- Email: [email protected]
Identification of an autoantibody to vascular endothelial cell-specific antigens in patients with systemic vasculitis
Identification Cell-Specific
of an Autoantibody to Vascular Endothelial Antigens in Patients with Systemic Vasculitis
LAUREN BRASILE, B.s.,JoELM. JAMESCERILLI, ~.~.~ocheste~
KREMER, M.D.,JOLENEL. CLARKE,B.S.A/~~~~, NewYork New York
PURPOSE: Although immunologic mechanisms have been postulated in the pathogenesis of vasculitis, an autoimmune process directed against specific autologous vascular wall antigens has not been previously documented. We examined the role of an immunologic response to vascular endothelial cell (VEX) antigens in patients with vasculitis, because of the observed immunogenic&y of VECs in renal and cardiac allograft recipients. PATIENTS AND METHODS: The study patients included 21 with systemic vasculitis and four with hypersensitivity vasculitis. A healthy, normal control group consisted of 51 young subjects and 61 older subjects. Thirty-two patients with connective tissue diseaseswere also evaluated. The presence of autoantibody to autologous monocytes and other cell types was determined with previously described standard crossmatch techniques. Patient sera exhibiting autoantibody to autologous monocytes were screened against a panel of allogeneic cells. The cell panel consisted of concordant T and B lymphocytes, monocytes, and VECs from 16 umbilical cord donors. Sera from four patients were analyzed for specificity to the vascular endothelium of frozen sections of vessels. RESULTS: An autoantibody to VEC antigens was detected in 18 of the 21 patients (86%) with confirmed systemic vasculitis, not of the hypersensitivity type. This autoantibody was highly cytotoxic, complement-fiig, and specific for antigens on the surface of the VEX. Findings on allogeneic VEX panel analysis support the existence of multiple allotypes in the VEX antigen system. In three patients, fluctuations in the titer of the autoantibody generally correlated with clinical symptoms. In the four patients screened for the specificity of the autoantibody to anatomically different cadaveric blood vessels, specific anatomic patterns of reactivity were observed. Autoantibody to VEX antigens was not found in young controls and was documented in low frequency in older controls and in patients with connective tissue diseases.The autoantibody was seen in one of four patients (25%) with hypersensitivity vasculitis. CONCLUSION: Our results indicate that an autoantibody to VEC antigens may be involved in the pathogenesis of systemic vasculitis or may be a diagnostic marker for the diseaseprocess.
74
July
1989
The American
Journal
of Medicine
Volume
87
T
he vascular endothelial cell (VEC) is a complex cell with specialized functions. VECs serve as antigen-presenting cells and are capable of phagocytosis [l]. VECs are able to inactivate hormones, modify compounds in circulation, and are partly responsible for the normal function of platelets [2]. The VECs express surface proteins that are important immunogens in organ transplantation. Vascular lesions are a common feature of allograft rejection [3]. The endothelium is always damaged whether the rejection is immediate or chronic [4]. The systemic antigens, such as ABH and major histocompatibility complex antigens, are expressed on VECs [5,6]. In addition, VECs express many tissue antigens with restricted expression. These include antigens shared with platelets, such as receptors for thrombin, fibrinogen, and von Willebrand factor [7-lo]. VEC antigens have been reported with shared expression with granulocytes such asthe human monocytic and human granulocytic antigens [11,12]. These tissue-restricted antigens reflect, in part, the specialized functions of this cell. We have previously reported that VECs and peripheral blood monocytes share the expression of a restricted antigen system that is linked to the major histocompatibility complex [13-151. The VEC/monocyte antigen is an important immunogen in organ transplantation in that most patients who exhibit preformed antibody to the VEC antigens of their donors lose their grafts to rejection even when the recipient and donor are HLA-identical [14,16-181. Because of the demonstrated immunogenicity of VECs in recipients of renal and cardiac allografts, the role of an immunologic responseto these VEC antigens was examined in patients with vasculitis. Although immunologic mechanisms have been proposed in the pathogenesisof vasculitis, an autoimmune processdirected against specific autologous vascular wall antigenshas not been previously demonstrated. Since the peripheral monocyte represents a convenient, alternative target cell for the VEC, noninvasive evaluation for the presence of anti-VEC antibody could be performed. This report indicates that patients with systemic vasculitis have a high prevalence of autoantibodies to the VEC antigen system-a prevalence that is not found in control populations.
PATIENTS AND METHODS Study Patients Twenty-five patients, including 21 with systemic vasculitis (Group I) and four with hypersensitivity vasculitis (Group II), were randomly studied through the Rheumatology Clinic of the Albany Medical College. Some patients in Group I were studied at the time of disease onset and others were studied at varying intervals after their initial presentation. Patients in Group I had the diagnosis established by a typical
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
TABLE I Clinical Features of Patients with Systemic Vasculitis at the Time of Disease Onset
Patient Number* : : z
f3 9 :: 12 13 14 :; :;
19 20 21 Average T's = Liver The patient
Sex and Age
Interval since Disease Onset (years)
F23 F22
0
Y4;; F43 M66 M61 F30 F29 M40 M60 F68 M60 F63 F72 F73 F33 F68 F66 F36 F49
i
Y2
+ +
t
Lung Disease
Abnormal LFT’s
+
+ +
t t t
i
21 ia E 4'h 'h
t t :
Azotemia
Abnormal Urine Sediment
t
+ t t t
t
:
Skin Disease
ENT Disease
+
+ t
t t t t
t t t
t t
t
t
t t sedimentation rate and headache sedimentation rate and headache sedimentation rate and headache t t t t
t
t
+ t t
Abdomlnal Pain
Anemia
t
t t t
z t
t t
t t t
t
t
t
t :
t t +
t t t
: : 4
Arthritis/ Arthralgia
t t t t t Increased Increased Increased
!
7M:14F/50.0 function number
Fever
Weight Loss
t
1
1
4'h
43%
t
t
29%.
33%
19%
19%
43%
48%
33%
57%
t
t
14%
57%
tests; ENT = ear, nose, and throat. identifies the same individual as in Table II.
