Search

The posttransplantation lymphoproliferative disorders (PT-LPDs) are a morphologically heterogeneous group of Epstein-Barr virus (EBV)-driven lymphoid proliferations of varying clonal composition. Some PT-LPDs regress after a reduction in immunosuppression, while others progress in spite of aggressive therapy. Previously defined morphologic categories do not correlate with clonality, and neither morphology nor clonality has reliably predicted the clinical behavior of PT-LPDs. We investigated 28 PT-LPD lesions occurring in 22 patients for activating alterations involving the bcl-1, bcl-2, c-myc, and H-, K- and N-ras proto-oncogenes and for mutations involving the p53 tumor suppressor gene. We correlated the results of these studies with the morphology of the lesions, their clonality based on Ig heavy and light chain gene rearrangement analysis, and the presence and clonality of EBV infection. We found that the PT-LPDs are divisible into three distinct categories as follows: (1) plasmacytic hyperplasia: most commonly arise in the oropharynx or lymph nodes, are nearly always polyclonal, usually contain multiple EBV infection events or only a minor cell population infected by a single form of EBV, and lack oncogene and tumor suppressor gene alterations; (2) polymorphic B-cell hyperplasia and polymorphic B-cell lymphoma: may arise in lymph nodes or various extranodal sites, are nearly always monoclonal, usually contain a single form of EBV, and lack oncogene and tumor suppressor gene alterations; and (3) immunoblastic lymphoma or multiple myeloma: present with widely disseminated disease, are monoclonal, contain a single form of EBV, and contain alterations of one or more oncogene or tumor suppressor genes (N-ras gene codon 61 point mutation, p53 gene mutation, or c-myc gene rearrangement). The PT-LPDs are divisible into three categories exhibiting distinct morphologic and molecular genetic characteristics. Alterations involving the N-ras and c-myc proto- oncogenes and the p53 tumor suppressor gene may play an important role in the development and/or progression of the PT-LPDs.

Recently reported cases of CD4+ CD56+ hematologic malignancies with a strong predilection for the skin correspond to the neoplastic counterpart of plasmacytoid dendritic cells. This rare, aggressive malignancy lacks pan-myeloid and pan-lymphoid markers and often presents with cutaneous lesions, splenomegaly, and involvement of lymph nodes and bone marrow. It has a poor prognosis, and many patients progress to acute myeloid leukemia. The proposed cellular origin is a CD56+ precursor cell related to plasmacytoid monocytes, which strongly expresses CD123 (IL-3Ra). We describe a 70 year-old man who presented with gray-brown truncal nodules and plaques, lymphadenopathy, and splenomegaly. His bone marrow demonstrated malignant CD4+ CD56+ mononuclear cells. Histologic sections of skin lesions showed an atypical infiltrate extending into the deep reticular dermis. Immunohistochemical staining of these cells for CD4 was diffusely positive. Moreover, the infiltrate strongly expressed CD56 and CD123 but showed only patchy or negative staining for other T and B cell markers. The combination of the patient’s clinical presentation and biopsy results best fits the diagnosis of this newly characterized, distinct clinicopathologic entity described in recent literature as agranular CD4+ CD56+ hematodermic neoplasm, plasmacytoid dendritic cell acute leukemia, and tumor-forming accumulations of plasmacytoid monocytes associated with myeloid disorders.

Waldeyer's ring is an uncommon, rarely reported primary site for Hodgkin's disease. We report a series of 16 such cases culled from the files of the Armed Forces Institute of Pathology and the National Cancer Institute. The patients' median age was 41 years (range, 14-74), and they presented with airway obstruction or unilateral tonsillar enlargement. The disease was localized to the Waldeyer's ring (stage I) in 46% of patients and extended to the cervical lymph nodes (stage II) in 39% and to the spleen (stage III) in 15%. Local radiation therapy, with or without chemotherapy, obtained a complete response in all but two patients. There was local recurrence in one patient and distant spread in three others. All patients for whom follow-up is available are alive without evidence of disease at 9 to 216 months (median, 20 months) except two who died of widespread Hodgkin's disease and two others who died of other causes. Histologically, eight cases were classified as mixed cellularity type (50%), four as nodular sclerosis (25%), and one as lymphocyte predominance, nodular (LPn; 6.3%); three others that showed interfollicular involvement were unclassified (18.7%). The Reed-Sternberg (RS) and atypical mononuclear cells in most cases of mixed cellularity and interfollicular types and all cases of nodular sclerosis had the classic immunophenotype (CD45-, CD20- and/or CD45RO-, CD15+ and/or CD30+). In the single case of LPn, they were of B-cell lineage (CD45+, CD20+, CD45RO-, CD15-, CD30-). In situ hybridization performed on routinely processed sections revealed Epstein-Barr virus (EBV) EBER1 mRNA in RS cells of eight of 12 cases studied (67%) only in mixed cellularity and nodular sclerosis, but not in LPn. We conclude that, however rarely, Hodgkin's disease of typical morphology and immunophenotype can originate in Waldeyer's ring. The incidence of EBV detection in the RS cells in our study is greater than that usually seen in nodal Hodgkin's disease in the United States. The greater prevalence of EBV-related Hodgkin's disease at this site is probably a reflection of the fact that the Waldeyer's ring is a reservoir for EBV.

The association of Dilantin (hydantoin) therapy and lymphadenopathy in the form of "pseudolymphoma" or malignant lymphoma has been reported primarily in the clinical literature in single case reports or small series, most of which contain very few pathologic details. We studied a series of 25 lymph node lesions collected at the Armed Forces Institute of Pathology from 1965 to 1991 that were suspected to be related to intake of the antiseizure medication Dilantin. Each case had been coded by the Environmental and Toxicology Department according to the so-called "operational degrees of certainty," which gives the possibility that a pathologic change is the result of a drug. Of the 25 cases, six were coded as probable, 17 as possible, and two as coincidental. The male:female ratio was 11:14, and patient ages ranged from 24 to 81 years (median, 53 years). Documented lymphadenopathy developed 1 week to 30 years (median, 5 years) after the start of Dilantin. The histologic features were reviewed in 25 of 25 cases, and the immunophenotypic data was available in 18 of 25. Fifteen of 25 cases showed a benign histology, which we primarily classified according to the presence or absence of immunoblastic hyperplasia. Seven cases were non-Hodgkin's lymphoma, and three were Hodgkin's disease. In two of five cases for which sequential biopsies were available for review, there was progression from paracortical hyperplasia to malignant lymphoma. Our report describes the clinicopathologic features of 25 cases of Dilantin-associated lymphadenopathy and confirms the heterogeneity and nonspecificity of these histologic patterns.

