Introduction

Major bleeding is one of the most feared, frequent and potentially life-threatening complication of anticoagulant therapy for venous thromboembolism (VTE). Previous data demonstrated a case fatality rate of 9% [1]. Unfortunately, cancer patients have an increased risk of anticoagulation-related bleeding when compared to non-cancer patients [2,3,4]. With the advent of new therapies, the incidence of major bleeding ranges from 6.5 to 18%, and it is not always explained by supratherapeutic levels of anticoagulant therapy [5]. The rationale is suspected to be multifactorial related to hematologic aberrancies, potentially cancer-related or therapy-related (drug interactions, dysfibrinogenemia), direct vascular invasion resulting in increased vascular fragility, bone-marrow suppression, tissue damage induced by radiation, and others [6, 7]. Unfortunately, VTE events are highly prevalent in the cancer population, and it portends significant morbidity and mortality. Given the prothrombotic nature of cancer and certain therapies, anticoagulation is often necessary in either a therapeutic or prophylactic manner.

As the population ages, there is a considerable increase in the use of antithrombotic agents in the management of atrial fibrillation or other ischemic events. Thus, many cancer patients may already be on anticoagulant or antiplatelet therapies. Use of these agents can contribute to delays in invasive interventions needed for diagnosis, and disruption to therapy can increase the risk of thromboembolic events. So, management of these agents presents both diagnostic and therapeutic dilemmas for the clinician. Our review will focus on considerations in cancer patients prior to initiation of anticoagulation, management of bleeding in an anticoagulated patient, and considerations for both cessation and re-initiation of therapy.

Considerations prior to anticoagulation

Clinical guidelines propose assessment of bleeding risk prior to initiation of anticoagulation [2, 8, 9]. Specifically, the clinician should identify any risk factors that may increase the risk of bleeding, select the optimal anticoagulant therapy, determine the ideal dosage, and identify the optimal treatment duration [2]. Close follow-up for modifiable risk factors including hepatic and renal function, hemoglobin and platelet levels, blood pressure, drug adherence, and use of other potentially interacting medications is imperative to minimize bleeding risk. For example, a decline in the estimated glomerular filtration rate, even to a small degree, has been associated with an increased risk of bleeding [10, 11]. Moreover, advancing stages of hypertension during antithrombotic medication are also associated with higher incidences of intracranial bleeding and bleeding in general [12, 13]. In addition, meticulous observation for the presence or development of absolute contraindications including invasive intervention, active intracranial or other life-threatening bleeding, and thrombocytopenia (< 25 × 109/L) is needed [2]. In some cases, use of a retrievable inferior vena cava (IVC) filter, serial platelet transfusion, or substitution of anticoagulant therapy may be alternative strategies. Resuming anticoagulation after the contraindication has resolved requires careful reassessment of risks and shared decision-making with the patient, caregivers, and providers.

Low molecular weight heparin (LMWH) had been the standard of care to both treat and prevent cancer-associated VTE over warfarin, a vitamin K antagonist (VKA). Major landmark trials (CANTHANOX, ONCENOX, CLOT, LITE, CATCH) have shown their superiority in preventing recurrent VTE with an improved bleeding profile [3, 14,15,16]. The emergence of direct oral anticoagulants (DOACs) has changed treatment paradigms, and trials have consistently shown non-inferiority or even superiority with regard to both efficacy and safety profiles [17,18,19,20,21]. The DOACs are a generally well tolerated oral option. Some populations however are not ideal for DOACs. Specifically, a few major randomized controlled trials, including HOKUSAI-VTE (edoxaban versus dalteparin), SELECT-D (rivaroxaban versus dalteparin), and ADAM-VTE (apixaban versus dalteparin), have identified cancer populations at increased risk for bleeding including those with gastrointestinal and genitourinary malignancies [18, 21,22,23,24]. In addition, use of a DOAC should be reconsidered in those with poor compliance, significant nausea and vomiting, or renal insufficiency. A sub-analysis of the CATCH trial between patients with and without renal failure noted a higher incidence of VTE recurrence and major bleeding in patients with renal impairment (defined as creatinine clearance < 60 mL/min) [21, 24]. Current guidelines advise dose adjustment of the DOAC for renal function, and in cases of progressive renal impairment, alternative anticoagulant therapy may be needed as all DOACs were not specifically studied in patients with a creatinine clearance < 30 mL/min [4, 14].

