Summary
Background
There is ongoing controversy as to whether surgical intervention to haematoma evacuation benefits patients with acute intracerebral haemorrhage (ICH). This study aimed to evaluate the association of surgical intervention to evacuate the haematoma and 6-month functional outcome in participants of the third Intensive Care Bundle with Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial (INTERACT3).
Methods
This was a secondary analysis of INTERACT3, which enrolled adults (age ≥18 years) spontaneous ICH patients within 6 h after onset. INTERACT3 was an international, multicentre, prospective, stepped-wedge, cluster randomised, blinded outcome assessed, clinical trial undertaken at 121 hospitals in 10 countries between December 12, 2017 and December 31, 2021. To limit heterogeneity in the results, we restricted analyses to participants in China. The primary outcome was poor functional outcome, defined by a score of 5–6 on the modified Rankin scale (mRS), at 6 months. Secondary outcomes include a mRS score of 4–6 and mortality at 6 months. Sensitivity analysis included propensity score matched analysis and the imputation of missing outcome variables. The effect of timing on surgical outcome was also evaluated. The INTERACT3 trial was registered at ClinicalTrials.gov (NCT03209258) and CHiCTR.org.cn (ChiCTR-IOC-17011787).
Findings
Of 5772 participants (mean age 62.0 ± 12.5 years) at 82 sites in China, 1411 (24.4%) received surgery in which craniotomy (72.6%) was the most common approach. After adjustment for confounding variables, surgery to evacuate the haematoma was associated with lower odds of a poor functional outcome (odds ratio 0.71, 95% CI 0.55–0.92; p = 0.010) and mortality (odds ratio 0.55, 95% CI 0.40–0.75; p = 0.0001) at 6 months. The association was consistent in propensity score matching analysis and sensitivity analysis by imputation. We did not detect significant differences in outcome between those who received surgery on the same day of hospital arrival compared to those who received surgery on the second or later days. In analysis limited to participants with supratentorial ICH and with a haematoma volume 30 mL or more, evacuation of the haematoma was associated with lower odds of poor functional outcome (n = 1234, odds ratio 0.68, 95% CI 0.46–0.99; p = 0.042) and mortality (n = 1291, OR 0.45, 95% CI 0.29–0.69; p = 0.0003).
Interpretation
This secondary analysis of the INTERACT3 indicates that evacuation of the haematoma is associated with better chances of surviving free of severe disability after acute ICH. With the evolution of instrument and techniques, further trial should address the role of haematoma evacuation in deep ICH patients, the time window and difference between mini-invasive techniques.
Funding
Joint Global Health Trials (JGHT) funding scheme from the Department of Health and Social Care, the Foreign, Commonwealth & Development Office, the Medical Research Council and Wellcome Trust; the West China Hospital Outstanding Discipline Development 1–3-5 programme; National Health and Medical Research Council of Australia; Sichuan Credit Pharmaceutical; and Takeda (China).
Keywords: Intracerebral haemorrhage, Haematoma evacuation, Secondary analysis, INTERACT3
Research in context.
Evidence before this study
We searched PubMed (from January 1, 1970, to April 20, 2025) and Embase (from January 1, 1947, to April 20, 2025), with a restriction to English-language publications using the following terms: “intracerebral haemorrhage OR haemorrhagic stroke” AND “surgery OR surgical treatment OR craniotomy OR endoscopy OR aspiration OR minimally invasive surgery” AND “multicentre OR multisite”. We identified several high-quality trials on haematoma evacuation for spontaneous intracerebral haemorrhage (ICH), while only the Early Minimally Invasive Removal of Intracerebral Hemorrhage (ENRICH) trial reported a positive effect of early haematoma evacuation in 300 ICH patients, which is appeared to be attributable to lobar but not deep ICH. The Swiss Trial of Decompressive Craniectomy vs. Best Medical Treatment of Spontaneous Supratentorial Intracerebral Haemorrhage (SWITCH) provides weak evidence that decompressive craniectomy without haematoma evacuation might have lower odds of poor outcome (modified Rankin scale [mRS] scores 5–6) in severe deep ICH. INTERACT3 was an international, multicentre, prospective, stepped-wedge, cluster randomised, blinded outcome assessed, clinical trial to estimate whether implementing a goal-directed care bundle including early intensive blood pressure lowering could improve outcomes in spontaneous ICH patients. The pragmatic design with systematic assessment of clinical outcomes of INTERACT3 allowed further analysis on the effect of haematoma evacuation on functional outcome in ICH.
Added value of this study
In this secondary analysis of INTERACT3 trial, we included 5722 patients with parenchymal ICH from 82 sites in China, with 1411 received surgical treatment for haematoma evacuation and 4776 (82.8%) patients were deep ICH. After adjustment for confounders, haematoma evacuation was significantly associated with a lower odds of mRS 5–6 and mortality at 6 months. Propensity score matched analysis and sensitivity analysis by imputation were consistent with the primary results. Furthermore, when focusing on those participants with a supratentorial haematoma volume ≥30 mL, which is a common indication for surgical intervention as well as an inclusion criteria for clinical trials, we also found the benefits of surgery from haematoma evacuation with an decreased odds of mRS 5–6 at 6 months.
Implications of all the available evidence
Findings from this study support the active treatment that haematoma evacuation could be considered in clinical practice for patients with spontaneous ICH. The effect of minimally invasive surgery including endoscopy and aspiration on the functional outcome in ICH patients, especially in deep ICH, should be tested in future trials. Furthermore, the time window as well as the difference between different techniques need to be estimated as well.
Introduction
Despite a considerable amount of research effort, there is ongoing controversy as to the role of surgery in the management of patients with acute spontaneous intracerebral haemorrhage (ICH).1,2 A meta-analysis of 21 randomised controlled trials showed that both open craniotomy and minimally-invasive surgery (MIS) can improve functional outcome and reduce mortality in patients with supratentorial ICH, but there was significant heterogeneity in the treatment effect across the individual trials.3 Differences in the design, intervention, and populations are likely explanations for the findings.2 Most recently, the Early Minimally Invasive Removal of Intracerebral Haemorrhage (ENRICH) trial reported a positive effect of a novel MIS approach in 300 ICH patients (69.3% with lobar ICH) using a Bayesian analytical approach.4 The effect of haematoma evacuation appeared to be attributable to lobar ICH, while it was not apparent for patients with anterior basal ganglia haemorrhage.4 For deep ICH, the Swiss Trial of Decompressive Craniectomy vs. Best Medical Treatment of Spontaneous Supratentorial Intracerebral Haemorrhage (SWITCH) showed a trend towards avoiding a very poor outcome (modified Rankin scale [mRS] scores 5–6) from decompressive craniectomy in 201 participants with a severe deep supratentorial ICH.5 The inspiring results from these clinical trials encourage further exploration of the surgical outcomes from haematoma evacuation for ICH.
The third Intensive Care Bundle with Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial (INTERACT3) showed clear benefits from the implementation of a care bundle with time- and target-based protocols that incorporated early intensive blood pressure (BP) lowering as a system of care in a broad range of patients with ICH.6 As the pragmatic stepped-wedge cluster randomised design allowed a large number of patients to be recruited with a broad range of characteristics, a high level of selection bias that arises from conventional individual patient randomised trials was avoided. Herein, we present results of secondary analysis that aimed to determine the relation of haematoma evacuation and functional outcome in patients with ICH.
