SHORT REPORT

Influence of genetic predisposition to thrombosison natural history of acute promyelocytic leukaemia

DAVID REE S,1 DAVI D GRIMWADE,2 STEPHEN LANGABEER,3 AL AN BURNE TT4

AND ANT HONY GOLDSTONE,3 on behalfof the MRC Adult Leukaemia Working Party 1Institute of Molecular Medicine, John Radcliffe Hospital, Oxford,2Imperial Cancer Research Fund, Lincoln’s Inn Fields, London, 3Department of Haematology, UCL Hospitals, London,and 4Department of Haematology, University of Wales, Cardiff

Received 5 August 1996; accepted for publication 12 November 1996

Summary. There is currently no means of identifying thesubgroup of APL patients who will succumb to haemorrhagicor thrombotic complications. We have investigated factor VLeiden and thermolabile methylene tetrahydrofolate reduc-tase (MTHFR) to determine whether these are of predictivevalue for thrombosis in the context of APL. Of 48 patientsdrawn from the MRC ATRA trial, two were heterozygous forfactor V Leiden (allele frequency 2·1%). 10 homozygotes and17 heterozygotes for thermolabile MTHFR were identified

(allele frequency 38·5%). Amongst these patients, onethrombosis occurred (thermolabile MTHFR heterozygote).In the group with no identified increased thrombotic risk,three episodes were recorded. This approach failed to predictthrombotic events in APL, although the exact implications ofspecific genotypes remain to be established by larger studies.

Keywords: thrombophilia, factor V Leiden, methylenetetrahydrofolate reductase, acute promyelocytic leukaemia.

Acute promyelocytic leukaemia (APL) is characterized by thet(15;17) leading to a PML/RARa rearrangement, a uniquesensitivity to all-trans retinoic acid (ATRA) and a potentiallydevastating coagulopathy (Grimwade & Solomon, 1996).The haemorrhagic tendency is well documented, but iscomplicated and incompletely understood, reflecting tovarying degrees increased procoagulant, fibrinolytic andproteolytic activities compounded by thrombocytopenia(Tallman & Kwaan, 1992). Less well appreciated than thebleeding tendency is the increased risk of thrombosis. ATRAleads to rapid amelioration of the bleeding, but there isincreasing concern that it may exacerbate thromboticcomplications (Rodeghiero & Castaman, 1994). Theinfluence of a genetic predisposition to thrombosis in APLis presently unknown. Therefore we have determined thefrequency of two common genetic markers of a prothrombotictendency, namely factor V Leiden and thermolabile methy-lene tetrahydrofolate reductase (MTHFR), amongst a groupof APL patients treated with ATRA and chemotherapy.

Factor V Leiden (Arg506→Gln), renders Va relativelyresistant to cleavage by activated protein C causing excessivethrombin generation and a presumed life-long prothrombotic

tendency (Bertina et al, 1994). It is the commonest knowncause of thrombophilia in Europeans, with an allelefrequency of up to 7% (Rees et al, 1995). Heterozygoteshave a 7-fold increase in risk of venous thrombosis andhomozygotes an 80-fold increase (Rosendaal et al, 1995).

Hyperhomocysteinaemia is recognized as an importantrisk factor for both arterial and venous thrombosis. Arecently identified polymorphism in the MTHFR gene,(nt677, C→T, Ala→Val) results in a thermolabile form ofthe enzyme, hyperhomocysteinaemia and thrombophilia.The allele causing the thermolabile enzyme appears to beremarkably prevalent in Europeans, at around 30% (Frosst etal, 1995). Furthermore, the MTHFR defect has been shownto interact with factor V Leiden to further increase the risk ofvenous thrombosis (Mandel et al, 1996).

PATIENTS AND METHODS

The present study investigated 48 consecutive patientsentered into the U.K. Medical Research Council ATRA trialfrom whom material was available for molecular analysis. Inall cases the clinical diagnosis of APL was confirmed bymolecular evidence for the PML-RARa rearrangement byRT-PCR. The patients were treated with ATRA andchemotherapy according to protocol (Grimwade et al, 1996).

