Cell Thymomas Research Articles

Immunohistochemical detection of bcl‐2 protein in thymoma

The protooncogene bcl-2 encodes a protein that inhibits apoptosis. The protein is expressed in most epithelial cells of the fetal thymic medulla but, to the best of our knowledge, no data are available on bcl-2 expression in thymoma. Expression of bcl-2 protein was analysed in 30 cases of thymoma by immunohistological staining of paraffin-embedded tissue. All cases were examined and classified according to the Salyer and Eggleston and the Müller-Hermelink classification. In four cases, the protooncogene bcl-2 was abnormally expressed in spindle cells of pure medullary thymoma, whereas the non-spindle cells in mixed and in cortical thymoma were negative. All the lymphocytes were also strongly positive in medullary thymoma while a few lymphocytes showed light staining in other thymomas.

Long-term Survival of Thymoma Patients by Histologic Pattern and Proliferative Activity

We performed DNA flow cytometry and analysis of the argyrophilic nucleolar organizer regions (AgNORs) in formalin-fixed, paraffin-embedded sections from 60 surgically resected thymomas. The results were correlated with histologic pattern, stage, associated clinical features, and survival to assess which parameters could best predict prognosis. On univariate analysis, the 10-year survival rates were 86% for predominantly lymphocytic type but only 42% for predominantly epithelial, mixed lymphoepithelial, or spindle cell thymomas (p = 0.006); survival rates were 85% for noninvasive but only 34% for invasive thymomas (p = 0.0002); 73% for diploid but only 38% for aneuploid cases (p = 0.005); 88% for thymomas with 5.75 AgNORs per cell or fewer but only 34% for thymomas with more than 5.75 AgNORs per cell (p < 0.0001). On multivariate survival analysis, tumor stage (p < 0.001) and AgNOR counts (p = 0.009) retained independent prognostic significance. The 16 patients with predominantly lymphocytic type and 5.75 AgNORs per cell or fewer were all alive at the end of the observation period. In conclusion, the histologic type of the American classification and the proliferative activity evaluated by AgNOR analysis are the best predictors of long-term survival for patients with thymoma. Both predictors can be easily evaluated in the same histologic section, are highly reproducible, and permit identification of a group of patients with a favorable outcome regardless of other clinicopathological features.

Accumulation of Immature CD3 -CD4 +CD8 - Single-Positive Cells That Lack CD69 in Epithelial Cell Tumors of the Human Thymus

Human thymomas are epithelial tumors that are associated with a large number of nonneoplastic T cells of mainly immature phenotype, suggesting that epithelial cells of thymomas retain the function of thymic cortical epithelium. We report here that a large proportion (16-85%, mean ± SD 58 ± 22%, n = 17) of CD4 + single-positive T cells within the thymoma lack cell surface expression of CD3 and also of CD69, an early activation antigen that is expressed by positively selected thymocytes. Normal thymus had this CD4 +CD8 -CD3 - population at much lower levels (12 ± 7%, n = 11). It has been reported that in the human the CD4 +CD8 -CD3 - cells are the predominant population that is intermediate between the CD4 -CD8 - double-negative and CD4 +CD8 + double-positive stages. In contrast to CD4 + single-positive cells, most of the CD8 + single-positive cells in the thymoma as well as in the normal thymus expressed CD69 and a high level of CD3. A partial explanation of these observations is that the epithelial cells of thymoma were unable to positively select all the immature thymocytes generated in the thymoma. This relative inefficiency of positive selection could not be attributed solely to the paucity of MHC class II expression in the thymoma.

A monoclonal antibody to cell surface antigen of human thymic epithelial cell.

The cell surface molecule identified by a monoclonal antibody(TE-1) to human thymic epithelial cell showed the specificity for thymic epithelial cells of both the cortex and medulla. TE-1 reacted with the epithelial cells of normal thymus and thymoma in fresh frozen tissues. The antigen recognized by TE-1 was mostly confined to the cell surface membrane and arranged in reticular network with long processes between thymocytes. On immunohistochemical analysis, TE-1 did not recognize normal epithelial cells of the uterine cervix, skin and stomach, and neoplastic cells of squamous cell carcinoma and gastric adenocarcinoma, all of which were stained with anti-cytokeratin monoclonal antibody. Among the tumor cell lines tested with flow cytometry, most of epithelial and all of hematopoietic cell origin were not labeled with TE-1. In summary, TE-1 appears to be a monoclonal antibody against a surface antigen of human thymic epithelial cell that is immunohistologically different from known epithelial cell surface antigens reported so far.

