Disseminated intravascular coagulation inacute leukemia: clinical and laboratoryfeatures at presentation

Disseminated intravascular coagulation (DIC) is acomplex syndrome characterized by systemic intra-vascular activation of the coagulation system (1–5).It leads not only to severe thrombosis, but also tosevere bleeding as a result of excessive consumptionof platelet and plasma coagulation factors. Labor-atory findings show abnormalities of theblood-clotting system, including thrombocyto-penia, hypofibrinogenemia, prolonged prothrom-bin time (PT), and elevated fibrin-related markers.DIC always occurs secondary to underlying dis-eases, such as infection, trauma, and malignancies.Especially, in cases with acute leukemia, life-threatening bleeding becomes a matter of concern

(3–7), as it is frequently accompanied by thromb-ocytopenia because of bone marrow involvement ofleukemic cells, and the start of chemotherapyoccasionally aggravates the pre-existing coagula-tion abnormalities (5, 8, 9). Although properdiagnosis of DIC is important for early initiationof supportive measures, lack of a specific diagnostictest has made this difficult, so that DIC is usuallydiagnosed clinically. For these reasons, severalgroups have constructed scoring systems basedupon laboratory tests and clinical manifestations(10–12); but the validity of these systems remains tobe confirmed, particularly for leukemia-associatedDIC. For this study, we evaluated 125 patients with

Yanada M, Matsushita T, Suzuki M, Kiyoi H, Yamamoto K, KinoshitaT, Kojima T, Saito H, Naoe T. Disseminated intravascular coagulationin acute leukemia: clinical and laboratory features at presentation.

Abstract: Background: Although there are two major scoring systemsfor the clinical diagnosis of disseminated intravascular coagulation(DIC), the validity of these systems for leukemia-associated DICremains to be confirmed. Methods: By analyzing 125 newly diagnosedacute leukemia patients, we investigated clinical and laboratory featuresof leukemia-associated DIC, and determined the validity of the twoestablished criteria. Results: A total of 36 patients (29%) werediagnosed with DIC according to expert opinion, a method regarded asthe de facto gold standard. Leukemia-associated DIC is characterized byrare manifestation of organ failure because of thrombosis and norelevance of the platelet count for the diagnosis. The results of receiveroperating characteristics analysis favored fibrin degradation product(FDP) rather than D-dimer as the fibrin-related marker test. Althoughprothrombin time, plasma fibrinogen, and serum FDP levels weresignificantly different for patients with and without DIC, multivariateanalysis identified FDP levels to be the only factor associated with DICdiagnosis. The cut-off level of 15 lg/mL for FDP was found to be themost effective to differentiate DIC from non-DIC, resulting in diagnosticsensitivity and specificity of 92% and 96%, respectively. The diagnosticresults for our patients produced with this FDP-based system were atleast comparable with or superior to those obtained with the twocurrently available scoring systems. Conclusions: Our findings suggestthat an FDP-based criterion may be applicable for the diagnosis ofleukemia-associated DIC. Although it appears to be simple andpracticable enough for clinical use, prospective validation of thiscriterion is needed.

Masamitsu Yanada1,2, TadashiMatsushita1, Momoko Suzuki1,Hitoshi Kiyoi2, Koji Yamamoto3,Tomohiro Kinoshita1, TetsuhitoKojima4, Hidehiko Saito5, TomokiNaoe1

1Departments of Hematology and Oncology; 2InfectiousDiseases; 3Transfusion Medicine, Nagoya UniversityGraduate School of Medicine; 4Department of MedicalTechnology, Nagoya University School of HealthSciences; 5National Hospital Organization NagoyaMedical Center, Nagoya, Japan

Key words: disseminated intravascular coagulation;diagnostic criteria; acute leukemia; fibrin degradationproduct

Correspondence: Masamitsu Yanada MD, Departmentof Hematology and Oncology, Nagoya UniversityGraduate School of Medicine, 65 Tsurumai-cho,Showa-ku, Nagoya, Aichi 466-8550, JapanTel: +81-52-744-2955Fax: +81-52-744-2801e-mail: myanada@med.nagoya-u.ac.jp.

