Myelofibrosis With Myeloid Metaplasia: Diagnosis, Prognostic Factors, and Staging

Myelofibrosis With Myeloid Metaplasia: Diagnosis, Prognostic Factors, and Staging Francisco Cervantesa and Giovanni Barosib Myelofibrosis with myeloid metaplasia (MMM) is an infrequent disease usually affecting elderly people. Due to the lack of a specific marker for MMM, efforts are being made to refine the diagnostic criteria. A prefibrotic form of the disease has recently been recognized and a set of diagnostic criteria proposed by the Italian Consensus Conference for the Diagnostic Criteria of MMM. Moreover, the number of circulating CD34ⴙ cells has recently been proven to be useful in the differential diagnosis of the disease. Median survival of patients with MMM is about 5 years, but there is wide variability. Hemoglobin level at diagnosis is the most important prognostic factor, whereas age, constitutional symptoms, low or high leukocyte counts, blood blast cells, cytogenetic abnormalities, and number of circulating CD34ⴙ cells are also of prognostic value. Based on some of the above factors, several prognostic systems of MMM have been proposed to identify at presentation subgroups of patients with a different risk profile. This is especially important in younger individuals, who may benefit from therapies that have curative potential but also involve a mortality risk (notably, allogeneic hemopoietic stem cell transplantation). Since no disease-oriented therapies for MMM are currently available and treatment is based on palliation and improvement of quality of life, therapy should be formulated according to the form of presentation of the disease and its prognostic implications. A staging system for MMM that meets treatment requirements and available therapeutic resources should be developed by considering parameters able to predict not only survival but also other disease outcomes, such as development of anemia, thrombocytopenia, splenomegaly, and blast transformation. Such a staging system should be derived from a large population of patients, systematically collected from diagnosis without selection bias and followed over time. Semin Oncol 32:395-402 © 2005 Elsevier Inc. All rights reserved.

M

yelofibrosis with myeloid metaplasia (MMM), also known as idiopathic myelofibrosis or agnogenic myeloid metaplasia, is a rare disease, with an estimated incidence in the Western countries of 0.4 to 0.7 new cases per 100,000 persons per year according to two epidemiologic studies1,2 or somewhat higher according to another.3 Like the other chronic myeloproliferative disorders (CMDs), MMM is a clonal disease arising in a pluripotent hemopoietic stem cell that, due to a somatic mutation, acquires a proliferative advantage over the benign hemopoietic progenitors.4-6 The resulting abnormal cell population releases several cytokines and growth factors in the bone aHematology bLaboratory

Department, Hospital Clínic, IDIBAPS, Barcelona, Spain. of Clinical Epidemiology, IRCCS Policlinico S. Matteo, Pavia,

Italy. Address correspondence to Giovanni Barosi, MD, Laboratorio di Epidemiologia Clinica, IRCCS Policlinico S. Matteo, Viale Golgi 19, 27100 Pavia, Italy. E-mail: [email protected]

0093-7754/05/$-see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1053/j.seminoncol.2005.04.011

marrow, leading to the appearance of the marrow fibrosis as a secondary phenomenon,4,7 and through the blood stream colonizes extramedullary organs such as the spleen and the liver. MMM is also characterized by progressive anemia, with tear-drop cells and myeloid and erythroid precursors in the peripheral blood. Due to the lack of a specific marker for the disease, the diagnosis of MMM remains one of exclusion and, therefore, efforts are being made to refine the diagnostic criteria of the disease so they may be uniformly applied in clinical practice. In addition, there is an increasing interest in the identification of the prognostic factors of MMM, and several proposals for prognostic stratification of patients have been published in recent years in attempts to create a common tool that could be of help in the design and evaluation of newer treatment strategies. The present report summarizes the current status of the diagnostic criteria of MMM, as well as the available knowl395

F. Cervantes and G. Barosi

396

Table 1 Updated Cologne Clinicopathologic Criteria for the Diagnosis of Myelofibrosis With Myeloid Metaplasia Clinical Criteria A1 No preceding or allied subtype of myeloproliferative disorders, chronic myeloid leukemia or myelodysplastic syndrome A2 Early clinical stages —Normal hemoglobin or anemia grade 1: hemoglobin >12 g/dL —Slight or moderate splenomegaly on palpation or >11 cm on ultrasound scan or computed tomography —Platelets >400 ⴛ 109/L A3 Intermediate clinical stage —Anemia grade II: hemoglobin >10 g/dL —Definitive leuko-erythroblastic blood picture and/or tear-drop erythrocytes —Splenomegaly —No adverse signs* A4 Advanced clinical stage —Anemia grade III: hemoglobin <10 g/dL —1 or more adverse signs*

Pathological Criteria B1 Megakaryocytic and granulocytic myeloproliferation and relative reduction of erythroid precursors Abnormal clustering and increase in atypical giant-sized megakaryocytes containing clumsy (cloud-like) lobulated nuclei and definitive maturation defects Staging of idiopathic myelofibrosis (IMF) MF0 prefibrotic stage IMF: no reticulin fibrosis MF1 early IMF: slight reticulin fibrosis MF2 manifest IMF: marked increase in reticulin and/or collagen fibrosis MF3 overt IMF: advanced collagen fibrosis-osteosclerosis (endophytic bone formation)

