Early single-center trials suggested microvascular and possible clinical benefits. In TAPAS (n = 1071), routine aspiration improved Myocardial Blush Grade 3 and ST-segment resolution, with fewer 1-year cardiac deaths and cardiac death/reinfarction events in prespecified analysis [1, 2]. In EXPIRA (n = 175, anterior STEMI), aspiration increased STR $\ge$70% and MBG $\ge$2, reduced CMR infarct size, and showed numerically lower 9-month cardiac mortality [3]. An earlier randomized PCI era study yielded mixed signals and emphasized the need for larger trials [16]. In contrast, pragmatic multicenter trials found no clinical advantage for a routine strategy: the registry-randomized TASTE trial (n = 7244) showed no reduction in 30-day or 1-year mortality versus PCI alone [4, 17] and the larger TOTAL trial (n = 10,732) showed no benefit for cardiovascular death, myocardial infarction, cardiogenic shock, or NYHA IV heart failure [6]. The TOTAL-OCT substudy failed to identify a mechanistic effect that could offset these neutral clinical results. Consistent with the neutral clinical findings, OCT revealed similar residual thrombus volumes with routine aspiration and with PCI alone, suggesting limited efficacy of first-generation aspiration catheters in reducing culprit lesion thrombus burden [8].

In TOTAL, 30-day stroke was higher with routine thrombectomy vs PCI alone (33/5033 [0.7%] vs 16/5030 [0.3%]; HR 2.06; 95% CI 1.13–3.75), with an early excess within 48 hours (0.30% vs 0.10%; HR 3.00; 95% CI 1.09–8.25). The absolute risk increase at 30 days was 0.34% (number needed to harm (NNH) $\approx$ 296), and a safety signal persisted to 180 days (52/5033 [1.0%] vs 26/5030 [0.5%]; HR 2.00; 95% CI 1.25–3.20; ARI $\approx$ 0.52%; NNH $\approx$ 194). At 180 days there were more ischemic strokes (37 vs 21; HR $\approx$ 1.7; 95% CI 1.03–3.00) and more primary hemorrhagic strokes (10 vs 2; HR 4.98; 95% CI 1.09–22.7) in the thrombectomy arm; strokes with significant disability or fatal outcome were also more common (0.7% vs 0.3%; HR 2.69; 95% CI 1.42–5.08) [7, 9]. Mechanisms may include aortic/ostial atheroma disturbance and air embolization during catheter exchanges. Major bleeding and access-site events were similar between groups [6, 8]. Although the absolute excess (~0.4%) was slight, it was statistically significant and clinically relevant. The NNH for stroke with routine thrombectomy can be calculated from the TOTAL trial data. With an absolute risk increase of 0.34% (0.66%–0.32%), the NNH is 296 patients (1/0.0034). This means that for every ~296 patients treated with routine thrombectomy rather than selective use, one additional stroke will occur. Given the 30.8% mortality rate associated with stroke, this translates to one extra death for ~961 patients treated with routine thrombectomy (296/0.308). Notably, in the PCI-alone group, no strokes occurred among the 354 patients (7%) who underwent protocol-allowed bailout thrombectomy; this observation is subject to selection bias and does not prove bailout safety [9]. Hypothesized mechanisms include athero- or air embolization related to catheter exchanges and aspiration maneuvers. Major bleeding and access-site complications did not differ significantly between groups [6]. Meta-analytic evidence reinforces this concern. The individual patient level metaanalysis of randomized trials (n = 18,306 primary PCI patients) showed a borderline increase in 30 day stroke/transient ischemic attack (TIA) with thrombectomy overall (0.8% vs 0.5%; odds ratio (OR) 1.43; 95% CI 0.98–2.10; p = 0.06), but a significant increase among patients with high thrombus burden (TIMI thrombus grade $\ge$3: OR 1.56; 95% CI 1.02–2.42) [18]. Complementing this, an extensive aggregated meta-analysis (14 RCTs; n  $\approx$  20,285) reported improved STR and MBG yet no reduction in all-cause mortality at $\ge$6 months (relative risk (RR) 0.92; 95% CI 0.81–1.04) and a significant increase in stroke (RR 1.59; 95% CI 1.07–2.35) [19].

Data imply that newer thrombectomy technologies could provide improved safety. In the multicenter, single-arm CHEETAH study of continuous aspiration (n = 400; 389 with 30-day follow-up), the rate of stroke within 30 days was 0.77% (3/389), and no device-related serious adverse events were reported. As CHEETAH was non-randomized, any cross-trial comparisons with TOTAL should be interpreted cautiously [20]. In the RETRIEVE AMI trial, there were zero cerebrovascular events with the stent-retriever arm [21, 22].

