The relationship between chronic lymphocytic leukaemia and prolymphocytic leukaemia. III. Evaluation of cell size by morphology and volume measurements

A rare case of B-lymphoproliferative disorder with villous lymphocytes harboring t(8;14)(q24;q32) translocation

Splenic lymphoma with villous lymphocytes (SLVL) or splenic marginal zone lymphoma with circulating villous lymphocytes is rare, and prolymphocytic transformation of SLVL is rarer. At present, only one case of SLVL with t(8;14)(q24;q32) translocation has been reported. In this study, we report a case of B-lymphoproliferative disorder with villous lymphocytes harboring t(8;14)(q24;q32) chromosome translocation that we inclined to SLVL with a prolymphocytic transformation. A 73-year-old female showed marked hepatosplenomegaly and high lymphocytosis (lymphocytes > 200 × 10⁹/L). The abnormal lymphocytes had short coarse villi and round nuclei with prominent nucleoli. The immunophenotypes showed CD19⁺, CD20⁺, HLA-DR⁺, CD22⁺, CD5⁺, Kappa⁺, CD25^dim, CD71^dim, Lambda^-, CD7^-, CD10^-, CD23^-, CD34^-, CD33^-, CD13^-, CD14^-, CD117^-, CD64^-, CD103^-, and CD11c^-. The karyotype showed complex abnormality: 46XX,+ 3,-10, t(8;14)(q24; q32)[11]/46XX[9]. The cytoplasmic projection, immunological characteristics, and trisomy 3 chromosome abnormality supported the diagnosis of SLVL. However, the presence of prominent nucleoli and high lymphocytosis suggested prolymphocytic transformation, probably as a result of t(8,14) chromosome translocation. In this report, we described an unusual case of B-lymphoproliferative disorder with villous lymphocytes harboring t(8;14)(q24;q32) translocation, which could provide help in the diagnosis and differential diagnosis of B-lymphocytic proliferative diseases.

Transformation in Low-grade B-cell Neoplasms

Low-grade B-cell leukemias/lymphomas are a diverse group of indolent lymphoproliferative disorders that are typically characterized by good patient outcomes and long life expectancies. A subset of cases, however, undergo histologic transformation to a higher-grade neoplasm, a transition associated with a more aggressive clinical course and poor survival. Transformation of follicular lymphoma to diffuse large B-cell lymphoma and Richter transformation of chronic lymphocytic leukemia/small lymphocytic lymphoma are best characterized in the literature. This article reviews clinical and pathologic characteristics of these most common forms of transformation, with an emphasis on salient histologic, immunophenotypic, and genetic features.

Pathology of chronic lymphocytic leukemia: an update

Chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) is a clonal lymphoproliferative disorder characterized by proliferation of morphologically and immunophenotypically mature lymphocytes. CLL/SLL may proceed through different phases: an early phase in which tumor cells are predominantly small in size, with a low proliferation rate and prolonged cell survival, and a transformation phase with the frequent occurrence of extramedullary proliferation and an increase in large, immature cells. Although some patients with CLL have an indolent disease course and die after many years of unrelated causes, others have very rapidly disease progression and die of the disease within a few years of the diagnosis. In the past few years, considerable progress has been made in our ability to diagnose and classify CLL accurately. Through cytogenetics and molecular biology, it has been shown that CLL and variants are associated with a unique genotypic profile and that these genetic lesions often have a direct bearing on the pathogenesis and prognosis of the disease. Similarly, the development of antibodies to new biologic markers has allowed the identification of a unique immunophenotypic profile for CLL and variants. Moreover, accumulating evidence suggests that CLL cells respond to selected microenvironmental signals and that this confers a growth advantage and an extended survival to CLL cells. In this article, we will review the progress in the pathobiology of CLL and give an update on prognostic markers and tools in current pathology practice for risk stratification of CLL.

Clinical and laboratory parameters that define clinically relevant B-CLL subgroups

B cell-type chronic lymphocytic leukemia (B-CLL) is a heterogeneous disease. This is reflected by the very wide-ranging clinical courses that B-CLL patients experience and by the marked variation in laboratory findings between patients. In this chapter, we will review the various clinical and laboratory parameters that divide B-CLL patients into "subgroups," and correlate the parameters that define them. When feasible, we will also link clinical features to the cellular and genetic characteristics recently defined for these leukemic cells. The discussion is limited to parameters that define phenotypes or subgroups that may relate to disease activity and clinical outcome.

