The clinical relevance of minor paroxysmal nocturnal hemoglobinuria clones in refractory cytopenia of childhood: a prospective study by EWOG-MDS

Myelodysplastic syndromes (MDS) are characterized by ineffective hematopoiesis, resulting in peripheral blood cytopenias, myeloid dysplasia and risk of leukemic transformation. MDS in childhood are rare with an estimated annual incidence of 0.8–1.8 per million children aged 0–14 years.1 Refractory cytopenia of childhood (RCC), defined as myelodysplasia without an increased blast count, is the most common subtype of childhood MDS. Karyotype is normal in the majority of patients with RCC, and, in contrast to adults with MDS-refractory anemia (RA), about 80% of children have a hypocellular bone marrow (BM).1 Although the exact pathophysiology of MDS remains unclear, immunosuppressive therapy (IST), consisting of antithymocytic globulin (ATG) and/or cyclosporine A (CsA), is effective in some adults with MDS, suggesting a T-cell mediated immune response directed against hematopoietic progenitor cells in a proportion of the patients.2 ATG and CsA might also be effective in RCC,3 but predictors of response to IST in RCC are yet unknown. In some,4, 5, 6 but not all,2 studies conducted in adult MDS patients, a predictor of good response to IST is the presence of a minor paroxysmal nocturnal hemoglobinuria (PNH) clone.

Minor PNH or glycosylphosphatidylinositol (GPI)-deficient clones are present in 20–57% of adult and pediatric patients with the immune-mediated BM failure syndrome aplastic anemia4, 7, 8, 9 and in 13–23% of adults with low-grade MDS.4, 6, 8, 10 The mechanisms by which PNH clones arise in immune-mediated BM failure syndromes are incompletely understood, but it is hypothesized that GPI-deficient cells have a conditional growth advantage by evading an immune attack directed against normal hematopoietic progenitor cells. Another recent hypothesis is that GPI-deficient cells, which can be detected at very low numbers in healthy individuals, can expand without a conditional advantage, especially in conditions with reduced stem cell numbers.11

No data are available on the frequency and clinical correlates of PNH clones in RCC. In a prospective multicenter study, we assessed the frequency of PNH clones in RCC with high-sensitivity flow cytometry, and correlated the presence of PNH clones with clinical characteristics and response to IST.

Peripheral blood samples for PNH analysis were obtained from 87 consecutive, treatment-naive primary RCC patients 18 years of age (Table 1). Patients were diagnosed according to the 2008 WHO criteria for pediatric MDS between June 2005 and December 2011, enrolled in the prospective, multi-center studies EWOG-MDS 2006 and EWOG-MDS RC06 ( NCT00662090 and NCT00499070) and treated, as in Supplementary Figure S1. GPI-deficient granulocytes and erythrocytes were defined as CD24FLAER and CD55CD59 cells, respectively. On the basis of healthy controls, patients were defined as PNH-positive when a GPI-deficient population >0.01% in erythrocytes and/or a population >0.03% in granulocytes were present (Supplementary Methods).

Table 1 Clinical and laboratory characteristics of PNH-negative and PNH-positive patients

PNH clones in erythrocytes and/or granulocytes were present in 36 (41%) of 87 RCC patients , which is higher than in adults with low-grade MDS 4, 6, 8, 10 but similar to studies in aplastic anemia in adults and children.4, 7, 8, 9 As PNH clones might be indicative of immune-mediated BM failure, this increased frequency might point towards a larger proportion of RCC caused by immune-mediated mechanisms than low-grade MDS in adults. Alternatively, PNH clones may more readily emerge by neutral evolution in conditions with reduced stem cell numbers,11 which might be reflected by BM hypocellularity, occurring more frequently in RCC than in low-grade MDS in adults.1

PNH clone size in our study ranged from 0.011–58% (median: 0.057%) in erythrocytes and from 0.03% to 86% (median: 0.93%) in granulocytes. Clinical symptoms of PNH (hemolysis) were present in one RCC patient (who consequently also fulfilled criteria for the PNH working classification subcategory of PNH in the setting of another BM failure syndrome12). PNH clone sizes in this patient, who was 16 years of age at diagnosis and who received a HSCT 3 months post diagnosis, were 58% in erythrocytes and 86% in granulocytes.

