sign in sign up
<<
Home , Coagulation screen, Coagulopathy, Fibrinogen, Thrombophilia

Leukemia (2013) 27, 553–559 & 2013 Macmillan Publishers Limited All rights reserved 0887-6924/13 www.nature.com/leu

REVIEW The and thrombotic risk associated with L- treatment in adults with acute lymphoblastic leukaemia

E Truelove1, AK Fielding2 and BJ Hunt3

The dramatic improvements seen in the outcome of paediatric patients with acute lymphoblastic leukaemia (ALL) have led to increasing incorporation of L-asparaginase (L-Asp) in adult treatment protocols. However, its use is associated with a disruption in the physiological balance between haemostatic and pathways, with the predominant clinical manifestation being . Although L-Asp is known to be associated with an acquired deficiency of (AT), the concurrent depletion of fibrinogen and other haemostatic means that the precise mechanism of thrombosis remains to be defined. In vitro assays are often prolonged but thrombosis rather than haemorrhage is the primary concern. Management of thrombotic events in these patients is based around agents that rely on AT for their anticoagulant effect, even though it is usually depleted. There is currently only limited evidence supporting the use of AT concentrates in either primary prevention or management following an established event. Evidence-based guidelines for prevention and management strategies are lacking.

Leukemia (2013) 27, 553–559; doi:10.1038/leu.2012.290

Keywords: L-Asparaginase; acute lymphoblastic leukaemia; thrombosis

10 INTRODUCTION potentially reversible cause of morbidity and mortality. L-Asp is In 1953, Kidd1 observed that guinea pig , but not rabbit or known to increase the thrombotic risk in patients with ALL, horse serum, could induce regression of transplanted mouse particularly in adults. This review discusses the clinical evidence lymphoma cells in vivo. In the 1960s, evidence emerged that the L- for this association and explores the available laboratory evidence asparaginase (L-Asp) activity of guinea pig serum was responsible to explain the underlying mechanism. for the anti-lymphoma effect.2 This experimental work paved the way for the subsequent clinical trial in which L-Asp was shown to be an active agent in children with relapsed acute lymphoblastic EPIDEMIOLOGY leukaemia (ALL).3 The pathogenesis of thrombosis in ALL is likely to be therapeu- L-Asp has been a standard component of paediatric ALL regimens tically related, for thrombosis rarely occurs at presentation, with since the 1980s and is usually combined with a corticosteroid, the majority of events occurring during remission induction.10 vincristine and intrathecal in induction for both Thrombosis is uncommon in children but the reported incidence paediatric and adult patients.4,5 Dramatic improvements in the in those receiving therapy for ALL ranges from 1.7 to 36.7%, 5-year survival for children with ALL have been seen over recent depending on study design, treatment protocol and whether decades, while the overall survival in adults has also improved, it asymptomatic events are included.10,11 remains much lower than in children.5,6 Outcome has continued to The PARKAA (Prophylactic Antithrombin Replacement in Kids improve without the introduction of novel agents, with large with ALL treated with Asparaginase) study reported a prevalence randomised clinical trial data identifying aggressive central nervous rate of thrombosis of 36.7% in 60 children during induction; system (CNS) prophylaxis and treatment, early post-induction however, the vast majority were asymptomatic, being diagnosed intensification and increased intensity of non-myelotoxic agents during extensive screening.12 The prevalence of symptomatic 7 as the key factors. The inclusion of L-Asp therapy has clearly played thrombosis was 5%. A meta-analysis of 17 prospective studies an important role and there is evidence to suggest a direct including 1752 paediatric patients found the incidence of 8 relationship between outcome in ALL and L-Asp dose intensity. symptomatic thrombosis to be 5.2%, with the majority of events 13 The preparations of L-Asp used in the treatment of ALL are occurring during induction. An incidence of 4.8% was observed derived from the bacteria Escherichia coli (E. coli) and in the 1280 patients receiving varying doses of L-Asp. A Erwinia chrysanthemi.9 Three product types are currently available: considerable number given the estimated annual incidence of native E. coli asparaginase, a pegylated form of the native E. coli in the paediatric population is 0.001%.13 asparaginase (polyethylene glycol-asparaginase) and Erwinia The incidence of symptomatic thrombosis appears to be higher asparaginase (Erwinase). in adult patients. A prospective study of 548 patients treated on Thrombosis is a complication associated with ALL and its Dana-Farber Institute (DFCI) ALL consortium protocols therapy in both paediatric and adult practice, and represents a including L-Asp diagnosed a total of 48 symptomatic clots in 43

