LYMPHOID HEMOPATHY PROGNOSIS METHOD

Abstract

A prognosis method including determining the quantity of B lymphocyte cells secreting interleukin 10 in a biological sample, such that a patient from whom the tumor sample is taken having a quantity of B cells secreting IL-10 below 5% will have more than a 50% chance of a good prognosis after a treatment.

Claims

1.-11. (canceled)

12. An in vitro prognosis method for a therapeutic response, mediated by a therapeutic antibody, in a tumor sample taken from a patient with a blood disease, said antibody being an antibody depleting the cells of said blood disease, the method comprising an in vitro determination of a quantity of B lymphocyte cells secreting interleukin 10 in said sample, such that a patient from whom the tumor sample is taken, having the quantity of B cells secreting IL-10 below 5% of a total cell quantity of said sample, will have more than a 50% chance of experiencing a depletion of more than 90% of cells of said blood disease after treatment with said therapeutic antibody.

13. The prognosis method according to claim 12, wherein the blood disease is a blood disease whose cells express CD20 surface marker.

14. The prognosis method according to claim 12, characterized in that the blood disease is chosen from among B-cell chronic lymphoid leukemia (CLL), diffuse large B cell lymphomas and all of the histological variants expressing CD20, follicular lymphomas and mantle cell lymphomas, marginal zone lymphomas, MALT (mucosa-associated lymphoid tissue) lymphomas, Burkitt lymphomas, lymphoplasmacytic lymphomas, Waldenstrm's disease, B-cell prolymphocytic leukemia, and all unclassifiable CD20+ B-cell lymphomas.

15. The prognosis method according to claim 12, wherein the quantity of B-cells secreting IL-10 in the sample below 5% is measured by flux cytometry, after activating the secretion of IL-10, or by measuring a circulating IL-10 level in serum.

16. The prognosis method according to claim 15, wherein the quantity of B-cells secreting IL-10 is measured by flux cytometry, after activating the secretion of IL-10 and deactivating this secretion.

17. The prognosis method according to claim 12, wherein said antibody is an IgG1 or IgG3 antibody.

18. The prognosis method according to claim 17, wherein said antibody is a CD20 antibody.

19. The prognosis method according to claim 17, wherein said antibody is a rituximab antibody.

20. An in vitro prognosis method for a therapeutic response, mediated by a therapeutic antibody, in a tumor sample taken from a patient with a blood disease, said antibody being an antibody depleting cells of said blood disease, the method comprising an in vitro determination in the cells of said sample: of the amino acid in position 176 of a sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 1, or in position 158 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 2, and of a quantity of B lymphocyte cells secreting interleukin 10 before a treatment, such that a patient, from whom the tumor sample is taken, having a valine in position 176 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 1, or in position 158 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 2, and having a quantity of B cells secreting IL-10 in the sample below 5% of the total cells of the sample will have more than a 50% chance of experiencing a depletion of more than 90% of the cells of said blood disease after treatment with said therapeutic antibody.

21. The prognosis method according to claim 20, wherein the blood disease is a blood disease whose cells express CD20 surface marker.

22. The prognosis method according to claim 20, characterized in that the blood disease is chosen from among B-cell chronic lymphoid leukemia (CLL), diffuse large B cell lymphomas and all of the histological variants expressing CD20, follicular lymphomas and mantle cell lymphomas, marginal zone lymphomas, MALT (mucosa-associated lymphoid tissue) lymphomas, Burkitt lymphomas, lymphoplasmacytic lymphomas, Waldenstrm's disease, B-cell prolymphocytic leukemia, and all unclassifiable CD20+ B-cell lymphomas.

23. The prognosis method according to claim 20, wherein the quantity of B-cells secreting IL-10 in the sample below 5% is measured by flux cytometry, after activating the secretion of IL-10, or by measuring a circulating IL-10 level in serum.

24. The prognosis method according to claim 23, wherein the quantity of B-cells secreting IL-10 is measured by flux cytometry, after activating the secretion of IL-10 and deactivating this secretion.