Accepted for publication 15 June 2006

Eur J Haematol 2006: 77: 282–287doi:10.1111/j.1600-0609.2006.00711.xAll rights reserved

� 2006 The AuthorsJournal compilation � 2006 Blackwell Munksgaard

EUROPEAN JOURNAL OF HAEMATOLOGY

282

acute leukemia referred to our hospital between1990 and 2004, and investigated clinical and labor-atory features of leukemia-associated DIC. We alsodetermined the validity of the two establisheddiagnostic systems, i.e. the Japanese Ministry ofHealth and Welfare (JMHW) criteria (10), and theInternational Society of Thrombosis and Haemo-tasis (ISTH) criteria (11).

Patients and methods

Patients

Adult patients were eligible for inclusion in theanalysis if they met all of the following criteria: theywere diagnosed with acute leukemia on the basis ofthe World Health Organization classification (13–15); their first presentation was between 1990 and2004 at Nagoya University Hospital; and measure-ments of platelet count, PT, plasma level offibrinogen, and serum level of fibrin degradationproduct (FDP) were performed before the start ofchemotherapy or anticoagulant therapy. Laborat-ory evaluations including the coagulation tests wereroutinely performed daily or every other day untilthe coagulation disorder had been resolved. As thiswas not a prospective study, any decision concern-ing the specific treatments for coagulopathy wasultimately left to the discretion of the attendingphysicians. However, consensus was reached on apolicy of prompt initiation of antileukemic therapy,intensive transfusion support to maintain plateletcount above 30 to 50 · 109/L and plasma fibrin-ogen level above 1.0 g/L, as well as concurrent use ofanticoagulants in the presence of significant coag-ulopathy. This study was reviewed and approvedby the institutional review board.

Evaluation of DIC

On the assumption that no proper gold standardexists for the diagnosis of DIC, the decisionwhether DIC was present or absent in a givenpatient was made at the time of final analysis on thebasis of expert opinion according to the method byBakhtiari et al. (16). Results of serial laboratorytests (white blood cell count, percentage of blasts inperipheral blood, platelet count, PT, fibrinogen,FDP, D-dimer, thrombin–antithrombin complex[TAT], and plasmin–plasminogen inhibitor com-plex [PIC] levels, any of these that were available)as well as clinical information (occurrence ofbleeding, organ failure, fever, and schedules ofchemotherapy, transfusion, and anticoagulant ther-apy) were presented to two experts in hemostasis(TM and TK). For assessment, status after the startof chemotherapy and scores based on the JMHW

and ISTH criteria were not taken into considera-tion. After the experts had independently deter-mined whether a patient had DIC or not, the resultswere compared. If there was disagreement, the finaljudgment was left to a third expert (KY). Theexperts were not given any other information, suchas the patient’s name, the leukemia subtype, and soon.

Citrated plasma was drawn from the patients andimmediately analyzed unless TAT and PIC weremeasured by using plasma stored at )80�C. One-stage PT and fibrinogen were determined with thegeneral method using the Simplastin Excel S(bioMerieux, Marcy l’Etoile, France) and MDAFibriquick (bioMerieux). Serum was collected inthe presence of thrombin and aprotinin and FDPwas measured with the standard quantitative latexaggregation assay using an anti-fibrinogen anti-body (LPIA FDP; ASKA Pharmaceuticals, Tokyo,Japan). Plasma levels of D-dimer, TAT, and PICwere measured with the enzyme-linked immuno-sorbent assay or the latex aggregation assays usingan LPIA Ace D-D Dimmer (Mitsubishi KagakuIatron, Tokyo, Japan), TAT Test-F (InternationalReagents Corporation, Kobe, Japan), and PICTest-F (International Reagents Corporation). Foran undisclosed study period, FDP had beenreplaced with FDP-E, which was assayed with thequantitative latex aggregation assay (LPIA FDP-E;ASKA Pharmaceuticals). For the analysis used inthis study, FDP-E values were converted to FDPvalues with the following formula: FDP ¼ (FDP-E + 21)/22.3. This formula is based on a regres-sion curve obtained from a total of 50 patients withor without DIC performed at one institution (17).

