SF3B1 mutation

What is measured: a hotspot mutation in the splicing factor SF3B1 (K700E most often, also K666, R625, H662, E622) and its marrow correlate, ring sideroblasts. How: a myeloid sequencing panel on marrow or blood, and an iron (Perls) stain counting ring sideroblasts, which WHO 2016 required at 15 percent (or 5 percent with the mutation) and WHO 2022 no longer requires, defining MDS with low blasts and SF3B1 mutation directly. Frequencies: 25 to 30 percent of MDS (some 80 percent of those with ring sideroblasts), RARS-T with JAK2, about 10 percent of CLL (adverse, often with ATM loss), about 20 percent of uveal melanoma (R625, associated with late metastases), mucosal melanoma, and rarely breast and pancreatic cancers. What a result changes: in MDS the IPSS-M treats an isolated SF3B1 mutation as favourable but not when it comes with del(5q) or other adverse mutations; the anaemia is managed with erythropoiesis-stimulating agents and then luspatercept (transfusion independence in 38 percent against 13 percent on placebo in MEDALIST, and superiority over epoetin first line in COMMANDS) or imetelstat, with iron chelation for overload; splicing modulators such as H3B-8800 failed in the clinic; in CLL it feeds prognostic models without a matched drug; in uveal melanoma it flags a class 1B tumour that needs longer surveillance. Where it matters: lower-risk MDS, CLL, uveal and mucosal melanoma, secondary AML and CMML.