Pooled analysis suggests benefit of catheter-based hematoma removal for intracerebral hemorrhage

The use of cloud based machine learning to predict outcome in intracerebral haemorrhage without explicit programming expertise

Machine Learning (ML) techniques require novel computer programming skills along with clinical domain knowledge to produce a useful model. We demonstrate the use of a cloud-based ML tool that does not require any programming expertise to develop, validate and deploy a prognostic model for Intracerebral Haemorrhage (ICH). The data of patients admitted with Spontaneous Intracerebral haemorrhage from January 2015 to December 2019 was accessed from our prospectively maintained hospital stroke registry. 80% of the dataset was used for training, 10% for validation, and 10% for testing. Seventeen input variables were used to predict the dichotomized outcomes (Good outcome mRS 0-3/ Bad outcome mRS 4-6), using machine learning (ML) and logistic regression (LR) models. The two different approaches were evaluated using Area Under the Curve (AUC) for Receiver Operating Characteristic (ROC), Precision recall and accuracy. Our data set comprised of a cohort of 1000 patients. The data was split 8:1 for training & testing respectively. The AUC ROC of the ML model was 0.86 with an accuracy of 75.7%. With LR AUC ROC was 0.74 with an accuracy of 73.8%. Feature importance chart showed that Glasgow coma score (GCS) at presentation had the highest relative importance, followed by hematoma volume and age in both approaches. Machine learning models perform better when compared to logistic regression. Models can be developed by clinicians possessing domain expertise and no programming experience using cloud based tools. The models so developed lend themselves to be incorporated into clinical workflow.

Thrombolysis experience in Costa Rica compared against individual patient data from two randomized controlled trials

OBJECTIVES

We sought to compare thrombolysis outcomes from the Costa Rican Stroke Registry Program (CRSRP) with published individual patient data from NINDS and CLOTBUST-ER trials using matching and outcome modeling from randomized clinical trials (RCTs).

MATERIALS AND METHODS

A retrospective observational study matching subjects on baseline characteristics, from the CRSRP, the control arm of CLOTBUST-ER, and the interventional arm of NINDS trials. Day 7-10/discharge modified Rankin Score (mRS), and early mortality was compared between matched subjects. A mortality model derived from RCTs was developed, and outcomes were compared at similar baseline NIHSS scores. CRSRP symptomatic hemorrhage (SICH) rate was compared with an Ibero-American cohort (IAC).

RESULTS

Of 540 CRSRP patients, 351 received rt-PA under 3 hours and were matched with NINDS subjects yielding 292 pairs; 161 CRSRP subjects treated within 4.5 hours were matched with CLOTBUST-ER subjects resulting in 151 pairs. The proportion of patients achieving excellent outcomes (mRS 0-1) did not differ between CRSRP and either NINDS or CLOTBUST-ER (CRSRP vs NINDS: 36.6% vs 32.9%, p=0.3; CRSRP vs CLOTBUST-ER: 26.5% vs 24.5%, p=0.8). Mortality was higher for CRSRP vs CLOTBUST-ER (7.3% vs 0.7%, p=0.006), but not vs NINDS (6.5% vs 4.5%, p=0.4). A pooled mortality model derived from 15 RCTs representing 4410 patients (R²=0.39) showed CRSRP and NINDS within expected mortality, while CLOTBUST-ER showed lower than expected mortality. CRSRP SICH rate equaled IAC (5.7% vs 5.7%; p=0.9).

CONCLUSIONS

Functional outcomes and SICH of thrombolysed Costa Rican patients compared favorably with published datasets, with a potential increase in early mortality.

Disability-Adjusted Life-Years Associated With Intracerebral Hemorrhage and Secondary Injury

Importance

Intracerebral hemorrhage (ICH) contributes significantly to the global burden of disease.

Objective

To examine the association of ICH and secondary injury with disability-adjusted life-years (DALYs) for the individual patient.

