Abstract

Post–polycythemia vera myelofibrosis (post-PV MF) is a late evolution of PV. In 647 patients with PV, we found that leukocytosis leukocyte count > (15 × 109/L) at diagnosis is a risk factor for the evolution of post-PV MF. In a series of 68 patients who developed post-PV MF, median survival was 5.7 years. Hemoglobin level less than 100 g/L (10 g/dL) at diagnosis of post-PV MF was an independent risk factor for survival. The course of post-PV MF, however, is a dynamic process that implies a progressive worsening of clinical parameters. Using a multivariate Cox proportional hazard regression with time-dependent covariates, we found that a dynamic score based on hemoglobin level less than 100 g/L (10 g/dL), platelet count less than 100 × 109/L, and leukocyte count more than 30 × 109/L is useful to predict survival at any time from diagnosis of post-PV MF. The resulting hazard ratio of the score was 4.2 (95% CI: 2.4-7.7; P < .001), meaning a 4.2-fold worsening of survival for each risk factor acquired during follow up. In conclusion, leukocytosis at diagnosis of PV is a risk factor for evolution in post-PV MF. A dynamic score based on hemoglobin level, and platelet and leukocyte count predicts survival at any time from diagnosis of post-PV MF.

Introduction

Post–polycythemia vera myelofibrosis (post-PV MF) is a recently named condition1  that represents the natural evolution of patients with polycythemia vera (PV).2  The criteria proposed for the diagnosis of post-PV MF1  set the time point of evolution along the natural history of the disease. Patients' survival after transition to MF, as well as the prognostic factors for survival, are not defined.

Post-PV MF is a delayed event in the course of PV. No risk factors for this condition have been identified so far. In patients with PV, the 15-year risk of evolution to myelofibrosis is estimated at 6% and the incidence is 5.1 × 1000 person-years.3  A similar figure is reported in young patients with PV.4  Patients with post-PV MF have a high rate of detection of the JAK2 (V617F) mutation ranging from 91%5  to 100%.6  Concerning the JAK2 (V617F) mutation burden, patients with post-PV MF have the highest proportion of mutant alleles in patients with chronic myeloproliferative disorders (CMDs).6  An abnormal stem cell trafficking has been reported in patients with post-PV MF.7-9 JAK2 (V617F) may activate circulating granulocytes, playing a role in the constitutive mobilization of CD34+ cells into peripheral blood. This phenomenon is particularly evident in patients with PV and post-PV MF.6 

Current treatments for patients with post-PV MF do not affect survival and are considered palliative.10  Allogeneic hematopoietic stem cell transplantation is the only curative treatment for post-PV MF. Only a few patients, however, have been treated with fully ablative11  or reduced-intensity conditioning allogeneic transplantation.12  Clinical trials on JAK2 inhibitors are still underway.13 

In a cohort of 647 patients with PV, 68 developed post-PV MF according to the criteria of the International Working Group on Myelofibrosis Research and Treatment (IWG-MRT).1  The aim of this study is to define survival of patients with post-PV MF and to identify prognostic factors for survival. We developed a dynamic prognostic model useful to predict survival at any time from diagnosis.

Methods

Patients

In 76 consecutive patients previously diagnosed as post-PV MF, systematic revision of clinical and histopathological records identified 68 patients fitting diagnostic criteria for post-PV MF established by the IWG-MRT.1  Patients were followed from 1982 to 2007 at the Division of Hematology of the Fondazione Policlinico San Matteo, University of Pavia, and at the Division of Hematology of the Niguarda Ca' Granda Hospital, Milan, Italy. Patients of the 2 institutions were well matched with regard to baseline demographic and disease characteristics. The diagnosis of PV and of primary myelofibrosis (PMF) was made in accordance with the criteria in use at the time of first observation.14-16  The study was approved by the institutional ethics committee of Pavia, and the procedures followed were in accordance with the Declaration of Helsinki. Samples for molecular analysis were obtained after patients provided written informed consent, also in accordance with the Declaration of Helsinki.

