Increased serum levels of vascular endothelial growth factor predict risk of progression in early B-cell chronic lymphocytic leukaemia

The present study is the first to evaluate serum levels of vascular endothelial growth factor (VEGF) in B-cell chronic lymphocytic leukaemia (CLL). All 68 B-cell CLL patients and 31 control subjects analysed had detectable serum levels of VEGF, with no statistically significant difference between two proups. An aberrant increase of circulating levels of VEGF was found in only 17.6% of cases. B-cell CLL patients whose serum VEGF levels were higher than the median (i.e. 194.8 pg/ml) or 75th percentile (i.e. 288.5 pg/ml) values were more frequently at an advanced clinical stage. In contrast, no correlation with other clinico-biological features representative of either tumour mass [bone marrow (BM) histology, peripheral blood (PB) lymphocytosis, beta-2 microglobulin (beta-2m), LDH, interleukin-6 (IL-6)] or disease-progression (DP) [lymphocyte doubling time (LDT)] was found. Serum levels of VEGF predicted the risk of DP in early CLL. Among 41 patients in Binet stage A, progression-free survival (PFS) was significantly shorter in those patients whose VEGF serum concentrations were above the median value. Interestingly, characteristics of stage A patients stratified according to the median value of VEGF were similar with respect to many clinico-biological features, thus suggesting a possible independent prognostic role for such a marker. Finally, when added to the Rai subclassification, VEGF serum levels identified two groups with different PFS within stages I-II. We conclude that increased serum levels of VEGF can be considered useful for predicting the risk of DP and add prognostic information to the Rai subclassification of stage A CLL.

Clinico-prognostic implications of simultaneous increased serum levels of soluble CD23 and beta2-microglobulin in B-cell chronic lymphocytic leukemia

Soluble CD23 (sCD23) and beta-2 microglobulin (beta2-m) are reliable prognostic parameters in B-cell chronic lymphocytic leukemia (CLL); however, their merit over well-established clinical variables such as clinical stages, bone marrow (BM) histology and lymphocyte doubling time (LDT) remains to be defined. Furthermore, information dealing with the impact on overall survival of the simultaneous increase of either beta2-m or sCD23 are lacking. In this prospective study based on 106 B-cell CLL patients, we propose a combination of beta2-m and sCD23 as a strong prognostic system whose statistical significance was mainly due to an excess of deaths in the subgroup displaying increased serum levels of either beta2-m or sCD23. Multivariate survival analysis confirmed the important dominant role of such a finding, thus excluding features with a high degree of codependence (i.e. clinical stages, LDT) and including variables with low association (i.e. BM histology) in the final regression model. The presence of increased serum levels of beta2-m/sCD23 and diffuse BM histology signified high-risk disease, whereas the absence of any adverse variable was associated with prolonged survival; in between there was a subgroup with only 1 characteristic which displayed an intermediate pattern of survival. Finally, on the basis combined increased serum levels of beta2-m and sCD23, a better stratification of low- and intermediate-risk patients could be obtained, thus allowing the formulation of a clinico-biological staging for CLL.

Clinicoprognostic implications of increased serum levels of vascular endothelial growth factor and basic fibroblastic growth factor in early B-cell chronic lymphocytic leukaemia

To assess the relative merit of increased serum levels of vascular endothelial growth factor and basic fibroblastic growth factor in predicting the risk of disease progression of patients with early B-cell chronic lymphocytic leukaemia we analyzed 81 Binet stage A patients whose sera were taken at the time of diagnosis and evaluated for the presence of vascular endothelial growth factor and basic fibroblast growth factor using an enzyme-linked immunosorbent assay. Serum levels of vascular endothelial growth factor positively correlated with Rai sub-stages (P=0.03), peripheral blood lymphocytosis (P=0.03), bone marrow histology (P=0.04) and beta2-microglobulin (beta2-m) (P=0.006). When dealing with basic fibroblast growth factor only a correlation with Rai sub-stages (P=0.02) could be found. Different cut-offs set on the basis of a stratification in quartiles, failed to demonstrate any correlation between serum levels of basic fibroblast growth factor and disease progression. In contrast, patients with increased serum levels of vascular endothelial growth factor (above median value, 203 pg ml(-1)) had a three times increased risk of disease progression, although, in multivariate analysis only Rai sub-stages (P=0.0001) and lymphocyte doubling time (P=0.002) retained their prognostic significance. Low levels of vascular endothelial growth factor were indicative of good clinical outcome in the subgroup of patients with either low (P=0.02) or high (P=0.03) beta2-m concentration. Finally, the highest prognostic power was obtained when serum vascular endothelial growth factor and beta2-m were examined in combination. Median of progression-free survival of patients who had both serum vascular endothelial growth factor and beta2-m higher than median value was only 13 months, in contrast median progression-free survival of patients with one marker increased (i.e. above the 50th percentile) was 40 months. Patients with both markers below the median experienced the best clinical outcome (median progression-free survival not reached at 40 months). In conclusion, serum levels of either vascular endothelial growth factor or basic fibroblast growth factor are high in patients with early chronic lymphocytic leukaemia, however, only vascular endothelial growth factor predicts behaviour of disease and helps to refine the prognosis of stage A patients.