Design, Setting, and Participants

This cohort study was conducted using data from the Universitätsklinikum Erlangen Cohort of Patients With Spontaneous Intracerebral Hemorrhage study. Consecutive patients admitted to a single tertiary care center from January 1, 2006, to December 31, 2015, were included. The sample comprised patients with oral anticoagulation-associated ICH (OAC-ICH) or primary spontaneous ICH (non-OAC-ICH). Statistical analysis was conducted from October 1 to December 31, 2020.

Exposures

ICH occurrence and secondary injury.

Main Outcomes and Measures

DALYs, years of life lost (YLL), and years lived with disability (YLD) were analyzed by hematoma location, ICH volume, and secondary injury (ie, hematoma expansion [HE], intraventricular hemorrhage [IVH], and perihemorrhagic edema [PHE]).

Results

Among 1322 patients with ICH, 615 (46.5%) were women and the mean (SD) age at hospital admission was 71 (13) years; ICH was associated with a mean (SD) of 9.46 (8.08) DALYs, 5.72 (8.29) YLL, and 3.74 (5.95) YLD. There were statistically significant differences in mean (SD) DALYs by extent of hematoma volume (< 10 mL ICH: 7.05 [6.79] DALYs; 10-30 mL ICH: 9.91 [8.35] DALYs; >30 mL ICH: 12.42 [8.47] DALYs; P < .001) and ICH location (deep location: 10.60 [8.35] DALYs; lobar location: 8.18 [7.63] DALYs; cerebellum: 8.14 [6.80] DALYs; brainstem: 12.63 [9.21] DALYs; P < .001). Regarding population-level disease burden of secondary injuries after ICH, there was a statistically significant difference in mean (SD) by injury type, with 0.94 (3.19) DALYs for HE, 2.45 (4.16) DALYs for IVH, and 1.96 (2.66) DALYs for PHE (P < .001) among the entire ICH cohort. Regarding individual-level exposure to secondary injuries after ICH, there were a mean (SD) 7.14 (6.62) DALYs for HE, 4.58 (4.75) DALYs for IVH, and 3.35 (3.28) DALYs for PHE among patients with ICH affected by secondary injuries.

Conclusions and Relevance

These findings suggest that there is a high burden of disability associated with ICH and secondary injuries, and the findings may guide public health strategies. The study findings further suggest that IVH and PHE may be relevant for the overall outcome of patients with ICH, that DALYs may represent a viable outcome parameter for studies to evaluate treatment outcomes in ICH research, and that IVH and PHE may represent potential treatment targets.

The Chemical Basis of Intracerebral Hemorrhage and Cell Toxicity With Contributions From Eryptosis and Ferroptosis

Intracerebral hemorrhage (ICH) is a particularly devastating event both because of the direct injury from space-occupying blood to the sequelae of the brain exposed to free blood components from which it is normally protected. Not surprisingly, the usual metabolic and energy pathways are overwhelmed in this situation. In this review article, we detail the complexity of red blood cell degradation, the contribution of eryptosis leading to hemoglobin breakdown into its constituents, the participants in that process, and the points at which injury can be propagated such as elaboration of toxic radicals through the metabolism of the breakdown products. Two prominent products of this breakdown sequence, hemin, and iron, induce a variety of pathologies including free radical damage and DNA breakage, which appear to include events independent from typical oxidative DNA injury. As a result of this confluence of damaging elements, multiple pathways of injury, cell death, and survival are likely engaged including ferroptosis (which may be the same as oxytosis but viewed from a different perspective) and senescence, suggesting that targeting any single cause will likely not be a sufficient strategy to maximally improve outcome. Combination therapies in addition to safe methods to reduce blood burden should be pursued.