1Department of Haematology, Centre for Clinical Haematology, Nottingham University Hospitals, Nottingham, UK; 2Department of Haematology, University College London, London, UK and 3Departments of Haematology, and , Kings College London, London, UK. Correspondence: Dr E Truelove, Centre for Clinical Haematology, Nottingham University Hospitals NHS Trust, Nottingham City Hospital, Hucknall Road, Nottingham NG5 1PB, UK. E-mail: [email protected] Received 5 September 2012; revised 2 October 2012; accepted 3 October 2012; accepted article preview online 9 October 2012; advance online publication, 26 October 2012 Risk associated with L-Asp treatment in ALL adults E Truelove et al 554 patients.11 Overall, 5% (27) of 501 paediatric patients and 34% (16) There was no difference in CNS thrombosis in a randomised of 47 adult patients experienced thrombotic events during comparison of native E. coli L-Asp and polyethylene glycol-L-Asp in 23 treatment phases that included L-Asp. However, in this cohort a study of 118 children, with 2 (3.4%) events in each arm. the majority of events occurred post-induction with a median time Children treated on the DFCI consortium protocol 05-01 received a of 3.5 months (0.5–10.1) from initiation of therapy to thrombosis. single dose of intravenous polyethylene glycol-L-Asp and venous A retrospective review of 54 adult patients treated consecu- thrombotic events occurred in 4 (2%) of patients, 2 being in the 24 tively with L-Asp containing regimens found an incidence of CNS. 18.5% symptomatic, objectively confirmed events during induc- tion.14 Caruso et al.15 published results of a meta-analysis of 13 prospective studies involving 323 adults with ALL, which EFFECT OF CORTICOSTEROID collectively identified a 5.9% incidence of thrombotic events Corticosteroids are an established component of ALL therapy and during induction. are usually given in combination with L-Asp and other chemother- There are no randomised controlled trials to allow accurate apeutic agents during induction. High levels of plasma corticos- assessment of the risk directly attributable to L-Asp therapy in ALL, teroids, such as those found in patients with Cushing’s syndrome, but it is notable that no episodes of thrombosis were reported have been associated with an increased risk of thrombosis. in a series of 204 adult patients treated with the intensive, multi- Elevated levels of coagulation factors (F) VIII, FIX and von agent chemotherapy regimen Hyperfractionated Cyclophosphamide, Willebrand’s Factor (vWF) are thought to contribute to the Vincristine, Doxorubicin and Dexamethasone (hyper-CVAD), which hypercoagulable state observed in these patients.25 Results from 16 does not include L-Asp. several in vitro studies have suggested that corticosteroids may directly activate coagulation and inhibit fibrinolysis by increasing the synthesis and secretion of vWF and plasminogen activator SITES OF THROMBOSIS inhibitor-1 (PAI-1).26 The majority of events occurring in association with L-Asp are Prednisolone has been associated with an increase in FVIII, vWF venous, but arterial thrombosis is reported.10,17 The PARKAA study and PAI-1.27,28 Dexamethasone has been shown to cause a of L-Asp-treated children with a central venous line (CVL), which modest but clear rise in the activity of FVII, FVIII, FIX and fibrinogen included asymptomatic events, reported 83% (19 of 23) in healthy volunteers, but to have no significant effect on PAI-1.29 thromboses to occur in the upper central venous system, 13% The use of corticosteroids during inflammatory states has been (3) in the right atrium and 4% (1) in the CNS.12 consistently associated with reduced levels of vWF and fibrinogen A meta-analysis of 1752 children reported 91 events, with 53.8% but increased PAI-1.26 Systemic inflammation is known to be a (49) involving the CNS, 42.8% (39) being venous in other sites and potent prothrombotic stimulus and the inflammatory and the remainder unspecified.13 Only 26 of the CNS events were coagulation systems are closely linked, therefore some of the clearly identified as being of venous origin with the others labelled prothrombotic changes induced by corticosteroid therapy may be as cerebral thrombosis, or . Of the non-CNS balanced by the anti-inflammatory actions.26,30,31 venous events, 64.1% (25) were identified as being CVL associated. There are no randomised trials comparing the thrombosis risk The overall risk of a CNS event was determined as 2.9% and a non- associated with different corticosteroids in L-Asp-exposed patients. CNS event as 2.3%. Of note, the site of event was not analysed in However, a comparison of paediatric patients treated according to relation to L-Asp therapy, but the majority of events did occur in different Berlin-Frankfurt-Munster protocols indicated a much patients receiving L-Asp. No peripheral arterial events were lower risk during induction with dexamethasone (1.8%) than with 32 reported. A further study in children receiving L-Asp focused prednisolone (10.4%). only on CNS events, finding 7 symptomatic CNS thromboses in A meta-analysis of prospective paediatric studies found a lower 719 patients, giving an overall rate of 1%.18 rate of thrombosis in those exposed to dexamethasone (2.0%) as The more recently published data from both adult and opposed to prednisolone (4.9%) during induction, but this paediatric patients treated according to DFCI consortium proto- difference was not significant.13 No difference in thrombosis cols identified 47 venous events, with 36% (17) being upper rates was seen between patients receiving L-Asp and extremity/CVL related, 19% (9) lower extremity, 19% (9) CNS and corticosteroids concomitantly and separately. 15% (7) pulmonary .11 The remainder were described as intra-cardiac (2%) and ‘other’ (8.5%). The sites of event were similar across adult and paediatric patients, but with a tendency MECHANISM OF ACTION OF ASPARAGINASE for adults to experience more and children L-Asp catalyses the hydrolysis of L-asparagine (ASN) to L-aspartic more CNS events. acid and ammonia, resulting in depletion of the circulating pool of A meta-analysis of adult studies involving 323 patients with 19 ASN.32 Although ASN is not an essential , human events found deep thrombosis originating in the lower limbs lymphoblasts typically have low levels of ASN synthetase activity the most common site (38.9%) with no arterial events.15 and are dependent on extracellular sources of ASN for synthesis.33 ALL cells require significant amounts of ASN for protein synthesis to maintain rapid proliferation, with a deficiency EFFECT OF L-ASP PREPARATION ON THROMBOSIS leading to cytotoxicity.34 L-Asp is a universal component of paediatric regimens used to ASN is required for dividing cells to proceed through the G1 treat ALL and is more often than not included in adult regimens. phase of the cell cycle, which ensures that the cell is ready for DNA However, the debate over the optimal preparation and dosage synthesis and is marked by high rates of synthesis of the enzymes continues. Two large randomised studies have shown superior required in S phase. Depletion of ASN is therefore not tolerated by efficacy for E. coli L-Asp when compared with Erwinia L-Asp in dividing cells and in vitro apoptotic cell death of leukaemia terms of complete remission rate, relapse rate and overall survival, cells induced by L-Asp is associated with cell-cycle arrest in the 19,20 35,36 but at the expense of increased toxicity. Erwinia L-Asp is often G1 phase. used as a second- or third-line agent in patients developing L-Asp is also known to catalyse the hydrolysis of glutamine hypersensitivity to E. coli L-Asp preparations. Erwinia L-Asp has also (GLU) to glutamic acid, resulting in depletion of circulating levels been linked to thrombosis in the literature but population sizes of this amino acid.37 GLU acts as a nitrogen donor for a number of have been small and doses given often differ from those currently enzymes involved in the de novo synthesis of purine and recommended.21,22 pyrimidine nucleotides, which are essential for the synthesis of

