SEARCH

SEARCH BY CITATION

Keywords:

  • child;
  • juvenile myelomonocytic leukaemia;
  • myelodysplastic syndrome;
  • incidence;
  • prognostic factors

Abstract

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Summary.  We aimed to identify and classify cases of paediatric myelodysplastic syndromes (MDS) occurring in Britain to estimate the incidence of this rare group of diseases, investigate the results of therapy and identify prognostic risk factors. Patients aged below 15 years at diagnosis were collected from England, Scotland and Wales, inclusively between 1990 and 1999. One hundred and thirty-five patients were accepted as de novo MDS or juvenile myelomonocytic leukaemia (JMML). The incidence for this period was 1·35 per million (age standardized rate) which is below that reported outside the UK. The overall survival was 45%[standard error (SE) = 4%] at 5 years: 40% (SE = 6%) for JMML and 50% (SE = 6%) for other MDS. Significant adverse prognostic factors for JMML were a platelet count < 40 × 109/l, raised fetal haemoglobin, FPC score and age above 2 years at diagnosis, for other MDS only monosomy 7 was significant. To conclude, the incidence of MDS/JMML in children in the UK appears to be lower than that reported outside the UK. This may be either a real difference in incidence or variation in reporting. Monosomy 7 is associated with poor outcome in MDS other than JMML. The prognosis of JMML depends on age, platelet count and fetal haemoglobin.

Myelodysplasia is uncommon in childhood but a number of recent publications have reviewed the clinical and laboratory features and attempted to improve classification (Hasle, 1994; Passmore et al, 1995). The classical Franco–American–British (FAB) system (Bennett et al, 1982) can be applied with some caveats to paediatric patients but more recent modifications, including those recommended by the World Health Organization (WHO) (Harris et al, 2000), have not been systematically applied in childhood. The best method for classifying paediatric myelodysplastic syndromes (MDS) thus remains uncertain and has been the subject of recent debate (Mandel et al, 2002; Hasle et al, 2003).

The incidence of paediatric MDS is also uncertain. The only published population-based studies are derived from small populations (Hasle et al, 1992, 1999a; Jackson et al, 1993). These incidence rates differ from each other quite markedly with annual incidences of 4·0/million, 3·2/million and 0·5/million observed, the latter being based on very small numbers. It remains unclear whether this variation is due to the differing diagnostic and inclusion criteria, incomplete ascertainment, a true variation in incidence or to chance. The larger studies are not truly population-based but include the largest United Kingdom (UK) series (Castro-malaspina et al, 1984; Creutzig et al, 1987; Brandwein et al, 1990; Tuncer et al, 1992; Passmore et al, 1995; Bader-Meunier et al, 1996; Groupe Français de Cytogénétique Hématologique, 1997; Hasle et al, 1999a; Lopes & Lorand-Metze, 1999; Luna-Fineman et al, 1999; Sasaki et al, 2001).

The UK paediatric MDS register was set up in 1991 to determine the incidence of these diseases in the UK, to obtain a population-based figure for survival, and to test the ‘FPC’ scoring system for paediatric MDS and juvenile myelomonocytic leukaemia (JMML), which is based on level of fetal haemoglobin (HbF), platelet count and cytogenetics (Passmore et al, 1995). We report here the results of the first 10 years of the register.

Case ascertainment.  The UK paediatric myelodysplasia register has collected cases of MDS occurring in children before age 15 years and from January 1990 onwards. Patients included in this study were diagnosed before 31 December 1999 and resident in England, Scotland or Wales at that time. They included those entered on the UK Medical Research Council (MRC) acute myeloid leukaemia (AML) trials 10 and 12 who were found on central morphological review to have bone marrow percentage of blasts less than 30% at the time of starting treatment. Other patients were ascertained through regular mailings to all members of the UK Children's Cancer Study Group (UKCCSG) and the paediatric haematology forum of the British Society of Haematology. A further check on registration was obtained through regional cancer registries (via the National Registry of Childhood Tumours), and the UK paediatric Bone Marrow Transplant (BMT) registry.

Diagnosis and classification.  Upon notification of a possible patient, forms were sent requesting clinical details, treatment plans and outcome data, and blood and bone marrow smears were requested for central review. The FAB classification was applied by two of the authors (J.M.C. and I.M.H.) with modification according to Neimeyer (Niemeyer et al, 1997), allowing an increased proportion of myeloblasts in the blood for chronic myelomonocytic leukaemia (CMML).