The posttransplantation lymphoproliferative disorders (PT-LPDs) are a morphologically heterogeneous group of Epstein-Barr virus (EBV)-driven lymphoid proliferations of varying clonal composition. Some PT-LPDs regress after a reduction in immunosuppression, while others progress in spite of aggressive therapy. Previously defined morphologic categories do not correlate with clonality, and neither morphology nor clonality has reliably predicted the clinical behavior of PT-LPDs. We investigated 28 PT-LPD lesions occurring in 22 patients for activating alterations involving the bcl-1, bcl-2, c-myc, and H-, K- and N-ras proto-oncogenes and for mutations involving the p53 tumor suppressor gene. We correlated the results of these studies with the morphology of the lesions, their clonality based on Ig heavy and light chain gene rearrangement analysis, and the presence and clonality of EBV infection. We found that the PT-LPDs are divisible into three distinct categories as follows: (1) plasmacytic hyperplasia: most commonly arise in the oropharynx or lymph nodes, are nearly always polyclonal, usually contain multiple EBV infection events or only a minor cell population infected by a single form of EBV, and lack oncogene and tumor suppressor gene alterations; (2) polymorphic B-cell hyperplasia and polymorphic B-cell lymphoma: may arise in lymph nodes or various extranodal sites, are nearly always monoclonal, usually contain a single form of EBV, and lack oncogene and tumor suppressor gene alterations; and (3) immunoblastic lymphoma or multiple myeloma: present with widely disseminated disease, are monoclonal, contain a single form of EBV, and contain alterations of one or more oncogene or tumor suppressor genes (N-ras gene codon 61 point mutation, p53 gene mutation, or c-myc gene rearrangement). The PT-LPDs are divisible into three categories exhibiting distinct morphologic and molecular genetic characteristics. Alterations involving the N-ras and c-myc proto-oncogenes and the p53 tumor suppressor gene may play an important role in the development and/or progression of the PT-LPDs.

Similarities have been noted in the histologic patterns of lymphomatoid granulomatosis and Epstein-Barr virus associated lymphoproliferative disease involving the lung. Epstein-Barr virus has also been identified by polymerase chain reaction in most cases of lymphomatoid granulomatosis; however, the precise cellular localization of Epstein-Barr virus sequences has not been extensively studied. We analyzed 10 cases of lymphomatoid granulomatosis involving the lung by immunohistochemistry and combined immunohistochemistry with in situ hybridization for Epstein-Barr virus, CD20, and CD45RO. All cases were selected from the files of the Armed Forces Institute of Pathology and met the clinical and histologic criteria for the diagnosis of lymphomatoid granulomatosis, grades 1 through 3. In all 10 cases, immunohistochemistry showed that most of the cells--small to medium-sized lymphocytes--were T cells (CD45RO+); however, a much smaller population of medium-sized to large atypical cells were B cells (CD20+). In each case, combined immunohistochemistry and in situ hybridization confirmed the presence of Epstein-Barr virus sequences within B (CD20+) cells and the absence of Epstein-Barr within T-cells (CD45RO+). Polymerase chain reaction analysis for immunoglobulin heavy-chain gene rearrangement identified a monoclonal pattern in six of nine cases tested, whereas analysis for T-cell receptor gamma-chain gene rearrangements was negative in three cases tested. On the basis of these findings, we hypothesize that most cases of lymphomatoid granulomatosis involving the lung represent a proliferation of Epstein-Barr virus infected B-cells with a prominent T-cell reaction and vasculitis, distinguishing these cases from angiocentric "T-cell lymphomas" in other sites, such as the head and neck.

The histopathologic features of tissue specimens from 16 patients with acute Epstein-Barr virus-induced infectious mononucleosis, which was confirmed by clinical and serologic methods, are described. The clinical course was usually self-limited (14 patients), but it resulted in the death of two patients, one of whom (patient with renal transplantation) was immunosuppressed. Each lymphoid tissue specimen, including those obtained from the lymph nodes (n = 9), tonsils (n = 5), spleen (n = 1), and appendix (n = 1), showed a nonuniform expansion of nonfollicular areas by a polymorphous population of lymphocytes, including transformed lymphocytes and immunoblasts. In situ hybridization demonstrated Epstein-Barr virus-infected lymphocytes in four of eight tissue specimens that were studied. Other histologic features included Reed-Sternberg-like cells, plasma cells, histiocytes, frequent mitoses, abundance of postcapillary venules, and necrosis. These histologic features should suggest a diagnosis of infectious mononucleosis rather than other processes, either benign or malignant, that can mimic it.

We examined 81 cases of Hodgkin's disease for evidence of the t(14;18) translocation, using the polymerase chain reaction assay on lysates of formalin-fixed, paraffin-embedded tissue. Seven of 74 amplifiable cases (9%) were positive for the translocation, which involves the bcl-2 oncogene and the immunoglobulin heavy chain gene. Two of these cases were sequenced and the breakpoints had the same pattern found in follicular lymphoma. The nuclei from one of the cases were sorted into large and small subpopulations. The t(14;18) signal was more intense in the large nucleus subpopulation, which contained a greater proportion of Reed-Sternberg-like nuclei. These results are consistent with the hypothesis that Reed-Sternberg cells carry the translocation, but they do not exclude the possibility that the translocation is found in cells representing the reactive component of Hodgkin's disease. The results also demonstrate that routinely processed material is suitable for polymerase chain reaction-based analysis of translocations, although the sensitivity is reduced 10- to 100-fold, compared with fresh tissue.

