Cardiovasc Pathol Vol. 2, No. 1 January-March 1993:21-34

21

Vascular Rejection in Cardiac Transplantation: Histologic, Immunopathologic, and Uitrastructural Features Elizabeth H. H a m m o n d , M D , Janet K. H a n s e n , Louise S. Spencer, A n n Jensen, D o n n a Riddell, Catherine M. Craven, M D , and R o b e r t L. Yowell, M D , P h D From the Departments of Pathology, LDS Hospital Primary Children's Medical Center, and

University of Utah School of Medicine, Salt Lake City, Utah

+4Although the majority of rejection found in cardiac transplant biopsies is cellular in type, a variety of vascular alterations occur in cardiac biopsies, constituting different forms of rejection that can be recognized using light microscopic and immunopathologic criteria. In this report, pathologic aspects of the vascular alterations associated with vascular and mixed rejection of cardiac allografts are described in detail. Methods and controls used in this report are identical to those previously reported. The histologic, immunopathologic, and ultrastructural findings associated with vascular rejection and other vascular processes in cardiac allografts are discussed. The relationship of these findings to chronic allograft rejection and potential pathogenetic mechanisms of these vascular changes are also detailed.

During the last few years, prophylactic immunosuppression has been increasingly used in cardiac transplantation. These protocols have included the use of cyclosporine A and antibodies directed against T cells (antilymphocyte globulin and antithymocyte globulin), as well as antibodies directed against the CD3 antigen (OKT3) (1-3). These protocols have led to an increase in cardiac transplant survival, in the experience of some centers, and have produced a spectrum of pathologic manifestations and complications different from that encountered in patients treated with standard triple therapy (3-5). Patients treated on these new regimens show a delayed onset of rejection episodes and, possibly, a decrease in the number of such episodes (1-3). It is disputed whether or not infectious complications are decreased using this therapy (3,6). Regardless of the mode of initial immunosuppression, the long-term outcome of allograft cardiac transplantation continues to be haunted by the problem of chronic rejection manifested as allograft coronary arManuscript received September 10, 1992; accepted November 20, 1992. Address for reprints: Elizabeth H. Hammond, MD, Department of Pathology, LDS Hospital, 8th Avenue and C Street, Salt Lake City, UT 84143.

©1993 by ElsevierSciencePublishingCo., Inc.

tery disease (7-15). Such coronary artery disease usually tends to be diffuse and concentric, leading to generalized myocardial ischemia and allograft loss. The incidence of this complication is reported to be anywhere between 33% and 60% of all recipients at five years posttransplant (9). The incidence of allograft coronary artery disease has not been significantly altered by the use of cyclosporine or newer immunosuppressive protocols, such as immunoprophylaxis (10,11,15,16). Allograft coronary artery disease has been shown to be more common in patients with the pathologic diagnosis of vascular rejection (10,11). Since 1985, we have studied endomyocardial biopsies from cardiac transplant recipients by light microscopy, immunofluorescence, and electron microscopy. As of March 1992 we had reviewed approximately 12,900 endomyocardial biopsies by light microscopy, 5,100 endomyocardial biopsies by immunofluorescence, and approximately 350 endomyocardial biopsies by electron microscopy. This report describes and classifies the vascular changes seen in these patients treated with immunoprophylactic protocols, including antilymphocyte globulin (ALG) and OKT3. Since 1985, the Utah Cardiac Transplant Program has treated 25 patients with ALG, 1054-8807/93/$6.00

22

HAMMOND ET AL. VASCULARRFAECTIONIN CARDIACTRANSPLANTS

Table

Cardiovasc Pathol Vol. 2 No. 1 January-March 1993:21-34

1. C o m p a r i s o n of U C T P and I S H L T R e j e c t i o n Criteria UCTP

ISHLT

Cellular rejection: Variations of ISHLT and UCTP grades Focal mild rejection: C2.5 ISHLT 1A or 2, depending on damage Mild rejection: C3 ISHLT 1B or 2, depending on presence of myocyte damage and infiltrate Moderate rejection: C4 ISHLT 3A or 3B, depending on extent of myocyte damage and infiltrate Severe rejection: C5 ISHLT 4, identical criteria Vascular rejection: All considered ISHLT 0, except as noted Mild vascular rejection: C1, V3 LM: Endothelial activation, edema, hemorrhage IF: Ig and C in vessels with or without interstitial fibrin Moderate vascular rejection: CI, V4 LM: Endothelial activation, edema, hemorrhage, ± vasculitis IF: Ig and C in vessels, interstitial fibrin prominent Severe vascular rejection: C5V5 LM: Like ISHLT4 ISHLT4 IF: Ig and C in vessels, Ig, C, fibrir in interstitium Mixed rejection: All considered as corresponding ISHLT cellular grade. Simultaneous occurrence of various UCTP C and V grades as defined above. Grades:

C3, C3, C4, C4,

V3 V4 V3 V4

u c r P = Utah Cardiac Transplant Program; ISHLT = International Society for Heart/Lung Transplantation; LM = light microscopy; IF = immunofluorescence; Ig = immunoglobulin; C = complement.

25 patients with 10-day OKT3 therapy, 261 patients with 14-day OKT3 therapy, and 12 patients with 21day OKT3 therapy. The morphologic observations in this report detail the types of vascular changes that can be encountered in endomyocardial biopsies, explanted hearts, and autopsies from such patients. The classification scheme used for cellular rejection in such patients has previously been reported (4). Changes strictly related to cellular rejection will not be discussed, given that they are the subject of numerous published reports (17-24).

Methods

Patient Population and Types of Immunosuppression From the outset of the program, all patients at LDS Hospital (n = 134) were prospectively followed by light microscopy and immunofluorescence. All patients at the Veterans Administration Medical Center (VAMC, n = 140) were followed routinely by light microscopy and occasional electron microscopy (first posttransplant biopsy and at least one other). From July 1988 to January 1992, all patients (n = 268) in the Utah Cardiac Transplant Program were routinely followed for four weeks by light microscopy and immunofluorescence based upon the findings that have been previously published (4,5). Immunofluorescence monitoring has been routinely done in our program for four weeks posttransplant and is continued only in those pa-

tients who show some propensity for vascular changes, as will be described below. By January 1992, 268 patients had been routinely followed by light microscopy and immunofluorescence. Electron microscopy was otherwise done sporadically and on selective biopsies where it was deemed appropriate.

