Drug-induced thrombosis and vascular disease in patients with malignancy:
INTRODUCTION — Cancer is often associated with a state of hypercoagulability, which can have a variety of clinical manifestations, including migratory superficial thrombophlebitis, “unprovoked” deep vein and other venous thrombosis, nonbacterial thrombotic endocarditis, and disseminated intravascular coagulation. Deep vein thrombosis is the most common vascular complication of antineoplastic therapy. Intracranial dural sinus vein thrombosis and thrombotic microangiopathies can also be seen. (See “Hypercoagulable disorders associated with malignancy”.)
This review will discuss vascular complications caused by the various chemotherapeutic agents used in the treatment of patients with malignant disease.
MECHANISMS AND RISK FACTORS — The proposed mechanisms for the increased incidence of vascular complications in patients with malignancy include release or expression of procoagulants by tumor cells (eg, tissue factor) and expression of procoagulant activity by normal host cells such as monocytes, platelets, and endothelial cells. (See “Pathogenesis of the hypercoagulable state associated with malignancy”.)
In addition, the usual treatments for cancer also significantly increase the risk of thrombotic events.
- Surgery is a major precipitating factor, with the risk of postoperative thrombosis being several-fold higher than in patients without cancer. (See “Prevention of venous thromboembolic disease in surgical patients”, section on ‘Risk factors for VTE’.)
- High dose chemotherapy and bone marrow transplantation for hematologic malignancies are associated with an enhanced risk of thrombosis, particularly hepatic veno-occlusive disease. (See “Pathogenesis and clinical features of hepatic sinusoidal obstruction syndrome (veno-occlusive disease) following hematopoietic cell transplantation”.)
- A number of chemotherapeutic regimens have been associated with an increased rate of thromboembolism . Two of the regimens most commonly complicated by thrombosis are L-asparaginase, often given for acute lymphoblastic leukemia, and the administration of tamoxifen and other agents in the treatment of breast cancer (see below).
- The presence of an indwelling central venous catheter is a known risk factor for the development of upper extremity venous thrombosis in both adult and pediatric cancer patients. (See “Catheter-induced upper extremity venous thrombosis” and “Pathogenesis and clinical manifestations of venous thrombosis and thromboembolism in infants and children”.)
Disease- and patient-specific risk factors may also contribute to the risk of chemotherapy-associated venous thromboembolic disease. In a study of 3003 patients receiving at least one cycle of chemotherapy, symptomatic VTE occurred in 1.9 percent over a median follow-up of 2.4 months, or a rate of 0.8 percent/month . The highest rates were seen in patients with upper gastrointestinal cancers, lung cancer, and lymphoma, with rates of 2.3, 1.2, and 1.1 percent/month, respectively.
On multivariate analysis, the following factors were significantly associated with development of symptomatic VTE :
- Upper gastrointestinal malignancy (odds ratio [OR] 3.9, 95% CI 1.4-10)
- Pre-chemotherapy platelet count ≥350,000/microL (OR 2.8, 95% CI 1.6-4.9)
- Use of white blood cell growth factors (eg, G-CSF, OR 2.1, 95% CI 1.2-3.6)
- Hemoglobin <10.0 g/dL or use of erythropoietin (OR 1.8, 95% CI 1.1-3.1)
L-ASPARAGINASE — Thrombotic events have been reported with induction chemotherapy regimens for acute lymphoblastic leukemia (ALL) that include L-asparaginase. Intracranial dural sinus thrombosis with hemorrhage is observed most frequently, but deep venous thrombosis and pulmonary embolism can also occur [3-7]. In one large series of children receiving L-asparaginase as part of induction chemotherapy for ALL, the incidence of thrombotic complications was 1.2 percent . Generalized bleeding episodes have rarely been observed. (See “Induction therapy for acute lymphoblastic leukemia in adults”.)
Asparaginase depletes plasma asparagine, thereby inhibiting protein synthesis in leukemic cells and the synthesis of many plasma proteins. The latter effect causes deficiencies of albumin, thyroxine-binding globulin, and various coagulation proteins, including prothrombin, factors V, VII, VIII, IX, X, XI, fibrinogen, antithrombin, protein C, protein S, and plasminogen [8-10]. This results in prolongation of the prothrombin time, activated partial thromboplastin time (aPTT), and thrombin time, and in hypofibrinogenemia with levels often less than 100 mg/dL. These coagulation abnormalities resolve within one to two weeks after cessation of the drug.
