Letters to the Editor
900 10 Sillaber C, Herrmann H, Bennett K, Rix U, Baumgartner C, Bohm A et al. Immunosuppression and atypical infections in CML patients treated with dasatinib at 140 mg daily. Eur J Clin Invest 2009; 39: 1098–1109.
11 Chen J, Schmitt A, Chen B, Rojewski M, Rubeler V, Fei F et al. Nilotinib hampers the proliferation and function of CD8+ T lymphocytes through inhibition of T cell receptor signalling. J Cell Mol Med 2008; 12: 2107–2118.
Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)
Leukocytosis is associated with poor survival but not with increased risk of thrombosis in essential thrombocythemia: a population-based study of 311 patients
Leukemia (2010) 24, 900–903; doi:10.1038/leu.2010.5; published online 4 February 2010
Essential thrombocythemia (ET) is a myeloproliferative neoplasm characterized by a sustained increase in platelet count and long-term survival. The long-standing threshold platelet count to define thrombocytosis of 600 109/l was recently lowered to 450 109/l by the World Health Organization in the 2008 classification. Thromboses, acute myeloid leukemia (AML) or myelofibrosis are the main complications of ET. Life expectancy of ET patients is good, with median survival ranging from 13 to 22.3 years.1–3 Consequently, long-term follow-up studies that focus on ET-related complications and/or survival of ET patients must be conducted to register a sufficient number of events to provide informative results and statistical analyses. Many studies on outcomes in ET have been published, including well-conducted large-scale studies. However, most were single-institution studies that focused on patients referred to hospitals, which does not necessarily reflect the course of ET in the general population. Indeed, in these studies, the median age at diagnosis was younger (50–57-years old) than usually noted for ET patients in population-based studies (67–73-years old).4,5 Moreover, long-term outcomes in ET patients from population-based studies, usually involving o100 patients, are rarely investigated. Finally, in these studies, survival was generally based on observed mortality, a good but incomplete representation of overall mortality: given the old age (465 years) of half of the patients at the time of diagnosis, more specific analyses such as relative mortality, that describes mortality directly associated with the disease, thus ruling out biases related to age or other illnesses, should be carried out. Taking the above into account, we carried out a long-term (1980–2007) retrospective study on 311 ET patients (174 females and 137 males) registered in the Registry of Hematological Malignancies of the Coˆte d’Or area, France, the world oldest registry specialized in hematology. The median age at diagnosis of 68.4 years (range: 15.8–94.4 years), and the World Standardized Population incidence rate was 1.4 per 100 000 inhabitants per year. The median follow-up was 9.5 years (range: 0–28). In all, 86% had received cytoreductive drugs, mainly hydroxyurea or pipobroman, during the follow-up. Altogether, 70% had received anti-platelet drugs alone (14%) or associated with cytoreductive treatment. The JAK2–V617F mutation assay was carried out for 129 out of 311 patients; 91 out of 129 (70%) were positive. In our study, 125 (40%) patients died during the 1980–2007 period, at a median age of 81 years, with a median of overall survival of 11 years. Relative survival was 93, 83 and 47% after 5, 10 and 20 years of evolution, respectively. Leukemia