Histopathologic Evaluation Specimen preparation. Light microscopic examination of consecutive serial endomyocardial biopsies was routinely performed according to the standard protocol used by the Utah Cardiac Transplant Program (4,5). One biopsy fragment was frozen in OCT freezing compound for immunofluorescent studies. The remaining fragments were immediately immersed in phosphatebuffered 10% formalin and rapidly processed for histologic evaluation by routine methods. Hematoxylin-eosin-stained sections and Masson's trichromestained sections were evaluated on each patient. At least five sections per piece were routinely examined on each of two slides. Histologie evaluation. Biopsies from all patients were examined using the standard Utah Cardiac Transplant Pathologic Criteria. These criteria are a modification of the standard Billingham Criteria and have been previously reported (4,5). After July 1990, all biopsies also received an International Society of Heart and Lung Transplantation (ISHLT) grade, in order to provide standard nomenclature (20). The relationship

Cardiovasc Pathol Vol. 2, No. 1 January-March 1993:21-34

H A M M O N D ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

23

Table 2. Antibodies Used for Cardiac Transplant Immunofluorescence Assay Antibody

Supplier

Usual Dilution

IgG-FITC IgM-FITC C3c,d-FITC

Protos Protos Dako

1.5 1:10 l:2,combined

Clq-FITC Fibrinogen-FITC albumin-FITC HLA-DR-PE Mouse IgG-FITC

Dako Dako Dako Becton-Dickinson Protos

straight 1:10 1:10 1:5 1:10

Catalog No. 311 313 c: F201 d: F323 F254 F11 l F117 7367 346

Protos: 1485 Bayshore Blvd., Suite 388, San Francisco, CA 94124. Dako: 6392 Via Real, Carpinteria, CA 93013. Becton-Dickinson: 2375 Garcia Ave., Mountain View, CA 94043.

of these grading schemes appears in Table 1. Because we have previously recognized a spectrum of vascular changes that we designate vascular rejection, we have devised a grading scheme for vascular rejection (5). Table 1 also shows the criteria for these grades and their corresponding ISHLT counterparts. Immunolluorescent method. Allimmunofluorescent examinations were carried out on 4-~tm thick frozen sections without prior fixation and subjected to 30-minute incubations using fluorescein isothiocyanate-labeled antibodies directed against IgG, IgM, fibrinogen (Fab' fragments), C3, Clq, albumin, and mouse immunoglobulin (absorbed against human IgG) by direct immunofluorescence. Monoclonal antihuman H L A - D R and monoclonal antihuman MHC Class I were detected by indirect immunofluorescence, until phycoerythrinlabeled H L A - D R became available, after which this antibody was sought directly. We graded vascular and interstitial immunofluorescent changes separately on each biopsy (scale of 0 to 3 +) to facilitate comparisons and to evaluate each patient over time. The details of this

method have been reported (4,5). Suppliers of antibodies and the dilutions used in this examination are shown in Table 2. Immnnofluorescence artifacts. As with all methods, certain precautions must be observed in the interpretation of the findings. Because IgG and albumin are more ubiquitous in tissue and generally show brighter staining, the threshold for positivity must be higher than for other reagents, such as complement components and HLADR. Furthermore, the pattern of staining, critical for interpretation, is more difficult to appreciate in biopsies showing both interstitial and vascular accumulation of reagents. Careful examination of the HLA-DR-stained slide, to detect the distribution of the vessels, will aid in this distinction. Because complement is almost never found in the interstitium (except in severe rejection), it will be either negative or in a vascular distribution that will aid in deciding whether vascular immunoglobulin (either IgG or IgM) is present. The presence of large amounts of albumin in the interstitium indicates that significant vascular leakage is present and confounds the evaluation of vascular complexes (Fig. 1). Normally albumin is confined to vascular spaces. Other common artifacts are summarized in Table 3. Ultrastructural processing. Tissue for electron microscopy was visually chosen from the fragments submitted and fixed in Karnovsky's fixative overnight (25). Routine processing, including osmication, was performed. Tissues were dehydrated through graded acetones and embedded in EMBED (25,26). Appropriate areas were chosen from toluidine blue-stained 1-~tmthick sections and sectioned at 700.& stained with uranyl acetate and lead citrate, and examined in a JEOL 100 SX transmission microscope (26). Results

Figure 1. In this immunofluorescence photograph, prominent leakage of albumin is seen in the interstitium, making positive identification of vascular profiles impossible. (Magnification x250.)

Histologic Evaluation All patients with vascular rejection diagnosed by light microscopy and immunofluorescence before 1.988

24

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

Table 3. Common Immunofluorescence Artifacts

Cardiovasc Pathol Vol. 2 No. I January-March 1993:21-34

show capillaries with flat endothelial cells and no interstitial edema or hemorrhage. Arterioles are rarely observed and have flattened endothelium (Fig. 2). Such biopsies are very common in patients at long intervals posttransplant but are exceedingly rare in the first month posttransplant, during which endothelial cell activation is commonly demonstrated. Biopsies taken in the early posttransplant period,

when ischemic changes may be present, often show areas of focal myocardial necrosis with hypereosinophilia of myocytes, increased focal contraction bands, and/or rare infiltration of polymorphonuclear leukocytes. These changes are also commonly associated with catecholamine effect or postsurgical hypoxia; thus they are regularly demonstrated and ignored in the first few biopsies posttransplant. At later intervals posttransplant, they may be a histologic sign of vascular rejection, as detailed below. Biopsies suggestive of vascular rejection (V2). The majority of biopsies in patients treated with immunoprophylaxis with OKT3 or A L G show a variety of endothelial cell changes. The most common change recognized in such biopsies is the presence of marked swelling of endothelial cells with increased nuclear size, hyperchromasia, and increase in numbers of nuclei per vessel examined (Fig. 3). Furthermore, the cytoplasm of such cells may be so vacuolated and swollen that they appear to occlude the lumen by light microscopy. In biopsies taken after the first week posttransplant, endothelial cell activation often persists, causing an obvious prominence of capillaries and venules even at a low magnification (Fig. 3). More than one week posttransplant, such changes are considered to represent endothelial activation and are called grade 2 vascular change (V2) in our nomenclature; vascular rejection cannot be excluded, but the changes are not diagnostic, considering that other processes can produce a similar light microscopic appearance. At times other than the first six weeks posttransplant, these changes are often but not always associated with vascular rejection diagnosed on the basis of the immunofluorescent findings. The prominence of these endothelial cell changes in our program is striking when compared with those

