Lymphoproliferative disease after lung transplantation: comparison of presentation and outcome of early and late cases

BACKGROUND

Post-transplantation lymphoproliferative disease (PTLD) after lung transplantation has not been fully characterized. In previous studies, the incidence has varied substantially, and most cases have been reported during the first year after transplantation. The purpose of this study was to review our center's experience with PTLD and to analyze the pattern of disease and determinants of outcome.

METHODS

Among 494 adult lung (n = 491) or heart-lung (n = 3) recipients, 30 cases of PTLD were retrospectively identified. The cases were classified by site(s) of involvement, histology and time of onset (early, < or =1 year, and late, >1 year after transplantation). The outcome of each case was ascertained, and risk factors for death were analyzed in a multivariate model.

RESULTS

PTLD was identified in 30 (6.1%) of the recipients during 1,687 patient-years (median 2.8 years) of follow-up. The incidence density was 1.8 cases per 100 patient-years. Fourteen cases were diagnosed during the first year after transplantation, and 16 cases in subsequent years. The incidence density was significantly higher in the first year than in later years (3.3 cases/100 patient-years versus 1.3 cases/100 patient years; p <.008). Presentation in the thorax and involvement of the allograft were significantly more common in the early cases (thorax: 12 of 14, 86%; allograft: 9 of 14, 64%) than in the late cases (thorax: 2 of 16, 12%; allograft: 2 of 16, 12%). There was no difference in survival after the diagnosis of PTLD between the early and late cases, but survival time after diagnosis was significantly longer in cases with, than those without, allograft involvement (median 2.6 years vs 0.2 year, respectively; log rank p = 0.007). The presentation and pattern of organ involvement of PTLD after lung transplantation is related to the time of onset.

CONCLUSIONS

Disease in the thorax and involvement of the allograft are common in the first year after transplantation, but other sites, especially the gastrointestinal tract, predominate later. PTLD that is confined to the allograft appears to have a somewhat better prognosis than disease that involves other sites.

Cyclin D1 as an aid in the diagnosis of mantle cell lymphoma in skin biopsies: a case report

Mantle cell lymphoma (MCL), an uncommon and aggressive form of non-Hodgkin lymphoma, typically involves lymph nodes. It usually only secondarily involves extranodal sites. We describe an unusual case of a MCL that presented and relapsed in the earlobes. Light microscopic findings were initially regarded as suggestive of small lymphocytic lymphoma, although subsequent analysis of fresh tissue by flow cytometry led to the diagnosis of MCL. Retrospective application of a broad panel of recently developed markers suitable for analysis of routinely processed tissue yielded results that also permitted a diagnosis of MCL. If these results had been available at the time of initial presentation, they would have obviated the need for rebiopsy. Greater awareness not only of the phenotypic criteria by which lymphomas are classified but of the lymphoma markers available for evaluation of routinely processed tissue should facilitate the accurate diagnosis of diseases like MCL and minimize the risk of misdiagnosis as an indolent disorder.

The spleen as a diagnostic specimen: a review of 10 years' experience at two tertiary care institutions

BACKGROUND

Few studies have examined the yield of the diagnostic splenectomy, and the relevance of these studies to the management of patients with unexplained splenomegaly or a splenic mass are limited by low number of cases, the use of selection criteria, and the lack of modern terminology and modern ancillary studies. The current study correlates clinical intent with preoperative clinical and radiologic studies and histologic findings in an assessment of the diagnostic yield of splenectomy.

METHODS

The medical charts, laboratory data, radiologic studies, and pertinent preoperative biopsies on all patients who underwent splenectomy between the years 1986 and 1995 were reviewed, and the clinical intent behind the procedure was correlated with histologic findings.

RESULTS

One hundred twenty-two of the 1280 patients underwent splenectomy for diagnosis, and in 116 patients a specific disease was identified histologically that explained the splenomegaly/splenic mass; malignancy was the most common cause of unexplained splenomegaly or splenic mass, though benign neoplasms and reactive disorders were documented in 25% of the cases. Primary splenic lymphomas were most commonly of large cell B-cell type.

CONCLUSIONS

In the setting of splenomegaly or splenic mass, splenectomy has a high diagnostic yield and usually discloses a malignancy. The clinical category of "primary splenic lymphoma" is biologically heterogeneous, and the diagnosis is usually an intermediate grade (not low grade) lymphoma. The range of conditions associated with splenic masses were quite commonly associated with diseases that are amenable to fine-needle aspiration (FNA) diagnosis, whereas those disorders associated only with splenomegaly included a large fraction of diseases for which FNA may yield either incomplete or misleading results.