Burkitt-type acute leukemia in a patient with B-prolymphocytic leukemia: evidence for a common origin

Burkitt-type acute leukemia cells were present in the bone marrow of a patient with B-prolymphocytic leukemia diagnosed from peripheral blood cell morphology. Immunophenotype analysis confirmed morphological patterns. Cytogenetic and fluorescence in situ hybridization (FISH) analysis showed an identical t(8;22)(q24;q21) with MYC locus rearrangement in blood and bone marrow cells, with additional chromosome abnormalities in the bone marrow. In addition, the loss of one copy of the TP53 gene and identical IGH DNA clonal rearrangements were shown with FISH and polymerase chain reaction analysis respectively in the two types of leukemic cells. These data indicated the common origin of the two coexisting leukemias and are the first example of such occurrence in a leukemic patient.

Albumin enhanced morphometric image analysis in CLL

BACKGROUND

The heterogeneity of lymphocytes from patients with chronic lymphocytic leukemia (CLL) and blood film artifacts make morphologic subclassification of this disease difficult.

METHODS

We reviewed paired blood films prepared from ethylene-diamine-tetraacetic acid (ETDA) samples with and without bovine serum albumin (BSA) from 82 CLL patients. Group 1 adhered to NCCLS specifications for the preparations of EDTA blood films. Group 2 consisted of blood films containing EDTA and a 1:12 dilution of 22% BSA. Eight patients were selected for digital photomicroscopy and statistical analysis. Approximately 100 lymphocytes from each slide were digitally captured.

RESULTS

The mean cell area +/- standard error was 127.8 microm(2) +/- 1.42 for (n = 793) for group 1 versus 100.7 microm(2) +/- 1.39 (n = 831) for group 2. The nuclear area was 88.9 microm(2) +/- 0.85 for group 1 versus 76.4 microm(2) +/- 0.83 for group 2. For the nuclear transmittance, the values were 97.6 +/- 0.85 for group 1 and 104.1 +/- 0.83 for group 2. The nuclear:cytoplasmic ratios were 0.71 +/- 0.003 for group 1 and 0.78 +/- 0.003 for group 2. All differences were statistically significant (P < 0.001).

CONCLUSIONS

BSA addition results in the reduction of atypical lymphocytes and a decrease in smudge cells. BSA also decreases the lymphocyte area and nuclear area, whereas nuclear transmittance and nuclear:cytoplasmic ratio are increased. A standardized method of slide preparation would allow accurate interlaboratory comparison. The use of BSA may permit better implementation of the blood film-based subclassification of CLL and lead to a better correlation of morphology with cytogenetics and immunophenotyping. Published 2003 Wiley-Liss, Inc.

Mature B-cell leukemias with more than 55% prolymphocytes. A heterogeneous group that includes an unusual variant of mantle cell lymphoma

We studied 20 cases of mature B-cell leukemia with more than 55% prolymphocytes in peripheral blood or bone marrow, fulfilling the French-American-British criteria for B-cell prolymphocytic leukemia (PLL). Cases segregated into 3 groups: de novo PLL, 6; PLL occurring in patients with a previous well-established diagnosis of chronic lymphocytic leukemia (PLL-HxCLL), 10; and t(11;14)(q13;q32)-positive neoplasms, 4. All cases expressed monotypic immunoglobulin light chain, and most were positive for CD5. All t(11;14)-positive neoplasms were CD23- and uniquely positive for cyclin D1. Cytogenetic abnormalities were present in 19; in all 19, the karyotype was complex, indicating clonal evolution and genomic instability. The most frequent cytogenetic abnormality in de novo PLL involved chromosome 7 in 4 cases. Trisomy 12 or add(12p) was present in 4 cases of PLL-HxCLL. We conclude that mature B-cell leukemias with more than 55% prolymphocytes are a heterogeneous group that includes t(11;14)-positive neoplasms, which we suggest are best classified as mantle cell lymphoma. We also suggest that prolymphocytic morphologic features are a common end-stage of transformation for several B-cell neoplasms.