Methods
Study design, participants, and setting
INTERACT3 was an international, multicentre, prospective, stepped-wedge, cluster randomised, blinded outcome assessed, clinical trial undertaken at 121 hospitals located in nine low- and middle-income countries (Brazil, China, India, Mexico, Nigeria, Pakistan, Peru, Sri Lanka, and Vietnam), and one high-income country (Chile), as outlined elsewhere.6 Participating hospitals either had no or inconsistent protocols for managing abnormal physiological variables in patients with acute ICH and were willing to implement the required interventional care bundle as part of routine care. Adult (age ≥18 years) patients with CT confirmed primary ICH who presented within 6 h after symptom onset were considered for enrolment. Details of the inclusion and exclusion criteria are described elsewhere.7 The trial is registered at ClinicalTrials.gov (NCT03209258) and the Chinese Clinical Trial Registry (ChiCTR-IOC-17011787). In this secondary analysis of INTERACT3, we included the patients with parenchymal ICH from China as: 1) there were approximately 90% of the participants were recruited in China in INTERACT3 study; 2) most of the surgery for haematoma evacuation were performed in China; 3) less variations in the surgical procedures and system of care in the same region. The surgical types for haematoma evacuation included craniotomy, endoscopy and catheter aspiration with or without thrombolysis, whereas patients who received external ventricular drainage or craniectomy were excluded.
Procedures
Demographic and clinical data, including the level of neurological impairment on the National Institutes of Health Stroke Scale (NIHSS, range 0–42, with higher scores indicating greater severity), were collected at the time participants were admitted to hospital. Clinical status and details of hospital management including surgery, were collected from baseline to Day 7 (or before discharge if earlier). Study outcomes were recorded centrally through telephone follow-up of participants at 6 months by independent staff blinded to randomised treatment allocation. The primary outcome was poor functional outcome, defined as a mRS score 5–6, at 6 months. The mRS is a standard 7-categorical global measure of functional outcome with scores ranging from 0 to 6, in which scores of 0–1 indicate a favourable outcome without or with symptoms but no disability; scores of 2–5 indicate increasing amounts of disability (and dependency); and a score of 6 indicates death. Secondary outcomes included mRS 4–6 and mortality at 6 months.
Statistical analysis
All analyses were undertaken at the patient level using logistic regression with adjustment as a random effect for cluster (hospital), a fixed categorical effect of time (4 randomised interventional time periods), and a fixed effect of the group assignment of each cluster at each period, as undertaken for the main results.8 Since the period lengths were not equal and the times of enrolment of the different clusters varied widely, a 6-month calendar period was used to adjust for the effect of time. Baseline characteristics and hospital management with a p-value of <0.1 shown in univariate comparisons were included as covariates in the multivariable analysis of the association of surgery and clinical outcome. A propensity score matching analysis was conducted as a sensitivity analysis to reduce baseline imbalances between the groups using a nearest neighbour matching with a caliper width of 0.20. Propensity scores were calculated for all the baseline covariates, with unbalanced covariates (standardised differences >0.1) further adjusted in the multivariable models to determine associations of surgery and outcomes. Subgroup analysis included age, sex, region, baseline neurological severity, history of hypertension, prior antithrombotic treatment, baseline haematoma volume, admission to an intensive care unit (ICU), randomised treatment and haematoma location. As outcome data were missing in >10% of participants, multiple imputation was conducted as a further sensitivity analysis with all covariates (including the outcome variable) in the mixed model (method outlined in Appendix 1 of Supplementary Material). A worse-case scenarios imputation, which assumes that all the missing primary outcomes had the worse outcomes was conducted as another sensitivity analysis. We further explored the impact of surgery by time to intervention using date (same day of hospital arrival vs. second day vs. later days) and type of surgery on the primary outcome. Further exploratory analysis was undertaken in patients with supratentorial ICH with a haematoma volume ≥30 mL, as this is popular criteria for surgery in routine practice. Data are reported with odds ratios (OR) and 95% confidence intervals (CI). All analyses were undertaken with SAS Enterprise Guide (version 8.2).
Ethics approval
This study is a secondary analysis of data originally collected under INTERACT3 study. The Biomedical Ethics Committee of West China Hospital approved the study before the commencement of any patient recruitment (Ethics Reference No. 22017 Review [217]). According to funding request from Medical Research Council, additional approval (Ethic Reference: 26596- tgr2r-ls: cardiovascular sciences, deptof) had been obtained from Research Ethics Committee of the University of Leicester, United Kingdom. Ethics approval was obtained at each site before site activation. All participants provided informed consent at the time of the original study. The secondary analysis of de-identified data was determined to be exempt from additional ethical review.
Role of the funding source
The funders of the study had no role in study design, data collection, data analysis and interpretation, or writing of the report.
Results
Of the 7036 INTERACT3 participants with acute ICH recruited between December 12, 2017, and December 31, 2021, 6356 (90.3%) patients were from China. After further screening, a total of 5772 patients from 82 sites in China with parenchymal ICH and surgical data available for these analyses in whom 1411 (24.4%) received surgical intervention for haematoma evacuation and 4361 (75.6%) received medical treatment alone (Fig. 1). Table 1 shows that in comparison to participants who did not receive surgery, those who had surgery were younger (mean age 58.8 vs. 63.1 years), had higher systolic BP (175.6 mmHg vs. 172.9 mmHg), higher median NIHSS scores (22 vs. 11), lower median GCS scores (9 vs. 13), larger haematoma mean volumes (39 vs. 10 mL) and greater intraventricular extension (33.2% vs. 21.8%).
Fig. 1.
Patient flow.
Table 1.
Baseline characteristics.