JAK2 (V617F) mutational analysis

Granulocytes were obtained from the neutrophil fraction by osmotic lysis of red cells. Genomic DNA was obtained using the Puregene Blood DNA isolation kit (Gentra Systems, Minneapolis, MN). A quantitative real-time polymerase chain reaction–based allelic discrimination assay was used to detect the V617F mutation of JAK2 gene.17 

Flow cytometric analysis of circulating CD34+ cells

Circulating CD34+ cells were enumerated by flow cytometry using a single-platform assay as previously described,18  following the cell-gating guidelines recommended by the International Society for Hematotherapy and Graft Engineering (ISHAGE)19  and the subsequent modifications of the European Working Group of Clinical Cell Analysis (EWGCCA).20  Daily instrument quality control, including fluorescence standardization, linearity assessment, and spectral compensation, was performed to ensure identical operation from day to day.

Statistical analysis

The cumulative probability of survival was estimated using the Kaplan-Meier method. Comparison between survival curves was performed using the Gehan-Wilcoxon test. Survival analysis was performed using Cox models with time-dependent covariates to assess the effect of the variables of interest on overall survival (OS). Cox regression models were also applied to carry out multivariate survival analyses. Standardized mortality ratios (SMRs) were calculated to compare the patients' mortality with the mortality of the general population in Italy. The Italian mortality rates by age, sex, and calendar year were provided by the Istituto Nazionale di Statistica (ISTAT; Rome, Italy). Statistical analyses were performed using Microsoft Excel 2000 (Redmond, WA), Statistica 7.1 (Stat-Soft, Tulsa, OK), and Stata 9.2 (StataCorp, College Station, TX).

Results

Disease information prior to post-PV MF

A total of 647 patients with PV were evaluated at the 2 institutions between 1970 and 2007. The median interval between the diagnosis of PV and that of post-PV MF was 13 years (range, 2.4-29.6 years). We found that the longer the follow-up of patients with PV, the higher the risk of developing post-PV MF (P < .001). During PV, myelosuppressive agents were given to 65 (96%) of 68 patients who developed post-PV MF and to 501 (86.4%) of 579 patients who did not, while the remaining patients received phlebotomy alone. The rate of patients receiving myelosuppression was significantly higher among those who developed post-PV MF (P = .01). On the other hand, patients receiving myelosuppression had a significantly longer follow-up than those treated with phlebotomy alone (7.1 years and 2.9 years, respectively; P < .001).

To investigate potential risk factors of transformation in post-PV MF present at diagnosis of PV, we evaluated the clinical features at diagnosis in the whole cohort of patients (n = 647). Para-meters taken into account were age, hemoglobin level, platelet count, white blood cell count, spleen size (all considered as continuous numeric variables), leukocytosis (white blood cell count > 15 × 109/L), calendar year at diagnosis, and institutional location. Univariate survival analysis showed that white blood cell count as numeric variable (P < .001) and white blood cell count more than 15 × 109/L (P = .002) were significant risk factors for transformation in post-PV MF.

Clinical features at diagnosis of post-PV MF

Table 1 summarizes clinical and hematologic data at diagnosis of 68 patients with post-PV MF. IWG-MRT criteria and patients' distribution per single criterion are outlined in Table 2. Regarding spleen, one patient underwent splenectomy before diagnosis of post-PV MF. Another patient with no spleen enlargement at diagnosis of post-PV MF showed anemia and leukoerythroblastic peripheral picture in addition to required criteria. Among 47 patients studied for the JAK2 (V617F) mutation at different intervals from diagnosis, all carried the mutation. In 27 patients evaluated at diagnosis, 21 (78%) had more than 50% mutant alleles. In all patients, the number of circulating CD34+ cells and serum lactate dehydrogenase (LDH) level exceeded the upper reference value (10 cells/μL for CD34+ cells and 450 U/L for LDH).

Table 1

Demographic and hematologic characteristics at diagnosis of 68 patients with post–polycythemia vera myelofibrosis