Of the 53 patients in whom the presence of GPI-deficient cells could be assessed with sufficient sensitivity in both erythrocytes and granulocytes, 23 were PNH-positive (43%). Of these 23 PNH-positive patients, 20 were PNH-positive in both cell types. The remaining three patients were positive in erythrocytes only (clone size: 0.012–0.026%) (Supplementary Figure S2). In patients with a detectable PNH phenotype in both lineages, clones were significantly larger (median difference: 0.76%; range: -0.15–84%) in granulocytes than in erythrocytes (Mann–Whitney-U test, P=0.0004).

The 36 PNH-positive RCC patients were significantly older than the 51 PNH-negative patients (median: 12.7 vs 7.5 years, P=0.005); male/female distribution was equal. PNH-positive patients were more often HLA-DR15-positive than PNH-negative patients (42% vs 21%, P=0.051) (Table 1), as described previously in adult MDS and AA patients.5, 6 This association was stronger in RCC patients with a PNH clone >0.1% in erythrocytes and/or granulocytes compared with patients without PNH clone or a clone 0.1% (67% vs 18% HLA-DR15-positive, P=0.000). All five patients with monosomy 7 in the study were PNH-negative (P=0.076), which is in line with a previously described low frequency of karyotypic abnormalities in PNH-positive MDS-RA patients.6 BM cellularity was similar in PNH-positive and -negative patients. At diagnosis, PNH-positive patients had lower leukocyte counts, lower hemoglobin levels, lower platelet counts (as in adult MDS-RA6), and tended to have a higher MCV than PNH-negative patients. Transfusion dependency for platelets and/or erythrocytes did not differ significantly between the two groups (Table 1).

Twenty-eight hypocellular RCC patients (14 PNH-positive, 14 PNH-negative) were treated with IST consisting of ATG and CsA (Supplementary Methods) and evaluated for response at day 180 after start of therapy. Two additional patients received IST but had a normo- or hypercellular BM. Therefore, they were not eligible for IST according to the protocol and were excluded from the analysis. Of the 28 hypocellular RCC patients treated with IST, one patient, PNH-negative, received a HSCT before day 180, and was regarded as a non-responder at day 180. At day 180, a partial response was reached by 10 of 14 (71%) PNH-positive, and 5 of 14 (36%) PNH-negative patients (P=0.058). When only PNH clones >0.1% in erythrocytes and/or granulocytes were considered, PNH-positive patients were more likely to respond to IST than patients with a PNH clone 0.1% or no PNH clone: 7 of 8 (88%) PNH-positive patients responded at day 180, as compared with 8 of 20 (40%) PNH-negative patients (P=0.038). Event-free survival (EFS) rates in the 28 hypocellular PNH-positive and -negative RCC patients were 63% (s.e.=14%) and 36% (s.e.=13%), respectively (log-rank P=0.085) at 2.5 years; EFS rates in patients with a PNH clone >0.1% and in patients with a PNH clone 0.1% or no PNH clone were 70% (s.e.=18%) and 40% (s.e.=11%), respectively (log-rank P=0.089) at 2.5 years. Median follow-up time of these patients was 12 months (range: 3–48 months). Of note, none of the patients reached a complete response by day 180. This might be explained in part by the use of rabbit-ATG instead of horse-ATG since 2007, resulting in inferior response compared with horse-ATG in aplastic anemia.13, 14 Two of 28 patients, both PNH-negative, received horse-ATG; in 2 patients, information on the type of ATG is unavailable; in the remaining 24 patients, rabbit-ATG (Thymoglobulin, n=23; ATG-Fresenius, n=1) was used. To adequately change or maintain current treatment recommendations3 for hypocellular RCC patients, with or without PNH clone, further evaluation of long-term outcome after IST and evaluation of factors other than PNH that might influence IST response are underway in a larger RCC cohort.