Comparison of younger versus older B-cell chronic lymphocytic leukemia patients for clinical presentation and prognosis. A retrospective study of 53 cases

Fifty-three patients affected with B-cell chronic lymphocytic leukemia (CLL) younger than 50 years and observed in two hematological institutions have been retrospectively evaluated in order to verify whether this disease has different clinico-hematological features at presentation and different prognosis as compared to older cases. In our experience young cases with B-CLL diagnosis, confirmed by immunophenotype in 90.5% of patients, accounted for 7.1% of the whole CLL population. Sex distribution, mean peripheral lymphocyte count, platelet count, distribution among Rai's and Binet's stages, total tumor mass (TTM) score, histological pattern of bone marrow infiltration and lymphocyte doubling time (LDT) were similar to a series of 201 CLL cases older than 50 years. Only hemoglobin mean level was significantly higher in younger patients (13.1 +/- 2.1 vs 12.2 +/- 2.6 g/dl; p < 0.01). The overall median survival was 7.1 years. Rai and Binet staging classifications and TTM score system retained their prognostic value in this CLL population. In addition, cases fulfilling criteria of "smoldering" CLL, had a very long survival (75% survival probability at 16 years). Life-expectancy of younger patients was significantly longer than that of older ones (median survival, 7.1 versus 4.1 years; p < 0.05). However, when the background mortality due to non-CLL related deaths (i.e., cardiovascular complications, epithelial cancers) was removed, survival advantage of young cases disappeared. In conclusion this study confirms that prognosis of young CLL patients can be easily assessed using the current well-defined criteria. Since age is not by itself a criterion for intensifying treatment, further efforts to identify those young CLL patients who qualify for more aggressive therapy should be made.

Elevated serum levels of soluble CD44 can identify a subgroup of patients with early B-cell chronic lymphocytic leukemia who are at high risk of disease progression

BACKGROUND

Although soluble CD44 (sCD44) is considered a reliable marker of both tumor burden and disease activity, to the authors' knowledge, its predictive and prognostic value in B-cell chronic lymphocytic leukemia (CLL) has not been addressed to date.

METHODS

The authors studied 94 previously untreated CD5-positive B-cell CLL patients whose sera was taken at the time of diagnosis, stored at - 70 degrees C, and analyzed for the presence of standard sCD44 (sCD44(std)) using a commercial enzyme-linked-immunoadsorbent-assay. The impact of the sCD44 level on the clinical outcome of the disease was assessed in 74 patients with early CLL (61 Binet Stage A patients and 13 asymptomatic Stage B patients). Because the time to disease progression appears to predict the survival time of patients with CLL, it was used as a surrogate endpoint in the current study.

RESULTS

Patients with higher than median sCD44 levels (i.e., 642 ng/mL) had a more advanced clinical disease stage (P = 0.04), higher peripheral blood lymphocytosis (P = 0.006), and increased circulating levels of either lactate dehydrogenase (P = 0.01) or beta(2)-microglobulin (P < 0.0001). In univariate analysis, seven of the nine parameters investigated predicted progression-free survival (PFS). In a stepwise multiple regression analysis, only 2 parameters provided independent prognostic information regarding PFS: Rai substages (0 vs. I-II) (P = 0.002) and serum sCD44 levels > 642 ng/mL (P = 0.01). When added to the classification of smoldering CLL versus nonsmoldering CLL, the sCD44 level distinguished two groups within the group of nonsmoldering Stage A patients; patients with a sCD44 level > 642 ng/mL had a median PFS of 36 months, whereas patients with a sCD44 level < 642 ng/mL experienced a longer PFS (median had not been reached at 8 years of follow-up). Furthermore, serum levels of sCD44 defined two different patterns of PFS within the group of patients with Rai disease Stages I-II (P = 0.01).

CONCLUSIONS

An increased serum level of sCD44 can be considered to be a promising parameter for predicting the risk of disease progression in patients with early CLL. Furthermore, sCD44 helps to refine the prognostic stratification of patients with either nonsmoldering CLL or Rai Stage I-II disease, thus enabling the identification of different prognostic subgroups in patients with early CLL.