Characteristic No. 
No. of patients 68 
Age at diagnosis, y† (range) 65 (44-81) 
Male/female 45/23 
WBC count, × 109/L,† (range) 12.2 (2.3-98) 
Hemoglobin level, g/L,† (range) 123 (78-148) 
PLT count, × 109/L,† (range) 369 (50-1827) 
Spleen size, cm below left costal margin 7 (0-25) 
Bone marrow findings 
    Marrow cellularity,† % (range) 90 (70-100) 
    Reticulin fibrosis, grade 2:grade 3 3:1 
    MK cluster, loose:dense 0.6:1 
    Marrow myeloblast,† % (range) 0 (0-5) 
Lactate dehydrogenase, U/L,† (range), n = 41 837 (460-3151) 
Circulating CD34+ cells/μL,† (range), n = 39 44.3 (12.1-1005) 
No. JAK2 (V617F) positive, (%), n = 27 27 (100) 
Proportion of JAK2 (V617F) alleles,† % (range) 87 (10-100) 
*Bone marrow karyotype (%) 
    Favorable 20 (80) 
    Unfavorable 5 (20) 
Characteristic No. 
No. of patients 68 
Age at diagnosis, y† (range) 65 (44-81) 
Male/female 45/23 
WBC count, × 109/L,† (range) 12.2 (2.3-98) 
Hemoglobin level, g/L,† (range) 123 (78-148) 
PLT count, × 109/L,† (range) 369 (50-1827) 
Spleen size, cm below left costal margin 7 (0-25) 
Bone marrow findings 
    Marrow cellularity,† % (range) 90 (70-100) 
    Reticulin fibrosis, grade 2:grade 3 3:1 
    MK cluster, loose:dense 0.6:1 
    Marrow myeloblast,† % (range) 0 (0-5) 
Lactate dehydrogenase, U/L,† (range), n = 41 837 (460-3151) 
Circulating CD34+ cells/μL,† (range), n = 39 44.3 (12.1-1005) 
No. JAK2 (V617F) positive, (%), n = 27 27 (100) 
Proportion of JAK2 (V617F) alleles,† % (range) 87 (10-100) 
*Bone marrow karyotype (%) 
    Favorable 20 (80) 
    Unfavorable 5 (20) 

Normal range of circulating CD34+ cells: less than 10/μL; normal range of LDH: less than 450 mU/mL.

*

Favorable indicates normal, 20q−; 13q−23 ; unfavorable, other than favorable.

Data are medians.

Table 2

International Working Group for Myelofibrosis Research and Treatment (IWG-MRT) criteria for the diagnosis of post–polycythemia vera myelofibrosis (post-PV MF) and the distribution of meeting criteria in 68 patients with post-PV MF

 No. of patients (%) 
IWG-MRT required criteria 
    1. Previous diagnosis of polycythemia vera (WHO criteria) 68 (100) 
    2. Bone marrow fibrosis grade 2-3 (on 0-3 scale) 68 (100) 
IWG-MRT additional criteria (2 are required) 
    1. Anemia* or 43 (63) 
        Sustained loss of requirement of phlebotomy or cytoreduction 25 (37) 
    2. Leukoerythroblastic peripheral blood picture 68 (100) 
    3. Increasing splenomegaly 66/67 (98)‡ 
        Palpable spleen more than 5 cm from left costal margin 54 (82) 
        Appearance of a newly palpable splenomegaly 12 (18) 
    4. Development of more than 1 of the constitutional symptoms† 26 (38) 
 No. of patients (%) 
IWG-MRT required criteria 
    1. Previous diagnosis of polycythemia vera (WHO criteria) 68 (100) 
    2. Bone marrow fibrosis grade 2-3 (on 0-3 scale) 68 (100) 
IWG-MRT additional criteria (2 are required) 
    1. Anemia* or 43 (63) 
        Sustained loss of requirement of phlebotomy or cytoreduction 25 (37) 
    2. Leukoerythroblastic peripheral blood picture 68 (100) 
    3. Increasing splenomegaly 66/67 (98)‡ 
        Palpable spleen more than 5 cm from left costal margin 54 (82) 
        Appearance of a newly palpable splenomegaly 12 (18) 
    4. Development of more than 1 of the constitutional symptoms† 26 (38) 
*

Defined as hemoglobin value less than 120 g/L (12 g/dL) for female and less than 135 g/L (13.5 g/dL) for male.

Defined as 10% or more weight loss in 6 months, night sweats, unexplained fever (>37.5°C).

One patient underwent splenectomy before diagnosis, so the calculation was provided on 67 patients.