Six RCC patients were followed by a watch-and-wait strategy and studied serially for the presence of PNH clones. Median follow-up time (time between first and last PNH measurement) of these patients was 260 days (range: 103–1159 days). Two patients were PNH-negative at first measurement and remained PNH-negative during follow-up. Four patients were PNH-positive at first measurement and remained PNH-positive during follow-up. Changes in clone size remained within one log-decade in both granulocytes and erythrocytes, and thus clone size was relatively stable (Figures 1a and b).

Figure 1

Evolution of GPI-deficient erythrocytes and granulocytes in patients followed sequentially. (a) Evolution of GPI-deficient erythrocytes in patients followed by watchful waiting (n=6). (b) Evolution of GPI-deficient granulocytes in patients followed by watchful waiting (n=5). Patient SC130 was severely neutropenic, thus limiting the sensitivity of PNH analysis in the granulocytic lineage. (c) Evolution of GPI-deficient erythrocytes in IST-treated patients (n=18). (d) Evolution of GPI-deficient granulocytes in IST treated patients (n=12). Median follow-up time from start of IST to last PNH measurement was 243 days (range: 60–725 days). Dotted lines indicate start of treatment; dashed lines indicate cutoff levels for PNH positivity (0.01% for erythrocytes, 0.03% for granulocytes). Six patients were studied before start of IST in the erythroid lineage only due to a limited number of cells in the granulocytic lineage. IST responders are indicated by circles; IST non-responders by squares. Patients without GPI-deficient cells were given an arbitrary value below 0.01% (a and c) and 0.03% (b and d).

Of the 30 RCC patients (either with a hypocellular (n=28) or normo- or hypercellular (n=2) BM) who were treated with IST, 18 were studied serially for the presence of PNH clones. Median follow-up time (time between first and last PNH measurement) of these patients was 255 days (range: 66–734 days). In PNH-positive IST responders, clone size remained relatively stable during follow-up, with changes in size within one log-decade. In one patient who was PNH-negative at diagnosis and IST non-responder, a PNH clone >0.1% (0.12%) emerged in the erythrocytes during follow-up; clone size in the granulocytes remained between 0.03 and 0.1%. No other significant increases in PNH clone size were noted (Figures 1c and d). That PNH clone size remained relatively stable in IST-responding patients seems at odds with the immune escape hypothesis, stating that GPI-deficient clones arise in conditions of T-cell mediated marrow destruction but has also been observed in adults with MDS and aplastic anemia.8, 15

In summary, we show that PNH clones are frequently present in RCC. PNH-positive RCC patients are older, have lower blood counts and are more frequently HLA-DR15-positive than PNH-negative patients. Although patient numbers were small, the presence of a PNH clone, especially when > 0.1%, might be associated with a greater likelihood of response to IST.


  1. 1

    Niemeyer CM, Baumann I . Classification of childhood aplastic anemia and myelodysplastic syndrome. Hematology Am Soc Hematol Educ Program 2011; 2011: 84–89.

  2. 2

    Sloand EM, Wu CO, Greenberg P, Young N, Barrett J . Factors affecting response and survival in patients with myelodysplasia treated with immunosuppressive therapy. J Clin Oncol 2008; 26: 2505–2511.

  3. 3

    Yoshimi A, Baumann I, Fuhrer M, Bergstrasser E, Gobel U, Sykora KW et al. Immunosuppressive therapy with anti-thymocyte globulin and cyclosporine A in selected children with hypoplastic refractory cytopenia. Haematologica 2007; 92: 397–400.

  4. 4

    Dunn DE, Tanawattanacharoen P, Boccuni P, Nagakura S, Green SW, Kirby MR et al. Paroxysmal nocturnal hemoglobinuria cells in patients with bone marrow failure syndromes. Ann Intern Med 1999; 131: 401–408.

  5. 5

    Maciejewski JP, Follmann D, Nakamura R, Saunthararajah Y, Rivera CE, Simonis T et al. Increased frequency of HLA-DR2 in patients with paroxysmal nocturnal hemoglobinuria and the PNH/aplastic anemia syndrome. Blood 2001; 98: 3513–3519.