Clinico-prognostic relevance of quantitative immunophenotyping in B-cell chronic lymphocytic leukemia with emphasis on the expression of CD20 antigen and surface immunoglobulins

Expression of CD20, evaluated as antibody binding capacity (ABC) (i.e. absolute number of molecules of antibody per cell), was analyzed using flow cytometry on leukemic cells of 93 previously untreated patients, all fulfilling strict criteria of "immunologically typical" (i.e. CD5+, CD23+) B-cell chronic lymphocytic leukemia (CLL). Although changes of CD20 antigen density did not correlate with clinical parameters representative of either tumor mass (i.e. clinical stage, histological pattern of bone marrow involvement, absolute peripheral blood lymphocytosis) or disease progression (i.e. lymphocyte doubling time), a trend toward a better life-expectancy was observed in the low CD20 expression group compared with the high CD20 expression group (p = 0.05; relative risk of death, 0.51, 95% confidence interval, 0.24-1.04). Given the correlation between CD20 ABC and mean fluorescence intensity (MFI) of light chain (LC) surface immunoglobulins (Sm Ig) (r = 0.481, p < 0.0001), as well as the impact of MFI of Sm Ig LC on overall survival (p = 0.01; relative risk of death 0.44; 95% confidence interval, 0.10 to 0.76), we tried to verify whether a combination of B-cell markers, evaluated in a quantitative manner, could have additive prognostic properties. To this purpose we gave a value of 1 or 0 to each B-cell marker according to whether it was expressed at a low (i.e. CD20 ABC < 17.9 x 10(3) molecules/cell, MFI of LC Sm Ig < 100) or high (i.e. CD20 ABC > or = 17.9 x 10(3) molecules/cell, MFI of LC Sm Ig > or = 100) level thus allowing patient stratification into two groups with scores of 2 and 0-1, respectively. Survival of patients who scored 2 was significantly longer respectively. Survival of patients who scored 2 was significantly longer than that of patients who scored 0-1 (p = 0.02; relative risk of death, 0.44; 95% confidence interval, 0.22-0.72). However, when quantitative changes of CD20 antigen and LC Sm Ig expression, either alone or in combination, were simultaneously analyzed in a Cox model which included usual clinico-hematological features, only absolute peripheral blood lymphocytosis (p = 0.0001) and Binet clinical stages (p = 0.0001) maintained their prognostic power unmodified. Although variability of CD20 and Sm Ig expression make it possible to appreciate biological heterogeneity of B-cell CLL better, however, they cannot substitute well-established clinico-hematological features in the prognostic assessment of B-CLL patients.

Surface CD14 positivity in B-cell chronic lymphocytic leukaemia is related to clinical outcome

The aberrant expression of the myelomonocytic antigen CD14 was investigated in 128 untreated patients diagnosed with B-cell chronic lymphocytic leukaemia (B-CLL). A cut-off value of 5 x 10(9)/l CD14-positive cells was chosen for statistical analysis because it showed the best discriminating power among patients with different clinical features. 56 cases had a CD14+ cell count >5 x 10(9)/l. A significant correlation was found between Rai and Binet stages and total tumour mass (TTM) score on one hand, and the absolute CD14+ cell cut-off, on the other. This relationship was more evident in Rai 0-II and Binet A-B stages, where a CD14+ cell count >5 x 10(9)/l was preferentially distributed among patients with a higher tumoral mass. In univariate analysis the survival probability at 5 and 10 years showed a significant correlation with Rai and Binet stages, TTM score, CD14+ absolute cell count and median age. The median overall survival (OS) was 63 months for patients with a CD14+ cell count >5 x 10(9)/l and 136 months for those with a CD14+ cell count < 5 x 10(9)/l. In the multivariate Cox regression model, Rai stage, age and CD14+ cell count were independent significant factors for the prediction of OS. Finally, when the same analysis was restricted to Rai stages 0-II, CD14+ cell count was the only significant independent parameter influencing OS, with a relative death risk of 3.8. In conclusion, these data reveal that CD14+ represents an important marker for predicting OS in B-CLL patients and, therefore, we suggest that it should be included in the immunological characterization of B-CLL.

Expression on leukemic cells and serum circulating levels of intercellular adhesion molecule-1 (ICAM-1) in B-cell chronic lymphocytic leukemia: implications for prognosis

