ALBERT

Description: Abstract 731 Paroxysmal Nocturnal Hemoglobinuria (PNH) is an acquired disorder of hemopoietic stem cells (HSCs). Affected individuals experience intravascular hemolysis and have a predisposition to thromboembolism, renal impairment and pulmonary hypertension. These symptoms vary in severity, but in general, the higher the proportion of PNH blood cells produced, the more severe the symptoms. The molecular pathogenesis in PNH is known to relate to a single gene mutation in the X-linked phosphatidylinositol glycan class A gene (PIG-A) which causes a complete or partial deficiency of glycophosphatidylinositol (GPI) anchored proteins leading to the symptoms of the disease. The recent development of eculizumab therapy in PNH has had a dramatic impact in reducing both morbidity and mortality in the disease but PNH remains incurable. A sub-optimal response to eculizumab, certainly when assessed by transfusion requirements, is often due to the underlying bone marrow failure that is considered to be universally present in PNH. The factors that lead from the development of a mutant clone to clonal expansion and symptomatic disease are poorly understood but are the key to improving responses and potentially to move towards a cure. Bone marrow failure appears to provide the environment necessary for expansion of PNH clones and small PNH clones are often detected in aplastic anemia and less commonly in myelodysplasia. There is no evidence that PNH HSCs have an intrinsic proliferative advantage compared to normal HSCs and an immune-mediated extrinsic suppression of normal hematopoiesis with a selective advantage for the PNH cells over residual normal stem cells is likely to explain the preferential development of PNH clones concurrent with bone marrow failure. To gain a better understanding of clonal expansion in PNH we have developed an in vitro PNH bone marrow culture model using a stromal cell line which allows PNH stem cells to be maintained in long term cultures and their capacity to form progenitor cells in myeloid colony forming assays to be assessed. We have evaluated bone marrow from 11 patients with PNH (median age 47 years, median granulocyte clone size 95.3%) and 10 normal controls (median age 42 years) within these long term bone marrow culture experiments. Unmanipulated bone marrow mononuclear cells (MNCs), CD34 selected cells and MNCs with their T-cell component depleted were used in this model. This in vitro model provides the environment for PNH stem cells to be maintained for up to eight weeks and, unlike previous studies, produce progenitor cells. When the patient's MNC's are used to seed the culture system there is poor growth of the culture which is only maintained for a median of 2 weeks. However if CD34 selected cells are used then the cultures are maintained for up to 8 weeks (similar to the normal controls) suggesting that there is a component within the MNC's that is responsible for suppressing the marrow culture which is removed by CD34 select. We next selectively removed the T-cells from the PNH MNC's and demonstrated that the marrow cultures now survived to the extent of the controls (see Figure). This demonstrates that the immune suppression in PNH resides in the T-cells. The progenitor cells produced in both the CD34 selected or the T-cell depleted MNCs over the course of the long term culture experiments show an increase in the proportion of normal progenitors the longer the cultures are maintained. This supports the hypothesis that PNH stem cells have no intrinsic proliferative advantage over normal HSCs and that T-cells are the critical cell suppressing the normal hematopoiesis in PNH. We are now examining the specific cell type that causes the myelosuppression in PNH and which will facilitate a targeted approach to the treatment of bone marrow failure in PNH and related disorders. In conclusion we have developed an in vitro PNH bone marrow culture model that allows the immune insult in PNH to be evaluated. Furthermore, this work confirms the important role that T-cells play in the etiology of PNH and provides a model with which to define the exact mechanism of suppression of hematopoiesis in PNH (and aplastic anemia). This information is essential to develop targeted therapies for the marrow failure seen in PNH and aplastic anemia. Disclosures: Kelly: Alexion Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Richards:Alexion Pharmaceuticals: Honoraria, Speakers Bureau. Arnold:Alexion Pharmaceuticals: Honoraria. Hill:Alexion Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Hillmen:Alexion Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

