. Tissue was also frozen and processed for immunohistochemistry. The degree of cellular rejection was evaluated according to previously published criteria using both the Texas Heart Institute (THI) and the International Society of Heart Transplantation (ISHT) gradations.10 11 The scoring of each section was performed in a blinded manner by an experienced cardiac pathologist. Sections for immunohistochemistry were cut at 5-mm intervals at −20°C in a cryostat, transferred to slides precoated with poly-d-lysine (Sigma Chemical Co), and fixed in cold acetone for 10 minutes. These slides were washed twice with PBS, pH 7.4, and then immersed in a 0.1% sodium azide with 1% hydrogen peroxide in water for 15 minutes at room temperature to neutralize endogenous peroxidase activity. The slides were then rinsed twice in PBS and flooded with RPMI 1640 media containing 5% horse serum. Sections were incubated for 60 minutes with primary mouse antibody against MHC class II antigen in a moist chamber followed by a predetermined optimum concentration of biotinylated anti-mouse Ig, followed by horseradish peroxidase avidin-biotin complexes (Vector Labs). The sites of antigen/antibody reaction were detected with the use of 3,3 diaminobenzidine (Sigma). The slides were counterstained with Mayer’s hematoxylin solution, coverslips were put on, and the slides were then examined under the microscope. Negative control slides were treated in an identical way except the primary monoclonal antibody was omitted. Frozen sections of dog lymph node served as a positive control. Findings were graded in a similar fashion to the THI grading scale: 0 (negative) and from 1 to 10 (positive). Scoring of the sections was performed by the same cardiac pathologists who scored the light microscopic slides.
A monoclonal antibody (TH14B; VMRD, Inc) to canine MHC class II antigens was used. This monomorphic IgG2a monoclonal antibody, which was obtained from mouse ascites fluid and purified by centrifugation and filtration through a 0.22-mmol/L filter, was originally developed by Davis et al.12 13 14 Radioiodination of intact monoclonal antibody (0.5 to 1.0 mg) was performed with 5 to 10 mCi of either 131I (n=2; DuPont) or 123I (n=1; Nordion) by incubation in a sterile vial coated with 1 mg Iodo-gen (Pierce Chemical Co).15 Alternatively, when 111In chloride (n=10) was used, a DTPA conjugation technique was used.16 Quality control assays included tests for protein-bound iodine (trichloroacetic acid precipitation), labeling efficiency by instant thin-layer chromatography in saline. The final product had a labeling efficiency and protein-bound iodine that were both >90%.
Beagles were imaged at baseline during full immunosuppression, ≈2 weeks after transplantation, and at least once thereafter at the end of the study period. On the day of percutaneous biopsy, the animals were injected intravenously with radiolabeled monoclonal antibody and imaged 48 hours after administration of 123I- and 111In-labeled antibodies and 72 hours after 131I-labeled antibody. Planar images were acquired in the anterior and left lateral projections. Each image was acquired for 900 seconds by use of a commercially available single-headed, rotating γ-camera equipped with the appropriate collimator with 20% windows centered on the relevant photo peaks of the isotope used. After acquisition of the immunoscintigraphic images, 4 to 6 mCi of 99mTc-labeled sestamibi was injected to verify the location of native and donor hearts. Regions of interest were drawn in the left lateral immunoscintigraphic images around the donor and native hearts. Background regions of interest were generated for both the donor and native hearts. For the donor heart, the background region of interest (2 to 3 pixels wide) was drawn adjacent to the anterior wall and apex. For the native heart, the background region of interest was drawn in the adjacent lung. Activity in the donor and native hearts was expressed as counts per pixel corrected for injected dose, background activity, and decay.
Results are expressed as mean±SD. Paired comparisons were made by use of paired Student’s t tests. ANOVA was used for multiple comparisons between groups. A value of P<.05 was considered statistically significant.