Disease complications and outcome

Patients with post-PV MF were observed for 181 person-years of follow up after diagnosis and received palliative treatments. During follow-up, the incidence of thrombosis was 42 × 1000 person-years (95% CI: 19-93.5): 3 patients had deep venous thrombosis, 2 had stroke, and 1 had myocardial infarction. Two patients had splenic infarction. The incidence of leukemia was 50 × 1000 person-years (95% CI: 26-115), and the 3-year leukemia-free survival was 82%. Univariate analysis performed on clinical parameters at diagnosis of post-PV MF identified as significant risk factors for leukemia the low platelet count (P = .041) and the high circulating CD34+ cell count (P = .016). In a multivariate Cox proportional hazard regression, only circulating CD34+ cell count retained a significant impact on leukemia-free survival (P = .036).

The median survival of patients with post-PV MF was 5.7 years. The standardized mortality ratio (SMR) was 6.5 (95% CI: 4.2-10.1), indicating a significantly higher mortality for patients with post-PV MF in comparison with the general Italian population matched for age, sex, and calendar year (P < .001). We compared the survival of patients with post-PV MF (mortality: 11.1 per 100 person-years) with the survival of 291 patients with PMF (mortality: 10.1 per 100 person-years). Gehan-Wilcoxon test showed that survival of patients with post-PV MF was not significantly different from that of patients with PMF (P = .32). In addition, after adjustment for white blood cell count, hemoglobin level, platelet count, spleen size, and age in a multivariate Cox proportional hazard regression model, there was no difference in survival between the 2 conditions.

Finally, to evaluate whether transformation to myelofibrosis affects the overall survival of patients with PV, a Cox proportional hazard regression model with transformation to myelofibrosis as time-dependent covariate was applied to the whole series of PV patients. We found that survival of patients with PV was significantly worsened after progression to post-PV MF (HR = 2.17; 95% CI: 1.27-3.72; P = .005). This finding retained statistical significance also after adjustment for age, white blood cell count, hemoglobin level, platelet count, and spleen size in a multivariate Cox proportional hazard regression model.

Prognostic factors at diagnosis of post-PV MF

The parameters we evaluated at diagnosis of post-PV MF to investigate potential predictors of survival were age, hemoglobin level, platelet count, white blood cell count, spleen size, year duration of PV, serum lactate dehydrogenase level, granulocyte JAK2-V617F mutation burden, circulating CD34+ cells (all considered as continuous numeric variables), hemoglobin value less than 100 g/L (10 g/dL),21  white blood cell count less than 4 × 109/L,21  white blood cell count more than 30 × 109/L,21  platelet count less than 100 × 109/L,22  and karyotype23  (according to the categorization in use for PMF). Univariate survival analysis showed that hemoglobin value less than 100 g/L (10 g/dL; P < .001) and circulating CD34+ cell count (P = .009) were significant risk factors for survival. Multivariate Cox regression model including the parameters available in all patients at diagnosis of post-PV MF (hemoglobin value, white blood cell count, platelet count, spleen size, age) indicated that only hemoglobin level less than 100 g/L (10 g/dL) was an independent risk factor for survival (P < .001). Using this hemoglobin level as cutoff, patients could be stratified into 2 risk categories with significantly different survival: 6.6 years for those with hemoglobin value 100 g/L (10 g/dL) or higher and 1.9 years for those with hemoglobin value less than 100 g/L (10 g/dL; P = .0001).

Time-dependent analysis of prognostic factors

Sixty-four patients with post-PV MF had longitudinal blood cell count measurements at regular intervals from diagnosis. We studied this cohort of patients to assess whether variation of hematologic parameters during follow-up may further help in predicting survival at any time from diagnosis. The acquisition of the following parameters was studied: hemoglobin level less than 100 g/L (10 g/dL),21  platelet count less than 100 × 109/L,22  and white blood cell count less than 4 × 109/L or more than 30 × 109/L.21  Modification of therapy was not involved in the acquisition of risk factors. During follow-up of post-PV MF, hemoglobin level dropped lower than 100 g/L (10 g/dL) in 17 (26%) patients, platelet count lower than 100 × 109/L in 23 (36%), and white blood cell count lower than 4 × 109/L in 7 (11%) and higher than 30 × 109/L in 14 (22%).

