Study of clinical and radiological profile and outcome in patients of intracranial hemorrhage.

BACKGROUND: Incidence of intracerebral hemorrhage (ICH) is twice as high as in Western countries. Prognostic factors for predicting function outcome and mortality play a major role in determining the treatment outcome.
METHODS: A prospective study of male and female patients ≥12 years with primary nontraumatic intracranial hemorrhage were included. Hemorrhage caused by trauma, anticoagulant or thrombolytic drugs, brain tumor, saccular arterial aneurysm or vascular malformation were excluded. Functional outcome of patients was determined by modified Rankin's scale. Glasgow Coma Scale (GCS) score and ICH score were calculated for each patient.
RESULTS: Hypertension was present in 45 out of 49 patients (92%) with ICH of basal ganglia. Hypertension was significantly associated with worst clinical outcome. Mortality was high if the patient was comatose/stuporous compared to drowsy state (P < 0.0001). Mortality was found to be high when the size exceeded 30 cm3. High ICH score, low GCS score at the time of admission, presence of intraventricular hemorrhage, and midline shift were significantly associated with poor clinical outcome.
CONCLUSIONS: Intracranial hemorrhage can be deleterious if present with low GCS score, high ICH score, intraventricular extension, and midline shift.

Introduction

Intracerebral hemorrhage (ICH) is a common devastating neurologic event that causes high morbidity and mortality with profound economic implication. ICH will seem to continue to be an important problem in both India and other developed countries. Nontraumatic ICH occurs due to bleeding from a vascular source directly into the brain substance. It is a major public health problem[1] with an annual incidence of 10–30/100,000 population,[12] accounting for 2 million (10–15%)[3] of about 15 million strokes worldwide each year.[4]

Approximately, 35–50% of patients with ICH die within the 1st month after bleeding.[56] Data from the Asian Stroke Advisory Panel revealed an incidence of ICH ranging from 17% to 33% of all strokes, twice as high as in Western countries.[7] Mortality depends on the size of the hematoma and on the location of the brain.

Intracranial hemorrhages include ICH, subarachnoid hemorrhage (SAH), and intraventricular hemorrhage (IVH). Hemorrhages occurring into extradural or subdural spaces are usually associated with trauma. Traumatic causes of hemorrhages were not included in present study. Primary or hypertensive (spontaneous) ICH, ruptured saccular aneurysm and vascular malformations and hemorrhages associated with bleeding disorders account for most of the hemorrhage that present as stroke.

Radiological studies such as computerized tomogram (CT) scan and magnetic resonance imaging have facilitated in locating and assessing the extent of insult precisely and deciding prognosis of the patient. Although several randomized therapeutic trials for ICH have been published, neither surgical nor medical treatments have been shown conclusively to benefit patients.[891011121314] However, early surgical intervention has shown mild statistically significant improvement in clinical outcome.[15] Prognostic factors for predicting function outcome and mortality thus play a major role in determining the treatment outcome.[56161718192021222324252627]

Hospital admissions for ICH have increased by 18% in the past 10 years,[28] probably because of increases in the number of elderly people,[29] many of whom lack adequate blood pressure (BP) control, and the increasing use of anticoagulants, thrombolytics, and antiplatelet agents. Incidence might have decreased in some populations with improved access to medical care and BP control.[303132]

Hence, we aimed to study, the clinical, radiological profile in patients of ICH. In clinical profile risk factors, symptomatology and physical findings were studied. In radiological profile, size, and location of hematomas, correlation of the size of hematomas with the short term prognosis (6 weeks) were also studied. We also studied mortality, morbidity, and predictive factors of clinical outcome of ICH.

Subjects and Methods

The study was started after obtaining approval from Institutional Review Board. The present study was conducted in the Medicine Department of a Tertiary Care Hospital in the Western part of India. Patients with intracranial hemorrhage during the period of November 2011–October 2013 were included. It is a prospective observational study. Hundred and nine patients with primary (nontraumatic) intracranial hemorrhage were included. Nine cases were excluded due to incomplete data. Male and females patients with age ≥12 years were included. Patients with intracranial hemorrhage (hemorrhage or bleeding within the skull) who have presented with a history of acute severe headache, altered sensorium, slurring of speech, acute hemiparesis, and accelerated hypertension – suggestive of acute cerebrovascular stroke and those who are presented with nontraumatic spontaneous hemorrhage were included. Patients whose hematoma was possibly caused by trauma, anticoagulant or thrombolytic drugs, brain tumor, saccular arterial aneurysm or vascular malformation were excluded. Patients with SAH, venous thrombosis, past history of intracranial hemorrhage, ischemic stroke, and those who underwent surgical intervention for it were excluded.