Figure 2. This endomyocardial biopsy shows no evidence of

Figure 3. Vessels in this biopsy show diffuse endothelial ac-

vascular or cellular rejection. The vascular endothelial cells are flat and all vessels are intact. (Magnification xl00.)

tivation; the vessels have a distinctive rope-like character. Compare with Figure 2. (Magnification ×250.)

Reagent

Artifactual Staining

IgG, IgM, C3

Neutrophil and eosinophil granules; macrophages; necrotic cells such as myocytes Blood clots on the endocardial surface, recognized by the uniform staining with fibrin Quilty endocardial infiltrates No staining visible with any reagent; usually this is attributable to inadvertent fixation of the specimen before freezing; sample cannot be interpreted Nonspecific staining of myocytes if dilution too low: prominent edge staining Pale, smooth staining of endothelium of small vessels; usually found in all heart biopsies

All All All

Albumin Clq

were reviewed and assigned a vascular grade. Since 1988, all patients have prospectively received a vascular grade. The pathologic observations reported here are a result of this prospective and retrospective analysis.

Histologic Features Associated with Vascular Changes and/or Rejection Biopsies without evidence of significant vascular abnormality (V1). Biopsies without vascular abnormality

Cardiovasc Pathol Vol. 2, No. 1 January-March 1993:21-34

Figure 4. Diffuse adherence of lymphocytes to the vessel wall gives a striking appearance diagnostic of vascular rejection. The grade must be determined by the amount of immunofluorescentaccumulation of immune complexes. (Magnification X400.) from programs using standard triple therapy (E.H., unpublished observations). Another change suggestive of vascular rejection by light microscopy is the presence of coagulative myocyte necrosis on a biopsy without histologic evidence of rejection and after the first six weeks posttransplant. Such necrosis is often seen in the subendocardium and may persist for several weeks. Later this area may be associated with histologic changes of repair that can, at times, resemble rejection, although the predominant infiltrating cells are macrophages. We have found such necrosis to be commonly associated with vascular rejection (detected on earlier or later biopsies by immunofluorescent demonstration of vascular immune complexes). Therefore, we believe that such necrosis is an indication of larger vessel damage with associated myocardial hypoxia. Mild vascular rejection (V3). The light microscopic features of mild vascular rejection include endothelial cell activation, as described above. Biopsies can be histologically identical to those without vascular immune complexes by immunofluorescence (V2). In addition, arterioles, when present in the sample, commonly show marked vacuolization of cell cytoplasm with occlusion of the lumina and increased PAS positive staining of basement membranes. Occasional infiltration of venules and capillaries by lymphocytes or the adherence of lymphocytes to the endothelial surface of such vessels may be seen, giving the microvasculature a ropelike character (Fig. 4). This appearance of vessels is more generalized and striking in moderate vascular rejection biopsies. Interstitial edema is common and appears as a basophilic, occasionally granular, irregular material around vessels. This edema can be discriminated from interstitial fibrous tissue on the basis of the

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

25

Figure 5. In this highly magnified view of the interstitium of an edematous biopsy, the fibrillar character of the edematous area is evident. Biopsy with vascular rejection. (Magnification ×400.) trichrome stain: trichrome staining discloses that the apparently edematous areas contain no collagen (Fig.

5). Moderate vascular rejection IV4). In the presence of ongoing or more severe vascular changes, capillaries, venules, and/or arterioles commonly display evidence of inflammation in the form of adherence of lymphocytes, macrophages, or polymorphonuclear leukocytes to the endothelial lumina and extravasation of such cells through the walls. Thus the rope-like character of vessels is more uniformly seen in moderate vascular rejection than in mild vascular rejection (Fig. 4). Pyknotic endothelial cells with occasional sloughing are common. Vessels are associated with appreciable interstitial edema and hemorrhage (Fig. 6). Nuclear dust and evidence of inflammatory cell necrosis are present in the area surrounding such vessels and seem to be an

Figure 6. A vein shows prominent endothelial activation and infiltration by lymphocytes. Perivascular edema is evident. Vasculitis such as this is light microscopic evidence of moderate vascular rejection. (Magnification x400.)

26

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

Cardiovasc pathol Vol. 2 No. 1 January-March 1993:21-34

Figure 7. In this photomicrograph, a small arteriole is vir-

Figure 9. Myocardium is reacted with antibody to factor

tually occluded by endothelial cell swelling. The surrounding interstitium is filled with particulate debris and fibrillar material suggestive of fibrin deposits. Moderate vascular rejection. (Magnification x400.)