NOTE. The combination of A1 ⴙ B1 establishes IMF; any other criterion confirms IMF. A1 ⴙ A2, B1 ⴙ MF0 is consistent with initial (prefibrotic) IMF. A1 ⴙ A3, B1 ⴙ MF1, MF2 is consistent with early manifestation of IMF. A1 ⴙ A4, B1 ⴙ MF3 is consistent with end stage (full-blown) IMF. *Adverse signs: age >70 years, hemoglobin <10 g/dL, myeloblasts >2% in peripheral blood, >2% erythroblasts in peripheral blood, leukocytosis >20 ⴛ 109/L, platelet counts < 300 ⴛ 109/L, severe constitutional symptoms, massive splenomegaly, cytogenetic abnormalities.

edge on the prognostic factors and prognostic classification of patients with this disease.

Diagnosis If analyzed singly, there is no biological, clinical, or pathologic characteristic of the disease that can be considered absolutely specific for MMM. Moreover, the characteristic lineage proliferation that allows other CMDs to be categorized with simple diagnostic criteria is lacking in MMM. Establishing uniform diagnostic criteria is a prerequisite to standardizing the conduct and reporting of clinical studies and is also of help for practitioners in their clinical practice. This justifies the efforts of researchers in the field to develop diagnostic criteria that can be widely accepted and utilized. The Polycythemia Vera Study Group (PVSG) first formalized a diagnostic procedure for MMM. In 1975, Laszlo8 set out the criteria for enrolling patients in cooperative group studies on MMM. These were: myelofibrosis involving more than one third of the sectional area of a bone marrow biopsy, a leukoerythroblastic blood picture, splenomegaly, and the absence of well-established diagnostic criteria for other CMDs (ie, the absence of increased red cell mass excluding polycythemia vera and the Philadelphia chromosome [Ph] excluding chronic myeloid leukemia). Systemic disorders were also excluded. In spite of the fact that the PVSG was a highly respected group, subsequent authors practically ignored the entire formulation of the definition of MMM as proposed by Laszlo. A literature analysis of the last 30 years aimed at collecting the

diagnostic criteria that the authors used for including patients in clinical studies on MMM showed very diverging results. Bone marrow fibrosis, leukoerythroblastic blood picture, and splenomegaly were necessary diagnostic criteria in only 97.2%, 81.5%, and 83.1%, respectively, of the 71 publications examined (personal observation). Moreover, new criteria were explicitly considered, such as the presence of anemia, myeloid metaplasia, abnormal platelet function, hepatomegaly, clusters of megakaryocytes, and megakaryoblasts in the bone marrow. The absence of monocytosis was also used to distinguish MMM from chronic myelomonocytic leukemia. Judging the set of criteria proposed by the PVSG to be insufficient, two clinical research projects were launched recently in Europe. The first originated from the strong hypothesis of a prefibrotic phase of the disease and includes megakaryocyte morphology, featured by the characteristic shape with plump lobulation of nuclei and asynchronous nuclear cytoplasm maturation, which is characteristic of the bone marrow biopsy in MMM . In the Cologne criteria, the authors9,10 proposed a diagnostic classification of MMM in which the presence of splenomegaly, leukoerythroblastic picture of peripheral blood, anemia, thrombocytosis, and clustering of megakaryocytes in bone marrow were used to define the disease (Table 1). These criteria have been accepted by the World Health Organization classification of MMM.11 The Italian Consensus Conference for the Diagnostic Criteria of MMM originated from the Italian Cooperative Group

Diagnosis, prognostic factors, and staging of MMM

397

Table 2 The Italian Criteria for the Diagnosis of Myelofibrosis With Myeloid Metaplasia Necessary Criteria A. Diffuse bone marrow fibrosis

B. Absence of Ph chromosome or BCR-ABL rearrangement in peripheral blood cells Optional Criteria

1. Splenomegaly of any grade 2. Anisopoikilocytosis with teardrop erythrocytes 3. Presence of circulating immature myeloid cells Diagnosis of MMM is acceptable if the following combinations are present: ● The 2 necessary criteria plus any other 2 optional criteria when splenomegaly is present ● The 2 necessary criteria plus any other 4 optional criteria when splenomegaly is absent

4. Presence of circulating erythroblasts 5. Presence of clusters of megakaryoblasts and anomalous megakaryocytes in bone marrow sections 6. Myeloid metaplasia

