Ruxolitinib is the only commercially available JAK2-inhibitor approved as therapy for myelofibrosis (MF). The registrative COMFORT (Controlled Myelofibrosis Study With Oral JAK Inhibitor Treatment) studies (Harrison et al, 2012; Verstovsek et al, 2012) enrolled only patients with palpable splenomegaly and demonstrated the clinical efficacy of ruxolitinib on splenomegaly and constitutional symptoms in this setting. The outcome of six patients treated with ruxolitinib without palpable splenomegaly were reported in 2012 and showed a significant clinical benefit (Benjamini et al, 2012). Since then, no additional similar experiences have been reported, and little is known on the clinical activity of ruxolitinib in this subset of patients. Between June 2011 and January 2015, 103 MF patients were followed in three Italian Haematology centres (Bologna, Roma, Torino) and treated with ruxolitinib. Thirteen (12Á6%) patients did not have splenomegaly at baseline but were burdened by constitutional symptoms (Table I). Median treatment duration was 24 months (range, 6–36). Within the first month of ruxolitinib therapy, all patients reported a clinical benefit on disease-related symptoms, specifically fatigue, night sweats, itching, and bone pain. Notably, titrated doses higher than 10 mg BID achieved better responses compared to lower doses. All patients showed weight gain (progressive weight loss was the ultimate cause for treatment decision in Patients 1 and 6: both recovered their ideal weight). Also, four patients (Patients 1, 2, 3 and 4) resumed their normal work activity during therapy. No patients developed hepatomegaly over the follow-up; in Patient 6, who presented mild hepatomegaly (3 cm below costal margin by palpation) after prior splenectomy, liver was not palpable after 6 months of therapy. Patients with marked thrombocytosis (Patients 1, 3, 4 and 10) normalized their blood counts. JAK2 V617F/CALR allele burden was evaluated at the start of treatment and at 12-month intervals: one case (Patient 5) showed a significant reduction (À71%) of mutation load. Treatment was generally well tolerated. No significant (grade 3 or 4) non-haematological adverse events were observed, although three patients experienced a total of 5 infectious complications requiring oral antibiotics (2 bronchitis, 1 pneumonia, 1 urinary tract infection). At ruxolitinib start, five patients were receiving red blood cell transfusions; in all these patients, transfusion need remained stable during treatment and one patient (Patient 3) became transfusion-independent after 2 months of therapy. Two patients (Patietns 5 and 6) eventually discontinued ruxolitinib due to persistent anaemia and thrombocytopenia, which did not allow an effective titrated dose to be maintained. After treatment discontinuation, all patients experienced a mild rebound of constitutional symptoms; in one case, cytopenias recovered. No patient progressed to acute myeloid leukaemia, either during or after ruxolitinib discontinuation; this is particularly relevant in Patient 6, who had a very high leucocyte count and complex cytogenetics. This experience confirms a positive effect of ruxolitinib on patient symptoms not related to reduction of spleen size in MF.