Leukemia (2013) 553 – 559 & 2013 Macmillan Publishers Limited Risk associated with L-Asp treatment in ALL adults E Truelove et al 555 RNA and DNA in proliferating cells.38 Furthermore, the synthesis of fibrin deposition. Most coagulation factors are serine proteases, ASN from L-aspartic acid requires GLU as the amide donor. which are regulated by serine protease inhibitors (Serpins), acting Adequate cellular concentrations of ASN and GLU are therefore as natural . necessary for RNA, DNA and protein synthesis in dividing cells, and Antithrombin (AT) is an important serpin, with a range of by depleting the availability of these amino acids L-Asp is able to inhibitory influences on the coagulation cascade, most notably disrupt the proliferation of lymphoblasts and induce apoptosis. (IIa) and FXa, but also FIXa, FXIa, FXIIa, tissue plasminogen activator, , and .45 Patients with a heterozygous deficiency of AT have a high risk of venous EFFECT OF L-ASP ON COAGULATION SCREENING TESTS thromboembolism (VTE)—as a general rule 50% will have a Currently available laboratory screening tests to assess the integrity venous VTE before the age of 50. Low levels of AT during L-Asp of the coagulation cascade are in vitro assays that cannot reliably therapy in childhood ALL have been reported by a number of 41,42,46 replicate in vivo haemostasis and therefore have limitations. They studies. The effect attributable to L-Asp specifically on AT have, however, been widely used to monitor haemostatic changes levels is difficult to interpret in some series due to the failure to occurring during L-Asp therapy. The most commonly employed measure other components of the haemostatic pathway. The tests are the activated partial thromboplastin time (APTT), deficiency observed could be a manifestation of failure in hepatic (PT), and fibrinogen. synthetic function, exacerbated by consumption during L-Asp can have profound effects on both the coagulation and formation. It is likely that the haemostatic balance of physiological fibrinolytic systems, resulting in reduced plasma levels of a anticoagulants versus coagulation factors is set at a lower level number of coagulation factors, including fibrinogen, FV, FVII, FVIII, following L-Asp. 42 FIX, FX and FXI as well as other key elements of haemostasis.39 The Homans et al. demonstrated that AT levels were reduced to APTT, PT and thrombin time are sensitive to decreased levels of 34% of pre-treatment levels when L-Asp was given as a single certain coagulation factors and prolongation of one or more of agent to children with ALL in remission and returned to baseline 14 these clotting times is often seen. days after cessation of therapy. The DFCI protocol 87-001 evaluated A number of studies have demonstrated that coagulation the in vivo effect of single agent L-Asp on AT levels in untreated screens in paediatric patients with ALL are sensitive to the effects paediatric patients with ALL. Levels at presentation were similar to 40–42 of L-Asp. The first study evaluated 26 patients during healthy age-matched controls, but significantly decreased after induction with vincristine, prednisolone and L-Asp and reported L-Asp therapy, returned to baseline after combination prolongation of the PT in 16, the APTT in 23 and the thrombin chemotherapy (without L-Asp) and again significantly decreased 46 time in 21 patients.40 All patients had reduced fibrinogen levels after consolidation (including L-Asp). The more recent CAPELAL and all parameters returned to normal within 2 weeks of study in adult patients reported that median AT levels of 120% 47 completion of L-Asp. Another study involving 13 patients during before first L-Asp infusion dropped to 59% by the fourth infusion. induction reported a similar pattern of results.41 However, a study During this study, 20 thrombotic events occurred in 9.3% of of patients exposed to single agent L-Asp once in remission patients from 2 to 35 days after the first L-Asp infusion. reported that although clotting times were prolonged from As with other serpins, AT has a structural flexibility that allows a baseline, the PT remained within normal limits and the APTT conformational change essential to its inhibitory function on 48 was only marginally prolonged.42 target proteases. However, this feature makes serpins susceptible to conformational changes induced by amino-acid modifications in structurally sensitive regions of the molecule. MECHANISMS OF THROMBOSIS Inherited AT deficiency is associated with a wide range of mutations, including those with conformational effects, which Haemostasis is a delicately balanced homeostatic process that 45,48 maintains fluidity and also the integrity of the cardiovascular tend to have a clinically severe phenotype. system after tissue injury by coordinating rapid and localised The acquired AT deficiency induced by L-Asp can be explained thrombus formation. The five major components of haemostasis— by a global reduction in hepatic protein synthesis. Two groups the vascular , , the coagulation cascade, have previously shown that L-Asp has no direct effect on mature AT in terms of antigen levels or activity, as detected by anti-FXa physiological anticoagulants and fibrinolysis—are tightly regulated 49,50 under physiologic conditions, but when pathological states over- activity. These groups have also demonstrated that whelm these regulatory mechanisms or the mechanisms are hepatocarcinoma (HepG2) cells treated with L-Asp show a 43 marked reduction in AT secretion but no reduction in AT mRNA defective, thrombosis can result. The association between L-Asp and thrombosis has been attributed to reduced biosynthesis of the levels, suggesting interference at a post-transcriptional stage. Further experiments to assess the mechanism of L-Asp on AT natural anticoagulant proteins; however, L-Asp also reduces levels of the coagulation proteins as evidenced by the fall in fibrinogen and secretion found intracellular accumulation of AT aggregates in increment in PT and APTT. This suggests that there may be a global HepG2 cells and observed a similar pattern in mice, with retention and aggregation of the protein within the lumen of dilated rough inhibition of production of both coagulant and anticoagulant 50 proteins in the , which may be exacerbated by consumption of cisterns. There is, however, some haemostatic factors in the formation of a thrombus but does not conflicting data within the literature, with results of an earlier study suggesting that L-Asp may have a direct effect on the AT reliably explain the mechanism of thrombosis. 51 There is evidence from a study of 23 children with ALL treated protein and other coagulation system proteins. cofactor II (HCII) is another serpin that inhibits with the Berlin-Frankfurt-Munster protocol that thrombin genera- 48 tion and peak thrombin were significantly higher during induction thrombin, protease nexin 1 and C1 inhibitor. Deficiency of HCII than re-induction.44 Interestingly, two of these patients developed has been linked to thrombosis, but the significance of its role at the time of peak endogenous thrombin remains unclear. L-Asp treatment of HepG2 resulted in decreased generation. secretion of HCII and Mitchell et al. found that HCII plasma concentrations were raised in children with ALL, but significantly 46,49 decreased following therapy with L-Asp alone. Physiological anticoagulants The system is the physiological pathway responsible The rapid and efficient response of the coagulation cascade to for degrading the activated coagulation cofactors, FVa and FVIIIa, haemorrhage is achieved by serial proteolytic cleavage of as well as reducing prothrombinase activity. Protein C is circulating inactive coagulation factors, which culminates in local activated by the thrombin-thrombomodulin (TM) complex on the