Thrombocytopenia is another challenge in cancer patients with VTE both in terms of incidence and management (Fig. 1). Interestingly both VTE and its recurrence is common in adult acute leukemia patients, and risk factors include catheter thrombosis, prior history of hematologic cancer, chronic lung disease, psychological disorder, and liver disease [25]. Anticoagulation in patients with platelets less than 50 × 109/L is permissible, but if the platelet count continues to decrease to below 25 × 109/L, then anticoagulation should be held [4, 26,27,28]. When the platelet count is less than 50 × 109/L, dose-reduced LMWH is the preferred alternative since safety data is not available for DOACs. Of note, caution in those with other underlying disease states prone to bleeding is also recommended, and these include liver disease or hypofibrinogenemia states including leukemia, disseminated intravascular coagulation (DIC), and other inherited and acquired disorders.

Fig. 1
figure 1

Right chest wall hematoma with bloody pleural effusion. Middle aged man with chronic myelogenous leukemia admitted with septic shock and atrial fibrillation on DOAC (A, initial CT scan, *diaphragm). During his hospital course, he developed swelling in his right chest without any trauma. Laboratory data revealed anemia (5.6 g/dL), worsened thrombocytopenia (52 K/mL), creatinine of 1.14 g/dL, and a normal coagulation profile. Imaging confirmed chest wall hematoma over the course of 7 days (B, yellow arrowhead identifies chest wall hematoma, white arrows show pleural effusion; yellow asterisk (*), diaphragm lowered due pleural effusion). Anticoagulation was stopped and blood products transfused. Bloody pleural effusion was drained and did not reaccumulate

Full size image

Brain tumors (Fig. 2) also have an increased thrombotic risk as well as varying propensities for spontaneous hemorrhage [29]. A meta-analysis estimated the risk of VTE to be 1.5 to 2.0% per month of survival in high-grade gliomas. Risk factors are patient-related (age, immobility, medical comorbidities), treatment-related (recent surgery, antiangiogenic therapy, corticosteroids), and tumor-related (tumor size, glioblastoma subtype, intra-tumoral characteristics) [30, 31]. In general, high-grade gliomas, pituitary adenomas and brain metastases from melanoma, choriocarcinoma, thyroid, and renal cell carcinoma all have an increased risk of spontaneous hemorrhage, whereas benign (or low grade) tumors and brain metastases from other primary tumors (such as breast) are lower risk [32, 33]. Furthermore, a more recent meta-analysis revealed a threefold increased risk in intracranial hemorrhage (ICH) in those with primary malignant brain tumors on anticoagulation, and this risk was increased even further in the setting of supratherapeutic anticoagulation or inadequate preoperative correction of coagulation abnormalities [14, 34, 35]. Thus, balancing the risks of thrombosis and hemorrhage is essential. The selection of agent between LMWH vs DOACs remains unclear, for with the exception of the Hokusai VTE-cancer study, most trials did not include brain tumor patients [23]. The American Society of Clinical Oncology guidelines recommend anticoagulation should be offered but acknowledge uncertainty in choice of agent due to limited safety data with DOACs; however, some treat with DOACs given the availability of an effective reversal agent [14, 31, 34]. A multidisciplinary approach with close clinical follow-up would be recommended.

Fig. 2
figure 2

Intracranial hemorrhage and tumor. CT brain demonstrating subdural hematoma (A, white arrows). CT brain demonstrating intraparenchymal hemorrhage within the right basal ganglia with surrounding edema (B, white arrowheads) and right-to-left midline shift (B, white arrow). MRI with T2* weighted sequence of glioblastoma (C, asterisk) involving the peri-atrial white matter and corpus callosum with hemorrhagic tumor necrosis (C, white arrowheads) related to bevacizumab. Saddle pulmonary embolism (D, white arrow highlighting clot in right and left pulmonary arteries)

Full size image

The potential for drug interactions (with VKAs and DOACs) can impact concomitant treatments and must always be kept in mind. The uptake of the DOACs depends on the P-glycoprotein system while also subject to being metabolized in varying degrees via the cytochrome P-450 (CYP) system [19]. Concurrent therapy with medications that have the potential to induce or inhibit these systems will likely impact the circulating drug levels. Furthermore, there are several anti-neoplastic therapies, both cytotoxic and targeted agents, that have been associated with increased thrombotic risk, and they include cisplatin, L-asparaginase, fluoropyrimidines, tamoxifen and aromatase inhibitors, antiangiogenic agents (bevacizumab, tyrosine kinase inhibitors), immunomodulatory agents (thalidomide in combination with anthracycline-based induction, lenalidomide with steroid), corticosteroids, and hematopoietic growth factors. Recognition of therapy-related increased thrombotic risk allows increased vigilance for VTE. Recent data from the AVERT and CASSINI trials showed that thromboprophylaxis with a DOAC in high-risk ambulatory patients with active cancer was effective; however, further studies to define the role for thromboprophylaxis therapy is needed [16, 36].