Immunoreactive thymosin α1 in human thymus and thymoma

Thymosin alpha 1-like immunoreactivity was assessed in human thymus and thymoma tissue extracts by means of a new radioimmunoassay that included an anti-thymosin alpha 1 mouse monoclonal antibody. Thymosin alpha 1-like immunoreactivity levels decreased with age in normal thymuses but not in thymomas. The average thymosin alpha 1-like immunoreactivity level was 45.0 +/- 52.1 ng/mg protein in normal thymuses and 273.9 +/- 205.0 ng/mg protein in thymomas. The average thymosin alpha 1-immunoreactivity level in thymomas was higher than that in normal thymuses. Thymosin alpha 1-like immunoreactivity levels in thymomas appeared to have no relationship to the clinical stage of the thymoma or associated diseases. When viewed according to histologic characteristics, the average thymosin alpha 1-like immunoreactivity level in polygonal cell thymomas (382.5 +/- 192.6 ng/mg protein) was significantly higher than that in the spindle cell thymoma (101.8 +/- 81.2 ng/mg protein). When viewed according to the degree of lymphocyte infiltration, thymomas could be classified according to four grades: absent, scant, moderate, and predominant. In predominant or moderate thymomas, the average thymosin alpha 1-like immunoreactivity level was higher than that in scant or absent thymomas. Also, thymosin alpha 1-like immunoreactivity levels in thymuses of patients with myasthenia gravis were relatively higher than those in patients with normal thymuses.

Cytokeratins in normal thymus and thymic epithelial tumors

Thymus consists of some distinct epithelial cells that contain different sets of cytokeratins (CK). Epithelium-derived tumors maintain the expression of some of the CK of the specific nontransformed cells. Therefore, it seems reasonable to hypothesize that thymic epithelial tumors may differentiate toward distinct subsets of intrathymic epithelial cells in terms of CK expression. Eighty-one thymomas and 14 thymic carcinomas were studied immunohistologically using monoclonal antibodies specific for a single CK or a CK pair. Thymic epithelial neoplasms could not be distinguished from each other on the basis of the profile of CK expression because the degree of overlap was extensive. However, polygonal cell thymomas differentiate toward a CK13-positive cortical subset that is rare in normal thymus. Spindle cell thymomas differentiate toward a CK13-positive medullary subset. Mixed cell thymomas are comprised of a CK13-positive medullary subset and a CK13-negative medullary subset, both of which are typical in normal thymus. CK18 was expressed to a greater extent on the epithelium of thymic carcinomas than on that of thymomas. Polygonal cell thymomas more frequently were invasive than spindle and mixed cell thymomas. There is a possibility that the epithelium of polygonal cell thymomas is immature because it is a phenotypically unusual subset in normal thymus. A thymic carcinoma arising in a thymoma has been reported, although the relationship between the thymoma and the thymic carcinoma was not clear. Nevertheless, given the similar cellular differentiation of thymoma and thymic carcinoma, CK18-positive epithelium in thymomas may be transformed into thymic carcinoma cells in certain conditions.

Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions.

We recently reported the purification of a lymphocyte granule protein called "fragmentin," which was identified as a serine protease with the ability to induce oligonucleosomal DNA fragmentation and apoptosis (Shi, L., R. P. Kraut, R. Aebersold, and A. H. Greenberg. 1992. J. Exp. Med. 175:553). We have now purified two additional proteases with fragmentin activity from lymphocyte granules. The three proteases are of two types; one has the unusual ability to cleave a tripeptide thiobenzyl ester substrate after aspartic acid, similar to murine cytotoxic cell protease I/granzyme B, while two are tryptase-like, preferentially hydrolyzing after arginine, and bear some homology to human T cell granule tryptases, granzyme 3, and Hanukah factor/granzyme A. Using tripeptide chloromethyl ketones, the pattern of inhibition of DNA fragmentation corresponded to the inhibition of peptide hydrolysis. The Asp-ase fragmentin was blocked by aspartic acid-containing tripeptide chloromethyl ketones, while the tryptase fragmentins were inhibited by arginine-containing chloromethyl ketones. The two tryptase fragmentins were slow acting and were partly suppressed by blocking proteins synthesis with cycloheximide in the YAC-1 target cell. In contrast, the Asp-ase fragmentin was fast acting and produced DNA damage in the absence of protein synthesis. Using a panel of unrelated target cells of lymphoma, thymoma, and melanoma origin, distinct patterns of sensitivity to the three fragmentins were observed. Thus, these three granule proteases make up a family of fragmentins that activate DNA fragmentation and apoptosis by acting on unique substrates in different target cells.