The criteria for the diagnosis of JMML were applied when adequate information was available (Emanuel, 1999). A small number of patients had no slides available for central review, and a judgement of the likely diagnosis was made based on data from the clinical records and authorized laboratory reports. When the JMML criteria were not fulfilled but this seemed the likely diagnosis, we classified these patients as CMML. Myeloid leukaemia in children with Down's syndrome (DS) is frequently preceded by a dysplastic prodrome, which may last many months. It is appropriate to classify all these patients separately as MDS/AML in DS, in contrast to other MDS, is extremely responsive to intensive chemotherapy, and this indicates that it is a distinct entity. We have excluded these patients from our analysis. Patients with granulocytic sarcoma and a low percentage of blasts in the marrow have been classified in some reports as MDS/JMML (Tuncer et al, 1992; Hasle et al, 1999a), however, these patients frequently have very favourable cytogenetics and respond as acute myeloid leukaemia (AML) patients so they were excluded.

Cytogenetics.  GTG-banded karyotype results were sent by referring hospitals, with no further review. In more recent years, some results showing monosomy 7 were confirmed by fluorescence in situ hybridization (FISH) using probes for the alpha-satellite region of chromosome 7. Several of the karyotypes submitted showing structural changes were not entirely characterized.

Management.  Patients with primary advanced MDS [refractory anaemia with excess blasts (RAEB) or RAEB in transformation (RAEBt)] were eligible for the MRC AML 10 and 12 trials and most were treated with intensive chemotherapy according to these protocols, although not all were registered on the trials. Where possible, information about BMT from a sibling or other donor was cross-referenced with the UK BMT registry database.

Incidence rates and survival.  Age-specific incidence rates were calculated using annual mid-year estimates of the population of Great Britain in the age groups 0, 1–4, 5–9, 10–14 years. Age standardized rates were calculated using the World Standard Population (Smith, 1992).

Patients were followed up to the end of 2001. Actuarial survival rates were calculated using the statistical package for social sciences (spss, Chicago, Illinois).

Overlap with other studies. Eleven of the patients included here were in the original series (Passmore et al, 1995), 13 were included in our previous report about chemotherapy in advanced MDS (Webb et al, 2002) and 17 are included in a multinational review of refractory anaemia (RA) (Kardos et al, 2003).

Incidence rates

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Two hundred and twenty-one patients were notified to the registry (see Table I). There were 135 patients with confirmed MDS/JMML. Incidence rates are shown in Table II, with those from Denmark and British Columbia for comparison. We also compared the combined incidence of MDS and AML to see if these differed between the countries. The age standardized incidence (ASR) of 1·35 per million was much lower than that reported from Denmark or British Columbia, so we performed a further calculation including 20 additional patients where the diagnosis was either in doubt, or not de novo MDS: Kostman MDS, Shwachman MDS, therapy-related MDS, patients where the diagnosis of MDS was probable but insufficient data were available, patients with chloroma and low percentage marrow blasts, patients with MDS transforming to AML in less than 2 months, and patients of Pearson's syndrome. This gave an ASR of 1·55/million/year and a crude rate of 1·43/million/year.

Table I.  Patients notified to the UK Paediatric Registry.
Disease categoryType/diagnosisNumber of patientsTotal
  • *

    AML, congenital dyserythropoietic anaemia, Philadelphia-positive chronic myeloid leukaemia, Fanconi's anaemia and severe aplastic anaemia.

  • RARS, refractory anaemia with ring sideroblasts.

de novoRA/RARS36135
RAEB19 
RAEBt13 
JMML/CMML67 
Additional patients included (see text)Kostman/Shwachman420
Therapy-related MDS6 
Probable MDS insufficient data4 
Chloroma blasts < 30%2 
MDS to AML in < 2 months2 
Pearson's2 
ExcludedDS MDS/AML/TAM2266
Non-DS TAM3 
Other*41 
Total  221
Table II.  Incidence of disease in the current study and two previous studies.
 United Kingdom de novo only (ASR)United Kingdom (ASR)*United Kingdom (crude)*Denmark (crude rate)British Columbia (crude rate)
  • *

    See text for details.

  • Based on average population of 10 879 479 children aged 0–14 years in Great Britain during the period of the study.

AML excluding DS 5·895·775·25·7
DS MDS/AML0·780·780·610·81·0
Total MDS/JMML1·351·551·433·52·5
JMML0·690·690·631·31·0
MDS other (all)0·660·810·802·11·4
MDS – RA0·350·390·360·420·35
Total non-DS 7·447·208·78·2

For the rest of the calculations, we included only the 135 patients with de novo MDS/JMML.

Clinical features

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

The median age of the whole group (135) was 38 months, for JMML/CMML was 23 months and for the other MDS was 57·5 months. There was a marked male predominance in the JMML/CMML group, with a male:female ratio of 49:18, but not in other MDS (29:39). There was a high incidence of associated abnormalities (see Table III) as has been noted previously (Hasle et al, 1995; Passmore et al, 1995; Groupe Français de Cytogénétique Hématologique; 1997), even after excluding the patients with DS.

Table III.  Associated abnormalities in de novo cases of paediatric MDS.
Type of MDSPatient numberAbnormalityNo.Percentage
  • *

    CINCA, chronic infantile neurological cutaneous arthritis; RARS, refractory anaemia with ring sideroblasts.

