The optimal duration of treatment with vitamin K antagonists (VKA) after venous thromboembolism (VTE) in patients with Philadelphia-negative myeloproliferative neoplasms (MPNs) is uncertain. To tackle this issue, we retrospectively studied 206 patients with MPN-related VTE (deep venous thrombosis of the legs and/or pulmonary embolism). After this index event, we recorded over 695 pt-years 45 recurrences, venous in 36 cases, with an incidence rate (IR) of 6.5 per 100 pt-years (95% confidence interval (CI): 4.9-8.6). One hundred fifty-five patients received VKA; the IR of recurrent thrombosis per 100 pt-years was 4.7 (95% CI: 2.8-7.3) on VKA and 8.9 (95% CI: 5.7-13.2) off VKA (P=0.03). In patients receiving VKA, the IR of recurrent thrombosis per 100 pt-years was 5.3 (95% CI: 3.2-8.4) among 108 patients on long-term VKA and 12.8 (95% CI: 7.3-20.7) after discontinuation among the 47 who ceased treatment (P=0.008), with a doubled risk of recurrence after stopping VKA (hazard ratio: 2.21, 95% CI: 1.19-5.30). The IR of major bleeding per 100 pt-years was 2.4 (95%: CI: 1.1-4.5) on VKA and 0.7 (95% CI: 0.08-2.5) off VKA (P=0.08). In conclusion, in MPN patients with VTE recurrent thrombosis is significantly reduced by VKA and caution should be adopted in discontinuation; however, the incidence of recurrence on treatment remains high, calling for clinical trials aimed to improve prophylaxis in this setting.

We retrospectively studied 181 patients with polycythaemia vera (n=67), essential thrombocythaemia (n=67) or primary myelofibrosis (n=47), who presented a first episode of splanchnic vein thrombosis (SVT). Budd–Chiari syndrome (BCS) and portal vein thrombosis were diagnosed in 31 (17.1%) and 109 (60.3%) patients, respectively; isolated thrombosis of the mesenteric or splenic veins was detected in 18 and 23 cases, respectively. After this index event, the patients were followed for 735 patient years (pt-years) and experienced 31 recurrences corresponding to an incidence rate of 4.2 per 100 pt-years. Factors associated with a significantly higher risk of recurrence were BCS (hazard ratio (HR): 3.03), history of previous thrombosis (HR: 3.62), splenomegaly (HR: 2.66) and leukocytosis (HR: 2.8). Vitamin K-antagonists (VKA) were prescribed in 85% of patients and the recurrence rate was 3.9 per 100 pt-years, whereas in the small fraction (15%) not receiving VKA more recurrences (7.2 per 100 pt-years) were reported. Intracranial and extracranial major bleeding was recorded mainly in patients on VKA and the corresponding rate was 2.0 per 100 pt-years. In conclusion, despite anticoagulation treatment, the recurrence rate after SVT in myeloproliferative neoplasms is high and suggests the exploration of new avenues of secondary prophylaxis with new antithrombotic drugs and JAK-2 inhibitors.

Chronic myeloid leukemia (CML) patients in sustained “deep molecular response” may stop TKI treatment without disease recurrence; however, half of them lose molecular response shortly after TKI withdrawing. Well-defined eligibility criteria to predict a safe discontinuation up-front are still missing. Relapse is probably due to residual quiescent TKI-resistant leukemic stem cells (LSCs) supposedly transcriptionally low/silent and not easily detectable by BCR-ABL1 qRT-PCR. Bone marrow Ph+ CML CD34+/CD38− LSCs were found to specifically co-express CD26 (dipeptidylpeptidase-IV). We explored feasibility of detecting and quantifying CD26+ LSCs by flow cytometry in peripheral blood (PB). Over 400 CML patients (at diagnosis and during/after therapy) entered this cross-sectional study in which CD26 expression was evaluated by a standardized multiparametric flow cytometry analysis on PB CD45+/CD34+/CD38− stem cell population. All 120 CP-CML patients at diagnosis showed measurable PB CD26+ LSCs (median 19.20/μL, range 0.27–698.6). PB CD26+ LSCs were also detectable in 169/236 (71.6%) CP-CML patients in first-line TKI treatment (median 0.014 cells/μL; range 0.0012–0.66) and in 74/112 (66%), additional patients studied on treatment-free remission (TFR) (median 0.015/μL; range 0.006–0.76). Notably, no correlation between BCR-ABL/ABLIS ratio and number of residual LSCs was found both in patients on or off TKIs. This is the first evidence that “circulating” CML LSCs persist in the majority of CML patients in molecular response while on TKI treatment and even after TKI discontinuation. Prospective studies evaluating the dynamics of PB CD26+ LSCs during TKI treatment and the role of a “stem cell response” threshold to achieve and maintain TFR are ongoing.