on Myeloproliferative Disorders,12 with the purpose of developing a definition of MMM according to the paradigm of “evidence-based medicine” using the “consensus” methodology. The results of the literature review were offered to a panel of 12 Italian experts and, through a questionnaire, they ranked the necessary and optional diagnostic criteria according to their preferences in order to identify a core set of criteria. Necessary criteria consisted of “diffuse bone marrow fibrosis” and “absence of Ph chromosome or BCR-ABL rearrangement in peripheral blood cells.” The six optional criteria in the core set consisted of splenomegaly of any grade, anisopoikilocytosis with tear-drop erythrocytes, presence of circulating immature myeloid cells, presence of circulating erythroblasts, presence of clusters of megakaryoblasts and anomalous megakaryocytes in bone marrow sections, and myeloid metaplasia. The diagnostic definition of the disease was as follows: obligatory presence of diffuse bone marrow fibrosis and absence of Ph chromosome or BCR-ABL rearrangement in peripheral blood cells, presence of two other of the core set criteria when splenomegaly is present, and presence of four of the core set criteria when splenomegaly is absent (Table 2). In the absence of biological markers for the specific diagnosis of the disease, the evaluation of the sensitivity and specificity of any proposed diagnostic criteria may only be assessed empirically. Among the cases reported to the community-based Italian Registry of MMM,13 which has been collecting incident cases of MMM around Italy since 1999, the false-negative rate of the Italian criteria is 2.1% (patients in whom the diagnostic criteria were not met but were assessed as having MMM). In these few cases, the limiting constraint for a diagnosis of MMM was the necessary criterion of the “presence of diffuse bone marrow fibrosis,” as they presented with none or minimal reticulin fibrosis in the bone marrow or with a fatty bone marrow in which the fibrosis was patchy and localized, representing a previously described pathologic variant.14 The rate of false-negative cases using the Cologne criteria was 3.5%. These were cases with no clustering of megakaryocytes in the bone marrow or no splenomegaly at diagnosis. In conclusion, neither a biological hypothesis on the natural history of the disease nor a nominalistic approach pro-

duce a set of criteria for the diagnosis of MMM that, when compared with the diagnostic skill of expert physicians, is infallible. This fact, however, does not preclude the use of such sets of criteria for the enrollment of MMM patients into clinical trials. However, the existence of atypical variants, like those with “prefibrotic myelofibrosis”15 or “myelofibrosis with fatty bone marrow,”14 may encompass new diagnostic features that may accommodate histologic and clinical heterogeneity. One biological feature, the number of circulating CD34⫹ hematopoietic progenitor cells, has recently called the traditional diagnostic criteria of MMM into question. CD34 is a surface antigen present on 1% to 3% of human bone marrow cells and on 0.05% of nucleated circulating cells. It serves as a marker for identifying and separating hematopoietic stem and progenitor cells because it is not found on fully differentiated, or mature, hematopoietic cells. Recently, it has been reported that the median number of CD34⫹ cells in peripheral blood in a large population of patients with MMM is 360 times higher than in a normal population and 18 to 30 times higher than in a selected population of patients with other Ph⫺ CMD.16 CD34⫹ cell number allowed an almost complete discrimination between MMM and other Ph⫺ CMD, when samples were taken from patients out of cytoreductive therapy. In fact, in the latter group of disorders the value of CD34⫹ circulating cells was only slightly increased. By using receiver operating characteristic (ROC) analysis, generated by calculating the sensitivities and specificities of data at several predetermined cut-off points,17 a threshold of 15 ⫻ 106/L CD34⫹ cells could discriminate MMM patients out of cytoreductive treatment from patients with other Ph⫺ CMD with a 98.4% positive and an 85% negative predictive value.16 Since 26% of the samples from MMM patients were obtained at diagnosis, a consistently increased number of circulating CD34⫹ cells could be considered a valuable indicator of the presence of MMM.16

Natural History and Survival Median survival of patients with MMM ranges from 3.5 to 5.5 years in modern series,18-27 with this disease currently being the CMD associated with the poorest prognosis. Figure 1

398

Figure 1 Actuarial survival curve of 161 patients diagnosed with MMM at the Hospital Clínic of Barcelona, Spain, during a 30-year period.

shows the survival curve of the series of 161 patients diagnosed with MMM during a 30-year period at the Hospital Clínic of Barcelona, Spain, which does not include cases of post-polycythemic and post-thrombocythemic myelofibrosis. As can be seen, the median survival of the series was 4.9 years, with a survival rate of 74% at 2 years and of 33% at 8 years. The shape of the curve reflects the wide variability in the survival of MMM patients, since a small proportion of them die within 1 year of presentation, whereas a few survive for more than two decades. When the survival of MMM patients has been compared with that of age- and sex-matched individuals from the general population, a 31% reduction in life expectancy has been observed.28 It has been suggested that in recent years there has been an increasing trend towards earlier diagnosis of MMM, which can be attributed to the generalized practice of routine laboratory studies.29 However, comparison of the survival of patients diagnosed in the same institution during different time periods indicates that such possible earlier diagnosis would not translate into longer survival.29 The main causes of death in MMM are transformation into acute leukemia (observed in 20% of cases at 10 years of diagnosis and usually of myeloid phenotype), infection and bleeding secondary to progressive bone marrow failure, portal hypertension or hepatic failure secondary to hepatic/splenoportal vein thrombosis or myeloid metaplasia of the liver, thromboses in other territories, and heart failure.23,25,30 With regard to young individuals with MMM, until recently the information on the proportion that they represented in the general MMM population, as well as on their characteristics and survival, was scarce and based on anecdotal case reports. As more intensive therapeutic options are becoming available for MMM, the interest in a better knowledge of the characteristics and prognosis of this patient subpopulation has increased, taking into account that it represents the natural target for some of the new treatment modalities, notably allogeneic hematopoietic stem cell transplantation.31,32 Morel et al,33 analyzing the subgroup of younger patients from their previously reported MMM series, observed a less aggressive course of the disease in individuals younger than 55 years of age at diagnosis. Thus, the median