constellation of clinical signs and symptoms, and by biopsy in 20 of 21 patients and angiography in one of 21 patients. All were hepatitis B surface antigen-negative. Patients with hypersensitivity vasculitis exhibited cutaneous palpable purpura without systemic visceral involvement. Control Subjects Controls consisted of healthy, normal volunteers divided into two groups: 51 young controls ranging in age from six months to 24 years (mean, 21 years); and 61 older controls ranging in age from 50 to 75 years (mean, 65 years) who were randomly selected from a blood pressure screening clinic at the Albany Medical College. Healthy
Control Subjects with Connective Tissue Diseases Outpatients followed in the Division of Rheumatology at Albany Medical College with the following diagnoseswere screened for the presence of autoantibody to VEC: 12 patients with rheumatoid arthritis; 10 patients with systemic lupus erythematosus (SLE); and 10 patients with progressive systemic sclerosis. Patients were screened as they becameavailable for testing based upon the random scheduling of their outpatient appointment. No patient refused to participate. Autologous Cell Isolation and Autoantibody Determination Blood was obtained from each patient and the mononuclear cells were isolated on Ficoll-Isopaque gradients. Autologous monocytes were isolated from the mononuclear cell population by adherence to plastic Petri dishes. In each case,the purity of the cell preparations was greater than 90% as determined by monoclonal antibody analysis. The presence of autoantibody to the monocytes and other cell types was tested
by the standard NIAID crossmatch techniques [14]. Cells and serawere mixed in wells of a Terasaki microtoxicity testing tray and incubated at room temperature for 30 minutes. Complement, which was prescreened for toxicity, was added to the trays and incubated at room temperature for 120 minutes. The trays were then stained with eosin and fixed with formalin. The trays were evaluated on an inverted phasecontrast microscope, using the following scale: less than 20% cell death = negative; 20%to 30% cell death = weak positive; 30% to 75% cell death = positive; more than 75% cell death = strongly positive. Cell Specificity Determination Patient sera exhibiting autoantibody to autologous monocytes were screened against a panel of allogeneic cells. The cell panel consisted of concordant T and B lymphocytes, monocytes, and VECs from 16 umbilical cord donors, and was evaluated by a technique previously reported [15]. Since the four cell types were obtained from each of the 16 donors, cell specificity of the autoantibody could be determined. This panel was not constant, as the serum samples were screened when cords became available. The aforementioned standard complement-dependent cytotoxicity technique was utilized to evaluate the panels. This allogeneic panel analysis identified the cell specificity of the autoantibody detected with the autologous monocyte as also being specific for VECs. VesselSpeciflcity Determination Frozen sections of some of the major abdominal blood vesselswere made at the time of organ donation for cadaveric renal transplantation. Sera from four patients, three positive and one negative for autoantibody to VECs, were tested for specificity to the vascular endothelium of frozen sections of vesselsusing a July
1989
The American
Journal
of Medicine
Volume
87
75
VEC ANTIGEN
SYSTEM
TABLE II Autoantibody
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
in Patients with Systemic Vasculitis
Patient Number
Autoantibody to VEC/Monocyte*
Diagnosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’sgranulomatosis Wegener’sgranulomatosis Polyarteritis nodosa Polyarteritis nodosa Polyarteritis nodosa Polyarteritis nodosa Temporal arteritis Temporal arteritis Temporal arteritis Mixed cryoglobulinemia Mixed cryoglobulinemia Takayasu’s arteritis Rheumatoid vasculitis Systemic lupus, large vessel Overlap syndrome of necrotizingvasculitis Primary central nervous system vasculitis
: : 5 ; i :: :i :“5 :; :; 4;
assessedby the following scale: negative = absenceof antibody binding, equal to the negative control of normal sera;weakly positive = weak intermittent binding; positive = continuous, strong binding; and strongly positive = binding equivalent to the positive control consisting of a monoclonal antibody to VEC antigens. Negative controls consisted of sera from normal male subjects not undergoing transfusion, and positive controls included anti-ABH, heterologous anti-VEC, and monoclonal anti-VEC sera. All specimenswere run on two occasions in duplicate and evaluated by two observers.
- % = % cell death of autologous monocytes. 7 Had received recent multiple transfusions. z Received serial plasmaphereses; the patient vidual in Table I.
Negativer Positive 30% Positive 75% Positive 75% Positive 50% Weakly positive Weakly positive Positive 75% Positive 65% Positive 50% Positive 30% Positive 90% Positive 50% Positive 50% Negative* Negative* Positive 25% Positive 50% Weakly positive Positive 75% Weakly positive
number
identifies
20% 20%
RESULTS
20% 20%
the same indi-
TABLE Ill Autoantibody
in Patients with Hypersensitivity
Vasculitis I
Palpable Patient
Purpura
Etiology Secondary Secondary to Drug to SLE
+
:
: -
: +
:
+ -
Unknown
VEC Autoantibody Positive
-
-
+
+
TABLE IV Summary of the Frequency of Autoantibody to VEC/ Monocyte Antigen in Patients with Vasculitis 1 Autologous Cytotoxicity Monocytes* Group I (Sysiemic Vasculitis) 3/21(14%) 5/21(24%) 8/21(38%) 5/21(24%)
Negative Weakly positive Positive Strongly positive I
Group II (Hypersensitivity l/4 (25%)
Vasculitis) Positive
I
1
T and B lymphocyte
ciossmatches
all negative.
biotin-avidin immunoperoxidase technique. Frozen sections 4 microns in thickness were fixed in cold acetone. The sera were overlayed on the tissue and incubated for 30 minutes. After washing, the sections were overlayed with a biotinylated goat anti-human IgG for 30 minutes. After additional washing, the slides were next overlayed with an avidinlbiotinylated peroxidase complex and incubated for 45 minutes. After washing, substrate was added for approximately four minutes. The reactions were then stopped and the sections were 76
July 1989
The American
Journal
of Medicine
Volume
of Patients with Vasculitis Seven patients had the diagnosisof Wegener’s granulomatosis, four had polyarteritis nodosa, three had temporal arteritis, two had essential mixed cryoglobulinemia, and one each had a diagnosis of Takayasu’s arteritis, rheumatoid vasculitis, lupoid vasculitis, the overlap syndrome of necrotizing vasculitis, and primary central nervous system vasculitis. The clinical signs and symptoms of the 21 patients studied are summarized in Tables I and II. Four patients with hypersensitivity vasculitis were also studied. All four patients had cutaneous manifestations of palpable purpura on the lower extremities without systemic vasculitic involvement (Table III). Two of the four had a hypersensitivity drug reaction that disappeared when the drug was discontinued. One of the four patients had a diagnosis of SLE and developed manifestations of cutaneous vasculitic involvement of the hypersensitivity type. Description
87
ALtoantibody Screening in Patients with Vasculitis Eighteen of 21 patients (86%) in Group I exhibited an autoantibody to autologous monocytes (Table IV). In no casewas antibody detected that wascytotoxic to the autologous T or B lymphocytes. Two patients with mixed cryoglobulinemia had negative results of testing for autoantibody to VEC (Table II). Both patients had multiple plasmapheresesprior to testing for autoantibody. In addition, Patient 1 (Table II), who was alsonegative, had received multiple transfusions prior to our evaluation. The cell specificity of the autoantibody was further determined by screening the sera from 18 of the autoantibody-positive patients with vasculitis against a panel of concordant allogeneic T or B lymphocytes, monocytes, and VECs. The autoantibody in the 18 autoantibody-positive patients in Group I reacted only with the panel monocytes and concordant VECs. In no instance,wasreactivity to the alldgeneic lymphocytes detected, nor did the sera react only with the monocytes. The sera from the patients with vasculitis exhibited various patterns of reactivity to the panel VECs. For example, patient sera labeled CL86, PR986, and PR287 were cytotoxic to VEC # 1027,891, 973, and 945, and negative to the rest of the panel VECs (Table V). These three sera reacted similarly and therefore may have been recognizing the same antigenic epitope on the endothelial cell. Five distinct “clusters” of reactivity were detected (Table V). The panel results indicate that the antibody to autologous monocytes was an anti-VEC antibody and not antiHLA, asit was never cytotoxic to lymphocytes or neutrophils. In contrast to the patients with systemic vas-
VEC ANTIGEN
SYSTEM IN SYSTEMIC
VASCIJLITIS
,’ BRASILE ET AL
TABLE V Cluster Analysis of Vasculitic Patient Sera to a Panel of VEC Cell Code Number 1,027 891 973 945
CL86
PR986
+
Cluster 1 +
+’ -
z +
PR287
961 947 1,021 1,004 996 1,007 1,026
BK485
+ z+
BK186
Patient Sera Code Number BM LY CN H1286
H191878
ER
ML
HL
Cluster 4 + +
Cluster
+ + + +
2
•i+ T
z + -
Cluster + + -
:
3 + + -
+ +
+
+
+
+
+
;
+ -
+ ---
+
+
BK7986
WR
Cluster 5 +
-
+
+
+
+
Sera: CL86. PR 986 and PR 287 representing
three different patients formed a cluster of reactivity. in that all three sera were cytotoxic to VEC numbers 1027.891, 973, and probably 945. Likewise, a second cluster of reactivity involved sera BK485. BK186. and BM. which reacted with the VEC numbers 961, 1021, and 1004. The third cluster involves sera LY. CN, and H1286 and reacted with VEC numbers 996.1007, and probably961. The fourth cluster involved sera ML and HL which reacted with VC numbers 1027 and 996. The fifth cluster includes sera BK 7986 and WR which reacted with VEC numbers 961.11026, and probably 891, These data suggest that the autoantibody was recognizing multiple distinct allotypes within the VEC/monocyte antigen system.
culitis, autoantibody was detected in only one of four patients in Group II (Table VI).
