A longitudinal analysis of paroxysmal nocturnal haemoglobinuria-type cells in patients with bone marrow failure: Results of a prospective multi-centre study in Japan.

To determine the prevalence and clinical relevance of glycosylphosphatidylinositol-anchored protein-deficient (GPI[-]) cell populations (paroxysmal nocturnal haemoglobinuria [PNH]-type cells) in patients with acquired aplastic anaemia (AA) or myelodysplastic syndrome (MDS), we prospectively studied peripheral blood samples of 2402 patients (1075 with AA, 900 with MDS, 144 with PNH, and 283 with other anaemia) using a high-sensitivity flow cytometry assay in a nationwide multi-centre observational study. PNH-type cells were detected in 52.6% of AA and 13.7% of MDS patients. None of the 35 patients with refractory anaemia (RA) with ringed sideroblasts or the 86 patients with RA with excess of blasts carried PNH-type cells. Among the 317 patients possessing PNH-type granulocytes, the percentage of PNH-type granulocytes increased by ≥10% in 47 patients (14.8%), remained unchanged in 240 patients (75.7%), and decreased by ≥10% in 30 patients (9.5%) during 3 years of follow-up. PNH-type granulocyte expansion occurred more frequently (27.1%) in the 144 patients who originally carried PNH-type granulocytes ≥1% than in the 173 patients with PNH-type granulocytes <1% (4.6%). This study confirmed that PNH-type cells are undetectable in authentic clonal MDS patients, and the presence of ≥1% PNH-type granulocytes predicts a higher likelihood of PNH-type cell expansion than with <1% PNH-type granulocytes.

The clinical significance of PNH-phenotype cells accounting for < 0.01% of total granulocytes detected by the Clinical and Laboratory Standards Institute methods in patients with bone marrow failure.

Small populations of glycosylphosphatidylinositol-anchored protein-deficient (GPI[-]) cells accounting for up to 0.01% of total granulocytes can be accurately detected by a high-sensitivity flow cytometry (FCM) assay established by the Clinical and Laboratory Standards Institute (CLSI method) and have a prognostic value in bone marrow failure (BMF); however, the significance of GPI(-) granulocytes accounting for 0.001-0.009% of granulocytes remains unclear. To clarify this issue, we examined the peripheral blood of 21 BMF patients in whom minor (around 0.01%) populations of GPI(-) granulocytes had been previously detected by a different high-resolution FCM method (OPTIMA method, which defines ≥ 0.003% GPI(-) granulocytes as an abnormal increase) using both the CLSI and OPTIMA methods simultaneously. These two methods detected an "abnormal increase" in GPI(-) granulocytes in 10 patients (48%) and 17 patients (81%), respectively. CLSI detected 0.002-0.005% (median, 0.004%) GPI(-) granulocytes in 7 patients who were deemed positive for PNH-type cells according to the OPTIMA method, which detected 0.003-0.012% (median 0.006%) GPI(-) granulocytes. The clone sizes of GPI(-) cells detected by each assay were positively correlated (r = 0.994, p < 0.001). Of the seven patients who were judged positive for PNH-type cells by OPTIMA alone, five received immunosuppressive therapy, and all of them achieved a partial or complete response. GPI(-) granulocytes detected in BMF patients by the CLSI method should thus be considered significant, even at percentages of < 0.01%.

Low-dose lenalidomide and dexamethasone therapy after melphalan-prednisolone induction in elderly patients with newly diagnosed multiple myeloma.

Lenalidomide (Len) and dexamethasone (dex) therapy is a standard therapy in patients with multiple myeloma. Elderly or unfit patients may reduce Len or dex doses to prevent toxicities that lead to treatment discontinuation. However, there have been few studies evaluating the efficacy and safety of lower doses of Len and dex. We conducted a phase II study of 1.5-year low-dose Len and dex therapy following melphalan and prednisolone (MP), the number of which cycles was determined by a response within 9 cycles. The Len dose was 10 mg daily and the dex dose was 20 mg weekly, which were continued for 1.5 years. Twenty-one patients were enrolled. The median number of cycles of MP was 3 (range, 2-9). The overall response rate was 81% and a very good partial response or better was achieved in 33.3% of patients. The median follow-up time for survivors was 70.5 months (range, 42-83 months), the median progression-free survival (PFS) was 27 months (95% CI, 21-33 months), and the median overall survival was not reached. Grade 3 or 4 adverse events were observed in 28.6% of patients. In conclusion, the low-dose Len and dex therapy safely achieved comparable efficacies to the standard-dose regimen in elderly patients with newly diagnosed multiple myeloma. UMIN000007889.