All patients presenting with and fulfilling the inclusion criterion were included in this study after obtaining informed consents. Detailed relevant history and clinical examination were done according to predesigned and pretested protocol. All patients underwent CT imaging. Hematological and biochemical investigations like; complete blood counts, random blood sugar, renal function test, serum electrolytes, fasting serum lipid profile, prothrombin time with international normalized ratio, activated partial thromboplastin time, chest X-ray, electrocardiogram (ECG), two-dimensional (2D)-echocardiography with Doppler examination (2D echo), urine complete examination, ophthalmic fundus, and other investigations as and when required by standard laboratory techniques were done. The cases were studied in detail with emphasis on (1) contribution of risk factors particularly hypertension, (2) the level of consciousness on admission and improvement or deterioration thereafter and it's relation to prognosis, (3) the degree of motor weakness and mental functions and their recovery.

CT scan was done on admission in all patients to confirm the diagnosis of ICH and to correlate the prognosis with the CT scan features according to the size of the hematoma in cm3 (length × width × thickness/2 in cm3) at various locations.[16] Other parameters in the CT scan were also included viz., the presence of ventricular extension, presence of midline shift and inferences correlated with prognosis. Management during the hospital stay included control of hypertension with appropriate antihypertensives, treatment of raised intracranial pressure with mannitol, maintenance of fluid and electrolyte balance, maintenance of airway, oropharyngeal suction, and Ryle's tube aspiration, general supportive measures to prevent bed sore, contractures. Treatment of intercurrent infections and associated complications were treated as per requirements. All the patients were followed up daily during the hospital stay and regularly thereafter until 6 weeks. Functional outcome of patients was determined by modified Rankin's scale (mRS).[33] Glasgow Coma Scale (GCS) score and ICH score[16] were calculated for each patient.

Intracerebral hemorrhage score[16]

The data obtained was tabulated and was analyzed using proportion and percentages. Statistical Package for Social Science software (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp.) was used for my statistical analysis. Different variables such as GCS score, ICH score, and its components, age, gender, BP, various risk factors of ICH were subjected to multiple regression analysis and were correlated with respect to clinical outcomes in term of mRS by GraphPad InStat software version 3.06 for Windows 95, (GraphPad Software, San Diego California USA, www.graphpad.com). Chi-square test was applied to calculate the mean value, 95% confidence interval (95% CI) of mean value and P value of hematoma size of all groups.

Results

Total 109 cases were taken for the study, but nine patients have to be excluded from study data due to insufficient data. The following observations were made from the data.

Age and gender

The mean age of the patients was 55.15 years. The majority of them were between 40 and 65 years. However, increased age was not associated with an increase in mortality. There were 62 males and 38 females.

Risk factors

Hypertension

This is the most consistent risk factor for ICH. In the present study, 88 patients had a diastolic BP of >100 mmHg on admission (88%). Mean BP calculated was 118.51 mmHg. Hypertension was present in 45 out of 49 patients (92%) with ICH of basal ganglia. In the case of lobar hemorrhages, 15 out of the 21 patients (71%) had hypertension while in the case of thalamic hemorrhages 12 out of 14 patients (86%) had hypertension. All the patients with cerebellar and brainstem hemorrhages were found to be hypertensive.

Diabetes mellitus

This is one of the major risk factors contributing to stroke. Hyperglycemia predisposes to hyperlipidemia and atherogenesis and thus perpetuates hypertension. In this study, 10 patients gave past history of diabetes mellitus. Mean 2-h postprandial blood sugar (PP2 BS) value was 153.99 mg/dl.

