THROMBOSIS RESEARCH 47; 533-540, 1987 0049-3848/87 $3.00 t .OO Printed in the USA.

Copyright (c) 1987 Pergamon Journals Ltd. All rights reserved.

DIABETESMELLITUSAS A HYPERCDAGUIA&E STATE: ITS RELATIONSHIP WITH FIBRIN FRAGMENTSAND VASCULARDAMAGE L.J. Garcia Frade, H. de la Calle, I. Alava, J.L. Navarro, L.J. Creighton*and P.J. Gaffney* Haematology and Endocrinology Departments, Ram% y Cajal Hospital,Madrid, Spain and *Divisionof Haematology, NationalInstitutefor Fiological Standardsand Control,South Mimms, Herts., U.K.
(Received 16.2.1987; Accepted in revised form 16.6.1987 by Editor M.J. Seghatchian)

Ha-static variableswere assessed in 43 patients,28 insulindependentand 15 non insulin-dependent. Maximum aggregation by low concentrations of adenosinedipnosphate(ADP)or arachidonic acid and elevatedplasma concentrations of TxB2, Factor VIII, vWF:Ag, ZoF and fibronectin(Fnct)indicateda hyperccagulable state. The manifestation of vasculopathy was associated with elevatedooncentrations of RCoF, Fnct, Hbalc, cholesterol and triglycerides, while impaired fibrinolysis was demonstrated by decreasedt-PA levels and the absenceof crosslinkedfibrindegradation products (XL-FDP).

INl'E&XXJCTION Diabetes mellitus is characterisedby vascular complications. Accelerated disease of the microcirculation and the rmjor blood vessels accounts &or blindness, renal failure,atherosclerosis of the larger vessels of heart, brain and lower limbs and shortenedlife expectancy(1,2). It has been suggested that a hypercoagulable state associated with metabolic disturbance(3), may contributeto the vascular lesions and ha-static abnormalities which have been reported in microvascularand macrovasculardisease (4,5,6,7). Otherwise,the long-termmanagementof diabetes requires avoiwce of hypoglycaemia since this conditionmay inducehypercoagulability (8). Keywords: Coagulation; Diabetes mellitus; Fibrinolysis; Platelet function; Vasculopathy.
533

534

DIABETES, XL-FDP, VASCULAR DAMAGE

Vol. 47, No. 5

The present study assessed variablesassociatedwith platelets,coagulationand fibrinolysis,together with the measurement of high molecular weight fibrin fragments in order to investigatewhether fibrin breakdown is associatedwith the hypercoagulable state, and to identifywhich variablesare involved in the development of vasculardisease.
MATERIALSAND-

Subjects Forty-three patients (29 males, 14 females,age 39.5 + 18.5 years, mean + S-D., range 15-73) were studied: of these, 28 were insulin-dependent (20 males, 8 females,age 28.6 -r12.5 years, mean f S.D., range 15-58, 10 of whom presented with micro and/or macrovasculardisease), while 15 were non insulin-dependent (9 males, 6 females, age 59.7 + 7.9 years mean f:S-D., range 43-73, all with diagnosedmicro and/or macrovascular disease). All patients were treated with insulin (32.28 + 2.24 units/day). The duration of diabetes was 8.45 + 1.24 years. Patients were investigatedfor clinical signs and symptomsof vascular disease, includinginspectionof the optic fundi. A group of 25 healthy volunteers, matched for age and sex, with normal glucose concentration and no family history of diabetescomprisedthe control group. Materials Blood samples were obtained while the patients were in a fasting state, wernight, before receiving their insulin and after an observed supine rest for a minimum of 15 minutes. Blood samples were withdrawnusing a two-syringetechnique with an l&gauge butterfly needle, without stasis. Blood from the first syringe was mixed with 3.2% sodium EDTA for measurementof Hbelc; serum concentrations of cholesterol, triglycerides and thromboxanes (TxB2) were also measured. 5 ml of blood from the second syringe were added to tubes containing0.5 ml 4.5 mM EDTA + 10 ug/ml indomethacin. 20 ml of blood were collectedinto tubes cantaining 3.8% trisodiumcitrate (9 vol:l vol). Platelet-rich plasma (PFU?) was obtained from titrated blood centrifugedat 170 x g for 12 minutes and platelet-poor plasma (PPP) was obtainedby centrifuging at 4000 x g for 15 minutes. The PPP was aliqmted and stored at -7OOC. Methods Platelet aggregation PBP was adjusted to 300,000 platelet/u1with autologous model. ADP PPP. Plateletaggregationwas measured in a m-2 Malin Electronics (100 PM) and arachidonicacid (10 mM) were used to determine the aggregation thresholds. Determination of TxB2 Plasma and serum TxB2 were measured by radioimmunoassay using the New England TxB2 (3H) BIA kit. Blood was held at 37*C for a period of 1 hour to obtain serum. Factor VIII, von Willebrand Factor, antithrombinIII and Fibronectin Factor VIII coagulantactivity (VIII:C)was determinedby a one-stageassay (9), while von Willebrand Factor related antigen (vWF:Ag),antithranbinIII (AT III) and fibronectin (Fnct) were measured using a Laurel1 rocket immutXX?lectrophoretiC technique (lo), using rabbit anti-humanantibodies (BehringDiagnostics). pir III was also measured by a chromogenicassay1 using the substrateS-2238 (11).