Usefulness of CD79b expression in the diagnosis of B-cell chronic lymphoproliferative disorders

We evaluated anti-CD79b for its usefulness in the diagnosis of B-cell chronic lymphoproliferative disorders (BCLPDs), particularly chronic lymphocytic leukemia (CLL) and mantle cell lymphoma (MCL). We analyzed 100 BCLPDs for CD5, CD19, CD20, CD23, CD79b, and surface immunoglobulin light chain (sIg) expression by 4-color flow cytometry. CD20, CD79b, and sIg expression were quantified. Correlational analysis and univariable and multivariable logistic regression models were used to determine the best combination of antigens for the immunophenotypic classification of CLL vs other BCLPDs. Positive and statistically significant Spearman pairwise correlations between CD20, CD79b, and sIg fluorescence intensity were demonstrated. In the simplest models in which a single variable was considered, cutoff points were chosen that gave misclassification rates for CLL of 16% for CD79b, 19% for sIg, and 18% for CD20. Low-intensity CD79b, CD20, and sIg are associated highly with CLL. A panel containing CD5, CD19, CD23, and sIg allowed correct classification of most cases. Addition of CD20 or CD79b improved diagnostic accuracy; CD79b was slightly better than CD20. CD79b seems to be a useful addition to a standard flow cytometry panel for the evaluation of BCLPDs.

Frequent deletions at 11q23 and 13q14 in B cell prolymphocytic leukemia (B-PLL)

Deletions of the long arm of chromosomes 11 and 13 are the most frequent structural chromosome aberrations in various types of lymphoproliferative disorders. However, these regions have not been studied so far in B cell prolymphocytic leukemia (B-PLL). We have investigated the incidence of 13q deletions in 18 B-PLL cases by fluorescence in situ hybridization (FISH), using molecular probes for the RB1 and D13S25 loci. Chromosome 11q deletions were evaluated by FISH using the yeast artificial chromosome (YAC) clone 755b11 from the chromosome 11q22.3-q23.1 region, which has been previously shown to be deleted in 20% of cases of chronic lymphocytic leukemia. Chromosome 11q23 deletions were found in 7/18 (39%) cases of B-PLL. Monoallelic loss of RB1, D13S25 and BRCA2 was present in 10/18 (55%), 6/18 (33%) and 3/18 (16%) of the cases, respectively. All the cases with D13S25 and BRCA2 deletion showed RB1 loss. Deletions of 13q14 and 11q23 are frequent chromosome aberrations in B-PLL and, in contrast to CLL, there is a preferential loss of RB1 with respect to the D13S25 locus suggesting that allelic loss of the RB1 gene may play a role in the pathogenesis of B-PLL.

Management of chronic lymphocytic leukaemia

Chronic lymphocytic leukaemia (CLL) is the most common form of adult leukaemia in Western countries. The diagnosis requires mature-appearing lymphocytes in the peripheral blood to >5 x 10(9)/L. The immunophenotype typically includes B cell antigens CD19, CD20 and CD23, low expression of surface immunoglobulin and CD5+, with other T cell antigens absent. Bone marrow biopsy, although not required for diagnosis, must show at least 30% lymphocytes. Cytogenetic abnormalities are frequent in patients with CLL, and may be associated with poor prognosis. Clinically, most patients are asymptomatic at presentation, with incidental lymphadenopathy and/or hepatosplenomegaly in the routine physical examination. Infections by opportunistic pathogens are the major cause of death. Aggressive transformation occurs in 10% of patients with CLL, most commonly prolymphocytic leukaemia (PLL) and Richter's syndrome. PLL de novo must be differentiated from PLL of an aggressive transformation. The incidences of autoimmune diseases and solid or haemopoietic secondary malignancies are increased in patients with CLL. Clinical stage is the strongest prognostic factor in CLL. There is no indication for early intervention. The current recommendation to start treatment includes disease-related symptoms, massive and/or progressive hepatosplenomegaly or lymphadenopathy, increasing bone marrow failure, autoimmune disease, and recurrent infections. Alkylating agents (e.g. chlorambucil) and nucleoside analogues (e.g. fludarabine) are the most active agents for CLL. Fludarabine induces higher response rates, but no improvement in overall survival has been observed. Fludarabine is the drug of choice for the majority of patients with CLL. Chlorambucil may be helpful for elderly patients with poor performance, and for patients who do not tolerate fludarabine. No drug combination is better than single agents. For patients refractory to initial treatment, referral to a clinical trial is the best choice. Other salvage therapy includes retreatment with the same initial agent (chlorambucil or fludarabine) if initial response was observed, or fludarabine for patients refractory to chlorambucil. Promising new approaches include cycle-active agents, nelarabine, biological therapy such as anti-CD52 monoclonal antibody, bone marrow transplantation, including the use of submyeloablative preparative regimens ('minitransplant') to induce graft-versus-leukaemia effect, and gene therapy. Prophylactic antibacterials and intravenous immunoglobulin should not be used routinely during supportive care. Epoetin may be helpful for patients who have anaemia without obvious cause. Assessment of response to therapy in CLL has been updated by the National Cancer Institute Working Group, and these guidelines are used worldwide for clinical trials.