Mechanisms of Impaired Neutrophil Migration by MicroRNAs in Myelodysplastic Syndromes.

In myelodysplastic syndromes (MDS), functional defects of neutrophils result in high mortality because of infections; however, the molecular basis remains unclear. We recently found that miR-34a and miR-155 were significantly increased in MDS neutrophils. To clarify the effects of the aberrant microRNA expression on neutrophil functions, we introduced miR-34a, miR-155, or control microRNA into neutrophil-like differentiated HL60 cells. Ectopically introduced miR-34a and miR-155 significantly attenuated migration toward chemoattractants fMLF and IL-8, but enhanced degranulation. To clarify the mechanisms for inhibition of migration, we studied the effects of miR-34a and miR-155 on the migration-regulating Rho family members, Cdc42 and Rac1. The introduced miR-34a and miR-155 decreased the fMLF-induced active form of Cdc42 to 29.0 ± 15.9 and 39.7 ± 4.8% of that in the control cells, respectively, although Cdc42 protein levels were not altered. miR-34a decreased a Cdc42-specific guanine nucleotide exchange factor (GEF), dedicator of cytokinesis (DOCK) 8, whereas miR-155 reduced another Cdc42-specific GEF, FYVE, RhoGEF, and PH domain-containing (FGD) 4. The knockdown of DOCK8 and FGD4 by small interfering RNA suppressed Cdc42 activation and fMLF/IL-8-induced migration. miR-155, but not miR-34a, decreased Rac1 protein, and introduction of Rac1 small interfering RNA attenuated Rac1 activation and migration. Neutrophils from patients showed significant attenuation in migration compared with healthy cells, and protein levels of DOCK8, FGD4, and Rac1 were well correlated with migration toward fMLF (r = 0.642, 0.686, and 0.436, respectively) and IL-8 (r = 0.778, 0.659, and 0.606, respectively). Our results indicated that reduction of DOCK8, FGD4, and Rac1 contributes to impaired neutrophil migration in MDS.

Establishment of a flow cytometry assay for detecting paroxysmal nocturnal hemoglobinuria-type cells specific to patients with bone marrow failure.

Minor populations of glycosylphosphatidylinositol-anchored protein-deficient (GPI[-]) cells in the peripheral blood may have a prognostic value in bone marrow failure (BMF). Our objective is to establish the optimal flow cytometry (FCM) assay that can discriminate GPI(-) populations specific to BMF from those of healthy individuals. To identify a cut-off that discriminates GPI(-) rare cells from GPI(+) cells, we determined a position of the borderline that separates the GPI(-) from GPI(+) cells on a scattergram by testing more than 30 healthy individuals, such that no GPI(-) dot fell into the upper left quadrant where fluorescein-labeled aerolysin (FLAER)-CD11b+ granulocytes and CD55-CD59- glycophorin A+ erythrocytes were positioned. This method allowed us to define ≥ 0.003% CD11b+FLAER- granulocytes and ≥ 0.005% glycophorin A+CD55-CD59- erythrocytes to be specific to BMF patients. Longitudinal cross-validation studies showed minimal (< 0.02%) inter-laboratory differences in the GPI(-) cell percentage. An analysis of 1210 patients with BMF revealed a GPI(-) cell population in 56.3% of patients with aplastic anemia and 18.5% of patients with myelodysplastic syndrome. The GPI(-) granulocyte percentages was 0.003-0.01% in 3.7% of patients. This FCM assay effectively identified an increase in the percentage of GPI(-) rare cells that are specific to BMF patients and allowed different laboratories to accurately detect 0.003-0.01% of pathological GPI(-) cells.

Genetic variants in C5 and poor response to eculizumab.