RA/RARS1Failure to thrive/developmental delay/renal tubular acidosis116/36 = 44%
2,3,4,5Syndromes: CINCA* (1), hypermelanosis of Ito (1), Cartilage hair hypoplasia (1), familial congenital nephrotic syndrome (1)4 
6Mild mental retardation/′coarse facial features′1 
7Factor XI deficiency1 
8Platelet storage pool disorder1 
9Asthma and diabetes1 
10Constitutional t(4;5)(q25,q33)1 
11,12Ex-premature (1 with intrauterine growth retardation and hepatosplenomegaly)2 
13Atrial septal defect/ventricular septal defect/patent ductus arteriosus1 
14Halothane hepatitis 4 months pre-MDS diagnosis1 
15Atrial septal defect/failure to thrive/ichthyosis1 
16Multiple café au lait spots1 
RAEB/t17Patent ductus arteriosus/short stature1 5/32 = 16%
18Platelet pool storage disorder1 
19Milroy's disease1 
20Multiple café au lait spots1 
21Dysmorphic features and immune deficiency1 
JMML/CMML22–28NF1 and/or juvenile xanthogranuloma716/67 = 24%
29Secundum ASD1 
30Sickle cell trait1 
31Cardiomyopathy/mitral valve and outflow tract abnormalities1 
32Immune deficiency/atopy/food intolerance1 
33Bilateral hearing impairment/renal abnormalities/third centile for height1 
34Thyroglossal cyst1 
35Leukodystrophy1 
36Congenital CMV1 
37Trisomy 8 (constitutional)1 
Total  3737/135 = 27%

Laboratory features

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Morphology review was available for 125/135 of the patients, 61 of whom were JMML/CMML and 64 of whom were MDS. Twelve of the 67 JMML/CMML patients did not strictly fulfil the diagnostic criteria for JMML. None of these 12 had the full range of investigations performed for JMML, in particular granulocyte–macrophage colony stimulating factor (GMCSF) hypersensitivity, so it is possible that they may have fulfilled the JMML criteria if these had been performed.

All but one of the 36 patients with RA had slides reviewed. Twenty-six of these 36 patients (72%) classed as RA had haemoglobin (Hb) less than 10 g/dl. Sixteen of the 36 had trilineage dysplasia, 14 had bilineage dysplasia, five had single lineage dysplasia and one had very poor films with minimal dysplasia. It was felt on balance that the latter patient should be classed as RA, in agreement with the local haematologists. Twenty-one of the 36 RA patients had platelet counts of less than 100 × 109/l (range 8–586) and 22 had neutrophils less than 1·5 × 109/l (range 0–15·7). As can be seen from Table IV, 16 of these patients with RA had no cytogenetic abnormalities and, of these 16, all but three had at least bilineage dysplasia. These three included the patient with poor films mentioned previously, a patient with chronic infantile neurological cutaneous arthritis (CINCA) syndrome (patient 2) and one with halothane hepatitis (Patient 14), who had dysplasia on the blood film, a relatively normal bone marrow appearance and was observed over a long time period before being accepted as RA.

Table IV.  Cytogenetics.
 JMML/CMML n = 67RA/RARS n = 36RAEB/t n = 32Total n = 135 (100%)
  • *

    Either 7q– or 7 abnormalities with other abnormalities also.

  • One failed.

  • RARS, refractory anaemia with ring sideroblasts; n, number of patients.

No clonal abnormality43 (64%)16 (44%)11 (34%)70 (52%)
Monosomy 7 alone11 (17%)12 (33%)8 (25%)31 (23%)
Other 7 abnormalities* 2 (3%) 3 (8·5%)3 (9%)8 (6%)
Other complex (≥ 2 abnormalities) 3 (4%) 04 (13%)7 (5%)
Trisomy 8 2 (3%) 2 (6%)2 (6%)6 (4%)
Others 6 (9%) 3 (8·5%) 4 (13%)13 (10%)
Details of others and complex Numbers of cells analysed shown in [] brackets where data entered. partial +8c45,X, –Y [20];17q–51–53,XX,+6,+8,+8,+10, +15,+19[cp13] 
 6q–46,XX,del(13)(q?13q22)[7]/46, XX,del(13),del(16)(q?13)[3] 
49,XY,+X,+9,+14 [10/30];t(4;5)(q23:q33)c50,XX+19,+21,+22,+mar 
t(1;5); 48,XX,+8,+21 
t(13;14) 9q– 
t(16;21) t(6:9)(p23:q34.3) 
45–46,XY,t(2:12:?),11, inc[29]/46XY[1] 46,xy 
46,XX t(9;11)(p22;q23), add(12)(p?12) t(9;11;13)(p21;q25;q22)der(11) ins(11;13)(p15;q14;q22) 
46,XX t(9;19)(p2?2;q13)   

Cytogenetics

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Cytogenetic analysis of the bone marrow had been attempted in all of the 135 de novo patients and failed in only one. The results can be seen in Table IV.