Finally, selection of anticoagulant therapy should also consider the patient’s level of activity, socio-economic situations, and general awareness. For example, those prone to falls or injury, with an inherent disability, or increased frailty may be at higher risk for bleeding (Fig. 3). If a procedure or surgery is planned, then the use of a LMWH in the short term followed by DOAC after the procedure may maximize the time treated without delay in procedure. Cost and insurance coverage are other important factors. Assessing the financial impact to the patient by sending a prior test claim and additional education by a pharmacist can clearly affect adherence. Pharmaceutical companies may have coupons and assistance programs for patients meeting the financial support criteria.

Fig. 3
figure 3

Left hip hematoma. Elderly woman with metastatic carcinoma to the peritoneum on LMWH presented to the emergency center after a fall. Exam revealed hematoma on her left buttock and lateral aspect of her thigh (A, CT scan with white arrows highlighting hematoma on left hip). She received reversal agent with protamine, and subsequently an IVC filter was placed. CT angiogram (B, black arrowheads indicate left hip hematoma) confirmed no active hemorrhage

Full size image

Workup

Bleeding may manifest externally or internally. For example, patients with gastrointestinal bleeding often present overtly with either hematemesis or melena, but those with musculoskeletal or retroperitoneal bleeding may report more subtle complaints such as fatigue or back pain. Significant bleeding may also present with hemodynamic instability, shock, or respiratory failure. Bleeding within a closed space such as ICH or within a small compartment (such as the spinal canal, pericardium, or orbit) can be particularly catastrophic, often necessitating reversal agents. Many trials have employed the International Society on Thrombosis and Haemostasis definition for major bleedings as follows: clinically overt bleeding accompanied by decrease in the hemoglobin level of ≥ 2 g/dL; requiring transfusion of ≥ 2 units of blood; occurring at a critical site; or resulting in death [37]. Critical sites include intracranial, intraspinal, intraocular, retroperitoneal, intraarticular, pericardial, or intramuscular with compartment syndrome (Fig. 4). Those not meeting the above criteria are considered minor, and frequently minor bleeding such as hematuria, epistaxis, or menorrhagia can typically be managed conservatively and often without interruption of anticoagulation.

Fig. 4
figure 4

Sites of bleeding with potential signs and symptoms

Full size image

Risk stratification of bleeding thus requires immediate assessment of hemodynamic stability, identification of bleeding source, and degree of blood loss [38]. A thorough history, concomitant review of other comorbid conditions (renal or hepatic impairment), medication profile, and current anti-neoplastic therapies all provide perspective for potential exacerbating or propagating factors. However, episodes of major bleeding may occur spontaneously without any triggering factors [39]. Identification of the exact dosage and regimen that the patient has been taking of the anticoagulant as well as the timing of medication ingestion is important, for in certain cases such as a recent overdosage, activated charcoal could be employed within the first hour of ingestion. Finally, the decision for reversal and/or antidote administration must also ensue, while recognizing increased risk for potential thrombotic sequelae. Clearly in cases with life-threatening or persistent bleeding (Table 1), reversal may be indicated, but in those without active bleeding, close observation and natural clearance of the agent is ideal [40].

Table 1 Indications for reversal of anticoagulant therapy [40]
Full size table

Laboratory assessment should begin with common hematologic testing including complete blood count; metabolic panel to assess liver and renal function; type and screen in case blood products are needed; and coagulation tests which include prothrombin time/international normalized ratio (PT/INR), activated partial thromboplastin time (aPTT), and fibrinogen levels [41]. Although the quantity of platelets may be helpful, qualitative function may be impacted by antiplatelet therapy or renal insufficiency. Light transmission aggregometry in platelet-rich plasma or whole blood is considered the gold standard for assessment of platelet function, but availability may be limited due to poor standardization and time consumption [42].