BACKGROUND: Eculizumab is a humanized monoclonal antibody that targets complement protein C5 and inhibits terminal complement-mediated hemolysis associated with paroxysmal nocturnal hemoglobinuria (PNH). The molecular basis for the poor response to eculizumab in a small population of Japanese patients is unclear. METHODS: We assessed the sequences of the gene encoding C5 in patients with PNH who had either a good or poor response to eculizumab. We also evaluated the functional properties of C5 as it was encoded in these patients. RESULTS: Of 345 Japanese patients with PNH who received eculizumab, 11 patients had a poor response. All 11 had a single missense C5 heterozygous mutation, c.2654G → A, which predicts the polymorphism p.Arg885His. The prevalence of this mutation among the patients with PNH (3.2%) was similar to that among healthy Japanese persons (3.5%). This polymorphism was also identified in a Han Chinese population. A patient in Argentina of Asian ancestry who had a poor response had a very similar mutation, c.2653C → T, which predicts p.Arg885Cys. Nonmutant and mutant C5 both caused hemolysis in vitro, but only nonmutant C5 bound to and was blocked by eculizumab. In vitro hemolysis due to nonmutant and mutant C5 was completely blocked with the use of N19-8, a monoclonal antibody that binds to a different site on C5 than does eculizumab. CONCLUSIONS: The functional capacity of C5 variants with mutations at Arg885, together with their failure to undergo blockade by eculizumab, account for the poor response to this agent in patients who carry these mutations. (Funded by Alexion Pharmaceuticals and the Ministry of Health, Labor, and Welfare of Japan.).

High prevalence of diffuse large B-cell lymphoma in occult hepatitis B virus-infected patients in the Tohoku district in Eastern Japan.

Occult hepatitis B virus (HBV) infection is a clinical challenge, but its relationship to clinicopathologic features and the risk of progression to malignant lymphoma (ML) are poorly defined. We estimated the prevalence of HBV infection of 1,358 patients with newly diagnosed ML. HBV infection was more prevalent in ML than in control patients. The occult HBV infection group had a higher median onset age, no liver or spleen involvement, and higher prevalence of diffuse large B-cell lymphoma than the other groups, indicating that occult HBV infection is a distinct clinicopathologic entity. J. Med. Virol. 88:2206-2210, 2016. © 2016 Wiley Periodicals, Inc.

Dysregulation of the MIRLET7/HMGA2 axis with methylation of the CDKN2A promoter in myeloproliferative neoplasms.

Overexpression of high mobility group AT-hook 2 (Hmga2), which is negatively regulated by MIRLET7 micro RNAs through 3'-untranslated region (3'UTR), causes proliferative haematopoiesis mimicking myeloproliferative neoplasms (MPNs) and contributes to progression of myelofibrosis in mice. Thus, we investigated HMGA2 mRNA expression in 66 patients with MPNs including 23 polycythaemia vera (PV), 33 essential thrombocythaemia (ET) and 10 primary myelofibrosis (PMF). HMGA2 mRNA expression, especially variant 1 with 3'UTR that contains MIRLET7-specific sites, rather than variant 2 lacking 3'UTR, is frequently deregulated due to decreased MIRLET7 expression in granulocytes from over 20% of PV and ET, and in either granulocytes or CD34(+) cells from 100% of PMF. Patients with deregulated HMGA2 mRNA expression were significantly more likely to show splenomegaly, high serum lactate dehydrogenase values, and methylation of the CDKN2A promoter compared with other patients without deregulation of HMGA2. A histone deacetylase inhibitor, panobinostat, significantly increased MIRLET7 expression and reduced variant 1 of HMGA2 mRNA expression, but not variant 2, in both U937 cells and PMF-derived CD34(+) cells. Moreover, both panobinostat and small interfering RNA of HMGA2 demethylated the CDKN2A promoter in U937 cells. In conclusion, the frequently dysregulated MIRLET7/HMGA2 axis could be a therapeutic target in MPNs.

A prospective analysis of clinical efficacy and safety in chronic myeloid leukemia-chronic phase patients with imatinib resistance or intolerance as evaluated using European LeukemiaNet 2013 criteria.

BACKGROUND: We conducted a phase II study to evaluate the efficacy and safety of dasatinib in Japanese patients with imatinib-resistant or imatinib-intolerant chronic myeloid leukemia (CML). METHODS: From 2009 to 2011, 54 CML-chronic phase (CP) patients with resistance (n = 40) or intolerance (n = 25) to imatinib were registered to undergo dasatinib treatment. Eleven patients showed both resistance and intolerance to imatinib. Coincidentally, the resistance criteria in this study were the same as a non-optimal response to tyrosine kinase inhibitors (TKIs) as defined in the European LeukemiaNet (ELN) 2013 recommendations. RESULTS: The overall incidence rate of major molecular response (MMR) at 12 months was 62.3% (n = 47). Forty patients with resistance to imatinib who were 'warning' and 'failure' patients based on the ELN 2013 recommendations were assessed; cumulative MMR and MR(4.5) rates were 62.5% (n = 39) and 21.0% (n = 40), respectively, at 12 months. Twelve patients who showed a BCR-ABL transcript level >1% on the international scale did not achieve a MMR or discontinued dasatinib treatment because of insufficient effects. With regard to safety issues, grade 3/4 non-hematologic adverse events (AEs) were infrequent. CONCLUSIONS: Patients with non-optimal responses (who meet ELN 2013 warning and failure criteria) to imatinib should be switched quickly to dasatinib, which is less toxic in CML-CP patients, to improve their prognoses. A BCR-ABL1 IS of <1% at 3 months of dasatinib administration is a landmark for good therapeutic outcome.