It is difficult to assess the role of the substantial reductions in the levels of natural anticoagulant proteins such as antithrombin, protein C, and protein S in the pathogenesis of the thrombotic events. Both procoagulant and anticoagulant protein synthesis in the liver are decreased by L-asparaginase, leading to uncertainty as to whether there is an alteration in the balance of the opposing forces of the hemostatic mechanism. Two mechanisms have been considered and are discussed below
Antithrombin — There are data suggesting that antithrombin supplementation given concurrently with L-asparaginase may suppress the prothrombotic state [11-13]. However, a drug-induced decrease in antithrombin is not predictive of subsequent thrombosis. In one study of children with ALL receiving L-asparaginase, prednisone, and vincristine, no correlation was found between protein C, protein S, or antithrombin levels and the presence or absence of thrombosis .
von Willebrand factor — Qualitatively abnormal von Willebrand factor (vWf) has been found in several patients at the time of L-asparaginase-induced thrombosis . L-asparaginase may lead to a transient increase in unusually large plasma vWf multimers that have enhanced platelet-agglutinating properties .
Erwinia-asparaginase is an alternate preparation which may have fewer effects on the coagulation system than Escherichia coli L-asparaginase. In a series of 11 adults with ALL, there was significant lowering of antithrombin but the levels of the vitamin-K dependent procoagulant factors II, VII, and X, remained within normal ranges .
TREATMENT OF BREAST CANCER — A number of large studies have found an increased incidence of thromboembolic events in women with breast cancer treated with chemotherapy, tamoxifen, or both. Additional risk factors may include postmenopausal status, prior mastectomy, increased body weight, presence of an indwelling central venous catheter, and evidence of coagulation activation .
Chemotherapy — Studies of adjuvant chemotherapy for stage II breast cancer or primary chemotherapy for advanced, metastatic disease have shown a high incidence of thrombosis, both arterial and venous, during but not after chemotherapy. The risk is higher in patients with metastatic disease, which is probably due to the increased tumor burden and the increased incidence of other predisposing factors such as immobilization.
Incidence — The incidence of thromboembolic disease following chemotherapy for breast cancer has varied widely in different reports, and appears to be related to the stage of the disease, the chemotherapeutic regimen given (eg, CAF versus CMF), and whether tamoxifen is given concurrently or following completion of chemotherapy (see ‘Tamoxifen and raloxifene’ below).
- In a randomized study of early breast cancer, patients assigned to perioperative chemotherapy with fluorouracil, doxorubicin and cyclophosphamide had a higher incidence of thromboembolic events within six weeks after surgery than a control group treated with observation alone (2.1 versus 0.6 percent) .
- In a Cancer and Leukemia Group B (CALGB) study, 433 patients were randomly assigned to receive one of three regimens based upon cyclophosphamide, methotrexate and 5-fluorouracil (CMF) as adjuvant therapy . Thromboembolic disease occurred in 5 to 7 percent. There was no difference between the three groups. No patient developed thrombosis after chemotherapy was completed.
A similar incidence (7 percent) was noted in a prospective study evaluated of 205 women with stage II breast cancer who were treated with one of two chemotherapy regimens: CMFVP (cyclophosphamide, methotrexate, fluorouracil, vincristine, and prednisone) or CMFVP plus doxorubicin and tamoxifen . Once again, no thrombotic episodes occurred during 2413 patient months of follow-up without therapy. There was no relationship between the development of thrombosis and estrogen or progesterone receptor status, age, number of involved lymph nodes, or subsequent tumor recurrence.
- A higher risk of thrombosis (18 percent) was noted in a series of 159 patients treated with CMFVP for advanced, stage IV disease . There were no differences in the presence of risk factors for thrombosis between the patients who had a thromboembolic event and those who did not.
