Prognostic value of hemostatic parameters after liver transplantation

Journal of Hepatology, 1992; 15: 125-128 @ 1992 Elsevier Science Publishers B.V. All rights reserved.

125

0168~8278/92/$05.00

HEPAT 01014

Marco Moia’,

run0 &;ridelli2, Martin Langer3, Ida Martinellil, and Pier Mannuccio Mannucci’

Dinangelo

Galmarini*

‘A. Bianchi Bonomi Hemophilia and Thrombosis Center and Institute of Internal Medicine, ’ Liver Transplantation Unit and ‘Institute of Anesthesiology and Intensive Care, IRCCS Maggiore Hospital and University of Milan, Italy

(Received 16 April 1991)

The prognostic value of hemostatic parameters after orthotopic liver transplantation was evaluated in 37 consecutive patients. Six simple hemostatic parameters (prothrombin time, activated partial thromboplastin time, thrombin time. thrombin coagulase time, plasma fibrinogen and platelet count) were obtained for each patient pre-transplantation and daily post-transplantation for at least 8 days. Using the results of these tests, the degree of hemostatic impairment was arbitrarily scored from 0 to 6. Starting from the first day post-transplantation, hemostatic parameters improved progressively, reaching plateau valdes on day 7 post-transplantation. On day 8 there were significant differences in the activated partial thromboplastin time, prothrombin time, and in the overall hemostatic scores between patients who survived at least 6 months and those who died. Comparing these hemostasis parameters with such liver function tests as AST, ALT and serum bilirubin, univariate analysis showed that activated partial thromboplastin time, coagulation score and AST were significant predictors of 6-month survival, but by multivariate analysis (Cox proportional hazard rate model) only the activated partial thromboplastin time was an independent predictor. Hence, a simple coagulation test is useful for predicting the survival of patients undergoing liver transplantation.

Orthotopic liver transplantation (OLTX) is performed in an increasing number of patients with end-stage liver disease (1). Since the liver plays a central role in the regulation of hemostasis, hemostatic parameters may be useful for monitoring patients undergoing OLTX (2). A previous study demonstrated that the degree of abnormality of hemostatic parameters pre-OLTX, which depends on the patient’s own liver function, is predictive of (j-month survival (3). In principle, hemostatic parameters post-OLTX should be even more useful for predicting patients survival since they depend on the function of the transplanted liver, and no correction or poor correction of the pre-operative hemostatic abnormalities would indicate poor graft function (4). During simple tests of hemostasis pre- and post-OLTX, we observed that a deterioration of results at the end of the first week postOLTX seemed to be predictive for poor medium-term outcome. Since there is little data about hemostatic parameters in the post-OLTX period (5), the predictive Correspondence.

P.M. Mannucci, Via Pace 9, 20122 Milano, Italy.

values of six hemostatic parameters and an overall score of hemostatic abnormalities were investigated in 37 consecutive patients undergoing OLTX.

Patients

and Methods

Patients

Thirty-seven consecutive patients who underwent OLTX between April 1984 and May 1988 at the University Hospital of Milan were investigated and, when possible, followed-up for at least 6 months. Four different groups of patients were transplanted: postnecrotic cirrhosis (n = 23, 13 men and ten women, age range 17-50 years), primary biliary cirrhosis (n = 4, all women, age range 40-56 years), hepatocellular carcinoma (n = 7, five men and two women, age range 38-54 years), and other miscellaneous diseases (one alcoholic cirrhosis, one Wilson’s disease, one fulminant hepatic failure). Six pa-

M. MQL4 et al.

126 tients underwent a second OLTX; one patient had three OLTX. In patients with two grafts, only one (either the first or the second) functioned for at least 8 days. In the last patient, the first and third grafts both functioned for at least 8 days. For the purpose of the study, we arbitrarily considered only the last OLTX of this patient, which lasted longest. The operative technique and surgical and anesthesiological personnel were the same for all patients. One patient died during the operation from cardiac failure due to pre-existing pulmonary hypertension and another

died postoperatively

week post-OLTX)

from cardiac

tients

1 month,

died within

within 3 months,

one within

failure.

one within 4 months.