In this session there seemed to be general agreement on the existence of anaplastic large cell lymphoma (ALCL) as an entity defined by a characteristic morphology and by diffuse expression of the Ki-1 (CD30) antigen. The discussion indicated the lack of specific immunophenotypic and genotypic markers for such a neoplasm and the variability of the clinical patterns associated with it: these include a childhood form, an adult cutaneous form, and an adult nodal disease. While typical cases of ALCL are clearly distinct (by pathologic, cytogenetic, and clinical criteria) from Hodgkin's disease (HD), there is a variety of histologic and immunophenotypic patterns that overlap those of ALCL and HD; most of these would be classified as HD, lymphocyte depletion (LD) or nodular sclerosis (NS), syncytial subtype. No agreed-upon criteria were found that could consistently define these patterns, nor was an agreement possible on whether they are part of a continuum unifying ALCL and HD or phenotypically similar expressions of different diseases.

The value of histopathologic parameters in predicting the long-term overall survival probabilities was studied in a series of 123 patients with pathologic stage IA, IB, IIA, IIB, or IIIA Hodgkin's disease, nodular sclerosis type who were treated with curative radiation therapy. The parameters that were studied included the relative proportion of atypical vs reactive cells, amount of eosinophils, presence of necrosis, degree of mitotic activity, intensity of different types of mesenchymal reactions, classification in three subtypes (ie, lymphocyte predominance, mixed cellularity, and lymphocyte depletion) or in two grades (ie, grades 1 and 2), and identification of the syncytial variant. For each parameter, the association with clinical risk factors was also analyzed. The results of this study show that there are no pathologic features that carry a significant predictive value of the overall survival.

Both immunophenotypic overlaps between Hodgkin's disease (HD) and non-Hodgkin's lymphoma (NHL), and evolution of one into the other have been reported. However, the underlying assumption that the antigenic expression of Reed-Sternberg (RS) cells is consistent in the same patient has not been evaluated. Such an evaluation was undertaken by immunophenotyping paraffin-embedded lymphoid tissue biopsies with HD from 56 patients in whom multiple specimens were obtained, either simultaneously from different sites or at different times. The panel of antibodies we used included: CD3 polyclonal antiserum, DAKO-M1 (CD15), L26 (CD20), BerH2 (CD30), MT1 (CD43), DAKO-LCA (CD45RB), UCHL1 (CD45R0), LN2 (CD74), and DAKO-EMA. The phenotype of RS cells was identical in simultaneous biopsies in only 11 of 39 patients (28%) and remained constant in consecutive biopsies in only 4 of 21 patients (19%). Major differences (relative to cell lineage specific antigens) were observed in 10 of 39 patients with simultaneous biopsies and in 10 of 21 patients over time; they mainly involved expression of T-cell antigens. Minor differences (relative to any other antigen) were observed in 22 of 39 patients with simultaneous biopsies and in 15 of 21 patients over time; these mainly involved CD15 or CD74. This striking variability of the immunophenotype of RS cells in the same patient may be due to aberrant marker expression, as a result of the neoplastic state, and/or to modulation of antigenic expression in relation to the host environment. This inconsistency suggests caution when interpreting the relationship between HD and NHL by paraffin immunophenotyping alone.

The authors examined paraffin-embedded lymph node biopsies from 65 cases of Hodgkin's disease for the presence of Epstein-Barr virus (EBV) DNA, using the highly sensitive polymerase chain reaction technique. Overall 40% of the cases were positive for EBV DNA; there were no statistically significant differences in the frequency of EBV positivity among the different subtypes of Hodgkin's disease. These results are in agreement with those of previous studies that employed less sensitive detection techniques and suggest that EBV either is present in pathologic tissues only in some phases of the evolution of Hodgkin's disease or is a pathogenetic factor involved in only a portion of cases.

The association of recurrent chromosomal abnormalities with certain types of non-Hodgkin's lymphomas (NHLs) provides important clues to their classification and pathogenesis. The t(11;14), an uncommon abnormality, has been reported in a variety of B-cell lymphomas, including the intermediately differentiated lymphocytic lymphoma (IDLL). The authors report here the histologic, immunophenotypic, and cytogenetic findings in four additional patients with such translocation. They represent 8% of 51 patients with NHL studied by cytogenetics in a provincial hospital of Western mainland Japan. Three of the patients had IDLL and one a follicular lymphoma of large cell type. These findings strengthen the association between t(11;14) and IDLL. Furthermore, the occurrence of t(11;14) in a follicular lymphoma, as the sole abnormality, along with histogenetic considerations, suggests the possible existence of a rarer subset of follicular lymphoma, not associated with the t(14;18) and histogenetically related to IDLL.

Clonal chromosomal abnormalities were found in tumor tissue of 43 (84%) of 51 patients with non-Hodgkin's lymphoma (B-cell, 32; T-cell, 15) from an adult T-cell leukemia/lymphoma-nonendemic area in western mainland Japan. Four tumors were tetraploid, and the other 39 had a chromosome number in the diploid range. Trisomies 3, 5, 7, 18, and X, monosomy 13, and loss of an X in female and a Y in male were found in more than three patients each. Structural abnormalities in arms 1p, 1q, 2q, 3q, 13q, 14q, and 18q were found in eight or more patients each. Clustering of breaks occurred in 3q25-29 (ten patients, nine of whom with a B-cell tumor), 11q13 (five patients), dir dup(12)(q13-15----q21-24) (four patients), 14q32 (12 patients), and 18q21 (seven patients). The 14q32 translocations were all associated with B-cell tumors. t(8;14) was seen in a small noncleaved cell lymphoma, t(11;14)(q13;q32) in one follicular and three intermediately differentiated lymphocytic lymphomas, and t(14;18)(q32;q21) in two follicular lymphomas and one diffuse small cleaved cell tumor. The translocation partner of 14q could not be determined in the other four patients, three of whom had der(18)t(18;?)(q21;?). The seven 18q21 abnormalities, including a novel translocation t(2;18)(p11;q21.3), all occurred in B-cell tumors; even in the absence of t(14;18), they were closely associated with lymphomas of follicular center cell origin (six of seven).