DNA was prepared using standard methods. A 220 bp

British Journal of Haematology, 1997, 96, 490–492

490 q 1997 Blackwell Science Ltd

Correspondence: Dr D. C. Rees, MRC Molecular Haematology Unit,Institute of Molecular Medicine, John Radcliffe Hospital, Headington,Oxford OX3 9DU.

fragment of the factor V gene was amplified and the PCRproduct digested with MnlI. Factor V Leiden was detected bythe loss of a restriction site (Rosendaal et al, 1995). The677C→T mutation in the MTHFR gene was identified in asimilar manner, using PCR and HinfI digestion. Themutation produces a single restriction site in a 198 bpfragment (Frosst et al, 1995).

RESULTS

The allelic frequency of factor V Leiden was 2·1% in thissmall sample (Table I), broadly in keeping with previousstudies of Europeans (Rees et al, 1995). The allele frequencyof thermolabile MTHFR was 38·5%, with 21% homozygotes;this does not deviate significantly from Hardy-Weinbergequilibrium, but is higher than the reported prevalence ofhomozygotes in Europeans of 12–15%.

Two individuals were heterozygotes for both factor VLeiden and thermolabile MTHFR. Neither experienced anydocumented thromboses – indeed one died from a cerebralhaemorrhage during induction. The other patient under-went an allogeneic bone marrow transplant (BMT) in firstcomplete remission; it is of note that this latter case did notsuffer from veno-occlusive disease.

Ten patients were homozygous for thermolabile MTHFR,and again experienced no proven thromboses. Two under-went autologous bone marrow transplants. The onlythromboses occurred in those with the lower genetic risk ofthrombosis, and in all cases were non-fatal. Thrombosis waspart of the clinical presentation in two patients: a calf deepvein thrombosis, and femoral artery occlusion. The othertwo thromboses involved the subclavian vein and wererelated to the presence of Hickman lines. All diagnoses wereconfirmed radiologically.

DISCUSSION

Factor V Leiden and thermolabile MTHFR have emergedrecently as two common genetic risk factors for venous andarterial thrombosis respectively in Europeans. Moreover,they have been shown to interact to further increase thelikelihood of venous thrombosis (Mandel et al, 1996). Itseems reasonable to suppose that these factors may also beimportant in diseases associated with secondary thrombosis,including neoplastic diseases such as APL. APL is

particularly pertinent, in that the key reason for failure toreach remission is early mortality secondary to haemor-rhage or thrombosis. An ability to identify those at particularrisk of clotting problems, using these DNA markers, maytherefore enable specific strategies to be adopted forindividual patients. A particular advantage of such genetictests is that, unlike clotting assays, DNA analysis is notconfounded by the acquired coagulopathy of APL. It ispossible that carrying these prothrombotic traits might alsoreduce the haemorrhagic tendency.

In the present study neither factor V Leiden northermolabile MTHFR were predictive of thrombosis. This iswell illustrated by the individual who underwent anallogeneic BMT without complications despite carryingboth factor V Leiden and thermolabile MTHFR. Conversely,three of the four thromboses occurred amongst patients withno identified genetic risk factors, the fourth in a thermolabileMTHFR heterozygote. Also, there is little to support the beliefthat either of these factors might protect against bleeding(two thermolabile MTHFR homozygotes and one factor VLeiden heterozygote suffered fatal haemorrhage).

Although this study was unable to show that these DNAmarkers were of predictive value for thrombotic (orhaemorrhagic) complications in APL, its power was limitedby the relatively small numbers. In particular, only twofactor V Leiden heterozygotes were found, making it difficultto draw any firm conclusions regarding its significance inAPL. Nevertheless, interpretation of DNA results in con-junction with more conventional plasma coagulation assaysmay offer additional information leading to more appropriatemanagement of individual patients. Clearly, the exactimportance of factor V Leiden and thermolabile MTHFR inAPL, and other malignant conditions, needs to be assessed inlarger numbers of patients, and this may best be achieved byincluding these analyses in on-going and future clinical trials.