Although most patients develop venous thrombosis, there is also an association between arterial thrombosis and cancer therapy among women with breast cancer [22,23]. A CALGB study, for example, found a 1.3 percent incidence of arterial thrombosis, either peripheral or cerebrovascular, in 1014 patients during treatment for stage II or III breast cancer on two separate chemotherapy protocols . All but one of the thrombotic events occurred while patients were receiving chemotherapy.
Mechanism — The mechanism of the thrombogenic effects of chemotherapy in patients with breast cancer is not well understood. One report noted statistically significant decreases in protein C and protein S levels during CMF chemotherapy which, in some patients, fell below the range of values seen in hereditary thrombotic disorders . Reductions in procoagulants factor VII and fibrinogen were also present and none of the patients in this small series had clinically evident thrombosis. In another report, there were significant declines in protein C concentration during CMF chemotherapy, which returned to baseline values after the completion of therapy .
Possible explanations for these chemotherapy-induced abnormalities include impairment of vitamin K metabolism and inhibition of DNA/RNA synthesis, leading to a reduction in protein synthesis by the liver. In addition, endothelial cell injury can lead to qualitative or quantitative abnormalities in vWf which may enhance the thrombotic potential. Other possible mechanisms include direct platelet activation, reduced fibrinolytic activity, and the release of procoagulant from tumor cells dying as a result of antineoplastic therapy.
Attempts to identify patients at risk by monitoring with coagulation tests have not been successful. In a study of 50 patients treated with adjuvant epirubicin and cyclophosphamide chemotherapy, the incidence of deep vein thrombosis was 10 percent . Preoperative levels of D-dimer, fibrinogen, and plasminogen activator inhibitor activity were significantly higher in the patients with breast cancer than in healthy women; however, monitoring during chemotherapy did not identify patients at higher risk for deep vein thrombosis.
Tamoxifen and raloxifene — A number of studies, including the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) overview analysis and the large Breast Cancer Prevention Trials, have demonstrated that tamoxifen use is associated with an increased rate of venous thromboembolic events (2.8 percent ), especially within the first few years of tamoxifen use [28,29], and that there is a significant additional procoagulant effect when tamoxifen is added to chemotherapy [23,30-37]. (See “Overview of the causes of venous thrombosis”, section on ‘Tamoxifen’.)
- In the NSABP P-1 Breast Cancer Prevention Trial (BCPT), which involved 13,388 women followed for an average of 3.6 years, the rates of pulmonary embolism and deep vein thrombosis were increased in older women receiving tamoxifen (risk ratio 3.0 [CI 1.1 to 11.2], and 1.6 [CI 0.9 to 2.9], respectively) . In a subsequent pooled analysis of 13 NSABP Breast Cancer Prevention trials, which involved 20,878 women, the risk of pulmonary embolism, deep venous thrombosis, and superficial phlebitis was increased two- to three-fold in those treated with tamoxifen and was increased 11- to 15-fold in those treated with tamoxifen plus chemotherapy . In this trial and in a study from the Marshfield Clinic Personalized Medicine Research Project , risk factors for VTE included increased age and body mass index. There were no differences in the rates of thrombosis between African Americans and white women in this study.
- In the five-year, randomized, double-blind, placebo-controlled International Breast Cancer Intervention Study (IBIS-1), involving 7139 patients, use of tamoxifen was associated with an increased risk of developing a major venous thromboembolic event (odds ratio 2.1, 95% CI: 1.1-4.1) . This risk was further increased in those patients who had surgery, immobilization, or fracture in the month prior to the event (odds ratio 4.7, 95% CI: 2.2-10.1).
- In a follow-up report from the IBIS-I trial, the risk of developing deep vein thrombosis or pulmonary embolus was significantly higher in the tamoxifen arm than in the placebo arm during the five years of active treatment (RR 2.3, 95% CI 1.4-3.9), but not after tamoxifen was stopped (RR 1.1, 95% CI 0.5-2.5) . There were no statistically significant differences between treatment groups in the rates of any cerebrovascular events or cardiovascular events, either during active treatment or after tamoxifen was stopped.
