Mechanical thrombectomy, in which physicians physically pull a blood clot out of a blocked brain artery using a stent retriever or aspiration catheter, is the most effective treatment available for acute ischemic stroke caused by large artery occlusion. Yet the procedure frequently leaves a trace on follow-up computed tomography: hyperdense, or unusually bright, areas of brain parenchyma appear on immediate post-operative scans in as many as 86% of acute ischemic stroke patients. These bright areas may represent two very different things — contrast dye that has leaked through a stressed blood-brain barrier, or hemorrhagic transformation, in which blood seeps into damaged tissue. Because severe bleeding is a feared complication, many clinicians delay or withhold antithrombotic drugs when they see hyperdensity, fearing that blood thinners could turn a small leak into a catastrophic hemorrhage.

The concept of PCHD was proposed to carve out a safer middle ground. It refers to non-solid, visually distinct hyperdense areas on the first post-thrombectomy CT that measure at least 0.1 cubic centimeters, are at least 5 Hounsfield units denser than the corresponding unaffected side, follow the gray-white matter boundaries of normal anatomy, and — crucially — exert no space-occupying mass effect. Lesions with mass effect, bleeding outside the brain parenchyma, or hemorrhage already visible before the procedure are excluded. In other words, PCHD captures a diffuse, anatomy-conforming brightness that pathologically resembles blood-brain barrier disruption rather than the rupture of small aneurysms that underlies classical intracerebral hemorrhage. That distinction matters, because blood leaking through a leaky barrier behaves very differently from blood under arterial pressure tearing into brain tissue.

The new study, conducted at Minhang Hospital, a tertiary teaching hospital affiliated with Fudan University in Shanghai, followed 141 consecutive patients with anterior circulation acute ischemic stroke who underwent mechanical thrombectomy between January 2019 and April 2023 and showed PCHD on their immediate post-procedure CT. Early antithrombotic therapy was defined as any antiplatelet or anticoagulant drug administered within 24 hours of the procedure. The treatment protocols were individualized by stroke mechanism: patients with stent implantation and severe residual narrowing received the intravenous platelet blocker tirofiban for roughly 24 hours without a loading dose, overlapping with oral dual antiplatelet therapy; other atherosclerotic patients received aspirin plus clopidogrel; and patients with cardioembolic strokes, often tied to atrial fibrillation, received antiplatelet or oral anticoagulant drugs such as warfarin, dabigatran, or rivaroxaban, timed according to infarct size and stroke severity.

Because different attending physicians at the center held divergent views on the safety of treating through hyperdensity, some patients received drugs within 24 hours while others waited until a 24-hour follow-up CT or later re-imaging cleared the way — a natural split that the researchers argue approximated randomization and reduced selection bias. Of the 129 patients with complete 90-day functional data, 59, or 45.7%, received early therapy. The primary endpoint was a favorable functional outcome at 90 days, defined as a modified Rankin Scale score of 0 to 2, a standard measure spanning no symptoms to moderate disability that still allows independent living.

The headline result was striking: 59.3% of patients in the early-treatment group achieved a favorable 90-day outcome compared with 37.1% of those who did not, an unadjusted difference of 22.2 percentage points that corresponds to a descriptive number needed to treat of roughly 4.5. Because the two groups were not perfectly balanced at baseline — patients who received early therapy had somewhat lower stroke severity scores before thrombectomy, and milder hyperdensity grades — the team built three nested logistic regression models. After adjusting for age and sex, early therapy carried roughly two and a half times the odds of a good outcome (odds ratio 2.47). In the fully adjusted model, which additionally accounted for low-density injury on post-procedure CT, early brain edema, prior intravenous thrombolysis, dyslipidemia, baseline stroke severity, number of retrieval attempts, and hyperdensity grade, the association held (odds ratio 2.72), and a third model using post-procedure stroke severity scores instead of baseline scores produced an even larger estimate (odds ratio 3.68). Sensitivity analyses using imputed data for the 12 patients missing 90-day scores, and subgroup analyses stratified by age, sex, atrial fibrillation, prior thrombolysis, residual stenosis, and stroke mechanism, all pointed the same direction, with no significant interactions.