Expression of intercellular adhesion molecule-1 (ICAM-1) has been correlated clinical and laboratory characteristics of 62 patients with untreated CD5+ chronic lymphocytic leukemia (CLL). ICAM-1 was detected on B-CLL cells from 19 out of 62 patients (30.6%) and its expression did not correlate with the majority of immunological markers. An important finding of this study was the association between ICAM-1 expression and mean fluorescence intensity (MFI) of Slg (r = 0.507; P < 0.001). As far as correlation with clinical parameters is concerned, a statistically significant association between Binet clinical stages and ICAM-1 expression was found (P < 0.05). Furthermore, an atypical CLL morphology was more frequently encountered among patients who expressed ICAM-1 (P < 0.005). To obtain more information on the role of ICAM-1 in CLL we measured serum levels with a sandwich enzyme immunoassay. In 47 B-cell CLL patients studied, the mean value of circulating ICAM-1 levels was significantly higher (561 +/- 201 ng/ml) than that observed in 25 normal controls (353 +/- 162 ng/ml; P < 0.005); a clear correlation being found with Binet clinical stages (P = 0.026) and bone marrow (BM) histology (P < 0.005). We conclude that circulating ICAM-1 is elevated in CLL and such an increase reflects tumor mass as defined by clinical stages and BM histology, rather than peripheral blood lymphocytosis (r = 0.240). Even if soluble ICAM-1 appears to be less specifically increased that soluble CD23 serum levels, these circulating molecules were not completely independent of each other (r = 0.512; P < 0.001).

Intercellular adhesion molecules (ICAMs) 2 and 3 are frequently expressed in B cell chronic lymphocytic leukemia

Using different monoclonal antibodies (moAbs) from the 5th International Workshop on Leukocyte Differentiation Antigens we studied the expression of intercellular adhesion molecules (ICAMs) 2 and 3 on a homogeneous group of 23 B cell chronic lymphocytic leukemia (CLL) patients. Our results show that either ICAM-2 or ICAM-3 are constitutively expressed on CD5+ B-CLL cells. Owing to the role of ICAM molecules in governing the migration and traffic of lymphocytes to lymph nodes, our findings need to be validated in a more consistent patient series to understand clinico-prognostic implications of such an expression.

Clinico-prognostic implications of increased levels of soluble CD54 in the serum of B-cell chronic lymphocytic leukemia patients. Results of a multivariate survival analysis

BACKGROUND AND OBJECTIVE

Although less specific than sCD23, sCD54 levels have clinico-prognostic relevance in B-cell chronic lymphocytic leukemia (CLL). Since serological markers are now emerging as potentially important in CLL, we tried to verify whether sCD54 might complement clinical stages.

METHODS

Serum levels of sCD54 were determined at the time of diagnosis in 115 previously untreated CLL patients. Results were correlated with clinicobiological parameters as well as with survival.

RESULTS

Life-expectancy was significantly shorter in patients with higher serum levels of sCD54 (p < 0.001); however, in a Cox's multivariate survival analysis, the only variables which entered the regression model at a significant level were bone marrow (BM) histology (p = 0.03) and lymphocyte doubling time (LDT) (p = 0.04). Interestingly, when LDT was excluded from analysis the only significant variables were clinical stages (p < 0.05) and sCD54 (p < 0.05). These results suggest that sCD54 and LDT give similar prognostic information.

INTERPRETATION AND CONCLUSIONS

In CLL, sCD54 is a reliable prognostic parameter whose value is independent of clinical stages. When investigated in relation to clinical outcome, serum levels of sCD54 were able to predict progression to a more advanced clinical stage. On the basis of these data, an integrated clinico-biological classification which separates intermediate risk into two prognostic subgroups is proposed.

Natural history of early chronic lymphocytic leukemia. A single institution study with emphasis on the impact of disease-progression on overall survival

BACKGROUND AND OBJECTIVE

Criteria for identifying patients with early chronic lymphocytic leukemia (CLL) who are likely to progress to a more advanced clinical stage rely on results of prospective clinical trials. Is not clear whether these same criteria apply to patients followed-up in the setting of clinical practice. With the aim of addressing this issue we investigated the clinical outcome of a series of patients with Binet stage A CLL.

DESIGN AND METHODS

Two hundred and four Binet stage A CLL patients observed at a single institution over an 18-year period form the basis of this study. Different proposals for subclassifying Binet stage A were validated by using our patients as test-set cases.

RESULTS

The survival of patients with early CLL (i.e., Binet stage A and Rai stage 0) was significantly different from that of an age- and sex-matched population. Three of 4 different criteria for subclassifying stage A (Rai substaging, Montserrat criteria, French Group proposal), when applied to our patients, gave similar results in terms of sample size, death rate and disease progression (DP) risk. The French Group proposal, based exclusively on blood counts and hemoglobin levels, was not effective in predicting the risk of DP. Forty-nine (23.5%) patients progressed to a more advanced clinical stage (30 to stage B and 19 to stage C); the risk of DP was 32.8% at 5 years and 49.6% at 10 years. When analyzed as a time-dependent variable (Mantel-Byar method), DP had a clear cut-impact on overall survival (p < 0.0001). Finally, outlook for survival of patients who experienced a change of clinical stage was similar to that of patients in that stage at the time of diagnosis.

INTERPRETATION AND CONCLUSIONS

As many patients are now being diagnosed while asymptomatic and at a younger age than previously, an accurate evaluation of prognosis is mandatory in early CLL. How prognostic information translates into a policy of early or delayed therapy is still unclear. Our results further support a conservative approach for CLL in early stage; patients who progress into a more advanced stage have similar survival to those in that stage at diagnosis.