Mechanistically, improved safety could be explained: (1) in aspiration with continuous negative pressure, less air entrainment occurs compared to intermittent, syringe systems, (2) en-bloc thrombus removal prevents fragmentation/thrombus embolism, and (3) fewer manipulations with the catheter minimize disturbance of aortic plaque. Still, these safety signals await verification in large randomized trials before the validity of newer aspiration techniques is determined to be better than manual aspiration.

Factors that are center-defined appear to be important in stroke prevention. Centers that use defined protocols in thrombectomy, areas solely for catheter preparation, and deliberate monitoring for ACT (anticoagulation time) will have lower complication rates regardless of the device used. Complete neurologic evaluation before and after any mechanical thrombectomy should be mandatory.

Both ESC 2023 and ACC/AHA 2025 discourage routine aspiration (Class III, LOEA). Yet, they permit case-by-case aspiration when large residual thrombus remains after initial flow restoration or when refractory no-reflow hampers reperfusion—generally a Class IIb scenario [13, 14]. Post-hoc analysis of high-thrombus subgroups within large RCTs did not show a clear outcome benefit for planned upfront aspiration [6, 7, 8].

AIMI (AngioJet rheolytic thrombectomy) found larger 30-day infarcts and more early adverse events with AngioJet than with PCI alone, raising concerns about the indiscriminate use of rheolytic energy and potential harm from worse distal embolization [10]. JETSTENT demonstrated better STR and lower 6-month major adverse cardiac events (MACE) rates with AngioJet before direct stenting, but no consistent infarct-size benefit, and was in tension with AIMI; overall, the evidence base is not consistent enough to support a case for routine use. Given inconsistent evidence and potential harm signals, rheolytic thrombectomy should not be used routinely in primary PCI [12].

In Enhanced Myocardial Efficacy and Recovery by Aspiration of Liberated Debris (EMERALD) (n = 501), distal protection captured debris but did not improve STR, infarct size, or 6-month outcomes; distal protection is not recommended in routine native-vessel primary PCI. Protection devices remain indicated in selected saphenous vein graft interventions but are not recommended for native-vessel primary PCI in STEMI [11].

Trials comparing intracoronary (IC) versus intravenous (IV) glycoprotein IIb/IIIa inhibition and testing low-dose IC fibrinolysis aimed to mitigate MVO. AIDA-STEMI showed no reduction in the composite of death, reinfarction, or new heart failure with IC abciximab versus IV [23]. In INFUSE-AMI (large anterior STEMI), lesion-directed IC abciximab reduced infarct size by MRI, whereas manual aspiration did not confer infarct-size benefit [24]. T-TIME found that low-dose IC alteplase did not reduce MVO overall and raised safety concerns among late presenters [25].

Deferring stent implantation to allow thrombus dissolution has intuitive appeal but has not been shown to translate into clinical benefit. In DANAMI-3–DEFER, deferred stenting did not reduce the composite of all-cause mortality, heart failure, or reinfarction compared with immediate stenting [26]. The strategy has therefore not been widely adopted. Accordingly, deferred stenting has not translated into reductions in mortality, recurrent MI, or HF, and should not be used routinely.

Recent data have raised doubts over the reliability of angiographic thrombus assessment, showing that there were considerable differences between TIMI thrombus grading and actual thrombus burden when assessed by OCT. The RETRIEVE AMI trial (2025) showed that in patients with a TIMI thrombus grade $\ge$4 (41% of the cohort), approximately 40% actually had low thrombus burden when quantified using an imaging modality, with median baseline thrombus volume differing vastly between groups (18.3 mm³ in the stent-retriever group compared to 7.7 mm³ in the manual aspiration group) [21, 22].

This observation brings into question the patient selection or exclusion processes used by larger randomized trials. A systematic approach using baseline OCT may be beneficial in determining which STEMI cases would genuinely benefit from thrombus modification approaches.

Corrected sentence: “In the TOTAL-OCT sub-study, the median pre-stent thrombus burden was just 2.36%, significantly less than the 12.96% in the RETRIEVE AMI cohort, which might explain the neutral result of routine manual aspiration [8].”