Ischemic heart disease, smoking, alcohol and family history

Ischemic heart disease (IHD), smoking and alcohol abuse and family history of cerebrovascular accident have been observed to be associated with an increased incidence of ICH, but not to the extent of hypertension.

Clinical presentation

There was considerable overlap of clinical features between all forms of ICH though there are different clinical features for different sites, the incidence of hypertension on admission in all forms of ICH correlated with other features of hypertension like left ventricular hypertrophy with or without strain pattern in ECG and hypertensive retinopathy. A headache was present in 41% of the patients more commonly seen with lobar hemorrhages. Vomiting occurred in 39% of patients, and 17% had seizures at onset. In most of them, the seizures were generalized tonic-clonic in nature. There was no association between seizures and hematoma size. A headache, vomiting and altered level of consciousness were seen in many patients, but their absence did not rule out ICH.

Level of consciousness on admission

Altered level of consciousness is one of the most common manifestations of ICH. Its severity varies according to the size and location of the hematoma. Stupor or coma was a bad prognostic sign. Altered level of consciousness was present in 63 patients. Bilateral upper motor neuron (UMN) signs and absent Doll's eye movement was significantly associated with high mRS score.

Fundoscopy, chest X-ray and electrocardiogram

Evidence of hypertensive retinopathy was present in 25 patients. Out of which 16 patients were expired (60%) (P = 0.0007). Papillodema was present in two patients, both of them get expired. The presence of hypertensive retinopathy may suggest prolonged, and uncontrolled hypertension and hence may be associated with a poor prognosis. Commonest chest X-ray finding was cardiomegaly (51%). The most common ECG findings were left ventricular hypertrophy (48%) followed by ST-T changes (16%) and sinus bradycardia (6%).

Various factors and outcome relation

Mean BP calculated: 118.51 mmHg was subjected to regression analysis in relation to mRS. (A: mRS) = −5.817 + 0.079 × (B: Mean BP). R2 = 11.80%. This is the percent of the variance in mRS explained by mean BP. The P < 0.0001, which was considered extremely significant. So hypertension was significantly associated with worst clinical outcome. PP2 BS value was also subjected to regression analysis in relation to mRS, the P = 0.019, which was also clinically significant. Hence, hyperglycemia was also significantly associated with clinical outcome.

Mortality was high if the patient was comatose/stuporous compared to drowsy state (P < 0.0001). Impaired consciousness lasting for more than 48 h had a major impact on prognosis.

Other features of significance from the prognostic point of view are bilateral UMN signs indicate a poor prognosis. In this study, 17 out of the 21 patients (81%) with this finding expired (P < 0.0001). The absence of Doll's eye movement was associated with increased mortality (P = 0.007).

Computerized tomogram scan aspects and outcome relation

Following parameters were studied in CT scan brain, site of the hematoma, size of the hematoma, midline shift, and intraventricular extension/IVH.

Site of hematoma

Most common site of ICH was basal ganglia (49%) followed by lobar (21%) and thalamus (14%). Mortality was 41% of all patients; it was highest in brain stem hematoma (67%) followed by a lobar hematoma (57%). Severe morbidity in term of mRS 4/5 was 15% in all patients; it was 21% in lobar, 14% in cerebellar, 14% in thalamus, 11% in the brainstem. IVH was highest in thalamus (43%) followed by lobar (38%) and basal ganglia (24%). Midline shift was highest in lobar (52%) and brainstem hematoma (44%) [Figure 1 and Table 1].

Figure 1

Site of hematoma

Table 1

Clinical outcomes, intraventricular hemorrhage and midline shift in relation to the location of intracranial hemorrhage

Location	Mortality		Morbidity		IVH		Midline shift
	n	Percentage	n	Percentage	n	Percentage	n	Percentage
All patients (n=100)	41	41	15	15	27	27	34	34
Basal ganglia (n=49)	17	35	8	16	12	24	14	29
Thalamus (n=14)	8	57	2	14	6	43	5	36
Lobar (n=21)	6	29	3	21	8	38	11	52
Brainstem (n=9)	6	67	1	11	1	11	4	44
Cerebellar (n=7)	4	57	1	14	0	0	0	0

IVH=Intraventricular hemorrhage

Size of the hematoma versus prognosis in general

Overall mortality was 41%, and morbidity (mRS 3/4/5) was 15%. In the present study 39 patients had a hematoma size of ≤ 5 cm3, of which five patients (12.8%) died. One point which is worth noting is that all those five patients who died had brainstem hemorrhages. In the 5.1–15 cm3 group, out of the 22 patients four patients died (18.18%). Mortality rate was found to be high when the size exceeded 30 cm3 (24 out of 25, i.e., 96%) [Table 2].