TABLEVI Frequencyof Autoantibody in Control Patients
Autoantibody Screening in Healthy Control Subjects The autoantibody was found in low Frequency in the control populations (Table VI). Autoantibody to VEC was not detected in 51 young controls ranging from six months to 24 years of age and was detected in eight of 61 (13%) of the controls greater than 50 years of age. Control Subjects with Connective Tissue Diseases The presence of autoantibody to VEC was found to be associatedwith other known connective tissue diseaseswith low frequency (Table VI). It was observed in one of 10 (10%) patients with SLE, three of 12 (25%) patients with rheumatoid arthritis, and zero of 10 patients with diffuse systemic sclerosis. Sequential Testing Using historical sera, additional screeningsfor autoantibody were performed on three patients. In Patient A, the development of autoantibody coincided with clinical manifestations of the disease(Figure 1). Using historical frozen sera, this patient was retrospectively found to be autoantibody-negative prior to the initial diagnosis of her disease.The autoantibody was present in high titer at the time when the patient was symptomatic. Patient B was first screenedat the time of diagnosisand exhibited high titers of autoantibody (Figure 2). The autoantibody disappeared and the patient exhibited marked clinical improvement after institution of a regimen of cyclophosphamide and prednisone. In Patient C, autoantibody titers to VEC were strongly positive at the time of diseaseonset and later reverted to negative during a prolonged period of clinical remission (Figure 3). Blood VesselSpecificity Sera from four patients with systemic vasculitis (Group I) were screened against a panel of frdzen cadaveric vesselsections from different anatomic sites in order to further confirm the cell specificity of the autoantibody (Table VII). Two patients (Wegener’s granulomatosis) were classified as strongly positive for
Number of Patients* Young controls, 25 years/age Older controls, 50 years/age
O/51 (0)
8/61(13)
Controls with connective tissue disease Rheumatoid arthritis SLE Scleroderma * Values
in parentheses
are percents.
autoantibody, one patient (temporal arteritis) was positive, and the fourth patient (mixed cryoglobulinemia) was negative for autoantibody to VEC monocyte antigens by cytotoxicity technique. The sera from the three positive autoreactive patients (Patients 1 through 3) bound to sections of vena cava, whereasthe sera from the fourth nonreactive patient did not. Interestingly, no sera (regardless of their cytotoxicity reactions to autologous monocytes and allogeneic VECs) bound to allogeneic aorta.
COMMENTS The VECs express tissue-specific antigens that can function asa potent immunogen in organ transplantation [6,19,20]. The peripheral blood monocyte also expressessome VEC antigens in. vitro. In random sera containing anti-monocyte antibody, VECs and monocytes exhibited equivalent reactivity in 91%of the concordant pairs tested [16]. There was no detectable cytotoxicity directed against concordant panel T or B lymphocytes or granulocytes [21]. The results of family segregation studies indicate that the genes coding for the VEC/monocyte antigen loci are in linkage with the major histocompatibility complex and may be a newly identified HLA antigen [22,23]. The observation that the monocyte, but no other resting cell types yet studied, can serve as an alternative target cell for the VEC-specific antibody provided an opportunity to perform noninvasive screenings for autoantibody to July
1989 The American
Journal
of Medicine
Volume
87
77
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULlilS
/ BRASILE ET AL
strong Pos it ive Y P z
Pos it ive
g $ g 2
Figure 1. Patient A (Patient 9 in Tables I and II). Relation of autoantibody response to VEC and clinical disease activity. Retrospective historical sera through January 1986. In June 1983 and March 1984; the patient complained of fatigue. In May 1984, the patient developed arthralgias, which persisted until June 1984. In January 1986, the patient developed vasculitic cutaneous nodules on the feet, which persisted until March 1986.
Weak Positive
Negative
marked disease
.
L
3184
6/83
5104
6104
l/86
3/86
strong H
Positive
r” 5
5 5
improvement PO.5 it ive
Weak Positive
1
Negative
5
strong Positive
I
marked disease 1
3105
4106
5186
.
E $ E 2 2
Weak Positive Positive Negative
t-remission+ :
T marked disease
.
1 2/a
78
July 1989
Figure 2. Patient B (Patient 3 in Tables i and II). Relation of autoantibody response to VEC and cljnlcal disease activity. Patient with an established diagnosis of Wegener,‘s granulomat,osis complained of arthralgias and nasal stuffiness in March 1985. In April 1986, there was a recurrence of hemoptysis with pulmonary infiltrates, cutaneous vasculitis, arthritis, microhematuria, and anemia. Treatment with cyclophosphamide 150 mg/day was resumed, and in May 1986, the pulmonary abnormalities, arthritis, and cutaneous vasculitis had cleared.
The American
Journal
I
9184
510 I
of Medicine
Volume
87
, 5185
4106
Fleure 3. Patient C (Patient 10 in Tables I &d II). Relation of autoaritibody response to VEC ahd clinical disease activity. Patient presented with arthralgias, microhematuria, hematochezia, and hypereosinophilia in February 1981. In May 1981, there was ah acute onset of absent pulses in the lower extremities. Treatment with cyclophosphamide was begun, and disease resolved. During the period from September 1984 to April 1986, the patient’s condition was in clinical remission without cyclophosphamide therapy.
VEC ANTIGEN
VEC. This report identifies an autoantibody to VEC antigens in 86% of the patients studied with systemic vasculitis. This may be an underestimate in that serial plasmaphereses and blood transfusions may have affected our ability to detect autoantibody tq VEC in three patients with systemic vasculitis in whom this antibody was not observed. The autoantibody was not present in healthy, normal, young control subjects and was uncommon in an older control population and in patients with other connective tissue diseases. Although autologous monocytes were used for the initial detection of autoantibody, the antibody is clearly directed against VEC antigens. In all patients, the autoantibody detected initially with autologous monocytes also reacted to the concordant monocytes and VECs in the panel analysis. In no instances were panel monocytes positive and the concordant VECs negative. The observation that concordant T and B lymphocytes and neutrophils do not react with the autoantibody indicates that the reaction is not directed against class I or II HLA antigens. The sera formed “clusters” of reactivity to the panel VECs, indicating the existence of at least five allotypes within the VEC monocyte antigen system (Table V). The patient sera positive for autoantibody exhibited reactivity to the panel VECs that ranged from 20% to 67%. The fact that sera from different patients formed clusters with distinctly different patterns of reactivity strongly supports the premise that multiple distinct allotypes are involved within the VEC system. The specificity of the autoantibody to VEC is further demonstrated by the crossmatching of sera from autoantibody-positive patients with vasculitis with the endothelial lining of cadaveric blood vessels. When this was performed, we observed that the autoantibody bound to specific anatomic vascular sites. That is, the autoantibody reacted weakly, or not at all, to the aorta, superior mesenteric artery, or iliac veiri, but reacted strongly and uniformly to inferior mesenteric and splenic arteries. The cytotoxicity method used for detection of autoantibody to VEC is a less sensitive technique than the biotin-avidin immunoperoxidase technique used in the staining of the cadaveric blood vessels. The detection of antibody to these blood vessels in the one patient who did not have autoantibody to monocytes (Patient 16, Tables II and VII) could be due to subthreshold antibody levels for detection by the cytotoxicity technique. These different serologic patterns cannot be explained by anti-HLA antibody, as the.sera either initially contained no anti-HLA antibody by panel analysis or were sequentially absorbed until T- and B-cell reactivity was eliminated prior to the panel screening. These results suggest that the different patterns of disease that are recognized clinically as specific vasculitic syndromes might be due to a varied antigenic expression on the surface of VECs within the vascular system. The VEC system described herein links an autoantibody directly to vascular endothelium for .the first time in patients with systemic vasculitis. Although an anti-endothelial cell antibody has been described in patients with SLE [24], it is characterized by reactivity with both B lymphocytes and monocytes and is found in patients with active disease without vasculitis. The SLE autoantibody thus appears to be serologically distinct from the VEC autoantibody. Anti-neutrophil cytoplasmic antibodies (PMN-
SYSTEM
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
TABLE VII Patterns of Reactivity to Abdominal Blood Vessels of Sera from Patients with Systemic Vasculitis
I
l
Vessels
3
Inferior mesenteric artery Superior mesenteric artery Splenic artery Iliac artery Aorta Inferior mesenteric vein Iliac vein Vena cava
+ + +
See Tables
I and II for diagnosis
Patient Number* 2 12 +;+ + -
+ and clinical
16
:
+ + -
+ +/-
‘+ +/-
+ -
presentation.
