Introduction

Acute intracerebral hemorrhage (ICH) can result in irreversible brain damage and impair long-term functional status of the patient. Primary intracerebral hemorrhage constitutes 78–88% of all cases of ICH and is caused by “rupture of small damaged vessels” usually due to arterial hypertension and cerebral amyloid angiopathy. Secondary ICH responsible for about 15% of the total ICH is seen in patients with cerebrovascular malformations (AVM’s, aneurysm, dural venous sinus thrombosis, AVF, vasculopathy, moyamoya), abnormal hemostasis due to anticoagulant use or thrombolytic therapy, and cerebral amyloid angiopathy. CTA spot sign can provide the necessary roadmap to identify the extent of microvascular damage associated with ICH and identify high risk cases that can be then urgently triaged for more preventive interventions such as with PCC. Acute hemorrhagic stroke is one of the most fatal forms of stroke, affecting 15% of the population worldwide with a 30-day mortality of up to 50% [1, 2]. Approximately, 50% of the cases that ultimately prove lethal, death occur within first 48 h of symptom onset [2]. For those who survive, only 20% are able to live independently at 6 months [2].

Acute ICH can present with neurosurgical emergency, requiring urgent surgical intervention. According to American Heart Association (AHA) guidelines, patients with suspected ICH should undergo non-contrast head CT (NCT) to evaluate the extent and etiology of ICH and CT angiography (CTA) to analyze the underlying vasculature [2]. CTA is also crucial in assessing the risk of hematoma expansion manifested by a positive CTA spot sign [3]. According to Demchuk et al. 2012, spot sign was defined as foci of contrast enhancement within an acute primary parenchymal hematoma, secondary to contrast extravasation, visible on the source images of CTA [3]. A positive spot sign found on CTA was shown to have a prognostic factor that has a positive predictive value (PPV) of 61% for hematoma expansion [4]. This can be used as a selection tool for treatment of patients with acute ICH who are at considerable risk for rapid hematoma expansion. CTA spot sign is an independent predictor of hematoma growth [5].

Baseline hematoma volume, location, and hematoma expansion during hospital stay have been shown to affect clinical outcomes and mortality [4]. Hematoma expansion which occurs in 40% of the patients has been explained by the “avalanche model” which attributes this growth to mechanical shearing of neighboring vessels secondary to growth of the initial hemorrhage [6]. This model is supported by multiple spot signs within one hematoma site suggesting several bleeding vessels secondary to mechanical shearing as opposed to one persistently bleeding vessel. Numerous studies in the past have highlighted the relationship between hematoma growth, poor clinical outcomes, and mortality [7, 8]; a 3-ml increase in hematoma volume has a PPV of 70% for a poor functional outcome such as disability or death [8]. Hence, hematoma expansion being the only modifiable predictor of clinical outcome appears to be an appropriate target for medical therapy. Serial computed tomography (CT) scanning has shown that 30–50% of the primary ICH will expand between the 6th and 12th hour [6, 9]. This fact implores one to use an agent which will act quickly and efficiently.

To date, there exists no definitive noninvasive treatment for acute ICH. In the past, several procoagulants such as fresh frozen plasma (FFP), factor 7, and prothrombin complex concentrate (PCC) have been studied to assess their role for this task. Prior clinical trials on recombinant factor 7 (rFVII) demonstrated a decrease in hematoma expansion within 4 h of ICH onset, but failed to decrease long-term mortality or improve functional outcome [10]. This may be secondary to known thromboembolic complications associated with rFVII and the fact that its short half-life may require multiple dosing for full effect [11]. As far as FFP is concerned, the large volume and long duration of onset make it a less suitable option particularly in cardiac or renal failure patients [10].

PCC is a promising noninvasive solution that can contain ICH expansion in critical patients. PCC is a procoagulant containing factors II, VII, IX, X, Protein C, and Protein S. The main indication for PCC is the urgent reversal of over-anticoagulation with warfarin [12]. PCC was originally developed in 1965 primarily for treatment of Hemophilia B, but now the main utility of PCC is as a replacement therapy for congenital or acquired deficiency of vitamin K-dependent clotting factors [13, 14]. Over the years, multiple variations of PCC have been developed, the latest being PCC4, which has slightly higher amount of FVII when compared to PCC-3. Several studies have shown superior efficacy of PCC compared to other procoagulants (such as FFP and rFVIIa) in quick reversal and prevention of hematoma expansion in patients on warfarin. PCC extrinsically replenishes the clotting factors blocked by warfarin [15], thereby reversing its effects [12]. Main advantages of PCC are smaller fluid volumes (unlike FFP), prompt administration, rapid reversal of INR <1.5 within 30 min of administration and better clinical outcome when compared to rFVII [13]. Long half-life of PCC translates into administration of a single dose to achieve the target goal of containment of hematoma expansion, thereby limiting the associated thromboembolic side effects [11, 13] unlike rFVII, which requires multiple dosing. Since ICH is a widespread issue affecting 15% of the world population, our study investigates role of PCC in all patients with acute ICH with active bleeding as evident by positive CTA spot sign, regardless of warfarin status.