In rare cases, primary malignant lymphomas may arise in the soft tissues. Only one previous case has arisen in the context of chronic lymphedema. Because of the clinical appearance of such lesions, which resemble violaceous nodular or plaquelike tumors, they may be confused clinically with lymphedema-associated angiosarcomas occurring after radical mastectomy (Stewart-Treves syndrome). Furthermore, the histologic appearance of some lymphomas and angiosarcomas may also be similar. We studied two women with primary postmastectomy lymphedema-related malignant lymphoma in the soft tissues of the upper arm. These tumors arose 11 and 30 years, respectively, after radical removal of ductal mammary carcinomas. Histologically, one neoplasm mimicked metastatic carcinoma or epithelioid angiosarcoma; whereas the other was initially confused with a variety of pathologic entities, including vasculitis, epithelioid hemangioma, and malignant fibrous histiocytoma. The lymphoid nature of both lesions was confirmed by immunoreactivity for leukocyte common antigen in addition to the B-lymphocyte marker, L26. Conversely, vascular and epithelial determinants were absent. One patient's disease pursued an indolent course; she died of unknown causes but with no evidence of lymphoma at last follow-up. The second patient is currently in remission on chemotherapy. Awareness of the existence of lymphedema-related malignant lymphoma and familiarity with methods used for its distinction from epithelioid vascular sarcomas should prevent unnecessary surgery.

Pretreatment lymph nodes, bone marrow, and blood were examined in 176 cases of non-Hodgkin's lymphoma. By the criteria of the Lukes and Collins functional--morphological classification, 158 (90%) were B-cell lymphomas and 17 (10%) were T-cell lymphomas. Bone marrow involvement was present in 53% of cases: 51% of B-cell types and 65% of T-cell types. Marrow involvement was most frequent in small lymphocyte (B) (89%), convoluted lymphocyte (60%), and small cleaved follicular center cell (FCC) lymphomas (55%). The pattern of bone marrow involvement was most frequently focal paratrabecular in B-cell lymphomas and diffuse in T-cell lymphomas. Blood involvement was present in 50% of cases with bone marrow lymphoma and generally reflected extensive bone marrow disease. There was a higher incidence of both bone marrow and blood involvement in pediatric patients than in adults.

Six cases of follicular lymphoma contained an abundant plasma cell component. With immunoperoxidase techniques, this was found to demonstrate monotypic cytoplasmic marking for either K or L Ig light chain in five cases, and for IgA heavy chain only in one case. A histogenetic relationship between follicular center cells and plasma cells was suggested by cell forms morphologically intermediate between these two types and by monotypic plasma cells in the neoplastic follicles. The progressive differentiation of follicular center cells into Ig-secreting cells in these cases is likely to be the result of an alteration of the immunoregulatory mechanisms that usually block the differentiation of follicular lymphomas. Four of our patients presented with disseminated disease, three had extranodal presentation and four manifested serum paraproteins. Their median survival was 40 months; two of them died of disease. The published data and our own suggest that follicular lymphoma with plasmacytic differentiation is a malignancy of intermediate grade, with survival and clinical features closer to lymphoplasmacytic/lymphoplasmacytoid lymphoma (LP immunocytoma) than to follicular lymphoma.

Blue naevi are reported for what is believed to be the first time in the capsule and fibrous trabeculae of lymph nodes. In one of the two cases that nodal blue naevus was accompanied by an ordinary blue naevus in the regional skin. These blue naevi are compared with the 'naevus-cell aggregates' in nodal capsule, the histogenesis of which is currently disputed. A developmental migratory arrest of melanocytes is thought to be the histogenetic mechanism involved in the production of these blue naevi in lymph nodes.

The clinical, morphologic, and immunohistochemical features of 5 cases of peripheral T-cell lymphoma with lymph node and skin involvement are reported. All patients were male and presented with generalized lymphadenopathy and skin lesions. Lymph node biopsy specimens showed a diffusely growing, pleomorphic, atypical lymphoid infiltrate composed of markedly irregular small and large cells. The stroma consisted of plasma cells, eosinophils, epithelioid histiocytes, and an increased number of small blood vessels. A similar neoplastic infiltrate was observed in the cutaneous nodules that developed in three patients. Immunohistochemistry on frozen tissue sections of lymph nodes using a broad panel of monoclonal antibodies revealed that the neoplastic cells were reactive with anti-T-cell antibodies, the majority of them belonging to the helper/inducer T-cell subset. Stromal plasma cells contained polyclonal immunoglobulins. It is concluded that the malignant lymphoma of peripheral helper/inducer T-cell origin observed in these five patients represents a variant of a mixed, blastic/cell T-cell lymphoma, characterized by a peculiar stromal component and skin involvement.

The identification of recurring chromosomal translocations has provided clues to the gene regions important in lymphoma development. Among 157 patients with non-Hodgkin lymphoma studied by cytogenetic analysis, four new recurring translocations have been identified--t(8;9) (q24;p13), t(11;18)(q21;q21), t(14,15)(q32;q15), and an unbalanced translocation giving rise to der(22)t(17;22) (q11;p11). Each translocation appeared twice. The t(11;18) was the only karyotypic abnormality in the two patients with it, and the t(14;15) was the sole karyotypic abnormality in one patient. All translocations were found in B-cell malignancies and were associated with both nodal and extranodal disease. Among the regions affected, only the immunoglobulin heavy- chain gene MYC, and BCL2, have thus far been associated with lymphoma. The breakpoint sites identified by these translocations warrant further investigation at the molecular level.

Several reports in the literature have documented an association of Langerhans' cell granulomatosis (LCG) with malignant lymphoma (ML). Seven additional patients are described, and different aspects of such an association are demonstrated. The diagnosis of LCG, made on the basis of routinely stained biopsy material, was supported by S-100 protein immunoperoxidase reactivity and/or the presence of Birbeck granules on electron microscopy. In six patients (cases 1 to 6) the LCG consisted of localized foci in nodes involved by ML. However, in one of these patients (case 1) focal LCG was again associated with ML in a later lymph node biopsy specimen from a different site, and in another patient (case 6) LCG subsequently developed in the skin and lungs. In the remaining case (case 7) pulmonary LCG developed one year after the diagnosis of Hodgkin's disease. Localized LCG in a lymph node involved by ML could represent an immune response to ML (cases 1 to 6 in the present study and 16 literature cases). The findings in case 6 indicate that such a response has the potential to become disseminated. Cases in which LCG and ML are found at different sites (case 7 in the present study and five literature cases) probably represent the coincidental association of two unrelated disorders. Finally, it has been postulated that, in rare instances, the development of Hodgkin's disease or malignant histiocytosis in a patient with LCG might represent malignant evolution of this hyperplastic process.

