AIDS-Associated Kaposi's Sarcoma - NCBI

2 downloads 95 Views 2MB Size Report
Oct 22, 1985 - Hairy cell leukemia cells. E-rosette- negative cells. 92,000. Vascular endothelium of all tissues including kidney, renal glomerular capil-.
Expression of Endothelial Cell AIDS-Associated Kaposi's Sarcoma The

Surface

Antigens

by

Evidence for a Vascular Endothelial Cell Origin

JOANNE L. RUTGERS, MD, ROSEMARY WIECZOREK, MD, FRANCO BONETTI, MD, KAREN L. KAPLAN, MD, DAVID N. POSNETT, MD, ALVIN E. FRIEDMAN-KIEN, MD, and DANIEL M. KNOWLES II, MD

From the Immunopathology Laboratory, Department of Pathology, the Kaplan Cancer Center, the Department of Dermatology, New York University School of Medicine; the Department of Medicine, Columbia University College of Physicians and Surgeons; and the Department of Medicine, Cornell University Medical College, New York, New York

The authors investigated 19 cases of Kaposi's sarcoma (KS) obtained from patients with the acquired immune deficiency syndrome (AIDS) for their expression of Factor VIII-related antigen (FVIIIRAg), HLA-DR (Ia) antigens, OKM1, and three distinctive vascular, but not lymphatic, endothelial-cell-associated antigens, E92, OKM5, and HCL. Antigen expression was demonstrated by immunoperoxidase staining of cryostat sections. FVIIIRAg is strongly expressed by the cells lining the vascular spaces (VCs) but is absent, weakly or focally, and variably expressed by the spindle cell (SC) component of KS. The VC component of each KS lesion examined strongly expressed E92, moderately expressed HCl, and weakly ex-

pressed OKM5. In contrast, the entire SC component of each KS lesion studied strongly expressed E92 and OKM5 and weakly expressed HCL. Neither the VCs nor the SCs expressed OKML. These studies provide strong and compelling evidence for the vascular endothelial cell histogenesis of both the vascular and spindle cell components of KS, demonstrate the intertumor and intratumor phenotypic heterogeneity of KS, and suggest that monoclonal antibodies OKM5 and anti-E92 are the best currently available immunohistochemical markers for identifying the spindle cell component of AIDS-associated KS in cryostat sections. (Am J Pathol 1986, 122:493-499)

THE HISTOGENESIS and biologic significance of Kaposi's sarcoma (KS) has remained controversial for many years. Many investigators have utilized Factor VIII-related antigen (FVIIIRAg), which is synthesized by endothelial cells' and megakaryocytes,2 as a marker for vascular endothelium,3 and for KS lesions,4-10 and have suggested a vascular endothelial cell origin for KS.4",'7,9"10 However, immunohistochemical staining for FVIIIRAg in KS has led to conflicting results. The consensus of previous studies is that while the cells lining the vascular spaces of KS consistently express FVIIIRAg, the spindle cells of KS exhibit absent6'8 or variable expression4,5,'7910 of FVIIIRAg, leaving unanswered the question of their histogenesis. Moreover, it is obvious that FVIIRAg is not a reliable and reproducible immunohistochemical marker for the spindle cell component of KS, and no other marker has emerged from multiple prior studies of KS. These conflicting results, combined with earlier observations11 and recent

histochemical and immunohistochemical studies,12 have led to the proposal that KS is derived from lymphatic

493

endothelium.12,13 Our laboratory has participated in the description

of monoclonal antibodies which detect three distinctive vascular endothelial-cell-associated antigens, E92,14 OKM5, 5I17 and HCl.18" 9 In the studies described here,

Supported in part by Grants EY03357 (D.M.K.), HL15486 (K.L.K.), HL21006 (K.L.K.), CA35463 (D.N.P.) and CA35982 (A.F.K.) from the National Institutes of Health and by grants from the Cancer Research Institute, the New York State AIDS Institute, and the Bernard and Frances Laterman Project Chai Philanthropic Trust (D.M.K.), and the Howard Gilman Foundation (A.EK.). Dr. Karen L. Kaplan is the Molly Berns Established Fellow of the New York Heart Association. Accepted for publication October 22, 1985. Address reprint requests to Dr. Daniel M. Knowles, New York University Medical Center, University Hospital, Dept of Pathology, 550 First Avenue, New York, NY 10016.