Table 2

Outcome in relation to the size of the hematoma in general

Size in cm3	Cases (n)	Midline shift	Ventricular extension	Outcome (mRS) at 6 weeks
				mRS 1	mRS 2	mRS 3	mRS 4	mRS 5	mRS 6	Total
<5	39	0	3	21	8	3	1	1	5	39
5.1-15	22	4	8	2	11	5	0	0	4	22
15.1-30	14	9	7	1	1	3	1	0	8	14
>30	25	21	9	0	0	1	0	0	24	25
Total	100	34	27	24	20	12	2	1	41	100

The P value for mRS <3 versus mRS >3 is <0.0001. There is a significant linear trend among the ordered categories. mRS=Modified Rankin's scale

Clinical outcome in relation to size of intracerebral hemorrhage

Size of the ICH (cm3) was high in patients who died (n = 41, mean size 37.63 cm3, 95% CI: 30.47–44.79) compared to patients who survived (n = 59, mean size 6.98 cm3, 95% CI: 5.38–8.58) (P < 0.0001) [Table 3].

Table 3

Size of intracerebral hemorrhage and its mean value in relation to mortality

Clinical outcome at 6 weeks	Number of patients	Mean value (cm3)	SDM	SEM	95% CI of mean	P
All patients	100	19.55	21.56	2.16	15.23–23.87	<0.0001
Patients who died	41	37.63	22.95	3.58	30.47–44.79
Patients who survived	59	6.98	6.19	0.80	5.38–8.58

CI=Confidence interval, SEM=Standard error of mean, SDM=Standard deviation of mean

The size of the ICH was high in patients who had significant disability mRS >3 (n = 44, mean size 35.81 cm3, 95% CI: 28.80–42.81) compared to patients who had mild or no disability mRS ≤3 (n = 56, mean size 6.71 cm3, 95% CI: 5.09–8.33) (P < 0.0001). The increase in the size of ICH was significantly associated with increased mortality and poor clinical outcome in the present study. Other studies Nag et al. 2012,[34] Hemphill et al. 2009,[35] Hallevy et al. 2002,[36] Zia et al. 2009[37] and Broderick et al. 1993[25] shows similar results [Tables 4, Figure 2].

Table 4

Size of intracerebral hemorrhage and its mean value in relation to clinical outcome (at 6 weeks) Modified Rankin's scale ≤3 and Modified Rankin's scale >3

Clinical outcome at 6 weeks	Number of patient	Mean value (cm3)	SDM	SEM	95% CI of mean	P
All patients	100	19.55	21.56	2.16	15.23-23.87	<0.0001
Patients with mRS <3	56	6.71	6.05	0.80	5.09-8.33
Patients with mRS >3	44	35.81	23.02	3.47	28.80-42.81

CI=Confidence interval, SEM=Standard error of mean, SDM=Standard deviation of mean, mRS=Modified Rankin's scale

Figure 2

Clinical outcome in relation to intracerebral hemorrhage score

Clinical outcome in relation to Glasgow Coma Scale score, intracerebral hemorrhage score, intraventricular hemorrhage, midline shift

Of all the 41 patients with ICH score 0, 39 had mRS <3, none died. Among 18 patients with ICH score 3, all patients died. All patients with ICH score 4 and 5 also died. The P value of mRS for ICH score is <0.0001 which is statistically highly significant. The increase in ICH score was associated with poor clinical outcome in the present study [Table 5].