Originally developed to assess the function of the “extrinsic pathway,” abnormalities in PT may reflect coagulopathy seen in liver failure, disseminated intravascular coagulopathy, trauma, and medication effect. The aPTT is commonly used to assess the “intrinsic pathway” of hemostasis, and it is most useful for monitoring the effect of unfractionated heparin (UFH) but does not reliably reflect the effect of other anticoagulants [43]. The conventional coagulation tests are plasma-based. Hence, they cannot measure interactions between clotting factors, tissue factor, and platelets, so they simply reflect a static evaluation of the coagulation cascade with clot formation as their endpoint rather than assessing the whole coagulation system [43, 44]. They have also been shown to correlate poorly with clinical bleeding and transfusion requirements, lack accuracy in detecting deficiencies in coagulation factors, fail to detect the effects of DOACs or antiplatelet therapy, and do not describe platelet function and fibrinolysis. More specific tests including anti-factor Xa assay and thrombin time should be sent to assess levels and function, but they may take time to return. Although PT and INR may be elevated in DOACs, it merely indicates the patient therapy but not specific to function.

Anticoagulant and antiplatelet therapies

Anticoagulation therapy

The main classes of anticoagulant drugs include vitamin K–dependent coagulation factor antagonists (VKA, e.g., warfarin), factor Xa inhibitors (e.g., fondaparinux, rivaroxaban, apixaban), direct thrombin inhibitors (DTI, e.g., argatroban, bivalirudin, dabigatran), and heparinoids (UFH, LMWH). Reversal strategies for anticoagulation-associated major hemorrhage are summarized in Table 2. Cancer patients were not excluded from many of the trials mentioned below, and although they did not compose the majority of patients, extrapolation to bleeding management in oncologic patients would be reasonable.

Table 2 Summary of common anticoagulant therapies [45, 46]
Full size table

Vitamin K antagonists

VKAs inhibit vitamin K dependent factors in the coagulation cascade: factors II, VII, IX, and X. VKA activity can be assessed by PT/INR. The risk of bleeding while taking a VKA increases with the duration of therapy and higher INR levels. For each increment in INR elevation above the therapeutic range, the risk of bleeding on a VKA doubles [47].

Vitamin K replacement is essential to replenish the vitamin K dependent factors and reverse VKA activity. It should be given promptly, and IV administration mitigates variability in oral vitamin K absorption. Although the risk profile is low, it may take 24 h or more to become effective [48].

Fresh frozen plasma (FFP) has been conventionally used to reverse an elevated INR due to VKA in conjunction with vitamin K. Although it is relatively inexpensive and widely available, its use is complicated by delays in administration, potential transfusion-related reactions, and large volumes which may be required for full INR reversal. This had led to the more widespread use of prothrombin complex concentrate (PCC). PCC is derived from plasma and contains factors II, VII, IX, and X in variable proportions in different preparations. It is rapidly administered in a small volume. PCC is very expensive and is reserved for life-threatening bleeding events or the need for surgery. PCC has been demonstrated to reverse INR to < 1.4 and maintain INR reversal for > 48 h in the majority of patients on VKA therapy [49]. In the INCH trial, a randomized trial comparing FFP and PCC for VKA reversal in ICH, PCC provided more rapid INR reversal with an effective reduction in hematoma expansion [50]. The trial was stopped early due to safety concerns with FFP therapy. Professional society guidelines currently recommend consideration of PCC over FFP for VKA reversal in ICH [48, 51].

Factor Xa inhibitors

Xa inhibitors prevent the conversion of prothrombin to thrombin. Detection of novel agents using calibrated chromogenic assays is expensive and not readily available, and the utility of existing anti-Xa assays has not been validated to identify the presence of the oral factor Xa inhibitors.

Andexanet alfa is a recombinant inactive form of factor Xa. It binds to Xa inhibitors with high affinity and sequesters the medication, resulting in reduced anti-Xa activity. In the ANNEXA-4 study of 352 patients with major bleeding, andexanet alfa reduced anti-Xa levels by 92% in rivaroxaban/apixaban treated patients and 75% in enoxaparin [52]. Andexanet alfa is approved by the FDA for the reversal of rivaroxaban and apixaban, but widespread use is limited due to the cost of the medication.

Four-factor PCC was commonly used for Xa inhibitor reversal prior to andexanet alfa, and it may be used if andexanet alfa is not available. Four-factor PCC has been demonstrated to effectively reverse rivaroxaban in healthy volunteers [53]. In the setting of intracranial bleeding, it is recommended to administer 25 to 50 units/kg if the medication was ingested within 3–5 half lives or if the time of last exposure is unknown. In patients with a known ingestion within 1 h, activated charcoal may also be utilized if this is deemed safe from an airway protection standpoint. Hemodialysis is not effective in removing Xa inhibitors [48].