494

RUTGERS ET AL

AJP * March 1986

we examined KS lesions for their expression of these vascular endothelial-cell-associated antigens in order to help delineate the histogenesis and better define the phenotypic characteristics of KS, and also to determine which, if any, of these markers is the best immunodiagnostic reagent currently available for the routine immunohistochemical identification of the spindle cell component of KS. Reactivity with these three monoclonal antibodies was compared with that of a heteroantiserum and a monoclonal antibody directed against FVIIIRAg, monoclonal antibody OKM1, and monoclonal anti-HLA-DR (Ta) in 19 cases of KS occurring in patients with the acquired immunodeficiency syndrome (AIDS).

Materials and Methods Patient Specimens Cutaneous punch biopsies (14), excisional lymph node biopsies (4), and an eyelid biopsy (1) were performed on 19 patients with AIDS after obtaining informed consent in order to obtain histologic documen-

tation of Kaposi's sarcoma. Each biopsy was divided into equal halves, one half fixed in 10% buffered formalin and routinely processed for histopathologic sectioning, and the remaining half snap-frozen in a mixture of isopentane and dry ice and stored at - 80 C. Heteroantisera and Monoclonal Antibodies The heteroantisera and monoclonal antibodies used in this study have been previously characterized.3'14-22 Their principal characteristics and distribution are summarized in Table 1.

Immunohistochemical Staining Cryostat sections were cut from each tissue block, fixed, and prepared for immunohistochemical staining as previously described.23 The sections were incubated with 40-90 M1 of each of the primary antibodies at previously determined appropriate dilutions in a moist chamber for 60 minutes at room temperature. The second step, for all antibodies except rabbit anti-human FVIIIRAg, was a 30-minute room temperature incu-

Table 1 -Heteroantisera and Monoclonal Antibodies Used in This Study Target antigen Endothelial cell (mol wt) distribution Antibody Isotype Immunogen Anti-FVIIIRAg

Heteroantiserum

FVIIIRAg

200,000

Anti-FVIIIRAg

IgG

FVIIIRAg

High molecular weight

Anti-HLA-DR

IgG1

B35M B-cell line

29,000 34,000

OKM5

IgG1

E-rosettenegative cells

88,000

CIV(anti-E92)

IgG1

Umbilical vein endothelial cells

92,000

Anti-HCI

IgG2a

Hairy cell leukemia cells

OKM1

IgG2b

E-rosettenegative cells

Glycolipid

170,000

Other distribution

Source

References

Vascular endothelium of all organs except renal glomerular capillaries Vascular endothelium of all organs except renal glomerular capillaries Vascular endothelium

Megakaryocytes; platelets

DAKO

Megakaryocytes

DAKO

B cells; monocytes; activated T cells

Dr. Wang

20,21

Vascular endothelium of all organs including splenic sinusoidal lining cells but excluding renal glomerular capillaries and renal cortical vasculature Vascular endothelium of all tissues including kidney, renal glomerular capillaries & splenic sinusoidal lining cells Vascular endothelium of all tissues, including kidney and renal glomerular capillaries, but excluding brain and splenic sinusoidal lining cells None

Monocytes; platelets

Ortho Pharmaceutical

15-17

Small proportion of monocytes; fibroblasts

Privately generated

14

Hairy cell leukemia; basal

Privately generated

18,19

Ortho Pharmaceutical

15,22

3

epidermal cells

Monocytes; granulocytes; null cells

KAPOSI'S SARCOMA

Vol. 122 * No. 3

495

Table 2-Results of Immunohistochemical Staining of 19 Cases of Kaposi's Sarcoma With a Panel of Heteroantisera and Monoclonal Antibodies Proportion of cells comprising cases of KS which are reactive with antibodies