ACKNOWLEDGMENTS

We are grateful to all the clinicians who entered patients intothe MRC ATRA trial and forwarded material for molecularanalysis, and data managers who supplied clinical informa-tion. D.R. and D.G. are supported by MRC clinical trainingfellowships. S.L. and DNA/RNAbanking facilities at Uni-versity College Hospital, London, are supported by the KayKendall Leukaemia Fund. We thank Professors John Clegg,

491Short Report

q 1997 Blackwell Science Ltd, British Journal of Haematology 96: 490–492

Table I. Distribution of genotypes and associated clotting problems. For factor V, G at 1691 is associated withthrombophilia (factor V Leiden). For MTHFR, T at 677 predicts a thermolabile enzyme and thrombophilia.

Thrombosis at Thrombosis duringFactor V, 1691 MTHFR, 677 No. presentation treatment Fatal haemorrhage

AG CT 2 0 0 1AA TT 10 0 0 2AA CT 15 0 1 1AA CC 21 2 1 1

Total 48 2 2 5

Sir David Weatherall, David Linch and Ellen Solomon foradvice and support.

REFERENCES

Bertina, R.M., Koeleman, B.P.C., Koster, T., Rosendaal, F.R., Dirven,R.J., De Ronde, H., Van Der Velden, P.A. & Reitsma, P.H. (1994)Mutation in blood coagulation factor V associated with resistanceto activated protein C. Nature, 369, 64–67.

Frosst, P., Blom, H.J., Milos, R., Goyette, P., Sheppard, C.A., Mathews,R.G., Boers, G.J.H., den Heijer, M., Kluijtmans, L.A.J., Van denHeuvel, L.P. & Rozen, R. (1995) A candidate genetic risk factor forvascular disease: a common mutation in methylene tetrahydro-folate reductase: isolation of cDNA, mapping and mutationidentification. Nature Genetics, 10, 111–113.

Grimwade, D., Howe, K., Langabeer, S., Burnett, A., Goldstone, A. &Solomon, E. (1996) Minimal residual disease detection in acutepromyelocytic leukemia by reverse-transcriptase PCR: evaluationof PML-RARa and RARa-PML assessment in patients whoultimately relapse. Leukemia, 10, 61–66.

Grimwade, D. & Solomon, E. (1996) Characterisation of the PML/RARa rearrangement associated with t(15;17) acute promyelo-cytic leukaemia. Current Topics in Microbiology and Immunology,220, in press.

Mandel, H., Brener, B., Berant, M., Rosenberg, N., Lanir, N., Jakobs,C., Fowler, B. and Seligsohn, U. (1996) Coexistence of hereditaryhomocystinuria and factor V Leiden: effect on thrombosis. NewEngland Journal of Medicine, 334, 763–768.

Rees, D.C., Cox, M.C. & Clegg, J.B. (1995) World distribution of factorV Leiden. Lancet, 346, 1133–1134.

Rodeghiero, F. & Castaman, G. (1994) The pathophysiology andtreatment of hemorrhagic syndrome of acute promyelocyticleukemia. Leukemia, 8, (Suppl. 2), s20–s26.

Rosendaal, F.R., Koster, T., Vandenbroucke, J.P. & Reitsma, P.H.(1995) High risk of thrombosis in patients homozygous forfactor V Leiden (activated protein C resistance). Blood, 85,1504–1508.

Tallman, M.S. & Kwaan, H.C. (1992) Reassessing the hemostaticdisorder associated with acute promyelocytic leukemia. Blood, 79,543–553.

492 Short Report

q 1997 Blackwell Science Ltd, British Journal of Haematology 96: 490–492