Increased Risk for Postoperative Hemorrhage After Intracranial Surgery in Patients With Decreased Factor XIII Activity
Implications of a Prospective Study
Background and Purpose— The functional integrity of the hemostatic system is a prerequisite for the safe performance of neurosurgical procedures. To monitor the individual coagulation capacity of each patient, standard tests are effective to detect deficiencies involving the generation of fibrin. However, fibrin clot strength depends primarily on coagulation factor XIII, which cross-links fibrin monomers and enhances clot resistance against fibrinolysis. Therefore, factor XIII is functionally involved in both the hemostatic and fibrinolytic systems. The objective of this prospective study was to determine the incidence and clinical relevance of perioperative decreased factor XIII with respect to standard coagulation parameters and the occurrence of postoperative hematoma.
Methods— In 876 patients, 910 neurosurgical procedures were performed. Prothrombin time (PT), partial thromboplastin time (PTT), platelet count, fibrinogen, and factor XIII were tested in each patient preoperatively and postoperatively.
Results— Postoperative intracranial hematoma (defined as requiring surgical evacuation) occurred after 39 (4.3%) of 910 surgical procedures. Patients with postoperative hematoma had significantly lower factor XIII and fibrinogen levels preoperatively and postoperatively than patients without hematoma. In patients with postoperative hematoma, PT and platelets differed significantly only postoperatively, whereas PTT was different neither preoperatively nor postoperatively. Of the 39 patients with a postoperative hematoma, 13 (33.3%) had a postoperative factor XIII <60% compared with 61 (7%) of 867 patients without hematoma (P<0.01, Fisher’s exact test). The relative risk of developing a postoperative hematoma is therefore increased 6.4-fold in patients with postoperative factor XIII <60%. The risk is increased 12-fold in patients who additionally have postoperative decreased fibrinogen levels (<1.5 g/L) and 9-fold in patients with platelet count <150×109/L and factor XIII <60%.
Conclusions— This is the first prospective study that demonstrates the association of decreased perioperative factor XIII with an increased risk of postoperative hematoma in neurosurgical patients. The risk is further increased in those patients with low factor XIII and additional abnormalities of fibrinogen, PT, platelets, and PTT. Factor XIII testing and specific replacement, as accepted for other clotting factors, may reduce the risk of postoperative hematoma.
Postoperative hematomas are a common complication of surgical procedures. However, this is rather dramatic after brain surgery and is frequently associated with severe neurological impairment or death of neurosurgical patients.1 Changes in hemostasis during intracranial surgery are miscellaneous and vary from disseminated intravascular coagulation2–4⇓⇓ to deep venous thrombosis.5,6⇓ These changes were attributed to alterations primarily caused by brain neoplasms7,8⇓ or to the influence of intracranial surgery itself.9 Although meticulous surgical hemostasis is of paramount importance to prevent a subsequent hematoma, impaired activation of coagulation and/or increased fibrinolytic activity may nevertheless cause postoperative bleeding. Standard coagulation tests including partial thromboplastin time (PTT), prothrombin time (PT), fibrinogen, and platelet count measure important parameters to monitor the perioperative hemostatic capacity in relation to the formation of fibrin clots. However, fibrin clot formation relies primarily on the activity of coagulation factor XIII, which cross-links fibrin monomers into stable polymers. Therefore, factor XIII is not only involved in the final step of the clotting process but also counteracts fibrinolytic degradation due to concentration-dependent fibrin clot stabilization.10–12⇓⇓ Thus, decreased enzymatic activity of factor XIII was discussed as a cause of postoperative hematoma in abdominal, gynecological, 13 plastic, and urological surgery14 and also recently in the field of neurosurgery.15 The objective of this prospective study was to determine the incidence and clinical relevance of perioperative decreased factor XIII with respect to standard coagulation parameters and the occurrence of postoperative hematoma requiring surgical evacuation in a large cohort of neurosurgical patients.
Subjects and Methods
In 876 patients, 910 intracranial neurosurgical procedures were performed between June 1997 and December 2000 (for details, see Table 1). This prospective study included 431 neurosurgical procedures in women (47.4%) and 479 neurosurgical procedures in men (52.6%). Mean ages were 53.6±15.4 years in women and 50.5±15.9 years in men. All surgical procedures were performed with patients under general anesthesia. The postoperative management was according to the surgically treated disease, which included a 24-hour observation period in the intensive care unit for patients with vascular and tumor surgery. Patients with shunt procedures and biopsies were followed in the intermediate care unit.