Type of

Case 1

2 3 4

5

6 7

8 9

10 11 12

13 14

15

16 17

18 19

Anti-

Rabbit Anti-

FVIIIRAg

FVIIIRAg

Anti-la

Anti-E92 VC SC

OKM5

Anti-HCI VC SC

OKM1 VC SC

lesion

VC

Skin plaque Skin plaque Skin plaque Skin plaque Skin plaque Skin nodule Skin nodule Skin nodule Skin nodule Skin nodule Skin nodule Skin nodule Skin nodule Skin nodule Eyelid nodule Lymph node, diffuse Lymph node, focal Lymph node, focal Lymph node, diffuse

+ +

-

+ +

+

+++

+

+++

++

+

+ + +

+++

+

+

+

+

-

+

+ + +

+

++

++

+ + +

+

+

++

+

++

++

++ +

+ ++

+ ++

++

++

++

++

+

++

++

++

++

++ +

+++

++

++

+

+ + +

+

++

+

+ + +

++

++

++

+

+ ++

++

+ + +

+

++ +

++

++ +

+++

+++

++

+

+ + +

++

+

-

-

+

++

++

++

++

+ + +

+++

+++

+

+++

+

+

-

-

++

+

+

+

+ ++

+ ++

+++

+++

-

+++

+

+

-

-

+

+

+++

+ ++

+

+++

++

+++

+

+++

+

++

-

-

++

+

+ ++

+

+++

++

+++

++

++

+++

+ ++

+

-

-

NA

+

NA

+++

NA

++

NA

+++

NA

+++

NA

+

NA

+

+

+

+++

++

+

++

ND

+

+++

++

+

-

+

+++

+

+++

++

+

+++

ND

++

++ +

+++

++

-

+++

+

+++

++

++

++

ND

+

+++

+++

++

-

++

++

++

+++

+

++

+++

+++

+

+++

+

+

-

++

+

++

+

+

++

++ +

+++

+

+++

+

+

-

+ ++

-

+++

+

+

+++

+++

++

+

+++

++

++

-

+ ++

-

+++

+

++

++

ND

++

++

++

+++

++

+++

++

+++

++

ND

++

+++

+

SC

VC

SC

VC

SC

ND

VC

SC

+

+

VC, vascular lining cells; SC, spindle cells; ND, not done; NA, not applicable. -, no or only rare cells positive; +, 50% of cells positive.

bation with KPL, a peroxidase-labeled goat anti-mouse IgG and IgM, diluted 1:40 (Kirkegaard & Perry Lab., Inc., Gaithersburg, Md). The second step for rabbit antiFVIIIRAg was a 30-minute room temperature incubation with a swine anti-rabbit IgG (Dako, Denmark) followed by a PAP (soluble complex of horseradish peroxidase and rabbit anti-horseradish peroxidase) incubation of 30 minutes. Negative controls included sections incubated with phosphate-buffered saline (PBS) alone, sections incubated with ascites obtained from mice injected with nonsecretory hybridoma cells, and sections incubated with KPL or PAP alone. The final brown reaction product was produced by incubation with 3,3'-diaminobenzidine (Sigma Chemical Co., St.

Louis, Mo) and H202 as previously described.23 All steps were separated by 5 minute room temperature washes in PBS. The slides were counterstained with hematoxylin, dehydrated, mounted in Permount, and examined with an American Optical light microscope. Results

Histopathology In each case, the tumor exhibited the typical histopathologic features characteristic of KS. Biopsies of five cutaneous KS lesions showed dermal tumors containing a variable mixture of spindle cells and irregular vas-

496

RUTGERS ET AL

cular spaces linked by bland-appearing endothelial-like cells, typical of the plaque stage of KS.24 25 Biopsies of the one eyelid and nine cutaneous KS lesions demonstrated predominantly intertwining fascicles of spindle cells commingled with extravasated erythrocytes and only occasional vascular spaces, consistent with the nodule stage of KS. Two lymph nodes were almost entirely replaced by a spindle cell proliferation with intermingled erythrocytes, whereas the other two lymph nodes exhibited focal involvement by both vascular and spindle cell components of KS.