As a consequence of both older age at diagnosis and longer follow-up, the mortality rate was higher (40%) than in other reports (from 8 to 27%)1–3,6 which underlines the differences between population-based and single-center studies (Table 1). Nevertheless, this high rate of mortality allowed us to better analyze survival on a sufficiently large number of events (n ¼ 125) compared with previous studies with a low number of deaths (ranging from 20 to 30).1,4 When relative survival was analyzed, that is focusing on deaths directly associated with the disease, mortality because of ET was low in the first 10 years of evolution, as survival was 83% at 10 years. However, it decreased to 47% after 20 years of follow-up, with a generally regular curve with no steep falls. The difference between median observed and relative survival (11 versus 19.5 years) underlines the need to carry out the appropriate analyses to study the effect of a single disease on mortality. Finally, contrary to previous results7 but in accordance with population-based studies,4,5 survival in ET patients was significantly lower than the expected survival of an age- and sex-matched control population, and the discrepancy was mainly observed after 10 years of evolution. A history of thrombosis was associated with higher mortality, but only in univariate relative survival. In contrast, neither thrombocytosis 41000 109/l nor the presence of the JAK2–V617F mutation at diagnosis were significantly associated with poorer survival. Three independent parameters were noted as prognostic factors for relative survival in multivariate analysis: age 460 years, a hemoglobin level below normal values and leukocytosis 411 109/l. Our study provides data showing that, in relative survival, age 460 years is still a risk factor of survival; in other words, there is excess mortality because of ET per se in people over 60 years compared with younger patients. A hemoglobin level below normal values was noted in 54 patients at diagnosis and was associated with a poor prognosis. One could imagine that patients with low level of hemoglobin had a myelodysplastic syndrome or a prefibrotic stage of primary myelofibrosis that could have previously been misdiagnosed, as these disorders are associated with worse overall survival and frequently anemia. However, at diagnosis there were no signs of myelodysplastic feature or fibrosis at bone marrow examination, either in morphology or histology in these patients. On the other hand, data such as increased blasts in the bone marrow or cytogenetic abnormalities during evolution were sometimes missing, leading to a possible underestimation of myelofibrosis transformation. In the same way, our results confirm that leukocytosis at the time of diagnosis is also associated with poorer survival; compared with the previous studies that reported a white blood cell count (WBC) of 15 109/l associated with a worse
94 (30%) Age 460 years Anemia
Relative survival 20 Age 460 years Anemia WBC411 109/l
Relative survival N/A N/A Reduced Compared with age-match controls
Reduced Compared with age-match controls
2 (2.1%) N/A
32 (33%) No ASA treatment
Relative survival N/A N/A
1 (2.5%) N/A
N/A N/A
6 1–22
67 18–87
96 28
Jensen, 2000
Overall survival 22.6 Male sex HT
6 (1.37%) Sequential use of different myelosuppressive drugs
84 (19%) N/A
9.3 N/A
55 N/A
435 38
Passamonti, 2004
Overall survival 13 Age 460 years
4 (1.7%) Previous MF Melpahalan exposure
53 (23%) Age 460 years
N/A N/A
65 18–90
231 48.5
Chim, 2005
Overall survival 18.9 Age 460 years WBC 415 109/l Tobacco use DM
N/A None
N/A Age 460 years WBC 415 109/l HT
13.6 N/A
54 12–88
322 32
Wolanskyj, 2006
Overall survival 18 Age 460 years WBC 415 109/l Anemia Venous thrombosis DM Smoking
20 (3.3%) Anemia PC 41000 109/l Age
N/A N/A
7 0–35
57 5–91
605 34
Gangat, 2007
Overall survival 22.3 Age 460 years HT
14 (2.3 %) Age 460 years
156 (25.8%) Age 460 years HT (before and at diagnosis)
5.6 0–26
50 16–90
605 37
Passamonti, 2008
Non population-based studies
Overall survival 19.5 Age 460 years WBC 415 109/l Hypertension DM Thrombotic event during follow-up
6 (1.5%) Sequential use of different cytotoxic agents CA MF
45 (12%) Age 460 years HT before diagnosis
9.5 3–28.5
64 16–91
386 37
Palandri, 2009
Abbreviations: ASA, acetylsalicylic acid; DM, diabetes mellitus; CA, cytogenetic abnormalities; HT, history of thrombosis; HU, hydroxyurea; MF, myelofibrosis; N/A, not available; PC, platelet count; WBC, white blood cells.