Deregulated expression of HMGA2 is implicated in clonal expansion of PIGA deficient cells in paroxysmal nocturnal haemoglobinuria.

Patients with paroxysmal nocturnal haemoglobinuria (PNH) have expanded clonal cells bearing a somatic mutation in the PIGA gene. Our previous study on two PNH patients with chromosome 12 rearrangements demonstrated the involvement of HMGA2 expression in clonal expansion. The present study investigated HMGA2 expression in PNH patients without chromosomal abnormalities. The expression of short HMGA2 with latent exon was significantly high in peripheral blood cells from 18 of 24 patients. Over-expression of truncated HMGA2 in mouse bone marrow cells caused expansion in recipient mice. These results support the idea that deregulated expression of HMGA2 causes expansion of PNH cells.

Long-term follow-up of patients with paroxysmal nocturnal hemoglobinuria treated with eculizumab: post-marketing surveillance in Japan.

All Japanese patients with paroxysmal nocturnal hemoglobinuria (PNH) treated with eculizumab were enrolled in post-marketing surveillance (PMS) between June 2010 and August 2019 to assess the long-term effectiveness and safety of eculizumab. The reduction in intravascular hemolysis, the change in hemoglobin (Hb) level, and the change in renal function were assessed to determine the effectiveness of eculizumab. The types and frequencies of adverse events (AEs) were assessed to determine its safety. A total of 632 patients were enrolled and the median treatment duration was 3.6 years. Treatment with eculizumab significantly reduced lactate dehydrogenase (LDH) levels and significantly increased Hb levels. These changes were maintained for up to 5 years of treatment. An estimated glomerular filtration rate ≥ 60 ml/min/1.73 m2 and higher LDH level at baseline were associated with increases in Hb levels during eculizumab treatment. The overall incidence of any AE was 69.92/100 patient-years. Hemolysis was the most common AE (6.43/100 patient-years). The incidence of infection-related AEs was 20.57/100 patient-years, and included meningococcal infection in three patients (0.12/100 patient-years). This long-term follow-up of patients with PNH demonstrated the sustained effectiveness of eculizumab and supports its well-established safety profile.

Assessment of dysplasia in bone marrow smear with convolutional neural network.

In this study, we developed the world's first artificial intelligence (AI) system that assesses the dysplasia of blood cells on bone marrow smears and presents the result of AI prediction for one of the most representative dysplasia-decreased granules (DG). We photographed field images from the bone marrow smears from patients with myelodysplastic syndrome (MDS) or non-MDS diseases and cropped each cell using an originally developed cell detector. Two morphologists labelled each cell. The degree of dysplasia was evaluated on a four-point scale: 0-3 (e.g., neutrophil with severely decreased granules were labelled DG3). We then constructed the classifier from the dataset of labelled images. The detector and classifier were based on a deep neural network pre-trained with natural images. We obtained 1797 labelled images, and the morphologists determined 134 DGs (DG1: 46, DG2: 77, DG3: 11). Subsequently, we performed a five-fold cross-validation to evaluate the performance of the classifier. For DG1-3 labelled by morphologists, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were 91.0%, 97.7%, 76.3%, 99.3%, and 97.2%, respectively. When DG1 was excluded in the process, the sensitivity, specificity, PPV, NPV, and accuracy were 85.2%, 98.9%, 80.6%, and 99.2% and 98.2%, respectively.

Results from multinational phase 3 studies of ravulizumab (ALXN1210) versus eculizumab in adults with paroxysmal nocturnal hemoglobinuria: subgroup analysis of Japanese patients.