& 2013 Macmillan Publishers Limited Leukemia (2013) 553 – 559 Risk associated with L-Asp treatment in ALL adults E Truelove et al 556 surface of endothelial cells and its effects are facilitated via the on the surface of the endothelium, such that the endothelium cofactor activity of . Both protein C and S are loses its ability to activate Protein C. Other changes include dependent and deficiency of either, as well as resistance to expression of (TF), cleavage of heparin sulphate activated protein C is associated with thrombosis.52 The (which usually potentiates AT) and increased expression of PAI-1. administration of L-Asp as a single agent to children with ALL in L-Asp has been shown to disrupt endothelial function. remission has been associated with marked reductions in protein P-selectin, a , and PAI-1 have been studied 42 59 C antigen levels. A significant decline in both protein C and free in children with ALL during induction inclusive of L-Asp. Levels of protein S occurred in paediatric patients treated with L-Asp on the both were noted to be increased at diagnosis compared with DFCI protocol 87-001.46 Two small studies in adult patients controls and continued to increase during induction. reported results consistent with the pattern seen in children, with In a study of 26 children treated on the FRALLE (French Acute marked reductions in the anticoagulant activities of protein C and Lymphoblastic Leukemia) group—2000 protocol, TF activity and 53,54 S following L-Asp. vWF antigen levels significantly increased with the number of 60 L-Asp infusions during induction. TM activity was also increased compared with controls but remained in the same range as values Fibrinolytic system seen at diagnosis. In vitro experiments indicate that L-Asp can is the process by which fibrin clots are dissolved and is induce a modest increase of TF expression on lymphoblasts, a requirement of normal haemostasis. The serine protease plasmin whereas a marked increase was observed on endothelial cell is the predominant fibrinolytic , but circulates as a lines.60 These in vitro observations again support the possible role proenzyme, plasminogen. Plasminogen is activated by tissue of endothelial and/or leukocyte activation in upregulation of TF plasminogen activator, released from endothelial cells as well as during the L-Asp-associated hypercoagulable state. by urokinase-type plasminogen activators, FXIIa and kallikrein. Plasmin generation at the site of vascular injury limits the extent of thrombus formation by directly degrading fibrin, fibrinogen, FV, PREVENTION AND MANAGEMENT OPTIONS FVII and FVIII. Inhibition of fibrinolysis is predominantly controlled Management of patients being treated with L-Asp presents a by two serpins, PAI-1, which inhibits tissue plasminogen activator considerable clinical dilemma. The APTT and PT are often and urokinase-type plasminogen activator, and a2-antiplasmin, 48 prolonged in association with reduced levels of haemostatic which inhibits plasmin. proteins and fibrinogen but thrombosis rather than haemorrhage Fibrinolytic parameters have been studied in L-Asp-treated ALL is the predominant problem. The thrombosis risk in adults is patients, with reduced fibrinogen levels being consistently 41,42,47,55 greater than in children receiving the drug; for adults have more reported. There is uncertainty whether depletion of prothrombotic physiological haemostasis than children, and this is fibrinogen is due to lack of production, generation of dysfibrino- compounded by additional risk factors in adults such as smoking, genaemia and/or excessive fibrinolytic activity, compounded by and poor performance status leading to a higher degree of increased consumption during thrombotic episodes. Plasminogen immobility. and a2-antiplasmin levels have also been shown to decline Since thrombosis is commonly related to an indwelling CVL, it is following L-Asp and again it is uncertain whether this reflects lack generally recommended that their placement should be post- of production by the liver, increased consumption, or both. poned until the first phase of induction therapy is complete; however, evidence to support this practice is lacking. Optimal Platelets timing for CVL insertion in patients with ALL is unknown. One A venous thrombus usually develops in conditions of low shear or retrospective study of paediatric ALL patients treated with L-Asp stasis and is predominantly composed of red cells and fibrin with during induction found that early CVL placement was not associated with an increased risk of thrombosis compared with relatively few platelets; thus, disorders affecting platelets are rarely 61 associated with increased risk of venous thrombosis. However, the delayed insertion. In addition, delayed CVL insertion is not always hypothesis that abnormal platelet aggregation in response to practical, particularly since adult therapeutic regimens often L-Asp may contribute to the associated thrombotic risk has been contain more doses of vesicant drugs, such as anthracyclines. investigated. Increased platelet sensitivity to ADP during induction In the past, treatment protocols frequently specified that with L-Asp has been reported in one paediatric study, while abnormalities in the be used to guide plasma another demonstrated normal platelet function studies through- replacement with fresh frozen plasma or . Evidence 41,42 out L-Asp therapy for patients in remission. This latter result to support this recommendation is lacking and available data now was supported by a study of 20 children, which found no argue against the practice. A paediatric study found that fresh difference in aggregation with standard agonists between ALL frozen plasma replacement had no demonstrable effect on 56 laboratory parameters of coagulation with only minimal improve- patients and controls. However, an in vitro agonist effect of L-Asp 62 on platelets from both leukaemic patients and controls was ments in fibrinogen. A retrospective analysis of paediatric patients documented, raising the possibility of a similar effect in vivo, but receiving L-Asp found that neither fresh frozen plasma nor this remains to be demonstrated. cryoprecipitate protected against CNS thrombosis and concluded that prophylactic plasma replacement is not warranted.18 Furthermore, there is evidence that fresh frozen plasma contains Endothelium ASN, which might compromise L-Asp therapy by replenishing the The endothelium has a vital role in the regulation of blood flow ASN pool the therapy is intended to deplete. and haemostasis by generating an antithrombotic surface due to concentrates are increasingly used in the management of expression of TM and heparan sulphate, as well as inhibiting hypofibrinogenaemia but have been associated with throm- 63 platelets via nitric oxide and prostacyclin (PGI2). However when bosis, and concern remains about the thrombotic potential that stimulated by agents such as 1, tumour necrosis factor may be associated with their use in L-Asp-treated patients. or thrombin, it undergoes a transformation—known as endothe- Thrombotic problems are much more common than haemor- lial cell activation—whereby it expresses a prothrombotic rhage but the presence of abnormalities of the APTT, PT and/or phenotype.57 TM exerts its effect by binding with thrombin, fibrinogen present a practical problem when patients require transforming it into a highly effective anticoagulant that activates invasive procedures. A particular concern is , protein C and reduces platelet activation.58 Endothelial activation since intrathecal drug administration is invariably a component of results in downregulation of TM production and pinocytosis of TM ALL management. Members of staff are reluctant to perform such