Survival (years) Median Risk factors
Leukemic transformation n 18 (5.8 %) Risk factors None
Thrombosis N Risk factors
5.3 1–17
9.5 0–28
Follow-up (years) Median Range
39 35
Mesa, 1999
72 17–90
311 44
Present study
Population-based studies
Population and non population-based studies on the outcome in essential thrombocythemia
Age at diagnosis (years) Median 68.4 Range 15.8–94.4
N Male sex %
Table 1
Letters to the Editor
901
Leukemia
Letters to the Editor
Leukemia
prognosis,2,6 the threshold of WBC in our work was lower (11 109/l corresponding to the mean WBC value of the entire population). Of course, the analysis remained statistically significant when a threshold of WBC 15 109/l was tested. Taking into account the three independent parameters, a model for relative survival was constructed in order to define three groups of risk: high for patients with 2 or 3 risk factors, medium if there was 1 risk factor and low when there were no risk factors. Median survivals were 10, 16 and 18 years in the high, intermediate and low-risk groups, respectively (Po0.001) in relative survival analysis (Figure 1), thus confirming previous results.2 To avoid the confounding effect of cytoreductive therapy on this model, we checked that the proportion of patients who received cytotoxic agents was the same in the three groups (P40.05). Indeed, in our population, most of them (86%) had been treated with such therapy. Low level of hemoglobin and high platelet count 41000 109/l were independent risk factors for overall survival in multivariate analysis restricted to elderly people (460 years old); however, in this population 460 years, prognostic significance was only maintained for hemoglobin level in relative survival analysis. Overall, at least one thrombotic event was recorded in 94 patients (30%), including 30 thromboses (18 arterial and 12 venous) before diagnosis, 37 (24 arterial and 13 venous) at the time of diagnosis and 53 (33 arterial and 20 venous) during follow-up. In all, 6 patients had both arterial and venous thrombotic events, whereas 23 patients had two or more thromboses. During follow-up, the cumulative risk for thrombosis was 24, 27 and 42% after 10, 15 and 20 years, respectively. In univariate and multivariate analyses, male sex was an independent factor for arterial occlusive events (Po0.05), whereas age over 60 years and hemoglobin level below normal values (12.5 g/100 ml or 13.5 g/100 ml for women and men, respectively) were independent factors for venous events (Po0.05). Past history of venous thrombosis was a risk factor of recurrence at diagnosis of both arterial and venous thrombosis (P ¼ 0.025). When all occlusive events, including arterial and venous events, in the past history, at diagnosis and during followup, were gathered, multivariate analysis showed that thromboses were significantly associated with age over 60 years and a hemoglobin level below normal values (Po0.05). The results were similar in a multivariate thrombosis-free survival analysis (Po0.05 for the hemoglobin level, P ¼ 0.06 for an age460 years). When adjusted to risk category (that is, age 460 years, a hemoglobin level below normal values and leukocytosis 411 109/l), the thrombotic risk was significantly increased in high and intermediate risk groups compared with the low one, whereas no other independent factor was noticed in multivariate analysis. In contrast, neither leukocytosis nor thrombocytosis nor the JAK2–V617F mutation was associated with an increased risk of thrombosis, in keeping with recent studies which failed to show a clear correlation between platelet counts and the risk of thrombosis.8 For WBC, previous studies concluded that a leukocyte count above 8.7 109/l was associated with thrombosis,9 whereas others did not.1,3,6 One may outline that the former study, which reported on a large number of occlusive events (4300) had stronger statistical power than the latter studies, which reported on only 45–156 thrombotic events. In this study, most patients (86%) were treated with cytoreductive drugs and the number of occlusive events was low compared with the previous studies; this may be the reason why we failed to find an association between WBC and occlusive events. Eighteen (5.8%) ET patients (8 males and 10 females) transformed into AML according to the WHO criteria based
Relative Survival
100
Survival (%)
902
Low intermediate High
50
0 0
5
10 15 Time in years
20
Figure 1 Relative survival curves for ET patients stratified according to the risk including the three risk-factors: age460 years, hemoglobin level below normal values at diagnosis and hyperleucocytosis 411 109/l (Po0.05). The risk was high if there were X2 risk factors, intermediate if there was 1 risk factor and low when there were no risk factors. Median survivals were 10, 16 and 18 years, in the high, intermediate and low-risk groups, respectively (Po0.001).