Direct thrombin inhibitors

DTIs directly inhibit the activity of factor IIa, which is the key factor in converting fibrinogen to fibrin. DTIs have an additional unique indication in the treatment of heparin-induced thrombocytopenia. Intravenous formulations are short acting and generally do not require reversal agents. However, the reversal of oral dabigatran was challenging before idarucizumab became available. Idarucizumab is a monoclonal antibody which binds to dabigatran with considerably higher affinity than factor IIa. In the RE-VERSE AD study of 503 patients with life-threatening bleeding (group A) or need for emergent surgical procedure (group B), idarucizumab reversed thrombin time to normal in 100% of patients, and this effect remained relatively stable 24 h after treatment. Normal intraoperative hemostasis was achieved in 92% of the patients in group B [54]. The rate of thrombotic events in this study was similar to those reported after major surgical procedures or hospitalization for uncontrolled bleeding [55, 56]. Idarucizumab is administered in 2 doses of 2.5 g given within 15 min. Activated charcoal is an additional option for oral agents such as dabigatran [48, 51].

If idarucizumab is not available, hemodialysis and activated PCC are alternative options. Dabigatran is renally excreted and is effectively removed by hemodialysis. However, there is a risk of worsening cerebral edema in patients with mass lesions which must be taken into consideration. The recommended dosing for PCC is 25 to 50 units/kg, the same as for Xa inhibitors. Administration of PCC beyond 3–5 half lives of dabigatran exposure may be considered in patients with renal insufficiency [48].

Heparin and low molecular weight heparin

UFH activates antithrombin III activity, which inhibits factors IIa and Xa. Heparin activity can be assessed utilizing aPTT or point-of-care activated clotting time. Heparin activity may also be assessed with thromboelastography; shortening of reaction time with heparinase implicates the presence of heparin in the sample as the cause of coagulopathy. Its effects may be reversed with protamine, which is a naturally occurring protein that binds to heparin.

LMWH has a similar mechanism of action but is longer acting and thought to have more predictable pharmacology in the setting of normal renal function. Assessment of LMWH activity requires the use of an anti-factor Xa assay. Protamine may be used to reverse LMWH activity, but this reversal is incomplete and estimated to be around 60% [48]. A novel approach to LMWH is andexanet alfa. Although this is a potential therapeutic intervention with a possibility of more complete LMWH reversal than with protamine, it is currently not approved for this indication and is still undergoing further study [52].

Antiplatelet therapy

Antiplatelet therapy can further increase the risk of bleeding in those on anticoagulation. The American College of Cardiology does not recommend anticoagulation in addition to dual antiplatelet therapy (“triple therapy”) [9]. In fact, the addition of a single antiplatelet agent to an oral anticoagulant can increase the risk of bleeding > 20%, with a further two- to three-fold increase if a second antiplatelet is also used. In the most recent guidelines, if triple therapy is necessary, then it should be used for the shortest period of time possible up to 30 days, with subsequent transition to anticoagulation with single antiplatelet therapy up to 12 months and followed by anticoagulation only [8, 17, 57].

The main classes of antiplatelet drugs commonly used in practice include cyclooxygenase-1 (COX-1) inhibitors (e.g., acetylsalicylic acid; ASA or aspirin), phosphodiesterase (PDE) inhibitors (e.g., dipyridamole and cilostazol), P2Y12 receptor inhibitors (e.g., clopidogrel, prasugrel, and ticagrelor), and glycoprotein IIb/IIIa inhibitors (e.g., abciximab, eptifibatide, and tirofiban). Reversal of antiplatelet therapy in the bleeding patient may be achieved through the administration of platelet transfusion. Desmopressin, also known as 1-deamino-8-D-arginie vasopressin (DDAVP), has also demonstrated to improve platelet function in patients on antiplatelet therapies in several tests of platelet function compared to those who had not received reversal agents [48, 58,59,60,61,62,63]. Given its low cost and relatively good safety profile, its administration should be considered in patients with major hemorrhage who were exposed to antiplatelet agents [48].