Patients were observed clinically, and if deterioration was seen, a postoperative cranial CT scan was done. A postoperative intracranial hematoma was defined as significant if surgical treatment was necessary because of relevant space occupation and/or significant neurological deterioration.
Standard coagulation tests including platelet count, PTT, PT, and fibrinogen were performed routinely preoperatively. Postoperative tests of these parameters were performed depending on the patient’s clinical course on the same day (patients with tumor and vascular surgery) and/or the first postoperative day. If necessary, testing was continued each day on the intensive care unit until referral to the ward. Additionally, intraoperative surveillance of these parameters was performed during surgery for patients with brain tumor and vascular surgery. PT >60% and fibrinogen >1.5 g/L were considered sufficient for hemostasis. If deficiencies were detected, the prothrombin complex concentrate and fibrinogen were administered. Prothrombin complex concentrate was substituted in 129 patients (14.2% of 910 procedures), and 34 patients (3.7% of 910 procedures) received fibrinogen replacement. Fresh frozen plasma was administered in 21 patients (2.3% of 910 procedures) intraoperatively and in 42 patients (4.6% of 910 procedures) postoperatively.
The factor XIII test was performed on admission and the day after surgery (790 patients), the second day after surgery (67 patients), and the third day after surgery (58 patients). In 4 patients with postoperative hematoma occurring the same day, factor XIII was tested in only 1 patient before hematoma evacuation but in the remaining 3 patients on the next morning.
To prevent deep venous thrombosis, all patients were treated with fractionated low-dose heparin (5000 U every 8 hours subcutaneously) starting within 24 hours after surgery plus intraoperative and postoperative compression stockings. This has been found to be safe and effective.16 Heparin administration was interrupted when PTT was increased >40 seconds.
The Sarstedt Monovette system was used for all blood sampling procedures. Samples (3 mL) used for all coagulation assays, including the factor XIII test, contained 0.3 mL of citrate solution. Samples were transferred immediately for further diagnostic processing. Before any of the tests were performed, the samples were centrifuged for 20 minutes at 3500 U/min at room temperature. Factor XIII assay was performed immediately, and samples were stored at −70°C.
Platelets were automatically counted according to standard procedures after blood was sampled in blood collection systems containing EDTA.
To characterize abnormalities in the fibrinolytic system, plasminogen, plasmin-antiplasmin (PAP) complex, and α 2-antiplasmin testing was additionally performed in 18 blood samples (46.2%) of patients with postoperative hematomas. Therefore, the samples that were obtained immediately before reoperation and stored at − 70°C were used.
All laboratory tests were performed according to the protocol of the manufacturer. For measurement of plasma factor XIII, a photometric test of Dade Behring (Berichrom FXIII) was used. This test system is not affected by postoperative heparin administration and is a validated procedure. The low number of platelets remaining in the plasma samples after centrifugation did not influence the accuracy of the test. Plasminogen and α2-antiplasmin were tested with the Berichrom plasminogen assay and Berichrom α2-antiplasmin assay, respectively. PAP complex was tested with the use of the Dade Behring Enzygnost PAP assay.
Statistical analysis was performed with the use of commercially available software (SPSS Inc, version 10.0). Data of all parameters were given as mean±SD. To compare independent preoperative and postoperative parameters, the unpaired Student’s t test was used. Probability values <0.05 were considered significant. Parameters compared with the t test were proven to be normally distributed.
Fisher’s exact test was used to compare the frequencies of postoperative hematoma and factor XIII, fibrinogen, platelets, PT, and PTT. Therefore, it was necessary to define cutoff values, which are given individually for each parameter in Results. The cutoff values for each of these parameters preoperatively and postoperatively were also used to calculate the relative risk of developing a postoperative hematoma. For all postoperative parameters, the lowest measured factor XIII, fibrinogen, platelet count, and PT and the highest PTT were used for statistical calculation up to the detection of a hematoma in this group and during the postoperative course in the group of patients without hematoma, respectively.
Preoperative and postoperative factor XIII, fibrinogen, platelet count, PT, and PTT were entered into a multivariate discriminant analysis (stepwise inclusion, Wilks λ method [probability for inclusion=0.05 and for exclusion=0.10]). A total of 760 (83.5%) of the 910 surgical procedures were included; 180 (19.8%) were excluded because of at least 1 missing parameter.
Logistic regression analysis (forward inclusion, Wald method) was performed to determine predictive factors for the occurrence of postoperative hematoma.
