STICH

Clinical Context

Spontaneous intracerebral hemorrhage (ICH) is a devastating form of stroke with high rates of mortality and long-term disability. The rationale for surgical evacuation of the hematoma is to reduce mass effect, limit secondary brain injury from hematoma products, and potentially improve neurological outcomes. However, the benefits of surgery, particularly early intervention, versus conservative medical management have long been debated due to the risks associated with surgery and the lack of definitive evidence.

Prior to STICH, surgical practices varied widely, and decisions were often based on surgeon preference, local protocols, and observational data, which suggested potential benefits for certain subgroups (e.g., superficially located hematomas). There was a clear need for a large, pragmatic randomized controlled trial to establish whether a policy of early surgical intervention offered an advantage over initial conservative management for patients with spontaneous supratentorial ICH.

Patient Population

The STICH trial enrolled 1033 adult patients (≥18 years) from 83 neurosurgical centers across 27 countries. Eligible patients had a CT-confirmed diagnosis of spontaneous supratentorial ICH and were either unconscious (Glasgow Coma Scale [GCS] score ≤14, or GCS 15 with a hematoma volume of ≥10 mL) or had a hematoma volume of ≥10 mL if conscious. A crucial entry criterion was clinical equipoise – the responsible neurosurgeon had to be substantially uncertain about the appropriateness of early surgery for that particular patient. Patients could be randomized if surgery could be performed within 72 hours of ictus, although earlier surgery (within 24 hours) was encouraged.

Exclusions included patients with intraventricular hemorrhage alone, hemorrhage secondary to a known underlying cause (tumor, aneurysm, AVM), infratentorial hemorrhage, or severe pre-existing medical conditions that would preclude a good outcome regardless of ICH treatment.

The median age of participants was 61 years. About 58% were male. The median GCS at randomization was 12. The median hematoma volume was 38 mL. Hematomas were predominantly lobar (51%) or deep (basal ganglia/thalamic, 46%).

Study Design

STICH was an international, multicenter, pragmatic, parallel-group randomized controlled trial.

Protocol Details

Patients meeting eligibility criteria were randomized 1:1 to:

• Early Surgery Group: Intended surgical evacuation of the hematoma within 72 hours of ictus (ideally within 24 hours), in addition to the best medical treatment available at the participating center. The type of surgical procedure (e.g., craniotomy, burr-hole evacuation, stereotactic aspiration, endoscopic removal) was left to the discretion of the operating neurosurgeon, reflecting the pragmatic nature of the trial.
• Initial Conservative Treatment Group: Best medical treatment according to local neurosurgical unit protocols. Surgery was reserved only for patients who experienced subsequent neurological deterioration deemed to require surgical intervention.

Best medical treatment typically included measures for blood pressure control, intracranial pressure management, and general supportive care in a neurosurgical unit.

Outcome Assessment

The primary outcome was assessed at 6 months via a postal questionnaire completed by the patient or their carer, or by telephone interview with a study researcher if the questionnaire was not returned. Researchers conducting telephone interviews were intended to be unaware of treatment allocation where possible, though the open nature of the intervention made complete blinding difficult.

Power Analysis

The trial was designed to detect an absolute difference of 12% in favorable outcomes between the two groups (assuming 30% favorable outcome in the conservative group and 42% in the surgery group), with 90% power at a 5% significance level, requiring approximately 1000 patients.

Risk of Bias Analysis

(Content for this section needs to be added if you plan to include it. Considerations: open-label design, pragmatic nature of surgical technique, potential for crossover, reliance on patient/carer reported outcomes for some.)

Results

A total of 1033 patients were randomized: 503 to early surgery and 530 to initial conservative treatment. In the early surgery group, 94% received surgery, with a median time from ictus to surgery of 30 hours. In the initial conservative treatment group, 26% (136 patients) eventually underwent surgical hematoma evacuation due to neurological deterioration, with a median time from ictus to surgery of 67 hours for these patients.

Primary Outcome: Favorable Outcome at 6 Months

A favorable outcome, defined as Good Recovery or Moderate Disability on the Extended Glasgow Outcome Scale (GOS-E) at 6 months, was achieved by:

• 26% of patients in the early surgery group (122 of 468 evaluable patients)
• 24% of patients in the initial conservative treatment group (118 of 498 evaluable patients)

This represented an absolute difference of 2.3% (95% CI -2.7% to 7.4%). The odds ratio for a favorable outcome with early surgery was 1.13 (95% CI 0.85–1.50; p=0.414). Thus, there was no statistically significant difference in favorable outcomes between the two treatment strategies.

Secondary Outcomes

• Mortality at 6 months: 36% in the early surgery group versus 37% in the initial conservative treatment group (OR 0.95, 95% CI 0.72–1.25; p=0.709).
• Rankin Scale scores at 6 months: No significant difference between groups.
• EuroQol (EQ-5D) quality of life: No significant difference between groups.
• Survival time: No significant difference.

Subgroup Analyses

Pre-specified subgroup analyses suggested a potential benefit for early surgery in patients with superficial hematomas (lobar hemorrhages without intraventricular hemorrhage) who were operated on within 24 hours. However, these were subgroup findings and should be interpreted with caution. This finding was explored further in the STICH-II trial.

Final Thoughts & Critical Appraisal

The STICH trial was a landmark study that provided crucial evidence regarding the role of early surgery for spontaneous supratentorial ICH. Its main finding was that a policy of early surgical hematoma evacuation did not result in a statistically significant improvement in overall favorable outcomes or mortality at 6 months compared to initial conservative management with surgery reserved for deterioration.

Key considerations:

• Pragmatic Design: The trial’s pragmatic design, allowing various surgical techniques and including a broad range of patients for whom surgeons were genuinely uncertain, enhances its generalizability to real-world practice.
• No Overall Benefit for Early Surgery: The primary conclusion is that, for the unselected population of ICH patients meeting inclusion criteria, routine early surgery is not superior to initial conservative management.
• Crossover: A significant proportion (26%) of patients in the conservative arm eventually underwent surgery, which might have diluted any potential benefit of a purely conservative strategy or any difference between the initial assignments.
• Timing and Type of Surgery: The trial allowed surgery up to 72 hours, and various techniques were used. Subgroup analyses hinted that very early surgery (<24 hours) and surgery for more superficial hematomas might be beneficial, but these were not definitive conclusions from the main trial.
• Impact on Guidelines: STICH significantly influenced clinical guidelines, generally leading to a more selective approach to surgical intervention in ICH, often reserving it for patients with neurological deterioration or specific hematoma characteristics (like large cerebellar hemorrhages, not studied in STICH, or superficial lobar hematomas based on subgroup hints).

While STICH did not support routine early surgery for all supratentorial ICH, it highlighted the need for further research into specific patient subgroups, optimal surgical timing, and minimally invasive techniques, which subsequent trials like STICH-II and ENRICH have aimed to address.

How do we reconcile the results with other studies?

(Content for this section needs to be added, discussing how STICH fits with later trials like STICH-II, MISTIE, ENRICH, etc.)

Found an error or want to suggest an improvement? Edit this page on GitHub.