Retrospective analysis of the surgical management of spontaneous supratentorial intracerebral hemorrhage: A single-center study

Background: Intracerebral hemorrhage (ICH) remains a devastating disease with high morbidity and mortality. The mortality rate ranges from 40% at 1 month to 54% at 1 year, and only 12%–39% achieve good outcomes and functional independence. The current management guidelines for spontaneous supratentorial ICH are still controversial. Objective: Patients who presented with ICH and underwent craniotomy with hematoma evacuation or minimal procedures from January 2016 to May 2020 were included in the analysis. Several clinical, radiological, and surgical variables were collected to identify the variables most likely related to lower mortality and better functional outcomes. Results: A total of 87 patients presented with HMC with ICH from January 2016 to May 2020. The mean age was 44.7 (42.2–47.2) years. There were 76 male (87.4%)/11 female (12.6%) patients, which reflect the population distribution in Qatar, which is mainly male predominant. Although Qatar is mainly a Middle-Eastern country, the ethnic distribution of patients was mainly of South Asian and Indian (60.9%) and Far-Eastern (20.7%) ethnicities because of diversity. The mean baseline Glasgow coma scale (GCS) was 8.2+/ − 3.7. The mean baseline functional independence measure (FIM) score was 59.4+/ − 36.7. Most hematomas were located in the basal ganglia (79.3%%). Baseline characteristics based on long-term outcomes are summarized in Table 1. The following variables were correlated with poor outcome: low GCS on postoperative day 1 (P = 0.06), low FIM score (P = 0.006), ICH location (P = 0.04), distance of the closest point to the surface (P = 0.009), and presence of uncal herniation (P = 0.04). The baseline characteristics based on mortality are outlined in Table 2. The following variables were correlated with mortality: diabetes mellitus (P = 0.02), baseline GCS (P = 0.04), GCS on postoperative day 1 (P = 0.002), unequal pupils (P = 0.05), and postoperative midline shift (P = 0.001). Conclusion: The preoperative clinical neurological status as well as mass effect (uncal herniation and midline shift) can be determinants of functional outcome and mortality. A deeper hematoma may result in poor functional outcome because of more surgical damage in functional brain tissues. Thus, the goal of surgery in spontaneous supratentorial ICH is to reduce mortality, but no evidence support that it can improve functional outcome. Although our findings are interesting, more prospective studies with a larger sample size are needed to confirm our results.

. Baseline characteristics based on long-term outcomes  . STICH-I found no difference between the two groups regarding mortality and functional outcome, unlike STICH-II, which suggested that early surgery may have a survival advantage if hemorrhage is superficial and no intraventricular hemorrhage (IVH) is present 1, 4-6 . Recently, some trials have compared Pupils preop reactive vs non Reactive Pupils preop equal vs non Equal Distance of the closest point to the surface, cm Brain stem Absent No evacuation only decompression Retrospective analysis of the surgical management of spontaneous supratentorial intracerebral hemorrhage Shaaban et al   I, MISTIE-II, MISTIE-III, CLEAR, and SLEUTH) [7][8][9][10][11] . The peroxisome proliferator-activated receptor gamma agonist pioglitazone has been evaluated in some MIS trials; theoretically, it enhances phagocytosis and improves oxidative stress and inflammation 12 . Therefore, based on the available literature, we are on the path but still have not reached our objective of establishing guidelines for the best management of spontaneous supratentorial ICH to provide the best care of our patients and thus achieve low mortality and better functional outcomes. Our state has a multicultural demography in its population, making it an excellent region to study the etiology and mechanism of ICH. In Qatar, 84% of the population are male because of the large majority of expatriates, and 82% are ,65 years old. In addition, the entire state is covered by a major tertiary hospital that receives most of the stroke cases in the country, allowing for a population-based data source. Thus, this study aimed to analyze and identify clinical, radiological, surgical variables probably associated with less mortality and better functional outcome by comparing our results with those of previous studies. We hope that the study findings can guide further research and help establish clear guidelines for the management of spontaneous supratentorial ICH.

METHODS
This study was a retrospective collection of the data of patients who presented with ICH to our hospital between January 2016 and May 2020. This study was approved by the local institutional review board and medical research committee. Patients were included in the study if they fulfilled the following criteria: 1) diagnosis of spontaneous supratentorial ICH in patients who underwent craniotomy with hematoma evacuation plus or minus decompression or patients who underwent decompression alone; 2) volume . 30 ml or midline shift (MLS

RADIOLOGICAL VARIABLES
Patients' plain head CT scans were analyzed preoperatively and postoperatively to identify the following data: location (basal ganglia, cortical, or thalamus); involved hemisphere (right vs left); ICH volume (cm 3 ) measured using the method with the largest length in three dimensions divided by two (equation abc/2); preoperative and postoperative distance of the closest point of hematoma to the brain surface (cm), MLS (mm) measured at the level of the septum pellucidum from a midline drawn at the level of the attachments of the falx cerebri anterior and posterior to inner skull tables; preoperative and postoperative edema (mild, moderate, or severe); brain stem involvement or noninvolvement; presence or absence of IVH preoperatively and postoperatively; presence or absence of hydrocephalus preoperatively and postoperatively; uncal herniation; and residual volume postoperative (no residual, near-total evacuation, partial evacuation, or no evacuation only decompression).