Ravulizumab demonstrated noninferior efficacy and comparable safety to eculizumab in two open-label, phase 3 studies in patients with paroxysmal nocturnal hemoglobinuria (PNH) who complement inhibitor-naive (Study 301) or were previously treated with eculizumab (Study 302). This subgroup analysis assessed ravulizumab's efficacy and safety in Japanese patients in Studies 301 and 302, who are known to have different clinicopathologic features from white patients. Patients were randomly assigned (1:1) to eculizumab every-two-weeks or weight-based dosing of ravulizumab every-eight-weeks for 26 weeks. Co-primary endpoints were transfusion avoidance and lactate dehydrogenase (LDH) normalization in Study 301 and percentage change in LDH levels from baseline to day 183 in Study 302. Thirty-three Japanese patients (n = 18 ravulizumab; n = 15 eculizumab) enrolled in Study 301; 12 enrolled in Study 302 (n = 5 ravulizumab; n = 7 eculizumab). In the Study 301 ravulizumab group, 83.3% (15/18) of patients avoided transfusion; the adjusted prevalence of LDH normalization was 52.1%. In the Study 302 ravulizumab group, the least-squares-mean percentage change from baseline in LDH was 8.34%. No deaths or meningococcal infections occurred during the 6-month primary evaluation period in either study. In conclusion, ravulizumab's efficacy and safety were consistent in the Japanese and global patient populations with PNH in the phase 3 studies. Clinical Trial Identifier: NCT02946463; NCT03056040.

Investigation of the cumulative number of chromosome aberrations induced by three consecutive CT examinations in eight patients.

In our previous study, we found that chromosomes were damaged by the radiation exposure from a single computed tomography (CT) examination, based on an increased number of dicentric chromosomes (Dics) formed in peripheral blood lymphocytes after a CT examination. We then investigated whether a cumulative increase in the frequency of Dics and chromosome translocations (Trs) formation could be observed during three consecutive CT examinations performed over the course of 3-4 years, using lymphocytes in peripheral bloods of eight patients (five males and three females; age range 27-77 years; mean age, 64 years). The effective radiation dose per CT examination estimated from the computational dosimetry system was 22.0-73.5 mSv, and the average dose per case was 40.5 mSv. The frequency of Dics formation significantly increased after a CT examination and tended to decrease before the next examination. Unlike Dics analysis, we found no significant increase in the frequency of Trs formation before and after the CT examination, and we observed no tendency for the frequency to decrease before the next CT examination. The frequency of Trs formation was higher than that of Dics formation regardless of CT examination. Furthermore, neither analysis of Dics nor Trs showed a cumulative increase in the frequency of formation following three consecutive CT examinations.

The role of Wilms' tumor gene peptide-specific cytotoxic T lymphocytes in immunologic selection of a paroxysmal nocturnal hemoglobinuria clone.

OBJECTIVE: To clarify an expansion mechanism of a paroxysmal nocturnal hemoglobinuria (PNH) clone with the Wilms' tumor gene (WT1). MATERIALS AND METHODS: In PNH patients with the HLA-A*2402 allele, frequencies of peripheral blood (PB) WT1 peptide-specific and HLA-A*2402-restricted CD8+ cells and WT1 peptide-stimulated interferon-gamma-producing mononuclear cells (MNCs), cytotoxicity of WT1 peptide-specific and HLA-A*2402-restricted cytotoxic T lymphocyte (CTL) clone (TAK-1) cells on bone marrow (BM) MNCs, and after co-incubation with TAK-1 cells, changes in colony-forming unit granulocyte-macrophage colony formation of CD34+ cells and in CD59 expression in viable CD34+ cells were investigated. RESULTS: The frequencies of PB WT1 peptide-specific and HLA-A*2402-restricted CD8+ cells (p < 0.005) and WT1 peptide-stimulated interferon-gamma-producing MNCs (p < 0.02) were significantly higher in 5 PNH patients than 8 healthy volunteers (HV). In 5 PNH patients or 3 HV, TAK-1 cells significantly killed BMMNCs and suppressed colony formations of CD34+CD59+ and/or CD34+CD59- cells in the absence and presence of a WT1 peptide or only in the presence of the peptide, respectively, in an HLA-restricted manner. After co-incubation with TAK-1 cells, reduction rates of colony formation of CD34+CD59- cells were significantly less than those of CD34+CD59+ cells in 5 PNH patients (p < 0.002) and proportions of viable CD34+CD59- cells from 5 PNH patients significantly increased in the absence (p < 0.01) and presence (p < 0.01) of a WT1 peptide in an HLA-restricted manner. CONCLUSION: WT1 peptide-specific and HLA-restricted CTLs may play an important role in expansion of a PNH clone during immunologic selection and/or in the occurrence of BM failure via interferon-gamma in PNH.