F. Cervantes and G. Barosi survival of younger patients with features of low-risk MMM was 13 years, ie, significantly longer than the survival of low-risk MMM patients overall. In the same direction, a survey of 121 patients who were 55 years old or younger at presentation from four European institutions34 found that they represented 22% of all cases of MMM and that their prognosis was substantially better than that of the general MMM population, since their median survival was 128 months or more than twice the survival overall of patients with the disease. MMM seldom affects persons under the age of 30.35 It must be remarked that some of the early reports on such patients probably corresponded to entities other than MMM, since the blood and marrow abnormalities were usually observed in the setting of infection, collagen vascular disease, renal failure, trisomy 21, or preleukemic syndromes.36 It is therefore possible that the above cases would not be considered as MMM according to the modern diagnostic criteria for the disease. On the contrary, in young individuals with true MMM the clinicohematologic picture was similar to that of the adult form of the disease.36 In a recent survey of 323 cases of MMM collected in two European institutions, nine patients younger than 30 years of age were identified, representing less than 3% of the total.35 In all cases, the disease presented without adverse prognostic factors and remained stable for years. Thus, with a median follow-up approaching 7 years, only two patients had died, at 10.7 and 9.9 years from diagnosis, while the remaining five were asymptomatic and without need for treatment at their last visit. The analysis of the characteristics of these very young patients showed that during the follow-up they all developed anemia and that splenomegaly invariably appeared in those who did not have a palpable spleen at disease presentation. The above observations would support the notion that the majority of young patients with MMM have the classical form of the disease that would have been, however, diagnosed at an earlier stage. Finally, the survival of patients in whom myelofibrosis develops following essential thrombocythemia does not seem to differ from that of individuals diagnosed de novo with MMM.37 On the contrary, the prognosis of patients with postpolycythemic myelofibrosis appears to be more unfavorable since, according to some studies, their median survival, once myelofibrosis is diagnosed, would be about 1.5 years.38,39

Prognostic Factors and Staging Systems The wide variability in the survival of patients with MMM often poses difficulties in making therapeutic decisions, as well as evaluating the efficacy of new treatment modalities for this disease. This fact has stimulated interest for the identification of those variables that, when present at diagnosis, allow prediction of patient’s survival. Table 3 summarizes the results of the main studies on prognostic factors of MMM published in the medical literature in which multivariate regression methods were employed. In addition, to allow appropriate comparison, those series including patients with

Diagnosis, prognostic factors, and staging of MMM

399

Table 3 Adverse Prognostic Factors in MMM Series Including More Than 100 Patients and Using Multiple Regression Analyses First Author Barosi Visani Dupriez Cervantes25 Reilly26 Tefferi40 Sagaster48 (n ⴝ 137) (n ⴝ 133) (n ⴝ 195) (n ⴝ 106) (n ⴝ 106) (n ⴝ 165) (n ⴝ 110) 19

Feature Age Gender Constitutional symptoms Spleen or liver size Hemoglobin <10 g/dL WBC count % blood myeloid precursors Blood blasts Thrombocytopenia Abnormal karyotype No. of circulating progenitors

21

ⴙ ⴚ

ⴚ ⴚ

ⴚ ⴙ ⴚ ⴙ

ⴚ ⴙ ⴙ* ⴙ

ⴚ

ⴚ

23

ⴙ ⴙ ⴙ ⴚ ⴙ ⴙ† ⴚ ⴙ ⴚ ⴙ

ⴚ ⴚ ⴙ ⴚ ⴙ ⴚ ⴚ ⴙ ⴚ

ⴚ ⴚ

ⴙ ⴚ

ⴙ ⴚ

ⴚ ⴙ ⴙ*

ⴚ ⴙ ⴚ ⴚ

ⴚ ⴙ ⴚ ⴚ ⴚ ⴚ

ⴚ ⴙ ⴙ

ⴚ ⴙ‡

ⴙ

*High. †Low or high. ‡Only ⴙ8 and 12p-.

the so-called prefibrotic form of MMM have not been considered. As can be seen, there are often discrepancies with regard to the prognostic significance of some of the features analyzed, whereas there is universal agreement on the importance of the degree of anemia in the prognosis of MMM. Among the clinical factors assessed for prognostic significance in MMM, an association between the presence of hypermetabolic or constitutional symptoms (ie, fever, night sweats, and weight loss) and shorter survival has been found in all studies in which it has been analyzed.18,22,23,25 Age at presentation is another variable that has been linked to the prognosis of MMM in several studies.19,21,22,25,26,28,40 As already mentioned, younger subjects with MMM survive significantly longer than older patients.34 However, although the latter statement applies to the individuals with MMM who are 55 or younger at diagnosis, the prognostic influence of having a more advanced age is less clear. It must be pointed out that at multivariate analysis, the poorer prognosis of older patients was maintained in only two studies,19,28 whereas it was not in the seen in the majority.21,23,25,26 Male sex was associated with worse prognosis in the univariate analysis of the Lille series, but it lost its prognostic significance at the multivariate study.23 A prognostic value for spleen size has been noted only occasionally,20 and the same applies to the presence of hepatomegaly.23 Initial haemoglobin concentration is the most consistent hematologic parameter associated with the prognosis of MMM, with 10 g/dL being the cut-off level usually determining a more unfavourable course of the disease.18,26,40,41 In one study, a poor prognostic influence of low reticulocyte counts was also observed.41 Low leukocyte count was a poor prognostic indicator in the Lille series,23 whereas high leukocyte count conferred poor prognosis in Lille and many other studies.21,23,26,28 According to haemoglobin level and leukocyte count, Dupriez et al23 proposed a score (the Lille Scoring System) that was able to identify three distinct prognostic groups (Table 4). In the low-risk group (hemoglobin ⬎10 g/dL and white

