When Considering a Patient for Mechanical Thrombectomy Whose Last Known Normal Is Greater Than 6 Hours, What Condition Would Deem the Patient Ineligible: Five Exclusion Gates That Decide the Case

A large vessel occlusion (LVO) stroke patient whose last known normal (LKN) is more than 6 hours ago becomes ineligible for mechanical thrombectomy (MT) the moment imaging shows a large established infarct core — typically an Alberta Stroke Program Early CT Score (ASPECTS) of 5 or less on non-contrast CT, or a core volume above roughly 50–70 mL on CT perfusion or diffusion-weighted MRI, depending on the trial criteria applied. Large core is the condition most often cited in exam-style questions because it fails the fundamental test: dead brain tissue cannot be saved by reopening the artery. Several other exclusion conditions also apply in the extended window, and they differ meaningfully from the 0–6 hour criteria most clinicians learned first.

This article covers the late-window (6–24 hour) thrombectomy decision as defined by the DAWN and DEFUSE 3 trials, the 2019 AHA/ASA guideline update, and the 2026 AHA/ASA acute ischemic stroke guideline that superseded earlier recommendations. It is general educational information, not a diagnosis or treatment plan.

Why the 6-Hour Boundary Changes the Exclusion List Entirely

The 0–6 hour window is forgiving because the ischemic penumbra is usually still large enough that selection relies mainly on non-contrast CT and clinical severity. Past 6 hours, the salvageable tissue has shrunk or died unpredictably, and the selection criteria shift from "exclude hemorrhage and score the CT" to "prove there is still salvageable tissue."

That shift is why the same patient can qualify at 5 hours 50 minutes and fail at 6 hours 10 minutes with no change in examination. Time is a proxy for tissue fate, not a direct measure of it. The DAWN trial (NEJM, 2018) and DEFUSE 3 (NEJM, 2018) both exploited this by using imaging-based selection rather than a rigid clock in the 6–24 hour window, and both showed substantial benefit in carefully selected patients.

The 2019 AHA/ASA guideline gave a Class I recommendation for thrombectomy in selected patients 6–16 hours from LKN (DAWN criteria) and a Class IIa recommendation for 16–24 hours (DEFUSE 3 criteria). The 2026 AHA/ASA guideline, published in Stroke, extended the upper bound to 24 hours with additional imaging-based selection and addressed posterior circulation separately. Criteria quoted from memory should always be confirmed against the current guideline text.

The Large Core: The Single Condition Most Often Named as the Disqualifier

A large established infarct core is the answer most board-style questions are looking for. The logic is straightforward: thrombectomy restores blood flow to tissue that is ischemic but not yet infarcted. Tissue that is already infarcted on imaging cannot be recovered, and reperfusing it exposes the patient to hemorrhagic transformation without clinical benefit.

DAWN required a clinical-imaging mismatch: patients 80 years or older needed a National Institutes of Health Stroke Scale (NIHSS) score of 10 or higher, an infarct volume under 21 mL, and a mismatch ratio of 1.8 or more. Patients under 80 needed NIHSS of 10 or higher with core under 31 mL, or NIHSS of 20 or higher with core 31–51 mL. These numbers matter because they set the working ceiling for "small enough core" in the extended window.

DEFUSE 3 used a different scheme: an internal carotid artery or M1 occlusion, NIHSS of 6 or higher, core volume under 70 mL, mismatch ratio of 1.8 or more, and a mismatch volume of 15 mL or more. The 70 mL ceiling versus DAWN's 51 mL ceiling is a real discrepancy that reflects different trial populations rather than a settled science.

A small but important failure case: a patient at 8 hours with an NIHSS of 22 but a CT perfusion core of 65 mL and mismatch ratio of 1.4 fails both trials. The core is technically under DEFUSE 3's 70 mL limit, but the mismatch ratio does not clear the 1.8 threshold, so there is too little penumbra to justify the risk.

ASPECTS ≤5 as a Practical Bedside Threshold

The Alberta Stroke Program Early CT Score grades ten regions of the middle cerebral artery territory on non-contrast CT, starting at 10 and subtracting one point per affected region. A score of 6 or higher is generally treated as favorable for reperfusion; 5 or lower signals a large core that most trials excluded.