  6. 6

    Wang H, Chuhjo T, Yasue S, Omine M, Nakao S . Clinical significance of a minor population of paroxysmal nocturnal hemoglobinuria-type cells in bone marrow failure syndrome. Blood 2002; 100: 3897–3902.

  7. 7

    Yoshida N, Yagasaki H, Takahashi Y, Yamamoto T, Liang J, Wang Y et al. Clinical impact of HLA-DR15, a minor population of paroxysmal nocturnal haemoglobinuria-type cells, and an aplastic anaemia-associated autoantibody in children with acquired aplastic anaemia. Br J Haematol 2008; 142: 427–435.

  8. 8

    Sugimori C, Mochizuki K, Qi Z, Sugimori N, Ishiyama K, Kondo Y et al. Origin and fate of blood cells deficient in glycosylphosphatidylinositol-anchored protein among patients with bone marrow failure. Br J Haematol 2009; 147: 102–112.

  9. 9

    Timeus F, Crescenzio N, Lorenzati A, Doria A, Foglia L, Pagliano S et al. Paroxysmal nocturnal haemoglobinuria clones in children with acquired aplastic anaemia: a prospective single centre study. Br J Haematol 2010; 150: 483–485.

  10. 10

    Iwanaga M, Furukawa K, Amenomori T, Mori H, Nakamura H, Fuchigami K et al. Paroxysmal nocturnal haemoglobinuria clones in patients with myelodysplastic syndromes. Br J Haematol 1998; 102: 465–474.

  11. 11

    Dingli D, Luzzatto L, Pacheco JM . Neutral evolution in paroxysmal nocturnal hemoglobinuria. Proc Natl Acad Sci USA 2008; 105: 18496–18500.

  12. 12

    Parker C, Omine M, Richards S, Nishimura J, Bessler M, Ware R et al. Diagnosis and management of paroxysmal nocturnal hemoglobinuria. Blood 2005; 106: 3699–3709.

  13. 13

    Scheinberg P, Nunez O, Weinstein B, Biancotto A, Wu CO, Young NS . Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl J Med 2011; 365: 430–438.

  14. 14

    Yoshimi A, Niemeyer CM, Fuhrer MM, Strahm B . Comparison of the efficacy of rabbit and horse antithymocyte globulin for the treatment of severe aplastic anemia in children. Blood 2013; 121: 860–861.

  15. 15

    Araten DJ, Bessler M, McKenzie S, Castro-Malaspina H, Childs BH, Boulad F et al. Dynamics of hematopoiesis in paroxysmal nocturnal hemoglobinuria (PNH): no evidence for intrinsic growth advantage of PNH clones. Leukemia 2002; 16: 2243–2248.

  16. 16

    Dallman PR, Siimes MA . Percentile curves for hemoglobin and red cell volume in infancy and childhood. The Journal of Pediatrics 1979; 94: 26–31.

Download references


AMA and this research were supported by the KiKa Foundation, Amstelveen, The Netherlands. In the Czech Republic RCC diagnosis and treatment were supported by a grant of the Ministry of Health for conceptual development of research organization 00064203 (University Hospital Motol, Prague, Czech Republic).

Author information

Correspondence to M M van den Heuvel-Eibrink.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Author Contributions

VHJV, AWL, CMN, MHE conceived and designed the study; AMA, AY, AF, PN analyzed the data; AMA, VHJV, AY, JJMD, RP, CMN, MHE interpreted the data; MD, HH, FL, BDM, MSL, JS, MZ, CMN, MHE treated patients and contributed patient samples; IB performed central review of bone marrow morphology; GG, BB performed central review of cytogenetics; AMA, VHJV, MHE wrote the paper; all authors approved the final version of the paper.

Supplementary Information accompanies this paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Aalbers, A., van der Velden, V., Yoshimi, A. et al. The clinical relevance of minor paroxysmal nocturnal hemoglobinuria clones in refractory cytopenia of childhood: a prospective study by EWOG-MDS. Leukemia 28, 189–192 (2014).

Download citation

Further reading