Studies on the components of the contact phase system in patients with acute nonlymphoblastic leukemia

The objective of the present study was to evaluate factors of the plasma kallikrein system in patients with acute nonlymphoblastic leukemia (ANLL), and compare the results to a normal control group. A prospective study was performed in the Tertiary Health Care Institution, Hemocentro, Campinas State University, Campinas, São Paulo, Brazil. Thirty-five patients, diagnosed as ANLL between 1988 and 1991, were considered for participation. Eleven patients were not elegible, according to the exclusion criteria: infection/ septicemia, previous treatment or blood transfusion. The study was performed with 24 ANLL patients, average age 34 years (16-69 years), 14 men and 10 women. Nineteen healthy volunteers, workers from the Hematology Center, average age 32 years (21-59 years), 11 men and 8 women, were .the control group. Plasmatic prekallikrein, C1-inhibitor, alpha 2-macroglobulin, activated partial thromboplastin time, prothrombin time, factor XII, factor XI, factor V and prealbumin were measured. Plasmatic prekallikrein (p=O.02) and prealbumin (p=O.03)were significantly decreased, and prothrombin time increased (p=O.003) in the patient group when compared to the control. Significantcorrelation (r=O.49,criticai value=O.43,p<O.05)between prekállikrein and prealbumin, and between prothrombin time and factor V (r=O.54, criticai value=O.44, p<O.05) was demonstrated in the patient group. No correlation was found between parameters analysed and circulant blast count or leukemia subgroups. Statistical analysis was performed by theWilcoxon tes1.Correlation between the parameters was also verified. These results suggest activation of the contact system or impaired Iiver synthesis in patients with ANLL, and could contribute to disease complications.


INTRODUCTION
F actorXII, plasmatic prekallikrein (PK) and HMWkininogen play an essential role in the initiation of imp9rtant biological reactions, including intrinsic coagulation pathway.Recently,-tecidual kallikrein and prekallikrein-kininogen complexes have been identified in neutrophils and myeloid precursors.I Studies "in vitro" demonstrate that granulocyiic elastase is capable of cleaving HMW-kininogen, PK and factors XII and XI.In addition, it can cleave Cl-inhibitor to an inactive form.I

Address for correspondence:
Joyce M. A nnichino-Bizza cchi, Hemocentro/UNICAMP, CP 6198 Campinas/SP -Brasil -CEP 13081-970 Disturbances of hemostasis are still a major problem in patients with acute leukemia, and activation of the contact phase or proteolysis of contact system components, if present, could contribute to these alterations.
Although a relationship exists between leukocytes and the kallikrein-kinin system, there are few clinicaI studies on leukemia patients.
Chiba et aI (1979) studied 14 acute myelocytic leukemia patients.Before and during the treatment they found decreased plasmatic PK and normal CI-INH.
Buzerak et aI (1982) studied 39 patients with acute leukemia and described increased kallikrein-kinin system activation in patients with a higher blast count.
.. This study was undertaken to evaluate some contact system parameters in 24 patients with ANLL at diagnosis, compared to a control group, and to verify correlation between possible alterations and bla~t count, some coagulation parameters and prealbumin.

PATIENTS AND METHODS
Blood samples were collected by clean venipuncture, in plastic tubes containing sodium citrate 3,8%, in a 9: 1 proportion.Following its centrifugation, the obtained plasma was kept at -80 D C until its use.
A prospective study, including all new cases diagnosed as ANLL in the Hemocentro at Unicamp between 1988 to 1991, was developed.Thirty-five cases of ANLL were considered for participation in this study.Eleven patients were not elegible, according to the exclusion cri teria: infection/septicemia, previous treatment or blood transfusion.Infection/septicemia was ruled out in the absence of fever, negative blood and urine culture, normal thorax X-ray and stable blood pressure.Finally, blood was analysed from 24 ANLL patients, average age 34 years (16-69 years), 14 men and 10 women.Diagnosis of renewed ANLL was based on blood count, bone marrow biopsy, cytology and cytochemistry.They were classified according to the FAB-system 4 (six Ml, nine M2, one M3, two M4, two M5, two M6 and two M7 cases).ClinicaI bleeding was present in 5 patients (21 % of cases) and characterized by epistaxis (cases 4 and 13), ecchymosis (cases 7 and 16) and hematuria (case 23).ClinicaI thrombosis was not presente Nineteen healthy volunteers, workers from the Hemocentro at Unicamp, average age 32 years (21-59 years), 11 men and 8 women, were the control group.
Plasmatic PK was measured after dextran sulphate activation, using an amidolytic assay~with Ac-Phe-Arg-pNA chromogenic substrate.~"Kallikrein-Iike"activity was evaIuated in the same way as the prekallikrein assay, substituting dextran sulphate as buffer.
Cl-inhibitor (CI-INH), alpha 2-macrogIobuIin (A2) and preaIbumin (PA) were measured by rocket immunoeIectrophoresis. 7cti vated partiaI thromboplastin time (APTT) was measured with kaolin and rabbit brain cephalin. 8Prothrombin time (PT) was obtained with rabbit brain thromboplastin. 9he APTT and PT results were expressed as' a ratio between coagulation time of patient plasma and a pool of 20 normal plasmas.Normal and prolonged APTT and PT reference plasmas were run daily as a laboratory quality controI.Factor XI and factor XII were measured by APTT corrrection time, using deficient plasma in factor XI and XII (Stago diagnosis) respectively.lOFactor V was .measuredby a described procedure. 11latelet count was determined automatically (CeII dyn 1600, Unipath, California, USA).
Statistical analysis was carried out using the Wilcoxon teste Patients and controIs were compared to all parameters studied.Patients were .alsoseparated according to leukemia subtypes, into group 1 (M1-6 cases), group 2 (M2-9 cases) and group 3 (M3 to M7-8 cases).This last group was made up from 4 subgroups, to achieve a great number of patients, making statistical analysis possible.Correlation between various parameters was performed: blast count or leucocyte count and PK; CI-INH or A2; PA and PK, PT or factor V; and PK and PT or factor V. The difference at p < 0.05 was considered significant.The results (tables 2 and 3) demonstrated that APTT, DISCUSSION CI-INH, A2, factor XI, factor XII and factor V were not different between patients and the controI group.Kallikrein-Disturbances ofhemostasis are still a frequent problem Iike activity was not increased in patients.PK was decreased in patients with acute Ieukemia.Prekallikrein or factor XII (p=0.02) and PT increased (p=0.003) in ANLL patients deficiency are associated with an in vitro coagulopathy when compared to the controI.PA was significantly Iower (prolonged APTT), in the absence of clinicaI bleeding.(p=O.03) in the patient group than in the controI one.When However, if activation of the contact system or proteolysis patients were separated into subgroups according to FABof the components of this system is present in leukemia, classification, no difference was verified in any variable intrinsic blood coagulation pathway could be compromised, measured.There was no correlation between circulant blast contributing to the coagulopathy seen in these patients.count or Ieucocyte count and PK, A2 or CI-INH.A positive