| Baseline characteristics | Overall N = 5772 | Surgery N = 1411 | No surgery N = 4361 | p-value |
|---|---|---|---|---|
| Age (years) | 62.0 (12.5) | 58.8 (11.9) | 63.1 (12.6) | <0.0001 |
| Sex | ||||
| Male | 3685 (63.8%) | 936 (66.3%) | 2749 (63.0%) | 0.0249 |
| Female | 2087 (36.2%) | 475 (33.7%) | 1612 (37.0%) | |
| Medical history | ||||
| Hypertension | 3984 (69.0%) | 961 (68.1%) | 3023 (69.3%) | 0.3924 |
| Previous stroke | 883 (15.3%) | 167 (11.8%) | 716 (16.4%) | <0.0001 |
| Coronary artery disease | 157 (2.7%) | 43 (3.0%) | 114 (2.6%) | 0.3843 |
| Other heart disease | 197 (3.4%) | 41 (2.9%) | 156 (3.6%) | 0.2273 |
| Atrial fibrillation | 64 (1.1%) | 11 (0.8%) | 53 (1.2%) | 0.1743 |
| Diabetes mellitus | 535 (9.3%) | 131 (9.3%) | 404 (9.3%) | 0.9818 |
| Hypercholesterolaemia | 118 (2.0%) | 26 (1.8%) | 92 (2.1%) | 0.5373 |
| Current smoker | 1169 (20.3%) | 319 (22.6%) | 850 (19.5%) | 0.0113 |
| Current alcohol consumption | 1161 (20.1%) | 316 (22.4%) | 845 (19.4%) | 0.0139 |
| Modified Rankin scale score of 0 before onset | 4434 (76.8%) | 1127 (79.9%) | 3307 (75.8%) | <0.0001 |
| Medication at arrival | ||||
| Antihypertensive medication | 2472 (42.8%) | 563 (39.9%) | 1909 (43.8%) | 0.0106 |
| Blood glucose lowering agents | 371 (6.4%) | 92 (6.5%) | 279 (6.4%) | 0.8704 |
| Statin or other lipid lowering agent | 119 (2.1%) | 26 (1.8%) | 93 (2.1%) | 0.5054 |
| Aspirin or other antiplatelet agent | 277 (4.8%) | 82 (5.8%) | 195 (4.5%) | 0.0407 |
| Anticoagulation agent | 43 (0.7%) | 6 (0.4%) | 37 (0.8%) | 0.1081 |
| Systolic Blood Pressure (mmHg) | 173.5 (27.4) | 175.6 (29.2) | 172.9 (26.7) | 0.0269 |
| Diastolic Blood Pressure (mmHg) | 98.8 (17.3) | 99.6 (18.0) | 98.5 (17.1) | 0.3415 |
| NIHSS at admission | 13.0 (6.0, 22.0) | 22.0 (14.0, 32.0) | 11.0 (5.0, 18.0) | <0.0001 |
| GCS score | 12.0 (9.0, 14.0) | 9.0 (6.0, 12.0) | 13.0 (11.0, 15.0) | <0.0001 |
| Randomised group, n (%) | ||||
| Intervention | 2598 (45.0%) | 631 (44.7%) | 1967 (45.1%) | 0.8009 |
| Control | 3174 (55.0%) | 780 (55.3%) | 2394 (54.9%) | |
| Brain imaging features | ||||
| Volume of haematoma at baseline | 15.0 (8.0, 30.0) | 39.0 (25.0, 50.0) | 10.0 (5.7, 20.0) | <0.0001 |
| Location of haematoma | ||||
| Deep, n (%) | 4776 (82.8%) | 1168 (82.9%) | 3608 (82.8%) | 0.9404 |
| Cortical, n (%) | 494 (8.6%) | 180 (12.8%) | 314 (7.2%) | <0.0001 |
| Cerebellum, n (%) | 302 (5.2%) | 91 (6.5%) | 211 (4.8%) | 0.0180 |
| Brainstem, n (%) | 289 (5.0%) | 17 (1.2%) | 272 (6.2%) | <0.0001 |
| Intraventricular haematoma, n (%) | 1420 (24.6%) | 468 (33.2%) | 952 (21.8%) | <0.0001 |
Data are n (%), mean (SD), or median (IQR).
BP, denoted blood pressure; GCS, Glasgow coma scale; NIHSS, National Institutes of Health stroke scale.
There were 977 (69.3%) participants who received surgery on the same day of admission to hospital and 336 (23.9%) who had surgery on the following day (Table 2). More than half of the surgical patients received a craniotomy (72.6%); the other interventions included catheter aspiration with or without thrombolysis (22.5%) and endoscopy (4.8%). Compared to participants without surgery, those who received surgery were more likely to receive more intravenous BP lowering (86.5% vs. 75.5%), antipyrexia treatment (17.2% vs. 4.7%), mechanical ventilation (64.6% vs. 5.9%), and assisted feeding (78.2% vs. 42.7%) (Table 2).
Table 2.
Management of patients in the first 7 days.
| Hospital management | Overall N = 5772 | Surgery N = 1411 | No surgery N = 4361 | p-value |
|---|---|---|---|---|
| Surgical procedure time from hospital arrival | ||||
| Same day | 977 (69.3%) | |||
| 2nd day | 336 (23.9%) | |||
| More than 2 days | 96 (6.8%) | |||
| Surgical procedure type | ||||
| Craniotomy | 1025 (72.6%) | |||
| Endoscopy | 68 (4.8%) | |||
| Catheter aspiration with or without thrombolysis | 318 (22.5%) | |||
| Admission department to hospital, n (%) | ||||
| Neurosurgery | 4707 (81.5%) | 1086 (77.0%) | 3621 (83.0%) | <0.0001 |
| Neurology | 324 (5.6%) | 18 (1.3%) | 306 (7.0%) | <0.0001 |
| Intensive care | 628 (10.9%) | 279 (19.8%) | 349 (8.0%) | <0.0001 |
| Emergency department | 23 (0.4%) | 8 (0.6%) | 15 (0.3%) | <0.0001 |
| Others | 90 (1.6%) | 20 (1.4%) | 70 (1.6%) | <0.0001 |
| Treatment during the first 24 h | ||||
| BP lowering treatment, n (%) | 4515 (78.2%) | 1221 (86.5%) | 3294 (75.5%) | <0.0001 |
| Intensive treatment for glucose control, n (%) | 371 (6.4%) | 92 (6.5%) | 279 (6.4%) | 0.8704 |
| Oral agents for glycaemic control, n (%) | 302 (81.4%) | 75 (81.5%) | 227 (81.4%) | 0.9728 |
| Insulin treatment for glycaemic control, n (%) | 89 (24.0%) | 20 (21.7%) | 69 (24.7%) | 0.5600 |
| Antipyrexia treatment, n (%) | 448 (7.8%) | 242 (17.2%) | 206 (4.7%) | <0.0001 |
| Management during 2–7 days | ||||
| Intravenous BP lowering, n (%) | 3979 (68.9%) | 1168 (82.8%) | 2811 (64.5%) | <0.0001 |
| Oral BP lowering, n (%) | 3926 (68.0%) | 807 (57.2%) | 3119 (71.5%) | <0.0001 |
| Insulin, n (%) | 701 (12.1%) | 264 (18.7%) | 437 (10.0%) | <0.0001 |
| Pyrexia treatment, n (%) | 1160 (20.1%) | 612 (43.4%) | 548 (12.6%) | <0.0001 |
| PCC administrated, n (%) | 548 (9.5%) | 219 (15.5%) | 329 (7.5%) | <0.0001 |
| Fresh frozen plasma, n (%) | 143 (2.5%) | 112 (7.9%) | 31 (0.7%) | <0.0001 |
| Vitamin K administrated, n (%) | 234 (4.1%) | 105 (7.4%) | 129 (3.0%) | <0.0001 |
| Mechanical ventilation, n (%) | 1170 (20.3%) | 911 (64.6%) | 259 (5.9%) | <0.0001 |
| Intensive care admission, n (%) | 2099 (36.4%) | 1051 (74.5%) | 1048 (24.0%) | <0.0001 |
| Assisted feeding, n (%) | 2965 (51.4%) | 1103 (78.2%) | 1862 (42.7%) | <0.0001 |
| Decision to withdraw active care, n (%) | 30 (0.5%) | 8 (0.6%) | 22 (0.5%) | 0.7766 |
Data are n (%), mean (SD), or median (IQR).
BP, denotes blood pressure; FFP, fresh frozen plasma; ICU, intensive care unit; PCC, prothrombin complex concentrate.
