Usefulness of Thromboelastography for Perioperative Evaluation of Hemostatic Profile in Patients with Primary Brain Tumors Undergoing Surgery.

Context: Hemostatic abnormalities are more common in patients with brain tumors than systemic malignant diseases. Conventional coagulation tests (CCT) are poor assays for dynamic assessment of clot strength in whole blood. Thromboelastography (TEG) gives us detailed information on the dynamics of clot development, stabilization, and dissolution reflecting in vivo hemostasis. TEG can assess both thrombosis and fibrinolysis. Aims: This study aimed to investigate the temporal trends in hemostatic profile occurring during surgery for primary brain tumors, using a combination of TEG and CCT, and to assess perioperative blood component support. Subjects and
Methods: A prospective, observational study was done on 40 patients with primary brain tumors larger than 4 cm in maximum diameter on computed tomography or magnetic resonance imaging. The tests (TEG and CCT [PT, INR, activated partial thromboplastin time, and platelet count]) were performed preoperatively (on the day of surgery), intraoperatively (2 h into surgery), and postoperatively (the day after surgery). Statistical Analysis: SPSS Version 21.0 statistical analysis software was used.
Results: We found a universal trend toward hypercoagulability (persistent decrease in R-time, K-time and increase in MA, α-angle, Coagulation Index) in all the TEG parameters measured intraoperatively and postoperatively even though the values were within normal limits. Results of CCT had poor correlation with TEG parameters. The mean intraoperative blood loss was 737.7 ± 185.6 mL, for which PRBC was transfused in 17 patients, FFP in 13, but no platelet transfusion was done intraoperatively. Conclusions: We found a trend toward hypercoagulability in our study in intraoperative and postoperative period using TEG which was not evident on CCT. TEG was a useful diagnostic tool to identify coagulation abnormalities and to guide perioperative blood transfusion. Copyright:

INTRODUCTION

Approximately 1.4% of all cancers are brain cancers, and they cause 2.3% of all cancer-related deaths.[1] The incidence of primary cerebral malignancies varies between 4 and 10/100,000 in the general population.[1] Although hemostatic abnormalities are common in patients with all types of malignancies, the incidence of thromboembolism is higher in patients with brain tumors than in those with systemic malignancies.[2]

Prothrombin time (PT), activated partial thromboplastin time (aPTT), and other plasma-based coagulation assays are traditionally utilized to assess hemostasis. However, they provide little information about the dynamics of clot formation or the quality of clot.[3] These conventional coagulation tests (CCTs) are nearly incapable in identifying the activation phase of the coagulation system and postoperative prothrombotic phase.[3] Therefore, they remain poor assays for dynamic assessment of clot strength in whole blood. Thrombelastography (TEG) is a laboratorial test that measures viscoelastic changes of the entire clotting process. It provides full information on the dynamics of clot development, stabilization, and dissolution which reflects in vivo hemostasis. Both thrombosis and fibrinolysis can be assessed using TEG.[4]

Aims and objectives

This prospective observational study was conducted in patients with primary brain tumors undergoing craniotomy for surgical management of the same. The primary objective was to observe the temporal trends in coagulation profile in the perioperative period in such patients and to identify specific patterns in coagulation abnormalities using TEG. Secondary objectives were to compare TEG with CCT and to assess and guide the blood component support during brain tumor surgery.

SUBJECTS AND METHODS

The study was conducted in the Department of Anesthesiology and Critical Care, in collaboration with the Department of Neurosurgery in our institute after getting ethical clearance from the institutional ethical committee (273/ETHICS/18) and registration in Clinical Trial Registry India (CTRI/2019/018219). Informed written consent was obtained from all the patients enrolled in the study. A total of 40 patients with primary brain tumors larger than 4 cm in maximum diameter on computed tomography (CT) or magnetic resonance imaging (MRI) undergoing elective surgery were enrolled in this study. Patients with a history of hematological or coagulation disorders, those taking anticoagulant therapy, and those with chronic diseases like liver cirrhosis or renal failure were excluded. Blood samples were taken for TEG and CCT in the preoperative (on the day of surgery), intraoperative (2 h after beginning of surgery), and postoperative (the next day after surgery) periods.