blood cell count [WBC] between 4 ⫻ 109/L and 30 ⫻ 109/L), patients had a median survival of 93 months, whereas those in the intermediate-risk (hemoglobin ⬍10 g/dL or WBC ⬍4 ⫻ 109/L or ⬎30 ⫻ 109/L) and high-risk (hemoglobin ⬍10g/dL and WBC ⬍4 ⫻ 109/L or ⬎30 ⫻ 109/L) groups had median survivals of 26 and 13 months, respectively. This prognostic score is very attractive due to its simplicity and discriminant power and has been widely used in stratifying patients for the analysis of clinical trials outcomes. A high proportion of immature myeloid precursors in peripheral blood had poor prognostic value in the series of Barosi et al19 and Visani et al,21 whereas the presence of circulating blasts in peripheral blood at MMM presentation conferred a poor prognosis in three other studies.23,25,27 Thrombocytopenia has also been associated with poorer prognosis in some studies.18,26,28 The relationship between the course of MMM and the results of ferrokinetic studies or the measurement of red blood cell and plasma volume has also been analyzed. Barosi et al19 found that patients showing features of decreased erythropoiesis had shorter survival, a correlation that was observed in two other studies.41,42 On the other hand, bone marrow histologic findings did not have prognostic relevance in several studies.18,25,43 However, rather surprisingly, in one study a longer survival was registered for patients displaying osteosclerosis in the initial marrow biopsy.20 It was recently re-

Table 4 Prognostic Classification According to Dupriez et al23 (Lille), Predicting Survival for MMM Patients Lille System Hemoglobin <10 g/dL Leukocytes <4,000/␮L or >30,000/␮L Their presence scores 1 point Median survival Score 0: 93 mo Score 1: 26 mo Score 2: 13 mo

400 ported that angiogenesis confers a poor prognosis in MMM.44 Finally, a pattern of red blood cell aplasia in a cellular bone marrow has been associated with a more unfavorable outcome in MMM.19 An abnormal karyotype has been linked to shorter survival of patients with MMM in two series,26,45 which is in contrast with the results of an earlier report by Miller et al.46 Actually, such an unfavorable influence could be indirectly inferred from the results of the study by Besa et al.47 More recently, Dupriez et al23 demonstrated the independent prognostic value of cytogenetic abnormalities, since this feature retained its unfavorable influence even in the subgroup of patients with low-risk MMM. Karyotypic abnormalities also conferred an adverse prognosis in the series by Reilly et al,26 in contrast to the findings of Tefferi et al,40 when all abnormalities were considered as a whole. In the latter study, the negative influence of chromosome changes was restricted to the presence of trisomy 8 and deletion of 12p, while deletions of 13q and 20q did not involve a shorter survival. It cannot be excluded that the lack of prognostic significance of cytogenetic abnormalities observed in other studies could be due to the low proportion of patients with available metaphases. A higher number of circulating CD34⫹ cells was recently associated with longer duration of disease.16 A correlation was also noted between the above parameter and an index that scored the myeloproliferative activity of the disease. With regard to prognosis, a clear correlation was found between circulating CD34⫹ cell count and risk groups of MMM, with a higher number of such cells indicating a more unfavorable patient risk group.16 The most important prognostic implication of the number of CD34⫹ cells in peripheral blood would be its strong correlation with the probability of transformation to acute leukemia. Thus, patients with MMM presenting with more than 300 ⫻ 106/L CD34⫹ cells in blood have a 50% probability of developing acute leukemia by 11 months from diagnosis.16 The adverse prognostic influence of an increased number of circulating colony-forming units granulocyte-macrophage (CFU-GM) and burst-forming units erythroid (BFU-E) hemopoietic progenitors was also recently reported.48. In individuals with MMM younger than 55 years, the factors associated with poor prognosis are anemia (hemoglobin ⬍10 g/dL), constitutional symptoms, and presence of blast cells in peripheral blood.34 Of note, in these younger patients no association has been found between presence of cytogenetic abnormalities and survival, although it must be stressed that only 17% of them displayed such a finding.34 Based on the above-mentioned three variables, two prognostic groups could be clearly identified among these younger patients: a “low-risk” group, including patients with none or one of the bad prognostic factors, and a “high-risk” group, defined by patients with two or three unfavorable factors (Fig 2). The “low-risk” group encompassed three quarters of the patients overall and median survival approached 15 years, whereas the “high-risk” group included a quarter of the patients, with a median survival of less than 3 years. This prognostic scoring system showed a high positive predictive value, sensitivity, and specificity in predicting survival in the series from which

F. Cervantes and G. Barosi

Figure 2 Actuarial survival curves of 121 patients ⱕ55 years old with “low-risk” and “high-risk” MMM.

it was derived and it is increasingly being used at the time of making treatment decisions in young patients with MMM. Beside the number of circulating CD34⫹ cells, prognostic factors for acute transformation of MMM include severe anemia, a high proportion of immature myeloid cells in peripheral blood, and a clinical picture of erythroid failure.49 An increased risk of evolution into blast transformation has been reported in MMM patients who have undergone splenectomy.30 In a large study of 223 splenectomized patients, the rate of blast transformation after splenectomy was 16.3%, and the risk was higher in the presence of increased spleen mass and preoperative thrombocytopenia.50 This could suggest that pre-splenectomy thrombocytopenia in MMM may be a surrogate for advanced disease and is associated with an increased risk of blast transformation.

