Leukemia Research Vol. 16, No. 10, pp. 1041-1048, 1992.

0145-2126/92 $5.00+ .00 © 1992 Pergamon Press Ltd

Printed in Great Britain.

B L E E D I N G C O M P L I C A T I O N S A N D C O A G U L O P A T H Y IN A C U T E LEUKAEMIA E. TORNEBOHM, M. BLOMB,~.CK,* D. LOCKNER, N. EGBERG* and C. PAUL Division of Clinical Haematology and Oncology, Department of Internal Medicine, Huddinge University Hospital, Huddinge, Sweden and *Department of Clinical Chemistry and Blood Coagulation, Karolinska Hospital, Stockholm, Sweden

(Received 10 December 1991. Revision accepted 10 February 1992) Abstract--Factors and inhibitors of coagulation and fibrinolysis were investigated on admission in 57 patients with acute leukaemia and they were correlated to the occurrence of haemorrhage. Coagulation disturbances were found in 98%. Seventeen of the patients with haemorrhagic symptoms had major bleeding. Severe thrombocytopenia (30 g/l) or caused haemorrhages in vital organs or an immediate need for transfusion because of excessive bleedings from mucosal membranes which did not respond to local coagulant therapy. All other bleeding episodes were classified as minor. In 7 patients the physician in charge suspected DIC. His suspicions were based on the presence of bleeding (5 patients) and septicaemia with organ failure (2 patients) together with pathological routine coagulation analyses (FDPs, prothrombin complex, activated partial thromboplastin time). Septicaemia was defined as a positive blood culture or fever ~>38.5°C for/>3 days. The study was approved by the Ethics Committee at Huddinge University Hospital. Blood sampling

Before transfusions and antileukaemic treatment were given, blood was drawn and immediately analysed for routine blood chemistry and haematological variables. At the same time citrated blood for the preparation of plateletpoor plasma was collected and stored at -70°C pending special coagulation analyses. Blood chemistry and haematological variables

The following analyses were performed: haemoglobin, leukocyte number with differential count, platelet count, aminotransferase ( A L A T ) and albumin. Blood coagulation variables

On admission plasma was taken for analyses of activated partial thromboplastin time ( A P T T ) , using reagents from Ortho Diagnostic Systems (Raritan, New Jersey, U.S.A.) Two methods were used to analyse the prothrombin complex; either with kits from Stago (Asni6res, France,

n = 42) or with the reagent Thrombotest from Nycomed (Oslo, Norway, n = 9). This was necessary because of a change in the routine methods used in the hospital. The following analyses were performed on frozen plasma samples employing chromogenic peptide substrates: Factor XII (F XII) according to the method of Egberg and Overmark-Berggren, Factor VII (F VII) activated by tissue thromboplastin using the method by Seligsohn et al. and Factor VIII (F VIII) by the method of Rosdn [20-22]. Factor V (F V) was determined by the clotting method of Kappeler et al. [23], von Willebrand Factor antigen ( v W F : A g ) by electroimmunoassay, as described by Holmberg and Nilsson [24], and the ristocetin cofactor activity of the von Willebrand Factor (vWF:rist. cof.) by the method of Barry et al. [25]. The ratios (RAs) for vWF: A g / F VIII and vWF: A f / v W F : rist. cof. were calculated for each patient. In 7 patients the levels of thrombin-antithrombin complex (TAT) were determined using the method described by Pelzer et al. [26]. Fibrinogen was analysed by the method of Vermylen et al. [27]. Coagulation inhibitors

Antithrombin (antithrombin or A T ) was measured as heparin cofactor using the chromogenic substrate S-2238 [28]. For the determination of protein C, plasma was activated with Protac and the activity was measured using S-2366 [29]. In the patients with very low levels of F XII (n = 4), the concentration of complement C~ esterase inhibitor was determined by Winman and Nilsson's method [301. Fibrinolytic variables

Fibrin/fibrinogen degradation products (FDPs) using Thrombo-Wellco reagents were analysed immediately in only 10 patients. A chromogenic peptide substrate S-2251 was used to analyse antiplasmin (antiplasmin or AP), [31]. Haemostatic score

To evaluate the presumptive roles of different coagulation and fibrinolytic variables a haemostatic score was created. The pathological results of the following analyses were regarded as possible causes of an increased bleeding diathesis: a high ratio between v W F : A g / F VIII and reduced levels of: antiplasmin, protein C, F VII, vWF : rist. cof., F V. fibrinogen and F XII. The haemostatic score ranged between 0.0 and 8.0 points.