As a first step, we evaluated univariate survival analysis with Cox regression models using hemoglobin value less than 100 g/L (10 g/dL), platelet count less than 100 × 109/L, white blood cell count less than 4 × 109/L, and white blood cell count more than 30 × 109/L as time-dependent covariates. The HRs were 5.8 (95% CI: 2.2-15.2; P < .001) for hemoglobin, 4.5 (95% CI: 1.67-12, P = .003) for platelets, and 8.2 (95% CI: 3-22; P < .001) for white blood cell count more than 30 × 109/L, while white blood cell count less than 4 × 109/L did not significantly affect survival (P = .115). After adjustment for age in a multivariate Cox proportional hazard regression with time-dependent covariates, hemoglobin value less than 100 g/L (10 g/dL), platelet count less than 100 × 109/L, and white blood cell count more than 30 × 109/L retained statistical significance on survival.

So, we defined a dynamic scoring system based on these 3 independent risk factors. As the 95% CIs of the 3 HRs did not differ, we assigned the same weight (presence = 1; absence = 0) to the 3 factors. As a consequence, the resulting score can be easily calculated by simply counting the number of risk factors acquired at any time during follow-up. The lower risk group includes patients who never acquire risk factors during follow-up (ie, hemoglobin level ≥ 100 g/L [10 g/dL], platelets ≥ 100 × 109/L, and white blood cells < 30 × 109/L). Conversely, higher risk categories include patients with 1, 2, or 3 risk factors, respectively. To assess the impact on survival of this dynamic scoring system, we analyzed the score as a continuous time-dependent covariate in a Cox survival regression model, obtaining an HR of 4.2 (95% CI: 2.4-7.7; P < .001). This implies a 4.2-fold increase of risk when the patient acquires one risk factor at any time from the diagnosis of post-PV MF. The time-dependent prognostic model retained statistical significance after adjustment for age (HR: 6.7, 95% CI: 3-14.7; P < .001). Figure 1 exemplifies the impact of this dynamic prognostic model on survival, showing the estimated survival curves for the resulting 4 risk groups according to the Cox time-dependent model.

Figure 1

Time-dependent survival estimation in post–polycythemia vera myelofibrosis. Survival curves estimated from the Cox proportional hazard regression with time-dependent covariates. According to the model, each patient is initially assigned to a risk group and followed in that group as long as no changes in risk factors take place. The patient is reassigned to another risk group whenever further risk factors are acquired. So, each patient may contribute with some observation time to the estimate of survival in different risk groups. Therefore, the upper curve includes patients who did not acquire any risk factors during the whole follow-up (ie, hemoglobin level ≥ 100 g/L [10 g/dL], platelet count ≥ 100 × 109/L, or white blood cell count < 30 × 109/L). The other curves include patients who acquired 1, 2, or 3 factors during follow-up.

Figure 1

Time-dependent survival estimation in post–polycythemia vera myelofibrosis. Survival curves estimated from the Cox proportional hazard regression with time-dependent covariates. According to the model, each patient is initially assigned to a risk group and followed in that group as long as no changes in risk factors take place. The patient is reassigned to another risk group whenever further risk factors are acquired. So, each patient may contribute with some observation time to the estimate of survival in different risk groups. Therefore, the upper curve includes patients who did not acquire any risk factors during the whole follow-up (ie, hemoglobin level ≥ 100 g/L [10 g/dL], platelet count ≥ 100 × 109/L, or white blood cell count < 30 × 109/L). The other curves include patients who acquired 1, 2, or 3 factors during follow-up.

Discussion

In this study, we evaluated 68 patients who developed post-PV MF in a cohort of 647 patients with PV. Diagnosis of post-PV MF is based on distinctive criteria, recently proposed by the IWG-MRT.1  These criteria combine histopathologic (bone marrow fibrosis), clinical (splenomegaly, constitutional symptoms), and hematologic findings (anemia, leukoerythroblastic peripheral blood picture).

In this series of 647 patients with PV, the analysis of risk factors that may predict transformation to post-PV MF showed that the presence of leukocytosis (white blood cell count > 15 × 109/L) at diagnosis of PV significantly correlates with post-PV MF occurrence. A correlation between leukocytosis and risk of acute leukemia has been recently reported in patients with PV.24  These data indicate that PV patients with leukocytosis are at higher risk of disease evolution. This suggests that in PV patients those with leukocytosis are the most appropriate candidates for clinical trials with JAK2 inhibitors.

A not-yet-defined issue in the natural history of PV concerns whether the development of post-PV MF has adverse prognostic implication on survival.2  Using a Cox proportional hazard regression model, with transformation to post-PV MF as time-dependent covariate, our data indicate a worsening of overall survival of patients with PV after fibrotic transformation.