DIC scores were also calculated with both theJMHW criteria applicable to patients with hema-tological malignancies (10) and the ISTH criteriafor overt DIC (11) (Table 1). For the scoring offibrin-related markers in the ISTH system, FDPvalues of <10, 10–40, and ‡40 lg/mL were con-sidered as indicating, respectively, no increase,moderate increase, and strong increase. Organfailure comprised myocardial infarction, shock,peripheral arterial obstruction, pulmonary infarc-tion, acute respiratory distress syndrome, ischemiccolitis, brain infarction, impaired consciousness,and grade 3 or higher elevation of creatinine,bilirubin, and glutamic pyruvic transaminaseaccording to the National Cancer Institute Com-mon Terminology Criteria for Adverse Eventsversion 3.0.

Statistical analysis

The distribution of baseline characteristics ofpatients with and without a diagnosis of DIC was

DIC in acute leukemia

283

compared by using Fisher’s exact test for categor-ical variables, and Mann–Whitney U-test for con-tinuous variables. A logistic regression analysis wasperformed to evaluate the factors predictive forDIC. Factors with P-values of <0.05 for univariateanalysis were included in the subsequent multi-variate analysis. The odds ratio was calculated inconjunction with a 95% confidence interval.Receiver operating characteristics (ROC) analysiswas also used to quantify the accuracy of diagnostictests, and areas under ROC curves (AUCs) werecompared between the two tests. Sensitivity,specificity, and positive and negative-predictivevalues were calculated with the aid of a 2 · 2 table,and differences between the two criteria werecompared with Fisher’s exact test. stat view 5.0

(SAS Institute Inc., Cary, NC, USA) and stata ver.8 software (Stata Corp LP, College Station, TX,USA) were used for all statistical analysis.

Results

Patients

Of the 151 patients diagnosed with acute leukemiaduring the study period, 26 were excluded becauseof insufficient laboratory data for full evaluation.Of the remaining 125 patients eligible for this study,36 (29%) were diagnosed with DIC at presentationaccording to the above-mentioned method ofreferral to two or three experts. The baseline

characteristics of patients with and without DICare shown in Table 2. Statistically significant dif-ferences were observed in the distribution of PTratio, fibrinogen and FDP levels, but no suchdifference was detected in the platelet count. Theproportions of patients with fever, liver dysfunc-tion, and renal dysfunction did not differ betweenthe two groups, while none of the patients hadorgan failure because of thrombosis. Table 3 showsthe prevalence of DIC by acute leukemia subtype.Not surprisingly, the incidence of DIC was 100%for acute promyelocytic leukemia, but otherwise noleukemia subtypes were associated with a remark-ably high DIC incidence. Early hemorrhagic deathoccurred in only one patient with acute monocyticleukemia who developed a cerebral hemorrhage onthe sixth day of admission.

Table 1. Criteria of two major scoring systems for DIC

JMHW criteria1 ISTH criteria2

Underlying disease Active: 1 RequiredInactive: 0

Thrombosis-related organ failure Present: 1 Not consideredAbsent: 0

Platelet count 650 · 109/L: 3 650 · 109/L: 250–80 · 109/L: 2 50–100 · 109/L: 1

80–120 · 109/L: 1 >100 · 109/L: 0>120 · 109/L: 0

Fibrin-related marker FDP level of Strong increase: 3P40 lg/mL: 3 Moderate increase: 220–40 lg/mL: 2 No increase: 010–20 lg/mL: 1

<10 lg/mL: 0PT PT ratio of Prolongation of

P1.67: 2 P6 s: 21.25–1.67: 1 3–6 s: 1

<1.25: 0 <3 s: 0Fibrinogen 61.0 g/L: 2 61.0 g/L: 1

1.0–1.5 g/L: 1 >1.0 g/L: 0>1.5 g/L: 0

Diagnosis of DIC Total score ‡4 Total score ‡5

DIC, disseminated intravascular coagulation; JMHW, Japanese Ministry of Healthand Welfare; ISTH, International Society of Thrombosis and Haemostasis; FDP, fibrindegradation product; PT, prothrombin time.1Criteria applicable to patients with hematological malignancies.2Criteria for overt DIC.