"Juvenile" xanthogranuloma: an immunophenotypic study with a reappraisal of histogenesis

The non-Langerhans histiocytoses, a nosologic category to which juvenile xanthogranuoma (JXG) belongs, represent a heterogenous collection of disorders related to the monocyte/macrophage lineage. The dermal dendrocyte was previously proposed as the cell of origin for JXG on the basis of Factor XIIIa reactivity, a suggestion that does not fully explain the occasional xanthogranulomatous proliferations localizing exclusively to extracutaneous sites. This study applies a panel of recently developed immunohistochemical markers to JXGs and relates the phenotype of this process to new concepts of monocyte/dendritic cell ontogeny. Twenty-seven JXG, ten dermatofibromas (DF), and ten age-matched normal skin specimens were stained using standard immunohistochemistry methods, and all JXGs were fascin+ and CD68+, although 26 of 27 were reactive for HLA-DR, 25 of 27 for Factor XIIIa, 25 of 27 for LCA, 21of 27 for CD4, and 8 of 27 for polyclonal s100. Six of those eight polyclonal S100+ cases were also reactive for monoclonal S100. None of those cases was reactive for CD1a, CD3, CD21, CD34, or CD35. Eight of ten dermatofibromas were FXIIIa+; all were negative for HLA-DR, LCA, CD4, and polyclonal s100. In controls, fascin+ dendritic cells were present but did not stain for Factor XIIIa, S100, or CD4. Based on the morphologic and phenotypic overlap of the lesional cells in JXGs and plasmacytoid monocytes, it would appear that the plasmacytoid monocyte might be considered the putative normal counterpart of the major cellular population of JXGs, a proposal that helps explain the extra-cutaneous, visceral, and soft tissue location that have been reported for occasional cases of JXG. We would also conclude that neither Factor XIIIa-nor S100+ results should preclude the diagnosis of JXG, and find that reactivity for CD4 and LCA may be used to distinguish JXG from DF when the latter is heavily lipidized or the former is not.

Posttransplant lymphoproliferative disease in children: correlation of histology to clinical behavior

PURPOSE

To determine whether the morphologic features of posttransplant lymphoproliferative disease (PTLD) correlated to a response to therapy.

PATIENTS AND METHODS

We reviewed our experience with PTLD in the pediatric population. We identified 32 patients with a total of 36 episodes of PTLD. The diagnosis was confirmed by tissue examination and classified according to the degree of monomorphic features of the lesion. Thirty-four of 36 episodes were managed with immunosuppression reduction, and the patients were assessed for their response to this strategy. Chemotherapy was used to treat 10 of 15 patients who had progressive disease, and their subsequent course was also analyzed.

RESULTS

Sixteen of 17 (94%) patients with polymorphic morphology responded to immunosuppression reduction compared with only 5 of 17 (29%) patients with monomorphic features (P < 0.001). All of the patients with progressive disease who did not receive additional therapy died. Standard chemotherapy regimens for lymphoma were administered to 10 patients with progressive disease, with a high response rate (90%), durable remissions, and acceptable toxicity.

CONCLUSIONS

We conclude that the morphologic characteristics of PTLD provide information to potentially help guide treatment strategies in the management of this disease. Standard chemotherapy regimens for malignant lymphoma appear to be a viable treatment option for patients with progressive disease, although further investigation is needed.

Recurrences in nodal T-cell lymphoma. Changes in histologic appearance and immunophenotype over the course of disease

We examined the patterns of relapse or persistence in 37 cases of nodal peripheral T-cell lymphoma (PTCL) to address the morphologic and immunophenotypic findings. Relapses were documented in lymph node (25 cases) and/or a variety of extranodal sites at a mean of 21 months after presentation; several cases recurred as late as 13 years. Persistent bone marrow involvement was a feature of angioimmunoblastic lymphoma (AIL) and histiocyte-rich and small-cell tumors. Relapses in anaplastic tumors often involved unusual extranodal sites. The majority of relapsed PTCLs retained a similar histologic appearance, pattern of nodal involvement, and immunophenotype. Histologic progression, as assessed by increased numbers of large cells, was seen in 3 cases of AIL, in 1 case with an initial small cell morphologic appearance, and in 2 cases of PTCL with an initial mixed small and large cell appearance. Immunostains for T-cell activation markers showed increased immunoreactive cells in 5 of the 6 cases, whereas increased numbers of p53-positive tumor cells were noted in 3 of the 6 cases. The discrete large cell transformation occasionally seen in B-cell lymphoma and extranodal T-cell lymphoma was not observed in these cases.