on a 20% blast threshold in the bone marrow, during the followup, after a median delay of 6.26 years (range: 1.9–20 years) from the diagnosis of ET; the median age at the time of AML was 74 years. Five patients were younger than 60 years at the time of diagnosis of ET, whereas 13 patients were older. All had been treated with cytoreductive drugs, mainly hydroxyurea as the first-line therapy, and two with pipobroman as the second line. Leukemic transformation appeared at a higher rate and more quickly after diagnosis than classically described in singleinstitution reports.1–3,10 This is probably related to the higher median age of our population and the longer follow-up. The cumulative risk of leukemia was 8.3 and 15% after 10 and 15 years, respectively, associated with a 1-year survival after AML transformation of 12%. Neither univariate nor multivariate analysis revealed any risk factors for leukemia at the time of diagnosis, which contrasts with other conflicting studies that reported risk factors such as a hemoglobin level below normal values, age, platelet count 41000 109/l, exposure to melphalan treatment or sequential cytotoxic agents.1–3,6 However, fortunately, leukemic transformations are uncommon events in ET, and most studies, including ours, report few cases of leukemia secondary to ET (from 4 to 20 in the literature, 18 in our work), which means that the statistical power may be too low to draw definite conclusions about risk factors for leukemia. In conclusion, our study, which is to our best knowledge the largest to date on long-term outcomes in ET from a populationbased registry, provides original data on ET. We confirm that ET is an indolent hemopathy with median relative survival of approximately 20 years. Leukemic transformation is a rare but severe event in the course of the disease, and occurs in o6% of the overall population (15% in cumulative risk after 15 years of evolution), but is responsible for 15% of deaths directly because of ET. People aged over 60 years and/or with a hemoglobin level below normal values at the time of diagnosis are at a higher risk for both thromboses and mortality, which underlines the usefulness of paying particular attention to these patients. Prospective studies are needed to confirm or to rule out the accuracy of our stratifying model based on age, hemoglobin and leukocytes.
Conflict of interest The authors declare no conflict of interest.
Letters to the Editor
903 Acknowledgements We wish to thank Mrs Dominique Bouchot, Martine Courtois, Isabelle Helot and Aure´lie Herry (CHU Dijon), for excellent technical help, and Philip Bastable for revisions. This work was supported by a grant from the Ligue contre le Cancer, Comite´ de Coˆte d’Or.
F Girodon1,2, F Dutrillaux1, J Brose´us1, M Mounier2, V Goussot1, P Bardonnaud1, ML Chre´tien3 and M Maynadie´1,2 1 Laboratoire d’He´matologie, Hoˆpital du Bocage, CHU de Dijon, France; 2 Registre des he´mopathies malignes de Coˆte d’Or, Universite´ de Bourgogne, Dijon, France and 3 Service d’He´matologie Clinique, Hoˆpital du Bocage, CHU de Dijon, France E-mail:
[email protected] References 1 Chim CS, Kwong YL, Lie AK, Ma SK, Chan CC, Wong LG et al. Long-term outcome of 231 patients with essential thrombocythemia: prognostic factors for thrombosis, bleeding, myelofibrosis, and leukemia. Arch Intern Med 2005; 165: 2651–2658. 2 Gangat N, Wolanskyj AP, McClure RF, Li CY, Schwager S, Wu W et al. Risk stratification for survival and leukemic transformation in essential thrombocythemia: a single institutional study of 605 patients. Leukemia 2007; 21: 270–276.