In this prospective study, 910 intracranial operations were performed in 876 patients (Table 1). In 39 of the procedures (4.3%), patients developed a postoperative intracranial hematoma, which was removed surgically. The location of the hematoma was epidural in 22 (56.4%), subdural in 10 (25.6%), and intracerebral in 7 (18%) cases. The hematoma was evacuated before heparin administration on the day of previous surgery in 4 (10.3%) of the 39 patients. All other patients underwent reoperation after the anticoagulation therapy was started (12 patients on day 1; 4 on day 2; 9 on day 3; 4 on day 4; 3 on day 5; and the remaining patients on days 6, 7, and 9).
Perioperative Changes of Factor XIII and Standard Coagulation Parameters
In the group of patients with postoperative hematoma, factor XIII was significantly lower preoperatively (97.89±24.98% versus 113.66± 27.31%) and postoperatively (71.45±20.57% versus 93.46±24.37%; P<0.01, t test) (Table 2) compared with the group without hematoma. As shown for factor XIII, fibrinogen also differed statistically significant preoperatively (2.92±1.22 versus 3.43± 1.17 g/L; P<0.01, t test) and postoperatively (2.07±0.93 versus 2.53±0.99 g/L; P<0.01, t test), respectively (Table 2). No preoperative significant differences were found for platelet count and PT. Postoperative platelet count was significantly lower in patients with hematoma (150.5±53.7×109/L) than in patients who did not suffer from rebleeding after surgery (185.6±68.3×109/L; P<0.01) (Table 2). PT was also significantly lower in the hematoma group (71.7±11.6%) than in the group without hematoma (76.4±12.6%; P<0.05) (Table 2). No difference was seen for PTT either preoperatively or postoperatively in either group.
The overall decrease in postoperative factor XIII, fibrinogen, platelets, and PT is likely to be due to perioperative blood loss or consumption. Therefore, to calculate the relative risk it seems rational to define a higher threshold of preoperative parameters that might be critical for postoperative hematoma formation than we have used for calculation of the same parameters postoperatively. For factor XIII, the average decrease in all patients is from 112.99±27.38% preoperatively to 92.50±24.66% postoperatively (approximately 18.1%), regardless of any postoperative bleeding complication. A preoperative factor XIII <80% was significantly associated with a postoperative hematoma (P< 0.01, Fisher’s exact test) (Table 3). Therefore, the risk of developing a postoperative hematoma is increased 3.9-fold in those patients. In contrast, a preoperative PT <80% was not associated significantly with hematoma formation. A preoperative fibrinogen <3.0 g/L also had a significant association with postoperative hematoma formation (P<0.01, Fisher’s exact test), resulting in a 2.9-fold increased risk for those patients (Table 3). Patients with platelet count >200×109/L preoperatively were not at risk of developing a postoperative hematoma.
Of the 39 patients with postoperative hematoma, 13 (33%) had a postoperative factor XIII <60% compared with 61 (7%) of 867 patients without hematoma. This difference was statistically significant (P<0.01, Fisher’s exact test). Therefore, those patients with a postoperative factor XIII <60% have 6.4-fold increased risk of developing a postoperative hematoma. Furthermore, patients with postoperative platelet count <150×109/L (P<0.01) and fibrinogen <1.5 g/L were at a 3.0- and 2.5-fold increased risk of developing a postoperative hematoma, respectively (P<0.01, Fisher’s exact test) (Table 4). In this series, no statistically significant association was found between decreased postoperative PT or prolonged PTT and the occurrence of postoperative hematoma.
Postoperative Hematoma and Complex Hemostatic Impairment
After hours of surgery, complex coagulation derangement may occur, and therefore it is important to know to what extent those abnormalities increase the overall risk of rebleeding. Postoperative factor XIII <60% and fibrinogen <1.5 g/L not only showed a significant association with postoperative hematoma formation, as it did for each of the factors alone, but increased the relative risk of developing a hematoma to 12-fold. Patients with postoperative platelet counts <150×109/L and factor XIII <60% were also at a significantly higher (9-fold) risk of having postoperative bleeding complications. Patients with postoperative decreased factor XIII and PT (both <60%) also had a 6-fold increased risk. Factor XIII <60% and PTT >35 seconds were associated with an almost 5-fold increased risk for postoperative hematoma (Table 5).