Towards a Definitive Staging System for MMM Dividing the natural history of a disease in stages of different severity is a necessary operational research tool in order to stratify the actual survival expectancy, to provide a correlation between treatment requirements and available therapeutic resources, and to identify unusual disease presentations that may prove to be clinically relevant. A direct estimate of prognosis, based on easily measurable parameters at diagnosis, is the most productive method for staging a disease. This, however, is elusive in MMM, in which the phenotypic complexity may take the form of longitudinal diversity, by which in the individual patient the disease may vary over time, or of group diversity, by which the disorder may vary in different patients over time. Moreover, at present, there are no diseaseoriented therapies for MMM, but only therapies focused on palliation and improvement of quality of life. Thus, therapy should be formulated according to the actual disease presentation and to the expectations of each type of presentation. The disease phenotypic diversity is not captured by prognostic parameters singled out at diagnosis, only by their capacity to predict survival. For example, the predictive power for the development of massive splenomegaly (⬎10 cm below the

Diagnosis, prognostic factors, and staging of MMM costal margin in 2 years from diagnosis) is not different between patients with the prognostic risk class 0 according to the Lille risk classification,23 and with the prognostic risk class 1 or 2 (29.5% v 33%, respectively) (personal observation). Thus, a staging system that meets treatment requirements and available therapeutic resources should be developed, giving consideration to parameters that also predict important intermediate outcomes of the disease, like development of anemia, splenomegaly, blast transformation, and thrombocytopenia. This seems even more important after the newly proposed World Health Organization disease definition, which includes patients with a prefibrotic stage of the disease.11 These patients have been rarely considered for inclusion in the previously published prognostic classifications, thus possibly producing a selection bias against the very early, indolent cases. One key element for tagging the staging evolution of MMM is the dynamics of bone marrow fibrosis. A staging system was proposed by Thiele et al9 in which the degree of bone marrow fibrosis, in association with megakaryocytic clustering in bone marrow, was used to define the progression of the disease in terms of the extent of splenomegaly and the appearance of anemia. In this case series, one fourth of the patients were diagnosed in a prefibrotic stage of the disease, which evolved towards mild to intense bone marrow fibrosis in 84%.51. A steady progression of myelofibrosis from a prefibrotic phase to fully developed myelofibrosis was also documented in sequential bone marrow biopsy specimens from pathologists’ case series.52 In a retrospective analysis of 100 patients with MMM, we were able to document that the predictive power for the evolution toward anemia (hemoglobin ⬍10 g/dL in 2 years from diagnosis) was 100% for having no or mild bone marrow fibrosis (personal observation). Another key parameter for staging the natural history of the disease is the number of CD34⫹ cells in peripheral blood. The value of CD34⫹ cells as a marker of the progression of MMM is supported by the result that the number of CD34⫹ cells correlates with the disease duration, with the extent of splenomegaly, and with the evolution towards blast transformation of the disease.16 The available results allow delineation of a model of evolution of the disease in which the tendency is to produce more and more displacement of CD34⫹ cells from bone marrow to peripheral blood. The egress of CD34⫹ cells from the bone marrow correlates with the displacement of progenitor cells to extramedullary sites, primarily the spleen and splenomegaly. A CD34 count greater than 300 ⫻ 106/L is associated with an accelerated phase of the disease and predicts evolution towards blast transformation.16 A valid staging system of MMM will be possible only with a large population of patients, systematically collected from the time of diagnosis without selection bias, and followed over time. This is lacking today, and it is not possible to provide a fully developed staging system for MMM. However, what is emerging from the literature is the existence of a class of patients with a very low tendency to evolve towards severe anemia or large splenomegaly and characterized by the lack of anemia, a low number of CD34⫹ cells in peripheral blood,

401 absence or low degree of bone marrow fibrosis, normal WBC, and normal or slightly high platelet count. This “smoldering MMM” probably encompasses what has been called “prefibrotic idiopathic myelofibrosis”15,51 or “atypical myeloproliferative disorder.”53 The hallmarks of this disease variant are young age, low number of immature myeloid cells without blasts in the peripheral blood, and absent or modest splenomegaly. Among the clinical features, a history of thromboembolic episodes, mainly in the splanchnic system, is remarkable. The importance of recognition of this variant is to provide guidance on therapy; these patients may be reassured of the slow evolution of their disease and should not be treated immediately.