Vol. 47, No. 5

DIABETES, XL-FDP, VASCULAR DAMAGE

535

Ristocetinco-factor (RCoF) was measured, using washed platelets according to the method of Weiss (12). Tissue plasminogen activator (t-PA) An enzyme-linked immunospecificassay (ELISA) described by Rijken et al. (13) was used with the exception that %well polyvinyl plates were coated with goat IgG antibodiesto t-PA and stored Each subsequent assay step was confined to 1 hour incubationsat at -4OY. 37Y. Assay of crosslinkedfibrin degradationproducts (XL-EDP) FDP levels in plasma were measured using specific monoclonal antibodiesto >[LEDP in an ELISA procedure (14,15). Hbalc was determinedby cationic interchangeresin column (Bio-Pad,Richmond, California). Cholesteroland triglycerides were measured using a calorimetric method (Testanar, Behring). Data analysis Values are presented as the msan + s.e.m. The significance of differencesbetween mean values was evaluatedusing Student'st-test. The chi squared test was used for evaluatingdifferencesbetween the diabetic subjects and controls and the effect of the various concentrations of ADP and arachidonic acid in the two groups. Correlationcoefficients were calculatedby the Pearson test. RESULTS Platelet aggregation In the diabetic patients maximum aggregationwas induced by a thresholdconcentration of ADP or arachidonicacid which was significantly lmer than in controls (Table1). Seventeenof 21 tested (81%) showed evidence of hyperaggregability in response to ADP and 12 of 23 (54%) in responseto arachidonicacid. TxB2 levels Plasma TxB2 levels were significantly higher in the patient group, while no differencewas found in the serum Tx82 levels when compared with the controlgroup (Table1). TABLE1 ThresholdConcentration of ADP and Arachidonic Acid for PlateletAggregation, Plasma and Serum Tx82 Levels Expressedin pg/O.l ml Minimum concentration of inducerto yield maximum plateletaggregation ADP umol/l: < 1.25 > 1.25 Arachiclonic acid ml/l:
0.5 1 12 < U.05

Controls

Diabetics

2 1U

17
4

< U-01

1 Serum TxB2 pg/O.l ml (n:lO) Plasma TxB2 pg/O.l ml (n:12)

11 60.9 + 4.9 2.6 f 0.1

11 52.6 + 4.0 7.3 f 1.9 n.s. < 0.05

536

DIABETES, XL-FDP, VASCULAR DAMAGE

Vol. 47, No. 5

Factor VIII, vWF':Agand AT III As is shown in Table 2, Factor VIII and vWF:Ag were significantly increased in diabetic patients, lothtype I andType II, and with or without vasculopathy. ICoF was slightly increased in the whole group, but in non insi-lin-dependent and vasculopathy groups the increase was highly significant. There was no difference in AT III levels between groups, and no significant difference was found between the two assays. TABLE 2 Concentrations of Coagulation Parameters (Results Expressed as Percentage f s.e.m.) Factor VIII % vWF:Ag % FCoF % AT III %x