Materials and methods

We retrospectively reviewed patients with acute ICH at our State Designated Stroke Center and Level I Trauma Center from Nov 2013 to Dec 2015. Approval from the Institutional Review Board was obtained and hospital protocols were followed. The hospital registry recorded all patients who were admitted with a diagnosis of ICH during this period and accounted to a total of 85 patients. Upon admission, all patients with the diagnosis of ICH received an initial non-contrast head CT. However, the timing of when the hemorrhage occurred and when the patient came to the emergency department is variable. Two independent reading radiologists were double blinded to the diagnosis and study group (control vs experimental) and accessed the patient’s charts through the medical record number. The blinded readers sequentially assessed all patients with ICH and isolated those with positive CTA spot sign. All of those patients with positive CTA spot sign were then followed up to evaluate percentage change in their hematoma size from the initial study without the knowledge of which patients received PCC (were experimental group) and which patients did not receive PCC (control group). Out of a total of 85 patients with ICH, 23 received PCC and from that group, there were 12 with positive CTA spot sign, which formed our experimental group. However, four of these patients had to be later eliminated since two of them underwent craniectomy urgently and two demised, leaving our net experimental group to eight (E = 8). From the initial group of 85 patients that were admitted with ICH, 62 patients did not receive PCC. From that group of 62 patients, 4 had a positive spot sign and survived, which formulated our control group. For those with positive CTA spot sign, PCC was administered right after the initial CTA. However, the exact time of PCC administration was not routinely documented in the charts.

Inclusion criteria for our study were acute ICH (within 24 h) and positive CTA spot sign (any age, sex, cause, or size). Exclusion criteria were non-active bleed demonstrated by a negative CTA spot sign, non-acute ICH, non-parenchymal ICH, surgical intervention before the follow-up CT at 5–24 h, lack of follow-up CT, death, or intervention with other procoagulants (i.e., fresh frozen plasma FFP, platelets, rFVII) as well as various antifibrinolytic agents.

At our institution, three indications for giving PCC include active acute ICH in the presence of warfarin, quick reversal of warfarin for an urgent surgical procedure, and lastly an active ICH with positive CTA spot sign. The latter criteria are followed by three of the four neurologists at our institution. Those patients with positive spot sign who received PCC formed our experimental group, while the ones who did not receive PCC comprised the control group. The degree of hematoma expansion was measured between the initial head CT and follow-up head CT at 5–24 h by using the ABC/2 formula (Fig. 1) for intracerebral hemorrhage volume [16] while accounting for the shape of the hematoma. This has been shown to be an accurate predictor of ICH volume [16]. The readers for this study were blinded and the only way they assessed the patient’s charts was through the medical record number. As the radiologists read each individual scan, they did not know which patient received PCC.

Fig. 1
figure 1

Non-contrast axial CT image (a) and ABC/2 formula (b) for ICH volume measurement. ABC/2 formula is as follows: A is the widest diameter, B is the diameter perpendicular to A, and C is the number of CT slices with hematoma multiplied by thickness (0.5)

Full size image

A t test was then done using the SPSS software to test statistical significance of association (Table 1). Outcome was measured in terms of percentage change in intracerebral hematoma at 24 h from the time of admission.

Table 1 Independent sample t test was performed using the SPSS software. Control group (n = 4) that did not receive PCC had an admission volume of 15.88 cm3 while the experimental group (n = 8) had an admission volume of 26.48 cm3. There was significant difference in the control group, with an increasing ICH mean volume of 46% (SD = 37.3%), compared to the experimental PCC group that had a mean volume decrease of 13% (SD = 29.9%) (p value = 0.012). P value of 0.012, which is particularly significant given our small sample size (n = 12)
Full size table

Results

Our retrospective cohort study favored our hypothesis. The control group (C) showed an increase in mean IPH volume of 46% (SD = 37.3%), whereas the experimental group (E) showed a decrease of 13% (SD = 29.9%) (p value = 0.012).