Table 2. Patient characteristic

Patients with DIC(n ¼ 36)

Patients without DIC(n ¼ 89) P-value

Age (years) 45.5 (16–83) 52 (16–79) 0.3994Sex (male/female) 24/12 62/27 0.8318Disease (AML/ALL) 29/7 71/18 0.9213WBC count (·109/L) 5.9 (0.7–351) 6.7 (0.7–814) 0.5818Hemoglobin (g/dL) 8.6 (3.7–15.0) 8.7 (3.7–16.1) 0.8300Platelet count (·109/L) 42 (8–165) 59 (3–1700) 0.1013PT ratio 1.20 (0.96–1.48) 1.13 (0.90–1.44) 0.0184Fibrinogen (g/L) 2.84 (0.71–8.47) 4.01 (1.49–7.99) 0.0140FDP (lg/mL) 47.7 (11.2–307) 5.0 (1.4–25.3) <0.0001

DIC, disseminated intravascular coagulation; AML, acute myeloid leukemia; ALL,acute lymphoblastic leukemia; WBC, white blood cell; PT, prothrombin time; FDP,fibrin degradation product.Values are presented as median (range) unless indicated.

Table 3. Prevalence of DIC by disease subtype

No. of patients (%)

AML with recurrent genetic abnormalities 15/29 (52)AML with t(8;21); (AML1/ETO) 1/13 (8)AML with inv(16) or t(16;16); (CBFb/MYH11) 0/1 (0)APL [AML with t(15;17); (PML/RARa) and variants] 13/13 (100)AML with 11q23 (MLL) abnormalities 1/2 (50)

AML with multilineage dysplasia 5/24 (21)AML and MDS, therapy-related 1/10 (10)AML not otherwise categorized 8/37 (22)

AML minimally differentiated 0/4 (0)AML without maturation 2/5 (40)AML with maturation 2/16 (13)Acute myelomonocytic leukemia 2/7 (29)Acute monoblastic and monocytic leukemia 2/4 (50)Acute megakaryoblastic leukemia 0/1 (0)

Precursor B-lymphoblastic leukemia/lymphoma 7/20 (35)ALL with t(9;22); (BCR/ABL) 2/5 (40)

Precursor T-lymphoblastic leukemia/lymphoma 0/5 (0)Total 36/125 (29)

DIC, disseminated intravascular coagulation; AML, acute myeloid leukemia; APL,acute promyelocytic leukemia; MDS, myelodysplastic syndrome; ALL, acute lym-phoblastic leukemia.

Yanada et al.

284

Factors associated with DIC diagnosis

Firstly, we assessed the individual relevance ofpotentially confounding factors for the diagnosis ofDIC. As shown in Table 4, multivariate analysisidentified FDP levels to be the only factor signifi-cantly associated with DIC development with astrong correlation (P < 0.0001). In fact, allpatients with FDP levels of 30 lg/mL or higherwere diagnosed as DIC, and all those with FDPlevels below 10 lg/mL were diagnosed as non-DIC.The significant relationship with the PT ratioobserved in the univariate analysis was not main-tained after adjustment for FDP levels. A similarresult was obtained when fibrinogen and FDPlevels were combined in a single model (data notshown). Fig. 1 compares ROC curves for FDP andD-dimer tests of 92 patients who had both testsperformed. Although both showed high discrimin-ation ability (AUC: 0.980 and 0.934, respectively),there was a trend for a larger AUC for FDP (P ¼0.0536). AUCs for other diagnostic tests are sum-marized in Table 5.Secondly, to determine the optimum cut-off

point for FDP levels, the probabilities of patientsdiagnosed with DIC were examined after they were

grouped by FDP level in increments of 5 lg/mL.The cut-off point of 15 lg/mL was found to be themost suitable for differentiating DIC from non-DIC, with a diagnostic sensitivity and specificity of92% and 96%, respectively.

