The hairy-cells (HC) of 10 patients with hairy-cell leukaemia were studied with several techniques to evaluate their phagocytic potential. Mononuclear cells from normal donors and from patients with acute monocytic leukaemia served as controls. Light microscopically HC seemed to have ingested bacteria or latex particles. Treatment of the cells with lysostaphin, an enzyme that kills extracellular Staphylococcus aureus, showed that almost all 'ingested' bacteria were extracellular. Lanthanum nitrate, added during the fixation procedure for electron microscopy, stained both the outer cell membrane and the membranes of the 'phagosomes' of the HC, also indicating that the 'ingested' particles were extracellular. HC showed no increased oxygen consumption on exposure to bacteria in the presence of serum. Furthermore, HC showed no lysozyme or peroxidase activity, whereas non-specific esterase activity was much weaker than in monocytes. These findings, which show that HC are essentially non-phagocytic, constitute strong evidence against a monocytic origin of the malignant cells of hairy-cell leukaemia.

AIM: To analyse the expression of multidrug resistance (MDR) related proteins at different steps in colorectal carcinogenesis. METHODS: The presence of three MDR related proteins (Pgp, MRP1, and LRP/MVP) was studied by means of immunohistochemistry in normal, adenomatous, and malignant colorectal epithelium. Formaldehyde fixed, paraffin embedded tissue sections of 17 samples of colorectal tissue were used (normal mucosa, n = 4; adjacent mucosa, n = 5; adenoma, n = 5; carcinoma, n = 3). RESULTS: For all three proteins, expression was found in the surface epithelium and the upper parts of the crypts in normal colon. In the adenomas, staining was seen along the complete length of the crypts. In the carcinomas analysed, all epithelium showed positive staining. Mucosa adjacent to either carcinoma or adenoma showed staining patterns mostly resembling those of normal mucosa, but sometimes some extension of staining was seen along the crypt. CONCLUSIONS: These proteins already show increased expression in the adenoma stage. In the absence of adequate mucin production in adenomas, MDR related proteins could be an important factor in protecting the epithelium against further environmentally induced genetic damage. This could be one of the reasons why only about 5% of colorectal adenomas will actually progress to carcinomas.

Cyclin D1 overexpression, detected by standard immunohistochemistry, was correlated with other prognostic variables and its prognostic value was evaluated in a group of 148 invasive breast cancers with long-term follow-up. Overexpression of cyclin D1 (59% of cases) was negatively correlated (chi 2 test) with histological grade (P = 0.0001), mean nuclear area (P = 0.004), mean nuclear volume (P = 0.02), and mitotic activity (P = 0.03) and positively correlated with estrogen receptor (P = 0.0001). There was a strong correlation between cyclin D1 overexpression and histological type (P = 0.0001). Positive cyclin D1 staining was seen in 11 of 13 tubular carcinomas, 3 of 3 mucinous carcinomas, 4 of 4 invasive cribriform carcinomas, and 17 of 20 lobular carcinomas. Of 102 ductal cancers, 52 were positive, and all 6 medullary carcinomas were negative. There were no significant correlations with lymph node status, tumor size, or DNA ploidy. In survival analysis, cyclin D1 overexpression did not provide significant univariate or multivariate prognostic value. In conclusion, cyclin D1 is mainly overexpressed in the well differentiated and lobular types of invasive breast cancer and is strongly associated with estrogen receptor positivity. It is negatively correlated with the proliferation marker mitoses count and with the differentiation markers nuclear area and nuclear volume. However, cyclin D1 overexpression does not seem to have prognostic value in invasive breast cancer when no adjuvant treatment is given.