Background After discontinuing ruxolitinib, the outcome of patients with myelofibrosis reportedly has been poor. The authors investigated whether disease characteristics before the receipt of ruxolitinib may predict drug discontinuation in patients with myelofibrosis and whether reasons for drug discontinuation, disease phase at discontinuation, and salvage therapies may influence the outcome. Methods A centralized electronic clinical database was created in 20 European hematology centers, including clinical and laboratory data for 524 patients who received ruxolitinib for myelofibrosis. Results At 3 years, 40.8% of patients had stopped ruxolitinib. Baseline predictors of drug discontinuation were: intermediate‐2–risk/high‐risk category (Dynamic International Prognostic Score System), a platelet count <100 ×109 per liter, transfusion dependency, and unfavorable karyotype. At last contact, 268 patients (51.1%) had discontinued therapy, and the median drug exposure was 17.5 months. Fifty patients (18.7%) died while taking ruxolitinib. The reasons for discontinuation in the remaining 218 patients were the lack (22.9%) or loss (11.9%) of a spleen response, ruxolitinib‐related adverse events (27.5%), progression to blast phase (23.4%), ruxolitinib‐unrelated adverse events (9.2%), and allogeneic transplantation during response (5.1%). The median survival after ruxolitinib was 13.2 months and was significantly better in the 167 patients who discontinued ruxolitinib in chronic phase (27.5 vs 3.9 months for those who discontinued in blast phase; P < .001). No survival differences were observed among patients who discontinued ruxolitinib in chronic phase because of lack of response, loss of response, or ruxolitinib‐related adverse events. The use of investigational agents and/or ruxolitinib rechallenge were associated with improved outcome. Conclusions The survival of patients with myelofibrosis after discontinuation of ruxolitinib is poor, particularly for those who discontinue in blast phase. Salvage therapies can improve outcome, emphasizing the need for novel therapies.

Abstract Essential thrombocythemia (ET) is characterized by abnormal megakaryopoiesis and enhanced thrombotic risk. Once-daily low-dose aspirin is the recommended antithrombotic regimen, but accelerated platelet generation may reduce the duration of platelet cyclooxygenase-1 (COX-1) inhibition. We performed a multicenter double-blind trial to investigate the efficacy of 3 aspirin regimens in optimizing platelet COX-1 inhibition while preserving COX-2–dependent vascular thromboresistance. Patients on chronic once-daily low-dose aspirin (n = 245) were randomized (1:1:1) to receive 100 mg of aspirin 1, 2, or 3 times daily for 2 weeks. Serum thromboxane B2 (sTXB2), a validated biomarker of platelet COX-1 activity, and urinary prostacyclin metabolite (PGIM) excretion were measured at randomization and after 2 weeks, as primary surrogate end points of efficacy and safety, respectively. Urinary TX metabolite (TXM) excretion, gastrointestinal tolerance, and ET-related symptoms were also investigated. Evaluable patients assigned to the twice-daily and thrice-daily regimens showed substantially reduced interindividual variability and lower median (interquartile range) values for sTXB2 (ng/mL) compared with the once-daily arm: 4 (2.1-6.7; n = 79), 2.5 (1.4-5.65, n = 79), and 19.3 (9.7-40; n = 85), respectively. Urinary PGIM was comparable in the 3 arms. Urinary TXM was reduced by 35% in both experimental arms. Patients in the thrice-daily arm reported a higher abdominal discomfort score. In conclusion, the currently recommended aspirin regimen of 75 to 100 once daily for cardiovascular prophylaxis appears to be largely inadequate in reducing platelet activation in the vast majority of patients with ET. The antiplatelet response to low-dose aspirin can be markedly improved by shortening the dosing interval to 12 hours, with no improvement with further reductions (EudraCT 2016-002885-30).