The distribution of epithelial membrane antigen in thymic epithelial neoplasms.

Thymic carcinomas arising within a thymoma have been reported, but the relationship between thymoma and thymic carcinoma is poorly understood. Epithelial membrane antigen (EMA) is known to be an effective marker for establishing the epithelial nature of neoplastic cells, and it is reported that staining of tumors is clearly related to the degree of tumor differentiation. Eighty-one thymomas (59 noninvasive, 22 invasive) and 14 thymic carcinomas were studied immunohistologically using antiepithelial membrane antigen (anti-EMA) monoclonal antibody. Thymic carcinomas tended to express much larger quantities of EMA than thymomas, and instances of EMA-positive thymoma were seen significantly more often in invasive thymomas than in noninvasive ones (P < 0.05). However, EMA positivity was also associated with gland-like structures, which were not necessarily associated with malignant disease. Nevertheless, in view of the concept that thymoma and thymic carcinoma show a similar cellular differentiation, EMA-positive epithelial cells in thymoma with no relation to gland-like configurations might represent a pool of cells having a latent potential for malignant disease and might be transformed into thymic carcinoma cells under certain conditions. Immunolabeling for EMA appears to be a useful tool for determining the degree of malignant disease among thymic epithelial neoplasms.

An immunohistologic study of the epithelial components of 81 cases of thymoma.

Eighty-one cases of thymoma were studied immunohistologically with the use of three mouse monoclonal antibodies: one was specific for subcapsular-cortical, one for intra-cortical, and one for medullary epithelial cells. Twenty-eight (60.9%) of 46 polygonal cell thymomas were of the cortical type and 1 (2.2%) was of the medullary type. Ten (55.6%) of 18 spindle cell thymomas and 7 (41.2%) of 17 mixed cell thymomas were of the medullary type, and 1 (5.6%) of 18 spindle cell thymomas was of the cortical type. Fourteen (17.3%) of 81 thymomas were composed of epithelial cells that were triple positive immunologically; although these are unusual, they also may be present in the normal thymus. Based on these findings, triple-positive epithelium in the normal thymus consists of common stem cells that can differentiate into subcapsular-cortical, intra-cortical, and medullary epithelium; these cells may be the target cells for tumorigenesis. Epithelium in polygonal cell thymoma tends to differentiate into cortical epithelium, whereas epithelium in spindle and mixed cell thymomas differentiates into medullary epithelium.

A striational muscle antigen and myasthenia gravis-associated thymomas share an acetylcholine-receptor epitope.

The coincidence of autoantibodies against the acetylcholine receptor (AChR) and muscle striational antigens (SA) is a characteristic finding in thymoma-associated myasthenia gravis (MG), but their origins are still unresolved. Some common muscle antigens that were shown to be targets of anti-SA autoantibodies in thymoma-associated MG have also been detected in normal or neoplastic thymic epithelial cells, suggesting that the release of (eventually altered) antigens from the thymic tumors could elicit SA autoimmunity. In contrast to this model, we report here that titin, which is a recently reported target of SA autoimmunity, is not expressed in thymomas. In addition, we show that skeletal muscle type-II fibers exhibit a striational immunoreactivity with monoclonal antibody mAb155, which was previously identified to label a very immunogenic cytoplasmic epitope of the AChR and neoplastic epithelial cells of MGassociated thymomas. We conclude from these findings that titin autoimmunity in thymoma-associated MG is either due to a molecular mimicry mechanism involving tumor antigens (other than titin) or is a secondary phenomenon following release of titin from muscle. Based on the common immunoreactivity of the AChR, a striational antigen and thymoma, we suggest as the pathogenetic mechanism of thymoma-associated MGa "circulus vitiosus" in which SA autoimmunity could help maintain the AChR autoimmunity that is primarily elicited by the thymomas.