Lymphocyte predominance Hodgkin's disease: the use of bcl-6 and CD57 in diagnosis and differential diagnosis

Distinction of lymphocyte predominance Hodgkin's disease (LPHD) from other forms of lymphoma often requires immunohistochemistry (IHC). Most previously published recommended panels include markers to define the large neoplastic cells (for example, CD20, J chain, CD45) as well as the non-neoplastic background cells (CD21, CD45RO, CD57, TiA 1). In the present study we examine the practical use of a double IHC method designed to look simultaneously at two germinal center specific cell types: bcl6+ cells and [bc16+, CD57+] co-positive cells. All 10 nodular LPHD had bcl6+ large cells and numerous CD57+ small background cells, including [bcl6+CD57+] cells in rosettes. One case of LPHD with large cell transformation contained numerous bcl6+ large cells both singly (in areas of typical LPHD) and in sheets (in foci of large cell transformation), many CD57+ small cells but few [bcl6+CD57+] co-positive cells and no rosettes. In none of the five cases of florid progressive transformation of germinal centers were true rosettes seen, although all contained variable numbers of bcl6+ large cells and CD57+ cells. Lymphocyte-rich classic Hodgkin's disease LRCHD cases were notable for bcl6 reactivity in Reed-Sternberg cells in all cases, numerous background small bcl6+ lymphocytes, and rare CD57+ cells. Two phenotypic profiles were associated with the 10 cases of T cell-rich B cell large cell lymphoma (TCRBCL). In the first, group "A," six of six cases had bc16- large cells and few CD57+ small cells, and none had significant numbers of [bcl6+, CD57+] co-positive cells. In the second, group "B," four of four cases had bcl6+ large cells with numerous CD57+ and [bcl6+, CD57+] co-positive cells. These findings not only show that LPHD can be distinguished from its morphologic mimics through identification of specific germinal center cell types, but also identifies a second group of TCRBCL (group "B") whose phenotype suggests it might be an architectural variant of nodular LPHD.

Expression pattern of T-cell-associated chemokine receptors and their chemokines correlates with specific subtypes of T-cell non-Hodgkin lymphoma

Chemokine receptors mediate the migration of lymphocytes through the binding of soluble ligands, and their expression is differentially regulated in lymphocyte subsets. The pattern of chemokine receptor expression in T-cell non-Hodgkin lymphoma has not been previously studied. Using a panel of mouse monoclonal antibodies, we studied the immunohistochemical expression of the Th1-associated chemokine receptor CXCR3 in 141 patients with T-cell lymphoma, and we studied the receptors CCR4 and CCR5 and some of their ligands in a subset of these tumors. Expression of CXCR3 was typical of the smaller T cells in angioimmunoblastic lymphoma (15 of 18 patients), angiocentric lymphoma (3 of 3 patients), histiocyte-rich tumors (4 of 5 patients), and unspecified T-cell lymphomas (17 of 39 patients). CXCR3 expression was seen in only 1 of 15 patients with anaplastic lymphoma kinase (ALK)-positive anaplastic large-cell lymphoma. In contrast, all ALK-positive tumors showed diffuse reactivity for the Th2-associated receptor CCR4 (5 of 5 patients). CCR4 expression was also a consistent feature of the large-cell transformation of mycosis fungoides. CCR5 expression showed no consistent association with any T-cell tumor type. The chemokines Mig (CXCR3 ligand), TARC (CCR4 ligand), and MCP-2 (CCR5 ligand) were detected in intratumoral blood vessels and histiocytes. Mig was also coexpressed by a subset of CXCR3-positive tumor cells in 6 of 20 lymphomas. MCP-2 was highly expressed in stromal cells in 3 patients with nodal involvement by cutaneous T-cell lymphoma. As with normal T-cell subsets, we demonstrated that there is frequent differential expression of chemokine receptors in T-cell tumors, which may explain, in part, the distinctive patterns of spread in different tumor subtypes. (Blood. 2000;96:685-690)

Light microscopic, immunophenotypic, and molecular genetic study of autoimmune lymphoproliferative syndrome caused by fas mutation

This case provides a complete light microscopic, immunophenotypic, and molecular genetic analysis of autoimmune lymphoproliferative syndrome (ALPS), a rare benign cause of dramatic lymphadenopathy with atypical histology and phenotype that may be mistaken for malignancy. The patient is 3-year-old child who was first clinically evaluated at the age of 16 months because of marked generalized lymphadenopathy and hepatosplenomegaly. Histologic sections of a cervical lymph node demonstrated a marked paracortical proliferation of occasional small and intermediate-sized lymphocytes and numerous large immunoblasts, the majority of which displayed a CD3(+), CD43(+), CD45RO(-) (OPD4, UCHL1) CD4(-), CD8(-) phenotype on paraffin sections, and which had a CD2(+), CD3(+), CD5(+), CD56(-), Tdelta1(-), [CD4(-), CD8(-)] double negative profile on flow cytometric analysis. Southern blot analysis did not identify a clonal T or B cell population, and sequencing of the fas gene identified a mutation that caused a single amino acid substitution in the intracytoplasmic death domain of this protein. An enriched population of CD45RO-negative naive T cells in the paracortex may explain the atypical histologic and immunophenotypic features of this case. Greater awareness of this heritable lymphoproliferative disorder will facilitate its recognition and minimize the possibility of misdiagnosis.