Studying the 68 patients of this series who developed post-PV MF, we found that all patients with available JAK2 (V617F) status carried the mutation with a high mutational burden. In fact, 78% of patients at diagnosis of post-PV MF had more than 50% mutant alleles, as previously reported.6  In PMF, the rate of homozygosity for JAK2 (V617F) has been recently reported to be 28%.25  Constitutive mobilization of CD34+ cells into peripheral blood represents a frequent phenomenon in post-PV MF,7,18  and all patients tested in this study had high circulating CD34+ cell count. Serum LDH measurement has been recently introduced in the proposed revision of WHO criteria for CMD as additional criterion to diagnose PMF.1  All patients in this series with post-PV MF had high levels of serum LDH, highlighting the clinical utility of this parameter in these patients.

Regarding disease complications of patients with post-PV MF, this study shows that thrombosis remains a relatively frequent complication in PV patients also after transition to myelofibrosis. Leukemia occurs with an incidence of 50 × 1000 person-years, and circulating CD34+ cell count at diagnosis of post-PV MF may predict leukemia-free survival. The frequency of leukemic transformation in patients with post-PV MF seems higher than that reported in patients with PMF.26 

In this study, the median survival of patients with post-PV MF was 5.7 years, slightly lower than that reported in a study including patients with post-PV and post–essential thrombocythemia MF.23  To better stratify patients, we studied potential risk factors for survival at diagnosis of post-PV MF. Using a multivariate Cox proportional hazard regression, we found that an hemoglobin level less than 100 g/L (10 g/dL) is an independent risk factor for survival. In fact, patients with hemoglobin value of 100 g/L (10 g/dL) or more had a median survival of 6.6 years, while those with hemoglobin value less than 100 g/L (10 g/dL) had a median survival of 1.9 years. The cutoff of 100 g/L (10 g/dL) for hemoglobin is also considered useful in the risk stratification of patients with PMF.21-23,27-29  Another common behavior between patients with post-PV MF and those with PMF is survival, which we found to be similar in the 2 conditions. Regarding the adverse impact of unfavorable karyotype on survival reported in a prior study,23  we did not find a significant correlation in our series of patients with post-PV MF. This may probably reflect the small number of patients with unfavorable karyotype or the different patient population.

The course of post-PV MF is a dynamic process during which progressive deterioration of clinical parameters occurs. This may imply the acquisition of additional risk factors. In fact, hemoglobin and platelets progressively tend to decrease, while leukocytes tend either to increase or to decrease. On this ground, we developed a time-dependent scoring system that can be used to predict survival at any time after diagnosis. According to this model, patients are classified into a risk group at diagnosis and remain in the same group until the acquisition of new risk factors. At this time point, patients enter a higher risk category. We provide evidence that a dynamic scoring system based on hemoglobin level less than 100 g/L (10 g/dL), platelet count less than 100 × 109/L, and white blood cell count more than 30 × 109/L is useful to predict survival at any time from diagnosis. In fact, the score predicts a 4.2-fold worsening of survival for each risk factor acquired at any time during follow-up of post-PV MF. The survival curves resulting from this dynamic model have to be interpreted differently from traditional survival curves. In fact, the survival curves of the dynamic model represent an estimated survival as long as the patient remains in the same risk group. A more accurate prediction of survival has potential clinical implications, as these patients are JAK2 mutated and may be candidates to clinical trials with JAK2 inhibitors.

In conclusion, this study demonstrates that patients with PV showing a white blood cell count more than 15 × 109/L at diagnosis have higher risk of developing post-PV MF. When patients with PV develop post-PV MF, a dynamic prognostic model based on hemoglobin level, platelet count, and white blood cell count may predict survival at any time after diagnosis.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Acknowledgments

This work was supported by grants from Fondazione Cariplo, Milan, Italy; Associazione Italiana per la Ricerca sul Cancro (AIRC), Milan, Italy; Fondazione Ferrata Storti, Pavia, Italy; Fondazione IRCCS Policlinico San Matteo, Pavia, Italy; and Ministry of University and Research, Rome, Italy.