Immunohistochemistry Each of the 19 KS lesions exhibited similar immunohistochemical staining patterns with differential staining of the cells lining the vascular spaces (VCs) and the spindle cells (SCs) (Table 2, Figure 1). In the majority of the lesions, the VCs exhibited consistent, uniform moderate to strong expression, whereas the SCs exhibited absent, weak, or focal and variable expression of FVIIIRAg consistent with the observations of other investigators.4'-I"2 The VC and the SC components also exhibited differential expression of the vascular endothelialcell-associated antigens E92, OKM5, and HCL. The VC component of each lesion tested reacted strongly with monoclonal anti-E92, slightly more weakly and more variably with monoclonal anti-HCl, and weakly with OKM5. In contrast, the entire SC component comprising each of the 19 lesions reacted strongly with OKM5 and with anti-E92 and weakly with anti-HCl. A subpopulation of the SCs, and to a slightly lesser extent, the VCs, expressed HLA-DR (Ia) antigens. Monoclonal antibody OKMl was essentially unreactive or only minimally reactive with both the VCs and the SCs of KS, in agreement with a previous report.7 Thus, in most cases of KS studied here, the majority of the cells lining the vascular spaces (VCs) strongly expressed E92 and FVIIIRAg, a slightly lesser number expressed HLA-DR and HCl, and a smaller population weakly expressed OKM5. The majority of the SCs strongly expressed OKM5, E92, and HLA-DR; a slightly smaller proportion expressed FVIIIRAg, and a minority weakly expressed HCl. An unexpected finding was that monoclonal antibody OKM5 reacted strongly with the keratinocytes immediately overlying each cutaneous KS lesion but not the adjacent keratinocytes. OKM5 was previously shown to be unreactive with normal epidermis (17). The keratinocytes did not express HLA-DR, OKMl, E92, HCl or FVIIIRAg.

AJP

*

March 1986

Discussion A plethora of electron-microscopic and histochemical studies have led investigators to propose various origins for the SC component typical of the plaque and nodule stage of KS,8'26-31 including current proposals that they are derived from lymphatic endothelium. I2 I3 The latter proposal is based upon early histochemical observations11 and a recent suggestion that the SCs of KS more closely exhibit phenotypic characteristics typical of normal lymphatic, rather than normal vascular,

endothelium.'2 The studies described here demonstrated that most SCs in each case of KS examined reacted strongly with monoclonal antibodies OKM5 and anti-E92. Both antibodies have been previously shown to detect distinctive antigens on the majority of vascular, but not lymphatic, endothelium in all human tissues and organs studied, including splenic sinusoidal lining cells.14'17 These studies further demonstrate that a variable proportion of the SCs of KS express HC1, an additional distinctive vascular endothelial-cell-associated antigen. Although we cannot unequivocally rule out the possibility that E92, OKM5, and HCO are not expressed by any lymphatic endothelial cells, prior studies with multiple samples from a diverse group of human tissues have failed to disclose an obvious expression of these antigens by lymphatic endothelium. 14"7"19 The cells lining dermal lymphatics and nodal lymphatic sinuses conspicuously lack E92, OKM5, and HCL. Additionally, our studies demonstrate that the SCs of KS consistently express HLA-DR antigens, as does normal vascular endothelium,21 in contrast with the results of Beckstead et al.12 This discrepancy may be related to technique or choice of monoclonal anti-HLA-DR. The monoclonal anti-HLA-DR used here detects the framework determinants of the p29,34 bimolecular HLA-DR complex and is comparable in reactivity to a well-known anti-Ia heteroantiserum.32'33 Together, these results provide the strongest and most compelling evidence thus far for the vascular endothelial cell origin of the SC component of KS. The intertumor and intratumor heterogeneity of endothelial cell surface antigen expression by KS lesions may reflect the heterogeneity inherent in normal vascular endothelium previously demonstrated by us14"17"19 using the monoclonal antibodies employed in the present study. For example, OKM5 reacts with the majority of normal vascular endothelium in all tissues and organs examined, including the splenic sinusoidal lining cells, but not with glomerular capillary endothelium, and exhibits differential reactivity with renal vascular endothelium.17 In contrast, monoclonal anti-E92 and

,;

Vol. 122 * No. 3

AO A

nk.

t .......