p53 Abnormalities in B-Cell Prolymphocytic Leukemia

B-cell prolymphocytic leukemia (B-PLL) is an aggressive disorder of mature B cells with distinct clinical and pathologic features. To determine the incidence of abnormalities of p53, we analyzed 19 cases of B-PLL by DNA blot to assess loss of heterozygosity (LOH) at 17p13.3, by immunocytochemistry to assess p53 expression, and by direct DNA sequencing of polymerase chain reaction-amplified exons 5 to 9 of the p53 gene. LOH was detected in 10 of 19 (53%) cases, p53 expression was detected in 8 of 17 (47%), and p53 mutations were detected in 10 of 19 (53%) cases. The pattern of mutations was distinct from that observed in other B-cell malignancies. Six cases exhibited missense mutations; 4 were transversions and 2 were transitions. The G:C → A:T transition at cathepsin G dinucleotides commonly reported in p53 mutations in chronic lymphocytic leukemia (CLL) and other hematologic malignancies was observed in only 1 case of B-PLL. Three cases exhibited deletions (ranging from 3 to 35 bp in length) and one case exhibited a 2-bp insertion. In 1 case, a 27-bp deletion resulted in the expression of a p53 protein lacking 9 amino acids from the DNA binding region. All samples with p53 mutation showed loss of germline p53 sequences. However, 3 of 10 showed no LOH by Southern blot, indicating a localized deletion around the p53 locus at 17p13.1. Five of the 10 cases with p53 mutation exhibited detectable p53 expression, including 4 cases with p53 missense mutation and 1 case with deletion. Two of 7 cases with no detectable mutation of p53 nevertheless overexpressed p53. Therefore, there was no correlation between protein expression and p53 mutation in B-PLL. Our data indicate that the overall abnormalities of p53 occurred in 14 of 19 (75%) cases of B-PLL. The frequency of p53 mutation (53%) in B-PLL is the highest reported in B-cell malignancies and may be responsible for the frequent resistance to therapy of this disease. In addition, the pattern of p53 mutation was different from that observed in CLL and other hematologic malignancies and may indicate that a distinct pathogenic mechanism operates in B-PLL.

Chronic lymphoid leukemias: recent advances in biology and therapy

There exists a wide variety of lymphoid leukemias derived from B and T lymphocytes. These diseases have distinct immunologic and biologic features as well as varied responses to therapeutics. The most common lymphoid leukemia is chronic lymphocytic leukemia (CLL) which is a clonal proliferation of a subset of B cells expressing the CD5 antigen. Prolymphocytic leukemia is usually derived from B cells and shares some features with CLL but is clearly a distinct entity. Hairy-cell leukemia is a B cell malignancy that is uniquely responsive to a variety of biologic and chemotherapeutic agents. Waldenström's macroglobulinemia is a B cell malignancy that secretes immunoglobulin M (IgM) and may present with the hyperviscosity syndrome. Other B cell malignancies that less commonly present as leukemias include non-Hodgkin's lymphomas such as follicular lymphoma or mantle zone lymphoma. Multiple myeloma may rarely present or evolve into a plasma cell leukemia, typically in far advanced disease. T cell malignancies that may present as chronic lymphoid leukemias, and in the past have often been referred to as T cell chronic lymphocytic leukemia, are large granular lymphocytic leukemia, adult T cell leukemia/lymphoma, Sézary cell leukemia and rare cases of non-Hodgkin's lymphoma that are T cell derived and may present or evolve into a leukemic phase. There is also a rare T cell counterpart of prolymphocytic leukemia. Distinguishing these diseases is critical for optimal care of these patients.