Sixty-four per cent of patients with JMML/CMML had a normal karyotype, as did 44% of patients with RA and 34% with RAEB/t. Monosomy 7 alone, found in 23% of patients, was the commonest abnormality both overall and in each subgroup. Other chromosome 7 abnormalities were found in a further 6% of patients and the next most common finding was trisomy 8 (4%). Other abnormalities are also shown in Table IV.

Treatment and survival

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Twenty-four patients (14 RA, three RAEB/t and 14 JMML/CMML) received no specific treatment, either because they had stable disease and were not transfusion dependent or because they were waiting for a suitable stem cell donor to be found. Results of treatment by broad category are given in Table V.

Table V.  Summary of treatment.
 No treatmentIntensive chemotherapyStem cell transplantAutologous transplantSplenectomy alone
nStatusnStatusnStatusnStatusnStatus
  1. n, number of patients.

RA (n = 36)149 alive 2·4–10 years40 alive158 alive 4 month-12 years10 alive21 alive 9 years
RAEB (n = 19)31 alive 6·3 years8 (1 of these BMT after relapse)3 alive 3·8–7·9 years 0 alive8 (in 1st CR)3 alive 3–8·7 years    
RAEBt (n = 13)  12 (3 of these BMT after relapse)6 alive 3·5–7·4 years 1 alive 4 years (total 7 alive)1 (in 1st CR)1 alive 9·4 year    
JMML (n = 67)145 alive 2·9–4·7 years10 (2 of these BMT after relapse)3 alive 4·3–10·9 years 1 alive 4·4 years (total 4 alive)3916 alive 1–10·3 years1 (cultured)1 alive 11·5 years31 alive 7·1 years

Some of the patients who received no specific treatment warrant further explanation: one child with RA (patient 5) received cyclosporine following a renal transplant and is alive more than 10 years from diagnosis, having improved on this treatment with no specific treatment for MDS.

One patient with RAEB (patient 21) who had intravenous immunoglobulin only is alive 3·8 years from diagnosis, another child received cytokines only, and a third child with RAEB had reduced cardiac function and went into cardiac failure on receiving preparative therapy for a low-intensity BMT.

Survival of JMML/CMML versus primary MDS is shown in Fig 1.

image

Figure 1. Survival of MDS and JMML/CMML patients.

Download figure to PowerPoint

Factors influencing survival

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Table VI shows the results of analysis of factors influencing prognosis that have been performed separately for patients with JMML/CMML and other MDS. We have attempted to apply the International Prognostic Scoring System (IPSS) (Greenberg et al, 1997) to these patients and found that it is not a good predictor of overall survival. There were very few children (nine JMML/CMML patients and two other MDS patients) placed in the lowest risk group for IPSS score. We did two analyses: the first (shown in Table VI) was overall survival and the second censored transplanted patients at the time of stem cell transplant. For both methods of analysis, the only component of the IPSS score that was significant was the karyotype. For MDS other than JMML/CMML, monosomy 7 either alone or in combination with other cytogenetic abnormalities was a prognostic factor for a poorer outcome (Fig 2), whereas for JMML/CMML it was associated with a better prognosis, although it was not quite statistically significant (P = 0·15). The FPC score, which is dependent on platelet count, fetal haemoglobin and cytogenetic complexity, remains of prognostic value in the JMML patients (Fig 3), with both HbF and platelet count being significant, but not in the other MDS. However, only 39 of the other MDS patients had all the necessary measurements for FPC score.

Table VI.  Prognostic factors.
FactorJMML/CMML (numbers in parenthesis = survival at 5 years)Other MDS (numbers in parenthesis = survival at 5 years)Total patients
  • *

    Monosomy 7 is a bad prognostic factor for other MDS but ‘good’ for JMML/CMML.

  • P-values were determined using the log-rank test. Mon 7, monosomy 7.

Age < > 2 yearsP = 0·015< 2 n = 34 (54%)P = 0·80< 2 n = 16 (50%)67:68
> 2 n = 33 (26%) > 2 n = 52 (50%) 
IPSSP = 0·55Low n = 9 (55%)P = 0·41Low n = 266:63
Int-1 n = 35 (35%) Int1 n = 29 (48%) 
Int-2 n = 16 (36%) Int2 n = 15 (46%) 
High n = 6 (56%) High n = 17 (47%) 
Cytopenias (IPSS)P = 0·08n = 7 (57%)P = 0·22n = 167:65
n = 24 (48%) n = 17 (68%) 
n = 34 (32%) n = 26 (39%) 
n = 2 (0%) n = 21 (52%) 
Cytogenetics (IPSS)P = 0·34Good n = 44 (39%)P = 0·006Good n = 28 (62%)66:67
Int n = 7 (38%) Int n = 11 (80%) 
Poor n = 15 (48%) Poor n = 28 (28%) 
FPC scoreP = 0·003n = 22 (56%)P = 0·39n = 14 (51%)65:39
n = 24 (44%) n = 17 (51%) 
n = 17 (18%) n = 8 (38%) 
n = 2 (50%)   
Monosomy 7*P = 0·15Mon 7 n = 13 (50%)P < 0·01Mon 7 n = 26 (30%)67:67
  No Mon 7 n = 54 (38%) No Mon 7 n = 41 (64%) 
image

Figure 2. Survival of other MDS patients by IPSS karyotype.