Table 5

Clinical outcome in relation to Glasgow Coma Scale score, intracerebral hemorrhage score, intraventricular hemorrhage, midline shift

	Number of patients	mRS <3	mRS >3	mRS 6	χ2	P mRS <3 versus mRS >3
ICH score
0	41	39	2	1	68.08	<0.0001
1	19	16	3	2
2	17	1	16	15
3	18	0	18	18
4	4	0	4	4
5	1	0	1	1
Total	100	56	44	41
GCS score
<8	36	1	35	34	61.33	<0.0001
>8	64	55	9	7
Total	100	56	44	41
IVH
Present	27	10	17	16	5.398	<0.05
Absent	73	46	27	25
Total	100	56	44	41
Midline shift
Present	34	7	27	25	26.217	<0.0001
Absent	66	49	17	16
Total	100	56	44	41

GCS=Glasgow Coma Scale, ICH=Intracerebral hemorrhage, IVH=Intraventricular hemorrhage, mRS=Modified Rankin's scale

Of all 100 patients, 36 patients had GCS score <8 at the time of admission, of which 35 had mRS score >3 and 34 died (P < 0.0001). IVH was present in 27 patients, of which 17 patients had mRS >3 and 16 died. 10 patients had mRS <3 [P = 0.005]. Midline shift was present in 34 patients of which 27 patients had mRS >3 and 25 were died (P < 0.0001).

Multiple regression analysis

With an objective to find out the impact of individual risk factor associated with higher mortality we have studied various variables such as age, size of ICH, mean BP, GCS score, ICH score, and they were subjected to multiple regression analysis in relation to clinical outcome in term of mRS score [Table 6].

Table 6

Multiple regression analysis of different variable with mortality

Variable	Beta coefficient	SE	95% CI
Constant	4.538	1.799	0.966-8.111
A: Age	0.014	0.012	−0.011-0.038
B: Mean BP	0.003	0.012	−0.020-0.026
C: GCS score	−0.269	0.049	−0.366-−0.172
D: ICH score	0.272	0.170	−0.066-0.610
F: Size	0.015	0.006	0.003-0.028

GCS=Glasgow Coma Scale, ICH=Intracerebral hemorrhage, CI=Confidence interval, SE=Standard error, BP=Blood pressure

Following results were obtained:

Estimated mortality = 4.538 + 0.014 × (A: Age) + 0.003 × (B: Mean BP) −0.269 × (C: GCS) + 0.272 × (D: ICH) + 0.015 × (F: Size).

Estimated mRS: 2.522 + 1.115 × (A: Age) + 0.258 × (B: Mean BP) −5.492 × (C: GCS) + 1.597 × (D: ICH) + 2.381 × (F: Size).

80.60% contribution in clinical outcome (mRS) at 6 weeks was made from five following variables: Age of the patient, mean BP control, the size of ICH, GCS score at admission and ICH score. That is to say, that size of ICH and GCS score at the time of admission were the most important predictor of clinical outcome [Table 7].

Table 7

Which variable(s) make a significant contribution to outcome?

Variable	Beta coefficient	P	Significance
Constant	2.522	0.013	Yes
A: Age	1.115	0.268	No
B: Mean BP	0.258	0.797	No
C: GCS score	−5.492	<0.0001	Yes
D: ICH score	1.597	0.114	No
F: Size	2.381	0.019	Yes

GCS=Glasgow Coma Scale, ICH=Intracerebral hemorrhage, BP=Blood pressure

Discussion

The incidence of ICH was highest in the age group 51–60 years (49% of the patients) in the present study. There was no significant gender difference observed. The distribution was similar to several other studies like Das Arjun study.[38]

Hypertension, diabetes mellitus, and ischemic heart disease

Hypertension was significantly associated with worst clinical outcome. Uncontrolled hypertension especially diastolic BP >100 mmHg was the most important risk factor in the present study. The majority of cases depending on the site of hemorrhage were associated with hypertension, especially cerebellar and brainstem hemorrhages. These findings were comparable to Das Arjun study.[38] Other factors, like left ventricular hypertrophy with or without strain pattern in ECG, hypertensive retinopathy indicated inadequately controlled hypertension were also with higher morbidity and mortality. Diabetes mellitus in our study was also a poor risk factor associated with worse clinical outcome. IHD, smoking, and alcohol abuse were less prominent risk factors in the present study.

Clinical presentation

Decreased level of consciousness on admission was associated with poor clinical outcome. The prognosis was worst in stuporous and comatose patients. Although symptoms such as a headache, vomiting, and altered level of consciousness were frequently seen, their absence did not rule out ICH with GCS <8.