25. The prognosis method according to claim 20, wherein the in vitro determination of the amino acid in position 176 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 1, or in position 158 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 2, is done by polymerase chain reaction (PCR).

26. The prognosis method according to claim 20, where said antibody is an IgG1 or IgG3 antibody.

27. The prognosis method according to claim 26, wherein said antibody is a CD20 antibody.

28. The prognosis method according to claim 26, wherein said antibody is a rituximab antibody.

29. A kit for the in vitro prognosis of a patient with a blood disease who may be treated with a therapeutic antibody, comprising: a detector for detecting an amino acid in position 176 of a sequence of a FcRIIIa receptor, as shown by sequence SEQ ID NO: 1, or in position 158 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 2, means for determining quantity of B lymphocyte cells secreting IL-10, and one or several control samples.

30. The prognosis kit according to claim 29, in which the detector for detecting the amino acid in position 176 of a sequence of a FcRIIIa receptor, as shown by sequence SEQ ID NO: 1, or in position 158 of the sequence of the FcRIIIa receptor, as shown by sequence SEQ ID NO: 2, comprise oligonucleotides with sequences SEQ ID NO: 3, SEQ ID NO: 4 and SEQ ID NO: 5.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0097] The invention will be better understood in light of the five figures and the example that follow:

[0098] FIG. 1 shows a graph showing that the frequency of competent IL-10 leukemia B-cells (CLL) correlates with the plasma IL-10 level on day D0, i.e., before treatment (p=0.0166, r=3969). The x-axis shows the quantity of IL-10 in the plasma in pg/mL and the y-axis shows the percentage of B10 cells expressed in %.

[0099] FIG. 2 shows a graph showing that the lymphocyte depletion observed after treatment with rituximab alone (D22) is statistically influenced by the frequency of competent IL-10 CLL B-cells on D0 (p=0.004). The x-axis shows the percentage of lymphodepletion in % and the y-axis shows the percentage of B10 cells expressed in %.

[0100] FIG. 3 shows a graph showing that the lymphocyte depletion is statistically greater for patients carrying the FcRIIIa-158V allele.

[0101] FIG. 4 shows ROC curves (Receiver operating curve) generated by logistical regression for the percentage of competent IL-10 CLL B-cells alone (A) (AUC=0.763), the carriers of V vs F/F alleles for the FCGR3A polymorphism (B) (AUC=0.675) and a combination of the percentage of competent IL-10 CLL B-cells and carriers of V vs F/F alleles for the FCGR3A polymorphism (C) (AUC=0.855).

[0102] FIG. 5 shows a graph comparing the number of regulator B-cells (listed by cytometryx-axis) to the assay of the IL-10 secretion (quick methody-axis). The rectangle with broken lines corresponds to individuals with a low risk of suboptimal response with a CD20 antibody. The rectangle with solid lines corresponds to individuals with a high risk of suboptimal response with a CD20 antibody. A good correlation is observed between the results obtained by the two methods (R=0.79).

DETAILED DESCRIPTION OF THE INVENTION

Examples

Example 1

[0103] The action mechanisms of rituximab (marketed under the brands MabThera, Rituxan) remain unknown and could be different depending on the subtype of B-cell lymphoproliferative disorder. Rituximab is known for causing in vitro apoptosis, complement-dependent cytotoxicity (CDC), antibody-dependent cellular cytotoxicity (ADCC) and antibody-dependent phagocytosis (ADPC), and some results tend to involve these mechanisms in vivo.

[0104] Certain factors affecting the response to rituximab have recently been discovered.

[0105] The inventors had previously observed that the polymorphism of the FcRIIIa V/F receptor affects the clinical response to rituximab treatment. Since this polymorphism alters the affinity of the constant part of IgG1 for the FcRIIIa receptor (expressed on the surface of NK cells and macrophages), the inventors put forth the hypothesis that the ADCC mechanism should be significant in the context of treating follicular lymphomas treated with rituximab.