Drug resistance plays an important role in chemotherapy failure in breast cancer. We studied the expression of MDR1, MRP, LRP, DNA topoisomerases, p53 and Ki-67 in different groups of breast cancer patients in relation to chemotherapy. Tissues from 6 normal breasts and 20 primary operable, 40 locally advanced and 10 anthracycline-resistant metastatic breast cancers were assessed. Sequential samples of the same patient were available from 17 patients with locally advanced breast cancer undergoing neo-adjuvant chemotherapy and in 7 metastatic patients undergoing paclitaxel treatment. Protein expression was investigated by immunohistochemistry. Significantly higher protein expression was observed for Pgp, Ki-67 and p53 in the locally advanced breast cancers than in primary operable breast cancers. No other significant differences in protein expression were found among the 3 breast cancer groups. Expression of none of the markers that could be assessed (Pgp, MRP, LRP, p53 and Ki-67) in locally advanced breast cancer had predictive value for pathological response. Interestingly, after chemotherapy a significant decrease in percentage of Ki-67 positive tumor cells was observed, whereas the other markers did not vary substantially. Furthermore, considering all breast cancer samples, a cumulative dose of doxorubicin >400 mg/m2 inversely correlated with Ki-67 positivity. However, 2 patients with a pathological complete remission had only 5-10% Ki67-positive tumor cells before chemotherapy, indicating that Ki67 negativity itself is not responsible for chemoresistance. In conclusion, none of the known proteins related to multidrug resistance predicted response to chemotherapy in breast cancer, and resistant clones left behind generally had a low proliferation rate.Int. J. Cancer 71: 787-795, 1997. © 1997 Wiley-Liss Inc.

PURPOSE To assess the prognostic value of p53, bcl-2, bax, and neovascularization in radically resected non-small-cell lung cancer (NSCLC) patients. PATIENTS AND METHODS Tumors from 116 patients were assessed by immunohistochemistry for expression of p53 (DO7 and PAb1081), bcl-2, and the quantification of microvessel density (CD-31). In addition, the expression of bax was assessed in 61 stage I tumors. The median levels of expression of each marker were used as cutoff points. RESULTS p53 was not correlated to any patient or tumor characteristic, whereas bcl-2 showed higher expression in squamous cell carcinomas (P < .001). bax expression was significantly related with male sex (P = .006) and adenocarcinoma type (P = .0013). p53 status, assessed with one monoclonal antibody (MoAb), was not predictive for survival; however, the combination of staining results obtained with two MoAbs identified the DO7-/PAb1801+ tumors as those with the worst prognosis. bcl-2 expression was associated with longer survival in stage I patients (P = .0169). The combined group expressing p53+(PAb1801)/bcl-2- had the worst survival in stage I patients (P = .034) and in the whole series in comparison with the other combinations of the two oncoproteins. bax expression alone had no influence on survival of stage I patients, but patients with bax+/bcl-2- tumors had the worst prognosis (P = .02 in comparison with bax+/bcl-2+). Tumor neovascularization was not related with other factors, and patients with CD-31+ tumors had a shorter survival duration than those with CD-31- tumors only in stage II (P = .0283). By multivariate analysis including all patients, the presence of p53+/ bcl-2- tumor expression and large tumor diameter (> or = 4cm) were independent prognostic factors for shorter survival duration. For stage I, only the presence of bax+/ bcl-2- tumor expression had a significant negative influence on survival. CONCLUSION The interaction and the regulation of new biologic markers, such as those involved in the apoptotic pathway, are complex. Combinations of the expression of several of them may give more valuable information than the study of just one. Prognostic influence of p53 staining varied depending on the choice of antibody and the combination of bcl-2- together with p53+ (PAb1801) or with bax+ had the worst influence on survival for patients with stage I NSCLC.

The presence of Epstein-Barr virus (EBV) was studied in specimens of 50 primary non-Hodgkin's lymphomas (NHL) of the salivary gland and the oral cavity and 11 solitary adenolymphomas of the parotid gland, using EBER-1/2 in situ hybridisation and by immunohistochemistry for the detection of latent membrane-protein-1 (LMP-1). None of the patients were tested for HIV-infection, nor were there any clinical signs to suspect HIV-infection. In one adenolymphoma, few reactive EBER-1/2 positive cells were detected. In this case staining for LMP-1 was negative. In one oral B-cell NHL, EBER-1/2 positive lymphoma cells were identified; these cells also expressed LMP-1. None of the 31 oral (30 B-cell and one T-cell) and 18 salivary gland (all B-cell) NHLs and none of the 10 adenolymphomas were EBER-1/2 positive or expressed LMP-1. These results indicate that EBV is not involved in the pathogenesis of oral and salivary gland primary NHL and adenolymphoma in immunocompetent patients.