Table 3 shows in univariate analysis, there were higher likelihood of mRS 5–6 in the surgery group compared to medical group (36.7% vs. 19.2%) at 6 months. However, after adjustment for baseline characteristics and hospital management variables, surgery for haematoma evacuation was associated with lower odds of mRS 5–6 (OR 0.71, 95% CI 0.55–0.92; p = 0.010), and mortality (OR 0.55, 95% CI 0.40–0.75; p = 0.0001) at 6 months, while there was no difference with regard mRS 4–6 (OR 1.04, 95% CI 0.82–1.33, p = 0.732). Similar results were found in propensity score matched analysis (matching characteristics have shown in Supplemental Table S1), with haematoma evacuation being associated with reduced odds of mRS 5–6 (OR 0.74, 95% CI 0.56–0.98; p = 0.034) and mortality (OR 0.52, 95% CI 0.37–0.72; p = 0.0001) (Table 3). The results were consistent in the sensitivity analysis by imputation (Supplemental Table S2). There were significant interactions between surgery and the pre-specified subgroups of sex, neurological severity, baseline haematoma volume, ICU admission, and randomised treatment (Fig. 2).
Table 3.
Functional outcomes in 6 months follow-up.
| Outcomes | Surgery | No surgery | Unadjusteda |
Adjustedb |
PSMc |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | OR (95% CI) | p-value | ICC | N | aOR (95% CI) | p-value | ICC | N | aOR (95% CI) | p-value | ICC | |||
| Primary outcome | ||||||||||||||
| mRS 5–6 | 449/1222 (36.7%) | 738/3853 (19.2%) | 5075 | 2.60 (2.24–3.03) | <0.0001 | 0.025 | 4962 | 0.71 (0.55–0.92) | 0.010 | 0.013 | 2126 | 0.74 (0.56–0.98) | 0.034 | 0.027 |
| Secondary outcomes | ||||||||||||||
| mRS 4–6 | 695/1222 (56.9%) | 1150/3853 (29.9%) | 5075 | 3.34 (2.90–3.85) | <0.0001 | 0.026 | 4962 | 1.04 (0.82–1.33) | 0.732 | 0.011 | 2126 | 1.06 (0.81–1.37) | 0.671 | 0.030 |
| Death at 6 months | 243/1278 (19.0%) | 436/3949 (11.0%) | 5227 | 2.08 (1.73–2.50) | <0.0001 | 0.034 | 5108 | 0.55 (0.40–0.75) | 0.0001 | 0.027 | 2214 | 0.52 (0.37–0.72) | 0.0001 | 0.044 |
Note: aOR, denotes adjusted odds ratio; CI, confidence interval; ICC, intraclass correlation coefficient; mRS, modified Rankin Scale; OR, odds ratio.
Un-adjusted analysis: logistic regression for death/disability with a random effect for cluster (hospital site), a fixed effect indicating the group assignment of each cluster at each step, and a fixed categorical effect of 6-month interval.
Model adjusted baseline characteristics including mRS before stroke, age, sex, baseline NIHSS score, baseline haematoma volume, baseline systolic blood pressure, location of haematoma, intraventricular haemorrhage, previous history of stroke, current smoker and current alcohol user, and hospital management variables during including admission ward, BP lowering treatment (intravenous or oral), insulin treatment, antipyrexia treatment, anticoagulant reversal treatment (administrated PCC, FFP or VitK), Mechanical ventilation, and assist feeding.
Model adjusted for unbalanced baseline variables SD > 0.10 include baseline NIHSS score, baseline haematoma volume, presence of intraventricular haemorrhage and further hospital management variables including admission ward, BP lowering treatment (intravenous or oral), insulin treatment, antipyrexia treatment, anticoagulant reversal treatment (administrated PCC, FFP or VitK), Mechanical ventilation, and assist feeding.
Fig. 2.
Subgroup analysis on primary outcome of mRS 5–6. Note: Multivariable logistic regression model with a random effect for cluster (hospital site), a fixed effect indicating the group assignment of each cluster at each step, and a fixed categorical effect of 6-month interval, adjusted baseline characteristics variables characteristics: mRS before stroke, age, sex, baseline NIHSS score, baseline haematoma volume, baseline systolic blood pressure, location of haematoma, intraventricular haemorrhage, previous history of stroke, current smoker and current alcohol user, and hospital management variables during including admission ward, BP lowering treatment (intravenous or oral), insulin treatment, antipyrexia treatment, anticoagulant reversal treatment (administrated PCC, FFP or VitK), Mechanical ventilation, and assist feeding.
There was no significant difference in the primary outcome according to the day of surgery (Supplemental Table S3). Compared with craniotomy, MIS including endoscopy and catheter aspiration with or without thrombolysis, was not associated with mRS 5–6 at 6 months after ICH (Supplemental Table S4). Among participants with supratentorial ICH and a haematoma volume ≥30 mL, haematoma evacuation was associated with lower odds of mRS 5–6 (n = 1234, OR 0.68, 95% CI 0.46–0.99; p = 0.042) and mortality (n = 1291, OR 0.45, 95% CI 0.29–0.69; p = 0.0003) (Supplemental Table S5). No significant differences were found on the date of surgery undertake in this group of patients (Supplemental Table S6).
Discussion
Haematoma evacuation has long been a theoretically promising strategy for ICH, considering the decrease of mass effect as well as the cascade of secondary injury induced by clot-derived factors. This study aimed to define associations of surgery from haematoma evacuation and clinical outcomes in a broad range of patients with acute ICH who participated in the INTERACT3 study. The key finding of our study was that the surgery from haematoma evacuation of ICH was superior to the medical treatment in relation to the lower likelihood of death or major disability (mRS 5–6). This result was confirmed in propensity score matched and sensitivity analyses. Furthermore, we failed to detect the difference in outcomes by the timing of surgery or the approach for haematoma evacuation.
Neurosurgeons as well as patients or their caregivers were very active in ICH management in China, where haematoma evacuation were frequently chosen as a treatment strategy when needed. Only nine patients outside of China received haematoma evacuation in INTERACT3. Although including only patients from China in these analysis raises issues of selection bias, the large-scale, pragmatic, stepped-wedge, cluster randomised controlled design of INTERACT3 allowed data to be obtained that closely matches those of routine systems of care.