(during

the first

Nine other

pa-

2 months,

two

Hence, 22 of 37

thrombin time, is not influenced by heparin. Fibrinogen (Fibrinogen Reagent, Bozhringer) is the end substrate of the coagulation cascade and platelets are the key component of primary hemostasis. The prothrombin time, aPTT, thrombin and thrombin coagulase times were expressed as the ratio of the clotting time of patient’s plasma to that of a reference plasma pooled from 30 normal donors. To give a comprehensive evaluation of hemostasis, an overall score was calculated for each patient as suggested by Bontempo et al. (3), arbitrarily assigning one point for each abnormal test independently of the degree of abnormality. The overall score ranges from 0 to 6, depending on the number of abnormal tests for each patient.

patients (59%) were alive at 6 months. Biochemical parameters Hemostatic parameters

Hemostatic parameters were measured in each patient pre-OLTX and daily post-OLTX for 8 days. These included prothrombin time (laboratory range: ratio 0.901.20), activated partial thromboplastin time (aPTT) (ratio 0.88-1.22), thrombin time (ratio 0.86-1.25), thrombin coagulase time (ratio 0.86-1.35), plasma fibrinogen assay (1.5-3.0 g/l) and platelet count (lSO-400e109/ 1). The prothrombin time (Manchester Reagent, Laboratori Baldacci, Pisa, Italy) is an overall measure of the tissue-factor induced coagulation pathway. The activated partial thromboplastin time (Automated aPTT, Organon Teknika, Milano, Italy) is an overall measure of the surface-induced coagulation pathway. Thrombin time (Thrombin Reagent, Boehringer Mannheim, Federal Republic of Germany) is an overall measure of the thrombin-tibcirogen reaction. Thrombin coagulase time (Thrombin Coagulate Boehringer) is an overall measure of the thrombin-fibrinogLn reaLdon, but unlike the

Biochemical parameters, measured in each patient at the same time as the hemostatic parameters, were serum bilirubin, aspartate aminorransferase (AST) and alanine aminotransferase (ALT) levels, measured by automated calorimetric tests. Blood sampling

For all hemostatic tests 4.5 ml of blood was collected by venipuncture into a Vacutainer containing 0.5 ml of 0.129 M trisodium citrate. Samples for platelet counts were collected into a Vacutainer containing EDTA. Statistical methods The Student’s t-test and Mann-Whitney U-test were used to evaleate the significance of differences between groups, depending on whether or not values were parametric and normally distributed. Post-OLTX survival was estimated with the Kaplan-Meier product-limit estimator and graphically recorded on the Kaplan-Meier curve.

7

L__--______

8

30

TO dayLc after

90

120

transplantation

150

180

Rg. 2. Kaplan-Meier survival o:rves for patients divided into two groups according to the results ot :)?e aPTT on day 8 post-GLTX. 0

1 days

after

6 7 transplantation

8

1.aPlT ratios (means f S.E.) from the day of OLTX (day 0) and until day 8 post-OLTX for patients who survived more than 6 months (solid line) and fess than 6 months (broken line),

Fig.

The solid line is the survival curve for @ems (n = 20) with normal aPTT (ratios equal to or less than 1.22); I:;- broken line is the survival curve for patients (n = 13) with abnorma.; STT (ratios greater than 1.22). The difference in survival between the two groups is statistically significant (p C 0.005).

HEMQSTASIS

AFTER

LIVER

TRANSPLANTATION

127

TABLE 1

Changesin hemostatic parameters before and soon after orthotopic liver transplantation. Test

Pre-OLT

Prothrombin time (ratio) aPTT (ratio) Thrombin time (ratio) Thrombin coagulase time (ratio) Plasma Itbrinogen (mg/di) Platelet count (xlO’/f) Score

1.88 1.81 1.20 1.50

Time post-QLTX day 1

f 0.83 f 0.9lJ + 0.23 dz 0.40

1.86 1.46 1.09 1.18

+ + f +

0.70 0.34 0.27 0.23

137*(68-420)

198(95-340)

48* (6-257) 4” (O-6)

34 (14-162) 3 (2-6)

day 7 ~---__~~ 1.40 t 1.14 + 1.04 + 1.23 f

day 8

~~ 0.37 0.17 0.17 0.24

1.46 1.23 1.08 1.17

264(121-500)

228(132-680)

-I2 (10-305) 2 (l-5) ____Values are means t one standard deviation, except those identified by an asterisk (median and range).