Background: Growth factors play a role in proliferation and motility of malignant glial cells, through autocrine and paracrine mechanisms. Also, proliferation of non-tumour cells, e.g., endothelial cells, is likely to be controlled by growth factors. Several growth factors with their appropriate receptors can be involved, but studies on tissue specimens evaluating this in glioma are rare.Materials and Methods: We evaluated the potential role of Transforming growth factor-alpha (TGF-alpha) and Epidermal growth factor receptor (EGF-R), the Platelet-derived growth factor A-and B-chain (PDGF-A and PDGF-B) and its receptors (PDGFR alpha and PDGFR beta), and basic fibroblast growth factor (bFGF) in gliomas by analysing 86 of these tumours on the single cell level for the presence of immunoreactive growth factors and receptors. In a few cases double-staining experiments were done to directly visualize co-expression of factor and receptor.Results: Multiple growth factors and their receptors are present in astrocytic tumours; the higher the grade, the more growth factors and the more positive cells are found. Oligodendroglial rumours and pilocytic astrocytomas showed little expression. Autocrine and paracrine mechanisms were frequently possible in the astrocytic tumours, often more than one loop could be involved. Interestingly, it was also frequently possible that non-tumour cells produced a growth factor for which the tumour cells expressed the receptor.Conclusions: Multiple growth factors appear to be involved in astrocytic tumours, with frequent autocrine and paracrine loops. Expression of these molecules seems to increase with increasing grade. The results argue for a contribution of non-tumour cells to the growth of a tumour.

To get insight into the failure of the immune system to eradicate Epstein-Barr virus (EBV) harboring Hodgkin and Reed-Sternberg cells (H-RS cells), expressing the latent membrane protein 1 (LMP1), we analyzed major histocompatibility complex (MHC) class I expression on H-RS cells in relation to the presence of activated cytotoxic cells, i.e., granzyme-B-expressing lymphocytes. H-RS cells in EBV+ cases of Hodgkin's disease (HD) were found to express significantly higher levels of MCH class I heavy- and light-chain molecules compared with EBV- HD cases. When low levels of MHc class I expression were found (mainly in EBV- cases), these were not associated with low levels of the transporter protein associated with antigen presentation 1 (TAP-1). The relatively high levels of MHC class I expression in H-RS cells in EBV+ HD cases were accompanied by significantly higher numbers of activated cytotoxic T lymphocytes (CTLs) as shown by the presence of increased numbers of CD8 and granzyme B+ lymphocytes. However, these cells were only sporadically detected in the close vicinity of the H-RS cells. These data suggest that mechanisms other than downregulation of MHC class I or TAP-1 expression on H-RS cells are involved in the failure of the immune system to eradicate EBV harboring H-RS cells. Probably, the function of activated CTLs is locally inhibited by the H-RS cells or by reactive cells in the vicinity of the H-RS cells.

Among non-Hodgkin's lymphomas occurring in childhood two major histologic subgroups can be identified: (1) Burkitt's lymphoma and (2) T-cell lymphoblastic lymphoma, an uncommon high-grade malignant non-Hodgkin's lymphoma. Although Burkitt's lymphoma with maxillofacial involvement is a well-documented disease, T-cell lymphoblastic lymphoma in the perioral region is rare. An unusual case of T-cell lymphoblastic lymphoma with initial oral manifestation in an 18-month-old child is presented.

Expression of both P-glycoprotein (P-gp) and mutant p53 have recently been reported to be associated with poor prognosis of breast cancer. The expression of P-gp is associated in vitro and in vivo with cross-resistance to several anti-cancer drugs. p53 plays a regulatory role in apoptosis, and mutant p53 has been suggested to be involved in drug resistance. Interestingly, in vitro experiments have shown that mutant p53 can activate the promoter of the MDR1 gene, which encodes P-gp. We investigated whether p53 and P-gp are simultaneously expressed in primary breast cancer cells and analysed the impact of the co-expression on patients prognosis. Immunohistochemistry was used to investigate P-gp expression (JSB-1, C219) and nuclear p53 accumulation (DO-7) in 20 operable chemotherapy untreated and 30 locally advanced breast cancers undergoing neoadjuvant chemotherapy with doxorubicin and cyclophosphamide. Double immunostaining showed that P-gp expression and nuclear p53 accumulation often occur concomitantly in the same tumour cells. A correlation between p53 and P-gp expression was found in all 50 breast cancers (P = 0.003; Fisher's exact test). P-gp expression, nuclear p53 accumulation, and co-expression of p53 and P-gp were more frequently observed in locally advanced breast cancers than in operable breast cancers (P = 0.0004, P = 0.048; P = 0.002 respectively. Fisher's exact test). Co-expression of p53 and P-gp was the strongest prognostic factor for shorter survival by multivariate analysis (P = 0.004) in the group of locally advanced breast cancers (univariate analysis: P = 0.0007). Only 3 out of 13 samples sequentially taken before and after chemotherapy displayed a change in P-gp or p53 staining. In conclusion, nuclear p53 accumulation is often associated with P-gp expression in primary breast cancer, and simultaneous expression of p53 and P-gp is associated with shorter survival in locally advanced breast cancer patients. Co-expression of P-gp and mutant p53 belong to a series of molecular events resulting in a more aggressive phenotype, drug resistance and poor prognosis.