SURGICAL VARIABLES
The following surgical data were collected: type of surgery (craniotomy with hematoma evacuation, craniotomy with hematoma evacuation þ decompression, or decompression alone), external ventri-cular drain was inserted or not inserted, intracranial pressure (ICP) monitoring was performed or not performed, brain navigation was included or not included during surgery, time from initial plain CT of the head to surgery (hours), and adverse events (redo surgery, infection, or none).

CLINICAL OUTCOMES
Good long-term outcomes were defined in this study: mRS of 0-2 at 3 months. The mortality rate was measured at 30 and 90 days. Functional independence measure (FIMs) is a widely accepted functional assessment measure that is used during inpatient rehabilitation. The FIM is an 18-item ordinal scale that is used in all diagnoses within a rehabilitation population. FIM scores range from 1 to 7 (1 ¼ total assistance; 7 ¼ complete independence). Scores , 6 require the help of another person for supervision or assistance. The FIM assesses independent performance in self-care, sphincter control, transfer, locomotion, communication, and social cognition. By adding the points for each item, the possible total scores for the level of independence range from 18 (lowest) to 126 (highest) points.

STATISTICAL ANALYSIS
By using SPSS for statistical analysis, data were coded with numbers (e.g., 0, 1, 2, 3 . . . ). Data were classified into either categorical or continuous numerical variables. Dependent variables were 30day mortality and modified Rankin scale (mRS) (further classified as good outcome [0-2] or poor outcome [3][4][5][6]). All other independent variables were related to 30-day mortality and mRS. The chisquare test was used for categorical variables, and the one-way analysis of variance was used for continuous numerical variables. Logistic regression analysis was performed to measure the odds ratio (ORs) of mortality and poor outcome, with adjustment for common risk factors, and presented as OR with confidence interval 95%. Continuous data were checked for normality using means^standard deviation. P value , 0.05 was considered significant. The procedures performed were craniotomy þ hematoma evacuation (78.2%), craniotomy þ hematoma evacuation þ decompression (13.8%), and decompression alone (8.0%). The mortality, survival, good outcomes (3-month mRS , 3 points), and poor outcome rates (mRS 3-6 points) were 8% (n ¼ 7), 92% (n ¼ 80), 11.5% (n ¼ 10), and 88.5% (n ¼ 77), respectively. Baseline characteristics based on long-term outcomes are summarized in Table 1. The following variables were correlated with poor outcome: low GCS on postoperative day 1 (P ¼ 0.06), low FIM (P ¼ 0.006), ICH location ICH (P ¼ 0.04), distance of the closest point to the surface (P ¼ 0.009), and presence of uncal herniation (P ¼ 0.04) ( Table 1). The baseline characteristics based on mortality are outlined in Table 2. The following variables were correlated with mortality: DM (P ¼ 0.02), baseline Odds ratio (OR) with 95% confidence interval GCS (P ¼ 0.04), GCS on postoperative day 1 (P ¼ 0.002), unequal pupils (P ¼ 0.05), and postoperative MLS (P ¼ 0.001) ( Table 2). In the multiple logistic regression analysis with poor outcome as the dependent variable, uncal herniation (P ¼ 0.06; adjusted OR, P ¼ 0.07) and location in the basal ganglia (P ¼ 0.02; adjusted OR, P ¼ 0.5) were significantly correlated with poor outcomes (Table 3).

RESULTS
In the multiple logistic regression analysis with mortality as the dependent variable, DM (P ¼ 0.021; adjusted OR, P ¼ 0.02), and equal vs unequal preoperative pupils (P ¼ 0.04; adjusted odds ratio, P ¼ 0.9) were significantly correlated with mortality (Table 4).

