Evaluación de la intervención quirúrgica en pacientes con hemorragia intracerebral espontánea supratentorial con extensión intraventricular

Resumen

TITLE: SURVIVAL ANALYSIS OF SURGICALLY EVACUATED SUPRATENTORIAL SPONTANEOUS INTRACEREBRAL HEMORRHAGE WITH INTRAVENTRICULAR EXTENSION AUTHORS: R. Rivera-Fernández (a) F. Guerrero-López (b) D. Rodríguez-Rubio (c) F.J. Gómez-Jiménez (d) F. Rodríguez-Vilanova (e) J. Mora-Ordóñez (e) V. Olea-Jiménez (e) M.D. Arias-Verdú (e) G. Quesada-García (e) F. Acebal-Blanco(f) E. Castillo-Lorente (f) M.A. Arráez-Sánchez ( g) (a) Intensive Care Medicine. Hospital de la Serranía, Ronda. Málaga. (b) Intensive Care Medicine. Hospital Virgen de las Nieves, Granada. Spain. (c) Neurosurgery Department. Hospital del Mar, Barcelona. Spain. (d) Professor of Medicine. Universidad de Granada. Spain. (e) Intensive Care Medicine. Hospital Carlos Haya, Málaga. Spain. (f) Intensive Care Medicine. Hospital Neurotraumatológico, Jaén. Spain. (g) Neurosurgery Department. Hospital Carlos Haya, Málaga. Spain Corresponding author: Castillo-Lorente, Encarnación. Neurotraumatological Hospital of Jaén. Carretera Bailén-Motril s/n, 23009, Jaén. Spain. e.mail:encarnacioncastillo@gmail.com; encarnacion.castillo.sspa@juntadeandalucia.es. Fax number: (34)953008174. Phone number: (34)622167868 This research has not received any funding. ABSTRACT AND KEYWORDS: OBJECT: Survival study and evaluation of surgical treatment in a cohort of patients with diagnosis of supratentorial spontaneous intracerebral haemorrhage (ICH). MATERIALS AND METHODS: All consecutive patients with supratentorial ICH admitted to the Intensive Care Units of three Spanish hospitals with Neurosurgery Department were included between 2009-2012. Data collection: age, APACHE-II, Glasgow Coma Score (GCS) and pupillary anomalies on admission, intracerebral hemorrhage (ICH) score, location/volume of haematoma, intraventricular haemorrhage (IVH), surgical evacuation ± external ventricular drain, 30-day survival and at hospital discharge. RESULTS: 263 patients. Mean age: 59.74±14.14 years. GCS: 8±4 points, APACHE II: 20.7±7.68 points. ICH Score: 2.32 + 1.04 points. 30% had pupillary anomalies. 30-day Mortality: 51.3% (45.3% predicted by ICH-score), and 53.2% at hospital discharge. A significant difference (p=0.004) was observed in hospital mortality rates between surgically treated patients (39.7%, n=78) versus conservatively managed (58.9%, n=185); specifically in those with IVH surgically treated (34.2%, n=38) versus non-operated IVH (67.2%, n=125), p<0.001. No significant difference was noticed between mortality rates in patients without IVH. Multiple logistic regression analysis showed a OR for surgery of 1.04 (95% CI 0.33-3.22) in patients without IVH versus 0.19 (95% CI 0.07-0.53) in patients with IVH (decreased mortality with surgical treatment). The propensity score analysis for IVH patients showed improved survival of operated group (OR 0.23, 95% CI 0.07-0.75), p=0.01. CONCLUSIONS: Hospital mortality was lower in patients who underwent surgery in comparison to patients conservatively managed, specifically for the subgroup of patients with intraventricular haemorrhage. KEY WORDS: Intracerebral haemorrhage. Surgery. Treatment. INTRODUCTION Non-traumatic, spontaneous intracerebral haemorrhage (ICH) remains a major public health problem, being responsible of 9% to 27% of all strokes worldwide with an annual incidence of 10–30 per 100,0001-2. It is also a growing concern for health economies, with hospital admissions for ICH having increasing by 18% over the past 10 years3, likely attributable to population aging4. Among stroke patients, the highest mortality rate is observed in those with spontaneous ICH, with a 30 day mortality close to 50%2. Intraventricular haemorrhage (IVH) is a factor significantly associated with worse prognosis in patients with spontaneous ICH, carrying a reported mortality rate of 50-80%5-6. Several studies have confirmed that IVH is a prognostic factor7-9 including the major randomized STICH trial (Surgical Trial in IntraCerebral Haemorrhage)7 reporting the presence of IVH in 42% of ICH patients (23% of them with associated hydrocephalus). Good outcome was observed in only 15% of patients with IVH in comparison to 31% of ICH patients without IVH in a further study8. Infratentorial location represents 10-15% of spontaneous ICH and is also an independent predictor of outcome. In infratentorial ICH the role of surgery has been well described including clearly defined surgical indications in the recent guidelines of the American Heart Association/ American Stroke Association (AHA/ASA)10. Surgery is generally accepted as playing a major role in the management of patients with infratentorial ICH. Factors influencing the surgical decision making include conscious level, size of haematoma, presence of hydrocephalus and degree of any brain stem compression11. The