BACKGROUND The recently discovered LRP protein has been shown to be involved in drug resistance and possibly in detoxification processes. MATERIALS AND METHODS To study the relation between LRP expression and exposure to cigarette smoke, LRP immunoreactivity was evaluated in 39 paraffin embedded normal lung tissues derived from patients operated on for pneumothorax, and related to amount of pack years smoked. We also studied the LRP protein expression in 36 non-small-cell lung cancer (NSCLC) samples and related the expression to patient characteristics and survival. Furthermore 17 lung tumor samples (10 NSCLC and 7 SCLC) derived from patients treated with chemotherapy were analysed in order to investigate the relation between LRP or MRP expression and the patient's response to chemotherapy. RESULTS In the normal lung tissues, LRP intensity levels were not correlated to the amount of pack years smoked, although a trend was seen for higher LRP intensity levels in patients who smoked more than 10 pack years. LRP expression was significantly higher in NSCLC samples than in SCLC samples, and all SCLC samples displayed very low LRP expression. Within NSCLC, squamous cell and adenocarcinomas had higher LRP expression than large cell undifferentiated and mixed tumors. In NSCLC patients LRP expression was not a prognostic factor for survival. At initial analysis LRP expression levels did not predict for the response to chemotherapy. Only 3 out of 17 patients expressed MRP, and all SCLC samples were MRP negative. CONCLUSIONS Striking different expression levels were seen between NSCLC and SCLC for both LRP and MRP. In a preliminary analysis LRP expression was not predictive for response to chemotherapy in lung cancer patients. In pneumothorax patients LRP levels were not correlated with the amount of pack years smoked.

Multidrug resistance (MDR) to anticancer drugs is a major cause of treatment failure in cancer. The lung resistance protein LRP is a newly described protein related to MDR in several in vitro models. LRP has been shown to be a strong predictor of poor response to chemotherapy and prognosis in acute myeloid leukemia and in ovarian carcinoma patients. Recently, based on a 57% and 88% amino acid identity with major vault proteins from Dictyostelium discoideum and Rattus norvegicus, respectively, we identified LRP as the human major vault protein, the main component of highly conserved cellular organelles named vaults. We have studied the immunohistochemical expression of LRP in freshly frozen normal human tissues and 174 cancer specimens of 28 tumor types. LRP was broadly distributed in normal and malignant cells, but distinct patterns of expression were noticed. High LRP expression was seen in bronchus, digestive tract, renal proximal tubules, keratinocytes, macrophages, and adrenal cortex whereas varying ing levels were observed in other organs. LRP was detected in all tumor types examined, but its frequency varied, fairly reflecting the chemosensitivity of different cancers. For example, low rates of LRP positivity were seen in testicular cancer, neuroblastoma, and acute myeloid leukemia; intermediate in ovarian cancer; and high in colon, renal, and pancreatic carcinomas. The wide occurrence of LRP in normal and transformed cells in humans, its similar distribution to that of vaults in other species, as well as the high level of conservation among eukaryotic cells of both the amino acid sequence of the major vault protein and the composition and structure of vaults, suggest that vault function is important to eukaryotic cells.

Reed-Sternberg (RS) and Hodgkin's (H) cells are considered to be the neoplastic cells in Hodgkin's disease. Although most data suggest a lymphoid origin, the nature of these cells still remains the subject of considerable controversy. Recently, monoclonal antibodies became available, directed against granzyme B, a serine protease specifically expressed by activated cytotoxic T cells (CTLs) and natural killer (NK) cells. Using two granzyme B-specific antibodies directed against different epitopes, we studied the expression of granzyme B in a well characterized group of Epstein-Barr virus (EBV)-positive and EBV-negative cases of Hodgkin's disease. Granzyme B expression was found in part of the H-RS cells in 11 out of 61 tested cases (18%, 9 of 46 cases of nodular sclerosing and 1 of 12 mixed cellularity Hodgkin's disease). In none of these cases did H-RS cells express B-cell markers, whereas in four cases, expression of either the T-cell marker CD3 or CD8 was found in a small minority of H-RS cells. The percentage of granzyme B-positive H-RS cells ranged from < 10% to > 50%. Granzyme B-positive H-RS cells were present in 6 of 26 EBV-positive cases and in 5 of 35 EBV-negative cases, indicating no relationship with the presence of EBV. Moreover, no significant differences were found regarding either stage at presentation or clinical outcome. We conclude that in a restricted number of cases of Hodgkin's disease, the H-RS cells express granzyme B, and therefore might be considered the neoplastic equivalent of either activated CTLs or NK cells.

