Primary myelofibrosis is a clonal disorder of a multipotent hematopoietic progenitor cell of unknown etiology; it is characterized by myeloid cell proliferation, megakaryocytic atypia, BM fibrosis, a leukoerythroblastic peripheral blood picture, extramedullary hematopoiesis (EMH), and splenomegaly. Primary myelofibrosis was previously called CIMF, MF with myeloid metaplasia (MMM), or agnogenic myeloid metaplasia (AMM). The disease can occur either de novo or as a late complication of PV or ET. In either case, it manifests as progenitor cell–derived clonal myeloproliferation accompanied by intense marrow stromal reaction, including collagen fibrosis, osteosclerosis, and angiogenesis.
Fibrogenesis and angiogenesis are thought to develop consequent to the release of growth-promoting factors (such as vascular endothelial growth factor [VEGF], PDGF, basic fibroblast growth factor [bFGF] and transforming growth factor β [TGF-β]) from proliferating atypical megakaryocytes in the BM. The JAK2V617F mutation is found in 50% to 60% of patients with PMF. Persistent JAK-STAT signaling, resulting in the overproduction of proinflammatory cytokines, has been observed in all patients with PMF (69,70). Proinflammatory cytokines have been associated with many of the symptoms of MF, including splenomegaly, transfusion dependence, thrombocytopenia, and shortened survival (71). Mutations in the thrombopoietin receptor (MPL) are found in 5% to 10% of patients, and CALR mutations in an additional 25% (45,46). Rare inactivating mutations in negative regulators of JAK-STAT signaling (eg, LNK, SOCS, and CBL) also contribute to the dysregulated JAK-STAT signaling in PMF (72).
The exact contributions of mutations in JAK2, MPL, and CALR to disease pathogenesis remain unclear. Recent studies suggested that the heterogeneity of mutations in PMF may underlie the heterogeneity of its clinical phenotype; that is, these mutations may be associated with distinct clinical features. In a study of 617 patients with PMF, those with CALR mutations had a lower risk of anemia, thrombocytopenia, and leukocytosis (73). In another series of 428 patients with PMF, CALR mutations were associated with younger age, lower leukocyte count, and higher platelet count, while MPL W515K/L mutations were associated with younger age and lower leukocyte count when compared with JAK2V617F mutations (53). A number of other mutations have also been found in PMF, albeit at much lower frequencies than JAK2 and CALR mutations (eg, mutations in ASXL1, EZH2, SRSF2, CBL, IDH1/IDH2, TP53, TET2, and DNMT3) (72). Unlike the JAK2/CALR/MPL mutations, which are mutually exclusive, the other less-frequent mutations may coexist with each other or with the three driver mutations.
PMF is a heterogeneous disorder with variable age of onset, presenting features, phenotypic manifestations, and prognosis. The incidence of PMF increases with age. In a series of 1,054 patients, the median age at diagnosis was 64 years; 17% of patients were younger than 50 years and 5% younger than 40 years (74). Clinical presentation can range from no or minimal symptoms, where disease is discovered during a workup for leukocytosis or splenomegaly, to severe symptoms. Severe fatigue is the most common presenting symptom. Constitutional symptoms (fatigue, weight loss, pruritus, low-grade fever, night sweats) are a prominent feature of PMF and can be debilitating. Myeloproliferation is one of the major features of the disease and can lead to sequestration of immature cells and production of blood cells in sites other than the BM, a phenomenon known as EMH. This commonly manifests as marked hepatosplenomegaly, with associated pain, early satiety, portal hypertension, and anemia and thrombocytopenia. Splenomegaly is present in 80% of patients and may extend into the pelvis. Hepatomegaly is seen in 40% to 70% of patients. EMH might cause symptoms in various other organs, leading to respiratory distress, pulmonary hypertension, ascites, pericardial tamponade, cord compression, and paralysis. Peripheral smear generally provides the first clue toward a diagnosis of PMF, with the presence of characteristic teardrop red cells and a leukoerythroblastic picture (presence of immature myeloid cells including blasts in the peripheral blood). Progressive anemia generally develops, requiring transfusions. Some patients may present with leukocytosis and thrombocytosis; however, most develop leukopenia and thrombocytopenia in later stages of the disease. Among the most feared complications of PMF is transformation to AML, occurring in 10% to 20% of patients in the first 10 years from diagnosis. The outcome after transformation is poor, with a median survival of 5 months. Transformation to AML is the most common cause of death in MF, followed by MF progression without acute transformation, thrombosis, and cardiovascular complications, infection, bleeding, and portal hypertension.