Factor V Leiden and prothrombin mutations were not associated with thrombosis in the BCPT  or IBIS-I  trials involving the use of tamoxifen for the prevention of breast cancer. However, in a case-control study from the CALGB, among women taking adjuvant tamoxifen for early stage breast cancer, those who had a thromboembolic event (TE) were nearly five times more likely to carry a factor V Leiden mutation than those who did not have a TE  .
A retrospective analysis of DNA from 220 subjects in the Marshfield Clinic Personalized Medicine Research Project found an association between tamoxifen-associated thromboembolic events and single nucleotide polymorphisms of the estrogen receptor 1 genotype, although this did not reach the level of statistical significance (hazard ratio 3.5; 95% CI 0.97-12) .
The possibility that tamoxifen may be associated with an increased incidence of arterial thromboembolism (ie, stroke) was raised in two NSABP randomized trials: P-1 and NSABP B-24, a trial of tamoxifen for intraductal breast cancer. However, the available data are conflicting:
- In the EBCTCG overview analysis of approximately 15,000 women randomly assigned to receive or not receive five years of tamoxifen, there was a trend toward increased stroke-related mortality that did not reach the level of statistical significance (54 versus 29 deaths, p = 0.07) .
- Tamoxifen use was not associated with higher stroke risk in a retrospective nested case control study of 11,045 women enrolled in a large HMO in the Los Angeles area who were diagnosed with breast cancer between 1980 and 2000 . When 179 who met the criteria for stroke were compared with 358 age and year of diagnosis-matched controls who had breast cancer but not stroke, there was no association between tamoxifen and stroke risk.
- In a systematic review of randomized controlled trials of tamoxifen use for breast cancer management and prevention published since 1980, the frequency of ischemic stroke during a mean follow-up period of 4.9 years was 0.71 percent with tamoxifen versus 0.39 percent for controls . It was concluded that, although the absolute risk of stroke was small, women with breast cancer who were treated with tamoxifen had an increased risk for ischemic stroke (odds ratio 1.82, 95% CI 1.4-2.4) or any stroke (odds ratio 1.40, 95% CI 1.1-1.7).
On the other hand, the increased risks of pulmonary embolus and stroke may be partially offset by a decreased risk of ischemic heart disease in women receiving tamoxifen. In the most recent overview analysis of randomized trials of adjuvant tamoxifen from the EBCTCG, the non-significant excess of stroke deaths (three extra per 1000 women during the first 15 years) in women treated with tamoxifen was balanced by a non-significant shortfall in cardiac deaths (three fewer per 1000 women during the first 15 years) . Thus, there was little net effect of tamoxifen on overall vascular mortality. This beneficial effect may be due to the favorable effect of tamoxifen on lipid profiles. (See “Managing the side effects of tamoxifen”, section on ‘Coronary heart disease’.)
Mechanism — The potential mechanisms for the tamoxifen procoagulant effect have not been identified. Tamoxifen is an antiestrogen that has weak estrogenic effects which may contribute to its prothrombotic activity. A number of studies have evaluated measurements of hemostasis in patients taking tamoxifen. The results have been conflicting and no major changes have been identified [46-49]. Some reports have found modest reductions in antithrombin and protein C [46,47], while a prospective double-blind study was unable to find consistent tamoxifen-induced changes in protein S or protein C activity .
Other agents — Aromatase inhibitors are being used as adjuvant hormonal treatment as an alternative to tamoxifen. In a systematic review and meta-analysis of seven trials, the use of aromatase inhibitors, as compared to the use of tamoxifen, was associated with a significantly lower incidence of venous thrombosis (1.6 versus 2.8 percent, respectively, OR 0.55; 95% CI 0.46-0.64) . (See “Adjuvant endocrine therapy for postmenopausal women with early stage breast cancer”.)
Second-line hormonal therapy for patients with advanced breast cancer who fail tamoxifen includes the aromatase inhibitor formestane and the progestational agent megestrol. In a phase III prospective randomized crossover trial, the two drugs showed similar antineoplastic activity but the incidence of deep vein thrombosis was significantly higher with megestrol than formestane (5 in 81 patients versus zero in 90 patients) .