References 1. McNally RJ, Rowland D, Roman E, et al: Age and sex distributions of hematological malignancies in the U.K. Hematol Oncol 15:173-189, 1997 2. Kutti J, Ridell B: Epidemiology of the myeloproliferative disorders: Essential thrombocythemia, polycythemia vera and idiopathic myelofibrosis. Pathol Biol (Paris) 49:164-166, 2001 3. Mesa RA, Silverstein MN, Jacobsen SJ, et al: Population based incidence and survival figures in essential thrombocythemia and agnogenic myeloid metaplasia: An Olmsted County study, 1975-1995. Am J Hematol 61:10-15, 1999 4. Jacobson RJ, Salo A, Fialkow PJ: Agnogenic myeloid metaplasia: A clonal proliferation of hematopoietic stem cells with secondary myelofibrosis. Blood 51:189-194, 1978 5. Buschle M, Janssen JW, Drexler H, et al: Evidence of pluripotent stem cell origin of idiopathic myelofibrosis: Clonal analysis of a case characterized by a N-ras gene mutation. Leukemia 2:658-660, 1988 6. Kreipe H, Jaquet J, Felgner J, et al: Clonal granulocytes and bone marrow cells in the cellular phase of agnogenic myeloid metaplasia. Blood 78:1814-1817, 1991 7. Castro-Malaspina H, Gay RE, Jhanwar SC, et al: Characteristics of bone marrow fibroblasts colony-forming cells (CFU-F) and their progeny in patients with myeloproliferative disorders. Blood 59:1046-1054, 1982 8. Laszlo J: Myeloproliferative disorders (MPD): Myelofibrosis, myelosclerosis, extramedullary hematopoiesis, undifferentiated MPD, and hemorrhagic thrombocythemia. Semin Hematol 12:409-432, 1975 9. Thiele J, Kvasnicka HM, Diehl V, et al: Clinicopathological diagnosis and differential criteria of thrombocythemias in various myeloproliferative disorders by histopathology, histochemistry and immunostaining from the bone marrow. Leuk Lymphoma 33:207-218, 1999 10. Michiels JJ, Thiele J: Clinical and pathological criteria for the diagnosis of essential thrombocythemia, polycythemia vera, and idiopathic myelofibrosis (agnogenic myeloid metaplasia). Int J Hematol 76:133-145, 2002 11. Thiele J, Pierre R, Imbert M, et al: Chronic idiopathic myelofibrosis, in Jaffe ES, Harris NL, Stein H, et al (eds): World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Haematopoietic and Lymphoid Tissues.Lyon, France, IARC Press, 2001, pp 35-38 12. Barosi G, Ambrosetti A, Finelli C, et al: The Italian Consensus Conference on diagnostic criteria for myelofibrosis with myeloid metaplasia. Br J Haematol 104:730-737, 1999 13. Barosi G, Marchetti M, Azzan C, for the Italian Registry of Myelofibrosis with Myeloid Metaplasia: The Italian Registry of Myelofibrosis is one year old. Haematologica 85:1121-1122, 2000 14. Polino G, Barosi G, Berzuini A, et al: Fatty bone marrow with severe myeloid hypoplasia in idiopathic myelofibrosis. Haematologica 71:117-121, 1986 15. Thiele J, Kvasnicka HM, Boeltken B, et al: Initial (prefibrotic) stages of idiopathic (primary) myelofibrosis (IMF)—A clinicopathological study. Leukemia 13:1741-1748, 1999 16. Barosi G, Viarengo G, Pecci A, et al: Diagnostic and clinical relevance of

F. Cervantes and G. Barosi

402

17.

18.

19.

20. 21.

22.

23.

24. 25.

26.

27.

28. 29.

30.

31.

32.

33.

34.

the number of circulating CD34⫹ cells in myelofibrosis with myeloid metaplasia. Blood 98:3249-3255, 2001 Zweig MH, Campell G: Receiver-operating characteristic (ROC) plots: A fundamental evaluation tool in clinical medicine. Clin Chem 39:561577, 1993 Varki A, Lottenberg R, Griffith R, et al: The syndrome of idiopathic myelofibrosis: A clinicopathologic review with emphasis on the prognostic variables predicting survival. Medicine (Baltimore) 62:353-371, 1983 Barosi G, Berzuini C, Liberato LN, et al: A prognostic classification of myelofibrosis with myeloid metaplasia. Br J Haematol 70:397-401, 1988 Hasselbach H: Idiopathic myelofibrosis: A clinical study of 80 patients. Am J Hematol 34:291-300, 1990 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 75:4-9, 1990 Rupoli S, Da Lio L, Sisti S, et al: Primary myelofibrosis: A detailed statistical analysis of the clinicopathological variables influencing survival. Ann Hematol 68:205-212, 1994 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 88:1013-1018, 1996 Tefferi A, Silverstein MN: Agnogenic myeloid metaplasia. Cancer Invest 15:177-179, 1997 Cervantes F, Pereira A, Esteve J, et al: Identification of “short-lived” and “long-lived” patients at presentation of idiopathic myelofibrosis. Br J Haematol 97:635-640, 1997 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 98:96-102, 1997 Thiele J, Kvasnicka HM, Werden C, et al: Idiopathic primary osteomyelofibrosis: A clinicopathological study on 208 patients with special emphasis on evolution of disease features, differentiation from essential thrombocythemia and variables of prognostic impact. Leuk Lymphoma 22:303-317, 2003 Kvasnicka HM, Thiele J, Werden C, et al: Prognostic factors in idiopathic (primary) osteomyelofibrosis. Cancer 80:708-719, 1997 Cervantes F, Pereira A, Esteve J, et al: The changing profile of idiopathic myelofibrosis: A comparison of the presenting features of patients diagnosed in two different decades. Eur J Haematol 60:101-105, 1998 Barosi G, Ambrosetti A, Centra A, et al: Splenectomy and risk of blast transformation in myelofibrosis with myeloid metaplasia. Blood 91: 3630-3636, 1998 Guardiola PH, Anderson JE, Bandini G, et al: Allogeneic stem cell transplantation for agnogenic myeloid metaplasia: A European Group for Blood and Marrow Transplantation, Sovieté Française de Greffe de Moelle, Gruppo Italiano per il Trapianto del Midollo Osseo, and Fred Hutchinson Cancer Research Center Collaborative Study. Blood 93: 2831-2838, 1999 Deeg HJ, Gooley TA, Flowers MED, et al: Allogeneic hematopoietic stem cell transplantation for myelofibrosis. Blood 102:3912-3918, 2003 Morel P, Demory JL, Dupriez B: Relevance of prognostic features in myeloid metaplasia to selection of patients for bone marrow transplantation. Blood 89:2219-2220, 1997(letter) Cervantes F, Barosi G, Demory JL, et al: Myelofibrosis with myeloid metaplasia in young individuals: Disease characteristics, prognostic