PURPOSE:The effects of estrogen only on prostate tissue were studied in 9 aging male transsexuals who had undergone orchiectomy. MATERIALS AND METHODS:Prostate volume was measured by ultrasound. Biopsies were stained for prostatic acid phosphatase (PAP), prostate specific antigen (PSA), Ki-67 and sex steroid receptors. RESULTS:Prostates were small, and biopsies showed PAP and PSA in most cases. No malignancy was found. Few androgen receptors were detectable but there were ample estrogen and progesterone receptors in stroma and epithelium. CONCLUSIONS:Prostates were small after long-term estrogens only. The sex steroid receptor content was considerably changed. PSA and PAP in the specimens suggested that their production does not depend on testicular androgens alone.

Twenty-two patients with primary non-Hodgkin's lymphoma (P-NHL) of the salivary glands have been studied with reference to age, sex, clinical symptoms, location of primary tumor, histological subtype, grade of malignancy according to the Working Formulation, stage of disease, treatment and follow up. The clinical features corresponded with previously reported data on salivary gland lymphomas. The most frequent histological diagnosis was MALT lymphoma. Most patients had a localised low-grade lymphoma. Overall survival was influenced by grade of malignancy according to the Working Formulation.

Using RT-PCR analysis of Epstein-Barr virus (EBV) latent gene transcription in EBV-harboring cell lines (JY and RAJI) and in post-transplantation lymphoproliferative disorders (PT-LPDs), we detected transcription of all tested latent genes (EBNA1, EBNA2, LMP1, LMP2A, and BARF0) in all cases, suggesting the presence of similar EBV expression patterns in both PT-LPDs and cell lines. In addition, the detection of immediate early (ZEBRA) and early gene (BHRF1) transcripts in cell lines and PT-LPDs indicates that activation of the virus lytic cycle occurs. To investigate EBV expression patterns at the single-cell level, a combination of immunohistochemistry and RNA in situ hybridization (including double-staining procedures) was used. In the JY and RAJI cell lines, the latency type 3 expression pattern was detected in 80 to 90% of the cells as shown by the co-expression of EBNA2 and LMP1. In contrast, in the three PT-LPDs that could be analyzed by double staining, cells expressing both EBNA2 and LMP1 were rarely detected. A mixture of at least three different cell populations were identified: (1) cells exclusively expressing EBER1/2 and EBNA1 (latency type 1); (2) cells expressing EBER1/2, EBNA1, and LMP1 (latency type 2); and (3) cells expressing EBER1/2, EBNA1, and EBNA2 in the absence of LMP1. Activation of the lytic cycle was observed in a small minority of cells, as demonstrated by detection of ZEBRA and EA-D in all cases and GP350/220 in two cases. Thus, in contrast to EBV-transformed cell lines, the observed EBV gene expression pattern in PT-LPDs reflects a mixture of multiple EBV-harboring subpopulations expressing different subsets of EBV-encoded proteins. These data indicate that the operational definitions of EBV latencies in vitro cannot easily be applied to PT-LPDs but that a continuum of different latency expression patterns can be detected at the single cell level in these lymphomas with, in a small minority of cells, progression to the virus lytic cycle.