A diagnosis of PMF is made using the 2008 WHO criteria (see Table 6-8) (6). Symptoms such as splenomegaly, leukoerythroblastosis, anemia, poor quality of life, and BM megakaryocyte hyperplasia are suggestive of PMF. Marrow fibrosis by itself is not specific for a diagnosis of PMF. Various degrees of fibrosis are observed in other MPNs, and MDS with fibrosis must be excluded. Morphologic features of the BM during the prefibrotic (cellular) phase of PMF are shown in Fig. 6-8, and those during the fibrotic phase are depicted in Figs. 6-9,6-10,6-11. Classical morphological features consistent with PMF and seen in the peripheral blood smear are demonstrated in Fig. 6-12. Bone marrow histology, especially megakaryocyte morphology, is a critical diagnostic criterion for PMF (Fig. 6-13). All patients suspected of having PMF should undergo bone marrow biopsy with reticulin and collagen staining and testing for JAK2V617F, CALR, and MPL mutations. Chronic myelogenous leukemia should be ruled out by testing for Bcr-Abl.
Table 6-82008 WHO Criteria for Diagnosis of Primary Myelofibrosis (PMF) ||Download (.pdf) Table 6-8 2008 WHO Criteria for Diagnosis of Primary Myelofibrosis (PMF)
|Major criteria |
|1. Presence of megakaryocyte proliferation and atypia, usually accompanied by either reticulin and/or collagen fibrosis, or, in the absence of significant reticulin fibrosis, the megakaryocyte changes must be accompanied by an increased bone marrow cellularity characterized by granulocytic proliferation and often decreased erythropoiesis (ie, prefibrotic cellular-phase disease) |
|2. Not meeting WHO criteria for PV, CML, MDS, or other myeloid neoplasm |
|3. Demonstration of JAK2617VF or other clonal marker (eg, MPL515WL/K) or, in the absence of a clonal marker, no evidence of bone marrow fibrosis due to underlying inflammatory or other neoplastic diseases |
|Minor criteria |
|1. Leukoerythroblastosis |
|2. Increase in serum lactate dehydrogenase level |
|3. Anemia |
|4. Palpable splenomegaly |
|Diagnosis of PMF requires meeting all three major criteria and at least two minor criteria |
It is difficult to distinguish the prefibrotic (cellular) phase of PMF from other types of chronic myeloproliferative neoplasms based on morphological criteria alone. However, careful microscopic examination of the bone marrow biopsy usually reveals scattered atypical megakaryocytes with morphological criteria classical for PMF in the fibrotic phase. As shown, some of the megakaryocytes in this bone marrow biopsy are remarkably variable in size and shape and characteristically contain markedly hyperchromatic nuclei (×200).
During the fibrotic phase of PMF, bone marrow hematopoietic cellular elements tend to decrease in number with interstitial infiltration of the bone marrow by fibroblasts that leads to a streaming effect. Characteristically, the megakaryocytes demonstrate variability in size and shape, and megakaryocytes containing hyperchromatic and hyperlobulated nuclei are frequently encountered during the fibrotic phase of PMF (×200).
Another common feature of the bone marrow during the fibrotic phase is marked expansion of bone marrow sinusoids, which are usually rudimentary under normal conditions (×100). Hematopoietic cellular elements can be detected within the bone marrow sinuses; a megakaryocyte is shown, comprising what is known as intrasinusoidal hematopoiesis (inset; ×400).
