Aspirin Plus Clopidogrel After Minor Stroke: Reading POINT and CHANCE

The 21-day rule for aspirin plus clopidogrel after a minor stroke or high-risk transient ischemic attack (TIA) is not the result of a trial that tested 21 days against every other number. It is what you get when you take two large trials, CHANCE and POINT, pool their patients, and ask a more specific question than either trial asked alone: not "does the combination work?" but "when does it work, and when does it only bleed?" The prevention benefit turned out to be crowded into the first weeks, while the bleeding cost kept mounting for as long as both drugs continued. Twenty-one days is roughly the point where continuing stops paying for itself.

Key points#

• CHANCE (2013) and POINT (2018) both showed that adding clopidogrel to aspirin early after a minor stroke or high-risk TIA reduces recurrent ischemic events.
• The two trials differed on purpose in how long they ran the combination: 21 days of overlap in CHANCE, a full 90 days in POINT.
• A pooled patient-level analysis split the follow-up into two windows and found nearly all the benefit fell in days 0 to 21, with none detectable afterward.
• Bleeding risk, unlike stroke risk, does not fade with time, so extra weeks on two drugs add harm without adding protection.
• Short-course dual therapy applies to genuinely minor strokes and high-risk TIAs, started within about 24 hours, in people without a competing reason for anticoagulation.

Two clocks running at different speeds#

Secondary stroke prevention after a minor event is a timing problem before it is a drug problem. The danger of a second stroke is highest in the hours and days right after the first one and then falls steeply. That is why every trial in this space insisted on starting treatment within roughly 24 hours: the window you are trying to protect is closing on its own.

Bleeding runs on a different clock. The risk of a serious hemorrhage on two antiplatelet drugs does not spike and fade; it accrues steadily, a little more with each day of treatment. So the two harms move in opposite directions over time. Recurrent-stroke risk is a fire that mostly burns out; bleeding risk is a meter that keeps ticking. The clinical question is when to stop feeding the meter without letting the fire flare back up.

CHANCE and POINT: one idea, one deliberate difference#

CHANCE, reported in 2013, enrolled 5,170 patients in China who had either a minor ischemic stroke (a National Institutes of Health Stroke Scale, or NIHSS, score of 3 or lower) or a high-risk TIA (an ABCD2 score of 4 or higher), all within 24 hours of onset. Patients got aspirin alone or aspirin plus clopidogrel. The two drugs overlapped for only the first 21 days, after which clopidogrel continued alone to day 90. Recurrent stroke at 90 days fell from 11.7 percent to 8.2 percent, a hazard ratio near 0.68, and serious bleeding stayed rare and roughly equal between the groups.

POINT, reported in 2018, asked almost the identical question in a largely North American and European population of 4,881 patients, using the same NIHSS and ABCD2 thresholds. Its one deliberate design change: both drugs ran together for the full 90 days. The composite of ischemic stroke, heart attack, or vascular death dropped from 6.5 percent to 5.0 percent, with recurrent ischemic stroke falling in step. But major hemorrhage roughly doubled, from 0.4 percent on aspirin alone to 0.9 percent on the combination, and the trial was halted early once that safety signal was clear.

Read side by side, the trials tell one story with a twist. Both confirmed that early clopidogrel prevents strokes. They diverged, by design, on duration, and the longer-running POINT paid the bleeding bill that CHANCE's shorter overlap had mostly dodged. Neither trial, on its own, could say where the tipping point sat.

Splitting the timeline to find the answer#

The resolution came from a pooled, patient-level analysis of both trials published in JAMA Neurology in 2019, covering just over 10,000 people. Rather than compare 90-day totals, the authors did something simpler and sharper: they cut the follow-up into an early window and a late window and counted events in each.

In days 0 through 21, dual therapy cut major ischemic events from 7.8 percent to 5.2 percent, an adjusted hazard ratio of 0.66 (95 percent confidence interval 0.56 to 0.77). That is a large and statistically firm reduction. In days 22 through 90, the hazard ratio was 0.94 (confidence interval 0.67 to 1.32), meaning no detectable difference at all. Essentially the entire ischemic benefit was earned in the first three weeks.

Bleeding told the mirror-image story. Major hemorrhage showed a nonsignificant upward drift even in the first 21 days and kept climbing after that, so every extra week of the combination past three weeks bought hemorrhage risk and nothing in return. The authors put it plainly: the benefit of the combination was confined to roughly the first 21 days.

Why the number lands near three weeks#

Line the two curves up and the logic is almost arithmetic. Early on, the recurrence risk is high, so the stroke-prevention benefit is large and dwarfs the small, slowly building bleeding cost. As recurrence risk decays, the benefit shrinks toward zero, but the bleeding cost keeps its steady pace. Somewhere the falling benefit and the rising cost cross. In the pooled data, that crossing sits near three weeks.

It helps to be clear about what 21 days is and is not. It is not a biological deadline, some moment when platelets or arteries suddenly change behavior. It is a population-level compromise: the duration that captures nearly all of the average benefit while limiting the accumulated bleeding risk that comes with longer treatment. The same reasoning also explains why CHANCE looked safer than POINT. CHANCE stopped the overlap at 21 days and saw little excess bleeding; POINT carried it to 90 days and saw exactly the harm that a longer course predicts. The pooled timing analysis did not overturn either trial; it reconciled them.

At the bedside#

Guideline bodies landed on short-course dual antiplatelet therapy: start within about 24 hours and continue for roughly 21 days, for people with a genuinely minor stroke or a high-risk TIA who lack a competing indication such as atrial fibrillation that calls for anticoagulation instead. The eligibility limits are not fine print. These trials studied low NIHSS scores and higher-risk TIAs, so the evidence does not stretch to larger strokes, where converting an ischemic area into a bleeding one is a bigger worry.

A few threads are still being pulled. Clopidogrel has to be activated by the liver enzyme CYP2C19, and people who carry loss-of-function versions of the gene respond poorly; the CHANCE-2 trial tested ticagrelor as an alternative in those carriers. Other work has probed whether treatment can start in a slightly wider time window. These refinements build on the CHANCE and POINT lesson rather than replacing it: fit the strength and length of treatment to when the risk actually lives, and stop before the cost overtakes the gain.