Differential expression of BCL-2 oncoprotein and Fas antigen on normal peripheral blood and leukemic bone marrow cells. A flow cytometric analysis

BACKGROUND

Fas antigen (Ag) has recently been identified as the putative surface molecule capable of transducing apoptotic signals into cells. Alterations in the expression of proto-oncogene bcl-2 have been implicated in the regulation of apoptosis.

MATERIALS AND METHODS

By employing a monoclonal antibody to bcl-2 protein (124 clone) and to Fas Ag (UB2 clone) the expression of these molecules was analyzed at flow cytometry on bone marrow (BM) and peripheral blood (PB) samples from patients suffering from different lymphoid and myeloid leukemic diseases (27 acute non-lymphocytic leukemia [ANLL]; 14 acute lymphocytic leukemia [ALL]; 19 B-cell chronic lymphocytic leukemia [CLL]; 2 Ph1+ chronic myeloid leukemia [CML]; one CD8+ T-cell chronic lymphoproliferative disorders). Results were compared with those observed on normal PB leukocytes and BM B-cell precursors from patients with non-neoplastic hematological disorders.

RESULTS

Fas Ag was constitutively expressed by both monocytes and neutrophils, while lymphocytes expressed bcl-2 with no difference between B and T cell subsets. Interestingly, bcl-2 expression was always absent on neutrophils. When dealing with ANLL patients, a relatively low bcl-2 and high Fas Ag phenotype characterized subtypes with granulocytic (M2) or promyelocytic (M3) differentiation. This observation was confirmed in a small number of patients for whom bcl-2 levels were quantified as antibody binding capacity (ABC) in molecules/cell. Leukemic cells from patients with ALL constitutively expressed bcl-2, the pattern of this expression being quantitatively lower than that of immature B-cell precursors. Finally, high bcl-2 and low Fas Ag expression represented a crucial part of the B-cell CLL immunophenotype.

CONCLUSIONS

Although based on a small number of patient and control samples, our results suggest that bcl-2 and Fas Ag are coordinately expressed on normal PB leukocytes. Fas Ag is expressed at low levels on B-CLL cells, generally considered long-surviving cells. The relatively lower bcl-2-expression detected in both M2 and M3 subtypes may explain, at least in part, the higher remission rates obtained in these forms of ANLL than in other less differentiated morphological variants.

Clinical and prognostic evaluation of bone marrow infiltration (biopsy versus aspirate) in early chronic lymphocytic leukemia. A single center study

BACKGROUND AND OBJECTIVE

Recently published studies dealing with chronic lymphocytic leukemia (CLL) patients in early clinical stage reported that bone marrow (BM) biopsies and aspirates can be considered complementary methods of evaluating the extent of BM involvement. Consequently, we designed the present study to investigate the clinical and prognostic implications of BM biopsies and aspirates in a series of stage A CLL patients followed-up in a single center.

PATIENTS AND METHODS

BM biopsy sections and aspirate smears obtained at the time of diagnosis from 102 CLL stage A patients were retrospectively evaluated. Results were correlated with clinical and hematological features as well as with survival and disease-progression risk.

RESULTS

Diffuse (D) BM histology was detected in 10 patients (9.8%) while 21 (20.5%) displayed lymphocyte infiltration (LI) > 80%. Twenty-six patients (25.4%) died with a 5- and 10-year survival probability of 85% and 50%, respectively. The survival of patients with D-BM histology was significantly shorter than that of patients with non-diffuse (non-D) histology (p < 0.05). Interestingly, when considering only CLL-related deaths (i.e. leukemia progression, infections) were considered, there was an increase in the statistical significance of BM histology (p = 0.01). There was no difference in life expectancy in cases with LI either using different cut-off levels (i.e. 70% and 80%) or excluding non-CLL related deaths. According to our experience, disease progression could only be predicted by BM histology (p = 0.008), while LI was not useful for forecasting progression to more advanced stages (p = NS).

INTERPRETATION AND CONCLUSIONS

In patients with early CLL, BM histology provides more reliable information regarding the clinical outcome of the disease than LI.