Histopathology of typhoid enteritis: morphologic and immunophenotypic findings

Enteric fever is a systemic illness caused by Salmonella infection, with S. typhi, S. paratyphi, and S. enteritidis being the most common serotypes. Humans are the only reservoir for S. typhi, and its predilection for the ileum is due to the fact that organisms enter the body by translocation across specialized Peyer's patch epithelium and then proliferate in the mucosal macrophages. The lesions in bowel and mesenteric lymph nodes are distinctive and mimic Kikuchi-Fujimoto disease and Rosai-Dorfman disease as well as infections caused by some non-salmonella bacteria. The four cases presented in this report, two culture-confirmed, all exhibited ileal mucosal hypertrophy caused by a neutrophil-poor monocyte/macrophage-rich hyperplasia. Though diffuse areas were present, much of the lesional proliferation was nodular, representing macrophage infiltration and colonization by the monocytes and macrophages. Immunophenotypic studies, which showed a CD68+, lysozyme+, UCHL-1+, OPD4-, CD4-, s100- profile, were helpful in distinguishing these lesions from other processes, including Kikuchi-Fujimoto disease and Rosai-Dorfman disease. Although rare in developed countries, enteric fever should be considered in any patient with recent travel to endemic areas and in the context of illness thought to be related to contaminated foods.

Benign vascular neoplasms of the spleen with myoid and angioendotheliomatous features

AIMS

To present the clinical light microscopic and immunophenotypic features of a distinctive vascular neoplasm of the spleen.

METHODS AND RESULTS

Two of the splenic lesions arose in children, and one was found in an adult. They ranged from 19 to 40 mm diameter and histologically were quite similar. Sheets of large epithelioid cells with a spectrum of nuclear configurations ranging from oval and vesicular to twisted and hyperchromatic were noted in each case. Distinct or prominent nucleoli were present in many cells, and occasional cells had nuclear pseudoinclusions. In two cases, bands of basophilic, fibroblast-rich stroma with scattered chronic inflammatory cells were present. The mitotic rate ranged from 0/10 high-power fields (HPF) to 0.5/10 HPF in these epithelioid cells. The vascular nature of these tumours was manifested as a sieve-like array of round, erythrocyte-filled spaces, most with attenuated and cytologically bland lining cells. The polygonal, epithelioid cells exhibited the following phenotype: smooth muscle actin (SMA)+, muscle specific actin (MSA)+, vimentin+, CD31-, CD34-, CD21-, CD8-, CD68- (2/3 cases), S100-, while the lining cells were CD34+, vimentin+ and SMA-, with variable CD31 and factor VIII related antigen expression. Elongated SMA+, MSA+ cell processes were evident in one case, reminiscent of previously characterized myoid elements of the normal spleen. An uneventful follow-up was noted for all three patients.

CONCLUSIONS

The histology and immunophenotype set these neoplasms apart from classic hamartomas, haemangiomas and previously characterized (haem)angioendotheliomas of the spleen, and may represent proliferations of myoid elements native to the spleen.

Primary gastric plasmacytoma: a rare cause of hypertrophic gastritis in an adolescent

BACKGROUND

This report describes a 16-year-old patient with gastric rugal hypertrophy caused by a primary gastric plasmacytoma. She had a 3-month history of nausea and burning abdominal pain. Radiographic studies showed giant rugal hypertrophy. Superficial endoscopic gastric biopsies showed mild inflammation with plasma cells of polyclonal origin in the mucosa. When symptoms persisted, she underwent laparoscopic full-thickness gastric biopsy. There was monoclonal plasma cell infiltration histologically diagnostic of plasmacytoma and inconsistent with Helicobacter pylori-associated mucosa-associated lymphoid tissue (MALT) lymphoma. There was no evidence for involvement of the bone marrow or regional lymph nodes. The tumor did not respond to radiotherapy, necessitating total gastrectomy.

METHODS

Blood samples were analyzed for interleukin (IL)-6 by enzyme-linked immunosorbent assay. Gastric biopsy and gastrectomy specimens were subjected to immunophenotyping for kappa and lambda light chains, CD45, CD20, and LN1 and to polymerase chain reaction analysis for herpes virus HHV8.