Two-color flow cytometric analysis of thymic lymphocytes from patients with myasthenia gravis and/or thymoma

Phenotypic characteristics of lymphocyte components of the thymus were determined by two-color flow cytometry using monoclonal antibodies to T cell-differentiation antigens and activation antigens, and B cells in eight myasthenia gravis (MG) thymuses and eight thymomas including four associated with MG to clarify the roles of thymus in the diseases. Fifteen normal thymuses and four thymuses from non-MG patients were used as controls. Phenotypes of lymphoid components in thymoma resembled those in normal thymuses, in which the majority of lymphoid cells were immature (common) thymocytes (CD4 +CD8 + or CD1 +CD3 −). The proportions of immature thymocytes, however, were relatively higher and those of mature thymocytes (CD4 +CD8 −, CD4 −CD8 +, CD1 −CD3 + or CD2 +CD3 +) were lower in thymomas than normal thymuses. The results speculate that functions which support further differentiations of immature thymocytes into mature thymocytes are deficient in neoplastic epithelial cells of thymoma. In thymomas, therefore, occasional T cells might escape from the thymic surveillance system which eliminates auto-reactive T cells. Between thymomas associated with MG and those without, however, no significant difference was found in the proportions of thymocytes at various stages of maturation. In nonthymomatous thymuses from MG patients, an increase of both B cells (CD2 −DR + or CD2 −CD21 +) and activated T cells (CD2 +DR + or CD2 +IL-2R +) was observed. Furthermore, immature thymocytes were significantly decreased and mature thymocytes were increased in the MG thymuses, as compared with normal thymuses, non-MG thymuses, and thymomas. These alterations were relevant to the histological findings observed in MG thymuses such as thymic involutions and germinal center formations: decreased immature thymocytes are considered to reflect thymic involutions, while germinal centers, in which lymphocytes from peripheral lymphoid organs are activated, are responsible for the increase of B cells, activated T cells, and mature T cells. In conclusions, our results suggest that neoplastic epithelial cells of thymomas are functionally deficient and that MG thymuses are immunologically active and resemble peripheral lymphoid organs.

Lymphocytes in thymomas are tolerant to self-MHC

Neoplastic epithelial cells of thymoma apparently retain the function of the cortical epithelial cells of normal thymus because a large number of nonneoplastic T cells in thymomas are often CD4 + 8+. However, the lack of medullary structure suggests that thymomas may lack some of the function of the normal thymus, especially the function of the medullary interdigitating cells to induce tolerance to self-antigens on T cells. Thymoma is often associated with autoimmune diseases, most frequently, myasthenia gravis. This suggests that the microenvironment of a thymoma may not be able to induce T cell tolerance to self-antigens. We addressed this question by testing the lymphocytes in thymomas for proliferative responses to mitogens and allogeneic or autologous stimulator cells. The lymphocytes in thymomas proliferated consistently in response to PHA and to allogeneic cells even when the response to OKT3 was undetectable. However, neither the thymoma lymphocytes nor the peripheral blood lymphocytes in these patients proliferated in response to autologous cells.

Primary Thymic Epithelial Neoplasms in Children

Primary epithelial neoplasms of the anterior mediastinum in children are very rare. We have studied 10 cases of thymic epithelial neoplasms in children aged 16 years or less and correlated their histologic features with the clinical outcome. The patients' ages ranged from one to 16 years (mean: 10.2); with a male:female ratio of 1.5:1. Nine patients had symptoms attributable to their tumors; one was asymptomatic. Four patients presented in clinical stage I, one in stage IIb, and five in stage IVb. Histologically, the tumors comprised a heterogenous group displaying a range of morphologic appearances: one tumor had the classic features of lymphocyte-rich thymoma of the adult; four were of the lymphocyte-rich type with associated unusual stromal features; two were spindle cell thymomas with cytologic and architectural atypia; and three displayed obvious cytologic features of malignancy (i.e., thymic carcinoma); two in the last group showed features of small cell carcinoma, and the other was an undifferentiated/anaplastic carcinoma. The epithelial nature of the tumors was supported in six cases by positive staining of the tumor cells with keratin antibodies and in two cases by electron microscopic demonstration of desmosomes and intracytoplasmic bundles of tonofilaments within the tumor cells. The prognosis for these patients correlated well with the degree of atypicality exhibited by the epithelial components; it was very poor in patients with small cell and undifferentiated/anaplastic carcinoma (8 months average survival), better for those with atypical spindle cell thymomas (multiple recurrences and metastases but no fatalities over a 15- to 72-month period), and best in those with lymphocyte-rich thymomas without cytologic atypia (no recurrences or metastases over an 8-month to 3-year follow-up).