Postoperative Factor XIII and Fibrinolysis
In a subgroup of 18 (46.2%) of the patients with a postoperative hematoma, plasminogen, α2-antiplasmin, and PAP were tested. Patients with postoperative hematoma and postoperative factor XIII < 70% had a lower plasminogen (85.6±23.56% versus 108±12.7%; P<0.05) and α 2-antiplasmin level (87.9±33.1% versus 117.6±14.8%; P<0.05, t test) than patients with postoperative hematoma and factor XIII >70%. Because PAP is a specific marker of fibrinolysis, it was chosen instead of the more global fibrinolytic marker D-dimer. Consistent with increased fibrinolytic activity, PAP was higher (1417.5±1793 versus 586.57±452.9 μg/L) in the group of patients with factor XIII <70% but showed a high variation. Although statistically not significant, elevated PAP levels in patients with decreased factor XIII and postoperative hematoma indicate increased fibrinolytic activity.
In a multivariate discriminant analysis of preoperative and postoperative factor XIII, fibrinogen, platelets, PT, and PTT, only postoperative factor XIII was included. Correct discrimination was 65.7% with an exact assignment to the group of patients with or without postoperative hematoma. In a logistic regression analysis, none of the parameters (preoperative and postoperative factor XIII, fibrinogen, platelets, PT, and PTT) was identified as an independent predictive factor for the development of a postoperative hematoma.
Perioperative Factor XIII Substitution
Because this study was not considered to prove the efficacy of normal plasma factor XIII concentration in preventing postoperative hematoma, factor XIII was replaced in only 23 patients. In 7 of these 23 patients, substitution was performed after hematoma evacuation to prevent a second rebleeding. Ten patients with postoperatively decreased factor XIII (<60%) received factor XIII replacement, and no postoperative hematoma occurred. Six patients with preoperative factor XIII <60% underwent substitution therapy; the postoperative plasma level remained >60%, and no hematoma occurred.
A postoperative hematoma is a potential life-threatening complication after intracranial surgery and is frequently associated with poor outcome and death of neurosurgical patients.1 Therefore, it is of vital importance to identify risk factors that can be avoided or specifically treated. In addition to coagulopathies, disseminated intravascular coagulation2–4⇓⇓ and drugs interfering with platelet function1,17,18⇓⇓ are identified as risk factors for bleeding complications. However, most of these abnormalities can be detected by routine coagulation tests or medical history. In this series of 39 postoperative hematomas, 1 patient was on antiplatelet medication, another patient was treated with oral anticoagulants, and a third patient had a severe liver dysfunction resulting in hemostatic impairment, which was specifically treated and controlled preoperatively.
Perioperative Coagulation Parameters and Risk of Postoperative Hematoma
Preoperative factor XIII and fibrinogen were significantly lower in the group of patients with postoperative hematoma than in patients without postoperative hematoma. Therefore, in these patients the presence of a hemostatic impairment to a certain degree can be assumed, which was clinically not relevant before operation. Postoperative factor XIII, fibrinogen, platelets, and PT were significantly lower in patients with hematoma, which may have amplified any preoperative impairment. Lower platelet count indicates diminished platelet adherence at damaged vessels, decreased generation of thrombin, and low platelet-associated factor XIII, which promotes clot stabilization and platelet interaction. Platelet-associated factor XIIIa increases the amount of cross-linking in a fibrin clot, thereby contributing to aging of the clot and reduction in the degree of platelet binding.19
Furthermore, the importance of platelet-mediated thrombin generation has recently been reported,20–22⇓⇓ and it was demonstrated that administration of recombinant activated factor FVII may compensate for a lowered platelet count with regard to thrombin generation.23 Decreased generation of thrombin, however, may also interfere with factor XIII activation and fibrin cross-linking. This may explain the exponential risk of postoperative hematoma in patients with combined decreased factor XIII, fibrinogen, platelets, and PT and increased PTT postoperatively. The correlation of fibrin clot stability with factor XIII concentration has been demonstrated in vitro12,24⇓ and was also of clinical relevance in patients who underwent cardiopulmonary bypass surgery.25 Therefore, the decrease of fibrinogen, the precursor of fibrin, which is the main substrate for factor XIII, as well as low plasma factor XIII increases the risk of postoperative hematoma because of a decreased number of cross-linked fibrin clots.