Prognostic features and therapeutical approaches in B-cell chronic lymphocytic leukemia: an update

In the past few decades important progress has been made in the understanding of chronic lymphocytic leukemia (CLL). Indeed, systematic studies of natural history and prognostic factors have made it possible to predict the outcome of disease. Although clinical stage (i.e. Rai and Binet stages) is the strongest predictor of survival, additional prognostic parameters, including patterns of bone marrow (BM) infiltration, lymphocyte doubling time (LDT), immunophenotype and cytogenetics, have now been identified. Furthermore, criteria of smoldering CLL (i.e. stage A, low lymphocyte count, non-diffuse BM histology, relatively high hemoglobin level, LDT > 12 months) allow identification of a subgroup of patients with indolent course and good prognosis for whom treatment should be delayed, unless progression occurs. Recent meta-analysis of clinical trials has demonstrated no survival advantage for immediate versus referred treatment in low clinical stages. The same considerations apply when comparing combination versus single-drug regimens. Purine analogues like fludarabine, 2'-chlorodeoxyadenosine and 2'-deoxycoformicin are active in CLL. Data on these drugs come from uncontrolled clinical trials; randomized studies are in progress. In addition, some issues concerning the relationship between response and survival, cross-resistance between purine analogues and eradication of the CLL clone, remain still unresolved. There are also increasing data on bone marrow transplants in CLL, although the high treatment-related mortality suggests that this procedure may have some benefit only in selected refractory young CLL patients with adverse features. This review will focus on recent progress in the prognosis and therapy of CLL. Issues that remain controversial will be a matter of discussion.

Combined Use of Interferon Alpha-2, Chlorambucil and Prednisone in Previously Treated Patients with Low-Grade Non-Hodgkin's Lymphomas. Results of a Phase II Study

Aims and Background

To investigate therapeutic activity and safety of alpha-interferon (alpha-IFN) in combination with chlorambucil (CLB) and prednisone (PDN), we treated 9 low-grade non-Hodgkin lymphoma patients with clinical evidence of relapsed (5 cases) or resistant (4 cases) disease with such an association.

Methods

In all instances, treatment consisted of alpha-2a IFN administered by subcutaneous route thrice weekly for 3 weeks, CLB, 5 mg/day for 21 days, and PDN, 30 mg three times a week for 3 weeks. Cycles were repeated every 28 days.

Results

A well-documented clinical response was observed in 6 (4 CRs+2 PRs) of 9 patients. Interestingly, 3 of 4 CRs were achieved in patients with histologically proven bone marrow involvement. Median duration of response was 18.5 months (range, 4-29 months). Myelosuppression was a common side effect. Two patients experienced grade 3 hematologic toxicity which did not preclude continuation of therapy.

Conclusions

As new purine analogues are not currently available, the combination of alpha-IFN, CLB, and PDN may represent, in our opinion, a valid therapy for patients not eligible for aggressive therapy such as autologous bone marrow transplantation.

Increased bcl-2/bax ratio in B-cell chronic lymphocytic leukemia is associated with a progressive pattern of disease

In order to evaluate clinical implications of altered expression of bcl-2 and bax proteins in B-cell chronic lymphocytic leukemia (CLL) we studied 27 patients with this disease. Cytofluorometric levels of bcl-2 did not reflect the status of disease. In contrast bax expression was lower in progressive than in non-progressive disease, therefore leading to a higher bcl-2/bax ratio in patients of the former group. If confirmed in longitudinal studies, quantitative cytofluorometric evaluation of bcl-2 and bax protein might help to identify patients with progressive disease who could possibly benefit from early therapy.

Clinico-biological implications of increased serum levels of interleukin-8 in B-cell chronic lymphocytic leukemia

BACKGROUND AND OBJECTIVE

Constitutive cellular expression and serum release of biologically active interleukin-8 (IL-8) has been reported in B-cell chronic lymphocytic leukemia (CLL). Given the autocrine role played by IL-8 in the process of cell accumulation characteristic of this disease we tried to investigate clinico-biological implications of increased serum levels of this cytokine in an unselected series of B-cell CLL patients.

DESIGN AND METHODS

Serum levels of IL-8 were determined at the time of diagnosis in 58 previously untreated B-CLL patients using an immunoenzyme assay. Results were correlated with main clinico-hematologic features as well as with the risk of disease progression. Finally, we looked for associations between IL-8 and molecules directly involved in apoptosis, such as intracellular bcl-2 and soluble APO-1/Fas.

RESULTS

Increased serum levels of IL-8 were found in 15 out of 58 (25.8%) B-cell CLL patients. Serum levels of IL-8 did not reflect clinico-biological features representative of tumor mass such as clinical stage, histopathologic pattern of bone marrow (BM) involvement, b2-microglobulin, sCD23 and sCD27 titers. Interestingly, circulating levels of IL-8 paralleled those of intracellular bcl-2 (r = 0.522; p = 0.01), thus confirming that the antiapoptotic effect of IL-8 can be exerted through a bcl-2 dependent pathway. Levels of IL-8 did not match those of soluble Apo-1/Fas (r = -0.013; p = 0.943). Finally, stage A patients with levels of IL-8 above the median value (i.e. 4.5 pg/mL) were more likely to progress to a more advanced clinical stage than those with levels below the median value (p < 0.05).

