Amongst the classic Philadelphia chromosome-negative myeloproliferative neoplasms (Ph-negative MPNs), primary myelofibrosis (PMF) is associated with the poorest outcome. Most recent estimates of median survival range from six to seven years. Approximately 90 percent of patients have an identifiable “driver” mutation—JAK2 V617F in 50 to 60 percent of patients, CALR exon 9 mutations in 20 to 30 percent, and MPL mutations in five to 10 percent.
About 10 percent of patients have “triple negative” disease, which has been associated with inferior outcomes. Pathologic hallmarks of PMF include megakaryocyte proliferation and atypia and varying degrees of reticulin or collagen fibrosis in the bone marrow and peripheral blood leukoerythroblastosis.
Typical clinical features include anemia, splenomegaly and a range of constitutional symptoms. Treatment is directed at improving these clinical features, as well as changing the natural course of the disease and improving survival. Management of patients with post-PV/ET myelofibrosis is approached in the same way.
Assessment of Prognosis and Symptom Burden
A number of prognostic scoring systems for PMF have been developed over the years. In the clinic, we typically use the International Prognostic Scoring System for the purposes of prognostication in patients at diagnosis and the Dynamic IPSS at later time points, as these models have been the most extensively validated and rely on easily available clinical information: age, hemoglobin level, leukocyte count, constitutional symptoms, and presence of circulating blasts.
However, it is important to recognize the importance of emerging information on the impact of both driver and other somatic mutations, both in the terms of genes affected and the number of mutations. For example, CALR mutations favorably impact prognosis, while ASXL1 mutations adversely affect survival, as does an increased number of mutations.
New prognostic models incorporating molecular information (e.g., MIPSS) have been presented, but are not yet ready for prime time. Because constitutional symptoms represent a major burden for patients with PMF or post-PV/ET MF and can be substantially alleviated with the use of JAK inhibitors, we routinely inquire about the presence and severity of fatigue, inactivity, night sweats, itching, fever, weight loss, bone pain, early satiety, and abdominal discomfort.
Our Overall Approach
For low-risk patients, we recommend observation only. For intermediate- and high-risk patients with symptomatic splenomegaly or troublesome MF-related systemic symptoms, we use ruxolitinib. In patients whose only disease manifestation is anemia, we use anti-anemia medications.
In accordance with the European Group for Blood and Marrow Transplantation (EBMT)/European LeukemiaNet (ELN) guidelines, we refer patients with intermediate-2- or high-risk disease for consultation regarding potentially curative allogeneic stem cell transplantation (allo-SCT). Selected intermediate-1-risk patients—such as those with high-risk genomic abnormalities; circulating blasts >5%; or refractory, transfusion-dependent anemia—are also referred for consultation regarding allo-SCT.
The optimal duration of ruxolitinib therapy before allo-SCT is currently unknown and is the subject of active investigation.
How We Use Ruxolitinib
It is critical to recognize the central role of the Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway and its universal activation across the spectrum of Ph-negative MPNs, irrespective of the mutation status of the JAK2 gene.
Ruxolitinib is highly effective in ameliorating constitutional symptoms and reducing spleen size in JAK2-mutated and—wild type patients alike. In the pivotal COMFORT-I study, as well as upon longer follow-up of the COMFORT-II study, ruxolitinib improved overall survival when compared with placebo or best-available therapy, respectively.
Of particular importance, reductions in spleen size, which are dose-dependent, correlate with the survival benefit of the drug. In general, we dose ruxolitinib on-label—i.e., according to the patient's platelet count.
In situations where it may be necessary to start with a lower dose of the drug—e.g., a patient with significant thrombocytopenia—we believe it is important to titrate up the dose as quickly as possible so as to achieve the best spleen response obtainable, whereas 10 mg twice a day appears to be sufficient to obtain relief of constitutional symptoms.
Although anemia may initially worsen on ruxolitinib, hemoglobin levels tend to improve and stabilize over time, and anemia is rarely a reason to discontinue the drug.
If ruxolitinib has to be discontinued in a patient responding well to it, it should not be done abruptly as this can lead to a rapid and profound return of symptoms. While occasional clonal remissions have been reported, we do not recommend serial measurement of the JAK2 V617F allele burden or bone marrow biopsies to check fibrosis grade during therapy with ruxolitinib.
How We Treat Anemia
Specific treatments for anemia of MF include erythropoietin-stimulating agents (ESAs), danazol, and immunomodulatory drugs (Imids). They may be combined with ruxolitinib in patients who have both symptomatic splenomegaly/constitutional symptoms and anemia.
A typical starting dose of danazol is 200 mg orally twice a day. The drug is contraindicated in men with a history of prostate cancer. Most responses to ESAs and danazol are seen within three to six months.
Among the Imids, in our experience, lenalidomide is the most active but causes significant cytopenias (particularly in combination with ruxolitinib) and needs to be given under strict supervision. Pomalidomide did not improve anemia response rates in a randomized placebo-controlled clinical trial; hence, we prefer low-dose thalidomide (50 mg/d) with or without prednisone (30 mg/d for the first month, 15 mg/d for the second month, 15 mg q.o.d for the third month). This dose is generally well-tolerated, even with long-term use. Anemia response rates to lenalidomide or thalidomide are in the range of 20 to 30 percent.
Overall, we try to avoid splenectomy because of its perioperative complication (infection, thrombosis, bleeding) rate, including a five to 10 percent mortality rate. However, splenectomy can be useful in selected situations—e.g., in patients with massive splenomegaly and refractory cytopenias not responsive to standard therapies. We do not usually recommend the use of splenic irradiation due to the prolonged and severe cytopenias this can cause.
For extramedullary hematopoiesis at other sites, as well as for pulmonary hypertension due to MF, low-dose radiation is the treatment of choice. Ruxolitinib has largely supplanted the use of hydroxyurea for splenomegaly.
For patients with elevated blood or bone marrow blasts, we use hypomethylating agents (HMAs), in combination with ruxolitinib if needed, to decrease blasts, followed by allo-SCT if a donor is available.
Combinations of ruxolitinib with other potentially effective anti-MF agents—e.g., histone deacetylase, hedgehog, DNA methyltransferase, and phosphatidylinositol-3-kinase inhibitors—are being evaluated in order to improve results seen with ruxolitinib alone and to obtain additional benefits (e.g., improvements in anemia or bone marrow fibrosis).
Other novel JAK inhibitors that are less myelosuppressive, such as pacritinib and momelotinib, are also in advanced phases of clinical testing.
Finally, new drugs for anemia—e.g., sotatercept, and novel compounds like immune checkpoint inhibitor—are being tested as well. Referring patients for clinical trials whenever possible continues to be of paramount importance.