Download figure to PowerPoint

image

Figure 3. Survival of JMML/CMML patients by FPC score.

Download figure to PowerPoint

Classification of paediatric MDS

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

There has recently been much interest in the classification of paediatric myelodysplastic and myeloproliferative disorders (Aricòet al, 1997; Harris et al, 2000; Mandel et al, 2002; Hasle et al, 2003; unpublished observations). It is clear that JMML has unique clinical and biological properties (Niemeyer et al, 1997; Emanuel, 1999). The recent WHO proposal for the classification of haematological malignancies identifies a subgroup of mixed proliferative dysplastic/proliferative disorders, which includes JMML. However, it remains uncertain whether all patients presenting in childhood with CMML morphologically fulfil these more stringent diagnostic criteria (we have retained this term, although some think that it should be dropped because CMML as seen in adults virtually never occurs in children). Eight-two per cent of our patients with CMML morphologically fulfilled the criteria for JMML and, in patients where criteria were not fulfilled, this may have been caused by missing data.

The new WHO classification recognizes two categories of MDS without excess blasts: RA and refractory cytopenia with multilineage dysplasia. Most of our patients had anaemia and those that did not had at least bilineage dysplasia, and we agree that refractory cytopenia was a more suitable phrase for use in childhood (Kardos et al, 2003). As we have noted before (Passmore et al, 1995), RA with ringed sideroblasts is excessively rare in childhood and no new patients were identified in this survey. The only confirmed patient, previously reported by us, had a very strong family history of AML/MDS/PSPD (platelet storage pool disorder).

Although the WHO proposes to reduce the bone marrow blast percentage cut-off for AML to 20%, we have kept these patients defined as RAEBt, as has Hasle (Hasle et al, 2003), until a time when there is good evidence to support an arbitrary cut-off of 20% rather than 30%.

A triple approach to classification based on underlying abnormalities, cytology and cytogenetics has recently been suggested by Mandel et al (2002). They applied this system retrospectively to a cohort of 40 patients but it is unclear how it would translate into a prognostic score. The population of patients studied was highly selected as only 16 of the 40 had primary MDS, the remainder having a variety of underlying disorders. In contrast, we have excluded all therapy-related and secondary MDS from the survival analyses in the series.

Cytogenetics

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

Cytogenetic analysis confirmed once again the dominance of monosomy 7 as a finding in paediatric MDS, and also the absence of abnormalities of 5q. We had previously postulated that a separate subgroup of infantile monosomy 7 should be identified: those children under 4 years with any type of MDS and monosomy 7. As most of these patients fulfil the diagnostic criteria for JMML, we have included them as such here. It is, however, of some interest that among the group of patients with JMML, monosomy 7 appeared to be associated with a better prognosis.

Among the abnormalities that did not involve chromosome 7 (Table V), several of the translocations that were submitted failed to identify the chromosomal breakpoints involved in the rearrangements described. The t(13;14)(q10;q10) has been described as a constitutional anomaly in at least two patients with MDS in childhood (Cerretini et al, 2002). It would, therefore, seem important to be able to discern whether the t(13;14) described in one of the JMML/CMML patients was acquired or constitutional. There is strong evidence that constitutional trisomy 8 predisposes to the development of malignancy (Mitelman, 1991; Brady et al, 2000), and one of the patients in the present study who was the subject of a previous publication (Brady et al, 2000) had constitutional, partial trisomy 8.

The incidence of t(6;9)(p23;q34) in haematological malignancies is rare in comparison with other chromosomal abnormalities, and is found in specific subtypes of AML and MDS, especially in RAEB (Lillington et al, 1993; Jadayel et al, 1995). A survey of patients with t(6;9) in the literature (Shapira et al, 1999) shows that patients with MDS and t(6;9) tend to have a prognosis at least as good as that of the general RAEB/RAEBt population (mean survival after diagnosis 15·1 months).

Calculations of incidence: inclusions and exclusions

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

We have encountered a number of problems in the comparison of incidence with other published series (Hasle et al, 1995, 1999b). These problems partly reflect the inclusion and exclusions of various types of MDS. Myeloid leukaemia in children with DS, and patients with granulocytic sarcoma and a low percentage of blasts in the marrow have been excluded from this report as explained in Patients and methods.