Validation of diagnostic criteria for DIC

We validated the JMHW and ISTH scoring systemsto determine how many of the 125 patients wouldbe diagnosed with DIC. The two criteria were alsocompared with the FDP-based criterion with a cut-off point of 15 lg/mL. The results are summarizedin Table 6 and Fig. 2. The FDP-based criterionshowed higher sensitivity and negative-predictivevalue than the ISTH criteria (P ¼ 0.0002 and0.0023, respectively), while specificity and posit-ive-predictive value were not impaired. The diag-nostic results produced with the FDP-basedcriterion were comparable with those obtained withthe JMHW scoring system.

Discussion

The aim of this study was to evaluate clinical andlaboratory features of leukemia-associated DIC.Our results suggested that leukemia-associatedDIC differed from DIC arising from otherunderlying diseases in that organ failure becauseof thrombosis was rare, and that platelet countplayed only a minor part if any, in the diagnosis.Of the other parameters included in the JMHWand ISTH DIC scoring systems, PT and fibrin-ogen levels initially appeared to differentiate ourpatients with DIC from those without DIC, butsuch correlations were not confirmed in multi-variate analysis. This can be explained by the factthat severe PT prolongation and hypofibrinigene-mia were almost exclusively seen in cases with theelevated FDP levels. As a fibrin-related marker,FDP is the most prevalent in Japan, but anotherwidely-used marker, D-dimer, was also studied

Table 4. Predictive factors for DIC diagnosis

Univariate analysis Multivariate analysis

P-value P-value OR (95% CI)

FDP <0.0001 <0.0001 1.33 (1.19–1.48)PT ratio 0.0108 0.1788Fibrinogen 0.0519Platelet count 0.3033

DIC, disseminated intravascular coagulation; OR: odds ratio; 95% CI: 95% confid-ence interval; FDP, fibrin degradation products; PT, prothrombin time.

Fig. 1. Receiver operating characteristics (ROC) curves forfibrin degradation product (FDP) and D-dimer tests (n ¼92). Although both showed high discrimination ability,there was a trend for a larger area under the curve for FDP(P ¼ 0.0536).

Table 5. Area under the ROC curve for the diagnosis of DIC

No. of patients AUC 95% CI P-value1

FDP 125 0.980 0.962–0.998 –D-dimer 92 0.934 0.883–0.986 0.0536TAT 50 0.950 0.898–1.000 0.2510PIC 49 0.894 0.798–0.990 0.0723Fibrinogen 125 0.641 0.512–0.769 <0.0001Platelet count 125 0.594 0.491–0.697 <0.0001PT ratio 125 0.635 0.519–0.751 <0.0001

ROC, receiver operating characteristics; DIC, disseminated intravascular coagula-tion; AUC, area under the curve; FDP, fibrin degradation product; TAT, thrombin–antithrombin complex; PIC, plasmin–plasminogen inhibitor complex; PT, prothrombintime.1Compared with the AUC for FDP.

DIC in acute leukemia

285

for the diagnostic significance. The results ofROC analysis favored FDP rather than D-dimer,although the AUC for D-dimer was 0.934 (95%CI: 0.883–0.986), indicating its usefulness indiscriminating the DIC from non-DIC.

Diagnostic criteria become a clinically useful toolfor syndromes such as DIC, which are difficult todiagnose directly. There are two major scoringsystems for the clinical diagnosis of DIC, i.e. theJMHW criteria and the ISTH criteria. Theirefficacy has been investigated for patients in criticalcare settings (16, 18–20). Several groups haveassessed the ISTH criteria retrospectively (18, 19)and prospectively (16) by studying the patientsadmitted to the intensive care unit (ICU), andfound that the fibrinogen score had little effect onthe diagnosis of DIC. Gando et al. (20) prospec-tively studied 273 ICU patients, and validated thecriteria of ISTH, JMHW as well as their own.Again, no predictive significance could be estab-lished for fibrinogen levels, which led them toexclude it from their revised criteria. On the otherhand, such studies have not been conducted forleukemia-associated DIC, although it should be