A total of 13 animals were studied. Two animals were studied by use of 131I-labeled antibody and 1 was studied with 123I-labeled antibody. The remainder were studied by use of 111In-labeled antibody because of the suboptimal quality of images obtained using radioiodinated antibodies. Two of 10 animals studied by use of 111In labeling were excluded from analysis. One animal rapidly developed clinical evidence of rejection and died suddenly before follow-up imaging could be performed. Another developed an area of regional akinesia, was found to have evidence of regional transmural infarction at autopsy, and was also excluded from analysis. In a third animal, intense activity was noted on baseline imaging without clinical or histological evidence of rejection. This activity was not present 2 weeks later. The reason for the transient increase in activity is unclear, but it may have resulted from pericardial reaction, and therefore these data were not included in the subsequent analysis.
In the remaining animals (both radioiodinated and 111In-labeled antibody), seven developed progressively worsening cardiac allograft rejection (group A); the remaining four animals showed no or minimal progression of allograft rejection (Table⇓). The animals in group A showed a significant increase in the amount of uptake of radiolabeled antibody to MHC class II antigen, increasing from 72.2±46.1 at baseline to 176.8±102.0 counts/pixel/mCi at the end of the study period (P<.009). In contrast, the animals in group B showed no significant change in uptake (74.4±43.8 and 60.2±37.4 counts/pixel/mCi, respectively; P=NS).
Fig 2⇓ summarizes the data obtained using 111In-labeled monoclonal antibody to MHC class II antigen. In group A animals (n=5), the uptake increased from 102.9±23.1 at baseline to 233.2±82.7 counts/pixel/mCi at the time of follow-up (P=.036). In group B animals (n=3), there was no significant change in the uptake of 111In-labeled antibody (91.1±34.9 and 75.9±24.9 counts/pixel/mCi, respectively; P=NS). Image quality was generally good, with rejecting allografts being easily visualized when 111In-labeled antibody was used (mean donor target-to-background ratio of 5.3±5.1). The amount of activity in the region of the native heart was minimal and often equal to or less than background activity in the adjacent lung (Table⇑; Figs 3⇓ and 4⇓).
Figs 3⇑ and 4⇑ show representative 111In-labeled images. Fig 3⇑ shows uptake of labeled antibody to MHC class II antigens at two different time intervals. The image on the left, taken 21 days after transplantation and having a biopsy sample showing THI grade 3 rejection, shows mild diffuse uptake. The right-hand image was obtained 34 days after transplantation; it reveals more intense uptake in the graft, with a corresponding biopsy sample showing THI grade 7 changes consistent with moderate rejection. Fig 4⇑ shows images with corresponding biopsy samples from an animal with moderate rejection (THI grade 7). There is diffuse uptake of radiolabeled antibody in the donor heart, with minimal residual blood pool activity seen in the region of the native heart. The histological sections show marked cellular infiltration with interstitial mononuclear cells and evidence of myocyte degeneration, and the immunohistochemically processed section shows fairly intense expression of MHC class II antigen occurring in the presence of moderate rejection.
Fig 5⇓ summarizes the relationship between uptake of 111In-labeled monoclonal uptake and histological severity of rejection from the corresponding myocardial biopsy samples. Using the THI grading scale, we graded biopsy samples as showing mild (grades 1 to 3), moderate (grades 4 to 8), and severe (grades 9 and 10) degrees of rejection and compared them with uptake of 111In-labeled antibody (Fig 5A⇓). Uptake was 107.5±35.7, 135.9±70.8, and 307.8±90.1 counts/pixel/mCi in biopsy samples showing evidence of mild, moderate, and severe rejection, respectively (P=.001). Uptake of 111In-labeled antibody was also evaluated using the ISHT histological gradation scale (Fig 5B⇓). For ISHT grades 1A, 1B, 2, 3A, 3B, and 4, uptake was 112.4±54.0, 98.8±18.9, 115.5±96.1, 134.3±56.8, 179.0±23.2, and 307.8±90.1 counts/pixel/mCi, respectively (P=.015). The relationship between uptake of radiolabeled monoclonal antibody and intensity of the immunohistological expression of MHC class II antigens was closer. Uptake in biopsy samples showing mild, moderate, and intense MHC class II antigen expression was 92.6±36.3, 158.5±54.7, and 307.8±90.1, respectively (P=.00004; Fig 5C⇓).