B-cell chronic lymphocytic leukaemia stage 0. An immunophenotypic study of 66 cases and comparison with B small cell lymphomas

The cells of 66 B-CLL stage 0 patients were analyzed using a large panel of monoclonal antibodies in order to better define the immunophenotype of B-CLL. The data were compared with the immunophenotypes of lymph node cells from 51 patients with diffuse B small cell lymphoma or leukaemia with lymph node enlargement. The most frequent immunophenotype of B-CLL stage 0 was SIgM(D)+ EmR+ CD5+ CD9+ CD21+ CD23+ CD35- CD38-. Among the lymphomas, EmR-positive and EmR-negative cases were identified. The vast majority of the EmR-positive cases usually showed the leukaemic pattern, immunophenotype and lymph node histology of B-CLL. The EmR-negative cases usually had immunophenotypes quite different from those of B-CLL and the histology of indented cell lymphoma (centrocytic or intermediate) or features of lymphoplasmacytoid/cytic lymphoma. More than 20% of EmR-positive cells proved to be the most important marker to distinguish B-CLL from other lymphocytic lymphomas. Indeed this was a sign of better prognosis. Lymphoplasmacytoid/cytic lymphomas were EmR-positive with the immunophenotype of B-CLL or EmR-negative with a definitely different immunological profile. This suggests two morphologically similar but biologically different subgroups of these diseases.

Cell sizing in chronic lymphoproliferative disorders: an aid to differential diagnosis

AIMS

To determine if leucocyte volume distribution analysis (LVDA), obtained using a Coulter Counter Model S Plus IV, can be used to aid differentiation of chronic lymphoproliferative disorder (CLPD) subtypes.

METHODS

Mean lymphocyte volume and lymphocyte distribution width were measured on each patient (n = 90) using a hard copy of an amplified LVDA histogram. The mean lymphocyte volume was taken as the mean of the values on either side of the peak at half maximum height. The lymphocyte distribution width was taken as the range of cell values between the two values used to calibrate the mean lymphocyte volume. A template showing typical histograms from commonly occurring CLPD was also produced on an acetate sheet. This was used to examine the histogram from each new patient to evaluate its usefulness as an alternative to the calculation of mean lymphocyte volume and lymphocyte distribution width.

RESULTS

Mean lymphocyte volume and lymphocyte distribution width were significantly higher in B cell lymphocytic leukaemia of mixed cell type (B CLL/PL), B cell non-Hodgkin's lymphoma with peripheral blood spill, hairy cell leukaemia and T cell prolymphocytic leukaemia than in B cell chronic lymphocytic leukaemia (B CLL). The mean lymphocyte volume, but not the lymphocyte distribution width, was also significantly higher in T cell chronic lymphocytic leukaemia than in B CLL. The template gave an immediate preliminary indication of possible subtype(s) of disorder and could be used as an alternative to measurement of mean lymphocyte volume and lymphocyte distribution width.

CONCLUSIONS

Electronic haematology analysers producing an LVDA provide a useful, cost effective cell sizing analysis which can aid the differentiation of subtypes of CLPD.

[Bilateral mental neuropathy revealing prolymphocytic transformation of chronic lymphocytic leukemia]

Authors report an original case of mental neuropathy which reveal prolymphocytoid transformation of Chronic lymphocytic leukaemia. They discuss various pathogenic hypotheses which can join these two events; these are evolutive markers of bad prognosis concerning this hemopathy. Clinical, etiopathological and prognostic features of mental neuropathy associated with hemopathies are studied through this case.

