INTRODUCTION

Although spontaneous intracerebral haematoma (ICH) accounts for approximately 10-15% of all acute cerebrovascular diseases¹, its importance stems from the generally severe neurological deficits it causes and its often grave prognosis².

To develop therapeutic strategies for patients with spontaneous ICH, we need to be able to predict their evolution particularly during the very acute stage (<24 hours after stroke onset)³. Several studies have focused on early survival after spontaneous ICH, and on the identification of its prognostic factors^4-7 but some used only univariate analysis which fails to account for interaction among prognostic factors^4-6, and other suggested a highly sophisticated model to predict outcome for routinely clinical purposes⁷.

Based on initial clinical and brain CT findings, our aim was to determine: (1) The predictors of 30-day mortality in ICH patients, (2) The prognostic factors that determine the probability of favorable functional outcome in patients who survive the acute stage of ICH.

METHODS

Data were collected from 67 consecutive patients admitted to the stroke unit of the Neurology Department of Tanta University hospital over a period of six months and who have met the following inclusion criteria:

1.      Admission within less than 24 hours of stroke onset.

2.      CT scan of the brain demonstrating non-traumatic ICH.

3.      Supratentorial site.

Thirty-eight patients (56.7%) were males and 29 (43.3%) were females. The age ranged between 34 and 80 (mean 53.5±10.19) years.

All patients were evaluated neurologically and clinically on admission and all variables were uniformly assessed. The level of consciousness was evaluated according to the (GCS)⁸. CT scan of the brain was performed within 24 hours of admission.

The following clinical variables were considered as potential prognostic factors suitable to influence the 30-day mortality and functional outcome: 1) patient's age, 2) gender, 3) mean arterial pressure (MAP) on admission, 4) level of consciousness on admission, 5) presence of gaze palsy and 6) pupillary abnormality. The following CT scan findings were also taken into account 1) haematoma location "site" 2) haematoma side, 3) haematoma size "volume" 4) midline shift "displacement on CT scan > 5mm" and 5) intraventricular extension (IVE) of the haematoma.

The total volume of parenchymal haematoma was calculated at bed side on CT films using ABC /2 formula⁹ (where the CT slice with largest area of the haematoma was identified and the largest diameter (A) of the haematoma on this slice was measured using the centimeter scale on the CT scan. The largest diameter 90° to A on the same slice measured next (B). Finally the approximate number of 10-mm slices on which ICH was seen was calculated (C). A, B and C were then multiplied and the product divided by 2 which yielded the volume of parenchymal haematoma in cubic centimeters).

Spontaneous ICH was classified on CT scan as deep (including putaminal and thalamic bleeds) or lobar. The haematoma was considered lobar, if was situated in the subcortical white matter, wholly outside the basal ganglia and thalamus, deep if the haematoma contacted these structures.

All patients were conservatively treated by a standard therapeutic protocol used in our stroke unit. The mortality rate was calculated 30 days following the stroke onset, whereas the survivors were evaluated for 2 months and their functional outcome then was classified according to the Rankin disability scale¹⁰ into the following grades (1=no significant disability-despite symptoms able to carry out all usual duties and activities, 2= slight disability- unable to carry out all previous activities, but able to look after own affairs without assistance, 3= moderate disability- requiring some help, but able to walk without assistance, 4= moderately severe disability- unable to walk without assistance and unable to attend to own bodily needs, 5= severe disability- bed ridden, incontinent and requiring constant nursing care and attention). Surviving patients' outcome at the end of 2 months was then classified into satisfactory "Rankin 1, 2 and 3" or unsatisfactory " Rankin 4 and 5".

Statistical Analysis:

The statistical analysis was performed by univariate methods (x² or Fisher exact probability test or t-test as appropriate) to correlate the patients' outcome and each potential prognostic factor. Then multiple logistic regression "multivariate" analysis was used to estimate the probability of a given outcome in an individual patient as a function of that patient's prognostic factor.

Outcome was expressed as 1) alive at 30 days versus dead at 30 days and 2) satisfactory outcome "Rankin 1-3" at 2 months versus unsatisfactory outcome "Rankin 4 and 5" at 2 months. Value of p < 0.05 was considered significant.

RESULTS

Among the 67 studied ICH patients, putaminal location was the site of preference (46.2%), lobar location was the second in frequency (28.4%) whereas thalamic location constituted the remaining (25.4%)

Arterial hypertension, the main vascular risk factor of spontaneous ICH, was present in (83.8%) of patients. 26.7% of our hypertensive patients had their haematoma in the lobar location, whereas the rest (73.3%) had their haematoma within the deep locations.