RESULTS

There was no elevation in circulating IL-6 levels, militating against a pathogenesis akin to that of Castleman's disease. There was no evidence for infection with the Kaposi's sarcoma-associated herpes virus HHV8, which has recently been found in patients with multiple myeloma.

CONCLUSIONS

This diagnosis and the characteristics of the tumor are very unusual, if not unique, for a patient of this age. The diagnostic evaluation of this patient also demonstrates the importance of deep endoscopic or full-thickness biopsies in some children with hypertrophic gastritis.

Angiomatosis with angiokeratoma-like features in children: a light microscopic and immunophenotypic examination of four cases

We have identified three patients with an initial clinical or biopsy diagnosis of angiokeratoma, all of whom were found to have a more extensive vascular lesion within the surgical excision. A fourth patient with identical histologic findings had no specified clinical diagnosis and his first procedure was excisional. The patients ranged in age from 7 to 16 years, and the lesions were located on the buttock, thigh, calf, and foot. Macroscopic appearances included mildly keratotic pink-red or blue-grey macules (three cases) and pink macules with focal ulceration (one case). In three of the four cases, a shave biopsy diagnosis of angiokeratoma had been made, and the extensive and deeply infiltrative nature of the vascular proliferation was recognized only at subsequent resection, at which point angiomatosis was diagnosed. In the fourth case. excisional biopsy was attempted at presentation, and the superficial morphology was angiokeratoma-like, but the vascular proliferation was present in the deep subcutaneous fat. CD31 and CD34 reactivity was present in the superficial and deep vessels in all cases, and lesional vessels were rimmed by a bland population of smooth muscle actin positive pericytes, findings that differentiate these cases from angiokeratoma, which has previously been reported to be CD34 negative. We conclude that the dilated vascular spaces that typify angiokeratoma may also be seen overlying a deep vasoformative process that is not amenable to resection, and suggest that caution should be exercised in evaluating small biopsies with angiokeratoma-like appearance.

The spleen in the Wiskott-Aldrich syndrome: histopathologic abnormalities of the white pulp correlate with the clinical phenotype of the disease

The Wiskott-Aldrich syndrome (WAS) is a X-linked hematologic disorder characterized by thrombocytopenia, eczema, and immunodeficiency of variable severity. Reported here are the results of a morphologic, morphometric, and immunophenotypic analysis of splenic lymphoid tissue in 12 WAS patients with documented molecular defect and with different disease severity. Spleens from 29 age-matched patients with different diseases were used as controls. Paraffin-embedded tissue (from all cases) and fresh-frozen samples (from 5 WAS patients and 4 control subjects) were used to study the different white pulp compartments by classic morphologic, immunophenotyping, and image analysis techniques. Data were statistically analyzed by both parametric and nonparametric tests. Spleens from WAS patients showed a significant depletion of the total white pulp (p = 0.0008), T cell (p < 0.05), and B cell (p = 0.0002) areas and marginal zone (MZ) thickness (p < 0.0001). Among WAS patients, a negative correlation was found between the score of severity of the disease and all variables considered (Spearman's rank correlation coefficient, r = -0.79, r = -0.73, r = -0.68, and r = -0.56, respectively). In conclusion, this study shows that in WAS a general depletion of the splenic white pulp occurs, supporting the evidence that WAS is characterized by a combined immune defect. The significant reduction of the MZ may explain the inability of WAS patients to mount a response to T-independent antigens.

Lymphoma presenting as a solitary bone lesion

Primary lymphoma of bone has characteristic clinical and radiologic manifestations; however, its histologic features and clinical outcome show considerable variability. The histologic and immunophenotypic features of 13 adult patients with lymphoma as a solitary bone lesion were compared with clinical outcome. All tumors studied were non-Hodgkin lymphoma of anaplastic or large cell type and included B-cell (9 cases), T-cell (3 cases), and null cell (1 case) phenotypes. All patients responded well initially to systemic chemotherapy (with or without radiotherapy); however, disease in 6 patients progressed or recurred, and 5 patients died of disease. Local disease progression was seen in 2 patients, with 4 patients experiencing relapse at distant sites. Expression of CD30, present in 7 cases, was associated with an anaplastic or large noncleaved histologic appearance. Absence of CD30 expression characterized 6 cases, including 4 with multilobate or cleaved morphologic features. Five of 6 cases that recurred were associated with CD30 expression, including 3 with anaplastic features. The 4 tumors with cleaved or multilobate nuclear morphologic features were associated with long disease-free survivals and may represent a distinct lymphoma subtype with a good prognosis.