We report a novel chromosome translocation--t(2;5)(p23;q35) or its variant, t(2;5;13)(p23;q35;q12)--in 3 patients with peripheral T-cell lymphoma. All 3 were female children who had peripheral lymphadenopathy without organomegaly and underwent complete remission with or without chemotherapy. Their tumors were characterized histologically by predominant large cells, at times showing phagocytosis, and immunologically by peripheral T-cell phenotype and expression of Ki-1 antigen and epithelial membrane antigen (EMA). Since the same translocation has been reported in the literature in 4 patients with malignant histiocytosis (MH), and our patients had histologic features suggestive of that disease, it is likely that many tumors previously interpreted as MH are actually phagocytic large T-cell lymphoma carrying this translocation.

Aldehyde-fuchsin, without prior oxidation, was noted to stain the insularβ-cell granules of pancreas sections briefly pretreated with trypsin. Furthermore, after that procedure, insularβ-cells were more intensely basophilic than in control sections not previously subjected to tryptic digestion.

In order to test the immunologic validity of the Lukes-Collins (L-C) classification of the non-Hodgkin's lymphomas (ML), 60 ML in adult patients were studied and their B- or T-cell nature was predicted on a morphologic basis, without knowledge of the clinical history or the results of surface marker (SM) studies (SIg, C', Fc and E-rosettes), which were performed on cell suspensions and cryostat sections from the same specimens used for histology. There was a good correlation between morphologic types and SM (97% for the nodular ML, 81% for the diffuse). The predictions were not confirmed in 6 instances: one nodular ML typed as T; one convoluted lymphocytic ML typed as histiocytic; and four diffuse ML, predicted to be of B-type, bore no detectable SM ("Null" ML). It is concluded that the L-C morphologic criteria do allow in most cases the recognition of the B- or T-cell nature of ML, but cannot detect variations of the SM pattern which seem to affect the clinical behavior of these neoplasias. Thus, while the L-C classification seems immunologically sound, its clinical relevance still needs to be demonstrated.

In a chromosome study of childhood lymphoblastic lymphoma, we found a novel translocation, t(9;17)(q34;q23), in three patients. They presented with mediastinal mass and no bone marrow involvement. Despite intensive chemotherapy, one patient had no response, the other two relapsed after a brief remission, and all progressed to death. The 9;17 translocation may have a clinical implication for lymphoblastic lymphoma patients in predicting a poor prognosis. Since, in addition to our cases, involvement of the 9q34 breakpoint, together with 2q33, 14q11, or 7q34, has been reported in the literature in four lymphoblastic lymphoma patients, a gene located in 9q34 and referred to as tcl-3 may participate in the genesis of the T cell malignancies carrying these translocations. Furthermore, as is the case in other lymphomas, the reciprocal breakpoint, 17q23, might be the site of a yet unidentified T cell function gene.

The lymphoproliferative processes that developed in five renal transplant recipients were studied in an attempt to characterize and classify them morphologically. Nine surgical specimens, hematological material on all patients, and autopsy specimens from three patients were available. Studies performed included: conventional histopathology; evaluation of cell markers (immunoglobulins and sheep erythrocyte, complement, and Fc receptors) and cytoplasmic immunoglobulins (peroxidase-antiperoxidase technique); ultrastructural examination; and karyotype analysis. The lymphoid lesions in our patients shared marked cytological polymorphism (small and large cells, of both follicular center and "medullary" type) and polyclonal B-cell features, which indicated a common reactive nonneoplastic origin. However, other features, such as morphological atypia of the immunoblasts, extensive necrosis, chromosomal aberrations, and an incipient monoclonal component suggested the development of lymphoma in some of these lesions. In contradistinction, the abundance of typical immunoblasts was a feature that seemed to correlate with the clinical activity of the disease rather than with the biological malignancy. The multiplicity of B-cell types and the presence of a follicular center cell component with diffuse distribution, as well as the extensive necrosis in the malignant forms, seem to differentiate morphologically the lymphoproliferative processes arising in transplant recipients from both the hyperplasias and the lymphomas developing in immunologically normal hosts. For the former, we propose the terms of "polymorphic diffuse B-cell hyperplasias" and "polymorphic B-cell lymphomas."

G-banded chromosomes were studied from involved lymph nodes or other tumor masses in 94 patients with malignant lymphoma. Clonal chromosome abnormalities were identified in 91 patients including all 81 B-lymphomas but only 6 of 9 T-lymphomas. Many recurring chromosome abnormalities were found. Most common numerical alterations involved gains of chromosomes 12 (19% of patients), 18 (13%), 7 (12%), and 21 (10%). Structural abnormalities, which were more frequent than numerical alterations, most commonly involved chromosome regions 14q (71% of patients), 18q (36%), 6q (31%), 1p (24%), and 8q (19%). Seven recurring translocations were identified, and all except one involved 14q32. The most frequent were t(14;18)(q32;q21) in 22 patients, t(8;14)(q24;q32) in 9 patients, and t(1;14)(q42;q32) in 3 patients. Deletions most frequently involved the long arm of chromosome 6 at band q21 (11 patients) or q23 (7 patients). The common recurring chromosome abnormalities were correlated with histology (International Working Formulation for Clinical Usage) and immunological phenotype. Four abnormalities were significantly associated with specific histologies. Eighteen (82%) patients with t(14;18)(q32;q21) were follicular. Similarly, 82% of patients with del(6)(q21) were large cell lymphomas. Lymphomas with trisomy 7 were either diffuse large cell or follicular, while patients with t(8;14)(q24;q32) were either diffuse large cell or small noncleaved cell. A significant association with immunological phenotype was seen for t(14;18) only. All patients were either B- or complement lymphomas, and the heavy chain(s) was more commonly gamma and less frequently delta mu than among the total B-lymphoma population. We conclude that essentially all lymphomas have cytogenetic abnormalities; further study is required to determine their significance. Particularly, it will be of interest to see if oncogenes are found in the regions of these chromosome abnormalities.