Description: In past 22 years, we have identified using flow cytometry 705 patients with detectable PNH (GPI deficient) populations of granulocytes, monocytes and red cells in the peripheral blood in samples sent for diagnosis. We undertook an analysis of presenting clinical features, blood count data and PNH clone sizes in order to better understand the natural history and provide a more objective classification of disease. Based on serial flow cytometry measurements of PNH clone sizes, we also studied disease stability, frequency of recovery and progression with an aim to guiding future management of individual patients. Clinical classification of patients at presentation was as follows; aplastic anemia (58%), hemolytic anemia (36.1%); myelodysplasia (2.5%); thrombosis (2.4%); hemolysis & thrombosis (0.6%), myeloproliferative neoplasm (0.3%); Fanconi anemia (0.1%). Median age at presentation was 45 years (observed range 0.5 – 90 years) and the Male:Female ratio was 1.05. Descriptive statistical analysis of presenting blood count data revealed novel gender related features not previously described in PNH. At presentation, pancytopenia was found in 61% of male and 47% of female patients; a normal blood count was present in only 0.3% of males and 4% of females. A combined low red blood cell count (RBC) and white cell count (WBC) was the most frequent bicytopenia affecting 19% males and 22% females. Leucopenia as a sole abnormality did not occur in males and was present in

Description: Paroxysmal nocturnal hemoglobinuria (PNH) is characterized by a lack of the terminal complement inhibitor CD59 on erythrocytes that renders these cells susceptible to chronic hemolysis. Eculizumab blocks terminal complement resulting in reductions in hemolysis, thrombotic events, renal impairment and transfusion requirement, as well as improvement in quality of life. The standard dosing regimen for eculizumab is 600 mg/week for 4 weeks (induction); 900 mg one week later; and then 900 mg every 14 ± 2 days (maintenance). This regimen maintains eculizumab levels 〉35 μg/mL, which is sufficient to completely and consistently block complement-mediated hemolysis in patients with PNH. In PNH clinical trials, 900 mg of eculizumab every 14 ± 2 days effectively and consistently blocked complement-mediated hemolysis in 98% of patients (n=195). During the studies, 10–15% of patients experienced an increase in hemolysis (elevation of LDH) near the end of the 14-day dosing interval with a return of pre-eculizumab symptoms such as hemoglobinuria, dysphagia, abdominal pain, or fatigue. The dosing interval was reduced to 12 days, as specified by label, resulting in sustained complement blockade, control of hemolysis and resolution of symptoms in nearly all patients. Three of the original 195 patients (2%) were not consistently blocked with the approved dosing regimen. Alternative eculizumab dosing regimens were investigated in these patients to assess their effectiveness and safety. Two different dosing regimens were employed; both included a maintenance phase with 1200 mg every 14 days. One regimen also included an induction period of 900 mg weekly for 5 doses. LDH, pharmacokinetics (PK), and clinical signs of complement breakthrough were monitored. The time from first eculizumab treatment to initial breakthrough on the 900 mg every 14 days ranged from 2 to 19 mo., and the reduction in the dosing interval to 900 mg every 12 days, as specified in the label, did not adequately control hemolysis in each of these 4 patients. Patient 1 was managed for 6 mo. with 900 mg every 12 days before experiencing additional complement breakthrough episodes (Figure, panel A). LDH levels (closed diamonds) reached 9234 U/L (ULN, 430-450 U/L) and breakthrough symptoms occurred 2 days prior to the next dose. The patient was re-induced with 900 mg eculizumab every 7 days for 5 weeks followed by 1200 mg every 14 days. Trough levels of eculizumab increased (open circles) each week during the induction phase (42.7 – 81.8 μg/ml) resulting in an immediate reduction in LDH to near normal levels. A maintenance dose of 1200 mg every 14 days in this patient resulted in sustained complement blockade. Patient 2 experienced breakthrough hemolysis after 19 mo. of standard dosing. Complement breakthrough occurred during a post-cholecystectomy infective endocarditis. After an adjustment to 900 mg every 12 days did not control complement breakthrough (Figure, panel B), the dose was changed to 1200 mg every 14 days without re-induction. This regimen resulted in sufficient levels of eculizumab to consistently reduce hemolysis to near normal levels. Further episodes of hemoglobinuria and other symptoms of hemolysis were not observed. Two additional patients received 1200 mg every 14 days without re-induction, one following complement breakthrough on the approved dose and the other due to the convenience of the 14 day interval with the 1200 mg dose. Complete complement blockade has been maintained in these patients for 8 mo. and 12 mo. to date, respectively. After 1 year of sustained complement blockade with the 1200 mg maintenance dose, patient 1 again demonstrated a breakthrough. Complement inhibition is now being maintained in this patient by a 1200 mg dose every 14 days with an additional 1200 mg dose in between the 14 day dosing interval every 4–5 doses. There were no reported adverse events in any of the four patients in which the 1200 mg dosing regimens were administered. In summary, these data demonstrate good correlation between eculizumab and LDH levels, suggesting that a breakthrough in complement activity due to insufficient drug levels can be monitored by levels of LDH near the end of the dosing interval. These results illustrate that two alternative-dosing regimens are well tolerated and can be effectively employed in the small percentage of PNH patients in which complement inhibition is not consistently maintained using the standard dose. Figure Figure