Flow cytochemical analysis of peripheral lymphocytes in chronic B-lymphocytic leukemia. Prognostic role of the blast count determined by the H*1 system and its correlation with morphologic features

Peripheral blood samples from 148 previously untreated patients with chronic B-lymphocytic leukemia (B-CLL) were analyzed with the Technicon H*1 flow cytometer. The absolute number and the percentage values of both LUCs (large unstained cells) and blasts were correlated with survival, as well as with well-known prognostic factors including morphological subtypes of lymphoid cells. Results showed that patients at the most advanced clinical stages (Rai: III and IV; Binet: C) had the highest percentage and count of both LUCs and blasts. Furthermore, the proportion of LUC positively correlated with the following prognostic factors: peripheral lymphocytosis (greater than 50 x 10(9)/l); marked splenomegaly (greater than 10 cm UCM); % of circulating prolymphocytes, % immunoblasts, and % LGL. Our data analysis further revealed that chemotherapy produced a greater reduction of both the LUCs and of the blast count than of that of small lymphocytes. An increase in LUC count was found to coincide with deterioration of clinical status (progressive changes in the clinical stages, occurrence of prolymphocytoid transformation). A rapid increase in blast count was found to occur in concomitance with the development of Richter's syndrome, and correlated positively with the number of peripheral immunoblasts determined by light microscopy. Moreover, a blast percentage higher than 7% had the strongest predictive relation to survival rate when compared with other hematological parameters (lymphocytosis greater than 50 x 10(9)/l, % of LUCs greater than 12%, LUC to lymphocyte ratio greater than 16%, LUCs count greater than 2.2 x 10(9)/l). In the light of these findings, it may be suggested that the presence both of larger proportions of LUCs and of blasts measured with the flow cytometry may be considered unfavorable prognostic factors in B-CLL. However, based on morphological and multivariate statistical analyses, the blast count proved to be the most important prognostic parameter determined by the H*1 system in B-CLL.

Immunophenotypic subclassification of chronic lymphocytic leukaemia (CLL)

To determine the significance of the immunophenotypic heterogeneity of B-cell chronic lymphocytic leukaemia (CLL), surface immunoglobulins (SIgs), mouse rosette assays (MR), and a panel of monoclonal antibodies for B cells, T cells and myeloid cells were performed on peripheral blood samples from 61 newly diagnosed cases. Four groups were observed: group I (SIg+, MR+, CD19/20+, CD5+, T antigen (Ag)-; 27 cases); group II (SIg+, MR+, CD19/20+, CD5+, T Ag+; 17 cases); group III (SIg+, MR+ CD19/20+, CD5-, T AG-; 12 cases); and group IV (SIg-, MR+, CD19/20+, Cd5+, T Ag-; 5 cases). Groups were compared according to French-American-British Cooperative Group subtypes, clinical and laboratory features, Rai staging, and survival. Typical CLL morphology (greater than 90% small lymphocytes) was present in 20/20 (100%) of group I cases and 23/27 (85%) group II, III and IV cases (P = 0.09). Expression of a myeloid antigen was seen in 5/27 group I cases (18%) and 1/16 group II cases (6%), but was not predictive of survival (P = 0.36). The CD5- group III had a lower haemoglobin level (P less than 0.0001), higher Rai stage (P less than 0.002), and poorer survival at 5 years (P less than 0.02) than the other groups. We conclude that at least four distinct immunophenotypic subgroups of B-cell CLL can be determined. Expression of myeloid or T-cell antigens does not appear to predict for patient survival; however, lack of CD5 antigen may be associated with more advanced stage of disease and poor patient survival.