During the fibrotic phase of PMF, the bone marrow is characterized by increased interstitial reticulin fibrosis (upper panel; ×100), which might be associated with the abnormal presence of collagen fibers that are detected by trichrome staining (lower panel; ×200).
Careful examination of peripheral blood smears from patients with PMF usually reveals teardrop red blood cells (arrows, upper panel; ×400). In addition, nucleated red blood cells (upper panel) and left-shifted granulopoiesis (lower panel; ×400) are seen, two morphologic criteria collectively described as leukoerythroblastosis.
Diagnostic algorithm for suspected primary myelofibrosis. (Reproduced with permission from Tefferi A, Vardiman JW. Classification and diagnosis of myeloproliferative neoplasms: the 2008 World Health Organization criteria and point-of-care diagnostic algorithms. Leukemia. 2008;22:14-22.)
The median survival is 5 years. In a review of 1,054 patients with PMF, the median survival was 69 months (74). Younger patients with good prognostic features may have a life expectancy exceeding 10 years. The most commonly used prognostic scoring systems are the International Prognostic Scoring System (IPSS), designed to be used at diagnosis, and the Dynamic IPSS (D-IPSS), which can be used at any point in the patient's disease course (74,75) (Table 6-9). Both prognostic scoring systems are based on clinical and laboratory characteristics: age (>65 years), constitutional symptoms (yes/no), hemoglobin (<10 g/dL), leukocyte counts (>25 × 109/L), and circulating blasts (≥1%). In the IPSS system, all factors are given a score of 1, while in the D-IPSS, hemoglobin below 10 g/dL is given 2 points. On the basis of these risk factors, patients are separated into four risk groups: low risk (0 points); intermediate-1 risk (1 point IPSS; 1-2 points D-IPSS); intermediate-2 risk (2 points IPSS; 3-4 points D-IPSS); and high risk (≥3 points IPSS; 5-6 points D-IPSS). Cytogenetic abnormalities are found in about half of the patients with PMF. Common abnormalities include del(13q), del(20q), trisomy 8 or 9, and abnormalities of chromosome 1 [partial trisomy and der(6)t(1;6)] (76). Tam et al analyzed 256 patients with PMF; 36% had chromosomal abnormalities (77). They categorized patients into those with favorable cytogenetics (sole deletion of 13q or 20q, trisomy 9 +/- one other abnormality); diploid cytogenetics; unfavorable cytogenetics (abnormalities of chromosomes 5 or 7, or complex [≥3] cytogenetics); and very unfavorable cytogenetics (any abnormality of chromosome 17). The median survival times (for patients with assessment at diagnosis) were 63, 46, 15, and 5 months, respectively. Gangat et al added unfavorable karyotype, platelet count below 100 × 109/L, and transfusion dependence as independent risk factors for inferior survival in another prognostic model (DIPSS-plus) (78).