After admission and adequate preoperative preparation, patients planned for surgery were taken into operation theater. Standard anesthesia monitors were attached. The patient's blood pressure was recorded and intravenous infusion of crystalloids started. Standard induction of GA was done. Intraoperative and postoperative blood samples were drawn from the central venous (internal jugular) line. Blood samples for TEG were collected into 1.8 mL vacutainer tubes and 1 mL was transferred to a vial containing kaolin (phospholipids). From this vial, about 340 μL was transferred to a 37°C prewarmed disposable cup containing 20 μL of calcium chloride and TEG was done on this sample using Thrombelastograph Hemostasis Analyzer or TEG® (Haemonetics Corporation, Niles, USA). The following TEG data were obtained: R-time (reaction time), K-time (kinetic time), α-angle (alpha-angle), MA (Maximum Amplitude), and Coagulation Index (CI). The following laboratory tests were obtained – complete blood counts (hemoglobin, TLC, DLC, and platelet count) and CCT (PT, INR, and aPTT).

The statistical analysis was done using Statistical package for the Social Sciences-21 (SPSS Inc., Chicago, Illinois, USA) statistical analysis software. The values were represented in number (%) and mean ± SD. To test the significance of two means, the Student “t”-test was used. A value of P < 0.05 was considered significant.

The sample size was calculated on the basis of preoperative and postoperative variations in K-time in a previous study by Goobie et al.[5] Keeping confidence level at 95%, power of study 90%, and considering data loss of 10%, The sample size came out to be 40. All the patients fulfilling inclusion criteria and giving consent were enrolled in the study till a total of 40 patients were enrolled.

RESULTS

All 40 patients enrolled in our study had supratentorial tumors, of which 22 were glioma and 18 meningioma. The mean age of patients was 45.8 ± 9.65 years. There were 20 male and 20 female patients. All the patients belonged to The American Society of Anesthesiologists (ASA) Physical Status Classes I, II, and III. All the patients had a Glasgow Coma Scale score of E4V5M6. Intraoperative bleeding was present in all our patients which ranged from 393 mL to 1000 mL and the mean blood loss was 737.7 ± 185.6 mL, for which PRBC was transfused in 17 patients and FFP in 13, but no platelets transfusion was done [Table 1]. The baseline parameters of all the patients were within normal range and they remained hemodynamically stable throughout perioperative period.

Table 1

Demographic and tumor characteristics

Admission parameters	Value
Age (years), mean±SD	45.8±9.65
Gender (male/female)	20/20
ASA Physical Status Class (I/II/III)	13/23/04
GCS E4V5M6	40
Diagnosis (glioma/meningioma/others)	22/18/0
Tumor size (cm), mean±SD	5.67±1.29

SD=Standard deviation, ASA=American society of anesthesiologists, GCS=Glasgow Coma Scale, TLC=Total leukocyte count

Coagulation parameters

We found a significant reduction in intraoperative and postoperative platelet count compared to preoperative values (P < 0.001). However, the counts were within the normal range [Table 2]. Significant changes were also found in PT (P = 0.002), INR (P = 0.005), aPTT (P = 0.002), Hb (P < 0.001), and TLC (P < 0.001) in the intraoperative and postoperative periods compared to the preoperative values. However, the values were within the normal range [Table 2]. Intraoperatively, the values of PT, aPTT, and INR had increased from the preoperative values, but they decreased again postoperatively to values even below the preoperative values [Table 2].