There is a real-world wrinkle the trials do not fully capture: inter-rater reliability for ASPECTS is moderate at best, and disagreement of one or two points between readers is common, especially with early ischemic change that is subtle. A patient scored 5 by an emergency physician may be scored 6 or 7 by a neuroradiologist. The practical implication is that a borderline ASPECTS should trigger CT perfusion or MRI rather than an automatic refusal.

The 2026 guideline reflects this by leaning on perfusion imaging and MRI-based core measurement rather than ASPECTS alone where those tools are available. Centers without perfusion capability still rely on ASPECTS plus clinical-imaging mismatch, but the confidence interval on that decision is wider.

A second wrinkle: ASPECTS is validated for anterior circulation. It does not apply to basilar artery occlusion, which the 2026 guideline addresses separately and where the posterior circulation ASPECTS (pc-ASPECTS) scale is used instead. Applying a standard ASPECTS to a basilar occlusion is a category error.

Pre-Morbid Function: The mRS 0–1 Requirement and Its Real-World Friction

Extended-window trials enrolled patients with a pre-stroke modified Rankin Scale (mRS) score of 0 or 1 — meaning no symptoms or no significant disability before the stroke. A patient already dependent before the event (mRS 3 or higher) was excluded, and the guideline carries that forward.

The clinical reasoning is that the benefit of thrombectomy is measured against the patient's prior functional baseline. If the baseline is severe disability, a good angiographic result does not translate into a meaningful functional gain, and the procedural risk is not offset.

The friction arises in practice because mRS is a coarse, observer-dependent scale. A patient with mild cognitive impairment who lives independently is mRS 1 by most readings; the same patient described as "needs some help with finances" may be scored 2 by a different clinician. The 2026 guideline language acknowledges this by focusing on "premorbid functional independence" rather than a strict numeric cutoff, but the trials' numeric cutoff remains the practical anchor.

A useful edge case: an 88-year-old nursing home resident with known vascular dementia, mRS 2, and a basilar occlusion at 10 hours. The strict trial criteria would exclude on mRS grounds. The 2026 guideline's posterior circulation recommendations are more permissive on some fronts but still require independent pre-morbid function. This is a genuine gray zone where a family meeting and shared decision-making are the appropriate response, not a protocol answer.

Intracranial Hemorrhage and Coagulopathy: The Absolute Barriers

Active intracranial hemorrhage on non-contrast CT is an absolute contraindication regardless of time window. This is not a subtle exclusion; it is the first thing the CT rules out, and the extended window does not change it.

Coagulopathy is the second absolute barrier, and it bites harder in the late window because the risk of hemorrhagic transformation rises with time and with core size. The standard thresholds used across trials and guidelines are an international normalized ratio (INR) above 3.0, a platelet count below 100,000/μL, and current use of a direct oral anticoagulant within the last 48 hours with an elevated or unmeasurable anticoagulant effect.

The direct oral anticoagulant piece is where real-world practice diverges from trial protocol. DAWN and DEFUSE 3 largely excluded patients on anticoagulation, but modern practice increasingly uses specific reversal agents (idarucizumab for dabigatran, andexanet alfa for factor Xa inhibitors) to make the patient eligible when the potential benefit is high. The 2026 guideline acknowledges this, but reversal is a center-specific decision with its own risk-benefit calculus.

A less obvious coagulopathy exclusion is severe thrombocytopenia from any cause. A platelet count of 80,000/μL in a patient with sepsis and an LVO at 10 hours is not an automatic exclusion by trial criteria in every interpretation, but most interventionalists would pause. The trial protocol said 100,000; the bedside decision may differ.

Large Territorial Infarct and Mass Effect: The Imaging Findings That Override the Clock

Beyond core volume, specific imaging patterns disqualify: a large established infarct involving more than one-third of the middle cerebral artery territory, a hyperdense middle cerebral artery sign with extensive early ischemic change, or any midline shift suggesting mass effect.

The one-third MCA territory threshold is roughly equivalent to an ASPECTS of 5 or lower in many cases, so the two criteria overlap. Where they diverge is in the pace of infarct evolution. A patient with a fast-evolving infarct can hit the one-third threshold in 3 hours, and a patient with slow collateral flow can be at 10 hours with an ASPECTS of 7. The point of the imaging criteria is to measure tissue fate directly rather than infer it from the clock.