We conducted a large international nested case-control study including 1,881 patients with Philadelphia-negative myeloproliferative neoplasms (MPN). Cases (n=647) were patients with second cancer (SC: carcinoma, non-melanoma skin cancer, hematological second cancer and melanoma) and controls (n=1,234) were patients without SC, matched with cases for sex, age at MPN diagnosis, date of MPN diagnosis and MPN disease duration. The aim was to evaluate the risk of SC after exposure to cytoreductive drugs. Patients exposed to hydroxyurea (HU) (median: 3 years) had a risk of SC similar to unexposed patients (OR=1.06, 95% CI 0.82-1.38). In contrast, in cancer-specific stratified multivariable analysis, HU had twofold higher risk of non-melanoma (NM) skin cancer (OR=2.28, 95% CI 1.15-4.51). A significantly higher risk of NM-skin cancer was also documented for pipobroman (OR=3.74, 95% CI 1.00-14.01), ruxolitinib (OR=3.87, 95% CI 1.18-12.75) and for drug combination (OR=3.47, 95% CI 1.55-7.75). These three drugs did not show excess risk of carcinoma and hematological second cancer compared with unexposed patients. Exposure to interferon, busulfan and anagrelide did not increase the risk. In summary, while it is reassuring that no excess of carcinoma was documented, a careful dermatologic active surveillance before and during the course of treatments is recommended.

We report the clinical presentation and risk factors for survival in 175 patients with myeloproliferative neoplasms (MPN) and COVID-19, diagnosed between February and June 2020. After a median follow-up of 50 days, mortality was higher than in the general population and reached 48% in myelofibrosis (MF). Univariate analysis, showed a significant relationship between death and age, male gender, decreased lymphocyte counts, need for respiratory support, comorbidities and diagnosis of MF, while no association with essential thrombocythemia (ET), polycythemia vera (PV), and prefibrotic-PMF (pre-PMF) was found. Regarding MPN-directed therapy ongoing at the time of COVID-19 diagnosis, Ruxolitinib (Ruxo) was significantly more frequent in patients who died in comparison with survivors (p = 0.006). Conversely, multivariable analysis found no effect of Ruxo alone on mortality, but highlighted an increased risk of death in the 11 out of 45 patients who discontinued treatment. These findings were also confirmed in a propensity score matching analysis. In conclusion, we found a high risk of mortality during COVID-19 infection among MPN patients, especially in MF patients and/or discontinuing Ruxo at COVID-19 diagnosis. These findings call for deeper investigation on the role of Ruxo treatment and its interruption, in affecting mortality in MPN patients with COVID-19.

It is judged safe to discontinue treatment with tyrosine kinase inhibitors (TKI) for chronic myeloid leukemia (CML) in experimental trials on treatment-free remission (TFR). We collected a total of 293 Italian patients with chronic phase CML who discontinued TKI in deep molecular response. Seventy-two percent of patients were on treatment with imatinib, and 28% with second generation TKI at the time of discontinuation. Median duration of treatment with the last TKI was 77 months [Interquartile Range (IQR) 54;111], median duration of deep molecular response was 46 months (IQR 31;74). Duration of treatment with TKI and duration of deep molecular response were shorter with second generation TKI than with imatinib ( P <0.001). Eighty-eight percent of patients discontinued as per clinical practice, and reasons for stopping treatment were: toxicity (20%), pregnancy (6%), and shared decision between treating physician and patient (62%). After a median follow up of 34 months (range, 12-161) overall estimated TFR was 62% (95%CI: 56;68). At 12 months, TFR was 68% (95%CI: 62;74) for imatinib, 73% (95%CI: 64;83) for second generation TKI. Overall median time to restart treatment was six months (IQR 4;11). No progressions occurred. Although our study has the limitation of a retrospective study, our experience within the Italian population confirms that discontinuation of imatinib and second generation TKI is feasible and safe in clinical practice.