role of medical versus surgical treatment in spontaneous supratentorial ICH remains controversial however, as does timing and type of neurosurgical intervention in this condition. Given the absence of strong scientific evidence to indicate surgery, this treatment is usually taken on a tailored manner based on surgeon preference in supratentorial locations12. Recent clinical guidelines10 state that early haematoma evacuation is not clearly beneficial when compared with haematoma evacuation when conscious level deteriorates (Class IIb; level of evidence A), and that supratentorial haematoma evacuation in deteriorating patients might be considered as life-saving measure (Class IIb; Level of evidence C). This was largely based on the findings of the STICH and STICH-II studies7,13, the two largest randomized trials comparing surgery to conservative management, neither which demonstrated a clear benefit for early surgical intervention. Three meta-analyses from the STICH authors14-16 have also reported that surgery improves prognosis when it is performed in selected groups of patients, the topic of surgical treatment in spontaneous supratentorial ICH therefore remains controversial. Based on this subgroup analysis our study hypothesis proposed that surgical treatment could improve survival outcomes in patients with supratentorial cerebral haemorrhage admitted to the intensive care unit (ICU). This hypothesis was supported by a previous non-published observation on a group of patients with supratentorial ICH admitted to Regional University Hospital of Málaga (Andalusia, Southern Spain) that showed better survival outcomes in the surgically treated group. We proposed a mutlicenter analysis of outcomes including two additional Hospitals with a Neurosurgery Department in the region. The objective of our study was to evaluate the benefit of surgery versus conservative management in survival outcomes in patients admitted to ICU with ICH. MATERIALS AND METHODS The study design is a multicenter cohort study including all patients presenting with a diagnosis of spontaneous supratentorial ICH from October 2009 to January 2012 to the ICUs of three Hospitals in Andalusia (Regional University Hospital of Málaga, University Hospital Virgen de las Nieves, Granada and Neurotrauma Hospital, Jaén). All participating Hospitals are tertiary referral centers for neurosurgical conditions in the region, providing comprehensive facilities for treatment of patients with acute stroke, including neurology, neuroradiology, neurosurgery and critical care. The inclusion criteria were patients over 14 years of age with supratentorial ICH confirmed on computed tomography (CT) imaging of the brain with or without intraventricular extension. Patients with a history of trauma, those with associated subarachnoid hemorrhage or underlying parenchymal lesions, and hemorrhages of infratentorial location were excluded. After admission patients underwent diagnostic imaging and treatment procedures according to the best clinical practice guidelines at the time of inclusion. Cardiovascular and respiratory stabilization of the patient was routinely performed on admission to the emergency department, standard assessment of clotting disorders was made and treated if required. General multimodal cardiovascular monitoring and intermittent pneumatic compression for prevention of venous thromboembolism beginning the day of hospital admission were considered standard of care. Intensive management of blood pressure with glycaemic, temperature and seizure control (no prophylactic anticonvulsant given) and early aggressive management of any medical complications were also routinely performed. Patients with a Glasgow Coma Score (GCS) ≤ 8, those with clinical evidence of transtentorial herniation, significant IVH or hydrocephalus were considered for intracranial pressure (ICP) monitoring. Management of raised ICP maintaining a cerebral perfusion pressure (CPP) of 50 to 70 mm Hg was performed depending on the status of cerebral autoregulation. The attending neurosurgeon was routinely consulted to assess the need for surgical haematoma evacuation and/or insertion of external ventricular drainage (EVD). The decision for surgical treatment was made following local guidelines for clinical practice or when the attending neurosurgeon considered it necessary. When a surgical decision was made for clot evacuation, a standard craniotomy with corticotomy and haematoma aspiration was carried out in all cases, with or without EVD placement. All patients were given the best available medical management, according to local policy and the updated AHA/ASA guidelines at time of patient inclusion in the study. For this cohort study data was collected by specifically-trained physicians in each hospital. Data collected at admission included age, admission GCS, Acute Physiology and Chronic Health Evaluation II (APACHE-II) scores, pupillary changes at ICU admission, haematoma volume17 (<30 ml, 30-50 ml, 50-100 ml and >100 ml) and location (lobar versus basal ganglia/thalamic), presence of IVH and ICH score18. Intervention and outcome data was subsequently collected including insertion of EVD and/or surgical evacuation of the clot, length of ICU admission and overall survival. Most of these variables were routinely registered in the clinical notes and easily identifiable from the electronic records with a low rate of exclusion of patients caused by incomplete or missed data. Data collection was prospective in Málaga (111 patients) and Granada (78 patients). Jaén Hospital (74 patients) joined the study later (January 2010) so data collection there was partially retrospective. We received approval from the Research Ethics Committee of each of the three participating hospitals. Data was analyzed anonymously. No consent form was provided because our study was observational and retrospective (in some cases as previously described). This issue was discussed and approved by the Ethics Committees of all three hospital involved in this study. Quantitative variables were expressed as means (± standard deviation) and qualitative variables as percentages and frequencies. The Student’s t-test was used to compare means and the chi-square test to compare proportions. Multivariate analysis was performed using a multiple logistic regression model. A propensity score matching was performed at the end of the study in patients with and without intraventricular extension of the haemorrhage to estimate the effect of the surgical treatment on both groups and to reduce bias due to confounding variables. Matching on the propensity score has previous been identified as a useful and reliable statistic tool that can provide an unbiased estimation of a treatment effect19. PSPP (psppire.exe 0.7.8.), and “R 2.14.2” were used for statistical analysis. A literature search was performed through Medline, with a search strategy using spontaneous, intracerebral, hemorrhage, supratentorial, surgical and treatment as medical subject headings (MeSH). The search was restricted to publications of the current century, and the original papers were obtained with authorized access to the electronic libraries of the Universities of Málaga and Granada. RESULTS The study included 263 patients: 111 from Málaga, 78 from Granada and 74 from Jaén Hospital. Mean age was 59.74 ± 14.14 years, GCS at ICU admission was 8 ± 4 points, and worst GCS on first ICU day was 7 ± 4 points. On admission APACHE-II severity score was 20.70 ± 7.68 points and ICH Score was 2.32 + 1.04 points. IVH was present in 163 (62%) of the 263 patients. 29.7% (n=78) underwent craniotomy and evacuation of the clot, and EVD was performed in 16.7% (n=44). Among 100 patients without IVH, 40 were operated through a standard craniotomy with corticotomy and haematoma evacuation. In the group of 163 patients with IVH, 38 were surgically treated (32 through craniotomy with clot evacuation alone and 6 with additional EVD insertion following clot evacuation). Of the 125 haematomas with IVH that were not surgically evacuated, 37 were treated with percutaneous insertion of EVD alone and 87 were managed conservatively. ICU stay was 8.74 ± 10.95 days, being ICU mortality 43.3%, with an overall hospital mortality of 53.2%. 30-day mortality rate was 51.3%, with a 30-day ICH score predicted mortality rate of 45.3%. 30% of patients (n=79), presented with pupillary anomalies (anisocoria or bilateral midriasis), with an observed 30-day mortality of 83% (ICH score predicted 30-day mortality of 62% in this group). In the majority of patients (n = 184) with no pupillary abnormalities on admission, the observed 30-day mortality was 37%, very close to the ICH score predicted mortality (38%). Patients who died during their hospital stay were older, had lower GCS on admission and on first day, higher APACHE-II scores and higher incidence of pupillary changes on admission. Additional factors increasing mortality were larger haematoma volume, the presence of IVH and conservative management (Table 1). Mortality rate significantly increased with age above 75 years and was significantly higher with GCS below 8. The relationship of these two variables with mortality was not linear (Table 2) therefore they were divided into intervals for the multivariate analysis. The global hospital mortality rate was 59.5% (97/163) in patients with IVH versus (vs.) 43% (43/100) in patients without (p=0.009), and it was 39.7% in the 78 patients who underwent clot evacuation surgery versus 58.9% in the 185 who did not (p=0.004). EVD was placed in 44 patients (1 patient with hydrocephalus without IVH and 43 patients with both ICH and IVH) and mortality rate was 51.2% in patients with EVD and 62% in patients without (not significant, n.s., p=0.1). Table 3 reports the characteristics of the