[0106] Regulator B-cells were recently identified in humans and mice as a cellular subtype capable of secreting IL-10. These B-cells, also called B10 cells or B10, are characterized by their ability to modulate inflammation, autoimmunity and the innate or adaptive immune response based on IL-10 production. In a murine model, the B10 cells would be capable of inhibiting the elimination of lymphoma cells induced by a CD20 antibody, by acting on the monocyte functions mediated by the constant parts of the immunoglobulins (Fc parts).

[0107] More recently, it has been shown that the cells of a clonal chronic lymphoid leukemia (CLL) had immunosuppressive properties and an IL-10 competence.

[0108] The inventors then considered that the competent IL-10 cells of CLL's could affect the efficacy of treatment with rituximab in patients with this type of leukemia.

[0109] Patients and Treatment

[0110] The Patients.

[0111] A prospective and randomized phase II study was conducted in 59 French centers including 140 patients between June 2012 and January 2013. The patients not having received any prior treatment (between 18 and 65 years of age) and whose chronic lymphoid leukemia diagnosis was confirmed by immunophenotyping according to the 2008 IWCLL criteria (Binet stage C, or Binet stage A or B with an active disease), were enrolled in this trial. An additional inclusion criterion was taken into account: the absence of 17p deletion, evaluated by FISH (<10% positive cores). All patients provided informed consent in writing before inclusion.

[0112] Randomization.

[0113] The patients were stratified based on their IGVH mutational status after FISH analysis (11q deletion) and assigned randomly with a 1:1 ratio to receive either the standard dose of FCR (fludarabine, cyclophosphamide and rituximab) chemotherapy for arm A or Dense-FCR with a rituximab pre-phase before standard FCR treatment for arm B.

[0114] Treatment.

[0115] The standard FCR treatment consists of 6 rounds separated by 28 days of rituximab (375 mg/m.sup.2 for the 1st round D1 and 500 mg/m.sup.2 for the other rounds), fludarabine (40 mg/m.sup.2/d D2-4), cyclophosphamide (250 mg/m.sup.2/d D2-4). For the experimental arm, the FCR treatment is preceded by a rituximab pre-phase made up of 4 successive infusions, distributed as follows: 500 mg on D0, and 2000 mg on D1, D8 and D15. Round 1 begins on D22 for the patients, and the following rounds are separated by 28 days.

[0116] Identification of Competent IL-10 Cells in CLL's, IL-10 Trial and FCGR3A Genotyping

[0117] The competent IL-10 cells of the CLL's are identified by flux cytometry from purified mononucleated blood cells after polyclonal stimulation.

[0118] The peripheral blood mononucleated cells (PBMC) have been purified from patients with CLL in arm B of the study using a Ficoll-Hypaque density gradient (Eurobio, Courtaboeuf, France).

[0119] The PBMC's were resuspended (910.sup.6 cells/mL) in medium (RPMI 1640Biotech GmbH, Aidenbach, Germany) containing 10% fetal calf serum (Eurobio, Courtaboeuf, France), 2 mM of L-glutamine (Eurobio, Courtaboeuf, France), 100 U/mL of penicillin, 100 g/mL of streptomycin, and 2.5 g/mL amphotericin (all antibiotics are obtained from Tebu-bio, Le Perray-en-Yvelines, France).

[0120] The clonal activation of the B lymphocytes was done with CpG (ODN 2006, 10 g/mL; InvivoGen, San Diego, USA), CD40L (50 ng/mL; R&D Systems, Minneapolis, Minn., USA) and anti-polyHistidine (500 ng/mL; R&D Systems, Minneapolis, Minn., USA) for 48 hours at 37 C. in a humid atmosphere comprising a 5% CO.sub.295% air mixture. Phorbol myristate acetate (PMA, 50 ng/mL; Sigma-Aldrich, St. Louis, Mo., USA) and ionomycin (1 g/mL; Sigma-Aldrich, St. Louis, Mo., USA) were added to stimulate IL-10 production. After 4 h at 37 C. in a humid atmosphere comprising a 5% CO.sub.2-95% air mixture, brefeldin A (1solution/mL; BioLegend, San Diego, Calif., USA) was added to block the IL-10 secretion in order to identify the B10 cells. The cells were marked using the following antibodies: anti CD19 BV421 (HIB 19), anti CD69 PE/Cy7 (FN 50), anti CD38 APC (HIT 2), anti IL-10 PE (JES3-9D7) from BioLegend (San Diego, Calif., USA), and anti CD45 KO (J.33) and anti CD5 FITC (BL1a) from Beckman Coulter (Brea, Calif., USA). The clonal CLL cells were identified as being CD19+CD5+CD20int lymphocytes. The analyses were done using a CyAn ADP flux cytometer (Beckman Coulter, Brea, Calif., USA).