The recent WHO classification has abolished the category of RAEBt and, whatever the merits of this suggestion, it emphasizes the difficulties of distinguishing advanced MDS from AML. In order to overcome this problem, we have examined the incidence of AML and MDS combined and the difference in incidence between the UK and Denmark and British Columbia remains. We have also excluded patients who had a blast percentage in the bone marrow of over 30% within 3 months of the initial marrow examination, as have others (Groupe Français de Cytogénétique Hématologique, 1997). However, some of the children in this study who had a diagnosis of MDS RAEB/t were treated immediately with chemotherapy and may, if left untreated for a few weeks, have developed AML.

MDS is a well-recognized complication of many congenital bone marrow disorders and, because of difficulties interpreting the marrows of many of these disorders, we excluded these patients from all our analyses except for the comparison of incidence.

After the above exclusions, there remains a group of ‘de novo’ MDS/JMML patients, which allows a more meaningful comparison of incidence. The incidence rates found in this study were considerably lower than those calculated by Hasle and colleagues (Hasle et al, 1995, 1999b), even after removing all the DS patients from the Danish series (see Table II), as a difference in incidence as a result of DS may be expected if the incidence of DS varied substantially between populations. It may be assumed that if any of these diseases were under-ascertained it would be RA, and that JMML was unlikely to be missed, however, the table shows that JMML rates are lower than those in British Columbia and approximately half those for Denmark, whereas the rates of RA are the same as for British Columbia and only slightly lower than Denmark. The rates of Philadelphia-chromosome-positive chronic myeloid leukaemia (CML) for the three countries are comparable (0·54 per million UK, 0·7 per million Denmark and 0·5 for British Columbia); unfortunately, we do not have accurate figures for the rates of aplastic anaemia in the UK which could be mistaken for MDS in some patients. It is possible that the rates of MDS and JMML are indeed dissimilar in the different countries.

Although not population based, the Japanese study (Sasaki et al, 2001) reported MDS in childhood to be 7·7% of all leukaemias (4·9% if the therapy-related and DS patients are excluded). This compares with 6% reported for British Columbia (4·6% if DS and therapy are excluded) and 9% for Denmark (6·8% if DS and therapy, and Fanconi's anaemia are excluded).

We have again confirmed the high incidence of associated abnormalities in children with MDS/JMML, even after excluding all the DS patients.

Prognosis

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

The scoring systems developed to guide prognosis in adult MDS have been of little help in paediatric cases and, in a previous paper (Passmore et al, 1995), the authors described a scoring system for paediatric MDS, using platelet count, fetal haemoglobin and cytogenetic complexity score at diagnosis, developed from a series of childhood MDS patients diagnosed between 1971 and 1991. Other authors have found age, fetal haemoglobin and platelet count to be significant factors in predicting outcome in these diseases (Castro-malaspina et al, 1984; Niemeyer et al, 1997). The IPSS scoring system for adult MDS was not useful in this group of paediatric patients. This was partly because very few children have low IPSS scores and partly because of the weight given to high blast percentage in the bone marrow. However, the IPSS cytogenetic score did prove to show a significant survival difference that was due to monosomy 7 having a poor prognosis. For JMML, IPSS was unhelpful. The FPC score continues to be prognostic in JMML but the only significant prognostic factor for MDS in childhood is the presence of monosomy 7. In contrast, monosomy 7 in JMML does not confer a worse prognosis than other karyotypes.

The study was not designed as a formal evaluation of treatment, although BMT was advocated for children with CMML, JMML and for those with symptomatic RA, especially in conjunction with cytogenetic abnormalities. The outcome for matched family donor stem cell transplantation is very good for RA without progression, but the results from alternative donor stem cell transplant in RA and JMML are much less satisfactory.

Patients with more advanced MDS were eligible for chemotherapy and, as we have shown, the outcome as for AML is dependent on cytogenetics.

The UK paediatric MDS register has collected data on 135 confirmed de novo patients in its first 10 years and case ascertainment is continuing. The data collected have been used to develop national guidelines for diagnosis and treatment. There are several useful prognostic indicators for JMML, but cytogenetics remain the only useful guide in other MDS.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References

This registry could not be run without the help of all the consultants and data managers at the UKCCSG centres, and the authors wish to thank Janette Wallis for secretarial and administrative help, and Mary Kroll for regression analysis. J.M.C .and H.K. are supported by the Leukaemia Research Fund.