ACPA) have been described by several investigators in association with Wegener’sgranulomatosis and microscopic polyarteritis nodosa [25-301. Speculation on the mechanism of vascular injury in these conditions has centered around endothelial damAge caused by the releaseof lytic enzymes and toxic oxygen speciesfrom autoantibody-activated neutrophils [29]. As others have pointed out, the exact pathobiologic implication of the PMN-ACPA system is highly speculative [31]. It is possible that the necrotizing lesions of Wegener’s granulomatosis precede the development of PMNACPA, and the antibodies may actually be a marker for the granulomatous disease process. The clinical lesionsthat characterize this disorder are undoubtedly associated with humoral and cellular interactions involving other immunologic systems f31). In summary, we have identified the presence of an autoantibody in 86%of patients with systemic vasculitis. This autoantibody is highly cytotoxic, complement-fixing, and exhibits spetificity for VEC. In three patients studied using historical sera (Figures 1 through 3), the presence and titer of the autoantibody appear to correlate with the clinical course of disease. Sera from patients with vasculitis exhibited distinct patterns of reactivity with panel endothelial cells, suggesting that allotypes of the VEC antigens exist. The sera also react in a non-uniform manner with different distinct anatomic vascular sites from cadaveric blood vesseltissue, implying a heterogeneity within the VEC antigen system. Taken together, these results may point to a role for the VEC autoantibody-VEC antigen system in the determination of the distinct patterns of vascular reactivity in different vasculitic syndromes. These interpretations are speculative and will require further correlations of the presence and titer of autoantibody to VEC with diseaseactivity in a variety of vasculitic syndromes. In addition, further characterization of the VEC antigen itself and the exploration of the POSsibility of varied antigen expression within the vascular tree will be needed in order to better elucidate its role in the diseasepathogeneis of multiple vasculitic syndromes.
REFERENCES 1. Cotran RS: Endothelial phagocytosis. An electromicroscopy study. Exp Mel Pathol 1965; 4: 217-231. 2. Gimbrone MA Jr, Spaet TH. eds: Progress in hemostasis and thrombosis, ~013.
Orlando, Florida: Grune and Stratton.
1976.
3. Baldwin WM Ill, Claas FHJ. VanEs IA. VanRood JJ: Distribution of endothelialmonocyte and HlA antigenson renal vascular endothelium. Transplant Proc 1981; 13: 103-106.
July 1989
The American
Journal
of Medicine
Volume
87
79
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULITIS./
BRASILE
ET AL
4. Williams GM, TerHarr A, Packs LC. et a/: Endothelial changes associated with hyperacute, acute, and chronic renal allograft rejection in man. Transplant Proc 1973; 5: 819-822. 5. Jaffe EA. Nachman RL. Becker CG, Minicki CR: Culture of human endothelial cells derived from umbilical veins. J Clin Invest 1973; 52: 2745-2756. 6. Moraes JR, Stasny P: A new antigen system expressed in human endothelial cells. J Clin Invest 1977; 60: 449-454. 7. Rodgers GM, Shuman MA: Prothrombin is activated on vascular endothelial cells by factor Xa and calcium. Proc Natl Acad Sci USA 1983: 80: 7001-7005. 8. Maruyama I, Salem HH. Majerus PW: Coagulation factor Va binds to human umbilical vein endothelial cells and accelerates protein C activation. J Clin Invest 1984; 74: 224-230. 9. Awbrey BJ, Hoak JC. Owen WG: Binding of human thrombin to cultured human endothelial cells. J Biol Chem 1979; 254: 4092-4095. 10. Bennet JS, Vflaire G: Exposure of platelet fibrinogen receptors by AOP and epinephrine. J Clin Invest 1979: 64: 1393-1401. ll.Thompson JS. Herbrick JM, Burns CP, etatGranulocyte antigens detected by cytotoxicity (GCY) and capillary agglutination (CAN). Transplant Proc 1978: 10: 885-888. 1ZThompson JS, Goeken NE, Brown SA. Rhoades JR: Antigens common to monocytes and endothelial cells: detection by monoclonal and allospecific antibodies. In: Progress in immunobiology: blood cell antigens and bone marrow transplantation, vol 149. New York: Alan R Liss, 1984; 169-175. 13. Cerilli J. Brasile L: Endothelial cell alloantigens. Transplant Proc 1988: 12: 3742. 14. NIAID Manual of Tissue Typing Techniques 1979-1980, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda. Maryland. 15. Cerilli J, Bay W, Brasile L: The significance of the monocyte crossmatch in recipients of living related HLA identical kidney grafts. Human lmmunol 1983; 7:45-50. 16. Brasile L, Galouzis T. Clarke J. Cerilli J: The 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. 17. Brasile L, Rodman E, Shield F Ill, Clarke J. Cerilli J: The association of antivascular endothelial cell antibody with hyperacute rejection: a case report. Surgery 1986; 99: 637639. 18. Cerilli J. Jesseph JE, Miller AC: The significance of antivascular endothelium
80
July
1989
The American
Journal
of Medicine
Volume
87
antibody in renal transplantation. Surg Gynecol Obstet 1972; 135: 246-249. 19. Paul CC, Class FHJ, VanEs LA: Accelerated rejections of a renal allograft associated with pretransplantation antibodies directed against donor antigens on endothelium and monocytes. N Engl J Med 1979; 300: 1258-1260. 20. Cerilli J. Brasile L. Galouzis T. DeFrancis B: Clinical significance of antimonocyte antibody in kidney transplant recipients. Transplantation 1981; 32: 495-497. 21. Clarke JL. Brasile L, Rhoades J, Thompson JS, Cerilli J: Serologic comparison of anti-granulocyte and anti-VEC cytotoxicity of human myelomonocytic antigen system (HMMA) and vascular endothelial cell system (VEC). Transplant Proc 1989; 21: 647-648. 22. Brasile L, Clarke J, Cerilli J: A probable mechanism for allograft rejection in HLA-identical combinations. Transplant Proc 1987; 19: 894-895. 23. Neppert J. Mueller-Eckhardt G, Stranzenbach W. Gottsche B. Mueller-Eckhardt C: Endothelial monocyte antigens are coded for by loci centomeric of H&B. VOX Sang 1985; 49: 221-229. 24. Cines DB, Lyss AP. Reeber M. Bina M, DeHoratias RJ: Presence of complement-fixing antiendothelial cell antibodies in systemic lupus erythematosus. J Clin Invest 1984; 73: 611-625. 25. Van der Woude FJ. Rasmussen N, Lobatto S, et al: Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet 1985; I: 425-429. 26. Gross WL. Ludemann G, Kiefer G. Lehmann H: Anticytoplasmic antibodies in Wegener’s granulomatosis (letter). Lancet 1986; I: 806. 27. Ludemann G. Gross WL: Autoantibodies against cytoplasmic structures of neutrophil granulocytes in Wegener’s granulomatosis. Clin Exp lmmunol 1987; 69: 350-357. 28. Savage COS. Winearls CG. Jones S, Marshall PD, Lockwood CM: Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vasculitis. Lancet 1987; I: 1389-1393. 29. Falk RJ. Jennette JC: Anti-neutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and crescentic glomerulonephritis. N Engl J Med 1988; 318: 16511657. 30. Specks U. Wheatley CL, McDonald TJ. Rohrbach MS, DeRemee RA: Anticytoplasmic autoantibodies in the diagnosis and follow-up of Wegener’s granulomatosis. Mayo Clin Proc 1989; 64: 28-36. 31. Cross CE. Liilington GA: Serodiagnosis of Wegener’s granulomatosis: pathobiologic and clinical implications. Mayo Clin Proc 1989: 64: 119-122.