groups with and without IVH, showing no difference in age, pupillary changes, APACHE-II score, and volume of haematoma. Although in patients with IVH, a lower GCS scores, higher ICH scores, higher ICU and hospital mortality rates, lower frequency of surgery (23.3% vs. 40%, p=0.004), and lower incidence on lobar haematoma was observed, in comparison to those patients without IVH. In the group of 163 patients with IVH, mortality was 34.2% (n=13) in the 38 who underwent craniotomy and evacuation of the clot versus 67.2% (n=84) in the 125 who did not (p<0.001). In the group of 100 ICH patients without IVH, mortality was 45% in the 40 who underwent clot evacuation surgery and 41.7% in the 60 who did not (n.s.; p=0.7). Therefore surgery did not diminish the mortality rate in patients without IVH but significantly reduced the mortality in those with IVH. Multiple logistic regression analysis of the data for the 163 patients with IVH showed that hospital mortality rate was significantly related to admission GCS, age, volume of haematoma, and surgery (odds ratio OR=0.14; confidence interval CI 95% 0.05-0.41). The variable “admission pupillary changes” (p=0.1) was not included in the multivariable model (Table 4). Mortality multiple logistic regression analysis of the data for all 263 patients in the study revealed a relationship between surgery and IVH: the OR for surgery was 1.04 (CI 95% 0.33-3.22) in patients without IVH (with no lower mortality observed comparing these with the medically treated group). The OR for surgery was 0.19 (CI 95% 0.07-0.53) for patients with IVH (lower mortality observed in the surgically treated group). This model also included the admission GCS, age, volume of haematoma and pupil abnormalities (Table 5). The “EVD” variable showed no significant relationship with mortality rate and therefore it was not considered for the multivariate analysis. A propensity score matching analysis was also performed in the group of 163 patients with IVH. For this purpose, we conducted a logistical multivariable analysis with a dependent variable of surgical treatment (yes/no) and different independent variables. These variables were GCS on admission, age, volume of haematoma (<30 ml, 30-50 ml, 50-100 ml and >100 ml), pupil changes (isochoric, anisocoria and bilateral non-reactive mydriasis), and haematoma site (lobar vs. basal ganglia/thalamic). With this model we assigned each patient a probability to receive surgical treatment (propensity score). The discrimination ability of this model was evaluated by the receiver operating characteristic (ROC) area under the curve. The ROC area under the curve had a value of 0.92 (0.88-0.97) for propensity score. In the group of 163 patients with IVH, 26 operated cases were matched with 26 from the conservatively managed group, with similar values of propensity score. No significant differences between both paired groups for potentially confounding variables was observed, as shown in Table 6. Mortality rate with this analysis was 30.8% in the surgical group (n=26), and 65.4% in the conservatively managed one (n=26), with a statistically significant p value of 0.01 (OR 0.23; CI 95% 0.07-0.75). A similar propensity score matching analysis was applied to patients with no IVH, showing no significant reduction in the mortality rate in the operated group (OR 0.83; CI 95% 0.25-2.76). DISCUSSION In our study mortality rate was lower in patients with spontaneous supratentorial ICH who underwent surgery in comparison to those conservatively managed particularly in the subgroup of patients with intraventricular haemorrhage which reached statistical significance. Indications of surgery for primary supratentorial ICH vary widely worldwide, reflecting uncertainty regarding best mangement20. The superiority of surgical treatment when compared to medical treatment is not well established, except in selected patient groups10-13, and early haematoma evacuation has not been shown to be beneficial in the two largest randomized trials regarding this topic7,13. In the STICH trial, with 1033 patients randomized to early surgery (< 24 hours) or initial conservative treatment, 26% of the patients in the surgical arm achieved a favorable outcome compared with 24% in the medical arm. STICH found no overall statistically significant difference not only in functional outcomes between treatment groups, but also in mortality. Subgroup analysis suggested that lobar haemorrhage within 1 cm of the cortical surface may benefit from surgery. The STICH II trial attempted to answer whether early surgery would be beneficial for conscious patients with superficial lobar haemorrhage of 10 to 100 cm3 within 1 cm of the cortical surface, without IVH and who were admitted within 48 hours of ictus. Favorable outcome was observed in 41% of patients in the early surgery group compared with 38% in the medical arm, and this difference was not