The large sample size with systematic assessment of clinical outcomes in INTERACT3 allow us to explore the controversial topic of the effects of surgery from haematoma evacuation in patients with spontaneous ICH. The first large-scale international multi-center, Surgical Trials in Intracerebral Haemorrhage (STICH) failed to show a benefit of early surgery in 1033 patients with supratentorial ICH, but a considerable number of them crossed over from medical treatment to surgery due to clinical deterioration.9 On the basis of a small but clinically relevant survival advantage in the subgroup of patients with lobar ICH, the follow-up STICH II trial did not show a clear result to influence practice.10 While the ENRICH trial now provides evidence of a benefit of early MIS on functional outcome, this seems most relevant to lobar rather than the more common, deep ICH.4 The SWITCH trial provided weak evidence that decompressive craniectomy without haematoma evacuation benefits patients with deep ICH,5 but subgroup analysis of the third Minimally Invasive Surgery Trial in Intracranial Haemorrhage (MISTIE III) showed only a survival advantage rather than benefit to functional outcome from MIS combined with thrombolysis for deep ICH.11 In our study, the majority of patients had deep ICH and received craniotomy. The high percentage of deep ICH receiving surgery reflected clinical work situation with proactive surgical intervention in China, despite guideline and ENRICH trial showed limited to no functional benefit. This discrepancy is partially driven by the high disease burden as well as the Chinese culture that life-saving is considered more valuable than the impact of residual severe disability. Previous guidelines provided recommendations based on the trials with functional outcome defined as mRS 0–2 or 0–3, which might ignore the potential acceptance of a mRS score of 4 by patients or caregivers. In the ENRICH trial, deep ICH were purposely excluded from further recruitment since they failed to meet a prespecified efficacy threshold in the interim analysis as part of an adaptive clinical trial design. Therefore, the size of the treatment effect in deep ICH, which is clearly less than in cortical ICH, is still worth exploring. Regarding the choice of surgery, it is mainly based on the neurosurgeon's preference and local protocol. The percentage of craniotomy was quite high, mainly because endoscopic haematoma evacuation was not widely adopted in China during the period of INTERACT3, between 2017 and 2021. It was not until the most recent 2022 AHA/ASA guideline recommendation that endoscopic haematoma evacuation was recommended as being useful to reduce mortality compared with medical management alone.12 The innovation of instrument and mini-invasive techniques could gradually shift the neurosurgeon's preference and might potentially produce a greater benefit than open craniotomy.
We performed analysis in those participants with a haematoma volume ≥30 mL from a supratentorial ICH as this is a common indication for surgical intervention as well as an inclusion criteria for clinical trials. Our results affirm the potential benefits of surgery from haematoma evacuation through an association with a decreased odds of mRS 5–6 at 6 months. Several ongoing trials are evaluating the effects of surgery to evacuate the haematoma in supratentorial ICH, including the Early Minimally Invasive Image Guided Endoscopic Evacuation of Intracerebral Haemorrhage (EMINENT-ICH, NCT05681988), Ultra-Early, minimally invasive intracerebral haemorrhage evacuation vs. standard treatment (EVACUTATE, NCT04434807), and the Dutch ICH trial (DIST, NCT05460793).
The surgical approach for haematoma evacuation continues to evolve and the approaches of endoscopy, MIS, and catheter evacuation with thrombolysis, were used according to interventional preference but evaluated in several trials.13,14 In one multicentre randomised controlled trial of 733 participants with supratentorial ICH, endoscopic surgery and stereotactic aspiration were found to be superior to craniotomy with a small bone flap, especially in deep ICH.13 In the Big data Observatory Platform for Stroke of China, Li et al. analyzed data from 7451 patients who received surgery for ICH between 2019 and 2021 to show that cranial puncture was associated with a lower odds of poor functional outcome compared to craniotomy (OR 0.84, 95% CI 0.70–1.01).15 In our study, where endoscopy and catheter aspiration with or without thrombolysis were combined as a MIS variable, we did not detect significant between-group difference in functional outcome at 6 months, which might be partially explained by the apparent imbalance of the sample size between both groups. Further trials are needed to explore the difference between approaches, especially those mini-invasive techniques.
There is insufficient data in relation to the time of surgery and outcome in ICH. The median time from the ictus to surgery were 26, 16.75 and 59 h (and 72 h to first dose of alteplase), in the STICH II, ENRICH, and MISTIE III trials, respectively.4,10,11 Our finding of no significant time relation for surgery was limited by the use of day rather than hour as the dependent variable. A recent systematic review showed patients undergoing surgery within 24 h have a higher likelihood of a good functional outcome, when compared to those undergoing surgery within 72 h after ICH onset.16 One of the concerns that hinders ultra-early evacuation was postoperative rebleeding. Recently, Ali et al. reported that ultra-early evacuation within 5 h of ictus utilizing a refined endoscopic technique is associated with increased intraoperative bleeding but not postoperative rebleeding or worse clinical outcomes, which could help to shorten the time window and explore the potential benefit of early surgical treatment for ICH.17
In line with the SWITCH trial and a pooled analysis of three clinical trials of decompressive surgery in malignant middle cerebral artery (MCA) infarction,5,18 we chose to define a very poor functional outcome according to mRS scores 5–6, which differs from the more common use of mRS scores 3–6 or 4–6 in ICH trials. However, this cut point is important in understanding that the potential benefit of surgery in reducing the chances of an extremely poor outcome, either death or having a bedridden level of disability, is offset by survivors having more severe disability (mRS 3–4). However, an analysis of 28 patients who received hemicraniectomy for malignant MCA infarction showed that surviving such a critical illness with an mRS 4 is acceptable to many patients and their relatives.19 Moreover, 77% (48/62) of survivors in the SWITCH trial would choose to undergo surgery again in light of their experience.5 Thus, a comparison of outcomes according to mRS scores 0–4 and 5–6 is reasonable for individualizing decisions over surgery in ICH.
Strengths of this study include the use of a large dataset from a broad range of patients with ICH, and the large proportion of patients with deep haematoma who received different types of surgery. This INTERACT3 cohort included 5722 patients at 82 secondary and tertiary hospitals, which could be representative of the proactive attitude towards surgical treatment for ICH in China, which is quite different from the routine clinical work in Western countries. Secondly, the cluster randomised design and the system of care intervention allowed the collection of routine data and avoided a large degree of selection bias. However, there are several limitations to the interpretation of our results. Firstly, we restricted the analysis in region of China, which might potentially influence the generalisability. However, the results could at least encourage future multicentre trials for haematoma evacuation in ICH treatment, especially for deep haematoma. Secondly, we did not standardise the surgical procedures, including the indications, instrument, trajectory, techniques, as well as length and times of the thrombolysis agent, which could bring potential heterogenicity in surgical performance and limit the adoption in routine clinical work. Future work exploring the effect of surgery in ICH should establish a standardisation of surgical protocols. Moreover, the competency of the neurosurgeon should also be taken into consideration. Thirdly, due to lack of data on an explicit time of surgery from hospital admission, the impact of timing on surgery needs further exploration. Fourthly, there might be potential unmeasured confounding by indication as well as selection bias due to analysis of a trial dataset. Considering this is an observational cohort study with propensity methods conducted in a subgroup of the matched participants, the result on surgery effect might not be generalizable to patients with different characteristics. Lastly, caution is required in the interpretation of subgroup analysis when the reduced sample size reduces the precision of results.
In summary, our study showed that the surgical treatment from haematoma evacuation of ICH is associated with a higher likelihood of survival and avoidance of a bedridden level of disability. Further clinical trials are necessary to allow for a better selection of surgical strategy for ICH, most notably in relation to the time, approach to treatment as well as deep ICH.
Contributors
Concept and design: CY, LS, CSA, JX. Acquisition, analysis or interpretation of data: XH, MO, LM, YL, XL, YJ, XC. Drafting the manuscript: XH, MO, LS, CSA. Critical review of the manuscript for important intellectual content: all authors. Raw data access: QL, MO. Data verification: QL, MO. Statistical analysis: MO. Funding: CY, LS, CSA. Administrative, technical, or material support: XH, MO, LM, YL, XL, YJ, XC. Supervision: CY, LS, CSA, JX. Final decision for submission: CSA.
Data sharing statement
Individual, de-identified participant data used in these analyses can be shared on request from any qualified investigator after the approval of a protocol and signed data access agreement via both the trial steering committee and the research office of The George Institute for Global Health, Sydney, Australia).
