Genes encoding the alpha (GNAS) and beta (GNB1) subunits of the heterotrimeric G-protein complex are recurrently mutated in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Alterations in G-protein coupled receptors can affect signaling via the PI3K/AKT/mTOR and RAS/MAPK pathways, suggesting that GNB1/GNAS mutations may function similarly to mutations in RAS and tyrosine kinases in myeloid disease progression. However, unlike RAS mutations, GNB1/GNAS mutations are among the most commonly affected genes in clonal hematopoiesis of indeterminate potential (CHIP), suggestive of a distinct functional role in myeloid disease initiation.

We evaluated 6343 unique patients who received gene panel sequencing at our institution as part of a diagnostic evaluation of known or suspected hematologic malignancy. We identified 68 patients that had at least 1 sample with a mutation in GNB1 (N =42), GNAS (N=24) or both (N=2) (Figure 1). Twenty-six patients had multiple samples (range 2-10) obtained at serial timepoints during the course of progression and treatment. GNB1 mutations affected codon 57 (K57E/M/N/T) in 38 out of 42 cases; the remaining were G53E, D76G, I81V, and K89R. GNAS mutations affected codon 201 (R201H/C) in 23 out of 24 cases.

Forty patients had active myeloid neoplasms [AML (N=10), MDS (N=13), MPN (N=11), MDS/MPN (N=5), BPDCN (N=1)]. The distribution of myeloid diagnoses was similar among patients with GNB1 or GNAS mutations. Analysis of serial samples showed that GNB1/GNAS mutations were frequently acquired at the time of leukemic transformation and clinical progression. Among 5 patients with secondary AML, 2 patients acquired GNB1/GNAS mutations at the time of transformation from MDS to AML, and 3 patients were found to have GNB1/GNAS mutations in high-risk MDS prior to subsequent transformation to AML. After induction chemotherapy for AML, two patients had expanded or newly acquired GNB1 mutations at time of treatment failure.

Notably, 15 out of the 40 (37.5%) patients with active myeloid disease had co-mutations in the MPN-associated genes JAK2, CALR, and MPL. Eleven of these patients had a clinical diagnosis of MPN, including myelofibrosis (n=5), essential thrombocytosis (n=3), polycythemia vera (n=2), and systemic mastocytosis (n=1). In the context of underlying oncogenic kinase alterations including JAK2 V617F, GNB1 mutations have been shown to promote resistance to kinase inhibitors in vitro. Fourteen out of fifteen patients had no history of ruxolitinib therapy, suggesting that these mutations arise during the course of disease progression in the absence of therapeutic selection pressures. The impact of GNB1/GNAS mutations on the clinical response to JAK inhibitors has not been evaluated.

Twenty-eight patients in the cohort (41%) had no evidence of active myeloid disease, including 14 patients with a lymphoid malignancy [mature B-cell neoplasms (n=8), mature T cell neoplasms (n=3), plasma cell neoplasms (n=2), acute lymphoblastic leukemia (n=1)], 6 patients with AML in complete remission, and 8 patients without a clinical diagnosis of hematologic malignancy. Four of the AML patients had a GNAS or GNB1 mutation that was newly acquired or persistent at the time of complete remission following induction chemotherapy. All 4 patients underwent allogeneic stem cell transplant (HSCT) based on disease risk at diagnosis and are alive with median follow up of 2.5 years. In 1 patient, a new GNB1 mutation was identified immediately after HSCT in the context of 100% donor chimerism, suggesting donor-engrafted CHIP. All 8 patients without hematologic malignancy were evaluated for cytopenias or cytoses at the time of the identification of the GNB1/GNAS mutation.

GNB1 and GNAS are recurrently mutated in diverse clinical and genetic contexts. In patients with active myeloid disease, MPN-associated mutations were common and GNB1/GNAS mutations were newly acquired at the time of clinical progression, suggesting a functional role similar to other gene mutations that activate mitogenic signaling pathways. However, many GNB1 and GNAS mutations are seen as sole mutations in the absence of active myeloid disease, suggesting that unlike other signaling mutations they can drive initiation of clonal hematopoiesis. The diverse clinical spectrum of patients with GNB1/GNAS mutations suggests that these mutations have distinct, context dependent effects in hematopoietic cells.


Kim:Aushon Biosciences: Consultancy; LabCorp, Inc.: Consultancy; Papgene, Inc: Consultancy. Lane:N-of-one: Consultancy; Stemline Therapeutics: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.