Univariate analysis (Mantel-Cox test) (6) was used to evaluate whether or not hemostatic parameters, AST, ALT and serum bilirubin on post-OLTX day 8 were predictors of survival. Differences were assumed to be significant by univariate analysis when p -=I0.005. Variables. with a probability of occurring by chance of less than 20% in univariate analysis were introduced into muhivariate regression analysis (Cox proportional hazard rate model) (7), to evaluate the discriminant power of hemostatic and liver function measurements in predicting survival. Odds ratios and confidence intervals were used to assess the power of the predictive value of hemostasic parameters for 6 month survival.

esults For all patients who underwent OLTX, the mean values of hemostatic parameters and median scores on day 0 (pre-OLTX) and on days 1, 7 and 8 post-OLTX are shown in Table 1 (values for days 2-6 not shown). The values of most measurements and of the overall score. &normal before OLTX, clearly improved from the first day post-OLTY, reaching a plateau on day 7. Only platelet counts did not slgttificz~+t;’ kprove during this period. Whereas baseline coagulation tests showed no significant difference between patients who survived at least

63 (12-320) 2 (O-6)

6 months and those who died within 6 months (data not shown), there was a difference between the two groups starting from day 7 post-OLTX, with progressive deterioration of most hemostatic parameters and of the overall score for the patients who died. Fig. 1 shows the changes over time for the mean aPTT values for the two groups. Prothrombin, thrombin and thrombin coagulase times, plasma fibrinogen and scores showed the same trends as aPTT, but for aPTT, prothrombin time and the overall score were there significant differences on day 8 postOLTX between patients who survived and those who died (Table 2). We compared aPTT, prothrombin time and the overall score on day 8 post-OLTX with AST, ALT and serum bilirubin levels to assess the predictive value of those tests for 6-month survival. This analysis included only patients who survived for at least 8 days post-operatively (rt = 35) but not two patients for whom results of hemostatic parameters on day 8 were not obtained. Hence analysis was carried out on a total number of 33 patients, 20 who survived at least 6 months post-OLTX and 13 who died. Univariate analysis showed that only the aPTT, coagulation score and AST were significant predictors of 6 month survival @ = 0.0056, p = 0.015 and p = 0.0255), whereas ALT and serum bilirubin were not @ = 0.56 and p = 0.07). However, multivariate analysis demonstrated that oniy the aPTT was an independent predictor of

TABLE 2 Results of hemostatic parameters Test

(means f S.D.) day 8 post-transplantation ~____ Patients

.__ Prothrombin time (ratio) aPlT (ratio) Thrombin time (ratio) Thrombin coagulase time (ratio) Plasma fibrinogen (mgldl) Platelet count (~10’0) Score ...- ____ Values are means + one standard deviation, except those

in relation to survival

___ Survivors at 6 months

1.29 I? 0.27 1.09 ir 0.16 1.05 + 0.10 1.07 + 0.18 226* (137-410) 77* (16-320) 2’ (O-3)

f 0.55 ?I 0.32 -c 0.16 f: 0.21

Dead within 6 months -1.73 + 0.76 I .40 z? 0.42 1.12 k 0.21 1.24 f 0.32 230 (132-680) 37 (12-126) 3 (l-6)

identified by an asterisk (median and range).