Hemorrhagic transformation of an established infarct, even a small one, is a relative to absolute exclusion depending on size. A petechial hemorrhage within an infarct does not automatically stop the procedure; a parenchymal hematoma does. The distinction matters at the bedside and is often missed in quick CT reads.

A failure case worth naming: a patient at 9 hours with a right MCA occlusion, ASPECTS 6, but CT shows early midline shift of 3 mm. The ASPECTS is technically above the threshold, but the shift indicates tissue loss that the score does not capture. Most interventionalists would treat the shift as exclusionary. The lesson: ASPECTS does not replace a reading of the whole scan.

Posterior vs Anterior Circulation: Different Exclusions, Different Rules

Anterior circulation (internal carotid artery, M1, M2) is the population the classic extended-window trials enrolled. Posterior circulation — basilar artery occlusion in particular — has a different evidence base, and the 2026 AHA/ASA guideline addresses it separately with recommendations that are more permissive on time in some scenarios and more restrictive on imaging criteria in others.

The 2022 ATTENTION and BAOCHE trials (NEJM) showed benefit for basilar artery thrombectomy out to 12 hours (ATTENTION) and 24 hours (BAOCHE) in selected patients, which is later than the anterior circulation trials generally enrolled. A patient at 10 hours with a basilar occlusion and a small pc-ASPECTS deficit is a different eligibility question than an MCA occlusion at the same time.

The practical implication is that a blanket statement like "thrombectomy is excluded after 6 hours if ASPECTS ≤5" is an anterior circulation statement. It does not directly apply to basilar occlusion, and applying it to a posterior circulation patient is a common source of inappropriate exclusion.

A second edge case: a patient with a basilar occlusion and an ASPECTS score of 8 but a pc-ASPECTS of 5. The standard ASPECTS is not the right tool. The decision should rest on pc-ASPECTS, brainstem involvement, and clinical severity, and the 2026 guideline directs clinicians to that distinction.

The Time-Cost Tradeoff: Why a 30-Minute Delay Can Flip Eligibility

The clock is a proxy, and in the late window a 30-minute delay can flip a patient from eligible to ineligible if the infarct is evolving fast. This is not a reason to rush — the imaging selection is what makes the procedure safe — but it is a reason to have a clear pathway for CT-to-groin time in extended-window patients.

According to DAWN's published trial data, the median time from imaging to groin puncture was 18 minutes, and the median time from symptom onset to groin puncture was 13 hours — a figure that reflects how late the window truly ran. DEFUSE 3 reported a median of 11 hours from LKN to groin puncture. These are the numbers that set the benchmark for real-world centers.

The practical computation: if a patient is at 8 hours from LKN and arrives at a center without perfusion imaging, transferring to a comprehensive center adds 45–90 minutes. That delays groin puncture to 9–10.5 hours, and if the core is evolving at the rate suggested by the ASPECTS change, the 30–60 minutes can cost a point or more. The decision to transfer versus treat locally is therefore an eligibility decision, not just a logistics decision.

A cost figure worth naming: the 2026 guideline cites door-in-door-out times as a key quality metric, but does not set a specific target for transfer in the extended window. Centers should track their own times and use them to decide whether transfer is feasible for late-window patients. A useful rule of thumb, drawn from the trial medians: a transfer that adds more than 60 minutes likely negates the benefit for a fast-evolving infarct.

Putting the Exclusions in Order: A Practical Checklist

Applying the exclusions in the wrong order wastes time. The practical sequence for an extended-window patient is: confirm LKN and current time, confirm the last known well is beyond 6 hours, verify CT rules out hemorrhage, check pre-morbid mRS, check coagulopathy labs, then move to perfusion imaging for core and mismatch.

The checklist below reflects the sequence most centers use, and it is written to be read at the bedside.

Two failure modes sit outside the table. First, a patient who meets every criterion but arrives at a center without perfusion imaging at 14 hours — the transfer decision itself may be the disqualifier. Second, a patient who meets every criterion but has an aortic dissection flap on the CT angiogram, which is rare but is an absolute contraindication to transfemoral access.

The 2026 guideline also flags kidney function as a consideration for CT perfusion contrast, though it is not a hard exclusion. A patient with an estimated glomerular filtration rate below 30 mL/min/1.73 m² may need MRI-based selection instead, which adds time. That is another practical pathway decision worth pre-planning.

FAQ

What is the single most common reason a late-window patient is excluded?