Although many recurring chromosome abnormalities have been found in malignant lymphoma (ML) in recent years, their relationship to histology remains largely undefined. We have correlated, in the same tumor mass, chromosome findings with histology, defined by the International Working Formulation for Clinical Usage, in 120 patients. We find differences among histologies in the frequency of normal metaphases and the modal number of the predominant abnormal clone. In addition, most histologies have been significantly (P less than .01) associated with specific chromosome abnormalities. In particular, ML, follicular, predominantly small cleaved cell was associated with t(14;18)(q32;q21); ML, follicular, mixed small cleaved cell and large cell with t(14;18)(q32;q21) and trisomy 8; ML, follicular, predominantly large cell with trisomy 7 and breaks in 17q21-q25; ML, diffuse, mixed small cell and large cell with breaks in 11p; ML, diffuse, large cell with trisomy 21 and breaks in 2q and 9q; ML, large cell, immunoblastic with breaks at 6q21; and ML, small noncleaved cell with t(8;14)(q24;q32). Only the associations with t(14;18) and t(8;14) have been previously reported. The associated chromosome abnormality usually occurred in 30% to 70% of a given histology, raising the possibility that cytogenetics may add important prognostic information in lymphoma as it does in the acute leukemias.

Antibodies to leukocyte common antigen (LCA) are valuable in surgical pathology in the diagnostic separation of malignant lymphomas from poorly differentiated neoplasms of other types. However, several publications have reported difficulty in obtaining adequate LCA labeling in paraffin-embedded specimens. To assess this problem further, we applied an amplified version of the avidin-biotin-peroxidase complex procedure to 315 formaldehyde-fixed hematopoietic malignancies, and 420 nonhematopoietic tumors that had been similarly processed. All non-Hodgkin's lymphomas were LCA-reactive with this method, whereas none of the nonhematopoietic neoplasms were stained. Twelve of 25 cases of Hodgkin's disease displayed LCA-positivity in Reed-Sternberg cells, but all contained reactive, benign inflammatory cells. Hence, the specificity of the amplified anti-LCA/avidin-biotin-peroxidase technique was 100%, and its sensitivity was 96%. This procedure should allow surgical pathologists to obtain reliable, reproducible staining of hematopoietic neoplasms in paraffin-embedded tissues.

Eight cases of follicular lymphoma (FL) were found amount 318 children with non-Hodgkin's lymphoma seen in two decades in four large institutions (overall incidence: 2.5%). The children's ages ranged from 3 to 13 years (median 8). Four patients presented with localized peripheral lymphadenopathy, two with tumors of the digestive tract, two with disseminated disease. Five were tentatively classified as stage I or II and 3 as stage IV. The existence of other diseases responsible for lymphadenopathy could satisfactorily be exclued. All patients are alive after follow-up periods of 1 to 14 years from diagnosis (median 4). Morphologically, 5 lymphomas were mixed and 3 histiocytic. The growth pattern was "expansile" in the younger patients (3 to 9 years), and "infiltrative," as in the adult disease, in the three older children (11 to 13 years). The histiocytic cytology correlated with stage IV disease. FL in children appears to be different from both its adult counterpart and the diffuse childhood lymphomas. Differences with the former include the absence of tumors of poorly differentiated lymphocytic type, the higher frequency of stage I-II disease and the better prognosis. This last feature, as well as the higher frequency of peripheral node involvement and the absence of leukemic conversion or CNS disease, differentiates the follicular from the diffuse childhood lymphomas.

The recent report of an immunoblastic lymphadenopathy (IBL)-like T cell lymphoma has rekindled questions about the nature, reactive or neoplastic, of IBL, angioimmunoblastic lymphadenopathy (AIL), and lymphogranulomatosis X (LgX) and blurred the criteria for their diagnosis. We looked in the literature and our own data for a categorization of AIL (IBL, LgX) and related disorders, needed for future prospective studies. Specific differences in the original histologic definitions and discordant immunophenotypic data may warrant the separate consideration of AIL, IBL and LgX and their subdivision into predominantly T cell or B cell lesions. DNA hybridization and cytogenetic studies of the processes sharing histologic features of AIL (IBL, LgX) demonstrate a continuum of disorders from purely reactive to frankly malignant, which may be categorized as follows: (1) those without evidence of clonality by any of three parameters (immunophenotypic, immunogenotypic, and cytogenetic), for which only the term AIL (IBL, LgX) might be reserved; (2) those with evidence of clonality by all parameters, or AIL (IBL, LgX)-like lymphomas; and (3) those that, due to any discordance among the three parameters, do not fit into either of the above categories, and for which the term AIL (IBL, LgX)-like dysplasias is proposed. This intermediate group seems to be composed of unstable lymphoproliferative conditions, in which a predominant component of normal cells coexists with clonal population(s) that may either disappear with time or selectively proliferate and develop into frank lymphoma.

The identification of recurring chromosomal translocations has provided clues to the gene regions important in lymphoma development. Among 157 patients with non-Hodgkin lymphoma studied by cytogenetic analysis, four new recurring translocations have been identified--t(8;9) (q24;p13), t(11;18)(q21;q21), t(14,15)(q32;q15), and an unbalanced translocation giving rise to der(22)t(17;22) (q11;p11). Each translocation appeared twice. The t(11;18) was the only karyotypic abnormality in the two patients with it, and the t(14;15) was the sole karyotypic abnormality in one patient. All translocations were found in B-cell malignancies and were associated with both nodal and extranodal disease. Among the regions affected, only the immunoglobulin heavy-chain gene MYC, and BCL2, have thus far been associated with lymphoma. The breakpoint sites identified by these translocations warrant further investigation at the molecular level.

The incidence, pattern, extent, and morphology of bone marrow involvement in 176 cases of non-Hodgkin's lymphoma (NHL) were studied in relationship to the Lukes--Collins classification. Ninety percent of the cases were B-cell lymphomas; 10% were T-cell lymphomas. In 53% of cases there was bone marrow involvement by lymphoma at diagnosis. Marrow involvement was most frequently found in the small lymphocyte (B), small cleaved follicular center cell (FCC), and convoluted lymphocyte lymphomas. Frequently, the extent of bone marrow biopsy replacement by lymphoma was less than 30%; the pattern of infiltration was predominantly focal (70%). Cytologic agreement between lymph node and bone marrow specimens was always present in small lymphocyte (B), small noncleaved FCC, convoluted lymphocyte, and lymphoepithelioid cell lymphomas. Cytologic diversity between lymph node and marrow was noted in 20% of small cleaved FCC, 40% (2/5) of large cleaved FCC, and 38% (3/8) of large noncleaved FCC lymphomas. In 79% of all involved cases, both bone marrow biopsy sections and aspirate smears were diagnostic of NHL; only biopsy sections were positive in 18%, and only smears were positive in 3%. The Lukes--Collins classification predicts a high incidence of bone marrow involvement for small lymphocyte (B), small cleaved FCC, and convoluted lymphocyte lymphomas.