RESULTS
The objective of this study was to evaluate some correlation between PK and PA (~0.49, criticaI value = components of the contact system in 24 ANLL patients.Patients 0.43, p<0.05), and between PT and factor V (r=0.54,criticaI were studied at diagnosis, without chemotherapy or infection, value=0.44,p<0.05), was observed.
to avoid conditions which could, per se, alter the contact system.Bleeding was present in 21 % of cases, a little less than that described by others, and thrombocytopenia was probably the major responsible factor.The exclusion of patients with infection/septicemia, a condition that increases the risk of bleeding in patients with thrombocytopenia, could explain the lower incidence of hemorrhagia in our patients.
We have demonstrated that ANLL patients have low biologic PK activity.This fact could suggest an activation of the contact phase, as described in previous reports.2,3  Otherwise, the lack of reduction of the main PK inhibitor, could disprove that the kallikrein-kinin system has been activated in leukemia patients.-13 The normal leveI of C 1-INH in our patients does not role out increased turnover, since synthesis could be increased, associated with simultaneous consumption, by kallikreinkinin system activation.C1-INH antigen frequently is not.,14-15 Kallikrein is present in leucocytes and myeloid precursors.50, in leukemia, a disease with an increased number of these cells, we could expected a correlation between blast cell count and PK, C 1 or A2 leveIs.However, our results did not confirm this hypothesis.
Kallikrein-like activity was not increased in our patients, suggesting there was no spontaneous proteolysis triggered by enzymes liberated from blasts cells.
PK is synthetized in li ver, and it can be a reliable index of liver function.'6Impaired liver function could contribute to the decreased PK leveI in our patients.The finding of decreased PA, a sensitive parameter of liver synthesis, and a significative correlation between PK and PA in our ANLL patients, corroborate this hipothesis.However, we have to take into account that prealbumin can be decreased in clinicaI situations without liver synthesis insufficiency, in which a modulation of protein synthesis occurs, with an increase in acute phase proteins and reduction in others.The great majority of these ~lterations are mediated by cytokines that act on hepatocytes. 17The evaluation of other proteins that are involved in acute phase response, like fibrinogen, reactive .protein C and alpha 1-antitripsin, would be an important aid in the analysis of our results.Previous studies described coagulation abnormalities in myelocytic Ieukemia, particularly in the .subtype M3.Blast cells are rich in procoagulants which can trigger blood coaguIation by direct factor X activation, or by tissue factor-like activity.In some patients the activation of blood coagulation can determine an increase of APTT, PT and thrombin time.Activation markers ofblood coagulation, like thrombin-antithrombin complexes, fibrinogen degradation products and Ddimmers, are of interest in confirming this activation or not.In our patients, PT and PTTA were measured and PT was the only parameter altered.
PT increases by reduction of factor 11, VII, X or fibrinogen.
50, impaired liver synthesis, vitamin K deficiency, isolated reduction of one of these factors, or increased consumption, could raise PT.
Therefore, impairment of liver synthesis could contribute to the increased PT verified .inour patients, favored by simuItaneous PK and PA decrease.An increased turnover, due to acti vation of blood coagulation could also have contributed to this finding.
Tbere .wasno correlation between PK and PT, suggesting the kallikrein-kinin system does not play an important role in this alteration.
In some patients (cases 1, 15 and 19) activation of the blood coagulation intrinsic pathway by kallikrein is suggested by simultaneous PK and factor XII decrease.

Table 2
Parameters of the contact system, coagulation and PA in ANLL (R) Ratio of time of coagulation in seconds of patient to a pool of normal plasma.
Parameters of the contact system and coagulation in ANLL and control group (R) Ratio of coagulation time in seconds of patient to a pool of normal plasmas.