Splenic pathology in myelodysplasia: a report of 13 cases with clinical correlation

Splenomegaly is uncommon in myelodysplasia (MDS) and, although cytopenias may be severe, therapeutic splenectomy is rarely performed. We report the histologic, histochemical, and immunophenotypic findings of nine cases of surgical splenectomy and four postmortem spleens from MDS patients. Four histologic patterns were identified: one dominated by erythrophagocytosis, one characterized by red pulp plasmacytosis, one with extramedullary hematopoiesis as the only salient finding, and one with marked red pulp expansion caused by a monocytic proliferation. Wright-Giemsa and histochemical stains were performed on touch preparations in three cases and played a critical role in the precise subclassification of one MDS patient's hematologic disorder, which ultimately proved to be chronic myelomonocytic leukemia. Splenectomy led to sustained improvement of cytopenias in three cases, but did not eliminate transfusion dependence for the remaining patients. Three splenectomy cases exhibited clustered Leder-negative mononuclear elements: two of these patients experienced disease progression to refractory anemia with excess blasts in transformation or acute myelogenous leukemia during post-splenectomy follow-up, whereas none of the three splenectomy patients without clustered mononuclear elements did. We conclude that splenomegaly in MDS usually reflects the sequelae of dyspoiesis rather than evidence of a proliferative phase, that clustering of Leder-negative large cells may correlate with either a substantial monocytic component or, possibly, increased risk of disease progression, and that the spleen can provide diagnostic as well as prognostic information in MDS patients with splenomegaly.

Relationship of Bcl-2 expression with apoptosis and proliferation in colonic graft versus host disease

Graft versus host disease (GVHD) occurs in up to 75% of patients who have had an allogeneic bone marrow transplant (BMT). However, the pathophysiology of this disorder, including the mechanisms responsible for enhanced apoptosis, are poorly understood. Bcl-2 is a cellular protein known to inhibit apoptosis in a variety of human tissues. Therefore, the aim of this study was to evaluate the expression of Bcl-2 in colonic GVHD and to determine its relationship to cell proliferation and apoptosis in this disease. Routinely processed colonic mucosal biopsy specimens from 47 allogeneic BMT patients with diarrhea were evaluated histologically for the grade of GVHD (0-4) and for the degree of apoptosis (apoptosis index). Immunohistochemical staining for Bcl-2 and MIB-1, a cell proliferation marker, was performed, and the results were correlated with the histological findings and with each other. Normal-appearing colonic mucosal biopsy specimens from 10 age-matched patients with noncolonic diarrhea served as controls. Also evaluated were 13 colonic biopsy specimens from 13 patients with chronic ulcerative colitis (three inactive, four mild chronic-active, six moderately severe chronic active) to test the specificity of our findings. When compared with controls, a slight trend toward increased Bcl-2 expression was noted in patients with high-grade GVHD (grade 3 or 4) (P=.09). However, Bcl-2 expression did not correlate with the degree of apoptosis in these patients. In contrast, the degree of Bcl-2 staining correlated positively with the crypt proliferative rate (P=.04). Furthermore, crypt proliferation was significantly higher in the GVHD patients in comparison with controls (MIB-1 index, 27.7+/-17.1 v 15.6+/-11.4, =.02), and increased progressively with each successively higher grade of GVHD, and with the degree of apoptosis. Similar to GVHD, Bcl-2 expression was increased in biopsy specimens of CUC patients with higher grades of active injury and epithelial regeneration. This immunohistochemical study does not provide support for Bcl-2 in the pathogenesis of GVHD-induced apoptosis in the colon, but instead, indicates that this protein may play a nonspecific role in the generalized response to cellular injury in GVHD.

T gamma/delta hepatosplenic lymphoma in a heart transplant patient after an Epstein-Barr virus positive lymphoproliferative disorder: a case report

BACKGROUND

An unusual case of a peripheral T-cell lymphoma of T gamma/delta hepatosplenic type (Tgamma/deltaHSL) that arose in a child 5 years after she received a heart transplant and 9 months after she developed Epstein-Barr virus (EBV) positive, B-cell lymphoid hyperplasia involving the tonsils is presented. The majority of the reported cases of Tgamma/deltaHSL have been described in young adult men without antecedent immunodeficiency; several well documented cases of Tgamma/deltaHSL in the posttransplant setting have been described previously, but none has been described in a child (or an adult) with a previously diagnosed EBV+ B-cell lymphoid hyperplasia.

METHODS

Standard histologic, immunohistochemical, flow cytometric, and molecular genetic techniques were used in the evaluation of diagnostic material.