Lymph node biopsies in a patient with follicular lymphoma showed rosette structures as seen in neuroepithelial neoplasms. These specimens were studied by histologic, immunoperoxidase (for immunoglobulins, intermediate filaments (IF), actin and neuron-specific enolase), immunofluorescence (for immunoglobulins, and with a panel of monoclonal antibodies), and electron-microscopic examination. The rosettes were formed by neoplastic lymphocytes arranged around eosinophilic fibrillary material. Ultrastructurally, this was composed of cytoplasmic processes, projecting from the lymphocytes and containing thin and intermediate filaments. Immunohistochemically, it stained for monoclonal IgM lambda, all other antigens present on the neoplastic cells, and weakly for vimentin and actin. Based on recent information about lymphocyte surface changes, it is speculated that the rosettes might represent an aggregation of neoplastic lymphocytes activated by a microenvironmental stimulus, perhaps antigen-antibody binding at the cell membrane. The practical implication of this hitherto unreported finding is that the presence of rosettes cannot be used to rule out a lymphoma.

Several steps in the clinical evolution of human neoplasia are associated with a variety of recurrent chromosomal defects that could prove essential to the understanding of cancer. We found 15 types of nonrandom chromosomal abnormalities in a study of 71 patients with follicular lymphoma; 10 of the types appeared to influence the histopathological findings, clinical course, or response to treatment. A translocation, t(14;18), observed in 85 percent of all patients appeared to be the main determinant of a follicular pattern. Ten patients with a t(14;18) as a single defect had the histologic features of follicular small cleaved-cell lymphoma. Most did not require treatment for one to four years, because their tumors had an initial indolent course. In contrast, patients with follicular small cleaved-cell lymphoma with t(14;18) and deletion 13q32 acquired the hematologic features of leukemia and had an acceleration of the disease. A deletion 6q together with a complete or partial trisomy 7 or trisomy 12 (or both) was associated with the clinically more aggressive follicular mixed small- and large-cell or large-cell histologic type, which often evolves from follicular small-cell lymphoma. A complete or partial trisomy 3, 18, or 21 correlated almost exclusively with follicular large-cell lymphoma. In all follicular stages, a trisomy 2 or duplication 2p often accompanied an accelerated clinical course and a poor response to treatment. This study suggests that several discrete genomic defects may govern the evolution of a patient's malignant disease.

Conditions were determined for successful subcutaneous tumor development of MOLT-4 human T-cell leukemia cells in BALB/c nude (nu/nu) female mice. MOLT-4 cells injected alone at one site resulted in tumors in none of 21 mice, whereas MOLT-4 cells injected with X-irradiated human fibrosarcoma cells (HT-1080) at another site resulted in tumors in 5 of 21 mice. When X-irradiated three times with 200 rad over a 3-week period and then inoculated with MOLT-4 cells and X-irradiated HT-1080 cells, 19 of 25 mice developed tumors, which indicated that a combination of X-irradiation of the recipients and inoculation of an admixture of the leukemia cells with X-irradiated fibrosarcoma cells markedly increases tumor incidence. The cells growing in the tumors had the characteristics of MOLT-4, as assessed by marker studies and histology.

Lymph node biopsy specimens from nine patients with a systemic lymphoproliferative disorder morphologically resembling Castleman's disease were studied for the presence of cytoplasmic immunoglobulins. After trypsinization of tissue sections, the avidin-biotin-peroxidase method was used to stain for IgG, IgA, and IgM heavy chains and kappa and lambda light chains. All cases showed polyclonal proliferation of plasma cells. The predominant immunoglobulin class observed was IgG. The IgM-containing cells tended to be localized concentrically around germinal centers. No notable changes were seen in multiple biopsy specimens from five patients. Polyclonal staining was also observed in lymph nodes obtained at autopsy from a patient who died of a disseminated atypical immunoproliferative process. The polyclonal staining pattern in these cases confirms the reactive nature of this systemic lymphoproliferative disorder.

A case of Castleman's disease of the plasma-cell type is reported in which central nervous system (CNS) involvement was diagnosed by cerebrospinal fluid (CSF) cytology. The patient had multicentric disease with constitutional symptoms, immunologic abnormalities and peripheral blood cytopenias requiring cytotoxic agents and steroids for treatment. CNS symptoms and diagnostic cytologic findings in CSF occurred in the absence of morphologic lesions demonstrable by computed tomography or magnetic resonance imaging of the brain.

Ten newly diagnosed patients, with localized diffuse histiocytic lymphoma classified according to the Lukes-Collins classification, were evaluated for response to treatment with extended-field radiation. All classified tumors were of follicular center cell type (four large cleaved, four large noncleaved, and one small noncleaved). There were nine patients with stage I disease and one with stage II disease. The patient with stage II disease had only two sites of involvement. The disease was nodal in six patients and extranodal in four. All patients are alive without evidence of disease recurrence from 19 to 102 months since diagnosis (median survival, 58 + months). No differences were found in response or survival rates between histologic categories. No significant immediate or long-term complications have occurred. These results suggest that radiotherapy alone is curative for localized diffuse histiocytic lymphoma and may be superior to recently advocated chemotherapeutic approaches.

In this study of 33 T cell acute lymphoblastic leukemia (T ALL) and 17 lymphoblastic lymphoma (LBL) patients, no relevant differences between the two groups were observed in clinical characteristics, response to therapy, and survival. We found translocations involving 14q11, 7q35, or 7p15, where T cell receptor alpha and delta, beta, and gamma subunit genes reside, in 20 patients (40%). Most of these translocations were seen with equal frequency in T ALL and LBL, indicating that, in a large proportion, the two diseases are different manifestations of the same lymphoblastic disorder. However, other translocations, such as t(9;17)(q34;q23), occurred only in LBL, perhaps pointing to the existence of subsets of LBLs that are distinct from T ALL. On the basis of karyotype, 50 patients could be classified into three groups: 20 patients with 14q11, 7q35, or 7p15 translocations (group A); 20 with other translocations, and/or deletions (group B); and 10 with normal diploidy (group C). There was no difference in survival time between any two of the three groups.