Declaration of interests
LS reports funding from the Medical Research Council of the UK, Sichuan Credit Pharmaceutic, and Takeda China; and speaker fees from Takeda China. CSA has received grants from the National Health and Medical Research Council of Australia, the Medical Research Council of the UK, and Takeda China. He is also the chair of the data and safety monitoring boards for several investigator-initiated trials, President-elect of the World Stroke Organisation; and Editor-in-Chief of Cerebrovascular Diseases. CY has received funding from West China Hospital. All other authors declare no competing interests. PMV receives research grants from ANID Fondecyt Regular 1221837, Pfizer and Boehringer Ingelheim.
Acknowledgements
The study is supported by an award (grant reference number MR/T005009/1) jointly funded by the Department of Health and Social Care, the Foreign, Commonwealth & Development Office, the Medical Research Council, and the Wellcome Trust (all London, UK); the West China Hospital Outstanding Discipline Development 1–3-5 programme (number ZY2016102); National Health and Medical Research Council of Australia (number APP1149987); Sichuan Credit Pharmaceutical; and Takeda (China). CSA is a senior investigator fellow for the National Health and Medical Research Council of Australia. We also thank the participants, their relatives, and their families.
Role of the Funder/Sponsor: The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Footnotes
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lanwpc.2025.101669.
Contributor Information
Lili Song, Email: lili_song@fudan.edu.cn.
Lu Ma, Email: alex80350305@163.com.
INTERACT3 Investigators:
Octavio Pontes-Neto, Millene Camilo, Francisco Dias, Octavio Vincenzi, Rodrigo Cerantola, Carla Moro, Renata Santos, Nara Texeira, Alexandre Longo, Rafaela Liberato, Sheila Martins, Arthur Pille, Bruna Chwal, Isabel Silva, Natacha Titton, Gustavo Weiss, Daissy Mora, Magda Ouriques, Leonardo Carbonera, Rodrigo Bazan, Gabriel Modolo, Fernanda Winckler, Luana Miranda, Juli Souza, Alexis Rojo, Wilhelm Uslar, Lorena Medel, Paula Munoz Venturelli, Javiera Lopez, Diego Herrero, Pablo Lavados, Barbara Vargas Latorre, Nathalie Conejan, Tomas Esparza, Patricio Sotomayor, Denisse Wenger, Juan Pablo Gigoux, Aldo Letelier, Lilian Acevedo, Vivianne Moya, Cristian Figueroa, Nicol Vallejos, Nathalie Conejan, Tomas Esparza, Patricio Sotomayor, Rodrigo Guerrero, Mauricio Velasquez, Jose Vallejos, Kimerly Pallauta, Tamara Santibanez, Angelo Queirolo, Andrea Lobos, Aralikatte Onkarappa Saroja, Ravishankar Naik k, Sandip Chindhi, Nakul Pampaniya, Kurubara Amaresh, Thomas Iype, Dileep R, Reeja Rajan, Praveen Panicker, Rupjyoti Das, Nupur Choudhury, Pankaja Gohain, Jemin Webster, Biyol Pakma, Lalbiak Sangi, Ivy Sebastian, Gaurav Aggrawal, Komal Raj, Deepankshi Rajoura, Sulena Singh, Varun Aggrawal, Amit Narang, Antonio Arauz, Vanesa Cano-Nigenda, Diego López-Mena, Héctor Valdez-Ruvalcaba, Roberto Toledo-Treviño, Reginald Obiako, Sani Abubakar, Oguike Emeka, Balogun Olayemi, Melika Lois, Ibinaiye Philip, Olurishe Comfort O, Njideka Okubadejo, Osigwe Agabi, Oluwadamilola Ojo, Kolawole Wahab, Abiodun Bello, Oyinloye Ibukun, Olufemi Sanayaolu, Sunday Adeniyi, Abdulraheem Jimoh, Mohammad Wasay, Dilshad Begum, Anila Anjum, Shahid Waheed, Ayeesha Kamal, Raja Farhat Shoaib, Fizza Orooj, Sadaf Majid, Taskeen Zehra, Abdus Salam Khan, Ravi Shanker, Nadir Ali Syed, Nashwa Ahmad, Carlos Abanto, Ana Valencia, Danny Barrientos, Jorge Ramirez, Pilar Calle, Dilum Palliyeguruge, Sumudu Muthucumarana, Shiroma Ratnayaka, Dilhara Ganihiarachchi, Arundathi Bandaranayake, S.D.B. Somaratne, Saumya Narayana, Sithara Gallage, Bimsara Senanayake, Udari Samarasiri, Dunya Luke, Mythily Sivapathasundaram, Vithoosan Sahadevan, Amani Rasmi, Yuran Deshaka, Nilukshi Fernando, Aruna Munasinghe, Kapilanga Rathnapriya, A.S. Nissanka, Kanchana Karunathilake, Isuru Gayan, Kaminda Wijenayake, Hasitha Gunasekara, Jagath Vidyarathne, Ajantha Keshavaraj, Kanagasabapathy Janarthanan, Arhivalaky Gerald Jeevathasan, Sivaram Sivamainthan, Mathyamuthan John Priyanth, Abirami John Priyanth, Thambippillai Rajendiran, Sanjeewa Alwis, Nushara Gunasekare, Vasundara Liyanarachchi, Athula Dissanayake, Wimalasiri Mewa Uluwattage, Gimhani Ratnayake, Charika Rajinee, Sakura Jayawardana, Janaka Peiris, Ranjith Wicramasinghe, Chamila Fernando, Jessie Abbas, Nethmini Withanage, Makaranda Bandara, Duy Ton Mai, Van Chi Nguyen, Viet Phuong Dao, Xuan Trung Vuong, Tien Dung Nguyen, Trung Hieu Dinh, Ha Quan Phan, Quoc Viet Bui, Dinh Tho Phung, Quang Tho Pham, Dinh Dai Pham, Duc Thuan Do, Phuc Duc Dang, Minh Duc Dang, Dang Hai Nguyen, Thi Phuong Nga Nguyen, Quoc Huy Nguyen, Quoc Dai Pham, Quoc Vinh Chau, Vinh Thy Van Tai, Tran Vinh Le, Cong Tri Le, Ha Mai Khuong Tran, Huu Khanh Nguyen, Hoang Minh Thao Ngyen, Duc Chien Vo, Thai My Phuong Nguyen, Trung Thanh Tran, Thi Hanh Vi Vo, Hao Nhien Cao, Ba Thang Nguyen, Thi Ngoc Suong Le, Thien Duc La, Chi Duc Pham, Huy Thai, Yongming Jiang, Weimin Li, Wei Huang, Ke Luo, Gangying Liu, Guanghai Tang, Guang Yang, Hongtao Jiang, Xu Zhang, Hongyan Jing, Sheng Zhu, Bo Pu, Dong Lv, Hui Kang, Qiuping Hu, Xiaochun She, Xiaoming Jiang, Yanli Chen, Shenghua Yang, Jianjun He, Zongping Li, Gang Cheng, Hailin Huang, Xiaoyi Wang, Jianqiong Lin, Minhui Chen, Chenghao Yang, Hao Ding, Yunliang