References

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Incidence rates
  6. Clinical features
  7. Laboratory features
  8. Cytogenetics
  9. Treatment and survival
  10. Factors influencing survival
  11. Discussion
  12. Classification of paediatric MDS
  13. Cytogenetics
  14. Calculations of incidence: inclusions and exclusions
  15. Prognosis
  16. Acknowledgments
  17. References
  • Aricò, M., Biondi, A. & Pui, C.-H. (1997) Juvenile myelomonocytic leukemia. Blood, 90, 479488.
  • Bader-Meunier, B., Miélot, F., Tchernia, G., Buisine, J., Delsol, G., Duchayne, E., Lemerle, S., Leverger, G., De Lumley, L., Manel, A.-M., Nathanson, M., Plantaz, D., Robert, A., Schaison, G., Sommelet, D. & Vilmer, E. for the French Society of Paediatric Haematology and Immunology (1996) Myelodysplastic syndromes in childhood: report of 49 patients from a French multicentre study. British Journal of Haematology, 92, 344350.
  • Bennett, J.M., Catovsky, D., Daniel, M.T., Flandrin, G., Galton, D.A.G., Gralnick, H.R., Sultan, C. & the French–American–British (FAB) Co-operative Group (1982) Proposals for the classification of the myelodysplastic syndromes. British Journal of Haematology, 51, 189199.
  • Brady, A.F., Waters, C.S., Pocha, M.J. & Brueton, L.A. (2000) Chronic myelomonocytic leukaemia in a child with constitutional partial trisomy 8 mosaicism. Clinical Genetics (Copenhagen), 58, 142146.
  • Brandwein, J.M., Horsman, D.E., Eaves, A.C., Eaves, C.J., Massing, B.G., Wadsworth, L.D., Rogers, P.C.J. & Kalousek, D.K. (1990) Childhood myelodysplasia: suggested classification as myelodysplastic syndromes based on laboratory and clinical findings. American Journal of Pediatric Hematology/Oncology, 12, 6370.
  • Castro-malaspina, H., Schaison, G., Passe, S., Pasquier, A., Berger, R., Bayle-Weisgerber, C., Miller, D., Seligmann, M. & Bernard, J. (1984) Subacute and chronic myelomonocytic leukemia in children (juvenile CML). Clinical and hematologic observations, and identification of prognostic factors. Cancer, 54, 675686.
  • Cerretini, R., Acevedo, S., Chena, C., Belli, C., Larripa, I. & Slavutsky, I. (2002) Evaluation of constitutional chromosome aberrations in hematologic disorders. Cancer Genetics and Cytogenetics, 134, 133137.
  • Creutzig, U., Cantù-Rajnoldi, A., Ritter, J., Romitti, L., Odenwald, E., Conter, V., Riehm, H. & Masera, G. (1987) Myelodysplastic syndromes in childhood. Report of 21 patients from Italy and West Germany. American Journal of Pediatric Hematology/Oncology, 9, 324330.
  • Emanuel, P.D. (1999) Myelodysplasia and myeloproliferative disorders in childhood: an update. British Journal of Haematology, 105, 852863.
  • Greenberg, P., Cox, C., Le Beau, M.M., Fenaux, P., Morel, P., Sanz, G., Sanz, M., Vallespi, T., Hamblin, T.J., Oscier, D.G., Ohyashiki, K., Toyama, K., Aul, C., Mufti, G. & Bennett, J. (1997) International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood, 89, 20792088.
  • Groupe Français de Cytogénétique Hématologique (1997) Forty-four cases of childhood myelodysplasia with cytogenetics, documented by the Groupe Français de Cytogénétique Hématologique. Leukemia, 11, 14781485.
  • Harris, N.L., Jaffe, E.S., Diebold, J., Flandrin, G., Muller-Hermelink, H.K., Vardiman, J., Lister, T.A. & Bloomfield, C.D. (2000) The world health organization classification of neoplastic diseases of the haematopoietic and lymphoid tissues: report of the clinical advisory committee meeting, Airlie house, Virginia, November 1997. Histopathology (Oxford), 36, 6986.
  • Hasle, H. (1994) Myelodysplastic syndromes in childhood: classification, epidemiology, and treatment. Leukemia and Lymphoma, 13, 1126.
  • Hasle, H., Jacobsen, B.B. & Pedersen, N.T. (1992) Myelodysplastic syndromes in childhood: a population-based study of nine cases. British Journal of Haematology, 81, 495498.
  • Hasle, H., Kerndrup, G. & Jacobsen, B.B. (1995) Childhood myelodysplastic syndrome in Denmark: incidence and predisposing conditions. Leukemia, 9, 15691572.
  • Hasle, H., Wadsworth, L.D., Massing, B.G., McBride, M. & Schultz, K.R. (1999a) A population-based study of childhood myelodysplastic syndrome in British Columbia, Canada. British Journal of Haematology, 106, 10271032.