Authorship

Contribution: F.P. and M.L. conceived the study, collected, analyzed, and interpreted data, and wrote the paper; E.M. and M. Cazzola analyzed and interpreted data; E.R. collected and analyzed data; M. Caramella, C.E., L.A., and C.D. collected clinical data; E.B. performed bone marrow evaluation; D.P. performed JAK2 mutation analysis; L.V. performed CD34+ cell count; P.B. performed cytogenetic analysis; C.P. did statistical analyses.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Francesco Passamonti, Department of Hematology, Fondazione IRCCS Policlinico San Matteo, Viale Golgi 19, 27100 Pavia, Italy; e-mail: f.passamonti@smatteo.pv.it.

References

References
1
Barosi
G
Mesa
RA
Thiele
J
, et al. 
Proposed criteria for the diagnosis of post-polycythemia vera and post-essential thrombocythemia myelofibrosis: a consensus statement from the international working group for myelofibrosis research and treatment.
Leukemia
 
Prepublished on August 30, 2007 as DOI 10.1038/sj.leu.2604314
2
Spivak
JL
Polycythemia vera: myths, mechanisms, and management.
Blood
2002
, vol. 
100
 (pg. 
4272
-
4290
)
3
Passamonti
F
Rumi
E
Pungolino
E
, et al. 
Life expectancy and prognostic factors for survival in patients with polycythemia vera and essential thrombocythemia.
Am J Med
2004
, vol. 
117
 (pg. 
755
-
761
)
4
Passamonti
F
Malabarba
L
Orlandi
E
, et al. 
Polycythemia vera in young patients: a study on the long-term risk of thrombosis, myelofibrosis and leukemia.
Haematologica
2003
, vol. 
88
 (pg. 
13
-
18
)
5
Tefferi
A
Lasho
TL
Schwager
SM
, et al. 
The JAK2(V617F) tyrosine kinase mutation in myelofibrosis with myeloid metaplasia: lineage specificity and clinical correlates.
Br J Haematol
2005
, vol. 
131
 (pg. 
320
-
328
)
6
Passamonti
F
Rumi
E
Pietra
D
, et al. 
Relation between JAK2 (V617F) mutation status, granulocyte activation, and constitutive mobilization of CD34+ cells into peripheral blood in myeloproliferative disorders.
Blood
2006
, vol. 
107
 (pg. 
3676
-
3682
)
7
Barosi
G
Viarengo
G
Pecci
A
, et al. 
Diagnostic and clinical relevance of the number of circulating CD34(+) cells in myelofibrosis with myeloid metaplasia.
Blood
2001
, vol. 
98
 (pg. 
3249
-
3255
)
8
Arora
B
Sirhan
S
Hoyer
JD
Mesa
RA
Tefferi
A
Peripheral blood CD34 count in myelofibrosis with myeloid metaplasia: a prospective evaluation of prognostic value in 94 patients.
Br J Haematol
2005
, vol. 
128
 (pg. 
42
-
48
)
9
Popat
U
Frost
A
Liu
E
, et al. 
High levels of circulating CD34 cells, dacrocytes, clonal hematopoiesis, and JAK2 mutation differentiate myelofibrosis with myeloid metaplasia from secondary myelofibrosis associated with pulmonary hypertension.
Blood
2006
, vol. 
107
 (pg. 
3486
-
3488
)
10
Tefferi
A
Myelofibrosis with myeloid metaplasia.
N Engl J Med
2000
, vol. 
342
 (pg. 
1255
-
1265
)
11
Deeg
HJ
Gooley
TA
Flowers
ME
, et al. 
Allogeneic hematopoietic stem cell transplantation for myelofibrosis.
Blood
2003
, vol. 
102
 (pg. 
3912
-
3918
)
12
Rondelli
D
Barosi
G
Bacigalupo
A
, et al. 
Allogeneic hematopoietic stem-cell transplantation with reduced-intensity conditioning in intermediate- or high-risk patients with myelofibrosis with myeloid metaplasia.
Blood
2005
, vol. 
105
 (pg. 
4115
-
4119
)
13
Pardanani
A
JAK2 inhibitor therapy in myeloproliferative disorders: rationale, preclinical studies and ongoing clinical trials.
Leukemia
2008
, vol. 
22