[0121] The plasma IL-10 level was determined using the LUMINEX technology on magnetic balls, through a laser double reading system, according to the manufacturer's recognitions (R&D Systems, Minneapolis, USA). The plasma of the 68 patients was incubated with superparamagnetic balls covered with IL-10 antibodies for 2 hours, at ambient temperature, diluted by half. The plasma IL-10 is quantified owing to the combination of two antibodies for detection: an IL-10 biotinylated antibody and a streptavidin antibody conjugated with phycoerythrin.

[0122] The FcRIIIA-158VF polymorphism is determined by nested PCR. Briefly, single-step allele-specific multiplex PCR tests were carried out as described in Dall'Ozzo et al. (Dall'Ozzo et al. J Immunol Methods. 2003; 277(1-2):185-192) with some modifications. The 25 L of reaction mixture comprises genomic DNA, 400 nM of forward primer (5-TCCAAA AGCCACACTCAAAGTC-3 (SEQ ID NO: 3)), 400 nM of reverse primer specific to allele V (5-AGACACATTTTTACTCCCATC-3 (SEQ ID NO: 4)) and 200 nM of reverse primer specific to allele F (5-GCGGGCAGGGCGGCGGGGGCGGGGCCGGTGATGTTCACAGTCTCTGA TCACACATTTTTACTCCCATA-3 (SEQ ID NO: 5)), 400 M of each nucleotide (dNTP), 2 mM of MgCl.sub.2 and 0.5 U of Taq DNA polymerase in its buffer (Promega, Madison, USA). The PCR conditions are: 3.5 min at 95 C. followed by 35 cycles, each cycle consisting of 95 C. for 20 sec, 56 C. for 20 sec, 72 C. for 30 sec. After amplification, the PCR products (137 bp for allele F and 81 bp for allele B) are separated on 8% acrylamide gel (Invitrogen, Carlsbad, USA) and viewed by marking with ethidium bromide.

[0123] Statistical Analyses

[0124] The data distributions were tested using the Shapiro-Wilk test. The .sup.2 and Fisher tests were used for categorical data.

[0125] The comparison of the medians was done using a Student T test or a Mann-Whitney test.

[0126] All of the variables with a value p<0.10 in the univariate analysis are included in an intermediate model. The variables of the final model were determined using the progressive elimination method using a Student T test (p<0.05 as significant model).

[0127] All of the statistical analyses were done by implementing, at a level a, 0.05 of a bilateral test using version 3.0.2.10 of the R software.

[0128] Results and Discussion

[0129] The lymphocyte depletion after monotherapy with rituximab was evaluated on day 22 (D22) to study the influence of IL-10 competent leukemia cells (B10) on the in vivo efficacy of rituximab, in 68 patients included in the experimental arm of the study.

[0130] The median of the number of leukemia lymphocytes before the four doses of rituximab (D0) was 91.13 G/L (range: 3.74-497.40) and was 2.60 G/L (range: 0.14-189.40) at the end of the pre-phase for treatment with rituximab alone (D22). Thus, the median lymphocyte depletion after the pre-phase for treatment with rituximab alone (D22) was 95.1% (range 77.0-99.9), among which 66% obtained a depletion of more than 90%.

[0131] The characteristics of the patients and their distribution based on 90% lymphodepletion are presented in table 1 below.

