JAK2 V617F and ringed sideroblasts: not necessarily RARS-T

To the editor:

In early 2005, shortly after the first recognition of JAK2 V617F mutations in Philadelphia-chromosome negative myeloproliferative disorders (MPD),¹  we reported in Blood²  that patients with myelodysplastic syndromes (MDS) and atypical MPD sometimes have JAK2 V617F, too, albeit uncommonly. These findings were quickly confirmed by others.^3,4 

Among the 101 MDS patients in our original cohort, we detected JAK2 mutations by DNA sequencing in 5 patients, and 2 of those 5 were reported as having refractory anemia with ringed sideroblasts (RARS).²  Since our 2005 publication, many readers have requested additional details about those 2 patients. This interest has been driven by the discovery that JAK2 V617F is detectable in up to 67% of patients meeting World Health Organization diagnostic criteria for the provisional entity RARS with thrombocytosis (RARS-T), an MDS-MPD overlap syndrome.^5-9  However, neither of our 2 patients with JAK2 V617F-positive RARS ever exhibited thrombocytosis or any other myeloproliferative features.

The first patient developed mild anemia at age 47, and was found to have a hypercellular marrow with mild dyserythropoiesis and “abundant” ringed sideroblasts. He was followed without specific therapy for 24 years; the platelet count was always within the normal range. He first presented to our institution at age 71 with a hemoglobin of 92 g/L (9.2 g/dL), mean corpuscular volume (MCV) 114 fL, white count of 4.4 × 10⁹/L with an unremarkable differential, and a platelet count of 371 × 10⁹/L. There was no organomegaly. The marrow was 95% cellular with 50% ringed sideroblasts, erythroid-predominant multilineage dysplasia without any fibrosis or megakaryocyte clustering, less than 5% blasts, and 46, XY karyotype. A research sample was obtained and later tested positive for JAK2 V617F. Multiple therapies were attempted, but all proved ineffective, and the patient died at age 79 of congestive heart failure.

The second patient presented at age 61with mild anemia and dimorphic erythrocyte morphology, shortly after completing radiotherapy for localized angiosarcoma of the vulva. Her initial bone marrow aspirate was normocellular and considered nondiagnostic, though 10% ringed sideroblasts were seen. Four years later, her hemoglobin had fallen to 70 g/L (7.0 g/dL) with MCV of 94 fL; her white count was 3.6 × 10⁹/L with an unremarkable differential, and platelet count was 153 × 10⁹/L. Organomegaly was always absent. Her marrow was 80% cellular with 60% ringed sideroblasts, striking dyserythropoiesis with normal granulopoiesis and rare hypolobated megakaryocytes, less than 5% blasts, and absent reticulin. Karyotype was 46, XX, and JAK2 V617F was present. After multiple treatments failed, the patient was maintained with transfusion support alone, and she died of congestive heart failure at age 70.

Our collective understanding of the diagnostic utility and pathophysiologic role of JAK2 V617F in myeloid disorders has evolved rapidly since 2005, but these 2 patients—both of whom had a mutant allele burden more than 30%—illustrate that much is not yet cut-and-dried. The significance of the JAK2 V617F clone in these 2 cases is unclear, because neither showed any hint of proliferative features. Just as JAK2 mutations do not alone make a diagnosis of MPD, and ringed sideroblasts do not necessarily mean RARS (or MDS), JAK2 V617F plus ringed sideroblasts does not automatically imply RARS-T.⁷  For instance, we and others^7,10  have cared for patients with a clinical picture otherwise consistent with primary myelofibrosis or essential thrombocythemia, who have also had some ringed sideroblasts - often not enough to meet the arbitrary 15% threshold necessary for a diagnosis of RARS or RARS-T, but present nevertheless.

Congenital sideroblastic anemia syndromes can be caused by diverse mutations that impair heme biosynthesis or mitochondrial iron processing (eg, ALAS2, ABCB7, and others).¹¹  It seems likely that the presence of ringed sideroblasts in MDS/MPD will eventually be linked to acquired, clonally-restricted mutations with similar effects.¹²  As with JAK2 V617F, these acquired mutations may not prove to be disease-defining, but might instead act as phenotype-driving clonal markers, with the specific phenotype dependent on the particular mutation repertoire.

Ringed sideroblasts are important because they provide an unambiguous marker of abnormal erythropoiesis. But their significance should not be overstated. As the JAK2 story has reminded us, little in hematology, or biology, is so straightforward.