A large core on perfusion imaging is the most frequent disqualifier, and DAWN's 51 mL ceiling and DEFUSE 3's 70 mL ceiling define the trial-based limits. In practice, a mismatch ratio below 1.8 is often the decisive number, because it means the penumbra is too small to justify risk. ASPECTS ≤5 on non-contrast CT is the equivalent bedside threshold when perfusion is unavailable. A borderline ASPECTS should trigger perfusion imaging rather than a reflex refusal.

Can a patient with a pre-stroke mRS of 2 still get thrombectomy at 10 hours?

The trials excluded mRS 2, and the guideline language focuses on premorbid independence, so an mRS 2 patient is outside the trial evidence base. That does not make the procedure impossible; it makes it a shared decision with the family and the interventionalist. Some centers will treat a high-functioning mRS 2 patient with a small core and large penumbra, especially if the mRS 2 reflects a single domain limitation. The 2026 guideline leaves room for this, but the evidence is weaker than for mRS 0–1.

Does a direct oral anticoagulant automatically exclude a late-window patient?

No, but it raises the bar. DOAC use within 48 hours is a trial exclusion, and the risk of hemorrhagic transformation is real. Reversal agents — idarucizumab for dabigatran, andexanet alfa for factor Xa inhibitors — are used at some centers to make the patient eligible when the potential benefit is high. The 2026 guideline acknowledges reversal but does not mandate it, so the decision is center-specific. A patient on a DOAC with a large core is a clearer exclusion than one with a small core and large penumbra.

How does the rule differ for a basilar artery occlusion?

Basilar occlusion is governed by a separate evidence base. ATTENTION (2022) and BAOCHE (2022) enrolled patients out to 12 and 24 hours respectively, later than the anterior circulation trials. The pc-ASPECTS scale replaces the standard ASPECTS, and brainstem involvement is a key exclusion criterion. Applying an anterior circulation ASPECTS cutoff to a basilar occlusion is a category error that can inappropriately exclude a treatable patient. Always confirm the current 2026 guideline text for the posterior circulation criteria.

Is an ASPECTS of 5 always an absolute exclusion?

No. ASPECTS reads have moderate inter-rater reliability, and a one-point disagreement is common. A score of 5 by one reader may be 6 by another, particularly with subtle early ischemic change. The trials rarely enrolled patients at 5 or below, so the evidence base is thin, but the guideline does not call it absolute. A borderline score should trigger perfusion imaging or MRI, and the final decision should rest on the full imaging picture plus the pre-morbid functional baseline.

Can a patient be too late even for the 24-hour window?

Yes. The 2026 guideline extends the upper bound to 24 hours for selected anterior circulation patients, but a patient at 26 hours from LKN is outside the trial evidence. That does not make thrombectomy impossible, but it moves the decision into individualized care with a much weaker evidence base. The 2026 guideline language should be checked for any update beyond 24 hours; as of that edition, 24 hours is the ceiling for the guideline-supported window.

What if the patient's family cannot confirm the LKN?

An uncertain LKN is treated as the earliest possible time, and that conservative reading often pushes the patient into the extended window or out of it entirely. If the patient woke with symptoms, LKN is the time they went to sleep, which can add several hours. In that situation, the imaging selection becomes even more important, because the clock cannot be trusted and the tissue fate has to be measured directly. A patient with a small core and large penumbra on perfusion can still qualify even with an uncertain LKN.

Does kidney function affect eligibility?

Kidney function is not a hard exclusion in the 2019 or 2026 guidelines, but an estimated glomerular filtration rate below 30 mL/min/1.73 m² complicates CT perfusion because of contrast exposure. MRI-based selection (diffusion-weighted imaging and fluid-attenuated inversion recovery mismatch) is an alternative that avoids iodinated contrast. The practical effect is a time cost, not an absolute bar, and the decision to use MRI should be pre-planned at centers that routinely see renal impairment.

This article is general educational information about late-window stroke selection criteria. It is not a diagnosis, treatment plan, or substitute for clinical judgment, and it makes no promise about outcomes. Eligibility decisions for mechanical thrombectomy must be made by a qualified local stroke team using the current guideline text, the patient's full imaging, and the patient's pre-morbid functional status. If you or someone nearby is showing stroke symptoms — face drooping, arm weakness, speech difficulty — call local emergency services immediately; time-critical stroke care cannot wait for an article.