Table 2

Comparison between preoperative, intraoperative, and postoperative coagulation parameters

Parameters, mean±SD	Time			Repeated measures ANOVA
	Preoperative	Intraoperative	Postoperative	F	P
Platelet (Lacs/µL)	2.56±0.74	2.37±0.71	2.14±0.73	247.6	<0.001
PT (s)	15.23±1.71	16.20±2.11	15.19±0.81	6.61	0.002
INR	1.17±0.13	1.23±0.17	1.16±0.07	5.57	0.005
APTT (s)	34.88±2.97	36.95±3.57	34.70±2.49	6.63	0.002
Hemoglobin (g/dL)	13.07±0.38	12.50±0.48	13.00±0.53	35.24	<0.001
TLC (×109/L)	9.10±1.81	8.78±2.01	8.94±1.90	9.74	<0.001

SD=Standard deviation, PT=Prothrombin time, APTT=Activated partial thromboplastin time, INR=International normalized ratio, TLC=Total leukocyte count

Thromboelastography parameters

The mean values of R-time (which represents the time to initial thrombin formation) were 5.18 ± 0.62 min, 4.85 ± 1.62 min and 4.06 ± 0.53 min at preoperative, intraoperative, and postoperative periods, respectively [Table 3]. We found a decrease in mean R value intraoperatively as compared to preoperative value, but the change was not significant (P = 0.201). However, there was a significant reduction in R-time postoperatively compared to the preoperative (P < 0.001) and intraoperative values (P = 0.005).

Table 3

Comparison between preoperative, intraoperative, and postoperative TEG parameters

Parameters, mean±SD	Time			Repeated measures ANOVA
	Preoperative	Intraoperative	Postoperative	F	P
R-time (min)	5.18±0.62	4.85±1.62	4.06±0.53	13.21	<0.001
K-time (min)	2.54±0.41	1.93±0.55	1.79±0.43	89.32	<0.001
α-angle (°)	54.80±5.40	61.94±7.87	68.57±5.70	162.93	<0.001
MA (mm)	58.77±4.51	61.7±6.13	65.45±6.01	33.59	<0.001
CI	0.33±1.59	0.68±1.61	1.07±1.63	10.19	<0.001

SD=Standard deviation, MA=Maximum amplitude, CI=Coagulation Index

The mean K-time (which represents the time to clot formation) measured at preoperative, intraoperative, and postoperative time points was 2.54 ± 0.41 min, 1.93 ± 0.55 min, and 1.79 ± 0.43 min, respectively. There was a significant decrease in K value intraoperatively as compared to preoperative values (P < 0.001). The K values decreased further in the postoperative period and the change was highly significant (P < 0.001) as compared to preoperative value but not significant (P = 0.059) when compared to intraoperative values.

α-angle (which represents the speed of solid clot formation) measured at preoperative, intraoperative, and postoperative time points had mean values of 54.80 ± 5.40 min, 61.94 ± 7.87 min, and 68.57 ± 5.70 min, respectively [Table 3]. We found a significant increase in intraoperative α-angle values as compared to preoperative values (P < 0.001). There was a further increase in postoperative value which was significantly higher than preoperative (P < 0.001) and intraoperative values (P < 0.001).

The mean values of MA (which is a measure of clot strength) at preoperative, intraoperative, and postoperative time points were 58.77 ± 4.51 mm, 61.7 ± 6.13 mm, and 65.45 ± 6.01 mm, respectively [Table 3]. There was a significant increase in intraoperative MA value as compared to preoperative value (P < 0.001). There was a further increase in postoperative value which was significantly higher than preoperative (P < 0.001) and intraoperative values (P = 0.001).

The mean values of CI (which represents the overall coagulation status) were 0.33 ± 1.59, 0.68 ± 1.61, and 1.07 ± 1.63, respectively, at the three time points [Table 3]. We found a significant increase in CI value intraoperatively as compared to preoperative value (P = 0.011). There was a further increase in CI postoperatively which was significantly higher than preoperative (P < 0.001) and intraoperative values (P = 0.001) [Table 4].