statistically significant. In the STICH-II trial a subgroup analysis showed that poor prognosis patients were more likely to have a favorable outcome with early surgery, but there was no advantage to early surgery for patients in the good prognosis category. Exclusion of patients at risk for brain herniation, high crossover rates of patients to surgical intervention (up to 26% in the STICH trial), narrow patient-based inclusion criteria, and the focus on early surgery leave unclarified whether surgery may benefit specific group of patients with supratentorial ICH in these two major trials. In 2007, Mendelow y Unterberg14, analyzing 12 prospective randomized trials, found a reduction of the mortality rate in surgically treated patients (OR 0.85; CI 95% 0.71-1.02). A significant benefit was observed both in survival and functional status (OR 0.58; CI 95% 0.36-0.92) in the three studies focusing on superficial (lobar) haematomas. Prasad et al.15 reported in his meta-analisys a significant reduction on mortality and dependence rates (OR 0.71; CI 95% 0.58-0.88) in surgically treated patients. In 2012 Gregson et al.16, analyzing 8 trials (2186 patients), found that surgery improved prognosis when performed in the first 8 hours after presentation (p =0.003), with a haematoma volume of 20-50 cm3 (p= 0.004), GCS 9-12 (p= 0.0009), and in the age range of 50-69 years (p= 0.01). These meta-analyses emphasize the need of specific subgroup indications when considering surgical evacuation of a spontaneous supratentorial ICH. In our study, unlike in the STICH trial, hospital mortality rate was significantly lower (p=0.004) in the ICH patients who underwent surgery (39.7%) than in those who did not (58.9%). With a surgical approach recommended only in cases with an admission GCS of ≥ 5 points patients in the STICH trial were only randomized if the neurosurgeon was substantially uncertain about the need for surgical treatment. This design may have excluded patients who might have benefited from surgery limiting extrapolation of the results. More than half of the patients included in STICH-II were fully conscious or confused which represents only a small group in the total number of patients admitted to ICU with spontaneous ICH. In addition to those patients others with more severe disease and a higher predicted mortality rate were admitted in our study including those with adverse prognostic factors such as GCS <8 or intraventricular haemorrhage. It is specifically in this subgroup where we found that surgical treatment could be beneficial. A more detailed analysis of our results revealed that the difference in mortality rates was significant in patients with IVH (mortality rate 34.2% in surgically treated vs. 67.2% in conservatively managed, p=0.001) but not in those without (45% vs. 41.7%, respectively, p=0.742). Multivariate analysis confirmed that surgery only showed a beneficial effect in the patients with IVH (OR for surgery 0.19; CI 95% 0.07-0.53). The propensity score matching revealed a statistically significant benefit in the surgically treated group of ICH with intraventricular extension in survival terms (p=0.01), but not in the group without IVH. Our results mirror those in the STICH study in terms of improved outcomes following surgery in the presence of IVH, with or without hydrocephalus. The STICH group reported more favorable outcomes as measured by Glasgow Outcome Score at 6 months in patients with IVH receiving early surgery vs initial conservative management, although this difference was not significant8. Two clinical trials have confirmed that intraventricular hemorrhage and hydrocephalus are independent predictors of poor outcome in spontaneous ICH7,21-22. Blood in the ventricular system has multiple morbid effects23 on cerebral physiology including damaging the reticular activating system and releasing pyrogens. Hydrocephalus may arise through direct mass-effect with impaired flow of cerebrospinal fluid or because of the toxic effects of blood degradation products on the Pacchionian granulations. External drainage of cerebrospinal fluid through ventricular catheters reduces intracranial pressure, but clots in the catheter and infection prevent sustained beneficial effects on hydrocephalus and neurosurgical status in many patients, even with the use of intraventricular fibrinolytic agents24. Careful aspiration of the intraventricular clot when ICH extends into the ventricular system can be performed at the same time as clot evacuation. This may reduce mass-effect, ease clearance of ventricular blood and permit accurate positioning of a ventricular catheter utilizing direct vision of the ependymal wall. In our study surgical evacuation only enhanced the survival of ICH patients with IVH (the majority did not require an associated EVD) which may suggest that changes produced by the corticotomy and haematoma evacuation may aid the drainage of