IVC filter

In those at risk for or with major bleeding or with an absolute contraindication to anticoagulation, inferior vena cava (IVC) filters provide an alternative therapy [8, 9]. Before the advent of implantable IVC filters, alternative methods to prevent embolization of DVT to the pulmonary arteries included femoral vein or IVC plication or ligation; however after the introduction of the Mobin-Uddin umbrella inferior vena cava (IVC filter) in 1967, placement of IVC filters steadily began to increase [64, 65]. Absolute contraindications for anticoagulation include active uncontrollable bleeding, high risk of major bleeding (e.g., coagulation defect, severe thrombocytopenia, recent intracerebral hemorrhage or cerebral lesion at high risk of bleeding), or urgent surgery requiring cessation of anticoagulation [66].

There are two general types of IVC filters currently in use. These are either permanent (or nonretrievable) or optional (or retrievable). Permanent IVC filters, available since the 1960s, are typically placed in patients with long-term absolute contraindications to anticoagulation and a need to mechanically prevent pulmonary embolism (PE) [67]. Retrievable IVC filters were developed in the 1990s and can be removed if contraindications to anticoagulation or risk of PE resolves. Published studies show that the majority of IVC filters inserted, although retrievable, are never removed. Retrieval rates range from 12 to 45% [68]. There is no data available to suggest that one type of filter is more effective than others. Complications can occur at the time of filter placement (technical issues, medication related) or at a later time related to the access site, the filter itself (thrombosis) or as a consequence of filter removal [69].

Considerations after bleeding event

There are no clear guidelines on restarting anticoagulation after a hemorrhagic event, and arguments have been made for and against resuming anticoagulation. The clinician must carefully review the original indication and weigh the risk for both repeat thrombotic and/or bleeding event. In some cases, there may be a rebound effect where VTE may be more likely to occur in the first weeks after stopping therapy [70]. The risk of recurrent VTE when a patient is off anticoagulation decreases with time from the initial event. The highest risk, estimated at 0.3% to 1.3% per day, is in the first 4 weeks, falling to 0.03%, to 0.2% per day in weeks 5 through 12, and 0.05% per day thereafter, so if the anticoagulation therapy is near completion, then it would be reasonable to stop it after a significant bleeding event [71].

Gastrointestinal bleeding is the most common type of bleeding associated with anticoagulation, and it complicates long-term anticoagulation therapy in 5 to 15% of patient [72]. Results from four large observational studies evaluating outcomes after resuming anticoagulation from gastrointestinal bleeding demonstrated that the risk of recurrent bleeding was not significantly increased in those that resumed anticoagulation. Anticoagulation was associated with a significant reduction in the likelihood of thrombotic events, and finally overall mortality was significantly lower in those that resumed anticoagulation [73,74,75,76]. The optimal time to resume anticoagulation remains unclear, but data to date suggests around 2 weeks. For some, alternate anticoagulant therapy may be considered [72].

Management of anticoagulation after an initial ICH in patients with primary and metastatic brain tumors is challenging. Data is limited and primarily derived from either observational studies or small studies [72, 77]. Within the first 24 h after ICH, 70% will develop some amount of hematoma expansion [78, 79]. Subsequently, the risk of hematoma expansion lessens, while the risk of arterial and venous thromboembolism accumulates [80, 81]. Typically most patients will receive antidote for anticoagulant, and often IVC filter placement will be considered. Some have recommended use of DOAC after ICH; however, we would encourage a multi-disciplinary approach and consensus based on the size of the bleed and clinical scenario [82].

If the risk of recurrent hemorrhage on anticoagulation is considered to be too great, aspirin may be an alternative [83]. However, conservative measures and close clinical follow-up with anticoagulant therapy are also reasonable. Finally, after a bleeding event, the possibility of resuming anticoagulant therapy or new thrombotic event often engenders significant anxiety for both the patient and their family, so shared decision-making is imperative.

Conclusion

Anticoagulation-related bleeding is higher in cancer patients than in the general population. Over the past decade, LMWH was the recommended agent for cancer-associated VTE. Recent major trials have shown non-inferiority of the DOACs in recurrent VTE as an endpoint, though the risk for major bleeding persists. Awareness of those with increased risk of bleeding prior to initiation of anticoagulation is essential. Given the prevalence of anticoagulant therapies in cancer patients, clinical providers must be able to recognize potentially critical bleeding sites and modalities to reverse major hemorrhage. Evidence for timing and need for re-initiation of anticoagulant therapy after a major bleeding remains sparse, but a multi-disciplinary approach and shared decision-making can be implemented in the interim.