Current Approaches to Diagnose and Treat Primary Myelofibrosis


During a Targeted Oncology™ Case-Based Roundtable™ event, Rami Komrokji, MD, discussed elements to diagnosing myelofibrosis and how to approach risk stratification before treatment.


Rami Komrokji, MD

Vice chair, Malignant Hematology Department

Head of the Leukemia and MDS Section

Moffitt Cancer Center

Tampa, FL

KOMROKJI: MF could be either primary de novo or coming from secondary from essential thrombocythemia [ET] or polycythemia vera [PT]. [Concerning] the major criteria listed [by the World Health Organization (WHO)], I always bring up the 2 points that not every fibrosis in the bone marrow is myelofibrosis.1 You can see it in lymphomas, hairy cell leukemia, connective tissue disease, etc, and you don't need fibrosis in the early stages of myelofibrosis to make the diagnosis. The classical megakaryocytic atypia is enough and in the prefibrotic MF, that's enough to diagnose the disease.

Prefibrotic MF is a relatively new entity that we talk about. Many patients in practice are labeled as ET, and sometimes it’s hard to tease those [differences] out. But those are the patients who we would think have ET, and in 3 to 4 years, they have overt MF. Usually, it will it take a decade to get there, but if a patient had ET and then in 3 or 4 years was in [overt] MF, those probably were patients with prefibrotic MF. There are few clues…most of the time, those patients will have high LDH [lactate dehydrogenase], on the bone marrow there will be more hypercellular granulocytic hyperplasia. There is more clustering of the megakaryocytes.

Currently, we manage them almost the same, but those are the patients who will transform earlier, at higher risk of leukemia. Maybe in the future, those are the patients we will target with some more interventions to try to prevent the overt MF.

The presence of a clonal marker excludes other diseases. [However], myelodysplastic syndrome [MDS] with fibrosis is sometimes hard to distinguish. Fibrosis can be seen in MDS; it's typically associated with bad outcomes and the new WHO classification with the blast increase has MDS with fibrosis [as a] category on its own. In the clinical phenotype, they typically don't have the hepatosplenomegaly as much as constitutional symptoms. They're cytopenic, more like MDS. If a good pathologist sees myeloid or erythroid dysplasia, that will favor MDS with fibrosis. The megakaryocytes are tricky because you always see megakaryocytic atypia in MPNs [myeloproliferative neoplasms], and it depends on how experienced the hematopathologist is. If they are mistakenly calling them dysplasia, that could be deceiving. There are some minor criteria: the anemia, leukocytosis, splenomegaly, LDH, and leukoerythroblastosis.

What is the role of risk stratification when treating patients with MF?

Once we establish the diagnosis, we want to risk stratify the patients and there are many models in MF, 3 or 4 clinical and 2 molecular. I like the MIPSS70 [MIPSS70: Mutation-Enhanced International Prognostic Score System for Transplantation-Age Patients With Primary Myelofibrosis] most because it's comprehensive and it was designed to look at the question of transplant or not in younger patients not counting the age as a factor.2 Anemia, transfusion dependency, thrombocytopenia, and leukocytosis… [lead to poor prognosis]. Circulating blasts, unfavorable karyotype, [etc], all of those are weighed in these models. Molecular models…account for bad mutations like ASXL1, SRFS2, or absence of calreticulin. But at the end, we are putting the patients into a spectrum of a low-risk disease, where the survival spans many years, to a high-risk disease where the survival is less than 2 years.

Why is it important to use prognostic models for MF?

The disease risk value in practice is deciding on transplant. If somebody is not eligible for stem cell transplant [SCT], you may argue that those models are not that important. Somebody who's very low risk will rarely be symptomatic, because if they have any symptoms, they probably move up to intermediate-1 risk.

If somebody's survival estimate is 2 to 3 years, or an intermediate-2 or higher risk by any of those models, we think of the SCT earlier on in the course of the disease [to consider if they are] eligible for transplant by functional status and comorbidities, not necessarily by age. The second thing is [having] enough disease risk to justify the SCT. In patients who have higher risk, the timing of the transplant is probably early on. In patients with lower risk, even if they are eligible for SCT, the optimal timing is probably to try to delay the SCT. It's always a hard decision because you don't want to go too early [because of] upfront transplant-related mortality. But you also never want to go into an MPN accelerated phase or acute myelocytic leukemia from MPN because those diseases have terrible outcomes.

What recommendations are there for treatment of higher-risk myelofibrosis?

Once we label the patients intermediate or higher risk, we are assessing the symptoms and deciding on treatment. We rarely see patients who just [have] transfusion-dependent anemia. Those patients are probably not the classical candidates for JAK2 inhibitors, at least the classical ruxolitinib [Jakafi] or fedratinib [Inrebic].

[For] most patients…you're treating either constitutional symptoms or splenomegaly. For those patients, JAK2 inhibitors are reasonable. The National Comprehensive Cancer Network guidelines split that choice of JAK2 inhibitor based on the platelet count.3 If it's below 50 × 109/L, pacritinib [Vonjo] is the choice; if it's above 50 × 109/L, [the choice is] ruxolitinib or fedratinib. Most [physicians] are more used to ruxolitinib, [it has] more data…but fedratinib is a reasonable option as well. Sometimes I think even a platelet cutoff of 100 × 109/L would be reasonable to consider pacritinib; the platelet cutoff of 50 × 109/L was for the truly unmet need and accelerated approval of pacritinib.

If patients are candidates for SCT, many times we do start the JAK2 inhibitors before the SCT because the SCT will still take 3 to 4 months to happen. If patients have a big spleen [and] poor performance from the disease, shrinking the spleen and getting them ready for SCT is reasonable.

The 3 available JAK2 inhibitors, ruxolitnib, fedratinib, and pacritinib…have different targets. Ruxolitinib targets JAK1/JAK2, [and has] potent JAK1 [activity]. Pacritinib has different targets; it doesn't have any JAK1 activity. It has some ACVR1 [activity] so some anemia response can be explained through that [and] other inflammatory pathways like IRAK1. Fedratinib also has some FLT3 activity and some JAK1 activity. Momelotinib has JAK1 and ACVR1 activity.

The choices are based on the cytopenia profile. Fedratinib most of the time is positioned as second line after ruxolitinib in patients that are still proliferative. Ruxolitinib is the first line in patients that are proliferative, not cytopenic. Pacritinib is for thrombocytopenia and when we have approval for momelotinib, hopefully that will be for the anemia phenotype.


1. Barbui T, Thiele J, Gisslinger H, et al. The 2016 WHO classification and diagnostic criteria for myeloproliferative neoplasms: document summary and in-depth discussion. Blood Cancer J. 2018;8(2):15. doi:10.1038/s41408-018-0054-y

2. Guglielmelli P, Lasho TL, Rotunno G, et al. MIPSS70: Mutation-Enhanced International Prognostic Score System for Transplantation-Age Patients With Primary Myelofibrosis. J Clin Oncol. 2018;36(4):310-318. doi:10.1200/JCO.2017.76.4886

3. NCCN. Clinical Practice Guidelines in Oncology. Myeloproliferative neoplasms, version 2.2023. Accessed September 7, 2023.

Recent Videos
Related Content