RESULTS

The patient's Tgamma/deltaHSL involved the spleen in a predominantly cordal pattern, and infiltrated the liver in an exclusively sinusoidal distribution. Bone marrow involvement was focal and interstitial. In all locations, malignant cells were of intermediate or large size and had oval nuclei with coarse chromatin, with a scant or moderate amount of eosinophilic cytoplasm. This Tgamma/deltaHSL expressed the characteristic CD2+, CD3+, [CD4- CD8-], Tdelta1+ phenotype, and malignant cells also expressed the natural killer cell marker CD56. Cytogenetic studies demonstrated isochromosome 7q with the addition of trisomy 8 as the tumor progressed. Southern blot analysis demonstrated clonal rearrangements of the gamma, delta, and beta loci of the T-cell receptor but did not identify EBV DNA within the tumor cells.

CONCLUSIONS

This case highlights the fact that a full range of lymphoid proliferations is possible in the posttransplantation period, and that a prior diagnosis of a B-cell disorder does not preclude the development of a subsequent T-cell posttransplant lymphoproliferative disorder (PTLD), which should be formally evaluated, especially if clinical circumstances appear atypical for a PTLD of the "usual" (EBV-related, B-cell) type.

Cytomegalovirus infection in the colon of bone marrow transplantation patients

The histologic distinction between cytomegalovirus (CMV) infection and graft-versus-host disease (GVHD) in the colon in bone marrow transplantation (BMT) patients relies heavily on the identification of viral inclusions, because the morphologic features of these two diseases are otherwise similar. The aim of this study was to assess (1) the prevalence of colonic CMV infection in BMT patients with the use of DNA in situ hybridization (ISH); and (2) the sensitivity and specificity of light microscopy in establishing a diagnosis of CMV infection in the colon of these patients. Fifty-five colonic mucosal biopsy samples from 50 consecutive allogeneic BMT patients with diarrhea were evaluated histologically for the presence of typical or atypical (suspicious, but not diagnostic) CMV inclusions and, if negative, for the grade of GVHD. CMV DNA ISH analysis was performed on all of the biopsy specimens and was correlated with the histologic and clinical findings. Histologic analysis revealed only one patient with morphologically typical CMV inclusions. Four other cases contained an isolated atypical mesenchymal cell with features considered suspicious, but not diagnostic, for CMV inclusions. All of these five cases exhibited histologic features that were otherwise indistinguishable from GVHD grades 1 to 2. The single case that was histologically positive for CMV was confirmed by DNA ISH. Of the four histologically atypical cases, only one was confirmed to be CMV positive by DNA ISH. Of the remaining 45 patients, 35 had GVHD, 1 had pseudomembranous colitis, 1 had ischemic colitis, and 8 had no abnormalities found. Light microscopic examination is a sensitive method of screening for CMV infection in the colon of BMT patients but is less specific than DNA ISH. CMV infection is an infrequent cause of colitis in our BMT population.

Clinical presentation and outcome in lymphocyte-predominant Hodgkin's disease

PURPOSE

The patterns of presentation, histologic pattern (nodular or diffuse), treatment, and long-term outcome were studied in patients with lymphocyte-predominant (LP) Hodgkin's disease (HD) to determine whether these patients should be treated differently than patients with other subtypes of HD.

PATIENTS AND METHODS

Pathology was reviewed for 97 patients with an initial diagnosis of LPHD made between 1970 and 1993. Seventy-five patients had LPHD on review: 55 had nodular LPHD, 14 had diffuse LPHD, and six had LP histology without subclassification. There were 60 males (80%) and 15 females (20%). Sixty-six patients (88%), presented with clinical stage (CS) I or II disease. Seventy-one patients were treated at the Joint Center for Radiation Therapy (JCRT) and were considered for analysis of treatment outcome. Sixty-one of these 71 were treated with radiation (RT) alone; 17 received mantle RT alone, 27 mantle and paraaortic RT, and seven total-nodal irradiation (TNI). Ten patients with subdiaphragmatic HD received pelvic and paraaortic RT. Of the 10 remaining patients, four were treated with RT and chemotherapy (CT) and six were treated with CT alone. The median follow-up time was 10.8 years.

RESULTS

The 10-year actuarial freedom-from-first-relapse (FFR) and 10-year overall survival rates for the 71 patients with LPHD treated at the JCRT were 80% and 93%, respectively. The 10-year actuarial FFR by nodular (n = 51), diffuse (n = 14), and unspecified (n = 6) histologic pattern was 74%, 100%, and 60%, respectively. Overall, 14 of 71 patients have relapsed: nine of 61 with stage IA, IB, or IIA disease and five of 10 with stage IIB to IVB disease have relapsed. The median time to relapse was 53 months. Nine of 71 patients have died. Only one death has been from HD: five patients died of second cancers, two of cardiac disease, and one of alcoholic liver cirrhosis. Of seven patients with second malignancies, five died. None of the second malignancies were non-Hodgkin's lymphoma (NHL).