Leukaemia as a manifestation of large cell lymphoma

A leukaemic phase is uncommon in large cell lymphoma, particularly at presentation of the disease, and very few cases have been reported since immunological markers became available. We have studied 24 cases presenting over a period of 15 years. 16 were B-lineage and eight T-lineage. In five patients the large cell lymphoma represented a transformation of a preceding low grade, small cell lymphoproliferative disease. Three other patients with T-lineage large cell lymphoma presented with skin infiltration (two cases) or lymphoma-associated nephrotic syndrome (one case) 3-9 months before leukaemia occurred. The other 16 patients (12 B-lineage and four T-lineage) presented with de novo large cell leukaemia/lymphoma. With the exception of skin infiltration, clinical features did not differ between T and B-lineage cases. Prognosis was generally poor although a minority of patients lived 1-2 years. Median survival was 7 months. In the B-cell lymphomas immunological markers were those expected but among the T-cell cases there were many unusual immunophenotypes with frequent failure to express markers which are usually positive in peripheral T cells. Without the availability of immunological markers diagnosis would have been difficult, with acute myeloid leukaemia being the most important differential diagnosis. T-lineage and B-lineage cases could not be distinguished on cytological features. In the majority of cases the cells were very pleomorphic with nuclear lobulation, prominent nucleoli and marked cytoplasmic basophilia being commonly observed. Marked nuclear lobulation was not confined to T-lineage cases but was seen also in seven of 12 cases of de novo B-lineage leukaemia/lymphoma; it was readily apparent in histological sections and on ultrastructural examination.

Proposals for the classification of chronic (mature) B and T lymphoid leukaemias. French-American-British (FAB) Cooperative Group

Peripheral blood, bone marrow films, and bone marrow biopsy specimens from 110 patients, well characterised by clinical and laboratory studies, including electron microscopy, were reviewed, to determine proposals for the classification of chronic (mature) B and T cell leukaemias. On the basis of cytology and membrane phenotype the following disorders were defined: (i) B cell type: chronic lymphocytic leukaemia (CLL); CLL of mixed cell type, which includes cases with more than 10% and less than 55% prolymphocytes (CLL/PL), and a less well defined form with pleomorphic lymphocytes but less than 10% prolymphocytes; prolymphocytic leukaemia (PLL); hairy cell leukaemia (HCL); HCL variant; splenic lymphoma with circulating villous lymphocytes; leukaemic phase of non-Hodgkin's lymphoma (follicular lymphoma, intermediate, or mantle zone lymphoma and others); lymphoplasmacytic lymphoma with peripheral blood disease (mostly Waldenström's macroglobulinaemia); and plasma cell leukaemia. (ii) T cell type: T/CLL, which was differentiated from reactive T/lymphocytosis; T/PLL; adult T cell leukaemia/lymphoma; and Sézary's syndrome. The recognition of distinct entities within the B and T cell leukaemias seems to have clinical and epidemiological connotations. It is hoped that these proposals may serve as the basis for further work, discussion, and improved management of patients.

The relationship between chronic lymphocytic leukaemia and prolymphocytic leukaemia. IV. Analysis of survival and prognostic features

The prognostic value of biological, clinical and laboratory features was analysed in a series of 265 patients with chronic lymphocytic leukaemia (CLL) and prolymphocytic leukaemia (PLL). On univariate analysis seven features were shown to influence significantly the survival of the whole group of patients: absolute prolymphocyte count (ABS PROL), percentage of prolymphocytes (%PROL), WBC, spleen size, age, intensity of surface-membrane immunoglobulin (SmIg) and mouse (M) rosettes. Multivariate regression analysis of these features showed that only ABS PROL and spleen size had independent prognostic significance. The survival in PLL (38 cases) was significantly shorter than in CLL (227 cases) (median survival = 3 and 8 years, respectively). Patients with CLL with an increased %PROL (11-55%), defined as CLL/PL, could be divided into two groups: those with ABS PROL less than or equal to 15 X 10(9)/l (26 cases) fell within the 'standard-prognostic risk' for typical CLL (i.e. less than or equal to 10% PROL), whereas the survival outlook for the cases with ABS PROL greater than 15 X 10(9)/l (40 cases) was as bad as for PLL. A scoring system was generated with the four features that showed high prognostic significance: ABS PROL, spleen size, SmIg and M-rosettes. The score proved to be superior to any single feature as a predictor of survival, being especially useful in the analysis of the CLL/PL group: cases with high scores (greater than 2) had a median survival of 2.5 years, while the median has not been reached for those with low scores (less than or equal to 2). We suggest that this scoring system may help to identify the cases of CLL/PL that behave as PLL, and as such may benefit from different treatment.