of an Autoantibody to Vascular Endothelial Antigens in Patients with Systemic Vasculitis
LAUREN BRASILE, B.s.,JoELM. JAMESCERILLI, ~.~.~ocheste~
KREMER, M.D.,JOLENEL. CLARKE,B.S.A/~~~~, NewYork New York
PURPOSE: Although immunologic mechanisms have been postulated in the pathogenesis of vasculitis, an autoimmune process directed against specific autologous vascular wall antigens has not been previously documented. We examined the role of an immunologic response to vascular endothelial cell (VEX) antigens in patients with vasculitis, because of the observed immunogenic&y of VECs in renal and cardiac allograft recipients. PATIENTS AND METHODS: The study patients included 21 with systemic vasculitis and four with hypersensitivity vasculitis. A healthy, normal control group consisted of 51 young subjects and 61 older subjects. Thirty-two patients with connective tissue diseaseswere also evaluated. The presence of autoantibody to autologous monocytes and other cell types was determined with previously described standard crossmatch techniques. Patient sera exhibiting autoantibody to autologous monocytes were screened against a panel of allogeneic cells. The cell panel consisted of concordant T and B lymphocytes, monocytes, and VECs from 16 umbilical cord donors. Sera from four patients were analyzed for specificity to the vascular endothelium of frozen sections of vessels. RESULTS: An autoantibody to VEC antigens was detected in 18 of the 21 patients (86%) with confirmed systemic vasculitis, not of the hypersensitivity type. This autoantibody was highly cytotoxic, complement-fiig, and specific for antigens on the surface of the VEX. Findings on allogeneic VEX panel analysis support the existence of multiple allotypes in the VEX antigen system. In three patients, fluctuations in the titer of the autoantibody generally correlated with clinical symptoms. In the four patients screened for the specificity of the autoantibody to anatomically different cadaveric blood vessels, specific anatomic patterns of reactivity were observed. Autoantibody to VEX antigens was not found in young controls and was documented in low frequency in older controls and in patients with connective tissue diseases.The autoantibody was seen in one of four patients (25%) with hypersensitivity vasculitis. CONCLUSION: Our results indicate that an autoantibody to VEC antigens may be involved in the pathogenesis of systemic vasculitis or may be a diagnostic marker for the diseaseprocess.
74
July
1989
The American
Journal
of Medicine
Volume
87
T
he vascular endothelial cell (VEC) is a complex cell with specialized functions. VECs serve as antigen-presenting cells and are capable of phagocytosis [l]. VECs are able to inactivate hormones, modify compounds in circulation, and are partly responsible for the normal function of platelets [2]. The VECs express surface proteins that are important immunogens in organ transplantation. Vascular lesions are a common feature of allograft rejection [3]. The endothelium is always damaged whether the rejection is immediate or chronic [4]. The systemic antigens, such as ABH and major histocompatibility complex antigens, are expressed on VECs [5,6]. In addition, VECs express many tissue antigens with restricted expression. These include antigens shared with platelets, such as receptors for thrombin, fibrinogen, and von Willebrand factor [7-lo]. VEC antigens have been reported with shared expression with granulocytes such asthe human monocytic and human granulocytic antigens [11,12]. These tissue-restricted antigens reflect, in part, the specialized functions of this cell. We have previously reported that VECs and peripheral blood monocytes share the expression of a restricted antigen system that is linked to the major histocompatibility complex [13-151. The VEC/monocyte antigen is an important immunogen in organ transplantation in that most patients who exhibit preformed antibody to the VEC antigens of their donors lose their grafts to rejection even when the recipient and donor are HLA-identical [14,16-181. Because of the demonstrated immunogenicity of VECs in recipients of renal and cardiac allografts, the role of an immunologic responseto these VEC antigens was examined in patients with vasculitis. Although immunologic mechanisms have been proposed in the pathogenesisof vasculitis, an autoimmune processdirected against specific autologous vascular wall antigenshas not been previously demonstrated. Since the peripheral monocyte represents a convenient, alternative target cell for the VEC, noninvasive evaluation for the presence of anti-VEC antibody could be performed. This report indicates that patients with systemic vasculitis have a high prevalence of autoantibodies to the VEC antigen system-a prevalence that is not found in control populations.
PATIENTS AND METHODS Study Patients Twenty-five patients, including 21 with systemic vasculitis (Group I) and four with hypersensitivity vasculitis (Group II), were randomly studied through the Rheumatology Clinic of the Albany Medical College. Some patients in Group I were studied at the time of disease onset and others were studied at varying intervals after their initial presentation. Patients in Group I had the diagnosis established by a typical
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
TABLE I Clinical Features of Patients with Systemic Vasculitis at the Time of Disease Onset
Patient Number* : : z
f3 9 :: 12 13 14 :; :;
19 20 21 Average T's = Liver The patient
Sex and Age
Interval since Disease Onset (years)
F23 F22
0
Y4;; F43 M66 M61 F30 F29 M40 M60 F68 M60 F63 F72 F73 F33 F68 F66 F36 F49
i
Y2
+ +
t
Lung Disease
Abnormal LFT’s
+
+ +
t t t
i
21 ia E 4'h 'h
t t :
Azotemia
Abnormal Urine Sediment
t
+ t t t
t
:
Skin Disease
ENT Disease
+
+ t
t t t t
t t t
t t
t
t
t t sedimentation rate and headache sedimentation rate and headache sedimentation rate and headache t t t t
t
t
+ t t
Abdomlnal Pain
Anemia
t
t t t
z t
t t
t t t
t
t
t
t :
t t +
t t t
: : 4
Arthritis/ Arthralgia
t t t t t Increased Increased Increased
!
7M:14F/50.0 function number
Fever
Weight Loss
t
1
1
4'h
43%
t
t
29%.
33%
19%
19%
43%
48%
33%
57%
t
t
14%
57%
tests; ENT = ear, nose, and throat. identifies the same individual as in Table II.
constellation of clinical signs and symptoms, and by biopsy in 20 of 21 patients and angiography in one of 21 patients. All were hepatitis B surface antigen-negative. Patients with hypersensitivity vasculitis exhibited cutaneous palpable purpura without systemic visceral involvement. Control Subjects Controls consisted of healthy, normal volunteers divided into two groups: 51 young controls ranging in age from six months to 24 years (mean, 21 years); and 61 older controls ranging in age from 50 to 75 years (mean, 65 years) who were randomly selected from a blood pressure screening clinic at the Albany Medical College. Healthy
Control Subjects with Connective Tissue Diseases Outpatients followed in the Division of Rheumatology at Albany Medical College with the following diagnoseswere screened for the presence of autoantibody to VEC: 12 patients with rheumatoid arthritis; 10 patients with systemic lupus erythematosus (SLE); and 10 patients with progressive systemic sclerosis. Patients were screened as they becameavailable for testing based upon the random scheduling of their outpatient appointment. No patient refused to participate. Autologous Cell Isolation and Autoantibody Determination Blood was obtained from each patient and the mononuclear cells were isolated on Ficoll-Isopaque gradients. Autologous monocytes were isolated from the mononuclear cell population by adherence to plastic Petri dishes. In each case,the purity of the cell preparations was greater than 90% as determined by monoclonal antibody analysis. The presence of autoantibody to the monocytes and other cell types was tested
by the standard NIAID crossmatch techniques [14]. Cells and serawere mixed in wells of a Terasaki microtoxicity testing tray and incubated at room temperature for 30 minutes. Complement, which was prescreened for toxicity, was added to the trays and incubated at room temperature for 120 minutes. The trays were then stained with eosin and fixed with formalin. The trays were evaluated on an inverted phasecontrast microscope, using the following scale: less than 20% cell death = negative; 20%to 30% cell death = weak positive; 30% to 75% cell death = positive; more than 75% cell death = strongly positive. Cell Specificity Determination Patient sera exhibiting autoantibody to autologous monocytes were screened against a panel of allogeneic cells. The cell panel consisted of concordant T and B lymphocytes, monocytes, and VECs from 16 umbilical cord donors, and was evaluated by a technique previously reported [15]. Since the four cell types were obtained from each of the 16 donors, cell specificity of the autoantibody could be determined. This panel was not constant, as the serum samples were screened when cords became available. The aforementioned standard complement-dependent cytotoxicity technique was utilized to evaluate the panels. This allogeneic panel analysis identified the cell specificity of the autoantibody detected with the autologous monocyte as also being specific for VECs. VesselSpeciflcity Determination Frozen sections of some of the major abdominal blood vesselswere made at the time of organ donation for cadaveric renal transplantation. Sera from four patients, three positive and one negative for autoantibody to VECs, were tested for specificity to the vascular endothelium of frozen sections of vesselsusing a July
1989
The American
Journal
of Medicine
Volume
87
75
VEC ANTIGEN
SYSTEM
TABLE II Autoantibody
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
in Patients with Systemic Vasculitis
Patient Number
Autoantibody to VEC/Monocyte*
Diagnosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’s granulomatosis Wegener’sgranulomatosis Wegener’sgranulomatosis Polyarteritis nodosa Polyarteritis nodosa Polyarteritis nodosa Polyarteritis nodosa Temporal arteritis Temporal arteritis Temporal arteritis Mixed cryoglobulinemia Mixed cryoglobulinemia Takayasu’s arteritis Rheumatoid vasculitis Systemic lupus, large vessel Overlap syndrome of necrotizingvasculitis Primary central nervous system vasculitis
: : 5 ; i :: :i :“5 :; :; 4;
assessedby the following scale: negative = absenceof antibody binding, equal to the negative control of normal sera;weakly positive = weak intermittent binding; positive = continuous, strong binding; and strongly positive = binding equivalent to the positive control consisting of a monoclonal antibody to VEC antigens. Negative controls consisted of sera from normal male subjects not undergoing transfusion, and positive controls included anti-ABH, heterologous anti-VEC, and monoclonal anti-VEC sera. All specimenswere run on two occasions in duplicate and evaluated by two observers.