Immunoglobulin gene rearrangement analysis is a sensitive method for determining clonality of B cell proliferations. We have examined tissue obtained from five renal and one cardiac allograft recipient with Epstein-Barr virus-associated B cell proliferations for immunoglobulin gene rearrangements. Biopsies from two patients with lesions that were hyperplastic morphologically, polyclonal by cellular immunoglobulin staining, and had diploid karyotypes, had no detectable gene rearrangements and were, therefore, consistent with benign reactive processes. These patients are alive without evidence of disease 37 and 57 months after diagnosis. In a biopsy from one patient with a lesion that was malignant lymphoma morphologically, monoclonal by cellular immunoglobulin staining, and had clonal cytogenetic abnormalities, clonal gene rearrangements were detected in a majority of cells, confirming their neoplastic nature. In biopsies from an intermediate group of three patients with morphologically malignant proliferations that were composed predominantly of a polyclonal population of B cells, clonal gene rearrangements were also found, consistent with early malignant transformation in a subpopulation of cells. These findings confirm the heterogeneity of the posttransplant EBV-associated lymphoproliferative diseases and have significant implications for our understanding of the pathogenesis of EBV-induced infections and related lymphomas.

A histologic review was undertaken of 35 lymphoreticular disorders that developed in primary immuno-deficiency patients from the Immunodeficiency Cancer Registry. Twenty-one (60%) of the lesions were non-Hodgkin's lymphomas: these included eight B-immunoblastic sarcomas. Eight (23%) of the lesions were Hodgkin's disease, with a high frequency of lymphocytic depletion type in an unusually young age group. Three lesions (8.5%) represented abnormal proliferative processes, which could not be definitely categorized as either benign or malignant. There were only two acute lymphoblastic leukemias (6%). Differences were found between lymphomas arising in Wiskott-Aldrich syndrome and those occurring in ataxia-telangiectasia; this suggests that different pathogenetic mechanisms might operate in their development. The lymphomas in Wiskott-Aldrich syndrome were all of non-Hodgkin's type, predominantly B-immunoblastic sarcomas, and presented as localized extranodal infiltrates. The lymphomas in ataxia-telangiectasia were either Hodgkin's disease, mostly of lymphocytic depletion type, or non-Hodgkin's lymphomas of the histologic subtypes associated with 14q translocations.

A new disease with a lymphoma-like Summary clinical presentation and a specific histological picture has been recognised in fifteen patients. Clinically, the disease, which seems to occur chiefly in the elderly, is characterised by an acute onset of constitutional symptoms, generalised lymphadenopathy, hepatosplenomegaly, and immunological abnormalities. The course is stormy, but in one third of the reported cases it may have been controlled by immunosuppressants; severe infectious complications are common and often fatal. Histologically, diffuse obliteration of the nodal architecture due to pronounced proliferation of small vessels and immunoblasts is the distinctive feature; there is similar proliferation in the spleen, bone-marrow, liver, and skin. Biopsy specimens and necropsy material indicate that the disease is a non-neoplastic process. The histological evolution, the widespread organ involvement, and the clinical pattern are very similar to those of a graft-versus-host reaction.

Independent review of slides of 668 cases of non-Hodgkin's lymphoma (NHL) by a panel of 4 experienced pathologists using the Working Formulation (WF) allowed determination of the agreement between reported diagnoses and panel review of slides. The panel agreed with the reported diagnosis of NHL in 93% of cases, but with the NHL subtype in only 55% of cases overall. The ability of the panelists to agree among themselves, however, was only slightly better than the panelists' agreement with the reported diagnosis (60% vs. 54%, respectively). Agreement of the panel with the reported subtype diagnosis varied from 14% to over 90%. The best agreement was with small lymphocytic lymphoma and follicular subtypes. Conclusions from this study are: 1) The WF functions well as common language for translation and comparison of diagnoses of subtypes of NHL. 2) Relative to time and cost involved, panel review using only light microscopy may not be useful in epidemiologic studies of NHL. 3) Small lymphocytic and follicular subtypes of NHL can be used more confidently in epidemiologic studies than can other subtypes whether the subtyping is done from abstracted reports or by panel review.

The clinical and pathologic findings in 24 patients with "angio-immunoblastic lymphadenopathy with dysproteinemia" (AILD) are presented. The patients' ages ranged from 44 to 80 years, with a median age of 68 years. The disease has an acute onset. In many respects, the clinical presentation is suggestive of malignant lymphoma. Generalized lymphadenopathy was always present. Hepatomegaly was found in 20 patients, splenomegaly in 17, constitutional symptoms in 20 and skin rashes in nine. Twenty patients had anemia, with positive Coombs' test in eight of 14 tested. Polyclonal hypergammaglobulinemia was found in 17 of 22 patients. Two patterns of evolution were recognizable: (1) long survival (24 to 67 months) without treatment or after the administration of intensive combination chemotherapy; and (2) rapid progression (one to 19 months) regardless of the treatment given. Sixteen patients died; postmortem examination in 10 cases showed the cause of death to be attributable to severe infection in eight patients, to renal disease in one and to cardiovascular disease in one. No evidence of malignant lymphoma was seen in any of these autopsies. Histologically, the disease is systemic, with specific lesions in the lymph nodes. The spleen, liver, bone marrow, skin and lung are also involved, but the changes are less characteristic than in the lymph nodes. In the patients in whom sequential biopsies were performed, a trend toward restoration of the nodal architecture was observed. AILD is a clinical-pathologic entity in a spectrum of yet to be defined immune reactions. The clinical, laboratory and pathologic manifestations of AILD are consistent with an autoimmune disorder, in which a deficiency of the T-cell regulatory functions probably predisposes to an abnormal proliferative and autoaggressive reaction of the B-cell system. Surgical staging procedures do not appear to be indicated. Intensive cytotoxic treatment may be hazardous in some patients, precipitating their death, but long survival after such therapy has been observed in others. Supportive therapy and small doses of steroids appear to be a safer therapeutic approach.