Deng, Fei Luo, Rongjun Zhang, Xiaofeng Wang, Hongbing Zhang, Xiaoliang Yang, Yang Zhang, Chengyi Yang, Yu He, Feng Liu, Rongjie Wang, Yuhui Zhang, Xiaodong Xin, Bin Feng, Wanru Hao, Chang Song, Yun Guo, Dehua Jiang, Jie Chen, Changtong Tang, Hongliang Zhu, Xin Li, Jin Cui, Haidong Xu, Boyang Li, Fusheng Tang, Yuanbin Li, Min Gao, Bo Yang, Xuejun Xu, Bing Deng, Yi Zheng, Yuanhong Ge, Keyu Chen, Yang Liu, Xinshen Li, Tingting Zhong, Jianfeng Xu, Hai Zhang, Jiyue Wang, Jianxin Zhu, Hanyu Sun, Fuhua Yu, Xueguang Zhang, Chao You, Lu Ma, Xin Hu, Jianguo Xu, Xi Li, Mingsen Zhang, Bin Wang, Yiming Ma, Donglin Jiang, Jun Zhou, Cong Liu, Wenhong Nie, Mingguo Li, Tao Tian, Yong Li, Mingfang He, Xiaolong Tu, Zhengjun Wu, Hong Liu, Dongsheng Zhong, Rongcai Jiang, Jian Sun, Ye Tian, Yingsheng Wei, Shuo An, Pingbo Wei, Le Luo, Bin Lin, Gang Liu, Yan Wen, Qiang Cai, Qianxue Chen, Pan Lei, Zhiyang Li, Meifang Zhang, Jiaquan He, Yan Chen, Jun Liu, Xinghai Liu, Junyan Li, Min Chen, Jing Wang, Bingzhi Zhou, Baichun Ye, Jiancheng Zhang, Manyuan Zhang, Xuming Pan, Xiaoxiang Yu, Jian Xu, Qingbao Xiao, Yuefei Wang, Liang Tao, Lin Shi, Niandong Zheng, Guoliang You, Bo Lei, Shu Chen, Honggang Wu, Jin Hu, Jianlan Zhao, Jian Yu, Qiang Yuan, Zhuoying Du, Xielin Tang, Qianke Li, Shenghua Liu, Feilong Yang, Kui Xiao, Chao Luo, Guang Wang, Xudong Che, Zhipeng Teng, Wenwu Wan, Jun Li, Yu Liu, Mingbo Fan, Tao Zhang, Lun Cai, Yuan Ma, Zhifeng Ma, Bin Li, Linlin He, Jinghui Li, Weibing Zhang, Shuxin Zhang, Hongzhen Zhang, Yingguang Dai, Jun Lei, Lei Mao, Yiyang Huang, Zhi Zhou, Ping Chen, Fang Chen, Pan Wei, Tiangui Li, Honglin Chen, Mengfei Zeng, Kejie Mou, Jun Xue, Yong Jiang, Xiaoping Tang, Tao Chen, Yalan Zhang, Yanbing Xu, Yuchen Gu, Lei Chen, Yujun Zhao, Bin Yang, Peng Kuai, Xi Wang, Yuwang Yang, Xueling Hu, Huitian Zhang, Yintao Yang, Weifeng Wang, Junyi Zhang, Wei Cheng, Xiaoxue Zhang, Xiaowen Ma, Qin He, Li Zhang, Rong Gao, Huixiang Liu, Jingwei Ye, Ping Xu, Xin Wu, Yuan Yuan, Peng Zou, Zhen Zhang, Jiyong Cheng, Zhangming Zhou, Yijun Zeng, Zhang Liang, Deming Du, Shui Yu, Yongjun Cao, Shoujiang You, Jiaping Xu, Zhichao Huang, Dongqin Chen, Wenfeng Xiao, Li Zhu, Miao Yuan, Yuhai Wang, Dongliang Shi, Xu Hu, Dingchao Xiang, Like Shi, Hongqin Wang, Liu Yang, Wang Miao, Yiyi Hu, Yuchun Zhao, Xi Hu, Yang Liu, Weiduo Zhou, Chao Sun, Tao Chen, Dong Tang, Kun Yao, Jin You, Shishi Chen, Jianmin Yao, Huanmei Li, Jinmei Liu, Ailin Bai, Yong Yi, Qingshan Deng, Peng Luo, Han Wang, Jingcheng Jiang, Qingwei Yang, Shunpo He, Jun Wang, Yu Chen, Hua He, Yuyang Deng, Zhikai Cao, Xuxia Yi, Jinbiao Luo, Shuang Luo, Min Gong, Li Liu, Xuejun Gao, Jia Liu, Li'e Wu, Jia Zhang, Hongying Sun, Xinhui Li, Lu Jia, Jianbing Wu, Jie Zhang, Huajun Zhang, Chunfu Du, Shun Li, Xiaobin Yang, Jie He, Lei Liao, Gezhi Zhou, Wentao Dong, Yunxiang Chen, Xiaofeng Lin, Xujian Shui, Peng Zhang, Yuan Zhao, Hongli Yang, Wenbin Zhao, Xiaoyi Zhang, Jincao Chen, Qian Wu, Xuan Dai, Baogui Tang, Yinjuan Wang, Tao Liu, Haixia Zhang, Faliang Duan, Ming Luo, Qingfang Jiao, Guoliang Lei, Dong Wang, Chunwang Song, Haopeng Tan, Feng Ye, Xinghu Qin, Xiaolong Liang, Junling Liu, Lang Yang, Jie Yang, Yapeng Lin, Qian Yang, Xuntai Ma, Yinkuang Qi, Baogen Pan, Caixia Jiang, Zhanying Ye, Ce Dong, Xiongfei Yue, Xiaopeng Yang, Tuoheti Maimaitiyiming, Jun Dong, Yonggang Wu, Feng Gao, Deqiang Zhao, Xinghai Zhang, PengJun Wang, Hongbo Jiang, Jianping Li, Wei Zhang, Jing Chen, Haibo Tong, Yonghong Wang, Kaipeng Qiao, Fuyou Guo, Mingchu Zhang, Yan Hu, Mengzhao Feng, Dengpan Song, Yi Zuo, Shangjun Chen, Chao Qian, Baoming Li, Jingku Ma, Sunfu Zhang, Bin Kong, Xingyu Dong, Qiang Li, Sheng Fang, Bin Lu, Yang Li, Zhen Zhang, Yongling Yang, Hong Yu, Huaiyu Sun, Yue Wang, Weimin Wang, Tong Li, Shengli Li, Zhiming Xu, Yongyi Wang, Qiang Dong, Yuping Tang, Heling Chu, Ying Lu, Zhong Wang, Xiaoou Sun, Jianhua Zhao, Shuaifeng Yang, Xiying Qian, Thompson Robinson, J. Jaime Miranda, Craig S. Anderson, Chao You, Lili Song, Adrian Parry-Jones, Nikola Sprigg, Sophie Durrans, Caroline Harris, Ann Bamford, Olivia Smith, Robert Herbert, Christopher Chen, William Whiteley, Rong Hu, Laurent Billot, Qiang Li, Jayanthi Mysore, Xin Hu, Yao Zhang, Feifeng Liu, Yuki Sakamoto, Shoujiang You, Qiao Han, Bernard Crutzen, Yunke Li, Emily Cheung, Stephen Jan, Hueiming Liu, Menglu Ouyang, Lingli Sun, Honglin Chu, Anila Anjum, Francisca Gonzalez Mc Cawley, Yan Wu, Lingling Feng, Jingjing Ni, Caixiu Du, Weiwei Fu, Alejandra Del Rio, Bruna Rimoli, Rodrigo Cerantola, Thanushanthan Jeevarajah, Madhushani Kannangara, Andrene Joseph, Chamath Nanayakkara, Xiaoying Chen, Alejandra Malavera, Chunmiao Zhang, Zhao Yang, Brook Li, Zhuo Meng, Menglu Ouyang, Leibo Liu, Yi Ning, Le Dong, Manuela Armenis, Joyce Lim, Helen Monaghan, Lu Ma, Xin Hu, Xi Li, Rui Luo, Guojuan Cheng, Yilin Dong, Ziqin Liu, Shuihong Wang, Ying Zhang, Jipeng Cheng, Hui Shi, Wenjing Li, Langming Mou, Ping Yi, Chen Chen, Xue Chen, Shalomi Weerawardena, Poornima Ellawala, Enalee Ranasinghe, Chrishmi Rodrigo, Kolawala Wahab, Sunday Adeniyi, Jeyaraj Pandian, Megha Khanna, Paula Muñoz Venturelli, Francisca González, Francisca Urrutia Goldsack, Alejandra Del RÃo, Mohammad Wasay, Dilshad Begum, and Anila Anjum
Appendix A. Supplementary data
References