Coagulopathy in leukaemia *3 3 I

reference value

I

"

3

r~ii

i~

I

I

2.0

3.0

n

• [

i I

4.0 5.0 ratio vWF:Ag/F VIII

FIG. 1. The ratio between vWF:Ag/F VIII in patients with acute leukaemia at diagnosis (n = 52). Each square represents a patient. The filled squares are patients with major bleedings. 3, acute promyelocytic leukaemia; *, acute lymphoblastic leukaemia. STATISTICAL METHODS Median values were calculated for all of the results. Differences between the subtypes of leukaemia were estimated using the Kruskal-Wallis test. The Mann-Whitney test was used to compare the group with bleeding complications with the group which did not. The same statistical comparisons were performed with the results from the patients with septicaemia, DIC and those without. Correlations were calculated using Spearman's coefficient (Ks). RESULTS

Blood coagulation factors, inhibitors and fibrinolytic variables Disturbances in the coagulation system were a common finding among all patients regardless of FAB group (56 of 57 patients). The median plasma concentrations of F VIII, v W F : A g and vWF:rist. cof. were 1.92, 2.36 and 1.78 U/ml, respectively (ref. range 0.50-1.50U/ml). In the majority of the patients (58%) all three variables were above the limit range. The correlations between F VIII and v W F : A g or vWF:rist, col. were 0.49 and 0.56, respectively (p < 0.01 and p < 0.001) and that between vWF: Ag and vWF: rist. cof. was as high as 0.74 (p < 0.001). The ratio between vWF : A g / F VIII was elevated in 30% of the patients (>1.6, Fig. 1). Moreover, the ratio between vWF: Ag and vWF :rist. col. showed a similar change, although the levels were not so high. In approximately one fourth of the patients the plasma levels of antithrombin, protein C and antiplasmin were below the reference values (Figs. 2-4). Low plasma levels of F .VII and F V were found in 15% and 7% of the patients.

Characteristics of different leukaemia patients groups The incidence of low antithrombin values was high in the patients classified as M2 (Fig. lb). The four M3 patients presented with the lowest platelet counts (26 x 109/1, Table 1) seen in any of the patients. The ratio v W F : A g / F VIII was high (2.0, Fig. 1). All of them had normal antithrom-

1043

bin values. They also had reduced levels of protein C (0.61 U/ml, Fig. 3), antiplasmin (0.58 U/ml, Fig. 4) and fibrinogen (1.8 g/l, Fig. 5). On the whole, they had significantly higher mean haemostatic scores (3.3) than the other FAB classes (1.0, p < 0.01). Three of the M3 patients suffered major bleedings on admission (Nos 5-7, Table 2). In the M4 group the concentrations of antithrombin were low (Fig. 2). Reduced levels of antiplasmin were also common (56% had levels < 0.75 U/ml, Fig. 4). Two patients with A L L presented with the highest ratio of v W F : A g / F VIII in all of the patients (Fig. 1) and very low levels of fibrinogen (Fig. 5). One also had a low antiplasmin value and both had major bleedings on admission (patients Nos 16 and 17, Table 2).

Haemorrhagic complications Minor bleedings, such as mucosal, petechial and skin haematomas were present in 24 (42%) of the patients. Of these, 17 patients also suffered major bleedings according to our defnition (Table 2). Mucosal haemorrhage (e.g. gingival, epistaxis) was the commonest major bleeding (9 of 17 patients). Of these 9 patients, 5 had epistaxis, 2 of whom needed massive transfusions. Other major bleedings came from the rectum, the genito-urinary tract and from the uterus. The highest incidence of serious haemorrhage occurred in 3 of the 4 M3 patients. Two had intracranial bleedings, one of which was fatal ( patient No. 6, Table 2). Otherwise the minor and major bleeding complications were evenly distributed. The patients with major bleedings had significantly lower platelet counts than the 'non-bleeders', 26 and 81 x 109/1 respectively (p < 0.001, Table 3). None had platelet c o u n t s < 10 x 109/1 (our limit for prophylactic platelet transfusion) and only 9 patients (16% had platelet counts < 20 x 109/1. Thrombocytopenia as a sole cause of major bleeding was found in one patient (patient No. 1, Table 2). Platelet dysfunction may have been the reason in two others (patient Nos 2 and 15, Table 2). The peripheral leukaemic cell count tended to be higher in patients with major bleedings, 33.8 x 109/1, than in those without, 5.9 x 109/1 (ns). The patients with serious haemorrhages had significantly lower levels of the prothrombin complex, protein C and fibrinogen than did those without major bleedings (Figs. 3, 4, Table 3). FDP was found to be elevated in all of the 10 patients in which it was measured. Five of them had haemorrhagic complications. No difference between the groups was seen for APTT. The mean haemostatic score was significantly higher in the group with major bleedings