Leukemia (2013) 553 – 559 & 2013 Macmillan Publishers Limited Risk associated with L-Asp treatment in ALL adults E Truelove et al 557 Proposed mechanism by which L-asparaginase causes thrombosis

L-asparaginase

Endothelial and/or white cell activation Reduced haemostatic protein synthesis Increased tissue factor expression

Thrombin generation, consumption of coagulation factors + physiological anticoagulants

External prothrombotic Hypofibrinogenaemia, prolonged APTT/PT & factors e.g reduced antithrombin and other physiological anticoagulants Increased age Immobility Infection

THROMBOSIS Underlying ? Corticosteroids

Figure 1. The proposed mechanism by which L-Asparaginase causes thrombosis. procedures without attempting to correct the coagulation seen.68 The new oral direct thrombin inhibitors and anti-Xa abnormalities and this still represents an unresolved problem. anticoagulants are currently being assessed in medical patients for thromboprophylaxis, and if these trials demonstrate efficacy and Prophylactic AT supplementation safety then study in L-Asp-treated patients would be warranted. It is likely that pharmacological thromboprophylaxis will have an Several studies have addressed the prevention of thrombosis. AT increasing role in the management of L-Asp-exposed patients. supplementation has been explored in paediatric and adult populations during L-Asp therapy, with initial studies concluding that although enhanced thrombin generation and d-dimer levels Management of thrombosis associated with therapy returned towards normal with AT Management of thrombotic events requires therapeutic antic- replacement, further randomised trials were required to demon- oagulation. A number of options are available. LMWH is usually 54,62 strate whether a clinical benefit existed. The more recent preferred to vitamin K antagonists in patients receiving concurrent PARKAA study showed a trend towards efficacy and safety with AT myelosuppressive therapy. Unfractionated heparin is preferable in concentrate in prevention of thrombosis in children treated with L- those with renal failure or at a high risk of haemorrhage, who may 64 Asp, but the study was not powered to answer this question. The need rapid anticoagulation reversal. Heparin is reliant on AT for its CAPELAL study reported lower rates of thrombosis in adults who therapeutic effect and resistance may be encountered with AT received prophylactic AT concentrate (4.5%) than in those who deficiency engendered by L-Asp. AT monitoring and replacement 47 had not (12.7%). However, this was a retrospective analysis and may be necessary to maintain therapeutic anticoagulation with again the authors concluded that further work is required to heparin and LMWH in L-Asp-treated patients. A more attractive define the role of AT in the management of L-Asp-treated patients. option would be to use an anticoagulant with a direct anti- coagulant effect not mediated through AT. The use of one of the Thromboprophylaxis new oral agents such as rivaroxaban or dabigatran is worthy of Pharmacological prophylaxis to prevent hospital-acquired VTE is exploration in the future, although their irreversibility may limit widely and safely used, and current guidelines recommend it be their use. offered to medical patients assessed to be at increased risk.65 It seems logical that use of anticoagulant agents for primary VTE prophylaxis in L-Asp-treated patients will reduce rates of SUMMARY thrombosis, but randomised clinical trial data are lacking. A non- Therapy with L-Asp is associated with an increased thrombotic risk randomised study of 41 paediatric patients with ALL using the low with alterations in the levels of multiple haemostatic proteins molecular weight heparin (LMWH) enoxaparin as prophylaxis being consistently demonstrated. Although the thrombotic risk in during L-Asp therapy reported no thrombotic events during 76 the past has been attributed to the acquired AT deficiency—which courses of L-Asp, with 1 brain infarct reported 7 days after clearly occurs, the levels of both coagulation factors and 66 cessation of prophylaxis. LMWH has also been used successfully anticoagulant proteins are reduced. We suggest that while L-Asp as secondary prophylaxis in patients re-exposed to L-Asp following reduces haemostatic protein production, it also seems biologically 67 an initial VTE without recurrence or excess . However, plausible that L-Asp exposure results in upregulation of TF through there remains concern that , which rely on AT for their white cell and/or endothelial activation leading to thrombosis mechanism of action, may not be the most appropriate option. initiation and consumption of haemostatic proteins. Clinical data are again lacking but in vitro experiments using In some patients, the presence of additional risk factors plasma from children with L-Asp-induced AT deficiency have including an underlying thrombophilia, , obesity, increasing demonstrated a consistent anticoagulant response with direct age and immobility will result in an overwhelming drive to thrombin inhibitors, independent of AT levels, whereas a profound thrombosis (see Figure 1). However, many questions remain effect on the action of LMWH related to AT concentration was unanswered and this area requires further study.