cerebrospinal fluid and ventricular blood postoperatively contributing to the improvement observed in this subgroup of patients. The pathophysiology of spontaneous supratentorial ICH is experimentally and clinically well documented25-27. There is some evidence to suggest the presence of a potentially recoverable penumbra of ischemic tissue around the ICH. The rationale for surgery in ICH is not only to remove the mass but also to reduce the toxic presence of intracerebral blood products, thereby promoting recovery of the penumbra. Our results suggest that surgery is effective against the mass effect of the haematoma, blood-clot mediated injury and assists the rapid clearance of intraventricular blood in ICH patients, thereby acting on the three key factors that determine prognosis in these patients. We would like to emphasize that our data is consistent with the most recent AHA/ASA guidelines which assert that supratentorial haematoma evacuation in deteriorating patients might be considered as a life-saving measure10. Latest European Stroke Organization (ESO) guidelines28 for the management of spontaneous intracerebral haemorrhages advises early surgery may be of value for patients with a GCS score 9–12 (moderate quality of evidence, weak strength of recommendation). The 30-day mortality rate in our study was 51.3%, with a 30-day ICH score predicted mortality rate of 45.3%. This difference can be explained by a higher rate of patients in our population with pupillary changes on admission with an observed mortality rate of 83% (ICH score predicted mortality rate of 62%) in this group. Patients without pupillary anomalies (70% of cases) showed a 30 day mortality rate of 37% in our study compared to 38% as predicted by the ICH score. The improvement in the survival rate observed with surgical treatment in our study is only applicable to a specific subgroup of patients (Table 6). The subgroup with intraventricular haemorrhage, mainly lobar haematomas, with a high severity score (as it is pointed by a ICH score predicted mortality rate 66.04%), with no cases of bilateral non-reactive mydriasis and an eldest age of 76 years, was the group of cases who did benefit from surgery in terms of survival. Limitations in our study include the lack of randomized assignment to conservative or surgical treatment although signs of irreversible cerebral herniation (GCS, pupillary changes) were controlled in multiple logistic regression analyses and with propensity score matching. Randomized controlled trials are generally considered the gold standard for assessing the efficacy of medications, medical procedures or clinical strategies. In observational studies, investigators do not control the treatment assignment, and large differences may exist between the two arms, in both observed and non-observed covariates. Rosenbaum and Rubin29 demonstrated that conditioning on the propensity score allows to obtain unbiased estimation of the treatment effect. The four most common techniques that may use the propensity score are: matching, stratification, regression adjustment and more recently weighting with the propensity score19. We have used matching on the propensity score, which is probably the most popular method and it allows the reader easily observe the characteristics of the treated and untreated patients. Our study did not analyze patient or relative wishes regarding medical or surgical treatment nor the attending surgeon decision making process in opting for surgery. We also only analyzed survival figures and no data regarding patients’ functional result was recorded. We considered the improvement in survival rates observed in patients surgically treated of sufficient interest to be reported before focusing on other longer term functional and neurological outcomes. CONCLUSIONS. Surgical haematoma evacuation might improve survival rates in patients with supratentorial spontaneous intracerebral haemorrhage particularly in those with intraventricular clot extension. This particular subgroup of patients is widely associated with poor outcome and any improvement in Neurosurgical management that may lead to better survival should be considered. REFERENCES 1.- Feigin VL, Forouzanfar MH, Krishnamurthi R et al. Global and regional burden of stroke during 1990–2010: findings from the Global Burden of Disease Study 2010. Lancet. 2014; 383:245–54. 2.- Labovitz DL, Halim A, Boden-Albala B, Hauser WA, Sacco RL. The incidence of deep and lobar intracerebral hemorrhage in whites, blacks, and hispanics. Neurology. 2005; 65:518–22. 3.- Qureshi AI, Suri MFK, Nasar A, Kirmani JF, Ezzedine MA, Divani AA, et al. Changes in cost and outcome among US patients with stroke hospitalized in 1990 to 1991 and those hospitalized in 2000 to 2001. Stroke. 2007;38:2180–4. 4.- Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol. 2003; 2:43–53. 5.