Retrospective Study Shows Inferior OS and Symptom Response in RAS/CBL-Mutant Myelofibrosis


The presence of RAS/CBL mutations was found to be associated with adverse phenotypic features and survival outcomes in patients with myelofibrosis.

The presence of RAS/CBL mutations was found to be associated with adverse phenotypic features and survival outcomes in patients with myelofibrosis, according to findings from a retrospective study published in Blood Advances.

At a median follow-up of 82 months (95% CI, 70-89), the median OS among the overall patient population (n = 464) was 123 months (95% CI, 103-141). However, univariate analysis revealed that patients with myelofibrosis who harbored RAS/CBL mutations had a median OS of 51 months (95% CI, 31-73). The median OS in patients with RAS/CBL wild-type myelofibrosis was 140 months (95% CI, 111-149; HR 2.30, 95% CI, 1.63-3.24; P < .0001).

In multivariate analysis, the former retained its significance (HR, 1.73; 95% CI, 1.10-2.71; P = .0177).

The median overall survival (OS) was 103 months (95% CI, 66-131) in patients with overt primary myelofibrosis, 193 months (95% CI, 124–not reached [NR]) in patients with pre-primary myelofibrosis, and 114 months (95% CI, 87-145) in patients with post-polycythemia vera (PV)/essential thrombocythemia (ET) myelofibrosis.

Notably, the presence of RAS/CBL mutations led to inferior OS among patients with overt primary myelofibrosis and pre-primary myelofibrosis. The median OS was 55 months and 31 months, respectively. In the RAS/CBL wild-type populations, the median OS was 110 months and 193 months, respectively.

Inferior OS was not observed in patients with post-PV/ET disease.

Pathogenesis in myelofibrosis is driven by dysregulation of the JAK/STAT pathway. Moreover, select mutations are associated with worse outcomes in patients with myelofibrosis and are incorporated into prognostic models.

The study included patients with diagnosed myelofibrosis that was retrospectively confirmed according to the 2016 World Health Organization criteria for primary myelofibrosis and the International Working Group-Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) criteria for post-PV/ET myelofibrosis. Of these patients, 29% (n = 132) had pre-primary myelofibrosis, 33% (n = 155) had overt primary myelofibrosis, and 38% (n = 177) had post-PV/ET myelofibrosis.

All patients underwent next-generation sequencing (NGS) to identify potential driver mutations in 29 myeloid-relevant genes. The assay used peripheral blood granulocytes that were collected at diagnosis or first referral.

NGS revealed that 12.7% of patients (n = 59) had RAS/CBL mutations, 5.4% (n = 25) had NRAS mutations, 2.8% (n = 13) had KRAS mutations, and 5.6% (n = 26) had CBL mutations. These RAS/MAPK pathway genes are the most commonly mutated in hematologic malignancies. Phenotype driver mutations included JAK2 (62%; n = 289), CALR (25%; n = 115), and MPL (7%; n = 32). Additionally, 9% of patients (n = 41) were triple negative. Co-occurring mutations most frequently identified were ASXL1 (33%), TET2 (20%), EZH2 (9%), SRSF2 (8%), and ZRSR2 (8%). High molecular risk mutations were found in 38% (n = 178) of patients, with 13% (n = 62) of patients harboring 2 or more of these mutations.

Patients’ response to JAK inhibitors was evaluated by the IWG-MRT and European LeukemiaNet criteria, and, at 6 months, predictors of response were identified through logistic regression analysis.

Among evaluable patients, the median age at diagnosis was 60 (range, 18-90) and 60% of patients were male.

Patients with RAS/CBL mutations were more likely to have a diagnosis of overt myelofibrosis than pre-primary myelofibrosis (P = .0021) or post-PV/ET (P < .0001) compared with patients with wild-type RAS/CBL. Additionally, mutated patients were more likely to be older at diagnosis (P = .0235), have white blood cell counts greater than 25 x 109/L (P = .0010), lower hemoglobin levels (P = .0016), lower platelet counts (P = .0031), higher CD34 positivity in peripheral blood (P = .0003), and higher peripheral blood blasts (P < .0001) compared with the wild-type population. Mutated patients were also more likely to have constitutional symptoms (P = .0088), be dependent on red blood cell transfusion (P = .0002), and develop extramedullary hematopoiesis (P = .0132) than unmutated patients.