VIII-related antigen using an immunoperoxidase technique with diaminobenzidene as a substrate. A vessel wall (arrows) is frayed and irregular, suggesting vascular damage. Compare with normal vessels at bottom of picture. (Magnification

ominous sign: it is much more commonly seen in patients with unrelenting vascular rejection and in patients with severe hemodynamic compromise in absence of interstitial lymphocytes (Fig. 7). In some cases, frank myocyte necrosis can also be demonstrated (focal areas of increased contraction band formation, myocyte eosinophilia, disruption of intercalated discs, or marked irregularity of myocyte borders). Myocyte necrosis often involves individual cells and appears as hypereosinophilia without associated inflammation. This necrosis, though difficult to document by light microscopy, is suggested by the presence of wavy myocytes, areas of myocytolysis characterized by perinuclear halos or vacuolization of myocytes, and a lack of striations appreciable on Masson's trichrome stain (Fig. 8). Less commonly, we have observed vascular wall necro-

sis, manifested by absence of capillary walls or gaps in walls through which cells or fluid exude(s). Such necrosis is seen after prolonged periods of vascular rejection. This is a difficult feature to quantify, but it can be highlighted by using immunoperoxidase detection of antibodies of endothelial cell specificity, such as antibodies to Factor VIII-related antigen or Ulex lectin (Fig. 9). Fibrinoid necrosis of vessels is not a feature in our experience. Severe vascular rejection (V5). The pathologic appearance of severe vascular rejection is identical to that of severe cellular rejection. In both cases, there is an interstitial inflammation which is pleomorphic and consists of polymorphonuclear leukocytes, eosinophils, and plasma cells and which is associated with severe edema and interstitial hemorrhage. Vessels in these biopsies

Figure 8. An area of myocardium shows some obvious contraction band formation and granularity of myocytes, suggestive of isolated myocyte necrosis (arrows). (Magnification ×400.)

Figure 10. A biopsy from a patient with severe rejectioa

x400.)

shows a pleomorphic infiltrate associated with vasculitis of the microvasculature. (Magnification x400.)

Cardiovasc Pathol Vol. 2, No. I January March 1993:21-34

Figure 11. Focal myocyte dropout is prominent in this biopsy from a patient with allograft coronary artery disease. This appearance on endomyocardial biopsy suggests ischemia. (Magnification ×400.) uniformly show endothelial activation, vasculitis, or mural necrosis (Fig. 10). Allografl coronary artery disease (global myocardial ischemia). In patients with long-standing vascular compromise, a distinctive group of morphologic features is demonstrated on endomyocardial biopsy. Such patients show focal areas of myocyte dropout in which myocytes are replaced by loose connective tissue. Surrounding myocytes are often hypertrophied. In these areas, focal inflammatory cells may remain. Capillaries are difficult to find, and no endothelial activation is present (Fig. 11). The patchy nature of this process and its subendocardial distribution suggest that the myocyte loss is attributable to damage of small arteries and arterioles outside of the field of examination. This observation has been documented on autopsy and explant evaluation of such hearts (11). If these changes are encountered on endomyocardial biopsy in patients with slowly worsening cardiac function, they are an ominous sign. Such patchy myocyte loss is distinctive from that loss usually associated with myocardial infarction, in which large zones of myocyte necrosis are ultimately replaced by dense scarring. When we recognize this ischemic change at intervals after the first weeks posttransplant, we note it in our reports as evidence of ischemia; we have observed these changes in 491 biopsies from 220 patients. When these patients come to allograft replacement and/or autopsy, they usually show uniformly narrowed coronary or epicardial arteries with or without inflammation in such vessels (11) (Figs. 12, 13).

Irnmunofluorescence lmmunofluorescence in the absence of vascular rejection (V1). In cases not displaying vascular rejection,

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

27

Figure 12. This artery from a patient four months posttransplant shows significant subintimal fibrosis. Scattered inflammatory cells are present in the fibrotic region of the wall. The patient had persistent vascular rejection that was not treated prior to death. (Magnification × 100.) weak staining of blood vessels with H L A - D R may be variably demonstrated (0-1 +; Fig. 14). Frequently, the positive staining is limited to arterioles and venules in the biopsy. Myocytes may or may not show major histocompatibility complex (MHC) Class I staining. Little or no immunoglobulin is demonstrated in the interstitium or in vessels; complement is never seen, and fibrin is rarely demonstrated. Albumin is also not present. Immunofluorescence suggestive of vascular rejection (V2). In biopsies demonstrating changes suggestive but not diagnostic of vascular rejection, positive M H C Class II staining is demonstrated (1 +). However, no immune complexes are localized to blood vessels; slight interstitial and/or vascular staining with IgG and/ or fibrinogen (1+) may be seen (Fig. 15). Often, al-

Figure 13. This acute vasculitis of an epicardial artery was found in a patient who died suddenly six months posttransplant. There was a large acute myocardial infarction in the distribution of the left anterior descending coronary artery. (Magnification X400.)

28

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

Cardiovasc Pathol Vol. 2 No. ! January-March 1993:21 34

Figure 14. A section of frozen myocardium from a negative biopsy reacted with phycoerythrin antihuman HLA-DR for 30 minutes. Faint vascular staining is seen in a patchy distribution. Compare with Figure 16. (Magnification x250, exposure time 30 seconds.)

Figure 16. Biopsy from patient with mild vascular rejection reacted with phycoerythrin antihuman HLA-DR. There is uniform, moderate expression by the vascular endothelium. Compare with Figures 14 and 20. (Magnification x250, exposure time 30 seconds.)

bumin is found in a similar distribution, suggesting that the p h e n o m e n o n is related to altered vascular permeability. Rarely, complement components are seen in vascular walls in the absence of immunoglobulin. This finding is never considered diagnostic of vascular rejection and frequently involves vessels in a patchy distribution, only arterioles, or only subendocardial vessels.

and/or IgM plus C3 and/or C l q in a vascular distribution in small quantities (1+: Fig. 17). In such patients, however, no significant fibrin leakage into the interstitial regions is seen. Because immunofluorescence appears to be much more sensitive than are light microscopic changes, in order to make a continued diagnosis of mild vascular rejection, one has to observe the previous and following biopsies on a patient in order to decide whether or not the process appears to be worse or better. Such immunofluorescent changes usually precede light microscopic evidence of vascular rejection (Figs. 4-7).

Immunofluorescent findings in mild vascular rejection (V3). In patients with mild vascular rejection, prominent H L A - D R positivity of blood vessels is always present and involves capillaries as well as arterioles and venules (1-2+); myocytes are most frequently M H C Class I positive (Fig. 16). Furthermore, the most striking and diagnostic feature is the presence of IgG

Figure 15. Section of frozen myocardium from a negative biopsy reacted with fiuoresceinated antihuman IgG for 30 minutes. Patchy trace staining is seen in a vascular distribution. Compare with Figures 17 and 18, similar tissue from patients with mild and moderate vascular rejection. (Magnification x250, exposure time 30 seconds.)