Safety and efficacy of high-dose ranimustine (MCNU) containing regimen followed by autologous stem cell transplantation for diffuse large B-cell lymphoma.

High-dose chemotherapy followed by autologous stem cell transplantation (ASCT) is widely used as a salvage therapy for relapsed or high-risk diffuse large B-cell lymphoma (DLBCL). To investigate the safety and efficacy of regimens including high-dose MCNU followed by ASCT for DLBCL, we analyzed the data from prospective multicenter trials. Twenty-nine patients were analyzed, and the median follow-up time for survival patients was 70 months. Fifteen patients received MCVC conditioning regimen, and fourteen patients received MEAM regimen. Major toxicities associated with these conditioning regimens included nausea (69%), anorexia (66%), febrile neutropenia (62%), diarrhea (59%), and mucositis (34%). One patient who developed severe sinusoidal obstructive syndrome and acute lung injury died without disease progression, and overall therapy-related mortality at 5 years was 3%. No patient developed therapy-related hematological malignancy. At 5 years, overall survival and progression-free survival in all patients were 82.8 and 58.2%, respectively. The 5-year OS in patients treated by the MCVC and MEAM regimens were 73.3 and 92.9%, respectively. These results suggest that regimens including high-dose MCNU followed by ASCT are feasible and effective for the treatment of relapsed or high-risk DLBCL. Further investigation is needed to evaluate of these regimens.

Correction to: Effects of eculizumab treatment on quality of life in patients with paroxysmal nocturnal hemoglobinuria in Japan.

In the original publication of this article, Tables 2, 3 and 4 were published incorrectly. The corrected tables 2, 3 and 4 are given in the following pages.

Effects of eculizumab treatment on quality of life in patients with paroxysmal nocturnal hemoglobinuria in Japan.

In paroxysmal nocturnal hemoglobinuria (PNH), various symptoms due to intravascular hemolysis exert a negative impact on patients' quality of life (QOL). To determine clinical factors related with improvements in QOL in PNH patients treated, we analyzed changes in QOL scales in PNH patients treated with eculizumab based on data collected from post-marketing surveillance in Japan. Summary statistics were obtained using figures from QOL scoring systems and laboratory values, and evaluated by t test. One-year administration of eculizumab improved the most QOL items in comparison with the baseline. In particular, significant improvement of EORTC QLQ-C30 was observed in fatigue, dyspnea, physical function, and global health status. Canonical correlation analysis revealed a high correlation between QOL and laboratory values. Changes in serum lactate dehydrogenase (LDH) and hemoglobin showed strong correlations with QOL improvement. Quality of life improvement was independent of patients' baseline characteristics of co-occurrence of bone marrow failure (BMF), or the degree of LDH. In this analysis, we found that the degree of QOL improvement was independent of the baseline LDH before eculizumab treatment and of co-occurrence of BMF. Paroxysmal nocturnal hemoglobinuria patients who have not received eculizumab treatment due to mild hemolysis may benefit from eculizumab treatment.

Microvascular thrombosis in the hepatic vein of a patient with paroxysmal nocturnal hemoglobinuria.

Paroxysmal nocturnal hemoglobinuria (PNH) is characterized by complement-mediated hemolysis, venous thrombosis, and bone marrow failure. In May 2003, a 33-year-old man was admitted to a hospital with right hypochondralgia and fever. He had a history of aplastic anemia. The patient's diagnosis of diffuse microvessel thrombosis in the hepatic vein due to an unknown cause was derived from the findings of a contrast-enhanced computed tomography examination of the abdominal region, angiographic evaluation of abdominal vessels, and pathohistologic examination of a liver biopsy sample. The patient was subsequently treated with warfarin. The abdominal pain and fever continued, however, and anemia gradually appeared. In April 2004, the patient was referred to our hospital to examine the cause of the thrombosis. On admission, slight anemia and a low serum haptoglobin level were observed. A flow cytometry evaluation of CD55 and/or CD59, CD59, and CD48 expression in erythrocytes, granulocytes, and monocytes, respectively, showed that the respective proportions of negative populations were 5.6%, 97.1%, and 96.2%. The patient then received a diagnosis of aplastic anemia/PNH syndrome, which had caused the hemolytic anemia and thrombosis, although no hemoglobinuria had been observed during his clinical course. This patient is, to our knowledge, the first reported case of a PNH patient with thrombosis present only in hepatic microvessels and not in hepatic large vessels, in spite of the presence of few hemolytic events.