Non-cerebral vasospasm factors and cerebral vasospasm predict delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage
Aneurysmal subarachnoid hemorrhage (aSAH)—a sudden bleed in the fluid-filled space around the brain—is a life-threatening neurosurgical emergency. For up to 30% of patients, the greatest danger comes not from the initial bleed, but from delayed cerebral ischemia (DCI): a condition where blood flow to the brain drops days later, causing new brain damage or death. While cerebral vasospasm (CVS)—narrowing of brain arteries—is widely seen as a key driver of DCI, some patients still develop DCI even after CVS is treated. This gap led researchers to ask: What other common clinical factors predict DCI—and how can combining these with CVS improve care?
In a 2022 study published in the Chinese Medical Journal, a team from the First Affiliated Hospital of Fujian Medical University (China) analyzed data from 711 aSAH patients to answer this question. Their goal: create a DCI prediction model using non-CVS factors (everyday clinical metrics) and test whether combining it with CVS improves accuracy.
What is aSAH, DCI, and CVS?
First, let’s break down the terms:
- aSAH: Bleeding from a ruptured brain aneurysm into the subarachnoid space (the area between the brain and its protective membranes).
- DCI: A decline in brain function (like confusion or weakness) or new brain infarcts (dead tissue) on CT scans—usually 3–14 days after the bleed. It’s a top cause of long-term disability or death in aSAH patients.
- CVS: Narrowing of brain arteries that reduces blood flow. It’s often linked to DCI, but not all DCI cases involve CVS.
The Study: Who, What, and How
The researchers looked at retrospective data from 711 aSAH patients (ages 18–60) who had surgery (clipping or coiling) at their hospital between 2013 and 2018. They excluded patients with pre-existing DCI, prior neurological conditions (like stroke or tumors), or serious systemic diseases (e.g., kidney failure, cancer).
To find DCI predictors, they analyzed:
- Admission factors: Loss of consciousness (LOC), high blood pressure (≥130/80 mmHg), fever.
- Clinical scores: Glasgow Coma Scale (GCS, measures consciousness), Hunt and Hess (severity of aSAH), VASOGRADE (a DCI risk scale combining bleeding and clinical status).
- Surgery details: Type (clipping vs. coiling), number of aneurysms.
- Post-op care: Mechanical ventilation (MV) duration, post-op GCS (consciousness on day 1 after surgery).
DCI was defined as new symptoms or brain infarcts not present on admission or immediate post-op scans. CVS was diagnosed via scans (DSA or CTA) showing persistent artery narrowing.
Key Findings: Non-CVS Factors Matter
Of the 711 patients, 57 (8%) developed DCI. The researchers found 7 non-CVS factors that predicted DCI:
- Loss of consciousness (LOC) at admission: 3.5x higher risk of DCI.
- Hypertension: 2.2x higher risk.
- Higher VASOGRADE score: Each point increase raised risk by 58%.
- More than 3 aneurysms: 11.5x higher risk (rare, but impactful).
- Surgical procedure: Coiling (endovascular) had 64% lower risk than clipping (open surgery).
- Post-op mechanical ventilation ≥3 days: 3x higher risk (a sign of severe brain dysfunction).
- Lower post-op GCS: Each 1-point drop raised risk by 17%.
Using these factors, the team created a non-CVS prediction model for DCI. They then compared its accuracy to a CVS-only model—and a combined model (non-CVS + CVS).
The Combined Model Wins: Better DCI Prediction
The results were clear:
- Non-CVS model: C-statistic = 0.805 (good accuracy; 85.5% sensitive, 63.2% specific).
- CVS-only model: C-statistic = 0.851 (better specificity, 91.3%, but lower sensitivity, 78.9%).
- Combined model: C-statistic = 0.933 (excellent accuracy; 94.5% sensitive, 82.9% specific).
In plain terms: The combined model was far better at both spotting DCI (sensitivity) and avoiding false positives (specificity). It also had the highest “Youden index”—a measure of overall diagnostic value (3.086 vs. 2.659 for non-CVS and 2.072 for CVS).
Why This Matters for Patients and Doctors
CVS is not a perfect DCI predictor. Some patients get DCI without CVS, and some with CVS never get DCI. This study shows non-CVS factors are just as important—and combining them with CVS drastically improves risk assessment.
For example:
- A patient with LOC, hypertension, and post-op MV for 3+ days has a high DCI risk even if CVS isn’t present.
- The VASOGRADE score—already used to predict DCI—proved critical here, reinforcing its value in clinical practice.
- Post-op MV duration is a simple metric that signals early brain dysfunction—an early warning sign for DCI.
Limitations to Consider
Like all studies, this one has gaps:
- It’s retrospective: Data was from past records, which can introduce bias.
- Small sample: Only 57 DCI cases—larger studies are needed to confirm findings.
- No external validation: The model hasn’t been tested in other hospitals, so its generalizability is unknown.
The researchers plan to address these with a prospective (forward-looking) study next.
Conclusion: A Better Tool for DCI Prevention
This study adds critical insights to aSAH care: DCI isn’t just about CVS. Common clinical factors—like LOC, hypertension, and post-op ventilation—help doctors spot risk earlier. The combined model (non-CVS + CVS) is easy to use (it relies on standard clinical data) and far more accurate than either model alone.
For patients, this means fewer missed DCI cases and less overtreatment. For doctors, it’s a practical tool to tailor care—like closer monitoring or early interventions—to high-risk patients.
The original study was led by Yue Chen, Guanmin Li, Xiaoyong Chen, Dengliang Wang, Wenhua Fang, Dezhi Kang, and Chenyu Ding from the First Affiliated Hospital of Fujian Medical University. It was published in the Chinese Medical Journal in 2022.
doi: doi.org/10.1097/CM9.0000000000001844
Was this helpful?
0 / 0