[0132] No significant correlation was found between 90% lymphodepletion and the clinical parameters of age, sex, Binet stage, IGHV mutation, cytogenetic anomaly or the 2 microglobulin.

[0133] A subgroup of B10 cells was identified in all of the tested patients (n=47, median: 3.06% of CLL cells, range: 0.12 to 29.55). The frequency of B10 cells among the leukemia cells correlates with the quantity of plasma IL-10 (FIG. 1, r=0.39, p=0.02), while the plasma IL-10 level does not correlate with the non-B10 leukemia cells (not shown). The frequency of B10 cells does not vary based on the characteristics of patients, and is also not significantly different in patients with CLL with or without IGVH mutation (median: 6.29%, range: 0.12 to 15.83 versus median: 1.85%, range: 0.23 to 20.81, respectively). Furthermore, such B10 cell frequencies also do not correlate with the cytogenetic alterations (del11q, del13q, trisomy 12).

[0134] A univariate analysis shows that the frequency of B10 cells affects the lymphodepletion above 90% after the pre-phase for treatment with rituximab alone (D22) (FIG. 2, p=0.004). The frequency of B10 cells also makes it possible to predict the 3 month clinical response (complete response) after the end of treatment with the fludarabine-cyclophosphamide combination and rituximab, considered to be the reference treatment for patients with chronic lymphoid leukemia (p=0.04).

[0135] These results show that all CLL patients have a subpopulation of B10 cells that represent a variable percentage of leukemia B-cells and that have a chemically significant in vivo inhibiting effect on the activity of rituximab.

[0136] Since the FcRIIIa-158V/F polymorphism correlates with the in vivo efficacy of rituximab in follicular lymphomas, the inventors have determined this polymorphism in the various patients of the cohort.

[0137] The FcRIIIa-158V/F polymorphism is significantly associated with a normal number of lymphocytes (<5 G/L) (p=0.03) and a lymphodepletion of 90% (p=0.03) at D22. This is also the case for carriers of the FcRIIIa-158V polymorphism (p=0.01) (FIG. 3).

[0138] Conversely, the FcRIIIa-158V/F polymorphism does not correlate with the 3 month clinical response after immuno-chemotherapy. Consequently, these results suggest that the immune functions mediated by FcRIIIa play a critical role in the activity of rituximab, but that the involvement of the FcRIIIa-158V/F polymorphism could be hidden either by the high activity of the immuno-chemotherapy, or by a direct inhibition of the effector immune cells by chemotherapy.

[0139] Logistical regression analyses show that only the frequency of B10 cells and the FcRIIIa-158V/F polymorphism are associated with a 90% lymphodepletion after the pre-phase for treatment with rituximab alone (D22) (close relative risk (OR)=0.83; confidence interval (CI): 0.72-0.3; p=0.002, and OR=4.95; 95% CI: 1; 07-27.48; p=0.04, respectively).

[0140] The receiver operating curve (ROC) built using the frequencies of the B10 cells and the FcRIIIa-158V/F polymorphism shows a very significant area under the curve (AUC) (AUC=0.855; 95% CI: 0.732-0.978), leading to a good distinction between patients who have a depletion of their lymphocytes of more than 90% and those who do not have such a depletion (FIG. 4).

[0141] From the ROC curve (FIG. 4), the following data can be extracted:

[0142] Data obtained from genotyping of the FcRIIIa receptor alone

[0143] Based on the population used during the multivariate analysis (n=44) (Area under the curve (AUC=0.669), 95% CI=0.514-0.824, the obtained results are as follows:

TABLE-US-00001 TABLE 2 Lymphodepletion > Lymphodepletion < 90% 90% Probability > 0.5 of obtaining 30 14 a lymphodepletion > 90% (true positives) (false positives) Probability < 0.5 of obtaining 0 0 a lymphodepletion > 90% (false negatives) (true negatives) Sensitivity = 100% Specificity = 0% Legend of table 2. Sensitivity and specificity for a lymphodepletion of 90% obtained by using the aforementioned computed probabilities.