35.

36. 37.

38.

39.

40.

41. 42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

factors and identification of risk groups. Br J Haematol 102:684-690, 1998 Cervantes F, Barosi G, Hernández-Boluda JC, et al: Myelofibrosis with myeloid metaplasia in adult individuals 30 years old or younger: Presenting features, evolution and survival. Eur J Haematol 66:324-327, 2001 Altura R, Head DR, Wang WC: Long-term survival of infants with idiopathic myelofibrosis. Br J Haematol 109:459-462, 2000 Cervantes, Alvarez-Larrán A, Talarn C, et al: Myelofibrosis with myeloid metaplasia following essential thrombocythemia: Actuarial probability, presenting characteristics and evolution in a series of 195 patients. Br J Haematol 118:786-790, 2002 Najean Y, Dresch C, Rain JD: The very long-term course of polycythemia: A complement to the previously published data of the Polycythemia Vera Study Group. Br J Haematol 86:233-235, 1994 Passamonti F, Malabarba F, Orlandi E, et al: Polycythemia vera in young patients: A study on the long-term risk of thrombosis, myelofibrosis and leukemia. Haematologica 88:13-18, 2003 Tefferi A, Mesa RA, Schroeder G, et al: Cytogenetic findings and their clinical relevance in myelofibrosis with myeloid metaplasia. Br J Haematol 113:763-771, 2001 Njoku OS, Lewis SM, Catovsky D, et al: Anaemia in myelofibrosis: Its value in prognosis. Br J Haematol 5479-5489, 1983 Najean Y, Cacchione R, Castro-Malaspina H, et al: Erythrokinetic studies in myelofibrosis: their significance for prognosis. Br J Haematol 40:205-217, 1978 Pereira A, Cervantes F, Brugués R, Rozman C: Bone marrow histopathology in primary myelofibrosis: Clinical and hematologic correlations and prognostic evaluation. Eur J Haematol 44:94-98, 1990 Mesa RA, Hanson CA, Rajkumar SV, et al: Evaluation and clinical correlations of bone marrow angiogenesis in myelofibrosis with myeloid metaplasia. Blood 96:3374-3380, 2000 Demory JL, Dupriez B, Fenaux P, et al: Cytogenetic studies and their prognostic significance in agnogenic myeloid metaplasia: A report on 47 cases. Blood 72:855-859, 1988 Miller JB, Testa JR, Lindgren V, et al: The pattern and clinical significance of karyotypic abnormalities in patients with idiopathic and postpolycythemic myelofibrosis. Cancer 55:582-591, 1985 Besa EC, Nowell PC, Geller NL, et al: Analysis of the androgen response of 23 patients with agnogenic myeloid metaplasia: The value of chromosomal studies in predicting response and survival. Cancer 49:308313, 1982 Sagaster V, Jager E, Weltermann A, et al: Circulating hematopoietic progenitor cells predict survival in patients with myelofibrosis with myeloid metaplasia. Haematologica 88:1204-1212, 2003 Barosi G: Myelofibrosis with myeloid metaplasia: Diagnostic definition and prognostic classification for clinical studies and treatment guidelines. J Clin Oncol 17:2954-2970, 1999 Tefferi A, Mesa RA, Nagorney DM, et al: Splenectomy in myelofibrosis with myeloid metaplasia: A single-institution experience with 223 patients. Blood 95:2226-2233, 2000 Thiele J, Kvasnicka HM, Schmitt-Graeff A, et al: Dynamics of fibrosis in chronic idiopathic (primary) myelofibrosis during therapy: A follow-up study on 309 patients. Leuk Lymphoma 44:949-953, 2003 Buhr T, Busche G, Choritz H, et al: Evolution of myelofibrosis in chronic idiopathic myelofibrosis as evidenced in sequential bone marrow biopsy specimens. Am J Clin Pathol 119:152-158, 2003 Barosi GB, Buratti A, Costa A, et al: An atypical myeloproliferative disorder with high thrombotic risk and slow disease progression. Cancer 68:2310-2318, 1991