Noninvasive Detection of Cardiac Rejection
Many techniques have been evaluated for noninvasively detecting rejection after cardiac transplantation. Currently, no noninvasive technique has found widespread acceptance for use in the detection of rejection in patients after cardiac transplantation. Several studies have shown that moderate to severe rejection after cardiac allograft transplantation can be detected with the use of a radiolabeled monoclonal antibody to myosin both in experimental models17 18 19 20 and in patients.21 22 These studies have shown a variable correlation between scintigraphic findings using 111In-labeled anti-myosin antibody and the histological extent of myocyte necrosis after transplantation. In patients with cardiac allografts, radiolabeled anti-myosin antibody uptake is reported to have a reasonable sensitivity (range, 80% to 95%), with a somewhat more questionable specificity (range, 33% to 80%).
MHC Class II Antigen Expression in Cardiac Rejection
The immunoglobulin superfamily of adhesion receptors, including the MHC class II antigens, plays an important role in the pathophysiology of allograft rejection.23 24 Immunocytochemical techniques have allowed identification of the different types of mononuclear cells associated with rejection and their interaction with antigens expressed on the graft. These techniques have revealed intense expression of MHC class II antigens in endomyocardial biopsy samples from patients with cardiac allograft rejection.5 6
Isobe et al,7 using a murine model of cardiac transplantation, were the first to show that it was possible to detect cardiac rejection using radiolabeled antibodies to murine equivalent MHC class II antigens. However, unlike in larger mammals, rejection in mice is primarily provoked by MHC class I antigens, not class II antigens, and spontaneous long-term tolerance has been reported using a murine vascularized graft model in the presence of class II mismatching.8 In the present study, we demonstrated that the upregulation of MHC class II antigen expression that occurs during rejection can be imaged successfully in this larger mammal in which MHC class II antigens are primarily responsible for rejection, as occurs in humans. In addition, by studying animals for a period of months, we were able to show that detection of increasing uptake of radiolabeled antibody to MHC class II antigen by the cardiac allograft on follow-up imaging was associated with development of worsening rejection of the allograft.
The uptake of radiolabeled antibody to class II antigen in the cardiac allograft correlated with the histological severity of rejection using either the THI or ISHT gradations (Fig 5⇑). Despite highly statistically significant differences in antibody uptake between these groups, there was significant overlap, particularly in the milder degrees of rejection. Discrimination between groups appeared to be better in the presence of more severe rejection; however, the number of data points within groups was often small. As might have been anticipated, the relationship between the immunoscintigraphic and immunohistological findings was closer than with the standard histological gradation. It has been shown previously that the upregulation of MHC antigen typically begins before histological evidence of rejection becomes obvious.5 Whether this has implications for clinical imaging is unclear; however, it may allow earlier detection of cardiac rejection before significant myocyte necrosis occurs.
The results of this study are encouraging and suggest that this technique may have potential in the clinical arena. However, there are several limitations to the present study. First, the main radioisotope used in this study was 111In chloride. The optimal time for imaging appeared to be 48 hours after administration of the radiolabeled monoclonal antibody. At this time, the blood pool and other background activities were minimal and allowed adequate visualization of the graft. However, in the clinical setting, this 48-hour wait may be too long and may delay institution of therapy if significant allograft rejection is present. Second, in this animal model, the cardiac allograft was located in the cervical region, which is advantageous for imaging. Imaging a cardiac allograft located within the thorax would be more problematic, particularly because of background activity from the liver. However, many of these limitations could be overcome by the use of a 99mTc-labeled peptide fragment rather than an intact antibody. The use of a peptide fragment would result in much more rapid blood clearance, allowing imaging within several hours rather than a few days, and the use of 99mTc would provide better imaging characteristics than 111In. Third, 1 of the animals studied demonstrated marked uptake of labeled antibody in the absence of clinical or histological rejection early after rejection while receiving full immunosuppression. This was possibly due to pericardial reaction after surgery. Although it only occurred in 1 of 11 animals, it does raise concerns about false-positives in the presence of a local inflammatory reaction when this technique is used. Finally, as eluded to earlier, due to the expense of this surgical preparation, the number of animals studied was small, which limited the statistical power of the study, especially with regard to comparisons between individual groups.
These data indicate that radiolabeled monoclonal antibodies to MHC class II antigens can be used to detect and evaluate the severity of cardiac allograft rejection in this large mammal model of cardiac allograft transplantation. Our results extend previous findings in a murine model and suggest that this technique may have a potential role in detection of rejection in patients after cardiac transplantation, and it therefore warrants further evaluation.