OCT-guided techniques have shown encouraging clinical outcomes. An extensive prospective registry of 3897 STEMI patients demonstrated that OCT-guided primary PCI (69.2% of cases) was associated with significantly lower 5-year all-cause and cardiovascular mortality rates compared to angiography-guided interventions [28]; the benefits were confirmed after propensity scoring. The oct-guided cohort of patients had a greater proportion of patients who underwent thrombus aspiration, as the strategy was tailored using imaging.

Clinical Implementation: OCT performance should be contemplated in STEMI cases where there is a suspicion of high thrombus burden for multiple reasons: (1) accurately assessing and quantifying accurate thrombosis volume, (2) to help select devices used, i.e., between manual aspiration, continuous aspiration systems, or stent-retrievers, (3) to assess standardized procedural success after thrombus modification. A personalized approach to patient selection may optimize outcomes and improve the risk-benefit of mechanical thrombectomy.

Fig. 2 illustrates how imaging-guided decision-making translates into procedural strategy during primary PCI for STEMI. The schematic highlights the transition from an occluded, thrombus-filled artery to restored flow through three main techniques—standard balloon–stent implantation, aspiration-based thrombectomy, and stent-retriever extraction. Integrating OCT-derived thrombus characteristics enables operators to choose devices purposefully, adjust procedural sequencing, and implement deferred stenting when distal embolization risk is high. This visualization underscores the shift toward precision-guided reperfusion rather than uniform device application.

Current European and American guidelines converge that routine manual aspiration during primary PCI in STEMI is not recommended, whereas selective bailout use may be considered in cases of large residual thrombus or refractory no-reflow despite standard measures [13, 14]. These recommendations reflect consistent neutral results across large, randomized trials and the observed stroke signal with aspiration [6]. Accordingly, both ESC 2023 and ACC/AHA 2025 discourage routine manual aspiration (Class III, Level A) while allowing strictly selective bailout use in the presence of large residual thrombus and refractory no-reflow, aligning the guidance with the neutral large-scale evidence and the stroke signal.

“Default pathway: Immediate reperfusion with primary PCI under guideline-directed anticoagulation/antiplatelets; do not use routine aspiration (Class III, LOE A; ESC 2023; ACC/AHA 2025) [13, 14]”.

After wiring and gentle lesion preparation:

If TIMI-3 flow and low thrombus burden $\to$ proceed to stent implantation without aspiration.

If large thrombus (e.g., TIMI 4–5) or refractory no-reflow/slow-flow despite vasodilators $\to$ consider single-pass bailout aspiration. Investigational options (use only in research/exceptional bailout): continuous aspiration systems (e.g., Penumbra CAT RX) or coronary stent retriever thrombectomy.

Perform aspiration (if chosen): one slow pass; maintain suction during withdrawal; reassess TIMI/MBG/STR.

Evaluate response:

If flow improves to TIMI-3 and the angiographic result is acceptable $\to$ complete PCI with stenting.

If no-reflow persists $\to$ treat pharmacologically; avoid repeated fragmentation; refrain from multiple passes unless clear incremental benefit is demonstrated.

Avoid routine rheolytic thrombectomy or distal protection (no proven outcome benefit bailout criteria [10, 11, 12].

Consider single-pass bailout aspiration when:

TIMI thrombus grade 4–5 persists after wiring/gentle lesion preparation;

Refractory noreflow/slowflow despite vasodilators and careful ballooning; or

Recurrent distal embolization despite minimal manipulation. Favor large-caliber, proximal culprit anatomy; avoid highly tortuous/ostial disease with aortic plaque”.

Adjunct imaging (optional): Post-flow OCT/IVUS to assess residual thrombus/plaque and optimize stent sizing/expansion in selected cases.

Expanded Patient Selection Criteria for Bailout Thrombectomy

Anatomical Considerations:

Vessel diameter $\ge$2.5 mm (to accommodate aspiration catheter).

Proximal to mid-segment location (avoiding distal vessels).

Absence of severe tortuosity ($>$90° angulation within 2 cm of lesion).

Non-ostial location or ostial with minimal aortic atheroma on angiography.

Thrombus Characteristics:

TIMI thrombus grade $\ge$4 with visible linear dimension $>$2 vessel diameters.

Fresh thrombus appearance (hazy, mobile, with contrast penetration).

Persistent after gentle wire manipulation and/or small balloon inflation.

Evidence of distal embolization during initial wire crossing.

Clinical Scenarios:

Acute presentation (6 hours) with large thrombus burden.

Hemodynamic instability with a large thrombus potentially contributing.

High-risk anatomy (left main, proximal left anterior descending artery (LAD) with large myocardial territory).

Failed pharmacological intervention (persistent TIMI 3 flow despite IC vasodilators).