E. TORNEBOHMet al.

1044

T A B L E 2. TYPE AND LOCALIZATION OF BLEEDING, PLATELET COUNT (PLT), TtlERAPY. THE MOST PRONOUNCED HAE~ MOSTATIC DISTURBANCES (HEMO. DIST.), HAEMOSTATIC SCORE (SCORE) AND OUTCOME ( C R - COMPLETE REMISSION) IN 17 PATIENTS WITH MAJOR BLEEDINGS

No.

Bleeding localization

FAB

1 2 3 4 5

M M M M M

6

1 1 2 2 3

PLT 109/1

Therapy*

Muscle Epistaxis Rectal Menorrhagia, skin Gingival, skin

11 24 52 23 26

M 3

Intracranial

10

7

M 3

Intracranial

29

8 9

M 4 M 4

Epistaxis, skin Epistaxis, skin, haemoptysis

24 50

10 11 12

M 5 M5 M 5

Mucosal, skin Skin Gingival

11 14 45

2 Ery, 6 Pit 2 Ery 2 Ery, 6 Pit 4 Ery, 6 Plt 4 Ery, 2 Plt, 1200FFP 4 Ery, t2 Pit 1200FFP, 2 AHF 4 Ery, 12 Plt 2400 FFP 4 Ery, 6 Pit 12 Plt cauterization 6 Plt 2Ery 2 Ery

13

M 5

Muscle

22

2 Ery

14

AUL

Epistaxis

26

15 16

ALL ALL

Gingival Muscle, haematuria

45 30

6 Ery, 18 Plt, 600 FFP, cauterization 6 Plt 4 Ery, 18 Pit 3000 FFP

17

ALL

Epistaxis, skin

26

8 Ery, 24 Plt, 3600 FFP

Hemo. $ dist.

Scores

CR

--F XII $ ratio t AP + fib $ C { FVI1 AP + fib ~ C + FV {

0 0 1 1 4

Yes Yes Yes Yes No

4

No

A P { fib $ ratio ~' AT $ C $ AP $ A T $ F V AP ~ A T $ AP { F V I I + C ~ ratio I' F XII J, A T ¢ F VII { C { ratio ]' A T { AT $ C {



Yes

1 2

Yes Yes

1 2 3

No No No

3

Yes

1

No

-AP { C { ratio 1" fib { FV{ C { ratio I" fib {

0 5

Yes Yes

3

Yes

* Therapy: Ery = units of erythrocyte concentrate (400-450 ml), P l t = units of random donor or single donor platelet concentrate (450-480 x 109/U), FFP = ml fresh frozen plasma, A H F = 600 U of F VIII concentrate (I-O also containing fibrinogen, vWF, fibronectin and F X I I I - - n o longer available). + C = protein C, A T = antithrombin, AP = antiplasmin, ratio = v W F : A g / F VIII, fib = fibrinogen. $ Score (range 0-8 points). § Only 4 of the 8 analyses could be performed.

TABLE 3.

M E D I A N VALUES FOR HAEMOSTATIC VARIABLES OF INTEREST IN PATIENTS WITH MAJOR BLEEDINGS (MB) AND TtIOSE WITHOUT ( n o MB). PROTHROMBIN COMPLEX (PROTR. COMP.)

No. Pit Protr. comp. Fibrinogen Antiplasmin Protein C Score *p +p ¢p §p

Bleeding complications and coagulopathy in acute leukaemia.

Factors and inhibitors of coagulation and fibrinolysis were investigated on admission in 57 patients with acute leukaemia and they were correlated to ...
660KB Sizes 0 Downloads 0 Views