Control CM IDIM NIDDM -VD +VD

104.5 +

5.5

107.5 +

4.6

108.3 +

5.5

98.5 + 2.2 105.6 ? 2.4 112.3 + 3.8 101.4 f 2.7 109.9 + 3.5 107.7 f 4.4

145.2 + 11.6* 142.0 + 12.5* 165.0 + 20.6* 139.0 + 13.9* 158.5 f 15.4*

213.0 + 20.4* 177.2 + 16.8* 271.0 f 45.8* 144.0 + 12.3* 254.2 i:27.3*

147.9 f 13.3+ 122.3 + 8.6

192.0 + 30.8* 118.0 + 11.2 169.9 f 20.3*

x assayed according to the method of Laurel1 (10) 1 P < 0.05 P < 0.001 IM - Diabetes mellitus (43) IDDM - Insulin-dependent diabetes mellitus (28) dIDDM - Non insulin-dependent diabetes mellitus (15) - Without vascular disease (18) -VD - With vascular disease (25) +VD Fnct An increase in mean plasma Fnct concentrations was observed in diabetic patents, with the vasculopathy group having the highest concentrations (Table 3). There was a relation between Fnct/Factor VIII r = 0.45 (P < 0.05) in the whole group of patients. The relations Fnct/vWF:Ag r = 0.58 (P < 0.01) and Fnct/FCoF r = 0.41 (P < 0.05) were significant only in the insulin-dependent diabetes mellitus patients. t-PA While there was no difference between the diabetic and control groups, diabetic patients with vasculopathy showed a decreased mean plasma concentration of t-PA (Table 3). XL-FDP Twenty-five patients were studied and abnormally high values were observed four. One of these presenting with carotid artery thrombosis (1129 ng/ml) and one with pyelonephritis (508 ng/ml) were excluded while two others had no apparent cause and were included. There was no difference between this 23-patient group and the control group (Table 3). Hbalc, cholesterol, triglycerides Patients with vasculopathy showed significant increases in the three variables in comlxrison with those witlxxt vasculopathy (Hbalc P < 0.01, cholesterol P < 0.01, t&glycerides P < 0.05). Hcwever, &&esterol and triglycerides values were still within the normal range (Table 4).

Vol. 47, No. 5

DIABETES, XL-FDP, VASCULAR DAMAGE

537

TABLE3 Concentrations of F'nct, t-PA and XL-FDP No. of subjects Control ml IDEM NIDDM -VD +VD 25 43 28 15 18 25 Ehct % t-PA iu/ml X-oligomers rig/d 147.1 i: 31.5 144.8 + 49.4 155.2 + 110.5 130.2 + 93.8 168.8 2 77.9 113.4 + 55.2

94.5 i: 3.8 119.7 i: 6.4* 120.6 + 8.4x 117.6 + 9.5+ 114.5 + 10.2+ 125.9 + 7.4* + P < 0.05 f P < 0.01 P < 0.001 TABLE4

2.4 + 0.5 1.6 + 0.3 1.6 f 0.4 1.1 f 0.2 2.0 + 0.5 0.9 -I0.1+

Concentrations of Hbalc %, Cholesterol(mg/dl)and Triglycerides(mg/dl)

No. of subjects

Hbalc

Cholesterol

Triglycerides

Control DM IDDM NIDDM -VD +VD

25 43 28 15 18 25

4.8 + 8.4 + 7.9 + 9.5 * 7.4 + 9.4 f

0.1 0.4* 0.5* 0.7* 0.6* 0.5*

200.5 + 202.8 + 202.6 + 203.3 t 175.6 + 226.9 ?

6.0 8.7 9.2 20.0 6.4 13.2+

90.2 -?: 6.0 106.8 f:14.5 85.1 + 15.1 154.5 + 28.3' 74.9 + 14.9 135.0 + 22.5x

+ P < 0.05 f P < 0.01 P < 0.001 DI!SCU!SION The diabetic patients passessad hyper-reactiveplatelets, as demonstratedby aggregationthresholdswith arachidonicacid and ADP. The higher plasma TxB2 concentrationsccnfirmed this observation,even taking into account the fact that 'basal' plasma TxB2 concentrationsmay have been affected by TxB2 synthesis during sample preparation. This confirms the data of many investigators who have described hyperactivityof platelets in diabetes msllitus and increased formationof TxB2 (16,17,inter alia). Factor VIII has been studied frequently in diabetic patients (3,6,7). mst authors found elevated values for Factor VIII, vWF:Ag and E&ZcF. The latterhas been described as higher in patients with vasculopathy (18) and this was confirmed in our study. Impaired fibrinolysis has been described at rest or after venous occlusionby some (19) while others have describedequal or higher than normal concentrations of t-PA. (3). We found decreased t-PA ccncentraticms only in the group with vasculopathy. The high plasma Fact concentrationsfound in the above patients corroborates