& 2013 Macmillan Publishers Limited Leukemia (2013) 553 – 559 Risk associated with L-Asp treatment in ALL adults E Truelove et al 558 The available evidence suggests that AT replacement may have nervous system thrombosis and haemorrhage in children with acute lympho- a role in prophylaxis against L-Asp-associated thrombosis but this blastic leukemia receiving asparaginase. Blood 2009; 114: 5146–5151. requires confirmation in randomised trials. It would seem logical 19 Duval M, Suciu S, Ferster A, Rialland X, Nelken B, Lutz P et al. Comparison of to use pharmacological anticoagulants for prophylaxis, particularly Escherichia coli-asparaginase with Erwinia-asparaginase in the treatment of those which do not exert their mechanism of action through AT. A childhood lymphoid malignancies: results of a randomized European organisation detailed understanding of the thrombotic mechanism will likely for research and treatment of cancer – children’s leukemia group phase 3 trial. be a key to defining the optimal strategy for prevention and Blood 2002; 99: 2734–2739. 20 Moghrabi A, Levy DE, Asselin B, Barr R, Clavell L, Hurwitz C et al. Results of the management of L-Asp-associated thrombosis. Dana-Farber Cancer Institute ALL consortium protocol 95-01 for children with acute lymphoblastic leukemia. Blood 2007; 109: 896–904. 21 Carlsson H, Stockelberg D, Tengborn L, Braide I, Carneskog J, Kutti J. Effects CONFLICT OF INTEREST of Erwinia-asparaginase on the coagulation system. Eur J Haematol 1995; 55: The authors declare no conflict of interest. 289–293. 22 Risseeuw-Appel IM, Dekker I, Hop WC, Hahlen K. Minimal effects of E. coli and Erwinia asparaginase on the coagulation system in childhood acute lympho- REFERENCES blastic leukemia: a randomized study. Med Pediatr Oncol 1994; 23: 335–343. 1 Kidd JG. Regression of transplanted lymphomas induced in vivo by means of 23 Avramis VI, Spencer S, Periclou AP, Sather H, Bostrom BC, Cohen LJ et al. normal guinea pig serum. I. Course of transplanted of various kinds in A randomized comparison of native Escherichia coli asparaginase and poly- mice and rats given guinea pig serum, horse serum, or rabbit serum. J Exp Med ethylene glycol conjugated asparaginase for treatment of children with newly 1953; 98: 565–582. diagnosed standard-risk acute lymphoblastic leukemia: a Children’s Cancer Group 2 Broome JD. Evidence that the L-asparaginase of guinea pig serum is responsible study. Blood 2002; 99: 1986–1994. for its antilymphoma effects. I. Properties of the L-asparaginase of guinea pig 24 Silverman LB, Supko JG, Stevenson KE, Woodward C, Vrooman LM, Neuberg DS serum in relation to those of the antilymphoma substance. J Exp Med 1963; 118: et al. Intravenous PEG-asparaginase during remission induction in children and 99–120. adolescents with newly diagnosed acute lymphoblastic leukemia. Blood 2010; 3 Ertel IJ, Nesbit ME, Hammond D, Weiner J, Sather H. Effective dose of L-aspar- 115: 1351–1353. aginase for induction of remission in previously treated children with acute 25 Van Zaane B, Nur E, Squizzato A, Dekkers OM, Twickler MT, Fliers E et al. Hyper- lymphoblastic leukemia: a report from children’s cancer study group. Cancer Res coagulable state in Cushing’s syndrome: a systemic review. J Clin Endocrinol 1979; 39: 3893–3896. Metab 2009; 94: 2743–2750. 4 Douer D. Is asparaginase a critical component in the treatment of acute lym- 26 Van Zaane B, Nur E, Squizzato A, Gerdes VE, Buller HR, Dekkers OM et al. Systemic phoblastic leukemia? Best Pract Res Clin Haematol 2008; 21: 647–658. review on the effect of use on procoagulant, anti-coagulant and 5 Pui CH, Evans WE. Treatment of acute lymphoblastic leukemia. N Engl J Med 2006; fibrinolytic factors. J Thromb Heamost 2010; 8: 2483–2493. 354: 166–178. 27 Isacson S. Effect of prednisolone on the coagulation and fibrinolytic systems. 6 Seibel NL. Treatment of acute lymphoblastic leukemia in children and adolescents: Scand J Haematol 1970; 7: 212–216. peaks and pitfalls. Am Soc Hematol Educ Program 2008; 1: 374–380. 28 Ozsoylu S, Strauss HS, Diamond LK. Effects of corticosteroids on coagulation of 7 Sallan SE, Hitchcock-Bryan S, Gelber R, Cassady JR, Frei III E, Nathan DG. Influence blood. Nature 1962; 195: 1214–1215. of intensive asparaginase in the treatment of childhood non-T-cell acute lym- 29 Brotman DJ, Girod JP, Posch A, Jani JT, Patel JV, Gupta M et al. Effects of short- phoblastic leukemia. Cancer Res 1983; 43: 5601–5607. term on hemostatic factors in healthy volunteers. Thromb Res 8 Silverman LB, Gelber RD, Dalton VK, Asselin BL, Barr RD, Clavell LA et al. Improved 2006; 118: 247–252. outcome for children with acute lymphoblastic leukemia: results of Dana-Farber 30 Esmon CT. Inflammation and thrombosis. J Thromb Haemost 2003; 1: 1343–1348. Consortium Protocol 91-01. Blood 2001; 97: 1211–1218. 