Decompressive Craniectomy in Spontaneous Intracerebral Hemorrhage: A Comparison with Standard Craniotomy Using Propensity-Matched Analysis

BACKGROUND

Spontaneous intracerebral hemorrhage carries a poor prognosis with a 30-day mortality rate of 35%-52%. There is no standardized surgical technique for treatment of spontaneous intracerebral hemorrhage. While minimally invasive techniques are popular, there has been renewed interest in decompressive craniectomy (DC). We compared surgical and functional outcomes of standard craniotomy and DC, both with hematoma evacuation, in the surgical treatment of supratentorial spontaneous intracerebral hemorrhage.

METHODS

This 4-year retrospective study compared outcomes of 2 surgical techniques: standard craniotomy in group A (n = 78) and DC in group B (n = 54). To minimize bias in case selection, propensity matching was performed to match preoperative Glasgow Coma Scale score and hematoma volume (group C).

RESULTS

Hematoma evacuation was performed in 132 patients. Mean age of patients was 53.3 years, 50.5 years, and 52.06 years in groups A, B, and C, respectively. Median preoperative Glasgow Coma Scale score was 9, 7, and 8 (P = 0.01; P = 0.45), and mean hematoma volume was 46.21 mL, 50.91 mL, and 49.90 mL. Overall mortality was 26.5%; 62.9% (n = 22) of deaths were in group A, and 37.1% (n = 13) were in group B (P = 0.69). Median modified Rankin Scale score was similar in both groups, both at discharge and at 3 months. After determining propensity scores, mortality and outcomes of matched groups remained similar.

CONCLUSIONS

DC with hematoma evacuation does not appear to provide a significant advantage over standard craniotomy with regard to functional outcomes and mortality. DC may overcome the need for subsequent surgery in accommodating postoperative mass effect in residual bleeds and rebleeds but is associated with greater blood loss and longer operative duration.

Surgery for spontaneous intracerebral hemorrhage - A comparative study with medical management in moderate to large sized hematomas

OBJECTIVES

Intracerebral Hemorrhage (ICH) is a devastating form of stroke and accounts for 10-15% of all cases. The management of ICH has predominantly been directed towards medical management. Multiple trials have failed to prove the superiority of surgical evacuation over conservative methods. However, surgery in a carefully selected set of patients is beneficial in reducing mortality and limiting disability. In this article, we retrospectively analysed our ICH register to compare the outcomes of surgical and conservative management of patients <70 years having a clot volume above 30 ml.

PATIENTS AND METHODS

We retrospectively analysed patients with ICH admitted at our centre between January 2015 and December 2017. A total of 119/624 patients with supratentorial hematoma volume >30 ml, GCS ≥ 5 and age less than 70 were included in this study.

RESULTS

The group was dichotomised into two groups A & B based on the management. Seventy-two (60.5%) patients underwent surgical intervention in group A and the remaining 47 (39.5%) were managed by best possible conservative methods in group B. The mean age in Group A was 51.01 years and 55.89 years in group B (P = 0.012). The volume of hematoma in the surgical group was 46.5 ± 14.9 ml in comparison to 38.53 ± 10.84 ml in the medically managed group (p = 0.002). Mortality at 90 days was 27/47 (57.44%) in the medically managed group while 23/72 (31.9%) in the surgical group (p = 0.006). Median mRS at discharge and 90 days were nearly identical and there was no significant difference in the dichotomized outcome among the two different management cohorts (p > 0.05). Mortality was the highest in the 30-50 ml medically managed group and >51 ml surgical group (p = 0.024). Age of the patient, GCS on presentation and medical management were independent predictors of mortality on logistic regression. The Cox Regression survival analysis of the two groups showed a clear survival advantage in the surgically managed group adjusting for age and GCS (p = 0.002) at 90 days.

CONCLUSION

Surgical Evacuation of spontaneous intracerebral haemorrhage has a survival advantage at 90 days in moderate to large sized hematomas. It, however, did not demonstrate any quantifiable improvement in functional outcome. Surgical evacuation of moderate-sized hematomas reduces mortality caused by delayed perihematomal oedema.