noted that the clinical presentations of DIC mayvary widely depending on the underlying disease.This situation prompted us to conduct this researchon the assumption that there was no proper goldstandard for the diagnosis of leukemia-associatedDIC. Instead, the presence or absence of DIC inindividual patients was determined on the basis ofexpert opinion according to the method describedpreviously (16). We decided to employ this methodbecause, at present, it seems to be the mostappropriate method in the absence of a goldstandard.

By using the data of 125 patients with acuteleukemia, we evaluated the JMHW and ISTHcriteria in comparison with the above-mentionedFDP-based criterion. As shown in Table 6, thediagnostic results for our patients produced withthe FDP-based system were at least comparablewith those obtained with the JMHW scoringsystem, and superior to those generated by theISTH DIC scoring system in terms of sensitivityand negative-predictive value. Because the weight-ing of the FDP score is relatively high in theJMHW criteria, the JMHW and FDP-based cri-teria may produce similar diagnostic results,although the simpler and more practicable featuresof the latter would render it more suitable forclinical use. When we interpret these results, itshould be remembered that this study has severallimitations, including the small sample size and itsretrospective nature. As this is a single center study,the sample size is inevitably limited, so that thereliability of the FDP-based criterion should betested for a larger number of patients, preferably ina multi-institutional study and in a prospectivefashion.

Despite such limitations, our findings suggestthat an FDP-based simple criterion may beapplicable for the diagnosis of leukemia-associ-ated DIC. A common diagnostic system isessential for the standardization of clinical prac-tices and beneficial when designing clinical stud-ies. If confirmed, this criterion may be quiteuseful and widely applicable to both clinicalstudies and clinical practice.

Table 6. Characteristics of criteria used for threescoring systems

Sensitivity SpecificityPositive-predictive

valueNegative-predictive

value

JMHW criteria 78% (64–92%)* 91% (85–97%) 80% (67–93%) 90% (84–96%)ISTH criteria 50% (34–66%)*,** 99% (97–100%) 95% (85–100%) 80% (72–88%)***FDP-based

criterion92% (83–100%)** 96% (92–100%) 89% (79–99%) 97% (93–100%)***

JMHW, Japanese Ministry of Health and Welfare; ISTH, International Society of Thrombosis and Haemostasis.95% confidence intervals are shown in parentheses.*P ¼ 0.0263.**P ¼ 0.0002.***P ¼ 0.0023.

Fig. 2. Distribution of patients with and without dissem-inated intravascular coagulation (DIC) based on the threediagnostic criteria. Numbers within the circles representpatients diagnosed with DIC according to each set ofcriteria.

Yanada et al.

286

References

1. Levi M, Ten Cate H. Disseminated intravascular coagu-lation. N Engl J Med 1999;341:586–592.

2. Levi M. Current understanding of disseminated intravas-cular coagulation. Br J Haematol 2004;124:567–576.

3. Tallman MS, Kwaan HC. Reassessing the hemostaticdisorder associated with acute promyelocytic leukemia.Blood 1992;79:543–553.

4. Tallman MS. The thrombophilic state in acute promyel-ocytic leukemia. Semin Thromb Hemost 1999;25:209–215.

5. Barbui T, Falanga A. Disseminated intravascularcoagulation in acute leukemia. Semin Thromb Hemost2001;27:593–604.

6. Rodeghiero F, Avvisati G, Castaman G, Barbui T,Mandelli F. Early deaths and anti-hemorrhagic treat-ments in acute promyelocytic leukemia. A GIMEMA ret-rospective study in 268 consecutive patients. Blood1990;75:2112–2117.

7. Di Bona E, Avvisati G, Castaman G, Luce Vegna M, De

Sanctis V, Rodeghiero F, Mandelli F. Early haemor-rhagic morbidity and mortality during remission inductionwith or without all-trans retinoic acid in acute promyelo-cytic leukaemia. Br J Haematol 2000;108:689–695.