- % = % cell death of autologous monocytes. 7 Had received recent multiple transfusions. z Received serial plasmaphereses; the patient vidual in Table I.
Negativer Positive 30% Positive 75% Positive 75% Positive 50% Weakly positive Weakly positive Positive 75% Positive 65% Positive 50% Positive 30% Positive 90% Positive 50% Positive 50% Negative* Negative* Positive 25% Positive 50% Weakly positive Positive 75% Weakly positive
number
identifies
20% 20%
RESULTS
20% 20%
the same indi-
TABLE Ill Autoantibody
in Patients with Hypersensitivity
Vasculitis I
Palpable Patient
Purpura
Etiology Secondary Secondary to Drug to SLE
+
:
: -
: +
:
+ -
Unknown
VEC Autoantibody Positive
-
-
+
+
TABLE IV Summary of the Frequency of Autoantibody to VEC/ Monocyte Antigen in Patients with Vasculitis 1 Autologous Cytotoxicity Monocytes* Group I (Sysiemic Vasculitis) 3/21(14%) 5/21(24%) 8/21(38%) 5/21(24%)
Negative Weakly positive Positive Strongly positive I
Group II (Hypersensitivity l/4 (25%)
Vasculitis) Positive
I
1
T and B lymphocyte
ciossmatches
all negative.
biotin-avidin immunoperoxidase technique. Frozen sections 4 microns in thickness were fixed in cold acetone. The sera were overlayed on the tissue and incubated for 30 minutes. After washing, the sections were overlayed with a biotinylated goat anti-human IgG for 30 minutes. After additional washing, the slides were next overlayed with an avidinlbiotinylated peroxidase complex and incubated for 45 minutes. After washing, substrate was added for approximately four minutes. The reactions were then stopped and the sections were 76
July 1989
The American
Journal
of Medicine
Volume
of Patients with Vasculitis Seven patients had the diagnosisof Wegener’s granulomatosis, four had polyarteritis nodosa, three had temporal arteritis, two had essential mixed cryoglobulinemia, and one each had a diagnosis of Takayasu’s arteritis, rheumatoid vasculitis, lupoid vasculitis, the overlap syndrome of necrotizing vasculitis, and primary central nervous system vasculitis. The clinical signs and symptoms of the 21 patients studied are summarized in Tables I and II. Four patients with hypersensitivity vasculitis were also studied. All four patients had cutaneous manifestations of palpable purpura on the lower extremities without systemic vasculitic involvement (Table III). Two of the four had a hypersensitivity drug reaction that disappeared when the drug was discontinued. One of the four patients had a diagnosis of SLE and developed manifestations of cutaneous vasculitic involvement of the hypersensitivity type. Description
87
ALtoantibody Screening in Patients with Vasculitis Eighteen of 21 patients (86%) in Group I exhibited an autoantibody to autologous monocytes (Table IV). In no casewas antibody detected that wascytotoxic to the autologous T or B lymphocytes. Two patients with mixed cryoglobulinemia had negative results of testing for autoantibody to VEC (Table II). Both patients had multiple plasmapheresesprior to testing for autoantibody. In addition, Patient 1 (Table II), who was alsonegative, had received multiple transfusions prior to our evaluation. The cell specificity of the autoantibody was further determined by screening the sera from 18 of the autoantibody-positive patients with vasculitis against a panel of concordant allogeneic T or B lymphocytes, monocytes, and VECs. The autoantibody in the 18 autoantibody-positive patients in Group I reacted only with the panel monocytes and concordant VECs. In no instance,wasreactivity to the alldgeneic lymphocytes detected, nor did the sera react only with the monocytes. The sera from the patients with vasculitis exhibited various patterns of reactivity to the panel VECs. For example, patient sera labeled CL86, PR986, and PR287 were cytotoxic to VEC # 1027,891, 973, and 945, and negative to the rest of the panel VECs (Table V). These three sera reacted similarly and therefore may have been recognizing the same antigenic epitope on the endothelial cell. Five distinct “clusters” of reactivity were detected (Table V). The panel results indicate that the antibody to autologous monocytes was an anti-VEC antibody and not antiHLA, asit was never cytotoxic to lymphocytes or neutrophils. In contrast to the patients with systemic vas-
VEC ANTIGEN
SYSTEM IN SYSTEMIC
VASCIJLITIS
,’ BRASILE ET AL
TABLE V Cluster Analysis of Vasculitic Patient Sera to a Panel of VEC Cell Code Number 1,027 891 973 945
CL86
PR986
+
Cluster 1 +
+’ -
z +
PR287
961 947 1,021 1,004 996 1,007 1,026
BK485
+ z+
BK186
Patient Sera Code Number BM LY CN H1286
H191878
ER
ML
HL
Cluster 4 + +
Cluster
+ + + +
2
•i+ T
z + -
Cluster + + -
:
3 + + -
+ +
+
+
+
+
+
;
+ -
+ ---
+
+
BK7986
WR
Cluster 5 +
-
+
+
+
+
Sera: CL86. PR 986 and PR 287 representing
three different patients formed a cluster of reactivity. in that all three sera were cytotoxic to VEC numbers 1027.891, 973, and probably 945. Likewise, a second cluster of reactivity involved sera BK485. BK186. and BM. which reacted with the VEC numbers 961, 1021, and 1004. The third cluster involves sera LY. CN, and H1286 and reacted with VEC numbers 996.1007, and probably961. The fourth cluster involved sera ML and HL which reacted with VC numbers 1027 and 996. The fifth cluster includes sera BK 7986 and WR which reacted with VEC numbers 961.11026, and probably 891, These data suggest that the autoantibody was recognizing multiple distinct allotypes within the VEC/monocyte antigen system.
culitis, autoantibody was detected in only one of four patients in Group II (Table VI).