-e:W.:

..l

i:.r.

t..... -''ss;. -KX.

.mIN

-It

Mlu:

A.

swlE,;;Jh

i >r'}

9*~ rM-A . .

~~5

-.1 X * _3

W

/

..dvY

.l. :. '

':

E~~~~ ~~~~~~

~

(p..

L~~,,,

497

>~~~~~~~~~~~~~~~~~~~al

.11.14

.

KAPOSI'S SARCOMA

~~~~~~~~~~~

.......

'i;...,.

S,1

SE

A 'Ia t

*

*6

.....

j.;

t%,

~~~~ ~~~~~~~x:a

g

F

.~~~~.4.. MI

5,V.1 A2

.

Figure 1- Immunoperoxidase staining of cryostat sections of nodular lesions of KS with rabbit anti-human FVIIIRAg (a), monoclonal anti-E92 (b), monoclonal antibody OKM5 (c), monoclonal anti-HCI (d), monoclonal anti-HLA-DR (la) antigens (e), and monoclonal antibody OKM1 (f). FVIIIRAg is strongly expressed by the cells lining the vascular spaces but is absent or only minimally expressed by the spindle cells component Both the SCs and the VCs strongly express E92. The SCs but not the VCs strongly express OKM5. The SCs and the VCs weakly express HCI.of AKS.subset of SCs and VCs express HLA-DR (la) antigens. Neither the VCs nor the SCs express OKM1.

..

498

AJP * March 1986

RUTGERS ET AL

anti-HCl react with the majority of normal vascular endothelium, including renal vascular and glomerular capillary endothelium. '4' 9 Anti-E92 reacts with splenic sinusoidal lining cells, but anti-HCl does not."4'-9 The heterogeneity of vascular endothelial cell expression may reflect states of differentiation or activation, or subsets which subserve distinct functions. The results of the present study suggest that the SCs of KS arise in an OKM5', E92' endothelial cell population which heterogeneously expresses HLA-DR, HCl, and FVIIIRAg. The consistent and strong reactivity of monoclonal antibodies OKM5 and anti-E92 with the majority of the SCs in every case of KS examined here suggests that they may be the best immunodiagnostic reagents currently available for the immunohistochemical identification of the SC component of AIDS-associated KS in cryostat sections. Both antibodies are far superior to rabbit anti-human FVIIIRAg, monoclonal antiFVIIIRAg, and anti-HCl, all of which exhibit considerable intertumor and intratumor heterogeneity of reactivity. Monoclonal antibodies E92 and OKM5 appear to be the reagents of choice for the immunodiagnostic identification of neoplastic proliferations of predominantly SCs of KS or possibly other neoplasms derived from vascular endothelium. The extent to which these monoclonal antibodies are useful in the identification of poorly differentiated and anaplastic angiosarcomas remains to be determined. It is well known that vascular endothelial cells and antigen-presenting cells share many properties. Endothelial cells express HLA-DR antigens,2' can be induced to express Fc and complement receptors34 and are capable of replacing monocytes in soluble antigeninduced T-cell proliferation35 and of stimulating mixed lymphocyte reactions.35 Patients with AIDS and/or KS are known to exhibit depressed proliferative responses in both autologous and allogeneic mixed lymphocyte reactions.36138 Therefore, it is of considerable interest and potential biologic significance that patients with AIDS and with other immunodeficiency states and profound immunologic dysfunction should develop a vascular endothelial cell neoplasm which exhibits the OKMl, OKM5+ phenotype. The OKMl, OKM5+ peripheral blood monocyte subset is capable of triggering allogeneic T-cell proliferation and is specifically responsible for triggering the autologous mixed lymphocyte reaction.16 The function of the OKM5 molecule in this and in other biologic systems is still unclear. In conclusion, the studies presented here demonstrate that both the cells lining the vascular spaces and the spindle cells of KS heterogeneously express three distinctive vascular endothelial cell associated antigens, E92, OKM5, and HCL. These studies provide strong evidence for the vascular endothelial cell histogenesis of

both the vascular and spindle cell components of KS, demonstrate the intertumor and intratumor phenotypic heterogeneity of KS, and suggest that monoclonal antibodies OKM5 and anti-E92 are the best currently available immunodiagnostic markers for the immunohistochemical identification of the SC component of AIDS-associated KS in cryostat sections.