The 30-day mortality rate was 11 deaths (16.42%), two patients died from non-neurological causes after one month, one patient died with bronchopneumonia and the other with renal failure and were included in the unsatisfactory outcome group at 2 months.

Regarding the 56 surviving patient after one month, 41 (73.21%) had a satisfactory functional outcome "Rankin 1-3" and 15 (26.79%) had unsatisfactory functional outcome "Rankin 4-5" at the end of the 2-months follow up period.

Univariate statistical analysis (Table 1) showed that patients age (P=0.0001), low admission GCS (P=0.00001), high admission MAP (P=0.005), gaze palsy (P=0.001), pupillary abnormality (P=0.003), ICH size (P-0.00001), IVE of the haematoma (P=0.01) and midline shift (P=0.04) to be significantly associated with 30-day mortality. On the other hand, gender (P=0.861), haematoma site (P=0.669) and haematoma side (P=0.47) were not significant. According to the results of univariate analysis, eight potential prognostic factors (patient's age, admission GCS, admission MAP, ICH site, ICH size, ICH side, IVE of the haematoma and midline shift) were selected for multivariate analysis by logistic regression.  The results are presented in (Table 2) and showed that admission GCS (P=0.014) and ICH size (P=0.04) were the only significant independent predictors of 30-day mortality independent predictors of 30-day mortality among our patients.

Using the same statistical model, univariate analysis found that a satisfactory outcome "Rankin 1-3" at 2 months to be inversely correlated with old age (P=0.00001), low admission GCS (P=0.0001), high admission MAP (P=0.002), ICH size (P=0.,0001), and the presence of IVE of the haematoma (P=0.01), and midline shift (P=0.001) whereas gender (P=0.942), presence of gaze palsy (P=0.14), presence of pupillary abnormality (P=0.24), ICH site (P=0.899), and ICH side (P=0.724) were not significant (Table 3).

In multivariate analysis, age (P=0.01), admission GCS (P=0.04), and ICH size (P=0.03), were significant independent predictors of functional outcome (Table 4).

Moreover, our study revealed by univariate analysis (Table 5), that the prognostic role of IVE of the haematoma is particularly related to the haematoma site, showing its significant adverse effect on the 2-months functional recovery of patients with lobar haematoma (P=0.02), in contrast to its insignificant effect on patients with both putaminal (P=0.32), and thalamic (P=0.37) locations.

In support to the above data, our study revealed also by univariate analysis (Table 6) a significant relationship between haematoma volume and IVE only for patients with lobar haematoma (P<0.05), whereas in both putaminal and thalamic locations, such relationship was lacking (P>0.05) in each.

Table 1. Univariate statistical analysis between variables and 30 day mortality.

CT Computerized tomography, F Fischer’s exact test, GCS Glasgow Coma Scale, ICH intracerebral hematoma, IVE Intraventricular extension,  MAP Mean arterial pressure,  t Student’s t test,  X² Chi square test

* Significant     

Table 2. Predictors of 30-day mortality in 67 Intracerebral hematoma patients by multivariate analysis.

CI confidence interval, CT computerized tomography, GCS Glasgow coma scale, ICH intracerebral hematoma, IVE intraventricular extension,  MAP mean arterial pressure, OR  odds ratio

* Significant.

Table 3. Univariate statistical analysis between patients with satisfactory outcome " Rankin 1-3" versus those with unsatisfactory outcome " Rankin 4-5" at 2 months.

CT computerized tomography, F Fischer’s exact test, GCS Glasgow coma scale, ICH intracerebral hematoma, IVE Intraventricular extension,  MAP mean arterial pressure,  t Student’s t test,  X² chi square test

* Significant

Table 4. Predictors of satisfactory outcome at 2 months in the surviving ICH patients’ multivariate analysis.

CI confidence interval, CT computerized tomography, GCS Glasgow coma scale, ICH intracerebral hematoma, IVE intraventricular extension,  MAP mean arterial pressure, OR  odds ratio

* Significant.

Table 5.  Relationship between IVE of parenchymal haematoma and functional recovery at 2 months regarding the surviving patients at various haematoma locations.

IVE intraventricular extension

Table 6. Relationship between mean hematoma volume (HV) measured on initial brain CT scan and IVE regarding the 67 studied patients at various supratentorial locations.

CT computerized tomography, HV hematoma volume,  IVE  intraventricular extension, SD standard deviation

* Significant

DISCUSSION

The 30-day mortality rate (16.42 %), in our study, is relatively low than that reported by others^7,11,12, but this difference may be attributed to the inclusion of infratentorial haematomas in other studies.