Prophylactic anticoagulation — In a prospective, double blind study, for example, 311 patients with stage IV breast cancer received either placebo or warfarin (1 mg PO daily for six weeks, adjusted to maintain an INR of 1.3 to 1.9) . Treatment was continued until one week after the end of chemotherapy, which lasted approximately six months. There were seven thromboembolic events in the placebo group and one in the warfarin group (4.4 versus 0.7 percent), a relative risk reduction of 85 percent. Major bleeding occurred in three patients, two of whom received placebo. However, the chemotherapeutic regimens were those in use in the early 1990s and are different from those commonly employed today.
In a separate study, the effect of low-intensity warfarin on markers of hypercoagulation was evaluated in a randomized trial of 32 patients with metastatic breast cancer who were undergoing chemotherapy . Before therapy, markers of clotting activation were increased in both groups, consistent with the presence of a hypercoagulable state. After starting chemotherapy, markers were progressively lower in the group receiving warfarin prophylaxis. These differences became statistically significant after the fourth course of chemotherapy. Deep vein thrombosis occurred in two of the 16 patients receiving placebo and in none of the 16 patients receiving very-low-dose warfarin.
TREATMENT OF TESTICULAR CANCER — Major cardiovascular issues that have been studied in connection with treatment for testicular cancer include hypertension, dyslipidemia, early atherosclerosis and coronary artery disease, Raynaud phenomenon, and thromboembolic events. These are discussed in detail separately. (See “Treatment-related toxicity in men with testicular germ cell tumors”, section on ‘Cardiovascular’ and ‘Miscellaneous agents’ below.)
THROMBOTIC MICROANGIOPATHY — Thrombotic microangiopathies have been associated with a number of cancer chemotherapeutic agents. This primarily occurs with one of four regimens: mitomycin C; cisplatin with or without bleomycin; gemcitabine; and the use of radiation and high dose chemotherapy prior to hematopoietic cell transplantation (HCT). (See “Causes of thrombotic thrombocytopenic purpura-hemolytic uremic syndrome in adults”, section on ‘Cancer and chemotherapy’.)
The syndrome more closely resembles the hemolytic-uremic syndrome (HUS) than thrombotic thrombocytopenic purpura (TTP) and usually develops within weeks to months after exposure to the drug. Detection may be delayed because chemotherapy can produce thrombocytopenia due to bone marrow depression and renal disease, thereby masking the presence of HUS. Affected patients typically present with slowly progressive renal failure, new or exacerbated hypertension, and a relatively bland urine sediment, often occurring in the absence of clinically apparent tumor.
Drug-related endothelial injury is presumed to be the initiating event. It has been proposed that von Willebrand factor (VWF) multimers derived from damaged or stimulated endothelial cells may be involved in the pathogenesis of intravascular platelet clumping. However, there are conflicting data on the role of VWF multimers in patients with drug-induced thrombotic microangiopathies. In one study, VWF multimeric patterns were normal during episodes of mitomycin C-induced microangiopathy . However, another report using a superior technique for detection of multimers found abnormalities in VWF in five of six patients . The sixth patient had a threefold elevation in the levels of VWF antigen.
Thrombotic microangiopathy confined to the kidney has been described in association with the use of inhibitors of vascular endothelial growth factor (VEGF) in cancer patients. (See “Extrinsic nonimmune hemolytic anemia due to mechanical damage: Fragmentation hemolysis and hypersplenism”, section on ‘Drugs’ and “Chemotherapy-related nephrotoxicity and dose modification in patients with renal insufficiency”, section on ‘VEGF pathway inhibitors’.)
MISCELLANEOUS AGENTS — A variety of other drugs that can be used in patients with malignancy can be associated with thromboembolic events. These are described below.
Estrogens — High doses of estrogens increase the plasma concentrations of clotting factors and the risk of thrombotic disease . Estrogenic prothrombotic effects are believed to be responsible for the dose-related cardiovascular complications of the estrogen analogue diethylstilbestrol (DES) in the treatment of prostate cancer . In comparison, the lower doses used for hormone replacement therapy or oral contraception are associated with only a small increase in thrombotic risk. (See “Postmenopausal hormone therapy and cardiovascular risk”, section on ‘Venous thromboembolism’ and “Initial hormone therapy for metastatic prostate cancer”, section on ‘Estrogens’.)