Table 4

Paired comparisons of Coagulation Indices at the three time points

CI	Mean	SE	t	P
Preoperative versus intraoperative	−0.35	0.824	−2.685	0.011
Preoperative versus postoperative	−0.74	1.096	−4.272	<0.001
Intraoperative versus postoperative	−0.39	0.605	−4.074	<0.001

CI=Coagulation Index, SE=Standard error

When the TEG parameters were correlated with CCT using Pearson correlation, we found significant correlation of R-time with PT (P < 0.001), INR (P < 0.001), and aPTT (P = 0.002) in intraoperative period. All other correlations were insignificant (P > 0.05) [Table 5].

Table 5

Correlation between conventional coagulation tests and thromboelastography parameters

Time	Parameter	R-time		K-time		α-angle		MA
		Pearson correlation	P	Pearson correlation	P	Pearson correlation	P	Pearson correlation	P
Preoperative	Platelet count	0.054	0.743	−0.101	0.537	−0.185	0.254	−0.152	0.348
	PT	−0.153	0.347	0.163	0.314	0.233	0.149	−0.005	0.977
	INR	−0.148	0.361	0.150	0.355	0.218	0.176	−0.023	0.889
	APTT	0.106	0.514	0.165	0.308	−0.234	0.147	−0.067	0.680
Intraoperative	Platelet count	−0.002	0.990	−0.076	0.643	−0.137	0.399	−0.273	0.088
	PT	0.715	<0.001	−0.054	0.739	0.025	0.878	0.215	0.183
	INR	0.694	<0.001	−0.043	0.793	0.014	0.934	0.227	0.160
	APTT	0.478	0.002	0.174	0.283	−0.163	0.316	0.080	0.625
Postoperative	Platelet count	0.069	0.670	−0.015	0.928	−0.217	0.178	−0.138	0.397
	PT	−0.092	0.573	0.052	0.748	−0.037	0.820	0.142	0.381
	INR	0.074	0.650	0.066	0.687	−0.063	0.699	0.187	0.247
	APTT	0.237	0.142	0.063	0.701	0.074	0.651	0.156	0.335

MA=Maximum amplitude, PT=Prothrombin time, APTT=Activated partial thromboplastin time, INR=International normalized ratio

None of the patients enrolled in this study had any evidence of intracranial hemorrhage in their postoperative CT scans or MRI. None of the patients developed deep-vein thrombosis (DVT) or pulmonary embolism (PE) during the study.

DISCUSSION

Patients with primary brain tumors are a major subset of patients requiring neurosurgical procedures.[6] Coagulation abnormalities have been described in literature to occur perioperatively in these patients.[7] The role of intrinsic tumor factors, mainly plasminogen activators, has been implicated and investigated.[89] However, the literature is not sufficient to provide a conclusive evidence on the type of coagulation abnormalities. Various studies describe hypercoagulability, hypocoagulability as well as normal coagulation in the perioperative period in this population.[1011] Hypocoagulability can lead to complications such as postoperative intracranial hematomas as well as hemorrhage outside the CNS, whereas hypercoagulability can lead to thromboembolic complications such as DVT and PE.[1213] As these disorders form complete opposite ends of the spectrum of coagulation-related complications occurring in the perioperative period, effective management requires accurate and prompt identification of these abnormalities across the spectrum of coagulation.

As traditional tests of coagulation have not been successful in identification of all types of coagulation abnormalities, viscoelastic tests of coagulation such as TEG have generated a lot of interest.[14] TEG has been used extensively in areas such as cardiac surgery, liver transplantation, and trauma, but its application in the field of neurosurgery and neurocritical care is limited.[1516171819]

This study was done to evaluate the usefulness of TEG in comparison with CCT in assessing hemostatic changes and bleeding in perioperative period in patients undergoing surgery for primary brain tumors. In this study, using TEG, we found a trend towards hypercoagulability in intraoperative and postoperative periods which was not evident on CCT and both showed poor correlation. Compared to CCT, TEG was a useful diagnostic tool to identify coagulation abnormalities in the perioperative period, and was useful in guiding treatment.