P

< 0.05 < 0.01 N.S. N.S. N.S. N.S. < 0.01

M. MOIA et al.

128 6-month survival (p = 0.001). The Kaplan-Meier survival curve showed better survival for patients who had normal aP’IT (ratio ~1.22) on day 8 post-OLTX than for patients with prolonged aP’TT (6 months survival: 80% vs. 31%, p c 0.005)(Fig. 2). For patients with normal aPlT on day 8, the chance of surviving at least 6 months was 9-times higher than for patients with prolonged aPlT (odds ratio = 9.0; confidence intervals: 1.9-42.3; chi square = 8.03, p < OAOS). Data for peri- and post-operative transfusion with blood derivatives showed that shorter term survivors were more intensively transfused than longer term survivors (medians and ranges of units transfused): 81 (26128) vs. 33 (12-136) red blood cell concentrates (p < 0.005); 149 (45-372) vs. 68 (23-286) fresh frozen plasma (p < 0.005); 18 (O-48) vs. 4 (O-36) platelet concentrates 0, < 0.05) and 0 (O-40) vs. 0 (O-36) cryoprecipitate (N.S.).

Discussion Our data indicate that post-operative hemostatic parameters may help predict the outcome of OLTX. The values of most measurements, abnormal before OLTX improved from the first day post-OLTX, reaching a plateau at day 7. To our knowledge, the only previous study evaluating the behaviour of coagulation tests post-OLTX is the study by Stahl et al. (5), which showed that in 30 consecutive patients undergoing OLTX, the aPTT and prothrombin time became normal on days 1 and 5 after OLTX. Hence, the recovery of normal prothrombin time :.;id aPlT was more rapid in Stahl’s patients than in ours (5 to 7 days). Since Stahl’s patients were not analyzed for survival, we do not know whether there was a relationship between survival and improvement of hemostasis tests. In this study, we did a number of liver function tests (AST. ALT and serum bilirubin) to look for possible References 1 Bismuth H, Ericzon BG. Rolles K, et al. Hepatic transplantation in Europe. First report of the European liver transplantation registry. Lancet 1987; ii: 674-6. 2 Mannucci PM. Diagnosis and assessment of bleeding tendency in chronic liver failure using three simple coagulation tests. Stand J Haematol 1970; 7. 364-7. 3 Bontempo FA, Lewis JH, Van Thiel DH, et al. The relation of preoperative coagulation findings to diagnosis, blood usage, and survival in adult liver transplantation. Transplantation 1985; 39: 532-6.

correlat.ions with hemostatic parameter:. Univariate analysis demonstrated that aP’lT, the coagulation score and AST levels on day 8 post-OLTX were predictors of survival, but only aPTT was an independent predictor by multivariate analysis. Patients with normal aPTT on day 8 post-OLTX had significantly better prognoses than patients with abnormally prolonged aPTT (6-month survival: 80% vs. 31%). In addition, calculation of the odds ratio showed that patients with normal aPTT values on day 8 post-OLTX had on average a 9-times greater chance of surviving at least 6 months than patients with prolonged aPTT values. The better hemostatic parameters observed in longer term survivors on day 8 postOLTX cannot be attributed to the corrective effect of transfusion with blood components, since longer term survivors were less intensively transfused. This supports previous findings (3). In our study there were no significant differences in baseline coagulation tests between patients who survived at least 6 months and those who died. This is different from data of Bontempo et al. (3), who found that the degree of abnormalities pre-OLTX could predict survival at 6 months. However, since the patients who died before day 8 were excluded from our analysis, no definite conclusion can be drawn about the predictive value of pre-OLTX coagulation tests, nor can our findings be validly compared with those of Bontempo. We conclude that the aPTT performed 8 days after OLTX is a simple index to predict patient survival independently of other liver function tests.

Acknowledgements We would like to thank Dr. A. Gringeri and Dr. A. Sangiovanni for their advice about the statistical analyses and Professor M. Colombo for critically reading our manuscript.

4 Grenvick A, Gordon R. Postoperative care and problems in liver transplantation. Transplant Proc 1987; 19: 26-33. 5 Stahl RL. Duncan A. Hooks MA, et al. A hypercoagulable state follows orthotopic liver transplantation. Hepatology 1990; 12: 553-8. 6 Benedetti J, Yuen K, Young L. Life tables and survival functions. In Dixon WJ. ed. BMDP !Statistical Software. Berkeley: University of California Press, 1983; 557-75. 7 Hopkins A. Survival anallysis with covariates. Cox model. In: Dixon WJ, ed. BMDP Statistical Software. Berkeley: University of California Press, 1983; 576-94.