Immunofluorescent findings in moderate vascular rejection (V4). Patients with moderate vascular rejection show large quantities of immune complexes

Figure 17. Frozen myocardium from patient with mild vascular rejection. Reacted with fluoresceinated antihuman igG. There is a punctate pattern of staining indicating vascular localization in round vessel profiles. Complement was present in an identical distribution. Compare with Figures 15 and 17. (Magnification x250, exposure time 30 seconds.)

Cardiovasc Pathol Vol. 2, No. 1 January-March 1993:21-34

Figure 18. Section of myocardium from patient with moderate vascular rejection. Vessels are cut longitudinally and show strong vascular accumulation of fluoresceinated antihuman IgG. Complement was present in an identical distribution. Compare with Figures 15 and 17. (Magnification x250, exposure time 30 seconds.) demonstrated within vascular walls (1-3+). Immunoglobulin demonstrated may be IgG and/or IgM, and C3 and/or C l q are very prominent (Figs. 18, 19). Blood vessels are strongly H L A - D R positive (2+), and myocytes are M H C Class I positive (Fig. 20). Furthermore, vessels may show M H C Class I staining as well as H L A - D R staining. No albumin is seen in a similar distribution. One of the most ominous features of this type of rejection is the presence of interstitial extravasated fibrin (2-3+), which may appear in an interstitial as well as vascular distribution (Fig. 21). When the distribution of the fibrin has increased over the previous biopsy, the grade of the biopsy from a vascular point of view is increased. An increase in the immunofluorescent findings over that of the previous biopsy

HAMMOND ET AL. VASCULAR REJECTION 1N CARDIAC TRANSPLANTS

29

Figure 20. Myocardial frozen section reacted with phycoerythrin antihuman HLA-DR. The section is overexposed in the usual 30 seconds used. Strong staining of all endothelium is seen. Compare with similarly exposed Figures 14 and 16. (Magnification x250.) will result in an upgrading of the biopsy grade to an V4 from a V3.

lmmunofluoreseence findings in severe rejection (V5). In patients with severe rejection, paradoxically, patchy loss of M H C Class II ( H L A - D R ) staining may be identified, and M H C Class I staining may also not be increased. Vessels may or may not show vascular immune complexes; interstitial IgG or IgM associated with C3 and/or C l q ( 1 - 3 + ) is almost uniformly demonstrated, and the amount of interstitial fibrin staining is striking (2-3+). Interstitial albumin can also be seen. This indicates loss of vascular integrity, which is also demonstrated by electron microscopic changes (see below and Fig. 25).

Immunofluorescenee in allograft coronary artery disease. By immunofluorescence, patients with chronic

Figure 21. Section of frozen myocardium reacted with fluFigure 19. Section of myocardium with moderate vascular rejection reacted with fluoresceinated antihuman C3. Longitudinal and cross sections of vessels show obvious accumulation. (Magnification x250, exposure time 30 seconds.)

oresceinated antihuman fibrin. Vascular and interstitial accumulation of fibrin is seen in this patient with moderate vascular rejection. Same patient as in Figure 19. (Magnification x250, exposure time 30 seconds.)

30

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

vasculopathy usually show increased MHC Class II (HLA-DR) staining of blood vessels (2-3+; Fig. 20). The vascular immunofluorescent findings of immune complexes may be decreased from those previously found, or they may be totally absent, and the amount of fibrin staining may also have decreased over previous biopsies. We believe that this decrease in staining (other than H L A - D R ) may reflect chronic damage of the microvasculature. In such patients, only the observation of serial biopsies plus a careful examination of the light microscopic findings will lead to the correct pathologic diagnosis of chronic global ischemia.

Ultrastructural Examination

Ultrastructural findings in the absence of vascular rejection (V1). In patients exhibiting no evidence of rejection, myocytes show no myofilament loss. The only regularly observed changes in such myocytes is the increase in the number of lysosomes present. The interstitium may be increased in area, but no inflammatory cells are observed. Capillary endothelial cells may show slight swelling; they regularly show an increase in the number of lysosomes. Basement membrane reduplication is uncommon, and no subendothelial fibrin is seen (Fig. 22). Ultrastructural findings in biopsies suggestive of vascular rejection (V2). In patients suspected of having vascular rejection, it is common to see marked endothelial cell swelling with occasional occlusion of vascular lumina by endothelial cell cytoplasm. However, in such patients endothelial cell destruction is not observed, subendothelial fibrin is not seen, and reduplication of basal lamina is uncommon. Myocyte changes are similar to those seen in grade 1 biopsies. Interstitial inflammatory cells are not encountered, though pericytes may be swollen and prominent (Fig. 23). Uitrastructural findings in mild vascular rejection (V3). In patients with mild vascular rejection, the histologic features of endothelial activation are prominent: endothelial cells are swollen and often occlude capillary lumina. The cytoplasm of such cells contains increased numbers of organelles, including pinocytotic vesicles, prominent Golgi zones, and increased polyribosomes. Reduplication of vascular basal lamina is common. Vascular contours are often distorted, suggesting recurrent damage and repair. Furthermore, occasional subendothelial fibrin is found; no convincing evidence of discontinuity between endothelial cells is regularly present (Fig. 24). Ultrastructural findings in moderate vascular rejection (V4). Patients with moderate vascular rejection show changes similar to those observed above. Necrotic endothelial cells may be seen as darker,