AIM--To compare the immunoreactivity of monoclonal antibodies S12 and CS1-4, which recognise different epitopes of the Epstein-Barr virus (EBV) latent membrane protein-1 (LMP-1), in EBV associated benign and malignant lymphoproliferative disorders and control tissues processed using different methods. RESULTS--Both monoclonal antibodies gave comparable results on frozen tissue sections and formalin fixed, paraffin wax embedded samples from cases with Hodgkin's disease and infectious mononucleosis. In all cases S12 stained more cells than CS1-4. For EBV associated B and T non-Hodgkin's lymphomas, frozen tissue sections yielded better LMP-1 staining results than formalin fixed material. Again, in all these cases S12 stained more cells and gave stronger results than CS1-4. For EBV negative tissues, both monoclonal antibodies showed cross-reactivity with melanocytic-like cells in the basal cell layer of the skin, synaptophysin-like staining in layers three and four of the cortex of the brain, and myelin-like staining in peripheral nerves and peripheral ganglion cells. Staining with S12 was always much stronger. Moreover, in contrast to CS1-4, S12 stained pancreatic islands in formalin fixed material but not in frozen tissue sections and sporadically stained solitary epithelial cells in the large bowel especially in formalin fixed tissue sections. CS1-4 also cross-reacted with myoepithelial cells around hair follicles and other adnexa of the skin. CONCLUSION--The results indicate that for optimal detection of LMP-1, S12 yields better results than CS1-4 and that tissue processing is very important especially when B and T non-Hodgkin's lymphomas are examined.

BHRF1, one of many Epstein-Barr virus (EBV)-encoded proteins, shows strong functional homology to the human bcl-2 proto-oncogene product, a protein involved in the pathogenesis of a subset of B-cell lymphomas, ie, follicle center cell lymphomas (FCCL). We have investigated the presence of possible latent and lytic transcripts of BHRF1 using a reverse transcriptase-polymerase chain reaction (RT-PCR)-based assay in a group of EBV-associated B-cell lymphomas in patients with (N = 5) or without overt immunodeficiency (N = 4), in T-cell lymphomas (N = 9), and in cases of Hodgkin's disease (N = 6). BHRF1 transcription was found consistently in EBV-associated (ie, diffuse EBER 1/2-positive) B- cell lymphomas in patients with or without immune deficiency, whereas in EBV-associated T-cell lymphomas or in EBV-associated Hodgkin's disease, BHRF1 transcription was only detected in two T-cell lymphomas and one case of Hodgkin's disease, which also harbored EBER 1/2- positive reactive cells. Moreover, weak BHRF1 signals were found in two T-cell lymphomas where EBER 1/2 expression was detected mainly in sporadic reactive lymphocytes and in one reactive tonsil with sporadic EBER 1/2-positive lymphocytes. BHRF1 transcripts were found to be generated by the C or W promoter (associated with viral latency) and/or by the H promoter (associated with the virus lytic cycle). In all cases with H promoter-derived BHRF1 transcripts, transcripts encoding ZEBRA were also detected, suggesting a reactivation of the virus lytic cycle. Analysis of other EBV genes revealed transcription of BARFO in all tested EBV-harboring tissues. Transcription of EBNA1 and LMP1 was usually detected, whereas EBNA2 transcription was found exclusively in B-cell lymphomas in immunocompromised patients. These data demonstrate that BHRF1 transcripts are exclusively found in EBV-associated B-cell lymphomas. When BHRF1 transcripts are detected in T-cell lymphomas or in Hodgkin's disease, it is probably due to the presence of reactive EBER 1/2-positive lymphocytes. The consistent transcription of BHRF1 in EBV-associated B-cell lymphomas suggests a possible pathogenic role for this gene product in EBV-positive B-cell lymphomas analogous to bcl-2.

In this study the distribution patterns of various extracellular matrix components and their receptors (i.e. beta 1 integrins) in B-cell non-Hodgkin lymphomas were examined and compared to those in reactive lymphoid tissue. Neoplastic follicles within follicular lymphomas showed similar patterns to that observed in reactive follicles, which appeared to be strongly associated with the presence of follicular dendritic cells. Diffuse lymphomas of low and intermediate malignancy grade revealed features comparable to those of interfollicular areas of reactive lymphoid tissue, irrespective to which compartment the tumour cells were related. Highly malignant lymphomas, however, displayed unique extracellular matrix configurations, resulting from active matrix degradation by macrophages; this may support rapid tumour growth. Extranodal lymphomas showed virtually the same matrix patterns as their nodal counterparts, suggesting that (malignant) lymphoid cells generate (at least partly) their own specific microenvironment. In reactive lymphoid tissue beta 1 integrins were mainly found on resident cells and except for alpha 4, alpha 5 (and beta 1) the lymphoid cells expressed very little, if any, beta 1 integrins. In comparison, expression of these integrins on lymphoma cells was reduced (follicular lymphomas) or could not be detected at all (diffusely growing lymphomas); this might contribute to the growth pattern and metastatic properties of the tumours.