Table 6-9Prognostic Scoring Systems for Myelofibrosis ||Download (.pdf) Table 6-9 Prognostic Scoring Systems for Myelofibrosis
|Factors ||IPSS ||D-IPSS ||D-IPSS-plus |
|Age >65 years ||1 ||1 ||1 |
|Constitutional symptoms ||1 ||1 ||1 |
|Hemoglobin <10 g/dL ||1 ||2 ||1 |
|Leukocytes >25 × 109/L ||1 ||1 ||1 |
|Blood blasts ≥1% ||1 ||1 ||1 |
|Platelet count < 100 × 109/L || || ||1 |
|Transfusion dependence || || ||1 |
|Unfavorable karyotype || || ||1 |
|Risk stratification (median survival) |
|Low ||0 points (11.2 y) ||0 points (not reached) ||0 points (15.4 y) |
|Intermediate-1 ||1 point (7.9 y) ||1-2 points (14.2 y) ||1 point (6.5 y) |
|Intermediate-2 ||2 points (4 y) ||3-4 points (4 y) ||2-3 points (2.9 y) |
|High ||≥3 points (2.3 y) ||5-6 points (1.5 y) ||≥4 points (1.3 y) |
Recent studies have explored the prognostic relevance of various mutations. In a study of 617 patients with PMF, CALR mutations were associated with longer survival (median 17.7 years) (73). Patients without any mutation in JAK2, MPL, or CALR (“triple negative”) had a higher incidence of transformation to AML and a shorter survival (median 3.2 years). When the CALR, JAK2, and MPL mutations were added to the IPSS, patients could be further subdivided into five risk groups, with significantly different median survivals (73). Another study of 428 patients with PMF reported similar findings (53). Patients with CALR mutations had the longest survival (median 15.9 years), while patients without any of these mutations (triple negative) had the shortest survival (median 2.3 years) (53). Mutations in epigenetic modulators (ASXL1, SRSF2, and EZH2) were associated with worse survival and increased risk of transformation to AML (72). The negative prognostic impact of ASXL1 was shown in another series of 570 patients (79). Patients who had ASXL1 mutations in the absence of CALR mutations had the worst survival (median 2.3 years). Future studies will explore further how to best implement the molecular information in everyday practice.
Before the approval by the FDA of the JAK1/2 inhibitor ruxolitinib in 2011, treatment of MF was unsatisfactory. Cytoreductive drugs such as HU or cladribine were used to control hyperproliferation, although their effects are transient and rarely result in complete spleen regression. Oral alkylating agents have also been used, but often induce severe myelosuppression and are associated with an increased risk of transformation to AML. Corticosteroids, erythroid-stimulating agents, and androgens have proven helpful in treatment of anemia. Patients with low serum Epo (<125 U/L) can be given subcutaneous injections of Epo (40,000 U/week). Corticosteroids (prednisone 0.5 to 1.0 mg/kg/day) or androgens (testosterone enanthate injections 400-600 mg/week; oral danazol 200 mg two or three times/day) have also been useful. Immunomodulatory agents (low-dose thalidomide and lenalidomide) have anticytokine and antiangiogenic effects and have been shown to reduce splenomegaly and improve anemia. They are usually used with tapering doses of prednisone for 3 months. Interferon alfa has been used with some success, but significant toxicity prevents its use in many patients. It may slow disease progression in patients with early MF, as well as reverse BM fibrosis in some patients (80). A recent retrospective study of Peg-IFN-α-2a in 62 patients with MF reported improvements in anemia and constitutional symptoms, normalization of platelet and leukocyte counts, and reduction in splenomegaly (81). In selected patients, splenectomy or splenic radiation may help with symptom control or may improve blood cell count but these procedures carry significant side effects.
Two pivotal phase III randomized trials provided evidence for the regulatory approval of the oral JAK1/2 inhibitor ruxolitinib. COMFORT-I randomized patients to ruxolitinib (n = 155) or placebo (n = 154) (82), while COMFORT-II compared ruxolitinib (n = 146) to BAT (n = 73) (83). Significantly more patients in the ruxolitinib arms had 35% or more reduction in spleen volume (approximately 50% reduction by palpation) from baseline at week 24 (COMFORT-I) or week 48 (COMFORT-II). Both studies showed significantly better improvements in MF-related symptoms and quality of life in patients treated with ruxolitinib. Thrombocytopenia and anemia were the most common toxicities associated with ruxolitinib therapy. These effects mostly appeared within the first 3 to 6 months of treatment and were managed with dose reductions or transfusions.
Long-term follow-up analyses have demonstrated that the effects of ruxolitinib are durable. After a median follow-up of 2 years, more than 80% of patients treated with ruxolitinib in COMFORT-I who had achieved a 35% or greater reduction in spleen volume had maintained at least a 10% reduction (84). In COMFORT-II, at 144 weeks, the Kaplan-Meier estimated probability of maintaining a spleen response was 50% (85). Ruxolitinib also improved survival (85,86). Long-term treatment with ruxolitinib may delay progression of or reverse BM fibrosis in some patients (87). Two case reports showed nearly complete resolution of marrow fibrosis and a reduction in JAK2V617F allele burden after long-term treatment with ruxolitinib (88,89).