Cardiovasc Pathol Vol. 2 No. 1 January-March 1993:21-34

shrunken cells occasionally disrupted from their neighbors. Interstitial organelles are often scattered in the edematous interstitium, suggesting that myocyte necrosis may have occurred. In such cases, the interstitium also includes rare lymphocytes, macrophages, and plump fibroblasts, recognized by their obvious ultrastructural differences. Myocytes appear unaffected, except for the occasional one that displays a paucity of myofdaments or increased lysosomes. Ultrastructurally, this process is striking because the vascular changes are so much more severe than those of the surrounding myocytes. It is difficult if not impossible to differentiate the ultrastructural changes of moderate vascular rejection from long-standing mild vascular rejection (Fig. 24). Ultrastructurai changes in severe rejection (V5). Severe rejection has been observed only in the postmortem material of one patient who died early after transplant of unrecognized rejection. Histologically, her infiltrate was typical of severe rejection, but the number of infiltrating cells was small. Immunofluorescence was not carried out on the tissue because of the interval from death to postmortem examination. U1trastructurally, the heart displayed marked edema, a paucity of findings in myocytes other than those typical in postmortem hearts, and a few scattered interstitial inflammatory cells. The ultrastructural appearance of the capillaries was striking. Many endothelial cells appeared necrotic, and endothelial activation of other remaining cells was prominent. Vascular basal lamina reduplication was not seen, probably because of the short interval posttransplant (Fig. 25). Ultrastructural findings in ailograft coronary artery disease with global ischemia. Ultrastructural observations of biopsies from patients with histological evidence of chronic damage of large and small blood vessels have shown changes analogous to those found in ischemic hearts of experimental animals; these hearts show a prominent loss of actin over myosin in myofilament bundles that are intact, giving a coarse appearance to the myofilaments (Fig. 26). In addition, large numbers of myocyte cytoplasmic organelles are scattered around in the interstitium, associated with a patchy and haphazard collection of collagen fibrils. The vessels usually have irregular profiles and may show irregular endothelial cell swelling.

Discussion The changes described in this report suggest that endothelial cell activation, permeability, and subsequent myocyte degeneration are prominent features in patients displaying light microscopic, immunofluorescent, and ultrastructural changes associated with vascular rejection. These morphologic features suggest that the

Cardiovasc Pathol Vol. 2, No. I January-March 1993:21-34

H A M M O N D ET AL. VASCULAR REJECTION 1N C A R D I A C TRANSPLANTS

31

Figure 22. Electron micrograph of myocardium of patient without rejection. Endothelium is flattened and contains lysosomes. (Original plate magnification x 3,000.)

Figure 23. Electron micrograph of myocardium from patient with endothelial activation (V2). Endothelial cell cytoplasm is swollen and partially occludes the lumen of the capillary. (Original plate magnification x3,000.)

endothelial cell plays an important role in this rejection process. In-depth investigations of endothelial cell biology have shown that rather than being nonspecific targets of injury, endothelial cells are capable of making diverse cytokines that can modulate the biologic behavior of cells in the myocardial tissue. These cytokines are produced in inflammation, ischemia, and many other circumstances commonly operative in transplantation, such as infection and lymphocyte activation (27-34). There is evidence that endothelial activation can be the result of immunosuppression or immunoprophylaxis utilizing monoclonal anti-T-cell antibodies, which can generate T-cell activation as a result of interaction of the antibody and the CD3 or T-cell receptor antigen on the lymphocyte surface (27). We have observed that the group of patients with the morphologic pattern of vascular rejection includes

patients sensitized to OKT3 while undergoing immunoprophylaxis. Although the pattern of rejection is similar, the mechanism of vascular alteration may be different. Virtually all patients with sensitization to OKT3 immunoprophylaxis show production of human antimouse antibody and exhibit morphologic changes in their endomyocardial biopsies indistinguishable from those seen in vascular rejection (5,10,11,35). Thus, at least some of these vascular rejection manifestations are related to humoral immune responses, with probable altered B-lymphocyte immune regulation or polyclonal B-lymphocytic activation. These morphologic changes are very similar to those described in reports detailing humorally mediated vascular rejection in renal transplant recipients, patients with serum sickness, and experimental animals with leukocytoclastic vasculitis or the Arthus phenomenon (36-44). The pathogenetic mechanisms responsible for these vascular

32

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

Cardiovasc Pathol Vol. 2 No. 1 January-March 1993:21 34

Figure 24. Electron micrograph of myocardium from patient with mild vascular rejection. Endothelial activation is prominent and vascular profile is irregular. Basal lamina is reduplicated. (Original plate magnification x8,000.)

Figure 25. Electron micrograph of interstitium of patient dying of vascular rejection in the early posttransplant period. Myocardium shows degenerative changes, and vessel endothelial lining is disrupted (arrow). (Original plate magnification

| ,9

x2,000.)

inflammatory processes (in which an antigen-antibody complex process is definitely implicated) involve complement activation, cytokine release and chemotaxis, and activation of neutrophils (36-44). The pathogenesis of the vasculitis and deposition of immune complexes demonstrated in vascular rejection and allograft coronary artery disease is unclear. Although it could result from humoral immune mechanisms to defined or undefined transplantation antigens, the possibility that this picture is related to delayed hypersensitivity and enhanced vascular permeability related to cytokine release cannot be differentiated by the present or previous studies (10,11,16,29,31,45). We have consistently seen up-regulation of H L A - D R on the large and small vessels of the allograft which, at least in experimental situations, is produced exclusively by interaction of endothelial cells with interferon gamma (2,3). This finding, as well as the

prominent fibrin deposition and endothelial activation, suggests that delayed-type hypersensitivity may be implicated in this process (16,31,45). The consequence of either a humoral or cellular immune response directed against the vascular endothelium would be compromised myocardial oxygenation. Important inflammatory participants in this process include polymorphonuclear leukocytes--which can be activated by complement components--endotoxin, and platelet activating factor (34,46--49). Such activation can result in the production of various leukotrienes, arachidonic acid metabolites, and a variety of cytokines that lead to vascular permeability, leukocyte adherence via various specifically induced adhesion molecules, and the activation of proteolytic enzymes such as protein kinase C (29-33,50). The ability of endothelial cells to express adhesion molecules (ELAM, ICAM, VLAM) in response to inflam-