Bevacizumab — Both arterial and venous thromboembolic complications have been associated with use of the angiogenesis inhibitor, bevacizumab and a number of the other antiangiogenesis inhibitors .
Arterial events — An increased incidence of potentially fatal arterial thromboembolic events (eg, transient ischemic attack, stroke, angina, myocardial infarction) has been associated with the use of a bevacizumab-containing chemotherapy regimen for advanced colorectal cancer . (See “Systemic chemotherapy for metastatic colorectal cancer: Completed clinical trials”.)
In randomized trials, the 4 to 5 percent incidence of serious arterial thromboembolic events in patients receiving bevacizumab in combination with a short-term infusional 5-FU-based chemotherapy regimen represents an approximately two- to threefold higher incidence than in the control groups receiving the same chemotherapy regimen but without bevacizumab .
The risk appears to be highest (17 percent) in patients with a prior history of an arterial thrombotic event and, in some reports, in those over age 65 . In such patients, bevacizumab should only be used with informed consent regarding these risks. Furthermore, the drug should be discontinued for all greater than grade 3 or any grade 2 new or worsened arterial thrombotic events during therapy . A discussion of the use of bevacizumab in elderly patients who have a history of an arterial thromboembolic event within the past 6 to 12 months can be found elsewhere. (See “Systemic chemotherapy for nonoperable metastatic colorectal cancer: Treatment recommendations”, section on ‘Bevacizumab’.)
The pathophysiology underlying these adverse effects remains unresolved .
Venous events — The use of bevacizumab may be associated with an increased risk of venous thromboembolic events (VTE) in cancer patients receiving this drug, although the degree of this risk is uncertain after two meta-analyses.
The first meta-analysis, published in 2008, included 7956 patients with a variety of advanced solid tumors from 15 randomized controlled trials in which standard antineoplastic therapy was used with and without bevacizumab . The following results were noted:
- When compared with controls receiving standard antineoplastic therapy alone, the addition of bevacizumab was associated with a significantly increased risk of VTE (RR 1.33; 95% CI 1.13-1.56)
- The increased risk for VTE was similar for both low dose (2.5 mg/kg per week; RR 1.31) and high dose (5 mg/kg per week; RR 1.31) bevacizumab.
- The risk for development of VTE following the use of bevacizumab differed among tumor types. It was highest for aerodigestive malignancies and lowest for renal cell and breast cancers.
The second meta-analysis, published in 2011, included 6055 patients from 10 randomized trials, and found the following :
- There were no statistically significant increase in the unadjusted or exposure-adjusted incidences of all-grade VTEs for bevacizumab versus controls in the overall population (OR 1.14; 95% CI 0.96-1.35), or by tumor type
- Several risk factors for VTE were identified, including tumor type, older age, poorer performance status, VTE history, and baseline anticoagulant use. No interactions between bevacizumab treatment and these factors were observed.
The 95 percent confidence limits of these two analyses overlap somewhat, suggesting that, if there is an increased risk associated with the use of bevacizumab, it is likely to be in the range of 13 to 35 percent.
Hypertension — Bevacizumab has been associated with the development of hypertension. Guidelines for pretreatment assessment, monitoring, and management of elevated blood pressure in patients receiving angiogenesis inhibitors such as bevacizumab are available (table 1A-B). This subject is discussed in more depth elsewhere.(See “Overview of angiogenesis inhibitors”, section on ‘Hypertension’.)
Sunitinib and sorafenib — Sunitinib and sorafenib are orally active multitargeted inhibitors of several tyrosine kinases, including vascular endothelial growth factor (VEGF). Similar to bevacizumab, they have been associated with the development of hypertension. (See “Overview of angiogenesis inhibitors”, section on ‘Hypertension’.)
As with bevacizumab, these agents are also associated with arterial thromboembolism. A systematic literature review and meta-analysis have indicated that use of the sunitinib and sorafenib was associated with an increased risk for the development of arterial thromboembolic events (RR 3.03; 95% CI 1.25-7.37) .