In the literature, the minimum factor XIII level that is necessary to achieve sufficient hemostasis ranges from a low level of 2%26 to moderately decreased levels of 30% to 50%.27,28⇓ Thus, the level of preoperative and postoperative factor XIII that is sufficient to achieve safe hemostasis after brain surgery is unclear. Whether a patient with postoperative decreased factor XIII develops a postoperative hematoma might be determined to a certain extent by normal levels of fibrinogen and platelets. Therefore, particularly those patients with factor XIII <60% and low fibrinogen or decreased platelets are at high risk of developing a postoperative hematoma. However, until now no report has described a correlation between plasma factor XIII and factor XIII activity in the wound bed, where fibrin cross-linking also occurs. Interestingly, it was shown in a rat model of bilateral frontal craniotomy and corticotomy that the local application of factor XIII significantly reduced the occurrence of postoperative hematoma.29
Consistent with the findings of a retrospective study,15 patients with postoperative factor XIII <60% had a 6.4-fold increased risk of developing a postoperative hematoma in this prospective study.
Factor XIII polymorphism may also be a cause of decreased plasma concentration. Recently, in a few studies a slightly higher incidence of factor XIII Val 34 Leu polymorphism in patients with primary intracerebral hemorrhage was related to impaired cross-linking of fibrin and/or coagulation proteins.30–32⇓⇓ However, in this study factor XIII polymorphism was not assessed, and no factor XIII genomic analysis was performed.
Factor XIII and Fibrinolytic Activity
Hyperfibrinolysis was reported to be associated with postoperative hematoma in patients who underwent surgery for brain tumors. 7,33⇓ Clot resistance to fibrinolysis is enhanced by platelet factor XIIIa–mediated cross-linking of α 2-antiplasmin to fibrin. 19 At low factor XIII concentrations, only γ-chain cross-linking of fibrin monomers occurs, while α-chain cross-linking is dependent on high factor XIII concentrations. 24
The almost 70% correct classification of patients with postoperative hematoma due to postoperative factor XIII concentration, as the only included parameter in a multivariate discrimination analysis, emphasizes the importance of factor XIII interaction with the hemostatic and fibrinolytic systems. Therefore, it would be likely to find abnormalities of the fibrinolytic system in patients with low factor XIII. Indeed, PAP complex testing in patients with postoperative hematoma revealed an almost 2-fold higher level and a significantly lower plasminogen and α2-antiplasmin level in those patients with decreased factor XIII. Although the increase of PAP was not statistically significant in patients with postoperative hematoma and low factor XIII compared with normal factor XIII, the elevated PAP demonstrates increased fibrinolytic activity in patients with low factor XIII. Moreover, increased plasmin-mediated fibrinolysis leads to a decrease of α 2-antiplasmin due to PAP formation and therefore to a decrease of α2-antiplasmin incorporation into the clot, which may have an additional effect on clot destabilization. Fujii et al9 found a parallel decrease of plasminogen and hematocrit, which could be a possible explanation for the significantly decreased plasminogen in the group of patients with hematoma and postoperative decreased factor XIII. However, no significant difference in hematocrit was seen in patients with postoperative hematoma and factor XIII <60% and >60% in this study (data not shown). A strong correlation between factor XIII, fibrinogen, and plasminogen was found in patients of different races,34 which may explain the significantly decreased plasminogen in patients with low factor XIII in this study, although such correlation has not yet been described for the European population. An impaired hepatic synthesis of factor XIII and plasminogen would also explain such a correlation.
Postoperative decreased PT alone was not associated with hematoma in this series. Because this is a known risk factor during the postoperative course, administration of fresh frozen plasma and prothrombin complex concentrates was commonly performed if the decrease was obvious by standard testing. None of the patients who received factor XIII, either preoperatively or postoperatively, developed a postoperative hematoma, and the patients who were underwent substitution after hematoma evacuation did not suffer from rebleeding again. Postoperative PTT was not associated with hematoma formation.
Although the incidence of postoperative hematoma of 4.3% in this series seems high, it must be considered that all patients were included prospectively. Therefore, the rate is consistent with other reports in which the occurrence of postoperative hematoma was between 1% and 12%. 1,7⇓ The patients underwent major surgical procedures for a variety of different pathologies (Table 1). As demonstrated in other studies,1 the incidence of hematoma formation was highest after meningioma surgery (30.8%); this was followed by surgery for brain injury (23.1%).35
Beside fibrin clot stabilization, factor XIII has recently been found to reduce endothelial permeability independently of its role in the coagulation cascade.36 Thus, decreased factor XIII may also influence the blood-brain barrier, especially in patients with brain neoplasm with respect to brain edema or postoperative bleeding complication. In this series >62% of all patients with postoperative decreased factor XIII (<60%) underwent surgery to treat meningiomas (14.5%), glioblastomas (18.4%), metastasis (15.8%), and severe head injuries (18.4%), which are conditions with at least a focal blood-brain barrier disruption. However, the correlation of plasma factor XIII with the extent of brain edema needs to be addressed in more detail in another study, using a specific protocol of preoperative MR image acquisition.