Secondary endpoints added mechanistic texture. Early therapy was associated with a greater shift toward improvement on the ordinal modified Rankin Scale after full adjustment, and 90-day mortality was numerically lower in the treated group (15.3% versus 27.1%), though that difference did not reach statistical significance. Importantly, there was no signal of harm: enlargement of the cerebral hyperdensity occurred at nearly identical rates in both groups (8.5% versus 11.4%). Persistent hyperdensity on the 24-hour follow-up CT, graded by the European Cooperative Acute Stroke Study III morphology criteria, was present in 79 of 135 evaluable patients and appeared less often in the early-treatment group on minimally adjusted analysis, though that difference dissolved after full adjustment. The researchers note that because their immediate scans were single-energy, they could not reliably distinguish contrast extravasation from hemorrhagic transformation, so the 24-hour findings are reported as hyperdensity by morphology rather than as presumed contrast staining.

Why might blood thinners help these patients rather than hurt them? The authors offer two complementary explanations grounded in the broader stroke literature. First, a recent meta-analysis indicates that 3 to 9% of acute ischemic stroke patients experience re-occlusion of the treated artery within 24 hours of thrombectomy, and early antithrombotic drugs can reduce that risk. Second, between 3 and 42.5% of patients suffer the so-called no-reflow phenomenon after a technically successful clot removal, in which stagnant blood components raise blood viscosity and slow flow, triggering microcirculatory problems such as capillary thrombosis even though the main artery is open. Antithrombotic therapy plausibly counters both threats. The clinical analogy that inspired the approach is cerebral venous sinus thrombosis, where significant hemorrhagic transformation often gradually resolves while patients remain on anticoagulation — supporting the idea that barrier-disruption bleeding and rupture-induced bleeding demand different management.

The study’s strengths include its focus on a precisely defined imaging phenotype, its use of covariates measured before treatment initiation, and the naturally split treatment practices that mimicked random assignment. But the authors are careful about its limits. This was a retrospective, single-center study with a modest sample size; residual confounding cannot be excluded despite multivariable adjustment; the number needed to treat is descriptive rather than a causal estimate; PCHD volume and location were not systematically recorded; and missing imaging and follow-up data introduce residual uncertainty. Subgroup findings suggesting larger benefits in atherosclerotic strokes and in older patients are explicitly hypothesis-generating, with imprecise estimates in the smallest strata.

Those caveats frame the takeaway. For roughly one in five acute ischemic stroke patients, neurological deterioration strikes within 72 hours of thrombectomy, and deterioration independently predicts poor three-month outcomes — making the post-procedure window a critical therapeutic frontier. By showing that PCHD patients treated early fared better without measurable increases in bleeding, the Shanghai team offers the first specific evidence that this underappreciated imaging finding could serve as a criterion for identifying patients in whom the benefits of early antithrombotic therapy outweigh the feared risks. The association, however compelling, remains observational; confirming it will require the larger, prospective, ideally randomized trials the authors themselves call for before practice guidelines change.

Subject of Research: Early antithrombotic therapy and 90-day outcomes after post-thrombectomy cerebral hyperdensity in acute ischemic stroke

Article Title: Early antithrombotic therapy improves 90-day functional outcomes in patients with postinterventional cerebral hyperdensity after mechanical thrombectomy: a retrospective cohort study

Article References: Luo, Y., Zhao, J., Huang, L., Xia, S., Wei, S., Ni, T., Cao, W., Shen, C., & Zheng, J. (2026). Early antithrombotic therapy improves 90-day functional outcomes in patients with postinterventional cerebral hyperdensity after mechanical thrombectomy: a retrospective cohort study. Frontiers in Neurology, 17, Article 1891162. https://doi.org/10.3389/fneur.2026.1891162

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1891162

Keywords: stroke, mechanical thrombectomy, antithrombotic therapy, postinterventional cerebral hyperdensity, hemorrhagic transformation, retrospective cohort study, modified Rankin Scale, blood-brain barrier, no-reflow phenomenon, endovascular neurology, cerebral hyperdensity, functional outcome