Description: Paroxysmal nocturnal hemoglobinuria (PNH) is characterized by intravascular hemolysis and venous thrombosis. Deficiency of the terminal complement inhibitor CD59 from PNH red cells results in complement-mediated hemolysis. Eculizumab, a humanized monoclonal antibody that inhibits terminal complement by binding to C5, is effective at controlling intravascular hemolysis in PNH. We now report that 10 of the 11 patients from an initial 3 month study have continued to receive 900mg eculizumab every other week for 2 years. The remaining patient stopped eculizumab after 23 months despite effective control of intravascular hemolysis, as the patient continued to be transfused even after erythropoietin therapy. This patient had the most severe hypoplasia at the start of eculizumab therapy with a platelet count below 30x109/l, suggesting that the ongoing transfusions were due to the bone marrow failure and not continuing hemolysis. Eculizumab was safe and well tolerated with two reported SAEs in the last year, neither of which were attributed to the drug. The dramatic improvement in various parameters of hemolysis persisted during the 2 year treatment period for all patients. Mean LDH levels decreased from 3111 +/− 598 U/L over the 12 months prior to treatment to 634 +/− 34 U/L up to 24 months following treatment (p=0.002). PNH red cells with a complete deficiency of GPI-linked proteins (Type III red cells) progressively increased during the treatment period from a mean of 36.7% to 58.9% (p=0.001) while partially deficient PNH red cells (Type II) increased from 5.3% to 8.7% (p=0.01). There has been no change in the proportion of PNH neutrophils in any of the patients during eculizumab therapy indicating that the increase in the proportion of PNH red cells is due to a reduction in hemolysis and transfusions rather than a change in the PNH clone(s) itself. The mean and median transfusion rates decreased from 2.1 and 1.8 units/patient/month to 0.4 and 0.3 units/patient/month respectively (p

Description: Analysis of immunoglobulin heavy chain (IgH) rearrangements in B-CLL differentiates subgroups of patients with significantly different clinical outcomes. Cases can be categorised according to mutational status of the variable (V) segment with unmutated VH regions linked to a worse prognosis. A restricted pattern of use of specific VH, DH and JH gene segments has also been reported in B-CLL. It has been hypothesised that B-CLL originates as a clonal expansion of B-cells that have been selected and activated by contact with self or foreign antigens, leading to those small clones to proliferate, mutate their IGH genes, acquire genetic lesions and eventually become malignant. B-CLL cells normally express low levels of Ig on the surface, normally IgM, although a proportion of patients express IgG or IgA, following the class-switch recombination (CSR) process. We have analysed the pattern of SHM and gene segment usage in this particular subgroup for 44 patients with IgG B-CLL. Successful PCR amplification of recombined Smu-Sgamma switch region DNA was achieved in 40 patients, confirming the presence of IgG class-switching. Mutational analysis of IgH V genes revealed 80% of patients contained more than 2% somatic hypermutation (SHM), with 63% of samples having a greater than 5% SHM rate. For VH gene segment usage, a significant predominance of the VH4 family was seen in 22 cases (50%), followed by VH3 in 17 cases (39%), while VH1 family was found in only 3 of 44 samples, this differs from classical IgM B-CLL where VH3 family usage predominates. Overall, VH4-34 was the most frequently used gene segment (34%), followed by VH3-07 (14%) and VH4-39 (9%). DH6-13 was the most frequently used DH segment (21%), followed by DH6-19 (13%). JH gene segment usage did not differ from normal B-cells, with JH4 being the most frequently used, followed by JH6 and JH5. There was a significant association between VH4-39, DH6-13 and JH5 in three samples all containing unmutated sequence. Together this data reveals a distinct pattern of IGH VDJ rearrangements in IgG B-CLL compared to classical IgM B-CLL. Firstly, the rate of SHM in IgG B-CLL (80%) is significantly higher than the 50% observed in IgM B-CLL. Secondly, VH segment usage pattern differs between the two subgroups with a significant under-representation of VH1 as well as an over-representation of VH4 family members in the IgG subgroup. Finally, there is a striking association between VH4-39 and DH6-13/JH5 in the very few unmutated rearrangements. This could be indicative of a different clonal history of these particular B cells in B-CLL. Together with recent published data, this latter finding suggests that this is a further sub-category exclusive to IgG B-CLL, where selection of a specific antigen receptor may have lead to B-CLL development in such cases. We conclude that class-switched IgG B-CLL contains different molecular features in the IgH genes compared with classical IgM B-CLL, and other B-cell malignancies. The clinical implications of these differences, especially the relationship between the mutational status of the VH genes and outcome in IgG B-CLL, will be further investigated.