TABLEVI Frequencyof Autoantibody in Control Patients
Autoantibody Screening in Healthy Control Subjects The autoantibody was found in low Frequency in the control populations (Table VI). Autoantibody to VEC was not detected in 51 young controls ranging from six months to 24 years of age and was detected in eight of 61 (13%) of the controls greater than 50 years of age. Control Subjects with Connective Tissue Diseases The presence of autoantibody to VEC was found to be associatedwith other known connective tissue diseaseswith low frequency (Table VI). It was observed in one of 10 (10%) patients with SLE, three of 12 (25%) patients with rheumatoid arthritis, and zero of 10 patients with diffuse systemic sclerosis. Sequential Testing Using historical sera, additional screeningsfor autoantibody were performed on three patients. In Patient A, the development of autoantibody coincided with clinical manifestations of the disease(Figure 1). Using historical frozen sera, this patient was retrospectively found to be autoantibody-negative prior to the initial diagnosis of her disease.The autoantibody was present in high titer at the time when the patient was symptomatic. Patient B was first screenedat the time of diagnosisand exhibited high titers of autoantibody (Figure 2). The autoantibody disappeared and the patient exhibited marked clinical improvement after institution of a regimen of cyclophosphamide and prednisone. In Patient C, autoantibody titers to VEC were strongly positive at the time of diseaseonset and later reverted to negative during a prolonged period of clinical remission (Figure 3). Blood VesselSpecificity Sera from four patients with systemic vasculitis (Group I) were screened against a panel of frdzen cadaveric vesselsections from different anatomic sites in order to further confirm the cell specificity of the autoantibody (Table VII). Two patients (Wegener’s granulomatosis) were classified as strongly positive for
Number of Patients* Young controls, 25 years/age Older controls, 50 years/age
O/51 (0)
8/61(13)
Controls with connective tissue disease Rheumatoid arthritis SLE Scleroderma * Values
in parentheses
are percents.
autoantibody, one patient (temporal arteritis) was positive, and the fourth patient (mixed cryoglobulinemia) was negative for autoantibody to VEC monocyte antigens by cytotoxicity technique. The sera from the three positive autoreactive patients (Patients 1 through 3) bound to sections of vena cava, whereasthe sera from the fourth nonreactive patient did not. Interestingly, no sera (regardless of their cytotoxicity reactions to autologous monocytes and allogeneic VECs) bound to allogeneic aorta.
COMMENTS The VECs express tissue-specific antigens that can function asa potent immunogen in organ transplantation [6,19,20]. The peripheral blood monocyte also expressessome VEC antigens in. vitro. In random sera containing anti-monocyte antibody, VECs and monocytes exhibited equivalent reactivity in 91%of the concordant pairs tested [16]. There was no detectable cytotoxicity directed against concordant panel T or B lymphocytes or granulocytes [21]. The results of family segregation studies indicate that the genes coding for the VEC/monocyte antigen loci are in linkage with the major histocompatibility complex and may be a newly identified HLA antigen [22,23]. The observation that the monocyte, but no other resting cell types yet studied, can serve as an alternative target cell for the VEC-specific antibody provided an opportunity to perform noninvasive screenings for autoantibody to July
1989 The American
Journal
of Medicine
Volume
87
77
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULlilS
/ BRASILE ET AL
strong Pos it ive Y P z
Pos it ive
g $ g 2
Figure 1. Patient A (Patient 9 in Tables I and II). Relation of autoantibody response to VEC and clinical disease activity. Retrospective historical sera through January 1986. In June 1983 and March 1984; the patient complained of fatigue. In May 1984, the patient developed arthralgias, which persisted until June 1984. In January 1986, the patient developed vasculitic cutaneous nodules on the feet, which persisted until March 1986.
Weak Positive
Negative
marked disease
.
L
3184
6/83
5104
6104
l/86
3/86
strong H
Positive
r” 5
5 5
improvement PO.5 it ive
Weak Positive
1
Negative
5
strong Positive
I
marked disease 1
3105
4106
5186
.
E $ E 2 2
Weak Positive Positive Negative
t-remission+ :
T marked disease
.
1 2/a
78
July 1989
Figure 2. Patient B (Patient 3 in Tables i and II). Relation of autoantibody response to VEC and cljnlcal disease activity. Patient with an established diagnosis of Wegener,‘s granulomat,osis complained of arthralgias and nasal stuffiness in March 1985. In April 1986, there was a recurrence of hemoptysis with pulmonary infiltrates, cutaneous vasculitis, arthritis, microhematuria, and anemia. Treatment with cyclophosphamide 150 mg/day was resumed, and in May 1986, the pulmonary abnormalities, arthritis, and cutaneous vasculitis had cleared.
The American
Journal
I
9184
510 I
of Medicine
Volume
87
, 5185
4106
Fleure 3. Patient C (Patient 10 in Tables I &d II). Relation of autoaritibody response to VEC ahd clinical disease activity. Patient presented with arthralgias, microhematuria, hematochezia, and hypereosinophilia in February 1981. In May 1981, there was ah acute onset of absent pulses in the lower extremities. Treatment with cyclophosphamide was begun, and disease resolved. During the period from September 1984 to April 1986, the patient’s condition was in clinical remission without cyclophosphamide therapy.
VEC ANTIGEN
VEC. This report identifies an autoantibody to VEC antigens in 86% of the patients studied with systemic vasculitis. This may be an underestimate in that serial plasmaphereses and blood transfusions may have affected our ability to detect autoantibody tq VEC in three patients with systemic vasculitis in whom this antibody was not observed. The autoantibody was not present in healthy, normal, young control subjects and was uncommon in an older control population and in patients with other connective tissue diseases. Although autologous monocytes were used for the initial detection of autoantibody, the antibody is clearly directed against VEC antigens. In all patients, the autoantibody detected initially with autologous monocytes also reacted to the concordant monocytes and VECs in the panel analysis. In no instances were panel monocytes positive and the concordant VECs negative. The observation that concordant T and B lymphocytes and neutrophils do not react with the autoantibody indicates that the reaction is not directed against class I or II HLA antigens. The sera formed “clusters” of reactivity to the panel VECs, indicating the existence of at least five allotypes within the VEC monocyte antigen system (Table V). The patient sera positive for autoantibody exhibited reactivity to the panel VECs that ranged from 20% to 67%. The fact that sera from different patients formed clusters with distinctly different patterns of reactivity strongly supports the premise that multiple distinct allotypes are involved within the VEC system. The specificity of the autoantibody to VEC is further demonstrated by the crossmatching of sera from autoantibody-positive patients with vasculitis with the endothelial lining of cadaveric blood vessels. When this was performed, we observed that the autoantibody bound to specific anatomic vascular sites. That is, the autoantibody reacted weakly, or not at all, to the aorta, superior mesenteric artery, or iliac veiri, but reacted strongly and uniformly to inferior mesenteric and splenic arteries. The cytotoxicity method used for detection of autoantibody to VEC is a less sensitive technique than the biotin-avidin immunoperoxidase technique used in the staining of the cadaveric blood vessels. The detection of antibody to these blood vessels in the one patient who did not have autoantibody to monocytes (Patient 16, Tables II and VII) could be due to subthreshold antibody levels for detection by the cytotoxicity technique. These different serologic patterns cannot be explained by anti-HLA antibody, as the.sera either initially contained no anti-HLA antibody by panel analysis or were sequentially absorbed until T- and B-cell reactivity was eliminated prior to the panel screening. These results suggest that the different patterns of disease that are recognized clinically as specific vasculitic syndromes might be due to a varied antigenic expression on the surface of VECs within the vascular system. The VEC system described herein links an autoantibody directly to vascular endothelium for .the first time in patients with systemic vasculitis. Although an anti-endothelial cell antibody has been described in patients with SLE [24], it is characterized by reactivity with both B lymphocytes and monocytes and is found in patients with active disease without vasculitis. The SLE autoantibody thus appears to be serologically distinct from the VEC autoantibody. Anti-neutrophil cytoplasmic antibodies (PMN-
SYSTEM
IN SYSTEMIC
VASCULITIS
/ BRASILE ET AL
TABLE VII Patterns of Reactivity to Abdominal Blood Vessels of Sera from Patients with Systemic Vasculitis
I
l
Vessels
3
Inferior mesenteric artery Superior mesenteric artery Splenic artery Iliac artery Aorta Inferior mesenteric vein Iliac vein Vena cava
+ + +
See Tables
I and II for diagnosis
Patient Number* 2 12 +;+ + -
+ and clinical
16
:
+ + -
+ +/-
‘+ +/-
+ -
presentation.