  • Hasle, H., Aricò, M., Basso, G., Biondi, A., Rajnoldi, A.C., Creutzig, U., Fenu, S., Fonatsch, C., Haas, O.A., Harbott, J., Kardos, G., Kerndrup, G., Mann, G., Niemeyer, C.M., Ptoszkova, H., Ritter, J., Slater, R., Starý, J., Stollmann-Gibbels, B., Testi, A.M., Van Wering, E.R. & Zimmermann, M. (1999b) Myelodysplastic syndrome, juvenile myelomonocytic leukemia, and acute myeloid leukemia associated with complete or partial monosomy 7. European Working Group on MDS in Childhood (EWOG-MDS). Leukemia, 13, 376385.
  • Hasle, H., Neimeyer, C., Chessells, J., Baumann, I., Bennett, J., Kerndrup, G. & Head, D. (2003) A pediatric approach to the WHO classification of myelodysplastic and myeloproliferative diseases. Leukemia, 17, 277282.
  • Jackson, G.H., Carey, P.J., Cant, A.J., Bown, N.P. & Reid, M.M. (1993) Myelodysplastic syndromes in children. British Journal of Haematology, 84, 185186.
  • Jadayel, D., Calabrese, G., Min, T., Van Rhee, F., Swansbury, G.J., Dyer, M.J., Maitland, J., Palka, G. & Catovsky, D. (1995) Molecular cytogenetics of chronic myeloid leukaemia with atypical t(6;9)(p23;q34) translocation. Leukemia, 9, 981987.
  • Kardos, G., Baumann, I., Passmore, S.J., Locatelli, F., Hasle, H., Schultz, K., Stary, J., Schmitt-Graeff, A., Fischer, A., Harbott, J., Chessells, J., Hann, I., Fenu, S., Cantu Rajnoldi, A., Kerndrup, G., Van Wering, E., Rogge, T., Nollke, P. & Neimeyer, C. (2003) Refractory anemia in childhood: a retrospective analysis of 67 cases with particular reference to monosomy 7. Blood (in press).
  • Lillington, D.M., MacCallum, P.K., Lister, T.A. & Gibbons, B. (1993) Translocation t(6;9)(p23;q34) in acute myeloid leukemia without myelodysplasia or basophilia: two cases and a review of the literature. Leukemia, 7, 527531.
  • Lopes, L.F. & Lorand-Metze, I. (1999) Childhood myelodysplastic syndromes in a Brazilian population. Pediatric Hematology and Oncology, 16, 347353.
  • Luna-Fineman, S., Shannon, K.M., Atwater, S.K., Davis, J., Masterson, M., Ortega, J., Sanders, J., Steinherz, P., Weinberg, V. & Lange, B.J. (1999) Myelodysplastic and myeloproliferative disorders of childhood: a study of 167 Patients. Blood, 93, 459466.
  • Mandel, K., Dror, Y., Poon, A. & Freedman, M.H. (2002) A practical, comprehensive classification for pediatric myelodysplastic syndromes: the CCC system. Journal of Pediatric Hematology and Oncology, 24, 596605.
  • Mitelman, F. (1991) Catalogue of Chromosome Aberrations in Cancer, pp. 575683. 4th Edn. Wiley-Liss, New York.
  • Niemeyer, C.M., Aricó, M., Basso, G., Biondi, A., Cantú Rajnoldi, A., Creutzig, U., Haas, O., Harbott, J., Hasle, H., Kerndrup, G., Locatelli, F., Mann, G., Stollmann-Gibbels, B., Van't Veer-Korthof, E.Th., Van Wering, E., Zimmermann, M. & Members of the European Working Group on Myelodysplastic Syndromes in Childhood (EWOG-MDS) (1997) Chronic myelomonocytic leukemia in childhood: a retrospective analysis of 110 cases. Blood, 89, 35343543.
  • Passmore, S.J., Hann, I.M., Stiller, C.A., Ramani, P., Swansbury, G.J., Gibbons, B., Reeves, B.R. & Chessells, J.M. (1995) Pediatric myelodysplasia: a study of 68 children and a new prognostic scoring system. Blood, 85, 17421750.
  • Sasaki, H., Manabe, A., Kojima, S., Tsuchida, M., Hayashi, Y., Ikuta, K., Okamura, J., Koike, K., Ohara, A., Ishii, E., Komada, Y., Hibi, S. & Nakahata, T. (2001) Myelodysplastic syndrome in childhood: a retrospective study of 189 patients in Japan. Leukemia, 15, 17131720.
  • Shapira, M.Y., Hirshberg, B., Amir, G. & Rund, D. (1999) 6;9 translocation in myelodysplastic syndrome. Cancer Genetics and Cytogenetics, 112, 5759.
  • Smith, P.G. (1992) Comparison between registries: age-standardized rates. Cancer Incidence in Five Continents, Vol. VI (ed. by D. M.Parkin, C. S.Muir, S. L.Whelan, Y.-T.Gao, J.Ferlay & J.Powell), pp. 865870. IARC Scientific Publications, Lyon.
  • Tuncer, M.A., Pagliuca, A., Hicsonmez, G., Yetgin, S., Ozsoylu, S. & Mufti, G.J. (1992) Primary myelodysplastic syndrome in children: the clinical experience in 33 cases. British Journal of Haematology, 82, 347353.
  • Webb, D.K.H., Passmore, S.J., Hann, I.M., Harrison, G., Wheatley, K. & Chessells, J.M. (2002) Results of treatment of children with refractory anaemia with excess blasts (RAEB) and RAEB in transformation (RAEBt) in Great Britain 1990–1999. British Journal of Haematology, 117, 3339.