538

DIABETES, XL-FDP, VASCULAR DAMAGE

Vol. 47, No. 5

with the findings of some previous studies (20), but not of others (21,22), with the concentrations higher in the presence of vasculardisease. The significant correlationbetween Fnct and von WillebrandFactor (vWF:Ag,RCoF) found in IDDM patientsmight indicatean endothelial stimulusas 00-n origin. With regard to vasculardisease, the elevatedconcentrations of F'nct and vWF:Ag could be a reactive phenomenonand/or contributeas a cause. The finding of elevatedconcentrations of Fnct, Factor VIII and vWF:Ag in the IDDM group without vasculopathysuggests that ha-static abnormalitiescould exist prior to clinicalvasculardisease,perhaps relatedto a defectivemetaboliccontrol. It is evident from in vitro work (23) that XL-FDP (X-oligomers) originate from crosslinked fibrin and the presence of X-oligomersduring abruptio placentae and in other overt hypercoagulable stateshas been demonstrated(24). An assay with great sensitivityand specificityis needed to detect the low concentrations of these crosslinkedfragmentswhich may be in the circulation before the onset of clinical thrombosis. While an increased concentrationof fibrir soluble complexeshas been reported in diabetes (25), our results shu&d that elevated concentrations of high molecularweight fibrin fragments(X-oligomers were not present in patientswith this condition. Diabeticpatientshave a variety of haemostaticabnormalities allowing the condition to be described as a hypercoagulable state (26). Nevertheless,in the patients in this study, high molecularweight fibrin fragmentswere not detected, as might be expectedduring hypercoagulability in the presence of a normal oi fibrinolyticresponse. It can be speculatedthat increasedconcentrations soluble fibrin (25), together with a poor fibrinolytic response (e.g. low t-P? concentrations and no increase in XL-EDP) may explain the diabetic vasculopathl accompaniedby fibrin deposition. The hypercoagulablestate may be present before the clinical manifestation of vascular damage. Markers of vasculopathy may include increasedconcentrations of FCoF and Fact and decreasedconcentrations of t-PA, although some of these alterationscould be partiallyrelatedtc age.

1. CO=, J.A. and HALUSHKA, P.V. Platelet function in diabetes mellitus. 44, 521-526,1980. Br.J.Haematol.2. MUSTARD,J.F. and PACKHAM,M.A. Plateletsand diabetesmellitus. N.-l. J.Med. 311, 665-666,1984. -3. HEH=, A., MCVERRY, B.A., WIIKINSON,L., GQLDSTONE, A.H., WIS, D. anti BLCOM, A. Diabetes: a hypercoagulablestate? Haemcstatis variables ir newly diagnosed type 2 diabetic patients. Acta Haematol.69, 254-259, 1983. 4. JONES, R.L. and PETERSCN, C.M. The fluid phase of coagulation and the accelerated atherosclerosisof diabetes mellitus. Diabetes, suppl.& 33-38, 1981. 5. ElUSlK!HI, J., CHHISTE, M., LlMlLEi, B., MARBEX, G.A., BERGER, W. an< UEUERT, F. Platelet aggregation,B-thromboglcbulin and platelet factor 1 in diabetes mellitus and in patients with vasculopathy. Thrombos. Hamtas. 52, 236-239,1984. -