INTERPRETATION AND CONCLUSIONS

IL-8 is an interesting marker in B-cell CLL, closely involved in the pathogenesis of disease. Furthermore, it is useful for predicting the pace of disease progression in early clinical stages.

CD11c expression in B-cell chronic lymphocytic leukemia. A comparison of results obtained with different monoclonal antibodies

In order to obtain more information on the pattern of CD11c-positivity in otherwise typical B-cell chronic lymphocytic leukemia (CLL), we analyzed immunological and clinico-pathological features of 99 such patients studied with two different monoclonal antibodies (MoAbs). Fifty-two out of 70 (74.2%) patients stained with IOM-11C MoAb and 3 out of 29 (10.3%) patients stained with Leu-M5 MoAb had more than 30% CD11c positive cells (P < 0.0002). The two groups were similar with regard to the expression of B-cell CLL-related antigens (CD5, CD20, CD23), as well as clinico-pathological features (i.e. Binet clinical stage and pattern of bone marrow involvement), thus suggesting that differences in CD11c expression were due to different reactivity patterns of the MoAbs utilized. In our experience, the use of different reagents may affect immunophenotyping results, thus providing conflicting data at times.

Infections in chronic lymphocytic leukemia. Analysis of incidence as a function of length of follow-up

BACKGROUND AND METHODS

Infections represent the major cause of death in chronic lymphocytic leukemia (CLL); however, clinical studies dealing with their incidence have yielded inconclusive results. In order to address this issue we reviewed the records of 125 CLL patients (mean age 65.6 yrs; 81 M/44 F; Stage A, 48; Stage B, 37; Stage C, 40) followed up at our institution over a 10-year period.

RESULTS

The 125 patients accrued 447 person-years, a mean of 3.8 years per person. There were 199 recorded infections: 47 severe (crude rate 9.8 per 100 person-years) and 72 moderate, respectively. The 5-year risk of developing a severe infection for the whole series was 26% (95% CI: 24.7-27.3%), and 21 out of 71 deaths (29.5%) could be attributed to infectious causes. Despite a linear trend toward increased risk (r = 0.98), hazard function analysis showed a constant pattern of risk (r = 0.30), suggesting a lack of correlation of this event with time. Furthermore, the 5-year risk of developing a severe infection increased to 57.1% (95% CI: 36.4-77.8%) for patients with low IgG levels (less than 6.5 gr/L), and to 68% for those with both low IgG levels and disease stage C. On the other hand, patients who experienced a severe infection more frequently had advanced clinical stage (P < 0.001), low IgG levels (P < 0.01) and diffuse bone marrow (BM) histology (P < 0.05).

CONCLUSIONS

Infection is a constant risk in CLL that is associated with shortened survival. Factors such as hypogammaglobulinemia and advanced disease appear to be the major predisposing factors.

Cellular expression and serum circulating levels of CD23 in B-cell chronic lymphocytic leukemia. Implications for prognosis

BACKGROUND

CD23 is a functionally relevant molecule in B-cell chronic lymphocytic leukemia (CLL) which mediates growth and differentiation signals in B-cells. An intriguing feature of CD23 is its ability to be cleaved from the cell surface and released into the serum.

MATERIALS AND METHODS

Serum levels of soluble CD23 (sCD23) were determined with a sandwich enzyme immunoassay at the time of diagnosis in 90 previously untreated CLL patients, in order to evaluate whether they reflected disease activity and tumor load. Results were correlated with those dealing with CD23 expression on leukemic cells to verify whether the cellular counterpart determines serum levels.

RESULTS

CD23 was detected on peripheral blood mononuclear cells (PBMC) from 78 out of 90 (86.6%) B-CLL patients, without correlation with clinical stage. Circulating levels of sCD23 in the serum of patients with CLL were highly elevated in comparison to 15 normal controls (p < 0.0005); this increase reflected tumor mass as defined by either clinical stage (p < 0.0005) or bone marrow (BM) histology (p < 0.0005). Neither percentage nor absolute number of CD23+ cells correlated with circulating levels. Interestingly, life expectancy was significantly shorter in patients with higher serum levels of sCD23 (p < 0.0005). When integrated into the Binet clinical staging system, sCD23 led to isolation of two subgroups with different prognosis among intermediate-risk patients. Furthermore, longitudinal studies support the idea that sCD23 can be utilized as an indicator of disease progression.

CONCLUSIONS

sCD23 is a highly sensitive and suitable marker with prognostic potential in B-CLL.

CD27 in B-cell chronic lymphocytic leukemia. Cellular expression, serum release and correlation with other soluble molecules belonging to nerve growth factor receptors (NGFr) superfamily

BACKGROUND AND OBJECTIVE

CD27, a transmembrane homodimer belonging to the nerve growth factor (NGF) receptor superfamily, is typically expressed on leukemic CD5+ cells in B-cell chronic lymphocytic leukemia (CLL) and found in soluble form in the serum of CLL patients. Therefore, we investigated clinico-biological implications of increased serum levels of sCD27 in an unselected series of B-CLL patients.