[0144] One can see that a large number of false positives are identified (about 33%).

[0145] 1Data obtained from the number of B10 cells alone [0146] a. With a threshold of 4.34%

[0147] Based on the population used during the multivariate analysis (n=44) (AUC=0.779), 95% CI=0.652-0.905, the obtained results are as follows:

TABLE-US-00002 TABLE 3 Lymphodepletion > Lymphodepletion < 90% 90% Probability > 0.5 of obtaining 21 2 a lymphodepletion > 90% (true positives) (false positives) Probability < 0.5 of obtaining 9 12 a lymphodepletion > 90% (false negatives) (true negatives) Sensitivity = 70.00% Specificity = 85.71% Legend of table 3: identical to the legend of table 2. b. With a threshold of 3%

[0148] Based on the population used during the multivariate analysis (n=44) AUC=0.729, 95% CI=0.598-0.589, the obtained results are as follows:

TABLE-US-00003 TABLE 4 Lymphodepletion > Lymphodepletion < 90% 90% Probability > 0.5 of obtaining 18 2 a lymphodepletion > 90% Probability < 0.5 of obtaining 12 12 a lymphodepletion > 90% Sensitivity = 60.00% Specificity = 85.71% Legend of table 4: identical to the legend of table 2. c. With a threshold of 5%

[0149] Based on the population used during the multivariate analysis (n=44) AUC=0.726, 95% CI=0.586-0.866, the obtained results are as follows:

TABLE-US-00004 TABLE 5 Lymphodepletion > Lymphodepletion < 90% 90% Probability > 0.5 of obtaining 21 3 a lymphodepletion > 90% Probability < 0.5 of obtaining 9 11 a lymphodepletion > 90% Sensitivity = 70.00% Specificity = 78.57% Legend of table 5: identical to the legend of table 2.

[0150] One can see that by imposing the threshold at less than 5% B10 cells, the false positive level falls to about 5%. B10 detection significantly improves the prognosis relative to the polymorphism.

[0151] 2Data obtained from genotyping of the FcRIIIa receptor and B10

[0152] Based on the population used during the multivariate analysis (n=44) (Area under the curve (AUC=0.0855), 95% CI=0.732-0.978, the obtained results are as follows:

TABLE-US-00005 TABLE 6 Lymphodepletion > Lymphodepletion < 90% 90% Probability > 0.5 of obtaining 26 5 a lymphodepletion > 90% Probability < 0.5 of obtaining 4 9 a lymphodepletion > 90% Sensitivity = 86.67% Specificity = 64.29% Legend of table 6: identical to the legend of table 2.

[0153] In the preceding tables 2 to 6, the sensitivity is the measurement of the test's ability to identify a condition when it is present. It is the proportion of true positives (TP) that is detected by the test as having the condition [Se=TP/(TP+FN)]. The false negatives are the patients having the condition, but not detected by the test. A high sensitivity is preferred in screening, since it is easier to reject patients not having the condition.

[0154] The specificity is the test's ability to exclude a condition when it is absent. It is the proportion of true negatives (TN) that is detected by the test as not having the condition [Sp=TP/(TP+FP)]. The false positives are the patients not having the condition, but detected by the test as being positive.

[0155] By combining the detection of the polymorphism and the number of B10 cells, one sees an improvement in the prognosis, and in particular a very significant reduction in the number of patients considered to be false negatives. Furthermore, the double detection makes it possible to obtain, for thresholds below 5% of B10, a good sensitivity (around 86%).

[0156] The inventors have therefore determined a B10 population in patients with chronic lymphoid leukemia and receiving rituximab. At the same time, they have determined the FcRIIIa-158VF phenotype of these patients. They have thus been able to show that the frequency of B10 cells and the FcRIIIa-158VF polymorphism were influencing the response to rituximab in patients with chronic lymphoid leukemia. The analysis of these two factors combined makes it possible to improve the sensitivity and specificity of prediction of the prognosis response.