In our experience, most patients with symptomatic splenomegaly or systemic MF-related symptoms, even those with transfusion-dependent anemia, can be successfully treated with ruxolitinib for long periods of time if the patient is carefully monitored (particularly during the first 3-6 months) and the dose adjusted to avoid therapy interruptions. Recommended starting doses are 20 mg twice per day in patients with platelets above 200 × 109/L, 15 mg twice a day in patients with platelets between 100 and 200 × 109/L, and 5 mg twice a day in patients with platelet counts below 100 × 109/L. The dose can be increased to a maximum of 25 mg twice a day if tolerated. Avoidance of treatment interruption is important for treatment success, as symptoms return to baseline within 7 to 10 days. Ruxolitinib doses of 10 mg twice a day or higher are effective maintenance doses. Other JAK inhibitors, which appear to be less myelosuppressive (pacritinib) and may possibly reduce the need for red blood cell transfusions (momelotinib), are in late-phase clinical development for MF.
Allogeneic Stem Cell Transplantation
Allogeneic stem cell transplant (alloSCT) is curative in MF; however, fewer than 10% of patients undergo alloSCT due to older age or severe comorbidities. Reduced-intensity conditioning regimens are an option in older patients and those with comorbidities (90). Spleen size influences the rate of engraftment after transplant, but splenectomy before alloSCT is not recommended. The use of ruxolitinib pretransplant to reduce splenomegaly is being evaluated. A study of 14 patients treated with ruxolitinib (median duration 6.5 months) before alloSCT showed that, at the time of transplantation, 7 of 11 patients with splenomegaly had a 41% median reduction in palpable spleen size as well as improvement in disease-related symptoms (91). Thirteen (93%) patients had engraftment, and 11 were alive after a median follow-up of 9 months. Treatment-related mortality was 7%. In a prospective study of 22 patients pretreated with ruxolitinib, 69% had reductions in spleen size, 86% had improvement in symptoms, engraftment was seen in all cases, and the estimated 1-year survival was 81% (92). Although the numbers were small, survival was longer in patients who responded to ruxolitinib (n = 10) than in those who did not (n = 10) (100% vs 60% estimated 1-year survival; P = .02).
Combination and Novel Therapies
Other targeted agents, such as epigenetic modifying agents (azacitidine, decitabine, panobinostat, and pracinostat); hedgehog inhibitors (LDE-225, IPI-926); PI3 kinase inhibitors (BKM120); antifibrotic agents (PRM-151); or telomenase inhibitor (imetelstat) have been tested. Most have not shown significant efficacy as single agents. Preclinical studies have shown synergistic effects when some of these agents were combined with ruxolitinib, suggesting a useful strategy to improve outcomes. Clinical trials testing ruxolitinib in combination with pracinostat, panobinostat, lenalidomide, decitabine, azacytidine, BEZ235, and LDE-225 are ongoing results of these trials are eagerly awaited.
Patients with MF should first be assigned to a risk category using one of the standard prognostic tools (IPSS, D-IPSS). For patients with low-risk disease, a watch-and-wait approach is acceptable. Patients in the intermediate or high-risk groups should be treated based on their symptoms. In most cases, with careful titration and monitoring, ruxolitinib can be safely used. For younger patients in the intermediate-2 and high-risk categories, alloSCT can be offered. For patients who are not eligible for alloSCT or are intolerant or lose their response to other therapies, enrollment in clinical trials is recommended. Ruxolitinib and other JAK inhibitors have been effective in reducing splenomegaly, in improving symptoms and quality of life, and in prolonging survival in patients with MF. JAK inhibitors have not been shown to eradicate the mutant clone, and patients lose their response to therapy over time. Results from ongoing trials of new targeted agents and their combinations are eagerly awaited.