References 1. Jaffe EA, Hoyer LW, Nachman RL: Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J Clin Invest 1973, 52:2757-2764 2. Jaffe EA: Endothelial cells and the biology of factor VIII. N Engl J Med 1977, 296:377-383 3. Mukai K, Rosai J, Burgdorf WHC: Localization of factor VIII-related antigen in vascular endothelial cells using an immunoperoxidase method. Am J Surg Pathol 1980, 4:273-276 4. Guarda LG, Silva EG, Ordonez NG, Smith JL: Factor VIII in Kaposi's sarcoma. Am J Clin Pathol 1981, 76:197-200 5. Nadji M, Morales AR, Ziegles-Weissman J, Penneys NS: Kaposi's sarcoma: Immunohistologic evidence for an endothelial origin. Arch Pathol Lab Med 1981, 105:274-275 6. Burgdorf WHC, Mukai K, Rosai J: Immunohistochemical identification of factor VIII-related antigen in endothelial cells of cutaneous lesions of alleged vascular nature. Am J Clin Pathol 1981, 75:167-171 7. Modlin RL, Hofman FM, Kempf RA, Taylor CR, Conant MA, Rea TH: Kaposi's sarcoma in homosexual men: An immunohistochemical study. J Am Acad Dermatol 1983, 8:620-627 8. Akhtar M, Bunuan H, Ali MA, Godwin JT: Kaposi's sarcoma in renal transplant recipients: Ultrastructural and immunoperoxidase study of four cases. Cancer 1984, 53:258-266 9. Flotte TJ, Hatcher VA, Friedman-Kein AE: Factor VIIIrelated antigen in Kaposi's sarcoma in young homosexual men. Arch Dermatol 1984, 120:180-182 10. Ordonez NG, Batsakis JG: Comparison of Ulex europaeus I lectin and factor VIII-related antigen in vascular lesions. Arch Pathol Lab Med 1984, 108:129-132 11. Dorfman RF: Kaposi's sarcoma, the contribution of enzyme histochemistry to the identification of cell types. Acta Unio Int Contra Cancrum 1962, 18:464-467 12. Beckstead JH, Wood GS, Fletcher V: Evidence for the origin of Kaposi's sarcoma from lymphatic endothelium. Am J Pathol 1985, 119:294-300 13. Dorfman RF: Kaposi's sarcoma revisited. Hum Pathol 1984, 15:1013-1017 14. Kaplan KL, Weber D, Cook P, Dalecki M, Rogozinski L, Sepe 0, Knowles DM II, Butler VP: Monoclonal antibodies to E92, an endothelial cell surface antigen. Arteriosclerosis 1983, 3:403-412 15. Talle MA, Rao PE, Westberg E, Allegar N, Makowski M, Mittler RS, Goldstein G: Patterns of antigenic expression on human monocytes as defined by monoclonal antibodies. Cell Immunol 1983, 78:83-99 16. Shen HH, Talle MA, Goldstein G, Chess L: Functional subsets of human monocytes defined by monoclonal antibodies: A distinct subset of monocytes contains the cells capable of inducing the autologous mixed lymphocyte culture. J Immunol 1983, 130:698-705 17. Knowles DM II, Tolidjian B, Marboe C, D'Agati V, Grimes M, Chess L: Monoclonal anti-human monocyte antibodies OKM1 and OKM5 possess distinctive tissue