Description: Monoclonal chronic lymphocytic leukemia (CLL)–phenotype cells are detectable in 3.5% of otherwise healthy persons using flow cytometric analysis of CD5/CD20/CD79b expression on CD19-gated B cells. To determine whether detection of such CLL-phenotype cells is indicative of an inherited predisposition, we examined 59 healthy, first-degree relatives of patients from 21 families with CLL. CLL-phenotype cells were detected in 8 of 59 (13.5%) relatives, representing a highly significant increase in risk (P = .00002). CLL-phenotype cell levels were stable with time and had the characteristics of indolent CLL. Indolent and aggressive clinical forms were found in family members, suggesting that initiation and proliferation involves distinct factors. The detection of CLL-phenotype cells provides a surrogate marker of carrier status, potentially facilitating gene identification through mapping in families and direct analysis of isolated CLL-phenotype cells.

Description: Paroxysmal nocturnal hemoglobinuria (PNH) is an acquired hemolytic anemia characterized by intravascular hemolysis, often resulting in the need for red blood cell (RBC) transfusions. PNH RBCs lack two complement regulatory molecules - CD59, a terminal complement inhibitor, and CD55, a C3 convertase inhibitor. Eculizumab, a humanized monoclonal antibody that inhibits terminal complement by binding to C5, effectively controls intravascular hemolysis as determined by a dramatic reduction in lactate dehydrogenase (LDH) to levels in or just above the normal range. Control of intravascular hemolysis in these patients led to a reduction in, or cessation of, RBC transfusions. During eculizumab treatment, a majority of patients demonstrate evidence of residual, low-level hemolysis; LDH levels remain slightly elevated, haptoglobin levels are low or undetectable, and bilirubin levels are above normal. We hypothesized that this low-level residual hemolysis may be due to clearance of PNH RBCs through a C3b-mediated mechanism. Therefore we investigated C3 deposition on RBC in PNH patients before and on eculizumab. A direct antiglobulin test (DAT) using monoclonal anti-C3d was positive in 29 out of 39 PNH patients on eculizumab. Of these 29 DAT-positive patients, who were all receiving transfusions, 25 had DAT testing prior to eculizumab therapy and only one of these was positive. DAT was negative in all of 8 normal volunteers. By two-color flow cytometric analysis with anti-CD59 and anti-C3, the majority of patients on eculizumab demonstrated three distinct RBC populations: CD59+/C3− (normal RBCs); CD59-/C3− (PNH RBCs without C3 coating); and CD59-/C3+ (PNH RBCs coated by C3). No CD59+/C3+ RBCs were observed. Of 21 DAT positive eculizumab treated patients tested, the median proportion of total RBCs that were C3b positive was 17.6%. 18 of 29 [62%] eculizumab patients with a positive DAT received at least one transfusion during eculizumab therapy compared with 1 of 10 [10%] for DAT negative patients (p=0.01), although even patients who did not become transfusion independent during eculizumab treatment showed a marked reduction in transfusion requirement. The median hemoglobin value for the 29 DAT positive eculizumab patients was 9.8 g/dL compared with 11.3 g/dL in the 10 DAT negative eculizumab patients (p= 0.08). No apparent relationship between LDH and DAT positivity was observed. It is proposed that resolution of intravascular hemolysis in PNH patients on eculizumab results in deposition of C3b on the surface of PNH RBCs which may explain, at least in part, the residual low level hemolysis occurring in some patients. This appears to be a previously undescribed mechanism of RBC clearance in PNH, most likely obscured by the rapidity of intravascular hemolysis in the absence of eculizumab therapy. Despite the low-level residual hemolysis, patients continue to receive significant benefit from eculizumab treatment.