- 1.Broderick J.P., Grotta J.C., Naidech A.M., et al. The story of intracerebral hemorrhage: from recalcitrant to treatable disease. Stroke. 2021;52(5):1905–1914. doi: 10.1161/STROKEAHA.121.033484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dammers R., Beck J., Volovici V., Anderson C.S., Klijn C.J.M. Advancing the surgical treatment of intracerebral hemorrhage: study design and research directions. World Neurosurg. 2022;161:367–375. doi: 10.1016/j.wneu.2022.01.084. [DOI] [PubMed] [Google Scholar]
- 3.Sondag L., Schreuder F., Boogaarts H.D., et al. Neurosurgical intervention for supratentorial intracerebral hemorrhage. Ann Neurol. 2020;88(2):239–250. doi: 10.1002/ana.25732. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pradilla G., Ratcliff J.J., Hall A.J., et al. Trial of early minimally invasive removal of intracerebral hemorrhage. N Engl J Med. 2024;390(14):1277–1289. doi: 10.1056/NEJMoa2308440. [DOI] [PubMed] [Google Scholar]
- 5.Beck J., Fung C., Strbian D., et al. Decompressive craniectomy plus best medical treatment versus best medical treatment alone for spontaneous severe deep supratentorial intracerebral haemorrhage: a randomised controlled clinical trial. Lancet. 2024;403(10442):2395–2404. doi: 10.1016/S0140-6736(24)00702-5. [DOI] [PubMed] [Google Scholar]
- 6.Ma L., Hu X., Song L., et al. The third intensive care bundle with blood pressure reduction in acute cerebral haemorrhage trial (INTERACT3): an international, stepped wedge cluster randomised controlled trial. Lancet. 2023;402(10395):27–40. doi: 10.1016/S0140-6736(23)00806-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Song L., Hu X., Ma L., et al. INTEnsive care bundle with blood pressure reduction in acute cerebral hemorrhage trial (INTERACT3): study protocol for a pragmatic stepped-wedge cluster-randomized controlled trial. Trials. 2021;22(1):943. doi: 10.1186/s13063-021-05881-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Billot L., Song L., Hu X., et al. Statistical analysis plan for the INTEnsive care bundle with blood pressure reduction in acute cerebral hemorrhage trial: a stepped-wedge cluster randomized controlled trial. Cerebrovasc Dis. 2023;52(3):251–254. doi: 10.1159/000526384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mendelow A.D., Gregson B.A., Fernandes H.M., 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(9457):387–397. doi: 10.1016/S0140-6736(05)17826-X. [DOI] [PubMed] [Google Scholar]
- 10.Mendelow A.D., Gregson B.A., Rowan E.N., et al. Early surgery versus initial conservative treatment in patients with spontaneous supratentorial lobar intracerebral haematomas (STICH II): a randomised trial. Lancet. 2013;382(9890):397–408. doi: 10.1016/S0140-6736(13)60986-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hanley D.F., Thompson R.E., Rosenblum M., et al. Efficacy and safety of minimally invasive surgery with thrombolysis in intracerebral haemorrhage evacuation (MISTIE III): a randomised, controlled, open-label, blinded endpoint phase 3 trial. Lancet. 2019;393(10175):1021–1032. doi: 10.1016/S0140-6736(19)30195-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Greenberg S.M., Ziai W.C., Cordonnier C., et al. 2022 Guideline for the management of patients with spontaneous intracerebral hemorrhage: a guideline from the American Heart Association/American Stroke Association. Stroke. 2022;53(7):e282–e361. doi: 10.1161/STR.0000000000000407. [DOI] [PubMed] [Google Scholar]
- 13.Xu X., Zhang H., Zhang J., et al. Minimally invasive surgeries for spontaneous hypertensive intracerebral hemorrhage (MISICH): a multicenter randomized controlled trial. BMC Med. 2024;22(1):244. doi: 10.1186/s12916-024-03468-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zhou H., Zhang Y., Liu L., et al. A prospective controlled study: minimally invasive stereotactic puncture therapy versus conventional craniotomy in the treatment of acute intracerebral hemorrhage. BMC Neurol. 2011;11:76. doi: 10.1186/1471-2377-11-76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Li Y., Yang H., Cao L., et al. Short-term surgical outcomes of spontaneous intracerebral hemorrhage in China from 2019 to 2021: a retrospective cohort study. Lancet Reg Health West Pac. 2023;39 doi: 10.1016/j.lanwpc.2023.100870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Huang J., Yao M., Ma Y., et al. Surgical interventions for spontaneous supratentorial intracerebral haemorrhage: a systematic review and network meta-analysis. eClinicalMedicine. 2025;79 doi: 10.1016/j.eclinm.2024.102999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ali M., Smith C., Vasan V., et al. Management of intracavitary bleeding during ultra-early minimally invasive intracerebral hemorrhage evacuation. J Neurosurg. 2024;142:1–11. doi: 10.3171/2024.6.JNS232985. [DOI] [PubMed] [Google Scholar]
- 18.Vahedi K., Hofmeijer J., Juettler E., et al. Early decompressive surgery in malignant infarction of the middle cerebral artery: a pooled analysis of three randomised controlled trials. Lancet Neurol. 2007;6(3):215–222. doi: 10.1016/S1474-4422(07)70036-4. [DOI] [PubMed] [Google Scholar]
- 19.Kiphuth I.C., Kohrmann M., Lichy C., Schwab S., Huttner H.B. Hemicraniectomy for malignant middle cerebral artery infarction: retrospective consent to decompressive surgery depends on functional long-term outcome. Neurocrit Care. 2010;13(3):380–384. doi: 10.1007/s12028-010-9449-8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.