Hemostasis is the physiological process that stops bleeding at the site of an injury while maintaining normal blood flow elsewhere in the circulation. The process of hemostasis was reported using “Classical Cascade Theory,” which was first described in 1964 by Davie and Ratnoff.[20] According to this theory, sequential activation of proenzymes by proteases in the blood leads to formation of thrombin, which, in turn, breaks fibrinogen into fibrin. This theory divides coagulation cascade into two arms: the extrinsic pathway and the intrinsic pathway. Till recent times, this model of coagulation was widely accepted. However, some drawbacks have been pointed out in this description of coagulation.[21] In a newer study, a cell-based pathway for coagulation has been described, which proposes that tissue factor is expressed on the surface of cells outside the vascular system, such as smooth muscle cells and fibroblasts surrounding blood vessels. When the tissue factors are exposed to blood following injury to vascular endothelium, the coagulation pathway is initiated and coagulation proceeds through phases of initiation, amplification, propagation, and termination.[22]

With our better understanding about cell-based model of coagulation as opposed to separate intrinsic and extrinsic coagulation pathways, the focus is gradually shifting from CCT (which work in isolation) to viscoelastic point-of-care tests such as thromboelastography (TEG), rotational TEG, and Sonoclot which provide in vitro assessment of global coagulation in the whole blood sample. These viscoelastic tests provide a better picture of in vivo hemostasis due to the interaction of all cellular and plasmatic components.[22]

Singh et al. were one of the first to study the coagulation abnormalities in brain tumors.[23] They evaluated the coagulation abnormalities in 25 patients with brain tumors perioperatively, using traditional coagulation parameters such as platelet count, aPTT, PT, thrombin time, and fibrinogen, fibrin degradation products and compared them with controls undergoing nonneurological surgeries. They found elevated platelet counts postoperatively. They found coagulation abnormalities in 40% of patients preoperatively, the most common ones being compensated DIC, fibrinolysis, and a hypercoagulable state. In contrast, we did not find coagulation abnormalities preoperatively in our patients. In their study, patients became more hypocoagulable during surgery, with compensated DIC becoming acute DIC and the hypercoagulability getting compensated or over-compensated. These findings are not in concurrence with the findings of our study as we found a trend toward hypercoagulability as reflected by TEG parameters, especially R and K values. However, the changes in TEG parameters did not correlate with the values of CCT.

Another study done by Prasad et al.[7] using similar coagulation parameters found coagulation abnormalities preoperatively as well as intraoperatively, though most of the abnormalities were hypocoagulable in nature. Coagulation profile showed a significant reduction in the intraoperative and postoperative platelet counts compared to preoperative values. However, the platelet counts were within the normal range throughout the study. Significant changes were also found in PT, INR, aPTT, Hb, and TLC in the intraoperative and postoperative periods compared to preoperative values. However, the values were within the normal range. Their findings are not in agreement with our study in which we found a trend towards hypercoagulability in intraoperative and postoperative period using TEG. These differences could be because the traditional tests of coagulation used in these studies may not have been able to identify a hypercoagulable state effectively. Even in our study, the conventional tests remained within the normal limit except for a slight prolongation in PT, INR, and APTT value intraoperatively.

In a study by Goh et al., TEG was done in 50 patients undergoing surgery for primary brain tumor on the day before surgery, intraoperatively, and 24 h after surgery.[6] The study of trends showed a tendency toward hypercoagulability, with a reduction in R- and K-times, and increase in α-angle and MA 24 h postoperatively. Our findings are in concurrence with the findings of this study, as our study also demonstrated a trend towards hypercoagulability in the postoperative period as evident by reduction in R-and K-times, and increase in α-angle, MA, and CI values. They suggested that TEG is useful both in the preoperative assessment of patients with large brain tumors and in the intraoperative stage for monitoring the hemostatic profile.