Description: Peripheral blood B cells in patients with paroxysmal nocturnal hemoglobinuria (PNH) comprise variable mixtures of normal B cells produced before the onset of disease and glycosylphosphatidylinositol (GPI)-deficient B cells derived from the PNH hematopoietic stem cell. In a detailed phenotypic analysis of 29 patients with PNH, this study shows consistent phenotypic differences between PNH B cells and residual normal B cells. In the majority of patients with active disease, PNH B cells comprised mainly naive cells with a CD27−IgM+IgDstrong+IgG−phenotype. The proportion of CD27+ memory cells within this compartment was related to disease duration (Spearman [rs] 0.403; P = .030). In PNH patients with predominantly GPI-deficient hematopoiesis, that is, a large granulocyte PNH clone, the residual normal B cells had a predominantly memory (CD27+) phenotype. Furthermore, the majority of these memory B cells were not immunoglobulin (Ig) class switched and had an IgM+IgD+IgG− phenotype. Using PNH as a novel model with which to study B lymphopoiesis, this study provides direct evidence that production of new naive B cells occurs throughout life and that the major population of long-lived memory B cells are IgM+IgD+. Moreover, studies of GPI− B cells in 2 patients in remission from PNH suggest that the life span of a B-cell clone can be more than 24 years.

Description: Flow cytometric analysis of GPI-linked antigens has had a major impact on the diagnosis of PNH. Significant numbers of patients with aplastic anemia have small PNH clones, and due to the precision in clone size measurement, reliable serial monitoring can now be undertaken although the clinical value of this is not proven. From our series of 234 PNH patients, we analysed clinical correlates between disease type and red cell and granulocyte peripheral blood clone sizes as determined by flow cytometry at presentation. For hemolytic patients (n = 99) the mean PNH clone sizes were: granulocytes 84.8%; red cells 45.3% (type III cells 33.6%). For aplastic patients (no macroscopic hemolysis) the mean clone sizes were: granulocytes 18.5%; red cells 6.4% (type III cells 4.5%). The two groups were statistically different (Mann Whitney U; P90% granulocyte clones (n = 30; mean follow up 48 months) with virtually all their hematopoiesis maintained from PNH stem cells have clone sizes that remain stable and their clinical behavior suggests that their PNH can persist for up to 40 years. The second group of patients (n = 16) were those with hemolytic PNH with granulocyte clones of 90% granulocyte clone group (2/30) and developed as a terminal event, one with GPI-MDS, and the second with a rapid emergence of GPI+MDS. One patient in the aplastic group showed progression to AML (1/34). 27% of patients had an improvement in cytopenias with concurrent decrease in PNH clone size. For hemolytic patients with granulocyte clones of

Description: Pulmonary hypertension (PHT) is an emerging common complication of hereditary hemolytic anemias. It has been mechanistically and epidemiologically linked to intravascular hemolysis and decreased nitric oxide (NO) bioavailability. While this complication has been described in approximately 30% of adult patients with sickle cell disease and thalassemia, the prevalence of PHT in patients with paroxysmal nocturnal hemoglobinuria (PNH), an acquired disease with the highest levels of intravascular hemolysis observed, has never been determined. PNH patients frequently have symptoms consistent with both hemolysis and PHT including severe fatigue and dyspnea on exertion. Therefore, we examined for the presence of PHT in PNH and explored potential mechanisms associated with its development by measuring the ability of plasma to instantaneously consume NO using ozone-based chemiluminescence. Doppler echocardiography was performed in 24 hemolytic PNH patients to estimate pulmonary artery systolic pressures. Systolic PHT was defined by a tricuspid regurgitant jet velocity (TRV) ≥ 2.5m/s at rest. Eleven (46%) patients had elevated pulmonary artery systolic pressures (mean TRV 2.7m/s ± 0.08) and one (4%) had severely elevated pressures (TRV 3.5m/s). Plasma from PNH patients (n=28) consumed 32.26 ± 8.74μM NO while normal subjects (n=9) consumed 2.42 ± 0.77μM NO (p=0.03). LDH levels correlated with NO consumption (p