TABLE-US-00006 TABLE 1 Lymphodepletion Lymphodepletion Univariate Multivariate >90% (n = 44) 90% (n = 23) analysis analysis Median Median OR Value AUC OR Value N (%) (IQR) N (%) (IQR) [95% CI] of p [95% CI] [95% CI] of p Age (years) 44 (100.00) 55.72 23 (100.00) 53.99 0.792 (51.31-58.12) (52.07-57.41) men 31 (70.45) 18 (78.26) 0.68 0.693 [0.19-2.16] Binet AB stage 36 (81.82) 16 (69.57) 0.51 0.404 [0.15-1.73] ECOG 0 31 (70.45) 5 (21.74) 1.50 0.572 [0.41-6.29] IGHV on 25 (56.82) 16 (72.73) 0.50 0.324 mutated [0.15-1.50] Cytogenetic anomalies 18 (51.43) 8 (44.44) 1.31 0.848 Del(13q) [0.41-4.29] Del(11q) 7 (16.67) 6 (26.09) 0.57 0.560 [0.16-2.07] Trisomy 12 2 (6.06) 2 (14.29) 0.40 0.572 [0.03-6.04] 2 39 (88.64) 2.90 22 (95.65) 2.76 0.857 microglobulin (2.33-3.66) (2.33-4.29) (mg/L) CD38+ (%) 34 (77.27) 2.00 17 (73.91) 10.50 0.133 (0.00-23.00) (1.75-26.00) Competent 32 (72.73) 2.30 15 (65.22) 9.51 0.004 0.763 0.83 0.002 IL-10 cells (%) (0.68-6.47) (5.35-15.70) [0.604-0.921] [0.72-0.93] FCGR3A 0.028 V/V 5 (11.91) 1 (4.55) V/F 25 (59.52) 7 (31.82) F/F 12 (28.57) 14 (63.63) FCGR3A V 30 (71.43) 8 (36.36) 4.23 0.014 0.675 4.95 0.043 carriers [1.43-13.42] [0.551-0.799] [1.07-27.48] Table 1: Parameters influencing the lymphocytosis caused by rituximab in monotherapy in 68 patients with chronic lymphoid leukemia (CLL) The group of patients having a lymphocyte inhibition of less than 90% was used as a reference for odds ratios.

Example 2

[0157] The inventors have also proposed a simple and quick method for analyzing the frequency of regulator B-cells. This method was compared to the reference method, which is based on the intra-cytoplasmic detection of the IL-10 synthesis influx cytometry.

[0158] The method consists of a first step for enriching the population of B lymphocytes, which normally make up the great minority among leukocytes, by negative selection from a peripheral blood sample. The blood is collected on a phase separator tube (of the CPT vacutainer BD type). Alternatively, the blood can be transferred from a primary tube containing an anticoagulant to the phase separator tube.

[0159] A depletion cocktail of the RosetteSep StemCell type is added to the blood sample in order to enrich the leukocyte phase with B lymphocyte. At the end of this step comprising an incubation, a separation by centrifugation and two cell wash centrifugations, a population containing more than 90% B lymphocytes is obtained.

[0160] The cells are counted by visual counting (cells under microscopy) or by cytometry in order to prepare a concentration of 6 million cells per mL and placed in culture in a medium containing RPMI+10% fetal calf serum. The cells are stimulated by a cocktail of polyclonal activators, such as a mixture of lignant-histidine+antibody CPG and CD40 patterns. An unstimulated well is done in parallel. The culture is kept for 18 h to 48 h at 37 C. without intervention.

[0161] At the end of this incubation, the cell culture is centrifugated and the culture supernatant is collected and kept at 20 C. or tested immediately through an enzyme-linked immunoabsorbent assay (EIA) in order to determine the concentration of IL-10 secreted.

[0162] The data obtained compared to the reference method in cytometry shows that the method by EIA assay of the IL-10 secretion provides well-correlated results between the two methods. Thus, the cell samples containing a proportion of regulator B-cells greater than 5% are correctly identified (FIG. 5).