In this study, the predictors of 30-day mortality, using univariate analysis were; age of the patient, admission GCS, admission MAP, presence of gaze palsy and pupillary abnormality, ICH size, IVE of the parenchymal haematoma and midline shift. Using the same statistical method, the 2-months functional recovery, utilized the same above predictors of 30-day mortality except the presence of gaze palsy and pupillary abnormality. Moreover, patient's gender, haematoma site and haematoma side were found to be insignificant predictors to both 30-day mortality and functional recovery at 2 months.

In the present study, high admission MAP was found, in univariate analysis, to be a main risk factor for 30-day mortality rate and unsatisfactory functional recovery at 2 months. Our data are going with those reported by Dondapani et al.¹³, who found that markedly elevated admission MAP > 145 mmHg in spontaneous ICH patients was associated with disruption of the cerebral autoregulation, which is accompanied by global cerebral hyperemia. They concluded that, once the diagnosis of hypertensive ICH has been made, it would appear that potent antihypertensive therapy to reduce blood pressure to a more normotensive range as quickly as possible is indicated, with recognition that the potential deleterious effect of hypoperfusion is outweighed by the effect of an elevated pressure within the intracranial circulation. On the other hand, Weisberg et al.¹⁴, had an opposite opinion, that most small haematomas and good outcome were due to hypertension, while large haematoma with poor outcome were often due to causes other than hypertension. The latter opinion was not going with our findings, where although lobar haematoma patients constituted only 26.7% of our hypertensive patients, no significant difference in either 30-day mortality rate and 2-months functional recovery was found between lobar haematoma patients and those of deep "putaminal and thalamic" locations which included the majority of our hypertensive patients.

With multivariate analysis, the size of haematoma as judged by initial CT findings and low admission GCS were significant independent predictors of both 30-day mortality and unsatisfactory functional recovery at 2 months among our patients, both parameters were already identified as important prognostic factors in previous studies^4,7,15,6, but most of these studies used only univariate analysis. The significance of these two parameters also stems from their easy to assess by simple bed side methods,

There is some controversy as to the importance of age as a prognostic factor⁷. Our study showed that old age was somewhat powerful significant predictor of both 30-day mortality and 2-months unsatisfactory functional recovery on using univariate analysis. On application of multivariate analysis, our data showed that old age was not significant independent predictor of 30-day mortality, but proved its significance as a powerful predictor of unsatisfactory functional recovery at 2 months. Our findings are going with those of a study⁷ reported that the 30-day mortality was correlated mainly with admission level of consciousness and the severity of the anatomic lesion on initial CT scan, whereas the later functional recovery was correlated mainly with the patient's age.

The prognostic role of IVE of the parenchymal hematoma is controversial¹⁶. Many authors^5,7,13,17 related IVE to a high case fatality rate and poor functional recovery. Our study demonstrated, in univariate analysis, a clear difference in outcome of IVE related to the haematoma location. Regarding lobar location, none of our patients with lobar hematomas who had a satisfactory functional outcome at 2 months had IVE, whereas IVE in such patients was associated significantly with unsatisfactory outcome. On the other hand, no significant difference in the functional recovery at 2 months was detected between patients with IVE versus those without IVE in both putaminal and thalamic locations.

Regarding the 30-days fatal outcome in relation to IVE, we found that lobar haematoma patients had the highest mortality rate, where among 5 patients with IVE in that location, 3 (60%) had a fatal outcome. Coming next, were those of putaminal hematomas who recorded a mortality rate of (27.27%), whereas thalamic haematoma patients with IVE were found to have the best chance for survival, where among 10 patients with IVE in that location, only 2 patients (20%) had a fatal outcome. Thus from our study, it can be argued that IVE is an adverse prognostic factor if associated particularly with lobar hematomas than with either putaminal or thalamic ones. The latter indeed was found to have the most optimistic outcome in association with IVE. Our results are in concordance with those reported by another study¹⁶.

The significant difference in outcome of IVE in relation to the haematoma location detected in our study may be largely correlated with the responsible parenchymal haematoma volume at both lobar and deep locations. Regarding lobar location, it was found that patients with IVE had a significantly larger mean parenchymal haematoma volume in comparison to that of patients without extension. On the other hand, regarding the putaminal and thalamic locations, no significant difference between mean parenchymal haematoma volumes of patients with IVE versus those without IVE was detected in either location.