This study was supported by a Grant-in-Aid (No. 93G-273) from the American Heart Association, Texas Affiliate, Inc. The authors would like to thank Edward T.H. Yeh, MD, for his help with review and criticism during the revision of this manuscript.
Guest editor for this article was Barry L. Zaret, MD, Yale University School of Medicine, New Haven, Conn.
- Received March 5, 1997.
- Revision received April 24, 1997.
- Accepted May 1, 1997.
- Copyright © 1997 by American Heart Association
Rose AG. Endomyocardial biopsy diagnosis for cardiac rejection. Heart Failure. 1986;2:64-72.
O’Connell JB, Bourge RC, Costanzo-Nordin MR, Driscoll DJ, Morgan JP, Rose EA, Uretsky BF. Cardiac transplantation: recipient selection, donor procurement, and medical follow-up. Circulation. 1992;86:1061-1079.
Isobe M, Narula J, Southern JF, Strauss HW, Khaw BA, Haber E. Imaging the rejecting heart: in vivo detection of major histocompatibility complex class II antigen induction. Circulation. 1992;85:738-746.
Carrel A, Guthrie CC. The transplantation of veins and organs. Am J Med. 1905;10:1101-1102.
Billingham ME, Cary NR, Hammond ME, Kemnitz J, Marboe C, McCallister HA, Snovar DC, Winters GL, Zerbe A. A working formulation for the standardization of nomenclature in the diagnosis of heart and lung rejection: heart rejection study group. J Heart Transplant. 1990;9:5887-5893.
Davis WC, Marusic S, Lewin HA, Splitter GA, Perryman LE, McGuire TC, Gorham JR. The development and analysis of species specific and cross reactive monoclonal antibodies to leukocyte differentiation antigens and antigens of the major histocompatibility complex for use in the study of the immune system in cattle and other species. Vet Immunol Immunopathol. 1987;15:337-376.
Ababou A, Goyeneche J, Davis WC, Levy D. Evidence for the expression of three different BoLA-class II molecules on the bovine BL-3 cell line: determination of a non-DR non-DQ gene product. J Leukoc Biol. 1994;56:182-186.
Ababou A, Davis WC, Levy D. The DA6-147 monoclonal antibody raised against the HLA-DRa chain identifies a cryptic epitope on the BoLA-DRa chain. Ann Rech Vet. 1993;24:402-407.
Hall TS, Baumgarter WA, Borkon AM, LaFrance ND, Triall TA, Norris S, Hutchins GM, Brawn J, Reitz BA. Diagnosis of acute cardiac rejection with antimyosin monoclonal antibody, phosphorous nuclear magnetic resonance imaging, two-dimensional echocardiography, and endocardial biopsy. J Heart Transplant. 1986;5:419-424.
Isobe M, Haber E, Khaw BA. Early detection of rejection and assessment of cyclosporine therapy in 111In antimyosin imaging in mouse heart allografts. Circulation. 1991;84:1246-1255.
Takeda K, Ueda K, Scheffel U, Ravert H, LaFrance D, Baumgartner WA, Reaitz BA, Herslowitz A, Wagner HN. Indium-111 myosin-specific antibodies and technetium-99m pyrophosphate in the detection of acute cardiac rejection of transplanted hearts: studies in a heterotopic heart model. Eur J Nucl Med. 1991;18:461-466.
Frist W, Yasuda T, Segall G, Khaw BA, Strauss HW, Gold H, Stinson E, Oyer P, Baldwin J, Billingham M, McDougall IR, Haber E. Noninvasive detection of human cardiac transplant rejection with indium-111 antimyosin (Fab) imaging. Circulation. 1987;76(suppl V):V-81-V-85.
Ballester M, Obrador D, Carrio I, Moya C, Auge JM, Bordes R, Marti V, Bosch I, Berna-Roqueta L, Estorch M, Pons-Llado G, Camara ML, Padro JM, Aris A, Caralps-Riera JM. Early postoperative reduction of monoclonal antimyosin antibody uptake is associated with absent rejection-related complications after heart transplantation. Circulation. 1992;85:61-68.