CONCLUSION

Patients with LPHD have different patterns of presentation, sex and age distribution, and likelihood of occult abdominal disease than patients with nodular-sclerosing (NS) or mixed-cellularity (MC) disease. The median time to relapse for LP patients was later than reported for other histologic subtypes; however, there was no pattern of continuous late relapse. With pathologic staging and standard treatment, mortality from LPHD is low; nearly all deaths have been cardiac- or second tumor-related. This suggests that less aggressive treatment for LPHD might continue to yield excellent results, while perhaps lowering the long-term risk of complications.

Well-differentiated inflammatory liposarcoma: an uncommon and easily overlooked variant of a common sarcoma

The presence of abundant inflammation is uncommon in well-differentiated liposarcomas (WDLPS). We describe ten cases of WDLPS in which an extensive lymphoplasmacytic infiltrate was present to such a degree that the differential diagnoses included inflammatory pseudotumor and Castleman's disease, and in which the lipogenic component could easily be overlooked. The median age (62 years) and tumor location (six retroperitoneal, three head/neck, one paratesticular) was comparable to usual WDLPS, as was the fact that six of ten cases recurred locally, but none metastasized during the period of follow-up (range 13 to 228 months, median 72 months). Key histologic features included (1) nodular lymphoplasmacytic aggregates; (2) intervening paucicellular stroma containing fibroblastic elements, frequently with plasma cell-rich zones and scattered atypical, often multinucleate cells; (3) merging of atypical adipocytic and inflammatory elements; and (4) adjacent clearly defined zones of lipoma-like or, more rarely, sclerosing-type liposarcoma. In recurrences, two cases "reverted" to purely lipoma-like liposarcoma, and two cases dedifferentiated. Immunoperoxidase studies demonstrated that B cells predominated in the lymphoid aggregates, and that the abundant plasma cells were polytypic in nature. We support the use of the term well-differentiated inflammatory liposarcoma for tumors of this type, which should be distinguished from high-grade liposarcoma containing inflammatory cells, and in order to identify a subgroup of WDLPS at high risk for misdiagnosis.

Mixed tumors and myoepitheliomas of soft tissue: a clinicopathologic study of 19 cases with a unifying concept

We report 19 unusual cases of mixed tumors and myoepitheliomas arising in soft tissues. The neoplasms occurred in 12 males and seven females. The age at diagnosis ranged from 2 to 83 years (mean 35, median 30). Eight tumors arose in the upper limb, six in the lower limb, three in the trunk, and two in the head and neck region. Three cases involved both dermis and subcutis; the remainder arose in subcutaneous (13 cases) or deep subfascial soft tissue (three cases). The most common presenting complaint was a painless swelling, with duration ranging from 2 weeks to 1 year (median 2.5 months). Microscopically, the tumors were predominantly well circumscribed and lobulated. Six cases showed a focally infiltrative margin. Cardinal morphologic features included nests, cords, and ductules of epithelioid cells and/or nests of spindled cells within a hyalinized to chondromyxoid stroma. One tumor was predominantly composed of myoepithelial cells and devoid of epithelial differentiation (i.e., ductules). Cytoplasmic hyaline inclusions were noted in two cases; squamous differentiation was seen in one case. Osteoid production and/or metaplastic bone was observed in three tumors. Chondroid differentiation (usually mature) was seen in four cases. Adipocytic differentiation was seen in two tumors. Mitotic activity was variable but generally scant; atypical mitotic figures were not identified. By immunohistochemistry, 16 of 16 cases expressed pan-keratin; 16 of 17 S-100 protein; six of 14 alpha smooth muscle actin (IA4); two of 10 muscle specific actin (HHF-35); two of 10 desmin; three of 11 glial fibrillary acidic protein; and three of 16 epithelial membrane antigen. Clinical follow-up was available in 10 patients and ranged from 6 months to 20 years (mean 4.25 years, median 2 years). Two patients developed local recurrence; metastasis to lung and lymph nodes were observed in two additional patients. Both of the latter patients died. We believe that these findings expand the concept of cutaneous mixed tumors to include neoplasms composed predominantly of myoepithelial cells and to include tumors arising in deeper subcutaneous and/or subfascial tissues. The clinical behavior of such neoplasms, when arising in soft tissues, may be difficult to predict but is most often benign; however, a minority of lesions metastasize. Until larger studies with longer follow-up are available, treatment and prognostication are probably best based on criteria used in comparable salivary gland tumors.