DESIGN AND METHODS

Serum CD27 (sCD27) levels were determined at the time of diagnosis in 82 previously untreated B-CLL patients using a sandwich enzyme-linked immunoassay (ELISA). Results were correlated with either clinico-hematological or biological features. Finally, quantitative flow cytometric analyses of membrane CD27 (mCD27) expression were carried out on peripheral blood (PB) cells of 22 B-CLL patients and 5 healthy controls, respectively.

RESULTS

CD27 was found to be expressed on the surface of both resting normal and leukemic B cells. sCD27 levels were significantly higher in B-CLL patients (median value 2150 U/mL) than in healthy controls (median value 220 U/mL) (p < 0.0001). There was a close relationship between sCD27 and soluble TNF-alpha, another molecule belonging to the NGF receptor superfamily. Changes in sCD27 level correlated with clinical stage, beta 2 microglobulin and LDH.

INTERPRETATION AND CONCLUSIONS

These findings indicate that sCD27 is a reliable marker of tumor mass in B-CLL. Its potential prognostic value should be tested in prospective studies.

Comparative flow cytometric evaluation of bcl-2 oncoprotein in CD5+ and CD5- B-cell lymphoid chronic leukemias

BACKGROUND AND OBJECTIVES

Levels of intracellular bcl-2 oncoprotein have been found to be increased in leukemic cells of CD5+ B-chronic lymphocytic leukemia (CLL) patients. However, it is not clear whether bcl-2 overexpression is a peculiar feature of CD5+ B-CLL. Based on this background we carried out a quantitative flow cytometric evaluation of intracellular bcl-2 levels on leukemic cells of CD5+ and CD5- B-CLL.

METHODS

We assessed in flow cytometry levels of bcl-2 protein using a quantitative indirect immunofluorescence assay (QIFI kit) on samples from 46 previously untreated CD5+ B-CLL patients. Results were compared with those obtained on either normal peripheral blood B-lymphocytes or leukemic cells from 7 CD5- B-CLL patients intentionally selected for statistical comparison.

RESULTS

A relatively homogeneous amount of bcl-2 protein which did not reflect either clinical-biological features at the time of diagnosis nor in vivo response to therapy was found. Results expressed as antibody binding capacity (ABC) accounted for a mean value of 12.2 +/- 1.5 x 10(3) molecules/cell (range, 6.4-13 x 10(3) molecules/cell). Levels of bcl-2 detected on CD5+ B-CLL leukemic cells were significantly lower than those of B peripheral blood lymphocytes from healthy donors (p = 0.0001). The same applied when comparing CD5+ and CD5- B-CLL patients (bcl-2 ABC, 8.07 +/- 0.26 x 10(3) molecules/cell vs. 12.2 +/- 1.5 x 10(9) molecules/cell; p = 0.0001).

INTERPRETATION AND CONCLUSIONS

According to the role of bcl-2 in preventing apoptosis, our results indicate that differences in the pattern of expression of such an oncoprotein, might, at least in part, explain the more aggressive clinical course of CD5- B-CLL forms.

Prophylaxis against infections with low-dose intravenous immunoglobulins (IVIG) in chronic lymphocytic leukemia. Results of a crossover study

BACKGROUND

In a recently reported study, low doses of intravenous immunoglobulins (IVIG) were shown to be as effective as high doses in protecting chronic lymphocytic leukemia (CLL) patients against infections, although a control group was not included. With this background we started a crossover study of low-dose IVIG prophylaxis aimed at investigating its superiority over empirical treatment of infections.

MATERIALS AND METHODS

Forty-two CLL patients with hypogammaglobulinemia (IgG < 600 mg/dL) and/or a history of at least one episode of severe infection in the 6 months preceding inclusion in the study were randomly allocated to receive either an infusion of 300 mg/kg IVIG every 4 weeks for 6 months or no treatment. Then they were switched to observation or IVIG for another 12 months; finally, they received IVIG or no therapy for 6 more months.

RESULTS

A significantly lower incidence of infectious episodes was observed during IVIG prophylaxis in 30 patients who completed the 6-month period of either observation or IVIG therapy. The same applied to the 17 patients who completed 12 months of either observation or IVIG prophylaxis. Interestingly, the restoration of serum IgG levels obtained in 17 out of 25 patients (mean percent value of IgG increase, 41.8%) did not parallel a decrease in the number of infectious episodes.

CONCLUSIONS

A protective effect against infections is demonstrated for low-dose IVIG in the present study. A benefit was shown in patients who completed either 12 or 6 months of IVIG prophylaxis; however, even this low-dose treatment is not a cost effective way to prevent infection in CLL patients.