The primary objective of the study is to assess efficacy and safety of different prophylactic or therapeutic antithrombotic approaches in patients with hematologic neoplasms and platelet count <50 x109/L, including unfractionated or low molecular weight heparin, fondaparinux, anti-vitamin K agents, antiplatelet agents, novel oral anticoagulants, fibrinolytic agents, with or without a policy of platelet transfusion. Cases with arterial or venous thromboembolism managed with observation or use of vena cava filters in patients with venous thromboembolism will be included too.

Illness treatment overview The incidence of venous thromboembolism (VTE) among patients with haematological malignancies has been recently reviewed (1). For patients with lymphoma, the incidence of VTE ranged from 1.5 to 14.6% and is 59% in patients with central nervous system lymphoma. The incidence of VTE in patients with acute leukaemia varies with time, being between 1.4 and 9.6% at diagnosis and between 1.7 and 12% during induction therapy. Notably, the highest rates of VTE were reported in patients with acute promyelocytic leukaemia, with values between 6 and 16% in the largest series. In patients with multiple myeloma who did not receive antithrombotic prophylaxis, the rate of VTE increased to 26% in those undergoing treatment regimens including immumodulatory drugs (thalidomide and lenalidomide) (1). Moreover, in patients who have undergone autologous or allogeneic haematopoietic stem cell transplantation (HSCT), the rate of VTE is between 2.9 to 9.9%, and was central venous catheter (CVC)-associated in the majority of cases (2-4). VTE has been shown to occur even during periods of severe thrombocytopenia; one-third of the events from a patient series occurred when the platelet count was <50 x109/L (3). Patients with acute leukaemia have a high risk of haemorrhage, mostly related to thrombocytopenia as a result of haematological disease and/or chemotherapy, such that the administration of anticoagulant drugs in this setting is problematic. No randomised controlled trials have addressed the issue of VTE treatment in patients with acute leukaemia, and the management of these cases is based only on small groups of patient series or experts' opinion. Five reports have described in detail cases of paediatric or adult patients with malignancies and thrombocytopenia who have been treated with heparin as a result of a VTE (5-9). Out of a total of 54 cases, 32 had haematological malignancies and a platelet nadir <50 x109/L (reviewed in ref. 10). The majority of patients (22 of 32) had CVC-related thrombosis. A full dose of low molecular weight heparin (LMWH) bid was administered to the majority of the patients with CVC-related thrombosis and to the totality of patients for whom the thrombosis was not CVC-related. All patients received platelet transfusions if the platelet count fell below <20-40 x109/L, and the dose of LMWH was halved when the platelet count was <20 x109/L in one report. None of the patients had major bleeding. Rethrombosis occurred in three of 32 (9.3%) patients (6,7). In a large series of 379 patients with acute leukaemia, treatment for 20 patients who had one (n=16) or two (n= 4) VTE events was essentially based on the administration of enoxaparin 100 U/kg bid; in the case of a platelet count <50 x109/L or in the clinical suspicion of bleeding risk the dose was reduced to 100 U/kg qd or 50 U/kg bid. Alternatively, the patients received a continuous i.v. infusion of unfractionated heparin (UFH) to obtain aPTTs in the lower therapeutic range (1.5 times greater than the basal value). Secondary prophylaxis after acute VTE was based on the administration of enoxaparin 100 U/kg qd in the case of ongoing chemotherapy or vitamin K-antagonists (VKA) (INR between 2 and 3) otherwise. In general, the length of secondary prophylaxis was reported to be not longer than six months (11). The safety of therapeutic anticoagulation treatment for the management of thrombocytopenic patients was also evaluated in two series of patients receiving HSCT (12,13). In 10 patients with multiple myeloma who had received autologous HSCT, anticoagulation was required following pre-transplant CVC-related subclavian vein thrombosis (n= 8), pulmonary embolism two months prior to transplant (n= 1), or a history of acute intermittent atrial fibrillation that was complicated by an arterial embolus to the leg (n= 1). Beginning on the first day of high-dose chemotherapy, the 10 patients received therapeutic UFH (a 5,000 U i.v. bolus followed by 1,000 U per h) to maintain aPTTs between 50 and 70 seconds and were switched to VKA treatment when their conditions stabilised. UFH treatment was interrupted once the VKA administration produced a therapeutic INR >2 for two consecutive days. Heparinised patients received platelet transfusions to maintain counts >30 x109/L. Three patients developed bleeding (haematuria, haematemesis, mucosal bleeding) that did not not require transfusion, and no thrombotic events occurred (12). In another series of 26 patients with HSCT who were given enoxaparin for the treatment of VTE, 21 patients had haematological malignancies. There were 25 VTE events recorded (four patients had two events at different sites) and 11 cases had upper extremity CVC-related deep venous thrombosis. During periods of thrombocytopenia (<55 x109/L), enoxaparin administration was reduced to a median value of 49 U/kg/day (range 34-75) and was withdrawn in some instances when the platelet count fell below 20 x109/L. Two major bleeding events (8%) occurred, in one case fatal (13). The aforementioned data are insufficient for the production of evidence-based guidelines. Experts and the AIEOP (Associazione Italiana di Ematologia e Oncologia Pediatrica) have suggested that the first two weeks of treatment should consist of the administration of full-dose LMWH (anti-factor Xa level 0.5-1 U/ml), maintaining the platelet count above 50 x109/L. After the first two weeks, halving the dose is recommended if the platelet count is between 20 and 50 x109/L. If the platelet count is below 20 x109/L, it is advised that the LMWH therapy be discontinued until the platelet count recovers to greater than 20 x109/L (14-16). Recent guidelines of the SISET (Società Italiana per lo Studio dell'Emostasi e della Trombosi) also suggested that in patients with haematological malignancies and VTE LMWH should be preferred over VKA either for the first six months or a longer time (17). Arterial thrombosis has been rarely reported in patients with acute leukemia as heralding manifestation or complicating the course of disease (18-21). However, details concerning antithrombotic treatment and the platelet count at the time of the event are lacking in the majority of the reported cases. There have been no published studies concerning the use of novel antithrombotic agents such fondaparinux, direct thrombin or factor Xa inhibitors, in patients with haematological malignancies or thrombocytopaenia. However, in vitro experiments that have been performed on plasma from children with ALL and antithrombin deficiency as a result of the administration of asparaginase have demonstrated that the direct thrombin inhibitor melagatran generates a consistent anticoagulant response that is independent of the antithrombin level. Therefore, this drug class may have important potential for use in this field (22).