The occurrence of a headache (41%), vomiting (39%), and seizures (17%) were comparable with other similar studies like Mohr[39] and Pillai study.[40] Mortality was high if the patient was comatose. It can be concluded that impaired consciousness lasting for more than 48 h had a major impact on prognosis which is also seen several other studies. Bilateral UMN signs indicate a poor prognosis (P < 0.0001). The absence of Doll's eye movement is associated with increased mortality (P = 0.007). The presence of hypertensive retinopathy may suggest prolonged, and uncontrolled hypertension and hence may be associated[41] with a poor prognosis.

Mortality, morbidity, and various parameters

The most common site for intracranial hemorrhage in the present study was basal ganglia (49%) followed by lobar (21%), thalamus (14%), brainstem (9%), and cerebellum (7%). These findings were comparable with several Indian studies except Das Arjun study[38] were basal ganglia, and thalamic hemorrhage was observed less than lobar, brain stem, and cerebellar hemorrhages. Prakash and Bhavani et al.[4142] have similar sites of hemorrhage as our studies. The most important factors predicting the final outcome were the size of hematoma, intraventricular extension, midline shift, GCS score and ICH score. These results are comparable to results of an Indian study Nag et al. 2012[34] which also shows mean hematoma volume associated with death before 30 days is 33.16 cm3 (P < 0.0001), with survival after 30 days is 15.45 cm3 (P < 0.0001). Ventricular extension depends on the location of the hematoma, and the size and per se does not carry a poor prognosis with it. In the present study, out of total 27 patients having IVH, maximum IVH were associated with basal ganglia hemorrhages (44%). In the present study, 6 weeks outcome was good in patients with small hematoma ≤5 cm3 without midline shift or ventricular extension. Outcome worsened as the size of the hematoma increased. Mortality increased to 82% (32 out of 39 patients died) in patients having hematoma size >15 cm3. In the present study, mortality was higher in the case of brainstem hemorrhages (67%), followed by thalamic (57%) and cerebellar hemorrhages (57%).

The most important factors predicting the final outcome were the size of hematoma, intraventricular extension, midline shift, GCS score and ICH score. These results indicated that high ICH score, low GCS score at the time of admission, presence of IVH, and midline shift were significantly associated with poor clinical outcome. Our these results are comparable to other studies like Nag et al. 2012,[34] Hemphill et al. 2009,[35] Hallevy et al. 2002,[36] Zia et al. 2009[37] and Broderick et al. 1993.[25] Ventricular extension depends on the location of the hematoma and the size and per se does not carry a poor prognosis. In this study, out of total 27 patients having IVH, maximum IVH were associated with basal ganglia hemorrhages (12/27 = 44%).

However, surgical interventions early in the course of illness may change the course of illness. A small clinically relevant advantage has been noted by surgical trial in the intracerebral hemorrhage-II study when the early surgical intervention has been done in these cases of ICH.[15] However, the advantage may depend on surgical skill, infrastructure, and patient education. Timely surgical intervention may change outcome positively. Cerebral amyloid angiopathy is also one of the risk factors for the development of ICH, which also need to be studied.

In the present study, 6 weeks outcome was good in patients with small hematoma ≤5 cm3 without midline shift or ventricular extension. Outcome worsened as the size of the hematoma increased. Mortality increased to 82% (32 out of 39 patients died) in patients having hematoma size >15 cm3. Mortality was relatively higher in the case of brainstem hemorrhages (67%), followed by thalamic (57%) and cerebellar hemorrhages (57%).

These results compared with other studies and according to these four predictors: GCS score, mRS, ICH score, and age were proposed as variables directly influencing hospital mortality in ICH.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Component	ICH score points
GCS score at presentation
13-15	0
5-12	1
3-4	2
ICH volume (cm3)
≥30	1
<30	0
IVH
Yes	1
No	0
Origin of ICH
Infratentorial	1
Supratentorial	0
Age
≥80	1
<80	0
Total ICH score	0-6

ICH=Intracerebral hemorrhage, IVH=Intraventricular hemorrhage, GCS=Glasgow Coma Scale