Fluorouracil — Cardiac toxicity occurs in 1.6 to 2.3 percent of patients treated with 5-fluorouracil , and myocardial ischemia and stroke occur in up to 10 percent of patients who receive continuous infusions of the drug . Induction of coronary spasm is the presumed mechanism . (See “Cardiotoxicity of nonanthracycline cancer chemotherapy agents”.)
Thalidomide and lenalidomide — Thrombosis has been noted in a significant percentage of patients with multiple myeloma treated with thalidomide and lenalidomide, most often in combination with other agents (eg, dexamethasone, doxorubicin). This subject is discussed separately. (See “Thrombotic complications following treatment of multiple myeloma with thalidomide and its analogues”.)
Cisplatin — A number of studies have found an increased incidence of arterial and venous complications in patients with a variety of malignancies treated with cisplatin-containing regimens. As an example, in a large retrospective analysis of all patients treated with cisplatin-based chemotherapy in 2008 for any type of malignancy at one medical center, 169 of the 932 patients (18.1 percent) experienced a thromboembolic event (TEE) within four weeks of the last dose of cisplatin . Of these TEEs, 89 percent were venous, 8 percent were arterial, and 3 percent were both venous and arterial.
Various combinations of bleomycin, cisplatin and vinblastine are associated with myocardial infarction and stroke [69-71]. In one study, for example, cisplatin-based multiagent chemotherapy for urothelial transitional cell carcinoma was associated with a 13 percent risk of vascular events (venous thromboembolism, arterial thromboses, and cerebrovascular events), most of which occurred during the first two cycles of chemotherapy .
In another prospective series of 108 patients treated with cisplatin and gemcitabine for non-small cell lung cancer, 19 had significant vascular events, including seven with lower extremity arterial thrombosis . (See “Neurologic complications of platinum-based chemotherapy”, section on ‘Cisplatin’.)
The use of cisplatin-bleomycin regimens for testicular cancer has been associated with Raynaud’s phenomenon in up to 40 percent of patients [73-75]. Vasospasm presenting as painful digits and paresthesias typically occurs 10 months after starting therapy and lasts indefinitely. The cumulative dose of bleomycin appears to be the major risk factor . Raynaud’s phenomenon associated with bleomycin and vinca alkaloids also occurs in patients with AIDS-related Kaposi’s sarcoma [76,77]. (See “Treatment-related toxicity in men with testicular germ cell tumors”, section on ‘Cardiovascular’.)
In series of patients with germ cell tumors, non-small-cell lung cancer, and urothelial cancers, thromboembolic phenomena associated with cisplatin-based chemotherapy were observed in 8 to 18 percent of treated patients, and rates of arterial thrombosis ranged from 3 to 9 percent [71,72,78]. (See “Treatment-related toxicity in men with testicular germ cell tumors”, section on ‘Thromboembolic events’.)
A phase III trial of oxaliplatin versus cisplatin added to infused fluorouracil for patients with advanced gastroesophageal cancer documented significantly more thrombotic events with the cisplatin- compared with the oxaliplatin-containing doublet (7.8 versus 0.9 percent) .
A prospective analysis of thromboembolic events in a randomized, controlled trial of 964 patients with advanced gastroesophageal cancer treated with four different triplet regimens, demonstrated significantly more thromboembolic events in those treated with cisplatin-containing regimens compared with those receiving oxaliplatin (15.1 versus 7.6 percent) .
Other chemotherapy combinations — Dacarbazine, alone or in combination with other drugs has been associated with the Budd-Chiari syndrome . In addition, veno-occlusive disease of the liver has been reported in patients with acute myelocytic leukemia who were treated with thioguanine and daunomycin with or without cytarabine [80,81].
- Cancer is often associated with a state of hypercoagulability, as evidenced by the complications of venous as well as arterial thromboembolic phenomena. (See “Hypercoagulable disorders associated with malignancy”.)
- The hypercoagulable state associated with malignancy can be aggravated by a number of factors, such as surgery, the use of indwelling catheters for the delivery of chemotherapy, and the chemotherapeutic agents themselves. (See ‘Mechanisms and risk factors’ above.)
- The chemotherapeutic agents most commonly associated with thrombosis and vascular disease (eg, L-asparaginase, tamoxifen, bevacizumab, thalidomide, cisplatin, estrogens) are discussed in this review