- Steiner T, Diringer MN, Schneider D, Mayer SA, Begtrup K, Broderick J, et al. Dynamics of intraventricular hemorrhage in patients with spontaneous intracerebral hemorrhage: risk factors, clinical impact, and effect of hemostatic therapy with recombinant activated factor VII. Neurosurgery. 2006; 59:767–73. 6.- Hallevi H, Albright KC, Aronowski J, Barreto AD, Martin-Shild S, Khaja AM, et al. Intraventricular hemorrhage: anatomic relationships and clinical implications. Neurology. 2008; 70:848–52. 7.- Mendelow AD, Gregson BA, Fernandes HM, Murray GD, Teasdale GM, Hope DT, et al. Early surgery versus initial conservative treatment in patients with spontaneous supratentorial intracerebral haematomas in the International Surgical Trial in Intracerebral Haemorrhage (STICH): a randomised trial. Lancet. 2005;365:387–97. 8.- Bhattathiri PS, Gregson B, Prasad KS, Mendelow AD. Intraventricular hemorrhage and hydrocephalus after spontaneous intracerebral hemorrhage: Results from the STICH trial. Acta Neurochir Suppl. 2006 ;96 :65-8. 9.- Mayer SA, Brun NC, Begtrup K, Broderick J, Davis S, Diringer MN, et al. Efficacy and safety of recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med. 2008; 358:2127–37. 10.- Hemphill JC, Greenberg SM, Anderson CS, Becker K, Bendok BR, Cusman M, et al. Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2015;46:2032-60. 11.- Pérez-Núñez A, Alday R, Rivas JJ, Lagares A, Gómez PA, Alén JF, Lobato RD. Tratamiento quirúrgico de la hemorragia intracerebral espontánea. Parte II: Hemorragia infratentorial. Neurocir. 2008; 19:110-12. 12.- Pérez-Núñez A, Lagares A, Pascual B, Rivas JJ, Alday R, González P, Cabrera A, Lobato RD. Tratamiento quirúrgico de la hemorragia intracerebral espontánea. Parte I: Hemorragia supratentorial. Neurocir. 2008; 19:12-24. 13.- Mendelow AD, Gregson BA, Rowan EN, Murray GD, Gholkar A, Mitchell PM. Early surgery versus conservative treatment in patients with spontaneous supratentorial lobar intracerebral haematomas (STICH-II): a randomized trial. Lancet. 2013;382:397-408. 14.- Mendelow AD, Unterberg A. Surgical treatment of intracerebral haemorrhage. Curr Opin Crit Care. 2007;13:169-74. 15.- Prasad K, Mendelow D, Gregson B. Surgery for primary supratentorial intracerebral hematoma. A meta-analysis of 10 randomized controlled trials. Stroke. 2009;40:624-26. 16.- Gregson BA, Broderick JP, Auer LM, Batjer H, Chen XG, Juvela S, et al. Individual patient data subgroup meta-analysis of surgery for spontaneous supratentorial intracerebral hemorrhage. Stroke. 2012;43:1496-504. 17.- Broderick JP, Brott TG, Grotta JC. Intracerebral hemorrhage volume measurement. Stroke 1994; 25: 1081. 18.- Hemphill JC 3rd, Bonovich DC, Besmertis L, Manley GT, Johnston SC. The ICH score: a simple, reliable grading scale for intracerebral hemorrhage. Stroke. 2001; 32(4):891-7. 19.- Gayat E, Pirrachio R, Resche-Rigon M, Mebazaa A, Mary J-Y, Porcher R. Propensity scores in intensive care and anaesthesiology literature: a systematic review. Intensive Care Med 2010;36: 1993-2003. 20.- Gregson B, Mendelow AD. STICH investigators. International variations in surgical practice for spontaneous intracerebral hemorrhage. Stroke. 2003; 34:2593-97. 21.- Mayer SA, Brun NC, Begtrup K, et al. Recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med. 2005; 352: 777–85. 22.- Mayer SA, Brun NC, Begtrup K, Broderick J, Davis S, Diringer MN, et al. Efficacy and safety of recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med. 2008; 358:2127–37. 23.- Qureshi AI, Mendelow D, Hanley DF. Intracerebral hemorrhage. Lancet. 2009;373: 1632-44. 24.- Morgan T, Awad I, Keyl P, Lane K, Hanley D. Preliminary report of the clot lysis evaluating accelerated resolution of intraventricular hemorrhage (CLEAR-IVH) clinical trial. Acta Neurochir Suppl. 2008; 105: 217–20. 25.- Nakamura T, Keep RF, Hua Y, Shallert T, Hoff TJ, Xi G. Deferoxiamine-induced attenuation of brain edema and neurological deficits in a rat model of intracerebral hemorrhage: the role of iron. J Neurosurg. 2006;104:305-12. 26.- Xi G, Keep RF, Hoff JT. Mechanisms of brain injury after intracerebral haemorrhage. Lancet Neurol. 2006;5:53-63. 27.- Nilsson OG, Polito A, Saveland H, Ungerstedt U, Nordström CH. Are primary supratentorial intracerebral hemorrhages surrounded by a biochemical penumbra? A microdialysis study. Neurosurgery. 2006;59:521-528. 28.- Steiner T, Al-Shahi Salman R, Beer R, Christensen H, Cordonnier C, Csiba L, et al. European Stroke Organisation (ESO) guidelines for the management of spontaneous intracerebral hemorrhage. Int J Stroke. 2014;9:840-55. 29.- Rosenbaum P, Rubin D (1983). The central role of the propensity score in observational studies for causal effect. Biometrika 70:41–55 Table 1.