Neoplastic epithelial cells express α-subunit of muscle nicotinic acetylcholine receptor in thymomas from patients with myasthenia gravis

We studied the expression of mRNAs coding for the α-subunit of the muscle nicotinic acetylcholine receptor (AChR)in thymomas from patients with myasthenia gravis (MG). Northern blot analysis failed to detect the expression, but amplification of mRNAs derived from thymomas by reverse transcription and polymerase chain reaction produced the DNA fragments containing the nucleotide sequence coding of part of the α-subunit. We further revealed that the α-subunit mRNA was derived from neoplastic epithelial cells of thymoma. Our results support the hypothesis that AChR expressed in thymoma is a candidate for the primary antigen which induces autoimmune responses to muscle AChR. The close relationship between MG and thymoma may be at least in part explained by this hypothesis.

Macrophages and interdigitating reticulum cells in normal human thymus and thymomas: immunoreactivity for interleukin‐1 alpha, interleukin‐1 beta and tumour necrosis factor alpha

Pairs of monoclonal/polyclonal antibodies directed against interleukin-1 (IL-1) alpha, IL-1 beta and tumour necrosis factor (TNF) alpha were used for immunocytochemical identification of cytokine-containing cells in cryostat sections of human fetal thymuses and thymomas. In the fetal thymus immunoreactivity for IL-1 alpha was mainly confined to the medulla and was detected in S-100 positive interdigitating reticulum cells. The pattern of immunoreactivity for IL-1 beta was similar to that for IL-1 alpha, but the number of positive cells was much lower. Cells positive for TNF alpha were extremely rare in the fetal thymus. In 11 thymomas macrophages were constantly present and were regularly distributed throughout the tumour, whereas S-100 positive interdigitating reticulum cells were fewer and were characterized by a zonal distribution. Thymoma-associated macrophages were negative for IL-1 beta and were poorly reactive for IL-1 alpha, only a few positive cells being detected in five of the cases. Some macrophages with immunoreactivity for TNF alpha were detected in seven cases; they formed rosettes with surrounding lymphocytes or were located in a perivascular position. A marked immunoreactivity for TNF alpha was constantly detected in mast cell granules, which were observed in nine thymomas but not in fetal thymus. Positive immunoreactivity of interdigitating reticulum cells for IL-1 alpha was confirmed in five reactive lymph nodes and was also observed in Langerhans' cells in dermatopathic lymphadenitis. Our findings suggest that IL-1 alpha is a crucial molecule for interdigitating reticulum cell and Langerhans' cell function.(ABSTRACT TRUNCATED AT 250 WORDS)

Lymphocytes in thymoma: association with myasthenia gravis is correlated with increased number of single‐positive cells

In the normal thymus the number of lymphocytes and the proportion of CD4+CD8+ double-positive cells decrease as a function of age. Six thymomas in elderly patients without myasthenia gravis had a large number of double-positive T cells, as seen in the normal thymic cortex. This suggests that the epithelial cells of thymoma have the function of cortical epithelium and can induce precursor T cells to differentiate to the double-positive stage but not further. In seven thymomas with associated myasthenia gravis, however, there was a conspicuous population of mature single-positive T lymphocytes. These mature T cells may include autoreactive T cells.