ACPA) have been described by several investigators in association with Wegener’sgranulomatosis and microscopic polyarteritis nodosa [25-301. Speculation on the mechanism of vascular injury in these conditions has centered around endothelial damAge caused by the releaseof lytic enzymes and toxic oxygen speciesfrom autoantibody-activated neutrophils [29]. As others have pointed out, the exact pathobiologic implication of the PMN-ACPA system is highly speculative [31]. It is possible that the necrotizing lesions of Wegener’s granulomatosis precede the development of PMNACPA, and the antibodies may actually be a marker for the granulomatous disease process. The clinical lesionsthat characterize this disorder are undoubtedly associated with humoral and cellular interactions involving other immunologic systems f31). In summary, we have identified the presence of an autoantibody in 86%of patients with systemic vasculitis. This autoantibody is highly cytotoxic, complement-fixing, and exhibits spetificity for VEC. In three patients studied using historical sera (Figures 1 through 3), the presence and titer of the autoantibody appear to correlate with the clinical course of disease. Sera from patients with vasculitis exhibited distinct patterns of reactivity with panel endothelial cells, suggesting that allotypes of the VEC antigens exist. The sera also react in a non-uniform manner with different distinct anatomic vascular sites from cadaveric blood vesseltissue, implying a heterogeneity within the VEC antigen system. Taken together, these results may point to a role for the VEC autoantibody-VEC antigen system in the determination of the distinct patterns of vascular reactivity in different vasculitic syndromes. These interpretations are speculative and will require further correlations of the presence and titer of autoantibody to VEC with diseaseactivity in a variety of vasculitic syndromes. In addition, further characterization of the VEC antigen itself and the exploration of the POSsibility of varied antigen expression within the vascular tree will be needed in order to better elucidate its role in the diseasepathogeneis of multiple vasculitic syndromes.
REFERENCES 1. Cotran RS: Endothelial phagocytosis. An electromicroscopy study. Exp Mel Pathol 1965; 4: 217-231. 2. Gimbrone MA Jr, Spaet TH. eds: Progress in hemostasis and thrombosis, ~013.
Orlando, Florida: Grune and Stratton.
1976.
3. Baldwin WM Ill, Claas FHJ. VanEs IA. VanRood JJ: Distribution of endothelialmonocyte and HlA antigenson renal vascular endothelium. Transplant Proc 1981; 13: 103-106.
July 1989
The American
Journal
of Medicine
Volume
87
79
VEC ANTIGEN
SYSTEM
IN SYSTEMIC
VASCULITIS./
BRASILE
ET AL
4. Williams GM, TerHarr A, Packs LC. et a/: Endothelial changes associated with hyperacute, acute, and chronic renal allograft rejection in man. Transplant Proc 1973; 5: 819-822. 5. Jaffe EA. Nachman RL. Becker CG, Minicki CR: Culture of human endothelial cells derived from umbilical veins. J Clin Invest 1973; 52: 2745-2756. 6. Moraes JR, Stasny P: A new antigen system expressed in human endothelial cells. J Clin Invest 1977; 60: 449-454. 7. Rodgers GM, Shuman MA: Prothrombin is activated on vascular endothelial cells by factor Xa and calcium. Proc Natl Acad Sci USA 1983: 80: 7001-7005. 8. Maruyama I, Salem HH. Majerus PW: Coagulation factor Va binds to human umbilical vein endothelial cells and accelerates protein C activation. J Clin Invest 1984; 74: 224-230. 9. Awbrey BJ, Hoak JC. Owen WG: Binding of human thrombin to cultured human endothelial cells. J Biol Chem 1979; 254: 4092-4095. 10. Bennet JS, Vflaire G: Exposure of platelet fibrinogen receptors by AOP and epinephrine. J Clin Invest 1979: 64: 1393-1401. ll.Thompson JS. Herbrick JM, Burns CP, etatGranulocyte antigens detected by cytotoxicity (GCY) and capillary agglutination (CAN). Transplant Proc 1978: 10: 885-888. 1ZThompson JS, Goeken NE, Brown SA. Rhoades JR: Antigens common to monocytes and endothelial cells: detection by monoclonal and allospecific antibodies. In: Progress in immunobiology: blood cell antigens and bone marrow transplantation, vol 149. New York: Alan R Liss, 1984; 169-175. 13. Cerilli J. Brasile L: Endothelial cell alloantigens. Transplant Proc 1988: 12: 3742. 14. NIAID Manual of Tissue Typing Techniques 1979-1980, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda. Maryland. 15. Cerilli J, Bay W, Brasile L: The significance of the monocyte crossmatch in recipients of living related HLA identical kidney grafts. Human lmmunol 1983; 7:45-50. 16. Brasile L, Galouzis T. Clarke J. Cerilli J: The 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. 17. Brasile L, Rodman E, Shield F Ill, Clarke J. Cerilli J: The association of antivascular endothelial cell antibody with hyperacute rejection: a case report. Surgery 1986; 99: 637639. 18. Cerilli J. Jesseph JE, Miller AC: The significance of antivascular endothelium
80
July
1989
The American
Journal
of Medicine
Volume
87
antibody in renal transplantation. Surg Gynecol Obstet 1972; 135: 246-249. 19. Paul CC, Class FHJ, VanEs LA: Accelerated rejections of a renal allograft associated with pretransplantation antibodies directed against donor antigens on endothelium and monocytes. N Engl J Med 1979; 300: 1258-1260. 20. Cerilli J. Brasile L. Galouzis T. DeFrancis B: Clinical significance of antimonocyte antibody in kidney transplant recipients. Transplantation 1981; 32: 495-497. 21. Clarke JL. Brasile L, Rhoades J, Thompson JS, Cerilli J: Serologic comparison of anti-granulocyte and anti-VEC cytotoxicity of human myelomonocytic antigen system (HMMA) and vascular endothelial cell system (VEC). Transplant Proc 1989; 21: 647-648. 22. Brasile L, Clarke J, Cerilli J: A probable mechanism for allograft rejection in HLA-identical combinations. Transplant Proc 1987; 19: 894-895. 23. Neppert J. Mueller-Eckhardt G, Stranzenbach W. Gottsche B. Mueller-Eckhardt C: Endothelial monocyte antigens are coded for by loci centomeric of H&B. VOX Sang 1985; 49: 221-229. 24. Cines DB, Lyss AP. Reeber M. Bina M, DeHoratias RJ: Presence of complement-fixing antiendothelial cell antibodies in systemic lupus erythematosus. J Clin Invest 1984; 73: 611-625. 25. Van der Woude FJ. Rasmussen N, Lobatto S, et al: Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet 1985; I: 425-429. 26. Gross WL. Ludemann G, Kiefer G. Lehmann H: Anticytoplasmic antibodies in Wegener’s granulomatosis (letter). Lancet 1986; I: 806. 27. Ludemann G. Gross WL: Autoantibodies against cytoplasmic structures of neutrophil granulocytes in Wegener’s granulomatosis. Clin Exp lmmunol 1987; 69: 350-357. 28. Savage COS. Winearls CG. Jones S, Marshall PD, Lockwood CM: Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vasculitis. Lancet 1987; I: 1389-1393. 29. Falk RJ. Jennette JC: Anti-neutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and crescentic glomerulonephritis. N Engl J Med 1988; 318: 16511657. 30. Specks U. Wheatley CL, McDonald TJ. Rohrbach MS, DeRemee RA: Anticytoplasmic autoantibodies in the diagnosis and follow-up of Wegener’s granulomatosis. Mayo Clin Proc 1989; 64: 28-36. 31. Cross CE. Liilington GA: Serodiagnosis of Wegener’s granulomatosis: pathobiologic and clinical implications. Mayo Clin Proc 1989: 64: 119-122.