Vol . 47, No. 5

DIABETES, XL-FDP, VASCULAR DAMAGE

539

6. CBRISTE, M., FRISTCHI,J., L,QMLE, B., TRAN, T.H., MARBET, G-A., BERGER, W. and DUCKEKT,F. Fifteen coagulation and fibrinolysis parametersin diabetes mellitus and in patients with vasculopathy. Thrombos.Haenkostas. 52, 138-143,1984. 7. BERN, M-M., CASSANI, M.P., FK)RTON, J., RAND, L. and QU'IS, G. Changes of fibrinolysis and factorVII coagulant,antigen, and ristocetincofactor in 19, 831-839,1980. diabetesmellitusand atherosclerosis.ThrombosisRes. J. Changes in platelet J., MADSBAD, S. and HILSI'ED, 8. ~NEIISON, function,blood coagulationand fibrinolysis during insulin inducedhypglycaemia in juvenile diabetic and normal subjects. Thrombos.Baemostas. 47, 254-258,1982. 9. GIDDINGS, J.C. Hereditary coagulationdisorders: laboratorytechniques. In: Blood Coagulaticn and Haenkostasis. J.M. Thomson (Ed.) Edinburgh, Iondon and New York: ChurchillLivingstone, 1980, pp. 117-157. in 10. IAURELL, C.B. Quantitativeestimationof proteins by electrophoresis agarose gel containingantibodies. Anal.Biochem. 15, 45-52, 1966. 11. ABILDGAARD,U., LIE, M. and ODEGARD,O.R. Antithrombin(heparincofactor) chrcmogenic substrates (S-2238 and chrcmozym-TH). a=aY with lXW ThrombosisRes. 11, 549-553,1977. A. and HYXRS, J. Quanti12. WEISS, H.J., mYER, L.V., RICKLES,F.R., -, tative assay of a plasma factor deficient in von Willebrand'sdisease that is necessary for platelet aggregation. J.Clin.Invest. 52 2708-2716, _' 1973. 13. RIJKEN, D.C., VAN HINSBERG, V.V.M. and SENS, E.H.C. Quantitation of tissue type plasminogenactivator in human @&thelial cultures by use of an enzymairmnunmssay. ThrcmbosisRes. 33, 145-153,1984. 14. RYIXIT, D.B., BLARE, A.S., CXTIS, L.E., MASSINGHAM,D.A., FLETCHER,W.A., MASCI, P.P., WBITAKER,, A.N., ELMS, M., BUNCE, I., WEBBER, A.J., WYATT, D. and BUNDIZZN, P.G. An immunoassayfor human D dimer using monoclonal antibodies. ThrombosisRes. 31, 767-778,1983. 15. PERRY, M.J., CALLUS, M., TIXXPE, R., SPITZ, M. and GAFFNEY, P.J. mm-. clonal antibodies specific for high molecular weight crosslinked fibrin degradation products. Br.J.Haematol. (submitted) 16. KUBISZ, P., ARABI, A., I-XX&N, J. and Cw)NBER(;, S. Investigationson platelet functionin diabetesmellitus. Haemostasis 14, 347-353, 1984. 17. HALUSI-KA, P-V., CURTIS RCGEBS, R., BilyDIOADHOLT, C. and COLlX'ELL, J.A. Increased platelet thromboxane synthesis in diabetes mellitus. J-Lab. Clin.Med.97, 87-96, 1981. 18. LUFKIN, E.G., FASS, D-N., O'FALKX, W.M. and BOWIE, E.J.W. Increasedvon WillebrandFactor in diabetesnrallitus. Metabolism28, 63-66, 1979. 19. -R, L-O., SUNMNIST, G. and LIIJA, B. Fibrinolytic activity,autonomic neuropthy and circulationin diabetesmellitus. Diabetes32, 4-7, 1983.

540

DIABETES, XL-FDP, VASCULAR DAMAGE

Vol. 47, No. 5

20.

SOLERTE, S.B., PIOVELLA, F., VIOLA, C., CARNEVALE, G., GAMBA, G., FIORAVANTI, M. and FERRARI, E. Plasma fibronectin, von Willebrand Factor antigen and blood rheology: association with diabetic microvascular disease. Acta Diabetol. 22, 239-246, 1985. LABAT-ROBERT, J., LEUT-, M., uu3PIS, G., RICARD, Y. and D-E, 2 J.C. Plasma and tissue fibronectin in diabetes. Clin.Physiol.Biocha. _' 39-48, 1984. LAMBERTON, R-P., GOaDMAN, A.D., KASSOF, A., RUBIN, C.L., TREBLE, D-H., Von Willebrand factor SAVA, T-M., MERIMEE, T.J. and DOODS, W.J. (VIIIRAg), fibronectin, and insulin-like graYJth factors I and II in diabetic retinopathy and nephropathy. Diabetes 33, 125-129, 1984. Giant fibrin fragments derived GRFENEY, P-J., JOE, F. and MA#UJD, M. from crosslinked fibrin: structure and clinical implication. Thrombosis Res. 20, 647-662, 1980. -L. Subunit and macromolecular strucGRAEFF, H., HAETER, R. and E!ACH'l&NN, ture of circulating fibrin from obstetric patients with intravascular 16, 313-328, 1979. coagulation. Thrombosis Res. GENSINI, G.F., ABBU'E, R., FAVILLA, S. and NERI SERNERI, G.G. platelet function and blood clotting in diabetes mellitus. Hamstas. 42, 983-993, 1979. SCHAFER, A.I. 1985. The hypercoagulable state. Changes of Thrcmbos.

21.

22.

23.

24.

25.

26.

Ann.lnt.Med. 102, 814-828,