Patients with mutated RAS/CBL were also more likely to have leukemic transformation compared with unmutated patients. At the last follow-up and across myelofibrosis subtypes, 36% of patients with RAS/CBL-mutant myelofibrosis transformed to acute leukemia versus 9% of unmutated patients (P < .0001). The 5-year cumulative incidence of leukemic transformation was 12% in the overall cohort. This measure was 29% in patients with RAS/CBL-mutant disease versus 9% in patients with RAS/CBL wild-type disease. In mutated versus unmutated patients respectively, the 5-year cumulative incidence was 24% versus 8% in overt primary myelofibrosis (P = .0066) and 45% versus 8% in pre-primary myelofibrosis (P = .0003).

Additional results showed that RAS/CBL mutations were found to be independent predictors of reduced response to JAK inhibitors, such as ruxolitinib (Jakafi).

In total, 26% of patients received a JAK inhibitor during their disease course; 57 patients were treated with ruxolitinib, 1 with pacritinib, 1 with momelotinib, and 2 with sequential JAK inhibitors (pacritinib-ruxolitinib vs ruxolitinib-momelotinib). Of these patients, 50% were treated in a clinical trial setting.

Of these patients, 15% had RAS/CBL mutations at baseline, whereas 8% acquired such mutations during treatment with JAK inhibitors.

Regarding responses to JAK inhibitors, 8% of patients, all with wild-type myelofibrosis, achieved an anemia response after a median treatment time of 30 months (range, 3-120). After a median time of 1 month of starting a JAK inhibitor, 90% of patients achieved a symptom response; however, the rate of response was significantly lower among RAS/CBL-mutated patients versus wild-type patients, at 67% and 94%, respectively (P = .0104). At 6 months of treatment, 77% of patients, 9% of which had RAS/CBL-mutant disease, obtained a symptom response (P = .0118).

Compared with 59% of wild-type patients, no patients with RAS/CBL-mutant disease derived a spleen response from JAK inhibitor therapy at 6 months of treatment.

Among JAK-inhibitor treated patients, those with RAS/CBL mutations had a significantly inferior OS rate of 30.4 months (95% CI, 11.1-71.3) compared with 91.4 months in the wild-type population (95% CI, 64.8–not reached; HR, 4.57; 95% CI, 1.98-10.55; P = .0001).

The absence of JAK2 (P = .0054) and the presence of CALR (P = .0352), as well as the presence of baseline RAS/CBL mutations (odds ratio [OR], 0.17; 95% CI, 0.04-0.75; P = .0194), correlated with a lower probability of the patient obtaining a symptom response at 6 months.

In multivariate analysis, RAS/CBL mutations (OR, 0.17; 95% CI, 0.03-0.86; P = .0323) and the absence of JAK2 (OR, 6.85, 95% CI 1.63-28.85; P = .0087) remained the only independent predictive markers of inferior symptom response at 6 months.

Although prospective studies are needed, the study authors concluded that the adverse phenotypic features associated with RAS/CBL mutations suggest that dual targeting of the JAK/STAT and RAS/MAPK pathways may provide an opportunity to improve responses in patients with myeloproliferative neoplasms, such as myelofibrosis.


Coltro G, Rotunno G, Mannelli L, et al. RAS/CBL mutations predict resistance to JAK inhibitors in myelofibrosis and are associated with poor prognostic features. Blood Advances. 2020;4(15):3677-3687. doi:10.1182/bloodadvances.2020002175

Related Videos
Akriti Jain, MD
Mikkael A. Sekeres, MD, MS
Somedeb Ball, MBBS, assistant professor, medicine, Division of Hematology Oncology, Department of Medicine, Vanderbilt University Medical Center
Mikkael A. Sekeres, MD, MS
Naseema Gangat, MBBS