DISCUSSION
This study showed that baseline clinical radiological and surgical findings can help predict clinical outcomes in patients with ICH. The presence of neuroimaging findings of ICH location and uncal herniation correlated with poor outcomes. However, the clinical status, especially GCS on postoperative day 1, of the patient was more related to both outcome and mortality. Our finding that the distance of the closest point to the surface in craniotomy correlated with outcome is an interesting observation.
In this study, we found that the deeper the hematoma, the worse the surgical outcome. This could advocate for minimally invasive procedures as the ideal method of surgical treatment, which is the prevailing concern in current randomized controlled trials.
Decisions regarding surgical intervention in spontaneous supratentorial ICH are still challenging among neurosurgeons in day-to-day practice. On patient evaluation, the most significant epidemiological, clinical, radiological, and surgical factors that can predict the overall mortality and functional outcome of every patient are still unknown. In our study, we tried to evaluate our experience and analyze as many factors as possible to identify the trends and possible associations between different factors and outcomes. In a previous study, Kim et al. compared patients who underwent craniotomy with hematoma evacuation and patients who underwent decompressive craniectomy in addition to hematoma evacuation. Factors associated with less 30-day mortality were preoperative GCS . 9, hematoma volume , 80, and time from ictus to surgery, while predictors of better  13 . Different theories can explain the finding that MLS on postoperative day 1 had a significant association with mortality (the more severe the MLS on postoperative day 1, the more likely the patient will die). First, hematoma evacuation will reduce all factors mentioned previously that are related to cytotoxic brain edema. Second, clotted hematoma has less mass effects and consequently low ICP 14,15 .
The distance from the closest point to the surface was 5.7 times more likely to be associated with a poor outcome (P ¼ 0.009), which is an interesting finding. This means that the closer the hematoma to the surface, the better the functional outcome. According to the STICH-II trial, early surgery can have a small clinical survival benefit for superficial hematoma without IVH, but no evidence can support the improvement in functional outcomes 5 .
Based on the trend that superficial hematoma is associated with better functional outcomes, we can suggest that the less normal brain tissue damage, the better the postoperative functional outcome and better prognosis. By translating that theory into the clinical practice, we can advocate for minimally invasive ICH evacuation, especially for deep-seated hematoma, to protect normal brain tissue as much as possible, in contrast to open surgery where dissection and retraction can cause significant brain damage 3 .
Recently, many trials, including MISTIE-II and MISTIE-III, have compared minimally invasive hematoma evacuation with conservative management. The MISTIE technique aimed to decrease the hematoma size to , 15 ml: a rigid cannula was inserted through a burr hole, and clot aspiration was performed with a 10-ml syringe until there was resistance. Thereafter, the cannula was replaced by a catheter under imaging guidance into the hematoma. The placement was confirmed by CT at 6 h after intrathecal administration of alteplase 1 mg in 1 ml of saline every 8 h, followed each time by a 3-ml saline flush and closure of the catheter for 1 h. The goal was to achieve a hematoma volume , 15 ml or to administer nine doses of alteplase, whichever came first. Although MISTIE proved to be a safe technique, it failed to provide better functional outcomes than conservative management 3,7,8 . Some trials are ongoing, including SCUBA (stereotactic ICH underwater blood aspiration, ENRICH trial (early minimally invasive removal of ICH), INVEST trial (minimally invasive endoscopic surgical treatment with Apollo/Artemis in patients with brain hemorrhage), and MIND trial. We hope that they can show promising results toward ICH management more in the direction of MIS 16,17 .
Postoperatively, DM and MLS were the two factors most significantly related to 30-day mortality. According to 2015 guidelines from the American Stroke Association, hyperglycemia and hypoglycemia should be avoided (Level C evidence: Class 1) 1 . Two previous studies have concluded that hyperglycemia on admission, even in patients with no DM history, increased the risk of mortality and poor functional outcome 18,19 . Based on our findings, patients with DM were 12.972 times more likely to die within 30 days than patients without DM. GCS on postoperative day 1 and preoperative equal or unequal pupils were two factors related to 30-day mortality. The same was true for uncal herniation and location in the basal ganglia in relation to functional outcome. However, these trends can be investigated by additional large studies. The type of surgery did not show any difference, but interestingly, no deaths were recorded in patients who had decompression þ hematoma evacuation or patients who underwent decompression alone. In a review by Mendelow 20 , a meta-analysis of three STITCH trials (STITCH I, STITCH II, STITCH Trauma) showed that patients treated with craniotomy with initial GCS 9-12 may benefit from surgery, especially with early intervention. Gregson et al. reviewed and performed a metaanalysis of trials conducted for surgical management of ICH until 2010 (14 trials included). They suggested that better outcomes can be achieved with early surgery within 8 h, hematoma 20-50 ml, GCS $9, and age 50 -69 years 21 . This study has several limitations. First, this study has a retrospective design and prone to the effects of confounders. Second, as it is a single-center study, the results cannot be generalized. Third, the sample size was small; a larger sample with a multicenter registry is needed to confirm our findings. Fourth, the sample size was not enough to study all potential factors or variables. Owing to the mentioned limitations, our study's findings can only be used for hypothesis generation rather than drawing firm conclusions.
In conclusion, clinical neurological status preoperative as well as mass effect (uncal herniation and MLS) can be determinants of functional outcome and mortality. In addition, a deeper hematoma may result in poor functional outcome because of more surgical damage in functional brain tissues. The goal of surgery in spontaneous supratentorial ICH is to reduce mortality, but no evidence supports it can improve functional outcome. Although our findings are interesting, more prospective studies with a larger sample size are needed to confirm our results.