31 Nowak-Gottl U, Ahlke E, Fleischhack G, Schwabe D, Schobess R, Schumann C et al. 9 Pieters R, Hunger SP, Boos J, Rizzari C, Silverman L, Baruchel A et al. L-aspar- Thromboembolic events in children with acute lymphoblastic leukemia (BFM aginase treatment in acute lymphoblastic leukemia. A focus on Erwinia aspar- protocols): prednisone versus dexamethasone administration. Blood 2003; 101: aginase. Cancer 2011; 117: 238–249. 2529–2533. 10 Payne JH, Vora AJ. Thrombosis and acute lymphoblastic leukaemia. Br J Haematol 32 Van Den Berg H. Asparaginase revisited. Leuk Lymphoma 2011; 52: 168–178. 2007; 138: 430–445. 33 Earl M. Incidence and management of asparaginase-associated adverse events in 11 Grace RF, Dahlberg SE, Neuberg D, Sallan SE, Connors JM, Neufeld EJ et al. The patients with acute lymphoblastic leukemia. Clin Adv Haematol Oncol 2009; 7: frequency and management of asparaginase-related thrombosis in paediatric and 600–606. adult patients with acute lymphoblastic leukaemia treated on Dana-Farber Cancer 34 Asselin BL, Ryan D, Frantz CN, Bernal SD, Leavitt P, Sallan SE et al. In vitro killing Institute consortium protocols. Br J Haematol 2011; 152: 452–459. of acute lymphoblastic leukemia cells by L-asparaginase. Cancer Res 1989; 49: 12 Mitchell LG, Andrew M, Hanna K, Abshire T, Halton J, Anderson R et al. A pro- 4363–4368. spective cohort study determining the prevalence of thrombotic events in chil- 35 Rytting M. PEG-asparaginase for acute lymphoblastic leukemia. Expert Opin Biol dren with acute lymphoblastic leukemia and a central venous line who are Ther 2010; 10: 833–839. treated with L-asparaginase. Results of the prophylactic antithrombin replace- 36 Ueno T, Ohtawa K, Mitsui K, Kodera Y, Hiroto M, Matsushima A et al. Cell cycle ment in kids treated with acute lymphoblastic leukemia treated with aspar- arrest and apoptosis of leukemia cells induced by L-asparaginase. Leukemia 1997; aginase (PARKAA) study. Cancer 2003; 97: 508–516. 11: 1858–1861. 13 Caruso V, Iacoviello L, Di Castelnuovo A, Storti S, Mariani G, de Gaetano G et al. 37 Grigoryan RS, Panosyan EH, Seibel NL, Gaynon PS, Avramis IA, Avramis VI. Thrombotic complications in childhood acute lymphoblastic: a meta-analysis of Changes of amino acid serum levels in pediatric patients with higher-risk acute 17 prospective studies comprising 1752 pediatric patients. Blood 2006; 108: lymphoblastic leukemia (CCG-1961). In Vivo 2004; 18: 107–112. 2216–2222. 38 Cory JG, Cory AH. Critical roles of glutamine as nitrogen donors in purine and 14 Elliot MA, Wolf RC, Hook CC, Pruthi RK, Heit JA, Letendre LL et al. Thromboem- pyrimidine nucleotide synthesis: asparaginase treatment in childhood acute bolism in adults with acute lymphoblastic leukemia during induction with lymphoblastic leukemia. In Vivo 2006; 20: 587–590. L-asparaginase-containing multi-agent regimens: incidence, risk factors, and 39 Zakarija A, Kwaan HC. Adverse effects on hemostatic function of drugs used in possible role of antithrombin. Leuk Lymphoma 2004; 45: 1545–1549. hematologic malignancies. Semin Thromb Hemost 2007; 33: 355–364. 15 Caruso V, Iacoviello L, Di Castelnuovo A, Storti S, Donati MB. Venous thrombotic 40 Ramsey NK, Coccia PF, Krivit W, Nesbit ME, Edson JR. The effect of L-asparaginase complications in adults undergoing induction treatment for acute lymphoblastic of plasma coagulation factors in acute lymphoblastic leukemia. Cancer 1977; 40: leukemia: results from a meta-analysis. J Thromb Haemost 2007; 5: 621–623. 1398–1401. 16 Kantarjian HM, O’Brien S, Smith TL, Cortes J, Giles FJ, Beran M et al. Results of 41 Pui CH, Jackson CW, Chesney C, Lyles SA, Bowman WP, Abromowitch M et al. treatment with Hyper-CVAD, a dose-intensive regimen, in adult acute lympho- Sequential changes in platelet function and coagulation in leukemic children cytic leukemia. J Clin Oncol 2000; 18: 547–561. treated with L-asparaginase, prednisolone and vincristine. J Clin Oncol 1983; 1: 17 Gugliotta L, Mazzucconi MG, Leone G, Mattioli-Belmonte M, Defazio D, Annino L 380–385. et al. Incidence of thrombotic complications in adult patients with acute lym- 42 Homans AC, Rybak ME, Baglini RL, Tiarks C, Steiner ME, Forman EN. Effect of phoblastic leukaemia receiving L-asparaginase during induction therapy: a ret- L-asparaginase administration on coagulation and platelet function in children rospective study. The GIMEMA group. Eur J Haematol 1992; 49: 63–66. with leukemia. J Clin Oncol 1987; 5: 811–817. 18 Abbott LS, Deevska M, Fernandez CV, Dix D, Price VE, Wang H et al. The impact of 43 Furie B, Furie BC. Mechanisms of thrombus formation. N Engl J Med 2008; 359: prophylactic fresh-frozen plasma and cryoprecipitate on the incidence of central 938–949.