Verma and Hemlata[24] had proposed that the TEG is useful for a rapid assessment of hemostasis and for goal-directed transfusion/drug therapy in a variety of clinical settings but not a substitute for conventional coagulation assays; rather, it complements the results of CCTs. According to them, there are compelling reasons for this tool to be adopted by anesthesiologists in the perioperative period as this can help to monitor hemostasis and to guide specific hemotherapy along with appropriate dosages of blood components based on TEG results in varied clinical settings.

Ahmad et al. in their study found that the TEG parameters have a good correlation with CCT and also showed excellent independent predictive efficacy for prediction of hypocoagulation.[25] In their study, PT, aPTT, and TT were directly proportional to R-time and K-time and inversely proportional to α-angle (P < 0.001). Platelet count showed a strong positive correlation with MA (P < 0.001). PT, aPTT, and TT showed a strong and significant positive correlation with R-time and K-time (r > 0.7), a strong inverse correlation with α-angle (r < -0.80), and a poor correlation with MA. Platelet count showed a weak correlation with R-time, K-time, and α-angle (r < 0.3) and a strong positive correlation with MA (r = 0.83). These findings are not in agreement with our study in which we found a poor correlation between the TEG parameters and CCT except for a significant correlation between R-time and the values of PT, APTT, and INR in intraoperative period. This might be because of the fact that they had studied a different patient population, i.e., pre-eclampsia/eclampsia patients, as against brain tumor patients in our study.

Abrahams et al. studied TEG parameters in 46 neurosurgical patients, including brain tumors, aneurysm surgery, and spine surgery, before, during, and after surgery.[26] The traditional parameters were normal, as we found in our study. In their study, the CI value suggested a hypocoagulable picture in the preoperative period, which increased in the postoperative period, suggesting a hypercoagulable picture. This hypercoagulable picture was not only seen in patients with brain tumors but also those with SAH and spine surgery, although the craniotomy patients appeared to be more hypercoagulable, with a higher CI. Therefore, it is not clear whether the hypercoagulability seen could only be attributed to brain tumors, or it could be due to other factors common to most neurosurgical procedures, such as long duration of surgery, immobilization, and stress response of surgery. We also found a trend toward hypercoagulability in the intraoperative and postoperative period in our study, as evident by changes in various TEG parameters (R-time, K-time, α-angle, MA, and CI). We did not evaluate patients with any other type of neurosurgical procedure. Therefore, the possibility that the other surgical factors mentioned above could influence the postoperative coagulation profile cannot be ruled out.

One of the limitations of our study was the sample size which was not large enough to evaluate the outcomes of coagulation abnormalities in terms of bleeding or thromboembolic complications such as DVT and PE. Outcomes of coagulation abnormalities are more important clinically than the mere presence of these abnormalities. Therefore, more studies are required to investigate this relationship. Another limitation was that the tests were done only up to 24 h postoperatively. As some of the studies done earlier have shown that hypercoagulability may continue much later into the postoperative period, we might not have been able to detect thromboembolic complications which might have developed later on. The impact of blood product transfusions on the TEG parameters, especially in the postoperative period, is unclear, as each of these blood products can directly affect the TEG parameters, especially R- and K-time, which represent the function of coagulation factors.

In our study, we found a trend towards hyper coagulability in intraoperative and postoperative period using TEG. The results of intraoperative CCTs were available only after the surgery and did not show a good correlation with TEG parameters. Blood products (PRBC and FFP) transfusion in the intraoperative period was based on TEG findings and resulted in good outcomes in all our patients.

CONCLUSIONS

It was concluded that TEG is a useful diagnostic tool to identify coagulation abnormalities in the perioperative period in patients with primary brain tumors, and may be useful in guiding treatment in the perioperative period. From our study, it is not clear when the coagulation status returns to normal in the postoperative period. Further studies are needed to address this issue. We would like to comment that TEG compliments the results of routine coagulation tests, providing reliable results immediately, and is an indispensable tool in assessing intraoperative blood loss and guiding management accordingly.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.