Myelofibrosis (MF), a subtype of the Philadelphia chromosome-negative myeloproliferative neoplasms (MPN), is characterized by Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway activation. The most common mutations reported in MF include the JAK2V617F mutation, and mutations in the calreticulin (CALR) and the thrompopoietin (TPO) receptor gene (encoded by the myeloproliferative leukemia virus (MPL) oncogene). The MPN Hub recently published a summary of a recent review of MF, which can be found here.
At the eighth annual meeting of the Society of Hematologic Oncology (SOHO), Raajit Rampal1 gave a presentation on the genomics in MF and their use in clinical practice. Rampal began by presenting the following case study, asking the audience to consider the risk of disease progression and implications for treatment response.
- A 69-year-old man seen in primary care for worsening fatigue, mild night sweats, minimal weight loss, and experiencing abdominal discomfort due to splenomegaly.
- Hemoglobin = 9.0 gm/dL; white blood cell count (WBC) = 12,000 per uL; platelets = 120,000 per uL.
- Peripheral blood analysis showed teardrop cells, early forms, reticulocytes, and 2% blasts.
- Bone marrow morphology consistent with primary MF, 2+ fibrosis genomic profiling shows JAK2V617F, ASXL1, DNMT3A mutations.
- Cytogenetic analysis reveals deletion 20q, trisomy 8.
- On treatment with ruxolitinib for disease-related symptoms and splenomegaly.
Rampal went on to highlight the International Prognostic Scoring System (IPSS) and the Dynamic International Prognostic Scoring System (DIPSS) as tools currently available to determine risk of disease progression by including variables such as patient age, WBC, hemoglobin levels, blast cells in the peripheral blood, and constitutional symptoms.
Despite the heterogeneity among MPN, the hallmark of these diseases is activation of the JAK-STAT pathway, including
- JAK2 mutation: 95% of polycythemia vera cases, 45–50% of essential thrombocythemia (ET) and MF cases,
- CALR mutation: 30–40% of ET/MF cases
- MPL mutation: < 5% of ET/MF cases
Impact of mutations on prognosis
With 40–50% of patients having more than one mutation of JAK-STAT signaling pathway (for example LNK, CALR, MPL, and JAK2 mutations), Rampal stated that this has implications for prognosis.
- Data from Klampfl et al.2 demonstrated that a CALR exon 9 mutation was associated with a longer overall survival (OS) than either the JAK2V617F or the MPL exon 10 mutations in patients with myelofibrosis.
- Vannucchi et al.3 showed that ASXL1, EZH2, SRSF2, and IDH1/2 seem to predict disease progression in myelofibrosis.
- Recent work by Santo and Getta et al.4 found that NRAS and KRAS seem to have a negative impact on survival and promote leukemic transformation.
- Analysing combinations of these mutations, Tefferi et al.5 found that the best OS was seen in patients with the CALR mutation and no ASXL1 mutation (CALR+ ASXL1−; median 10.4 years), whereas patients with no CALR mutation combined with an ASXL1 mutation had the worst OS (CALR− ASXL1+; median 2.3 years).
The MIPSS70-plus score assesses prognosis based on mutations
Rampal went on to discuss scores that encompass some of these mutations to enable the stratification of patients into risk groups. The MIPSS70-plus score, for example, includes genetic information as well as clinical variables similar to IPSS and DIPSS.
According to the MIPSS70-plus score, the following variables are associated with a reduced OS6:
- Hb < 100 g/L
- PB blasts ≥ 2%
- constitutional symptoms
- absence of CALR Type 1
- high molecular risk
- ≥ 2 high molecular risk mutations
- unfavorable karyotype
When new mutations that may have a prognostic value are discovered, these tools are updated. For example, the U2AF1 mutation has been found to be associated with anemia and thrombocytopenia; it has been incorporated into MIPSS70-plus v2.0 as well as in the Genetically Inspired Prognostic Scoring System (GIPSS).
Applying this information to the aforementioned case study, the variables suggest that, using the MIPSS70-plus v2.0 tool, the patient would be at very high risk with a 10-year OS of less than 5%, while the MIPSS-70 risk score tool would give a 34% probability of survival for 5 years. Rampal highlights that this is important information can be obtained using a simple online calculator (http://www.mipss70score.it/).
Rampal then asked whether genetics could inform us how well the patient in the case study would do on ruxolitinib. A study by Patel et al.7 demonstrated that the type and number of mutations could predict how long patients would respond to treatment. The presence of ASXL1, EZH2, and DNMT3 mutations specifically suggested a shorter time to treatment failure, which would be of concern with the patient in the case study.
Predicting outcome after transplantation
Another score, the Myelofibrosis Transplant Scoring System (MTSS)8 enables the stratification based on the outcomes of allogeneic hematopoietic stem cell transplantation (allo-HSCT). The following parameters are assessed:
- leukocyte and platelet count
- performance status and age
- HLA mismatched unrelated donor
- absence of MPL or CALR mutations
- presence of ASXL1 mutations.
The patient in the case study would be unlikely to benefit from ruxolitinib for a prolonged period, however, using the MTSS tool, the 5-year OS following allo-HSCT is in the range of 77–90%. In contrast, other studies9, 10 have assessed posttransplant outcomes and have found that ASXL1 and other mutations have no impact while U2AF1 mutations may be associated with worse transplant outcomes. Therefore, tools to predict outcome after transplantation are evolving and more studies are needed to consolidate these findings.
Rampal concluded his presentation by stating that assessing the genomic profile of patients with myelofibrosis is vital for the optimization of treatment and that, as genomic alterations change over time, recurrent profiling should be utilized. He stressed that several other mutations, aside from JAK-STAT activating mutations, can be present and may have prognostic value, such as predicting duration of response to ruxolitinib treatment. MIPSS-70 is a useful tool for risk stratification; however, as genomic alterations are not static but change over time, recurrent profiling should be utilized to inform treatment decisions, such as referral for allo-HSCT.