Rationale Data about treatment of arterial or venous thromboembolism in patients with haematological malignancies and thrombocytopenia are mainly anecdotal. The very limited knowledge in this setting does not allow to plan a randomized controlled trial, lacking a standard of care and being quite uncertain the benefits and risks of different strategies. Therefore we planned an observational study either retrospective and prospective to gain information about efficacy and safety of different therapeutic strategies in patients with hematologic neoplasms and platelet count <50 x109/L having had diagnosis of arterial or venous thromboembolism.

Inclusion Criteria:

Potential subjects must satisfy all of the following criteria to be enrolled in the study:

• diagnosis of hematologic neoplasm (acute leukemia, myelodysplastic syndrome, lymphoma, multiple myeloma, chronic myeloid leukemia, Ph-negative chronic myeloproliferative neoplasms) independently of the stage of disease or treatment (including transplant procedures);
• platelet count <50 x109/L at the time of starting antithrombotic prophylaxis or
• platelet count <50 x109/L at the time of diagnosis of arterial or venous thromboembolism objectively proven or
• platelet count >50 x109/L at time of thrombosis but subsequent thrombocytopenia <50 x109/L while receiving antithrombotic treatment;
• diagnosis of arterial thrombosis include acute coronary syndrome, ischemic stroke (including major and minor stroke), peripheral arterial thrombosis, retinal arterial thrombosis;
• diagnosis of venous thrombosis include thrombosis of deep veins of the limbs and the abdomen, superficial veins of limbs, cerebral and splanchnic veins, retinal vein, and pulmonary embolism. Splanchnic venous thrombosis include occlusion of hepatic, portal, mesenteric, and splenic veins.

Exclusion Criteria:

The following situations will not be criteria of inclusion neither outcomes of interest:

• transient ischemic attack without CT and/or NMR signs;
• superficial vein thrombosis without Doppler ultrasound examination showing evidence of thrombosis;
• antithrombotic prophylaxis only local for central venous lines (i.e. CVC flushing with heparin);
• occlusion of the central venous catheter (notice that CVC-related deep venous thrombosis, i.e. thrombosis of the deep veins where the central line is placed, will be a criterion of inclusion or an outcome of antithrombotic prophylaxis).

Because this study is designed to provide as wide a picture of the different clinical management strategies in unselected patients with hematologic neoplasms, inclusion criteria are set deliberately wide.

All consecutive subjects that present to the centre and satisfy the inclusion criteria will be considered as potential candidates for enrolment.

There is no age limit for including the patients in the study.