Contraindications to Consider:

Presentation $>$12 hours (organized thrombus, higher fragmentation risk).

Severe calcification is visible on angiography.

Vessel diameter 2.5 mm.

Extreme tortuosity requiring excessive catheter manipulation.

Heavy aortic arch atheroma (if transfemoral approach).

Technical Requirements:

Operator experience with $>$20 aspiration procedures.

Availability of multiple aspiration catheter sizes.

ACT confirmed ACT $\ge$250 seconds before attempt.

Cardiac surgical backup available (though rarely needed).

Device selection for modern bailout techniques.

OCT-guided evaluation: When available, always perform OCT in the initial wire crossing to help quantify the actual thrombus burden. Thrombus volume $>$15 mm³ by OCT predicts a high likelihood of enjoying the benefits of mechanical extraction [22].

Continuously aspirating systems (CAT RX): There are many considerations for use:

Large proximal vessel thrombus ($\ge$3.0 mm vessel diameter).

Fresh thrombus characterized by high mobility on angiography.

Cases requiring a more extended aspiration time.

Operators with limited manual aspiration techniques.

The technique involves a single pass with continuous –29 inHg suction, advancing and crossing the lesion slowly, maintaining negative suction during the withdrawal phase, and allowing for controlled back bleeding.

Stent retriever systems (Solitaire X): There are many considerations for use:

Very large, organized thrombus ($>$20 mm³ by OCT if available).

Has previously failed manual or continuous aspiration.

A thrombus that appears laminated or resilient on imaging.

The technique: nets should be deployed distal to the thrombus; allow 2–3 minutes of integration; retrieve under continuous aspiration vacuum.

Maintain anticoagulation, ACT $\ge$250 s.

Device proficiency and availability: We suggest that institutions optimize health care by standardizing on a single advanced system (either continuous aspiration OR stent-retriever), rather than attempting to remain proficient in multiple devices. Operators with $>$10 cases of advanced thrombectomy have better outcomes and fewer adverse events

Several non-mutually exclusive mechanisms could explain why aspiration may be harmful in an all-comer STEMI strategy:

Aortic/ostial plaque disturbance and air embolization: Catheter exchanges at the coronary ostia and in the aortic root may dislodge atheroma or introduce microbubbles if the system is not perfectly purged, providing a plausible pathway for cerebral emboli (consistent with the stroke signal in TOTAL) [6].

Thrombus fragmentation with distal embolization: Negative pressure applied to laminated thrombus can shear and disperse clot fragments into the microcirculation when engagement is suboptimal, paradoxically worsening MVO despite epicardial patency.

Pressure-gradient and flow-field effects: Rapid aspiration can transiently alter local hemodynamics, potentially collapsing fragile channels or dislodging debris downstream.

Device-specific concerns (rheolytic): High-velocity jets can cause hemolysis and endothelial trauma. In AIMI, a larger infarct size raised the possibility that indiscriminate rheolysis exacerbates distal embolization [10].

PCI procedure time was modestly longer with routine thrombectomy in TOTAL (median 39 vs 35 min), and more aggressive guide manipulation or catheter exchanges may predispose to aortic plaque disturbance and paradoxical systemic embolization [9].

These mechanisms underscore why a selective, carefully executed bailout approach—rather than routine use—is more defensible.

Progress likely depends on smarter selection and better tools:

Enriched RCTs in very-high thrombus burden: Randomize after initial wiring/flow restoration, using standardized TIMI thrombus grading and core-lab adjudication. Hard outcomes plus mechanistic endpoints (CMR infarct size, MVO) should be co-primary.

Imaging-guided selection: Evaluate OCT- or IVUS-confirmed large thrombus and plaque morphology to identify phenotypes most likely to benefit from mechanical extraction.

Device innovation and technique: Test single-pass, continuous-vacuum protocols; compare modern large-lumen aspiration vs stent-retriever approaches in prespecified high-risk anatomies; build in systematic stroke surveillance with neurological assessment.

Therapy integration: Optimize combinations with potent antithrombotics and vasodilators; pre-specify rescue pharmacology for no-reflow to minimize repeated device passes.

Without proven clinical benefit in unselected populations, with a small but non-zero stroke risk, routine aspiration is unlikely to be cost-effective. Device costs, additional procedure time, and potential downstream costs of periprocedural stroke may outweigh any theoretical gains in surrogate markers. A selective bailout strategy concentrates device use in patients with the highest pretest probability of net benefit, which is economically and clinically rational pending definitive randomized evidence.