Thus from our study, we could conclude that for a lobar haematoma to rupture into the ventricular system, it should attain a large volume, in contrast to the deep putaminal and thalamic hematomas where even a small sized haematoma can gain an access to the ventricular wall.

The above findings are going with that of Massaro et al.¹⁸, who found that deep hematomas had a greater frequency of IVE which remained constant with increasing volume, while for lobar hematomas, the frequency of IVE increased with enlarging haematoma volume.

On application of multivariate analysis, IVE was found to be an insignificant predictor to both functional recovery and case fatality rate. Our data are not going with that reported by Daverat et al.⁷, who found that IVE to be a poor prognostic factor in their multivariate analysis on 166 ICH patients.

The correlation that we observed can be used to answer some management questions during the acute stage. CT and clinical findings by themselves can help to predict which patients are at high risk of short term mortality and may require intensive treatment.

REFERENCES

1.      Kase CS, Caplan LR. Intracerebral hemorrhage. Boston: Butterworth-Heinemann; 1994.

2.      Thrift AG, Donnan GA, McNeil JJ. Epidemiology of intracerebral hemorrhage. Epidemiol rev. 1995; 17 (2): 361-81.

3.      Masè G, Zorzon M, Biasutti E, Tasca G, Vitrani B, Cazzato G. Immediate prognosis of primary intracerebral hemorrhage using an easy to use model for prediction of survival. Acta Neurol Scand. 1995; 91: 306-9.

4.      Lisk Dr, pastur W, Rhoades H, Putnam R, Grotta J. Early presentation of hemispheric intracerebral hemorrhage: prediction of outcome and guidelines for treatment allocation. Neurology. 1994; 44: 133-9.

5.      Portenoy R, Lipton R, Berger A, Lesser M, Lantos G. Intracerebral hemorrhage: a model for prediction of outcome. J Neurol Neurosurg Psychiatry. 1987; 50: 976-9.

6.      Broderick JP, Brott TG, Tomack T, Husler G. Volume of intracerebral Hemorrhage. A powerful and easy to use predictor of 30-day mortality. Stroke.  l993; 24: 987-93.

7.      Daverat P, Castel JP, Dartigues JF, Orgogozo J. Death and functional outcome after spontaneous ICH: a prospective study of 166 cases using multivariate analysis. Stroke. 1991; 22: 1-6.

8.      Jennet B, Bond M. Assessment of outcome after severe brain damage. Lancet. 1975; 1 (7905):480-4.

9.      Kothari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR, Zuccarello M, et al. The ABCs of measuring intracerebral hemorrhage volumes. Stroke. 1996; 27:1304-5,

10.    Rankin J. Cerebral vascular accidents in patients over the age of 60. III. Diagnosis and treatment. Scott Med J. 1957; 2 (6): 200-15,

11.    Ricci S, Celani MG, La Rosa F, Vitali R, Duca E, Ferraguzzi R, et al. SEPIVC: A community based study of stroke incidence in Umbria, Italy. J Neurol Neurosurg Psychiatry. 1991; 54: 696-8.

12.    Anderson CS, Jamrozik KD, Burvill PW, Chakera TM, Johnson GA, Stewart-Wynne EG. Determining the incidence of different subtypes of stroke: results from Perth community stroke study, 1989-1990. Med J Aust. 1993; 158: 85-9.

13.    Dondapani BK, Suzuld S, Kelley RE, Reyes-Iglesias Y, Duncan R. Relation between blood pressure and outcome in intracerebral hemorrhage. Stroke. 1995; 26: 21-4.

14.    Weisberg LA, Stazio A, Shamsnia M, Elliott D. Non-traumatic parenchymal brain hemorrhages. Medicine (Baltimore). 1990; 69: 277-95.

15.    Rådberg JA, Olsson JE, Rådberg CT. Prognostic parameters in spontaneous intracerebral hemorrhage with a special reference to anticoagulant therapy. Stroke. 1991; 22: 571-6.

16.    Lampl Y, Gliad R, Eshel Y, Sarvora-Pinhas I. Neurological and functional outcome in patients with supratentorial hemorrhages. A prospective study. Stroke. 1995; 15: 2249-53.

17.    Rosenow F, Hojer C, Meyer-Lohmann C, Hilgers RD, Mühlhofer H, Kleindienst A, et al. Acta Neurol Scand. 1996; 96:174-82.

18.    Massaro AR, Sacco RL, Mohr JP, FouIkes M, Tatemichi T, Price T, et al. Clinical discriminators of lobar and deep hemorrhages: the Stroke Data Bank. Neurology. 1991; 41: 1881-5.