This is the first prospective study that emphasizes the importance of plasma factor XIII for sufficient fibrin clot stabilization after intracranial surgery. The results indicate that decreased factor XIII is currently underestimated and is associated with postoperative hematoma formation. In particular, those patients with decreased factor XIII, fibrinogen, and platelets are at high risk of developing a postoperative hematoma. Therefore, factor XIII testing and specific replacement may reduce the risk of postoperative hematoma.
- Received November 26, 2001.
- Revision received January 31, 2002.
- Accepted March 7, 2002.
- ↵Hamilton MG, Hull RD, Pineo GF. Venous thromboembolism in neurosurgery and neurology patients: a review. Neurosurgery. 1994; 34: 280–296;comment 296.
- ↵Goh KY, Tsoi WC, Feng CS, Wickham N, Poon WS. Haemostatic changes during surgery for primary brain tumours. J Neurol Neurosurg Psychiatry. 1997; 63: 334–338.
- ↵Robinson BR, Houng AK, Reed GL. Catalytic life of activated factor XIII in thrombi: implications for fibrinolytic resistance and thrombus aging. Circulation. 2000; 102: 1151–1157.
- ↵Francis CW, Marder VJ. Increased resistance to plasmic degradation of fibrin with highly crosslinked alpha-polymer chains formed at high factor XIII concentrations. Blood. 1988; 71: 1361–1365.
- ↵Burchhardt P, Tilsner V, Hupe W. Faktor XIII-Mangel als Ursache postoperativer Nachblutungen. Aktuelle Urologie. 1978; 9: 29–33.
- ↵Holden MP. Dangers of aspirin before cardiac surgery. BMJ. 1992; 305: 365–366.
- ↵Duckert F, Jung E, Shmerling DH. A hitherto undescribed congenital haemorrhagic diathesis probably due to fibrin stabilizing factor deficiency. Thromb Diath Haemorrh. 1961; 5: 179–186.
- ↵Egbring R, Seitz R, Gürten GV, Köther M, Barthels M, Fuchs G, Lerch L, Kröninger A. Bleeding complications in heterozygotes with congenital factor XIII deficiency.In: Mosessen MW, ed. Fibrinogen 3. Amsterdam, Netherlands: Elsevier; 1988: 314–346.
- ↵Egbring R, Kailing A, Kehl HG, Seitz R, Maasberg M, Holst F, Fuchs G, Wallin R, Saldeen T. Disorders with severe acquired factor XIII deficiency: lack of synthesis or increased consumption (DIC): efficacy of factor XIII replacement in bleeding complications.In: McDonagh J, Seitz R, Egbring R, eds. International Conference on Factor Thirteen. Stuttgart, Germany: Schattauer; 1993: 216–240.
- ↵Laohaprasit V, Edwards MW, Mayberg MR. Prevention of postoperative intracerebral hemorrhage with topical recombinant factor XIII in the rat. Neurosurgery. 1993; 32: 630–633;comment 633–634.
- ↵Catto AJ, Kohler HP, Bannan S, Stickland M, Carter A, Grant PJ. Factor XIII Val 34 Leu: a novel association with primary intracerebral hemorrhage. Stroke. 1998; 29: 813–816.
- ↵Gemmati D, Serino ML, Ongaro A, Tognazzo S, Moratelli S, Resca R, Moretti M, Scapoli GL. A common mutation in the gene for coagulation factor XIII-A (VAL34Leu): a risk factor for primary intracerebral hemorrhage is protective against atherothrombotic diseases. Am J Hematol. 2001; 67: 183–188.
- ↵Corral J, Iniesta JA, Gonzalez-Conejero R, Villalon M, Vicente V. Polymorphisms of clotting factors modify the risk for primary intracranial hemorrhage. Blood. 2001; 97: 2979–2982.
- ↵Palmer JD, Francis DA, Roath OS, Francis JL, Iannotti F. Hyperfibrinolysis during intracranial surgery: effect of high dose aprotinin. J Neurol Neurosurg Psychiatry. 1995; 58: 104–106.
- ↵Noll T, Wozniak G, McCarson K, Hajimohammad A, Metzner HJ, Inserte J, Kummer W, Hehrlein FW, Piper HM. Effect of factor XIII on endothelial barrier function. J Exp Med. 1999; 189: 1373–1382.