3 Passamonti F, Rumi E, Arcaini L, Boveri E, Elena C, Pietra D et al. Prognostic factors for thrombosis, myelofibrosis, and leukemia in essential thrombocythemia: a study of 605 patients. Haematologica 2008; 93: 1645–1651. 4 Jensen MK, de Nully Brown P, Nielsen OJ, Hasselbalch HC. Incidence, clinical features and outcome of essential thrombocythaemia in a well defined geographical area. Eur J Haematol 2000; 65: 132–139. 5 Mesa RA, Silverstein MN, Jacobsen SJ, Wollan PC, Tefferi A. Population-based incidence and survival figures in essential thrombocythemia and agnogenic myeloid metaplasia: an Olmsted County Study, 1976–1995. Am J Hematol 1999; 61: 10–15. 6 Palandri F, Catani L, Testoni N, Ottaviani E, Polverelli N, Fiacchini M et al. Long-term follow-up of 386 consecutive patients with essential thrombocythemia: safety of cytoreductive therapy. Am J Hematol 2009; 84: 215–220. 7 Passamonti F, Rumi E, Pungolino E, Malabarba L, Bertazzoni P, Valentini M et al. Life expectancy and prognostic factors for survival in patients with polycythemia vera and essential thrombocythemia. Am J Med 2004; 117: 755–761. 8 Carobbio A, Finazzi G, Antonioli E, Guglielmelli P, Vannucchi AM, Delaini F et al. Thrombocytosis and leukocytosis interaction in vascular complications of essential thrombocythemia. Blood 2008; 112: 3135–3137. 9 Tefferi A, Gangat N, Wolanskyj A. The interaction between leukocytosis and other risk factors for thrombosis in essential thrombocythemia. Blood 2007; 109: 4105. 10 Wolanskyj AP, Schwager SM, McClure RF, Larson DR, Tefferi A. Essential thrombocythemia beyond the first decade: life expectancy, long-term complication rates, and prognostic factors. Mayo Clin Proc 2006; 81: 159–166.
Bimodal distribution of genomic MLL breakpoints in infant acute lymphoblastic leukemia treatment Leukemia (2010) 24, 903–907; doi:10.1038/leu.2010.14; published online 18 February 2010
In contrast to favorable cure rates in childhood acute lymphoblastic leukemia, prognosis in infant leukemia occurring within the first year of life remained poor, despite intense effort in therapy optimization in this critical subgroup. The early clinical manifestation, detection of leukemic clonespecific gene rearrangements in neonatal blood spots and high concordance rates in monochorial twins argue for a prenatal origin in most, if not all cases of infant leukemia.1 A hallmark of infant leukemia is the high prevalence of chromosomal rearrangements involving the MLL gene at chromosome band 11q23, detectable in about two-thirds of cases. The most frequent partner genes are AF4 (MLLT2), AF9 (MLLT4) and ENL (MLLT3), accounting for B85% of all MLL-positive cases.2 With respect to clinical course and biological features, infant leukemia exhibits different characteristics to a phenotypically and cytogenetically similar disease in older age.3,4 A high rate of breakpoints located in the telomeric portion of the MLL breakpoint cluster region (BCR), similar to therapyinduced acute leukemia associated with topoisomerase II inhibitors, suggested a similar mechanism for MLL rearrangement generation in infant leukemia. Other DNA features, such as scaffold-attachment regions (SAR), DNAse I hypersensitivity
sites and initiation sites of apoptotic nucleases, have been discussed to be involved in alternative break mechanisms.1 The aim of this study was to assess breakpoint distribution within the MLL gene in a cohort of 59 infants, in comparison with pediatric and adult patients, to evaluate age-dependent breakpoint distribution and association of recombination-related sequence motifs. Leukemic cells from infants enrolled in treatment protocols of the German and Austrian ALL-BFM 95 or the Interfant-99 study groups were analyzed. Patients’ characteristics are shown in Supplementary Table 1. Parents or legal guardians gave informed consent for treatment and DNA analysis. The distribution of MLL partner genes AF4, AF9 and ENL was similar to the allocation in the complete Interfant-99 study population.2 Breakpoint distribution of our cohort was compared with previously published MLL-AF4, MLL-AF9 and MLL-ENL breakpoints with information of the subject’s age.5–9 Methods for amplification of genomic fusions are described in the Supplementary Information. Amplicons spanning chromosomal breakpoints were directly sequenced and fusion sequences were aligned to reference sequences, including the MLL, AF4, AF9 and ENL BCR (extracted from the ENSEMBL database), to define breakpoint positions for subsequent analysis. Breakpoint clusters were defined as regions in which lower limit of 95% confidence band, determined by bootstrapping Leukemia