 (pg. 
23
-
30
)
14
Berk
PD
Goldberg
JD
Donovan
PB
Fruchtman
SM
Berlin
NI
Wasserman
LR
Therapeutic recommendations in polycythemia vera based on Polycythemia Vera Study Group protocols.
Semin Hematol
1986
, vol. 
23
 (pg. 
132
-
143
)
15
Vardiman
JW
Harris
NL
Brunning
RD
The World Health Organization (WHO) classification of the myeloid neoplasms.
Blood
2002
, vol. 
100
 (pg. 
2292
-
2302
)
16
Barosi
G
Ambrosetti
A
Finelli
C
, et al. 
The Italian Consensus Conference on Diagnostic Criteria for Myelofibrosis with Myeloid Metaplasia.
Br J Haematol
1999
, vol. 
104
 (pg. 
730
-
737
)
17
Rumi
E
Passamonti
F
Pietra
D
, et al. 
JAK2 (V617F) as an acquired somatic mutation and a secondary genetic event associated with disease progression in familial myeloproliferative disorders.
Cancer
2006
, vol. 
107
 (pg. 
2206
-
2211
)
18
Passamonti
F
Vanelli
L
Malabarba
L
, et al. 
Clinical utility of the absolute number of circulating CD34-positive cells in patients with chronic myeloproliferative disorders.
Haematologica
2003
, vol. 
88
 (pg. 
1123
-
1129
)
19
Keeney
M
Chin-Yee
I
Weir
K
Popma
J
Nayar
R
Sutherland
DR
Single platform flow cytometric absolute CD34+ cell counts based on the ISHAGE guidelines: International Society of Hematotherapy and Graft Engineering.
Cytometry
1998
, vol. 
34
 (pg. 
61
-
70
)
20
Brando
B
Barnett
D
Janossy
G
, et al. 
Cytofluorometric methods for assessing absolute numbers of cell subsets in blood: European Working Group on Clinical Cell Analysis.
Cytometry
2000
, vol. 
42
 (pg. 
327
-
346
)
21
Dupriez
B
Morel
P
Demory
JL
, et al. 
Prognostic factors in agnogenic myeloid metaplasia: a report on 195 cases with a new scoring system.
Blood
1996
, vol. 
88
 (pg. 
1013
-
1018
)
22
Dingli
D
Schwager
SM
Mesa
RA
Li
CY
Tefferi
A
Prognosis in transplant-eligible patients with agnogenic myeloid metaplasia: a simple CBC-based scoring system.
Cancer
2006
, vol. 
106
 (pg. 
623
-
630
)
23
Dingli
D
Schwager
SM
Mesa
RA
Li
CY
Dewald
GW
Tefferi
A
Presence of unfavorable cytogenetic abnormalities is the strongest predictor of poor survival in secondary myelofibrosis.
Cancer
2006
, vol. 
106
 (pg. 
1985
-
1989
)
24
Gangat
N
Strand
J
Li
CY
Wu
W
Pardanani
A
Tefferi
A
Leucocytosis in polycythaemia vera predicts both inferior survival and leukaemic transformation.
Br J Haematol
2007
, vol. 
138
 (pg. 
354
-
358
)
25
Barosi
G
Bergamaschi
G
Marchetti
M
, et al. 
JAK2 V617F mutational status predicts progression to large splenomegaly and leukemic transformation in primary myelofibrosis.
Blood
2007
, vol. 
110
 (pg. 
4030
-
4036
)
26
Mesa
RA
Li
CY
Ketterling
RP
Schroeder
GS
Knudson
RA
Tefferi
A
Leukemic transformation in myelofibrosis with myeloid metaplasia: a single-institution experience with 91 cases.
Blood
2005
, vol. 
105
 (pg. 
973
-
977
)
27
Reilly
JT
Snowden
JA
Spearing
RL
, et al. 
Cytogenetic abnormalities and their prognostic significance in idiopathic myelofibrosis: a study of 106 cases.
Br J Haematol
1997
, vol. 
98
 (pg. 
96
-
102
)
28
Cervantes
F
Pereira
A
Esteve
J
, et al. 
Identification of ‘short-lived’ and ‘long-lived’ patients at presentation of idiopathic myelofibrosis.
Br J Haematol
1997
, vol. 
97
 (pg. 
635
-
640
)
29
Visani
G
Finelli
C
Castelli
U
, et al. 
Myelofibrosis with myeloid metaplasia: clinical and haematological parameters predicting survival in a series of 133 patients.
Br J Haematol
1990
, vol. 
75
 (pg. 
4
-
9
)