Vol. 122 * No. 3

distributions including differential reactivity with vascular endothelium. J Immunol 1984, 132:2170-2173 18. Posnett DN, Chiorazzi N, Kunkel HG: Monoclonal antibodies with specificity for hairy cell leukemia cells. J Clin Invest 1982, 70:254-261 19. Posnett DN, Marboe CC, Knowles DM II, Jaffe EA, Kunkel HG: A membrane antigen (HCl) selectively present on hairy cell leukemia cells, endothelial cells, and epidermal basal cells. J Immunol 1984, 132:2700-2702 20. Winchester RF, Kunkel HG: The human Ia system. Adv Immunol 1979, 28:221-292 21. Hirschberg H, Moen T, Thorsby E: Specific destruction of human endothelial cell monolayers by anti-DRW antisera. Transplantation 1979, 28:116-120 22. Breard J, Reinherz EL, Kung PC, Goldstein G, Schlossman SF: A monoclonal antibody reactive with human peripheral blood monocytes. J Immunol 1980, 124: 1943-1948 23. Knowles DM II, Tolidjian B, Marboe CC, Halper JP, Azzo W, Wang CY: A new human B-lymphocyte surface antigen (BL2) detectable by a hybridoma monoclonal antibody: Distribution on benign and malignant lymphoid cells. Blood 1983, 62:191-199 24. Ackerman AB: Subtle clues to diagnosis by conventional microscopy: The patch stage of Kaposi's sarcoma. Am J Dermatopathol 1979, 1:165-172 25. Gottlieb GJ, Ackerman AB: Kaposi's sarcoma: An extensively disseminated form in young homosexual men. Hum Pathol 1982, 13:882-892 26. Pepler WJ, Theron JJ: An electron microscopic study of Kaposi's hemangiosarcoma. J Pathol Bacteriol 1962, 83:521-525 27. Becker BJP: The histogenesis of Kaposi's sarcoma. Acta Unio Int Contra Cancrum 1962, 18:477-486 28. Hashimoto K, Lever WF: Kaposi's sarcoma: Histochemical and electron microscopic studies. J Invest Dermatol 1964, 43:539-549 29. Niemi M, Mustakallio KK: The fine structure of the spindle cell in Kaposi's sarcoma. Acta Pathol Microbiol Scand 1965, 63:567-575

KAPOSI'S SARCOMA

499

30. Harrison AC, Kahn LB: Myogenic cells in Kaposi's sarcoma: An ultrastructural study. J. Pathol 1978, 124:157-160 31. Waldo ED, Vuletin JC, Kaye GI: The ultrastructure of vascular tumors: Additional observations and a review of the literature. Pathol Ann 1977, 12:279-308 32. Winchester RJ, Wang CY, Halper JP, Hoffman T: Studies with B cell allo- and hetero-antisera: Parallel reactivity and special properties. Scand J Immunol 1976, 5:745-757 33. Halper JP, Knowles DM II, Wang CY: Ia antigen expression by malignant lymphomas: Correlation with conventional cell markers. Blood 1980, 55:373-382 34. Ryan US, Schultz DR, Ryan JW: Fc and C3b receptors on pulmonary endothelial cells: Induction by injury. Science 1981, 214:557-558 35. Hirschberg H, Evensen SA, Henriksen T, Thorsby E: Stimulation of human lymphocytes by allogenic endothelial cells in vitro. Tissue Antigens 1974, 4:257-261 36. Gupta S, Safai B: Deficient autologous mixed lymphocyte reaction in Kaposi's sarcoma associated with deficiency of Leu3 + responder T cells. J Clin Invest 1983, 71:296-300 37. Seligmann M, Chess L, Fahey JL, Fauci AS, Lachmann PJ, L'Age-Stehr J, Ngu J, Pinching AJ, Rosen FS, Spira TJ, Wybran J: AIDS-An immunologic reevaluation. N Engl J Med 1984, 311:1286-1292 38. Friedman-Kien AE, Laubenstein LJ, Rubinstein P, Buimovici-Klein E, Marmor M, Stahl R, Spigland I, Kim KS, Zolla-Pazner S: Disseminated Kaposi's sarcoma in homosexual men. Ann Int Med 1982, 96:693-700

Acknowledgments The authors wish to acknowledge their sincere appreciation to Dr. Chang Yi Wang for the monoclonal anti-HLADR, to Dr. Gideon Goldstein for monoclonal antibodies OKMl and OKM5, to Mr. Thomas Taylor for excellent technical assistance and to Ms. Margaret Walden for help in preparation of the manuscript.