Cardiovasc Pathol Vol. 2. No. 1 January-March 1993:21-34

HAMMOND ET AL. VASCULAR REJECTION IN CARDIAC TRANSPI,ANTS

33

Figure 26. Electron micrograph of heart from patient with chronic allograft rejection and global ischemia at explant. The myocytes show a preponderant loss of actin over myosin, giving a coarse appearance to the myofilaments. This appearance is typically seen when ischemic damage is present. Compare with the myofilaments in Figures 22-25. (Original plate magnification × 10,000.)

matory stimuli or cytokine release such as I L l can create the morphologic expression of vascular rejection, including endothelial activation and capillaritis (29-33). Furthermore, complement components are similarly capable of inducing expression of such adhesive molecules, which can lead to the adherence and migration of polymorphonuclear leukocytes through vessels, with consequent destruction of adjacent cells such as myocytes by the above-described mechanisms (34,39). Thus it appears likely that the manifestations of vascular rejection, as well as severe rejection, represent the consequences of endothelial cell activation, secretion of cytokines, and increased endothelial cell adherence of leukocytes, migration of such leukocytes, and ischemic damage to the myocardium. The long-term consequence of long-standing or intermittent vascular rejection, if it does not lead to acute allograft loss or demise of the patient, appears in our studies to be allograft coronary artery disease in most patients. In a previous prospective study, we showed a significant and independent correlation with the vascular and mixed patterns of rejection diagnosed in the early posttransplant period and the subsequent development of chronic allograft vasculopathy (10,11). The light microscopic and immunofluorescent findings that allow patient stratification into these patterns of rejection are, therefore, important to recognize because they may lead to allograft loss in a slow and insidious way that is largely untreatable (10,11) The difference in allograft coronary artery disease in patients who manifest vascular or mixed rejection offers a compelling reason to seek and utilize more sensitive predictive methods, such as biopsy immunofluorescence, for earlier detection of these rejection patterns (11). This paper has described the variety of vascular lesions that can occur in patients with cardiac allograft

rejection. Because of the advent of new lmmunosuppressive regimens utilizing agents such as mouse monoclonal antibodies, the important effects of cytokine release by endothelial cells and the action of these factors on vessels can no longer be ignored. The effect of such cytokine release produces a spectrum of morphologic injuries that can be seen histologically by immunofluorescence and confirmed ultrastructurally. Using this information, we may be able to design treatments in which patients are stratified by their rejection patterns and, hence, alter acute risk of allograft loss as well as risk of allograft coronary artery disease (10,11). The authors thank the other health care professionals involved in the Utah Cardiac Transplant program for their help and support. They also thank the A. Lee Christensen Fund of the Deseret Foundation, LDS Hospital, for support of this research. They are grateful to Sherree Terry for her excellent assistance in preparation of the manuscript.

References l. Bristow MR, Gilbert EM, O'Connell JB, et al. OKT3 monoclonal antibody in heart transplantation. Am J Kidney Dis 1988; 11:135-140. 2. Renlund DG, O'Connell JB, Gilbert EM, et al. A prospective comparison of murine monoclonal CD-3 (OKT3) antibody-based and equine antithymocyte globulin-based rejection prophylaxis in cardiac transplantation. Transplantation 1989;47:599~505. 3. Woodley SL, Renlund DG, O'Connell JB, Bristow MR. Immunosuppression following cardiac transplantation. Cardiol Clin 1990;8:83-96. 4. Hammond EH, Yowell RL, Nunoda S, et al. Vascular (humoral) rejection in heart transplantation: pathologic observations and clinical implications. J Heart Transplant 1989;8:430-443. 5. Hammond EH, Wittwer CT, Greenwood J, et al. Relationship of OKT3 sensitization and vascular rejection in cardiac transplant patients receiving OKT3 rejection prophylaxis. Transplantation 1990;50:776-782. 6. Graham MT, Moreno-Cabral CE, Starnes VE, et al. Cytomegalovirus infection is associated with cardiac allograft rejection and atherosclerosis. J A M A 1989;261:3561-3565.

34

HAMMONDET AL. VASCULAR REJECTION IN CARDIAC TRANSPLANTS

7. Alexander JW. The cutting edge--a look to the future in transplantation. Transplantation 1990;49:237-240. 8. Behrendt DM, Billingham ME, Boucek MM, Marxmiller JM, Rose EA. Rejection/infection: the limits of heart transplantation success. In: Proceedings of the Loma Linda International Conference on Pediatric Heart Transplantation. J Heart Lung Transplant 1991: 10;(Part 2):841-847. 9. Billingham ME. Cardiac transplantation: The Pathology of Organ Transplantation. In: Sale GE, ed. Boston; Butterworths 1990: 133. 10. Hammond EH, Ensley RD, Yowell RL, et al. Vascular rejection of human cardiac allografts and the role of humoral immunity in chronic allograft rejection. Transplant Proc 1991;23:26-30. 11. Hammond EH, Yowell RL, Price GD, et al. Vascular rejection and its relationship to allograft coronary artery disease. J Heart Lung Transplant 1992;11 (3):11 l-119. 12. Normann SJ, Salomon DR, Leelachaikul P, et al. Acute vascular rejection of the coronary arteries in human heart transplantation: pathology and correlations with immunosuppression and cytomegalovirus infection. J Heart LungTransplant 1991;10:674-688. 13. Hess ML, Hastillo A, Mohanakumar T, et al. Accelerated atherosclerosis in cardiac transplantation: role ofcytotoxic B-cell antibodies and hyperlipidemia. Circulation 1983:68;(Supp111):94-101. 14. Palmer DC, Tsai CC, Roodman ST, et al. Heart graft atherosclerosis. Transplantation 1985;39:385-388. 15. Uretsky BF, Murali SP, Sukhakar R, et al. Development of coronary artery disease in cardiac transplant patients receiving immunosuppressive therapy with cyclosporine and prednisone. Circulation 1987;76:827-834. 16. Rose EA, Smith CR, Petrossian GA, Barr ML, Reemtsma K. Humoral immune responses after cardiac transplantation: correlation with fatal rejection and graft atherosclerosis. Surgery 1989;106:203-208. 17. Billingham ME. Dilemma of variety of histopathologic grading systems for acute cardiac allograft rejection by endomyocardial biopsy. J Heart Transplant 1989;9:272-276. 18. Billingham ME. Cardiac transplantation. Cardiovasc Clin 1988: 18:185-199. 19. Billingham ME. Diagnosis of cardiac rejection by endomyocardial biopsy. J Heart Transplant 1982;1:25-30. 20. Billingham ME, Cary NRB, Hammond ME, et al. A working formulation for the standardization of nomenclature in the diagnosis of heart and lung rejection: heart rejection study group. J Heart Transplant 1990;9:587-592. 21. Kemnitz J, Cohnert T, Schafers H, et al. A classification of cardiac allograft rejection. Am J Surg Pathol 1987;11:503-515. 22. McAllister HA. Histologic grading of cardiac allograft rejection: a quantitative approach. J Heart Transplant 1989;9:277-282. 23. Schoen F. Interventional and Surgical Cardiovascular Pathology: Clinical Correlations and Basic Principles. Philadelphia: WB Saunders, 1989. 24. Zerbe TR, Arena V. Diagnostic reliability of endomyocardial biopsy for assessment of cardiac allograft rejection. Hum Pathol 1988;19:1307-1314. 25. Dvorak AM, Hammond ME, Dvorak HF, Karnovsky MJ. Loss of cell surface material from peritoneal exudate cells associated with lymphocyte mediated inhibition of macrophage migration from capillary tubes. Lab Invest 1972;27:561-566. 26. Hammond EH, Anderson JL, Menlove RL. Predictive value of immunofluorescence and electron microscopic evaluation of endomyocardial biopsies in the diagnosis and prognosis of myocarditis and idiopathic dilated cardiomyopathy. Am Heart J 1987; 114:1055-1061. 27. Breisblatt WM, Sehulman DS, Stein K, et al. Hemodynamic response to OKT3 in orthotopic heart transplant recipients: evidence of reversible myocardial disfunction. J Heart Lung Transplant 1991;10:359-365.