8. Sarris AH, Kempin S, Berman E, et al. High incidence ofdisseminated intravascular coagulation during remissioninduction of adult patients with acute lymphoblastic leu-kemia. Blood 1992;79:1305–1310.

9. Chojnowski K, Wawrzyniak E, Trelinski J, Nie-

wiarowska J, Cierniewski C. Assessment of coagulationdisorders in patients with acute leukemia before and aftercytostatic treatment. Leuk Lymphoma 1999;36:77–84.

10. Kobayashi N, Maekawa T, Takada M, Tanaka H,Gonmori H. Criteria for diagnosis of DIC based on theanalysis of clinical and laboratory findings in 345 DICpatients collected by the Research Committee on DIC inJapan. Bibl Haematol 1983:265–275.

11. Taylor FB, Jr., Toh CH, Hoots WK, Wada H, Levi M.

Towards definition, clinical and laboratory criteria, and ascoring system for disseminated intravascular coagulation.Thromb Haemost 2001;86:1327–1330.

12. Wada H, Gabazza EC, Asakura H, Koike K, Okamoto

K, Maruyama I, Shiku H, Nobori T. Comparison ofdiagnostic criteria for disseminated intravascular coagula-tion (DIC): diagnostic criteria of the International Society

of Thrombosis and Hemostasis and of the Japanese Min-istry of Health and Welfare for overt DIC. Am J Hematol2003;74:17–22.

13. Brunning R, Matutes E, Flandrin G, Vardiman J.

Acute myeloid leukaemia: introduction. In: Jaffe ES,Harris NL, Stein H, Vardiman JW, eds. World HealthOrganization Classification of Tumours. Pathology andGenetics: Tumours of Haematopoietic and Lymphoid Tis-sues. Lyon: IARC Press, 2001:77–80.

14. Brunning R, Borowiz M, Matutes E, Head D, Flandrin

G, Swerdlow S, Bennett J. Precursor B lymphoblasticleukaemia/lymphoblastic lymphoma. In: Jaffe ES, Harris

NL, Stein H, Vardiman JW, eds. World Health Organ-ization Classification of Tumours. Pathology and Genetics:Tumours of Haematopoietic and Lymphoid Tissues. Lyon:IARC Press, 2001:111–114.

15. Brunning R, Borowiz M, Matutes E, Head D, Flandrin

G, Swerdlow S, Bennett J. Precursor T lymphoblasticleukaemia/lymphoblastic lymphoma. In: Jaffe ES, Harris

NL, Stein H, Vardiman JW, eds. World Health Organ-ization Classification of Tumours. Pathology and Genetics:Tumours of Haematopoietic and Lymphoid Tissues. Lyon:IARC Press, 2001:115–117.

16. Bakhtiari K, Meijers JC, de Jonge E, Levi M. Pros-pective validation of the International Society of Throm-bosis and Haemostasis scoring system for disseminatedintravascular coagulation. Crit Care Med 2004;32:2416–2421.

17. Ohtomo M, Nishikata J, Emi Y, et al. Basic investigationfor measurements of FDP-E and whole FDP fractions byusing LPIA-100 (in Japanese). Igakutoyakugaku1988;19:1229–1236.

18. Sivula M, Tallgren M, Pettila V. Modified score fordisseminated intravascular coagulation in the critically ill.Intensive Care Med 2005;31:1209–1214.

19. Angstwurm MW, Dempfle CE, Spannagl M. New dis-seminated intravascular coagulation score: a useful tool topredict mortality in comparison with Acute Physiology andChronic Health Evaluation II and Logistic Organ Dys-function scores. Crit Care Med 2006;34:314–320.

20. Gando S, Iba T, Eguchi Y, et al. A multicenter, pros-pective validation of disseminated intravascular coagula-tion diagnostic criteria for critically ill patients: comparingcurrent criteria. Crit Care Med 2006;34:625–631.

DIC in acute leukemia

287