In the experimental group (E = 8), five out of the eight patients showed a reduction in ICH volume from the initial head CT to the follow-up head CT at 5–24 h with the average reduction in hematoma size being −22.35% (−33.33, −0.41, −19.96, −39.87, −18.18%). Three of the eight showed an increase in ICH volume with the average of 45.39% (9.55, 31.63, and 95%). Two of the three experimental patients who showed increase in hematoma size at follow-up head CT, both of them eventually showed a decrease in relative hematoma size at 27th hour. One of these two patients was found to have initial SBP of 200, who showed an increase in volume from their initial 6th hour follow-up; however, in their 27th hour follow-up, there was an overall reduction of −7.57% when compared to their initial exam. High SBP has been shown to be an independent risk factor for hematoma expansion and remains a confounding factor. The second of these patients had an increase of less than 10% (9.55%) in overall ICH volume from initial to the follow-up NCCT at 5 h which may be secondary to natural expansion of hematoma in the first 12 h; however, from their 5-h follow-up to their 27th hour follow-up, there was an overall decrease of 7.1%, with minimal increase in the mean hematoma size by 2.54% at 27th hour from the time of the initial head CT, which may be physiological.

In the control group, three out of the four patients showed an increase in ICH volume with the average change being 63.37% (44.96, 83.54, and 61.60%), whereas only one showed a relatively mild reduction in ICH volume (−4.21%). Therefore, the control group showed an overall quantitative increase in hematoma size of approximately 46%, with majority of patients showing mean hematoma volume increase of 63.37% and only one showing mean decrease of 4.21. However, our study is limited due to small sample size affecting the power of the study and is in need of a large scale multicentric double-blinded randomized controlled study.

Discussion

Our study results suggest that use of PCC in high-risk individuals with acute ICH manifested by positive CTA spot sign (Fig. 2) offers statistically significant reduction in the mean hematoma size at 5–24 h. Early containment of hematoma expansion is crucial in preventing secondary volume expansion which can ultimately result in brain herniation and irreversible brain damage.

Fig. 2
figure 2

CT angiogram series in two different patients above in sagittal, coronal, and axial planes demonstrate CTA spot sign shown as focal contrast extravasation, which is an independent predictor of hematoma growth

Full size image

In the past, multiple approaches to arrest hematoma expansion have been tested, including surgical evacuation, control of blood pressure, reversal of coagulopathies or anticoagulants, and hemostatic therapy [17]. PCC prevents acute hematoma expansion but does not address the underlying etiology for such, which can be treated after initially stabilizing the patient. Literature review suggests that elevated SBP increases the risk of hematoma enlargement [18] and therefore serves as an independent risk factor [18]. According to a trial (INTERACT) which explored lowering the blood pressure in order to resolve hematoma expansion, no benefit in overall clinical outcome [19]. According to a study in South Korea, positive CTA spot sign was used as a guide for early surgical evacuation of hematomas [20]. Early surgical intervention was shown to lower the mortality rates at 90 days but increased the length of inpatient stay and complications due to it being an invasive procedure [21]. Other trials focusing on early evacuation of hematoma also (STICH I) trial failed to show overall clinical benefits over conservative methods [4, 21]. The invasive nature of surgical intervention is associated with complications including risk of infection, hemorrhage, and systemic complications. As of now, there can be no general recommendations for surgery alone being a treatment for ICH [17, 22, 23].

There are several procoagulants available, such as different forms of prothrombin-dependent coagulation factors, which comprise of a concentrated form (PCC), unconcentrated form (FFP), and single factors (i.e., factor VII/rFVII). The FAST trial explored the effect of rFVIIa in reducing the growth of ICH as well as its effect on overall survival and clinical outcomes [11, 24]. This study showed reduction in hematoma growth with the use of rFVIIa within 4 h of ICH onset [24], and its Phase IIb trial showed a decrease in mortality and improved functional outcomes. However, in the phase III trial, no change in 90-day mortality or functional outcome was observed [11], which may be due to thromboembolic complications such as myocardial ischemia or cerebral infarction [11]. Another study funded by the NIH called the STOP-IT trial looked at the role that rFVIIa on non-warfarin induced ICH. This study concluded in April 2016, pending results.