Cardiovasc Pathol Vol. 2 No. 1 January-March 1993:21-34

28. Cines DB, Lyss AP, Bina M, Corkey R, Kefalides NA, Friedman HA. Fc and C3 receptors induced by herpes simplex virus on cultured human endothelial cells. J Clin Invest 1982;69:123-128. 29. Cotran RS. New roles for the endothelium in inflammation and immunity. Am J Pathol 1987;129:407-413. 30. Cotran RS, Pober JS, Gimbrone MA, et al. Endothelial activation during interleukin 2 immunotherapy: a possible mechanism for the vascular leak syndrome. J Immunol 1987;139:1883-1888. 31. Libby P, Salomon DD, Payne FJ, Schoen FJ, Pober JS. Functions of vascular wall cells related to development of transplantationassociated coronary arteriosclerosis. Transplant Proc 1989;21: 3677-3684. 32. Pober JS. Cytokine-mediated activation of vascular endothelium. Am J Pathol 1988;133:426-433. 33. Pober JS, Collins T, Gimbrone MA, Libby P, Reiss CS. Inducible expression of class II major histocompatibility complex antigens and the immunogenicity of vascular endothelium. Transplantation 1986;41(2):141-146. 34. Yasuda M, Takeuchi K, Hiruma M, et al. The complement system in ischemic heart disease. Circulation 1990;81:156 163. 35. Hammond EH, Yowell R, Greenwood J, Hartung L, Wittwer C. Monitoring of patients for OKT3 sensitization prevents adverse outcome. Transplantation (in press). 36. Talbot D, Givan AL, Shenton BK, et al. Rapid detection of low levels of donor specific IgG by flow cytometry with single and dual color fluorescence in renal transplantation. J Immunol Methods 1988;112:279-283. 37. Brasile L, et al. Clinical significance of the vascular endothelial cell antigen system: evidence for genetic linkage between the endothelial cell antigen system and the major histocompatibility complex. Transplant Proc 1985;17:2318-2321. 38. Andres G, Brentjens JR, Caldwell PRB, Camussi G, Matsuo S. Biology of disease: formation of immune deposits and disease. Lab Invest 1986;55:5:510-520. 39. Hugli TE. Complement and cellular triggering reactions: introductory remarks. FASEB J 1983;43:2540-2557 (minisymposium). 40. Lawley TJ, Bielory L, Gascon P, Yancy KB, Young NS, Frank MM. A prospective clinical and immunologic analysis of patients with serum sickness. N Engl J Med 1984;311:1407-1413. 41. Mannik M. Mechanisms of tissue deposition of immune complexes. J Rheumatol 1987;13:14:35-42. 42. Vaughan JH, Barnett EV, Leadley PJ. Serum sickness: evidence in man of antigen-antibody complexes and free light chains in the circulation during the acute reaction. Ann Intern Med 1967; 67:596-602. 43. Miltenburg AM, et al. Induction of antibody-dependent cellular cytotoxicity against endothelial cells by renal transplantation. Transplantation 1989;48:681-688. 44. Hayry P, et al. Generation and breakdown of a vicious cycle in context of acute allograft rejection. Transplant Proc 1986;18:52-62. 45. Dvorak HF, Galli SJ, Dvorak AM. Cellular and vascular manifestations of cell-mediated immunity. Hum Pathol 1986;17:122-137. 46. Colvin RB. Clinical applications of monoclonal antibodies in renal allograft biopsies. Am J Kidney Dis 1988;11(2):126-130. 47. Bishop AG, Hall BM, Duggin GG, Horvath JS, Sheil AG, Tiller DJ. Immunopathology of renal allograft rejection analyzed with monoclonal antibodies to mononuclear cell markers. Kidney Int 1986;29:708-717. 48. Hall BM. Cellular infiltrates in allografts. Transplant Proc 1987; 19:50-56. 49. Mason DW, Morris PJ. Effector mechanisms in allograft rejection. Annu Rev Immunol 1986;4:119-145. 50. Dinarello CA, Mier JW. Lymphokines. N Engl J Med 1987;317: 940-945.