Distribution of fibronectin and laminin in human thymoma

The distribution of fibronectin and laminin, two extracellular matrix and basement membrane components, was studied in 55 cases of thymoma using immunohistochemistry. The results were compared according to histologic type or clinical stage of the tumor. In addition, electron microscopic observation was done to clarify the extracellular structure of thymoma. Two staining patterns were seen. First, a diffusely or partially intricate network of fibers that contained fibronectin and laminin surrounded tumor cells in 21 out of 55 cases. Most of these cases were spindle cell thymomas and showed low invasive tumors. Second, fibers that contained fibronectin and laminin were restricted only to the septa, blood vessels, and perivascular spaces that did not show a network in the remaining 34 cases. Polygonal cell thymomas showed the latter staining pattern and these were more invasive tumors. We conclude that this network is a characteristic structure of spindle cell thymomas and is related to the invasiveness of the tumor.

The expression of a tissue-specific self-peptide is required for allo-recognition.

We have compared the functional properties of I-Ad expressed on different cell types. Specifically, we have transfected I-A alpha d and I-A beta d cDNA into a panel of T cell thymomas of various phenotypes. Excellent class II surface expression was achieved in all T cell tumors, equivalent in level to that found on the B cell lymphoma A20. Interestingly, however, two allo-I-Ad-specific Thy differed in their recognition of the transfected tumor cells: whereas the 42H11 T cell hybridoma (THy) was stimulated very efficiently by all transfectants, the RK38.2.2 Thy did not react to any of them. Both THy responded equally well to I-Ad on A20 B lymphoma cells. Purified macrophages isolated from various sources were also differentially recognized by the two THy, although there was only a quantitative difference in stimulation. Taken together, these results are best interpreted to show that the TCR of the RK38.2.2 THy is specific for I-Ad in the context of a B cell-specific determinant, possibly a self-peptide that is naturally associated with Ia. A cross-reactive molecule could be expressed by macrophages and COS-1 cells, but not by T cells.

Immunohistochemical Demonstration of H Antigen, Peanut Agglutinin Receptor, and Saphora japonica Receptor Expression in Infant Thymuses and Thymic Neoplasias

Ten infant thymuses and 13 primary thymic tumors obtained from archived paraffin-embedded tissue were examined for the presence of tissue blood group O antigen (H), peanut agglutinin receptor antigen (PNA-r), Saphora japonica agglutinin receptor antigen (SJA-r), carcinoembryonic antigen (CEA), cytokeratin (CK), and epithelial membrane antigen (EMA). In the thymuses studied, Hassall's corpuscles contained abundant immunoreactive CK, PNA-r, and H antigens, whereas CEA, SJA-r, and EMA were present focally in Hassall's corpuscles. Immunoreactive CK, PNA-r, and CEA were demonstrated focally in the subcapsular region, cortical nurse cells, and subcapsular-perivascular monocytic cells, respectively. PNA-r was present in all 12 epithelial type tumors, including all eight thymomas. CEA was present in nine tumors, including six thymomas. Six thymomas contained H antigen and SJA-r; five continued CK and EMA. SJA-r and EMA were also present in one carcinoid tumor of thymic origin. In epithelial thymomas, the antigens stained nests of epithelial cells resembling the pattern of staining in Hassall's corpuscles. Membrane staining of spindle cells of both spindle cell and epithelial thymomas was less intense than staining of epithelial type cells.

Myoid cells in the human thymus and thymoma revealed by three different immunohistochemical markers for striated muscle.

Specimens of human thymus and thymoma obtained from 81 patients including 22 with myasthenia gravis (MG), ranging in age from 10 weeks of gestation to 75 years were examine by the indirect immunoperoxidase technique using three different striated muscle markers: myoglobin, CPK-mm and beta-enolase. Myoid cells ovoid or elongated in shape were distributed in the medulla and showed positive immunoreactions for myoglobin, CPK-mm and beta-enolase by light and electron microscopy. The antigens were first detectable immunohistochemically in the fetus at the 10th week of gestation and showed a relative increase in number and varying distribution patterns with age involution in non-myasthenic thymuses. In MG cases, the number of myoid cells was smaller than in non-myasthenic cases, especially in markedly involuted thymuses. Myoid cells were occasionally observed within thymomas in areas showing medullary differentiation. Small numbers of cells showing a positive immunoreaction for beta-enolase but not for myoglobin and CPK-mm were scattered in the area predominantly composed of epithelial cells. These cells were occasionally positive for keratin and showed desmosome-like structures but no apparent myofilaments, suggesting that muscle-like differentiation of epithelial cells might occur in the neoplastic process of thymoma.