PCC is FDA approved for warfarin-related ICH; its role in hemorrhage in non-warfarin patients has not been established. Our literature review failed to discover definitive completed trials evaluating the role of PCC in hematoma expansion in non-warfarin population. On molecular level, use of PCC in patients on anticoagulants like warfarin results in a predictable outcome of decreasing hematoma size, as PCC replenishes the missing clotting factors caused by warfarin. However, our results show beneficial role of PCC in preventing ICH hematoma expansion even in non-coagulopathic patients. This leads one to infer that surplus of essential clotting factors provided by PCC has a role in precipitating coagulation cascade by supplementing inherent clotting factors even in the absence of warfarin.

Our study is not without biases and limitations. Small sample size makes it harder to ensure that a proper representation of the population is made, so that the data analysis can be generalized to other populations. It also results in low statistical power lowering the chance of detecting true effect even when the study is statistically significant. Another limitation can be the subjective bias involved while measuring the hematoma volume. Although the ABC/2 formula has been shown to be a good indicator of obtaining accurate hematoma volume, measurement error cannot be excluded. Subjective error is possible while taking measurements and volume calculations (Fig. 3). We minimized this bias by concealing the subject’s identifying information and blinding the reader to their group status (control vs experimental). We minimized this bias by ensuring double reads on each patient to foster reproducibility and accuracy. While using the ABC/2 formula, attention was also paid to the shape of the hemorrhage such as whether it was well-defined elliptical or round vs multinodular and irregular to correctly assess the hematoma size. Another potential bias that we encountered during the study were the mimics of acute hemorrhage for example a large hyperdense neoplasm such as an atypical meningioma, which can enhance and show nodular vascularity, can be mistaken for positive CTA spot sign. Lastly, our study being a retrospective study, selection bias cannot be excluded. The study was double blinded to objectively assess the experimental and control groups simultaneously. Since we included all patients with acute ICH (<24 h), it allowed for a more broad variation in the symptom onset to CT time, which may have ultimately affected the way PCC had on the hematoma expansion. Even if a hemostatic agent works superbly, the hematoma may grow at a rate as high as 26% within the 6th to 12th hour [24]. We would advocate administering PCC within the first 12 h of symptom onset [6, 9].

Fig. 3
figure 3

Pictorial bar graph analysis demonstrates overall decrease in ICH volume in the experimental PCC group and overall increase in ICH volume in the control group

Full size image

PCC is an aged old agent which has survived the test of time. It has been further refined over the years and is now available in the form of PCC4. Our study provides an initial platform to evaluate our preliminary experience at this single institutional stroke center. This study aims to lay the ground work for further research in this field, the benefits of which will be valued across multiple stroke centers. Potential directions for future research include multicentric large-scale double-blinded randomized controlled trials to further evaluate the optimal therapeutic uses of PCC, assess its functional outcomes, and determine the extent of known associated complications such as thromboembolic events. Functional status can be assessed by several means such as by evaluating hematoma size on non-contrast head CT at 90 days, assessing Rankin score at 90 days, and length of hospital stay. Since the long-term effects of PCC are currently not clear, we recommend use of PCC only in patients with warfarin and patients with acute hemorrhage (<24 h) with active bleeding (positive CTA spot sign). A multinational study incorporating multiple stroke centers can further legitimate this study by removing the cultural and regional factors and model a more diverse sample population that can be a better representation of all patients. Further studies comparing utility of PCC in acute (<24 h.) vs subacute (24 h to 2 weeks) ICH may be of great clinical importance in patient management.

Ultra-early hemostatic agents (given within 3–4 h of onset) include antifibrinolytic agents such as aminocaproic acid, tranexamic acid, and aprotinin. On a more molecular level, several interesting studies are in the pipeline. According to one such study, inhibition of the gelatinase matrix metalloproteinase-9 by the broad-spectrum inhibitor GM6001 had shown several beneficial effects done on a mouse model which resulted in decreased injury volume and improved clinical outcomes [25]. These studies are promising and target to solve the problem inside out.

Conclusion

Acute intracerebral hemorrhage can be a catastrophic neurologic event and must be addressed in a timely manner. Expansion of intracranial hemorrhage can lead to mental deterioration and decreased functional status in patients who survive. Since urgent neurosurgical evacuation so far has failed to translate into overall improvement in clinical outcome, we anticipate great promise in PCC as a noninvasive quick-acting tool to stabilize and control acute active ICH in all patients even if they are not on warfarin. A universal standard protocol for the safe use of PCC in acute ICH should be adapted across institutions like ours with clear guidelines.