- Relationship between mortality and rest of variables (age expressed in years, GCS and APACHE-II in points, length of ICU stay in days). Variables Survivors (n=123) Non Survivors (n=140) p Age 55.16 + 14.76 63.76 + 12.27 <0.001 Admission GCS 10.25 + 3.78 6.51+ 3.42 <0.001 Worst GCS on first ICU day 8.84 + 3.93 5.31 + 2.87 <0.001 Length of ICU stay 11.41 + 12.26 6.41 + 9.08 <0.001 APACHE-II 17.06 + 7.18 23.91 + 6.61 <0.001 ICH score 1.82 + 0.94 2.75 + 0.91 <0.001 Admission pupillary changes <0.001 Normal 90.2%(n=111) 52.1% (n=73) Anisocoria 9.8% (n=12) 26.4% (n=37) Bilateral non-reactive mydriasis 0% (n=0) 21.4% (n=30) Volumen of haematoma* 0.001 <30 ml 36.5% (n=42) 20.5% (n=26) 30-50 ml 25.2% (n=29) 20.5% (n=26) 50-100 ml 33.9% (n=39) 42.5% (n=54) >100 ml 4.3% (n=5) 16.5% (n=21) Site of haematoma 0.6 Lobar 55.7% (n=68) 52.5% (n=73) Basal ganglia/thalamic 44.3% (n=54) 47.5% (n=66) Intraventricular haemorrhage 53.7% (n=66) 69.3% (n=97) 0.009 External ventricular drain 17.1% (n=21) 16.4% (n=23) 0.8 Surgical treatment 38.2% (n=47) 22.1% (n=31) 0.004 *Haematoma volume estimation was missed in 19 patients Table 2.- Hospital mortality rates according to cathegorized GCS and age, confronted to patients with and without intraventricular haemorrhage (IVH). Variables Total (n=263) No IVH (n=100) IVH (n=163) Admission GCS (a) 14 - 15 41 (17.1%) 21 (9.5%) 20 (25%) 9 - 13 69 (33.3%) 30 (26.7%) 39 (38.5%) 6 - 8 71 (63.4%) 24 (54.2%) 47 (68.1%) < 6 82 (79.3%) 25 (80%) 57 (78.9%) Age (b) <50 62 (35.5%) 32 (34.4%) 30 (36.7%) 50 - 74 163 (52.1%) 54 (38.9%) 109 (58.7%) > 75 38 (86.8%) 14 (78.6%) 24 (91.7%) (a),(b): Statistically significant differences in mortality among the different admission GCS intervals (p<0.001) and different age intervals (p<0.05) in patients with and without IVH. Table 3.- Relatioship between intraventricular haemorrhage (IVH) and rest of variables. Variables No IVH (n=100) IVH (n=163) p Age 58.09 + 14.61 60.75 + 13.78 0.1 Admission GCS 9.06+ 4.18 7.77 + 3.75 0.01 Worst GCS on first day in ICU 7.63 + 4.09 6.56+ 3.62 0.02 APACHE-II 19.54 + 7.82 21.42 + 7.55 0.05 ICH score 1.61 + 0.84 2.75 + 0.9 <0.001 Length of ICU stay 6.26 + 6.87 10.27 + 12.61 0.004 Surgical treatment 40% (n=40) 23.3% (n=38) 0.004 Admission pupillary changes 0.3 Normal 72% (n=72) 68.7% (n=112) Anisocoria 20% (n=20) 17.8% (n=29) Bilateral non-reactive mydriasis 8% (n=8) 13.5% (n=22) Volume of haematoma* 0.3 <30 ml 20.9% (n=18) 32.1% (n=50) 30-50 ml 26.7% (n=23) 20.5% (n=32) 50-100 ml 40.7% (n=35) 37.2% (n=58) >100 ml 11.6% (n=10) 10.3% (n=16) Site of haematoma <0.001 Lobar 68.7% (n=68) 45.1% (n=73) Basal ganglia/thalamic 31.3% (n=31) 54.9% (n=89) ICU mortality 35% (n=35) 48.5% (n=79) 0.03 Hospital mortality 43% (n=43) 59.5% (n=97) 0.009 *Haematoma volume estimation was missed in 19 patients Table 4.- Multivariant analysis of mortality in patients with intraventricular haemorrhage (n=163). Variables OR CI 95% Surgical evacuation 0.14 (0.05-0.41) Age <50 1 50 - 74 2.09 (0.78-5.58) > 75 18.72 (2.66-131.14) GCS on admission 14-15 1 9 - 13 2.87 (0.6-13.68) 6 - 8 9.07 (1.98-41.38) < 6 10.19 (1.92-53.96) Volumen <30 ml 1 30-50 ml 4 (1.32-12.15) 50-100 ml 3.45 (1.2-9.97) >100 ml 19.87 (2.46-160.07) Table 5.- Multivariant analysis of mortality in patients with supratentorial spontaneous intracerebral haemorrhage (n=263). Variables OR CI 95% GCS on admission 14-15 1 9 - 13 2.32 (0.78-6.95) 6 - 8 5.71 (1.9-17.16) < 6 7.23 (1.99-26.56) ICU admission pupils * 3.52 (1.47-8.43) Volumen of haematoma <30 ml 1 30-50 ml 2.21 (0.88-5.51) 50-100 ml 3.51 (1.42-8.699 >100 ml 7.40 (1.7-32.31) Age (years) <50 1 50 - 75 1.38 (0.64-2.97) > 75 8.78 (2.39-32.26) IVH 3.84 (1.47-10.04) Surgical evacuation IVH 0.19 (0.07-0.53) No IVH 1.04 (0.33-3.22) *ICU admission pupil are codified as 0 “Normal” and 1 “Anisocoria or Bilateral non-reactive midriasis” Table 6.- Propensity score matching in two paired groups of 26 patients with intraventricular hemorrhage (surgical vs. medical treatment). No significant difference can be seen between both groups in considered variables except in mortality. Variables Medical group (n=26) Surgical group (n=26) p Age * [1] 59.92 + 11.64 58.31 + 12.65 0.6 Admission GCS (A) * 8.5 + 3.5 8.6 + 3.4 0.9 Worst GCS on first ICU day (W) 6.9 + 3.5 6.9 + 3.4 0.9 Difference A-W 1.6 + 2.2 1.7 + 3.5 0.9 APACHE-II 19.27 + 7.1 20.85 + 5.7 0.4 ICH score 2.88 + 0.65 2.92 + 0.56 0.8 Volume of hematoma * < 30 ml 30-50 ml 50-100 ml >100 ml 7.7% (n=2) 34.6% (n=9) 50% (n=13) 7.7% (n=2) 3.8% (n=1) 26.9% (n=7) 61.5% (n=16) 7.7% (n=2) 0.8 Site of hematoma * Lobar Basal Ganglia/thalamic 69.2% (n=18) 30.8% (n=8) 73.1% (n=19) 26.9% (n=7) 0.8 Left side of hematoma 42.3% (n=11) 46.2% (n=12) 0.8 ICU admission pupils * Normal Anisocoria Bilateral non-reactive mydriasis 88.5% (n=23) 11.5% (n=3) 0% (n=0) 88,5% (n=23) 11,7% (n=3) 0% (n=0) 1 Mortality Predicted by ICH score 63.43 + 24.82 66.04 + 21.47 0.7 Hospital Mortality 65.4% 30.8% 0.01 *Variables included in propensity score calculation are shown with an asterisk. [1] Eldest age was 76 years in the surgically treated group, and 78 years in the conservatively managed one.