Leukemia (2013) 553 – 559 & 2013 Macmillan Publishers Limited Risk associated with L-Asp treatment in ALL adults E Truelove et al 559 44 Leihanacova-Tousovska K, Zapletal O, Vytiskova S, Strbackova P, Sterba J. Profile 56 Jaime-Perez JC, Gomez-Almaguer D. Platelet aggregation is stimulated by of thrombin generation in children with acute lymphoblastic leukaemia treated L-asparaginase in children with acute lymphoblastic leukemia and in normal by Berlin-Frankfurt-Munster (BFM) protocols. Blood Coagul Fibribolysis 2012; 23: individuals. Haematologica 2002; 87: 891–892. 144–145. 57 Hunt BJ, Jurd KM. Endothelial cell activation - a central pathophysiological pro- 45 Patnaik MM, Moll S. Inherited antithrombin deficiency: a review. cess. BMJ 1998; 316: 1328. 2008; 14: 1229–1239. 58 Gresele P, Momi S, Migliacci R. Endothelium, venous thromboembolism and 46 Mitchell L, Hoogendoorn H, Giles AR, Vegh P, Andrew M. Increased endogenous ischaemic cardiovascular events. Thromb Haemost 2010; 103: 56–61. thrombin generation in children with acute lymphoblastic leukemia: risk of 59 Giordano P, Molinari AC, Del Vecchio GC, Saracco P, Russo G, Altomare M et al. thrombotic complications in L-asparaginase-induced antithrombin deficiency. Prospective study of hemostatic alterations in children with acute lymphoblastic Blood 1994; 83: 386–391. leukemia. Am J Haematol 2010; 85: 325–330. 47 Hunault-Berger M, Chevallier Delain M, Bulabois CE, Bologna S, Bernard M et al. 60 Schneider P, Van Dreden P, Rousseau A, Kassim Y, Legrand E, Vannier JP et al. Changes in antithrombin and fibrinogen levels during induction chemotherapy Increased levels of tissue factor activity and procoagulant phospholipids during with L-asparaginase in adult patients with acute lymphoblastic leukemia or treatment of children with acute lymphoblastic leukemia. Br J Haematol 2009; lymphoblastic lymphoma. Use of supportive coagulation therapy and clinical 148: 582–592. outcome: the CAPELAL study. Haematologica 2008; 93: 1488–1494. 61 McLean TW, Fisher CJ, Snively BM, Chauvenet AR. Central venous lines in children 48 Hernandez-Espinosa D, Ordonez A, Vicente V, Corral J. Factors with conforma- with lesser risk acute lymphoblastic leukemia: optimal type and timing of pla- tional effects on haemostatic serpins: implications for thrombosis. Thromb Hae- cement. J Clin Oncol 2005; 23: 3024–3029. most 2007; 98: 557–563. 62 Nowak-Gottl U, Kuhn N, Wolff JE, Boos J, Kehrel B, Rath B et al. Inhibition of 49 Bushman JE, Palmieri D, Whinna HC, Church FC. Insight into the mechanism of hypercoagulation by antithrombin substitution in E. coli L-asparaginase-treated asparaginase-induced depletion of antithrombin III in treatment of childhood children. Eur J Haematol 1996; 56: 35–38. acute lymphoblastic leukemia. Leuk Res 2000; 24: 559–565. 63 Gollop ND, Chilcott J, Benton A, Rayment R, Jones J, Collins PW. National audit of 50 Hernandez-Espinosa D, Minano A, Martinez C, Perez-Ceballos E, Heras I et al. the use of fibrinogen concentrate to correct hypofibrinogenaemia. Transfus Med L-asparaginase-induced antithrombin type 1 deficiency. Implications for 2012; 22: 350–355. conformational diseases. Am J Pathol 2006; 169: 142–153. 64 Mitchell L, Andrew M, Hanna K, Abshire T, Halton J, Wu J et al. Trend to efficacy 51 Nowak-Gottl U, Boos J, Wolff JE, Lill H, Veltmann H, Werber G et al. Asparaginase and safety using antithrombin concentrate in prevention of thrombosis in chil- decreases clotting factors in vitro: a possible pitfall? Int J Clin Lab Res 1995; 25: dren receiving L-asparaginase for acute lymphoblastic leukemia. Results of the 146–148. PAARKA study. Thromb Haemost 2003; 90: 163–164. 52 Walker ID, Greaves M, Preston FE. Investigation and management of heritable 65 CG92 Venous thromboembolism – reducing the risk: NICE guideline 2010. thrombophilia. Br J Haematol 2001; 114: 512–528. 66 Elhasid R, Lanir N, Sharon R, Weyl Ben Arush M, Levin C, Postovsky S et al. Pro- 53 Vigano-D’Angelo S, Gugliotta L, Mattioli Belmonte M, Cascione ML, Pattarini E, phylactic therapy with enoxaparin during L-asparaginase treatment in children D’Angelo A. L-asparaginase treatment reduces the anticoagulant potential of the with acute lymphoblastic leukemia. Blood Coagul Fibribolysis 2001; 12: 367–370. protein C system without affecting vitamin K-dependent carboxylation. Thromb 67 Qureshi A, Mitchell C, Richards S, Vora A, Goulden N. Asparaginase-related venous Res 1990; 59: 985–994. thrombosis in UKALL 2003- re-exposure to asparaginase is feasible and safe. Br J 54 Gugliotta L, D’Angelo A, Mattioli Belmonte M, Vigano-D’Angelo S, Colombo G, Haematol 2010; 149: 410–413. Catani L et al. Hypercoagulability during L-asparaginase treatment: the effect of 68 Kuhle S, Lau A, Bajzar L, Vegh P, Halton J, Cherrick I et al. Comparison of the antithrombin III supplementation in vivo. Br J Haematol 1990; 74: 465–470. anticoagulant effect of a direct thrombin inhibitor and a low molecular weight 55 Nowak-Gottl U, Werber G, Ziemann D, Ahlke E, Boos J. Influence of two different heparin in an acquired antithrombin deficiency in children with acute lympho- Escherichia coli asparaginase preparations on fibrinolytic proteins in childhood blastic leukaemia treated with L-asparaginase: an in vitro study. Br J Haematol ALL. Haematologica 1996; 81: 127–131. 2006; 134: 526–531.

& 2013 Macmillan Publishers Limited Leukemia (2013) 553 – 559